Cover Page The handle http://hdl.handle.net/1887/74526 holds various files of this Leiden University dissertation. Author: Bogarin Chaves, D.G. Title: Evolutionary diversification and historical biogeography of orchidaceae in Central America with emphasis on Costa Rica and Panama Issue Date: 2019-07-02 https://openaccess.leidenuniv.nl/handle/1887/1 http://hdl.handle.net/1887/74526 https://openaccess.leidenuniv.nl/handle/1887/1� EVOLUTIONARY DIVERSIFICATION AND HISTORICAL BIOGEOGRAPHY OF ORCHIDACEAE IN CENTRAL AMERICA with emphasis on Costa Rica and Panama Diego G. Bogarín Chaves 2019 Diego G. Bogarín Chaves. 2019. Evolutionary diversification and historical biogeography of Orchidaceae in Central America with emphasis on Costa Rica and Panama Ph.D Thesis at University of Leiden, the Netherlands, 2019 Cover design: Fenna Schaap and Diego Bogarín. Cover photograph: Lepanthes genetoapophantica by Franco Pupulin. Layout: Diego Bogarín Printed by: Proefschriftmaken.nl This Ph.D research was made possible with financial support of: The Office of International Affairs and External Cooperation, University of Costa Rica Naturalis Biodiversity Center, Endless Forms group, the Netherlands Lankester Botanical Garden, University of Costa Rica Alberta Mennega Foundation, the Netherlands Evolutionary diversification and historical biogeography of Orchidaceae in Central America with emphasis on Costa Rica and Panama Proefschrift ter verkrijging van de graad van Doctor aan de Universiteit Leiden, op gezag van de Rector Magnificus prof.mr. C.J.J.M. Stolker, volgens besluit van het College voor Promoties te verdedigen op dinsdag 2 juli 2019 klokke 11:15 uur door Diego Gerardo Bogarín Chaves Geboren te San José, Costa Rica in 1982 Promotor: Prof. Dr. Erik F. Smets Naturalis Biodiversity Center, Leiden University & KU Leuven Copromotors: Dr. Barbara Gravendeel Leiden University & Naturalis Biodiversity Center Prof. Franco Pupulin Lankester Botanical Garden, University of Costa Rica Promotiecommissie: Prof. Dr. Gilles P. van Wezel (Chair) Leiden University, The Netherlands Prof. Dr. Menno Schilthuizen (Secretary) Leiden University, The Netherlands Prof. Dr. Paul Kessler Leiden University, The Netherlands Prof. Dr. Hans ter Steege Vrije Universiteit Amsterdam & Naturalis Biodiversity Center Prof. Dr. Renate Wesselingh Université Catholique de Louvain, Belgium Prof. Dr. Alexandre Antonelli Royal Botanic Gardens, Kew, London, UK This thesis is dedicated to my parents Gerardo and Inés; my brother, Sergio; my sister, Marielos; and my wife Maricruz Table of Contents General Introduction Chapter 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Aims of the thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Outline of the thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Taxonomy and systematics of Lepanthes and allies Chapter 2 Phylogenetic comparative methods improve the selection of characters for generic delimitations in a hypediverse Neotropical orchid lineage . . . . . . . . . . . . . . . . . . . . 25 Chapter 3 Genus-level taxonomical changes in the Lepanthes affinity . . . . . . . . . . . . . . . . . . . 49 Chapter 4 Anchored Hybrid Enrichment generated nuclear, plastid and mitochondrial markers resolve the Lepanthes horrida (Orchidaceae: Pleurothallidinae) species complex . 59 Chapter 5 Two new Lepanthes (Orchidaceae: Pleurothallidinae) from Panama . . . . . . . . . . . 91 Pollination biology Chapter 6 Pollination of Trichosalpinx (Orchidaceae: Pleurothallidinae) by biting midges (Diptera: Ceratopogonidae) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Chapter 7 Floral anatomy and evolution of pollination syndromes in Lepanthes and close relatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Evolutionary diversification and biogeography of Lepanthes and allies Chapter 8 Recent origin and rapid speciation of Neotropical orchids in the world’s richest plant biodiversity hotspot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Chapter 9 Speciation and biogeography of the hyperdiverse genus Lepanthes (Orchidaceae: Pleurothallidinae) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 General discussion and conclusions Chapter 10 General discussion and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Summaries Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Samenvatting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 Resumen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 References and acknowledgments References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 Curriculum vitae Curriculum Vitae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 List of publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 Abstracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 General Introduction Chapter 1 Introduction 1.1 Evolutionary diversification and historical biogeography of Orchi- daceae in Costa Rica and Panama The floristic richness of the Neotropics has a complex origin. The most diverse plant family in the American continent is made up of the Orchidaceae with more than 13,000 species (Ulloa et al., 2017) and the most diverse genera of angiosperms are Piper L. (1,804) and Peperomia Ruiz & Pav. (1,133) (Piperaceae), Miconia Ruiz & Pav. (1,110) (Melastomataceae) and the orchid genera Epidendrum L. (1,459 species), Lepanthes Sw. (1,125) and Stelis Sw. (1,128). In contrast, most genera of American angiosperms (5,975) contain less than 100 species (Ulloa et al., 2017). Orchidaceae is also the most diverse plant family in Central America, concentrating 13% of the species and the number of species is triple that of other well-represented angiosperm families. Although we are still far from knowing the exact number of orchid species extant in both coun- tries nowadays, at present Costa Rica (1,620 spp.) and Panama (1,372 spp.) together contain more than 2,000 species of orchids; representing about 8.0% of all orchid species on just about 1% of the Earth’s land surface. In this region, Cymbidiae, Pleurothallidinae and Laeliinae are the most diverse groups and contain the largest genera: Maxillaria Ruiz & Pav. s.l., Lepanthes, Oncidium Sw., Pleurothallis R.Br., Stelis and Epidendrum showing the same global pattern ob- served in the Neotropics. Historically, the isthmus of Costa Rica and Panama has been a source of fascination for its strategic position linking North America to South America. The geological events that led to the closure of the isthmus that started with the formation of a volcanic arc, dating from the Cretaceous to Eocene, 67 to 39 million years ago (Ma) (Montes et al., 2015), have been studied extensively but are still controversial . There is no consensus about when the isthmus closed the Central American Seaway (CAS) separating the Pacific from the Atlantic Ocean and favoring the Great American Biotic Interchange (GABI). Traditionally, it was assumed that this closure was established between 3.5-5.0 Ma, but other studies that include new information suggest a closure much earlier, between 13-15 Ma in the middle Miocene (Bacon et al., 2015; Montes et al., 2015). Despite this controversy, it is clear that with the initial emergence of a volcanic arc in the Creta- ceous, orchids had millions of years to colonize some of these oceanic islands by wind dispersal of seeds and evolve there. According to a phylogenomic analysis and net diversification regimes across lineages using BAMM analysis, Givnish et al., (2015) proposed that Orchidaceae arose around 112 Ma in the Cretaceous, long before the formation of the arc and subsequent closure of the Isthmus of Panama. However, the most diverse Neotropical subtribes Laeliinae, Oncidi- 11 inae, Maxillariinae and Pleurothallidinae probably diversified between 10-25 Ma after the last acceleration of net diversification rate that occurred about 25 Ma, overlapping with the possible closure of the Isthmus proposed recently (Bacon et al., 2015; Givnish et al., 2015; Montes et al., 2015). Indeed the flora of the Isthmus is dominated mainly by species of Cymbidiae, Laeliinae and Pleurothallidinae that diversified in the past 20 Ma. Consequently, we can assess some of the factors that shaped this extraordinary diversity in the isthmus by analyzing the current floristic composition of selected orchid groups with phylogenetics, floral trait evolution, pollination evi- dence and biogeographical analyses (Fig. 1.1). 1.2 Orchid diversity in the hotspot of Costa Rica and Panama Updated floristic inventories of Costa Rica and Panama (both countries treated as a biogeographic unit) resulted in the detection of 2012 species of orchids of which 934 are shared (Bogarín et al., 2014b). From these figures, 784 (39%) species are endemic to the Isthmus (Table 1.1). A strategy to analyze the current evolutionary and floristic relationships of the Orchidaceae of the Isthmus is to study the most diverse groups in the region. An analysis of the various genera in both Costa Rica and Panama shows that Epidendrum L., Lepanthes Sw. and Stelis s.s. Sw. contain the highest number of species and the highest percentages of endemism (Tables 1.2–1.3). These genera are also monophyletic (Pridgeon et al., 2001) and therefore there is no bias due to the use of different nomenclatural circumscriptions that might cause variations in the number of species assigned to a genus. One factor that can affect the interpretation of evolutionary and biogeographic data is the intensity of the alpha-taxonomic work. The most diverse genus in the Isthmus is Epidendrum, the taxonomy of which has been developed in detail and consistently by Hágsater and colleagues. Lepanthes is a diverse genus, and despite extensive work by (Luer, 2003a) it is expected that there are still many undiscovered species new to science, especially in the relatively unexplored areas of the Cordillera de Talamanca and Panama. If this expectation is correct, Lepanthes may exceed Epidendrum in number of species recorded in the Isthmus (Pupulin and Bogarín, 2014). Figure 1.1. Geography of southern Central America (Costa Rica and Panama) showing the main ranges: Talamanca crossing both countries and San Blas-Darien on the southeast of Panama towards Colombia. 12 Chapter 1 Moreover, the taxonomy of Stelis s.s. is the least developed, and conclusions based on these data are likely biased (Luer, 2003b). Botanical exploration and alpha-taxonomy are therefore tasks that must be promoted with impetus in the region. Some other diverse groups in the Isthmus are Camaridium Lindl., Dichaea Lindl., Oncidium, Pleurothallis, Scaphyglottis Poepp. & Endl., So- bralia Ruiz & Pav., Specklinia Lindl. and Telipogon Kunth. (Fig. 1.2) These groups also maintain a tendency to hold many endemic species. The taxonomic work in these genera has also revealed new species and expanded knowledge on their geographic distributions, encouraging more po- tential case studies to understand the evolution and diversification of Orchidaceae in the Isthmus (Bogarín et al., 2014a; Dressler and Pupulin, 2015; Pupulin et al., 2012). 1.3 Biogeography and endemism of orchids in Costa Rica and Panama About 40% of the orchid species are endemic to the Isthmus. The highest percentages of ende- mism recorded could be related to geological events of its volcanic arc, vicariance and in situ spe- ciation produced by the lifting of the Cordillera de Talamanca. For example, allopatric speciation in Lycaste bruncana Bogarín and L. tricolor Rchb.f. (Fig. 1.3), among other examples found in Brassia R.Br., Epidendrum, Kefersteinia Rchb.f., Oncidium, Pleurothallis and Stelis, indicate an important role of the altitudinal division in vicarance speciation induced by the Talamanca and its climate barrier effect blocking the Caribbean trade winds (Bogarín, 2007; Pupulin, 2001; Pupulin and Bogarín, 2009). The highest percentages of endemism are found in the most diverse genera. For instance, 90% of the species of Lepanthes are endemic and about 50% of the species of Ste- lis and Epidendrum (Table 1.4) occur nowhere else. The study of the factors favoring this high Costa Rica Panama Total Species 1574 1372 2012 Endemics 485 299 784 Genera 199 187 211 Costa Rica Panama Genus Number of species Genus Number of species Epidendrum 207 Epidendrum 221 Stelis 88 Lepanthes 151 Lepanthes 66 Stelis 103 Pleurothallis 54 Camaridium 48 Camaridium 48 Pleurothallis 48 Scaphyglottis 39 Specklinia 44 Sobralia 39 Scaphyglottis 38 Specklinia 34 Sobralia 38 Oncidium 32 Telipogon 37 Dichaea 26 Masdevallia 34 13 Table 1.2. The most diverse genera in Costa Rica and Panama Table 1. 1. Number of species, endemics and genera in Costa Rica and Panama Introduction endemism in Lepanthes is key to understanding its diversification and will be further discussed in the upcoming chapters of this PhD thesis. Other genera also deserve more attention because, although not as diverse, they show high rates of endemism; one of these is Telipogon, in which more than 70% of the species are endemic. Current floristic relationships with other groups of orchids of the Andes is evident. For example, Telipogon is a diverse genus in the highlands of the Isthmus, and its northern distribution is limited. Other genera of South American affinities are Brachionidium Lindl., Fernandezia Lindl. and Pterichis Lindl. that almost reach their northern- most distribution in the Cordillera de Talamanca. About 10 genera are present in Panama but not in Costa Rica. These genera have a strong South American affinity: Discyphus Schltr., Eloyella P.Ortiz, Koellensteinia Rchb.f., Neomoorea Rolfe, Rudolfiella Hoehne and Selenipedium Rchb.f. They range from Central Panama to the southeast of Darien and towards Colombia, indicating a common geological history of this area but different from western Panama and southeast Costa Rica. The geological formation of foothills of Maje, Darien and San Blas in Panama and western Colombia is reflected in the species composition data. Geographical distributions of Dinema Lindl., Euryblema Dressler, Helleriella A.D. Hawkes and Horichia Jenny suggests that these genera might be present in Costa Rica (Bogarín et al., 2014b). On the other hand, 18 genera pres- ent in Costa Rica are still not recorded from Panama. Some of them show a northern distribution such as Arpophyllum La Llave & Lex. and Restrepiella Garay & Dunst. However, Epistephium Kunth, Funkiella Schltr., Lankesterella Ames, Trevoria F.Lehm., Tropidia Lindl. and Warmingia Rchb.f., all with representatives in South America, might be distributed in Panama after all. The bias resulting from less floristic and alpha-taxonomic work in Panama should be reduced in the upcoming years (Bogarín et al., 2013). Costa Rica Panama Genus Endemic species Genus Endemic species Lepanthes 102 Epidendrum 53 Epidendrum 80 Pleurothallis 23 Stelis 37 Stelis 23 Telipogon 31 Lepanthes 21 Pleurothallis 15 Telipogon 17 Sobralia 15 Sobralia 16 Camaridium 14 Masdevallia 9 Specklinia 14 Camaridium 8 Masdevallia 13 Specklinia 7 14 Table 1.3. Genera with most endemic species in Costa Rica and Panama. Genus Species in the Isthmus % endemic species Lepanthes 155 90.12 Epidendrum 133 46.18 Stelis 60 43.80 Table 1.4. The most diverse genera and the % of endemis in the Isthmus of Panama Chapter 1 1.4 Evolutionary diversification and orchid floristic composition There are many factors that can enhance orchid species diversifications such as orogeny, past climatic fluctuations, interactions with other organisms such as mycorrhiza, pollinators, seed dis- persers or key innovations such as colonization (extrinsic) or trait evolution (intrinsic). The main aim of my PhD thesis consisted of studying the factors that led to the formation of the current species composition of Orchidaceae in the Isthmus. Based on our taxonomic experience we have selected Lepanthes and closely related genera as a model group to study the extraordinary species richness and evolution in Costa Rica and Panama and its relationship with the Andean flora. We intend in the future to extend this model to other diverse groups such as Stelis. Epidendrum is Figure 1.2. Some representatives of the major groups of Orchidaceae present in Lower Central America. From left to right: Camaridium campanulatum, Epidendrum nocturnum, Epidendrum (Oerstedella) wal- lisii, Lepanthes matamorosii, L. bradei, Pleurothallis anthurioides, Scaphyglottis pulchella, Stelis trans- versalis, Telipogon panamensis. Photographs by Diego Bogarín. 15 Introduction Figure 1.3. Lankester Composite Digital Plate of Lycaste bruncana, a species from Costa Rica and Pana- ma restricted to the Pacific watershed of Cordillera de Talamanca. Photograph by Diego Bogarín. 16 Chapter 1 Figure 1.4. Lankester Composite Dissection Plates (LCDP) of some representative species of the highly diverse genus Stelis from Costa Rica and Panama. Species are currently under taxonomic review. Photographs by Diego Bogarín. 17 Introduction another interesting group, and it is being evaluated by Hágsater and co-workers so there will be information available in the future. Although Stelis s.s. remains an excellent group as a candidate to study their high evolutionary diversification, the limited taxonomic expertise and little eco- logical information available so far prevented us to address this group (Fig. 1.4). However, some clues about its pollination mechanism (hitherto little-known) indicate that it may be pollinated by gall midges of Cecidomyiidae under conditions that we are still exploring. 1.5 The orchid genus Lepanthes Lepanthes is one of the major genera in the Pleurothallidinae. With over 1,000 spp., the genus ranges from southern Mexico and the Antilles to Peru and Bolivia, with few species in the Gui- anas and Brazil. Plants grow mostly from 1,500 to 3,000 m elevation in humid, often shady places. Highest diversity is found in the Andean region of Colombia and Ecuador with more than 300 species in each country (Luer, 1996b; Luer and Thoerle, 2012) (Fig. 1.6-1.7). Lepanthes is represented in Costa Rica and Panama with about 150 spp. Only two species are shared with Colombia and Ecuador. This may reflect the floristic influence of the Andean region in Costa Rica and Panama at the genus level but not the species level . Species are usually restricted to specific ranges or mountains, and endemism is high. Plants are recognized by the monophyllous ramicauls, enclosed by a series of lepanthiform sheaths and congested, distichous inflorescences. Floral morphology distinguishes Lepanthes from other genera with lepanthiform sheaths (Dra- conanthes (Luer) Luer, Trichosalpinx Luer and Lepanthopsis (Cogn.) Hoehne among others). Flowers are characterized by the ovate to elliptic sepals and the transversely bilobed petals. Lip morphology is complex (Fig. 1.6); the lip is usually bilaminate with the two blades supported by connectives that often lift the blades above the column. The central part of the lip is made up by the body, which is attached to the column. The appendix is developed from the sinuous between the connectives and varies morphologically among the species in different combinations of lobes, hairs, projections, trichomes and membranes. Lepanthes taxonomy has been studied by Luer and Thoerle (2012) and particularly in Costa Rica and Panama by Luer and Dressler (1986), Luer (2003a), Pupulin et al . (2009), Pupulin and Bogarín (2014). Givnish et al., (2015) pointed out that the role of limited dispersal of seeds and ineffective pollinators, limited gene flow, pop- ulation bottlenecks and genetic drift deserve to be further studied, and that Lepanthes would be one of the best study cases for that so we fo- cused on its systematics in Chapters 2 and 3. Although Lepanthes is considered a monophy- letic group, it has been poorly sampled phyloge- netically (Pridgeon et al., 2001). Phylogenetic analyses of the Pleurothallidinae showed that Andinia (Luer) Luer (including Neooreophilus Archila) is not closely related to Lepanthes, and 18 Figure 1.6. Scanning electron microscopy (SEM) of a flower of Lepanthes horichii showing the complex morphology in detail. A. Sepal. B. Petal (lower lobe). C. Lip (lobe). D. Column showing the apical anther. Photographs by Diego Bogarín Chapter 1 flower similarities are homplastic (Wilson et al., 2017) (Fig. 1.8). Neooreophilus species have a similar flower morphology as Lepanthes, and there is some evidence of its pollination by pseudo- copulation (S. Vieira-Uribe, pers. comm. 2015). Neooreophilus is absent in Mesoamerica, and it might be a younger group when compared to Lepanthes, which is widespread in the Neotropics. Phylogenetics of these two groups could help to shed light on this hypothesis. Furthermore, the floral morphology of Lepanthes varies astonishingly around the same scheme in all the >1,000 species known. The flowers are developed above or beneath the leaves or sometimes in inflores- cences surpassing the leaves and the petals and lip tend to be reduced or almost absent in some species. The most common colors of flowers are yellow, red, orange, purple (rarely green) or a combination of these. The appendix of the lip plays an important role in pollination of Lep- anthes flowers. Blanco and Barboza (2005) described the first case of pseudocopulation in the 19 Figure 1.7. Some species of Lepanthes from Costa Rica and Panama. Species are mostly endemic and show a wide range of morphological variation around the same scheme. Note the coloration of the flowers, which might be involved in attraction of pollinators. Photographs by Diego Bogarín. Introduction genus in which male fungus gnats of Bradysia floribunda (Diptera: Sciaridae) visit flowers of L . glicensteinii Luer, apparently attracted by sexual pheromones. The male adheres to the flower appendix during copulation. In this attempt the insect removes the pollinarium with the abdo- men. Calderón-Sáenz (2012) observed the same phenomenon in L. yubarta E.Calderon, which is visited by another species of Bradysia in Valle del Cauca, Colombia. Sciaridae flies, commonly known as dark-winged fungus gnats, are a diverse group of flies with more than 8,000 species worldwide. Eggs are deposited between the lamina of sporocarps of fungi, and the larvae feed on sporocarps and other decaying organic matter such as rotten trunks or plant roots or leaves. Some species are pests of important economic crops such as mushrooms. Blanco and Barboza (2005) and Calderón-Sáenz (2012) clearly described the pollination of Lepanthes but left many evolu- tionary questions un answered. We are studying more cases of pollination in other Lepanthes species where morphological evidence indicates that other parts of the body of fungal gnats are being used such that the pollinia are not always attached to the abdomen. Probably, flowers pro- duce pheromone-like compounds to attract pollinators and we have some preliminary evidence that flowers indeed use this strategy to attract males. The anatomy of the flower was studied in order to find possible secretory structures in- volved in pollinator attraction (Fig. 1.7). Sciarids are attracted by yellow colors. Special traps were designed to catch flies in greenhouses made up by yellow cardboard and petroleum jelly. Although this method proved to be less effective in studying Lepanthes pollination (Godden, 2002), the approach works well in large populations of plants to increase the probabilities of catching gnats carrying pollinia. Sciarid flies have short life cycles (Wilkinson and Daugherty, 1970). Adults usually live less than 7 days, and they are considered poor flyers. Thus the chance to deceive inexperienced males may be high. Sciaridae is a highly diverse group but poorly known. The behavior and natural history of Sciaridae are key to understanding the evolution of Lepanthes. Why is Lepanthes more diverse than closely related genera such as Anathallis Barb. Rodr., Draconanthes, Lankesteriana Karremans, Lepanthopsis, Trichosalpinx and Zootrophion Luer? A hypothesis is that pseudocopulation triggered the high speciation levels in Lepanthes. To study the evolutionary diversification of Lepanthes and the possible triggers of speciation, it was necessary to extend the molecular phylogenetic sampling of the “Lepanthes clade” as de- scribed by Pridgeon et al., (2001) in order to find answers to the evolutionary success of Lepan- thes as compared to its sister genera (see Chapters 2-5). However, the pollination mechanisms 20 Figure 1.8. Floral convergence among the species of Neooreophilus. A-B. and Lepanthes C. Photos: A-B. by Sebastián Vieira-Uribe. C. by Diego Bogarín. Chapter 1 that operate in the sister genera are also important for comparisons with Lepanthes. Observa- tions on the pollination of Trichosalpinx revealed a frequent visitation by biting midges of the Ceratopogonidae family (see Chapters 6-7) . Finally, we also used biogeographical areas within the Neotropics in order to draw accurate conclusions about endemism and species distribution in biogeographical analyses (Chapters 8-9). Aims of the thesis In this thesis, I targeted the orchid genus Lepanthes, one of the six genera of angiosperms that surpasses 1,000 species in the Neotropics, as a study model to investigate the evolutionary processes that promoted species diversifications. To investigate some of the possible factors that shaped the diversification in Lepanthes and related genera we improved the taxonomy of the group by providing a solid phylogenetic framework combined with ancestral state recon- structions, assessing inter-specific relationships in species complexes with hundreds of molec- ular markers, and describing new species, (Chapters 2-5), disclosed a new pollination system, identified morphological characters associated with similar pollination mechanisms (Chapters 6-7) and discussed the impact of biogeographical events and orogeny (formation of the Andes and Central America) on the extant species richness and biodiversity of Lepanthes (Chapter 8-9). This thesis provides new insights in the complex evolution of one of the most species-rich angiosperm lineages in the Neotropics. Outline of the thesis Lepanthes contains more than 1,130 species and new species are constantly being discovered in the Neotropics. An approximate number of the actual species diversity is not yet known and this number tends to increase partially due to the extreme diversity of the genus but also because several regions of the Neotropics continue to be explored and the boost of alpha-tax- onomic studies (Luer and Thoerle, 2012; Pupulin et al., 2018; Pupulin and Bogarín, 2019). In addition, the phylogenetic relationships of the Lepanthes and allied genera were problematic at the start of my PhD project, not because of the lack of sufficient DNA markers but because of insufficient taxonomic sampling and the widespread convergences in reproductive characters. Therefore, in Chapter 2 (Bogarín et al. in review) we presented the integral discussion on the phylogenetics of the Lepanthes clade integrating phylogenetics and morphological evolution of character states. Consequently, in Chapter 3 we proposed a new classification of the Lepanthes clade based on a more extensive taxonomic sampling and the information obtained in Chapter 2 (Bogarín et al., 2018). Similar to the poor understanding of inter-generic relationships, some inter-specific relationships are difficult to understand because of the high morphological simi- larity, especially in floral traits. In addition, these species complexes are challenging to resolve using standard DNA barcoding markers such as nrITS or matK. Therefore, in Chapter 4 we as- sessed the performance of hundreds of innovative molecular markers derived from an anchored hybrid enrichment approach (AHE) to resolve phylogenetic relationships and improve species recognition in the Lepanthes horrida species group (Bogarín et al., 2018). Further, some areas of the Neotropics are rich in Lepanthes species but much floristic work still needs to be done. This 21 Aims and outline of the thesis is for instance the case for Panama, where an underestimation of species is well known but an increase of taxonomic studies is revealing new species or new records from neighboring regions (Bogarín et al., 2013). In this way, in Chapter 5 we revealed two new species of Lepanthes detected during fieldwork (Bogarín et al., 2017). In addition to the systematics and the evolution of morphological traits, pollination studies are key in understanding homoplastic characters in closely related genera and the role of pollinators as drivers of species diversity. However, this is largely unknown because knowledge of pollination systems in the group is still scarce and only the pollination system of Lepanthes is known. Therefore, in Chapter 6 we addressed the pollination of Lepanthes’ closely related genus Trichosalpinx through study of floral anatomy, pollinator behaviour and floral traits shared with other angiosperms to elucidate its pollination mechanism (Bogarín et al., 2018). The similar floral morphology and homoplastic characters described in Chapter 5 among Trichosalpinx and the closely related genera Anathallis and Lankesteriana suggest that they are pollinated by a similar system as shown in Chapter 6. Hence, in Chapter 7 we assessed the micromorphological and histochemical features of floral organs to test a hypothesis on floral convergence in this clade (Bogarín et al., 2018). And finally, to understand the role of abiotic factors such as the impact of the Andean mountains in the diver- sification of Lepanthes in Chapter 8-9 we inferred the biogeographical history and diversifica- tion dynamics of the two largest Neotropical orchid groups (Cymbidieae and Pleurothallidinae), using two unparalleled, densely sampled phylogenies coupled with geological and biological datasets (Pérez-Escobar et al., 2017a). In Chapter 10, I discuss further steps needed to com- pliment the findings presented in my PhD thesis to fully understand and better protect orchid species radiations in the Neotropics. 22 Chapter 1 Taxonomy and systematics of Lepanthes and allies Chapter 2 Phylogenetic comparative methods improve the selection of characters for generic delimi- tations in a hypediverse Neotropical orchid lineage Diego Bogarín, Oscar A. Pérez-Escobar, Adam P. Karremans, Melania Fernández, Jaco Kruizinga, Franco Pupulin, Erik Smets and Barbara Gravendeel Scientific Reports. In review Abstract. Taxonomic delimitations are challenging because of the convergent and variable nature of phenotypic traits. This is particularly evident in species-rich lineages, where the ancestral and derived states and their gains and losses are difficult to assess. However, phylogenetic comparative methods help to evaluate the parallel evolution of a given morphological character, thus enabling the discovery of traits useful for classifications. In this study, we investigate the evolution of selected traits to test for their suitability for generic delimitations in the Neotropical species-richest orchid lineage Lepanthes. We evaluated every generic name proposed in the Lepanthes clade producing densely sampled phylogenies with Maximum Parsimony, Maximum Likelihood, and Bayesian ap- proaches. In addition, we assessed with Ancestral State Reconstructions 18 phenotypic characters that have been traditionally used to diagnose the genera. Our results support the recognition of 14 monophyletic genera and provide solid morphological delimitations. We identified 16 plesio- morphies, 12 homoplastic characters, and 7 synapomorphies, the latter of which are reproductive features mostly related to the pollination by pseudocopulation and possibly correlated with rapid diversifications within Lepanthes. Furthermore, the ancestral states of some reproductive characters suggest that these traits are associated with similar pollination mechanisms promoting homoplasy. Our methodological approach enables the discovery of useful traits for generic delimitations in the Lepanthes clade. This offers various other testable hypotheses for future research on Pleurothalli- dinae orchids because phenotypic variation of some of the characters evaluated here also occur in other diverse genera. 25 2.1 Introduction Taxonomic delimitation is essential to understand, document, and quantify earth’s biodiversity. This is particularly true for species, which are regarded as the fundamental units of biological systems. Species delimitations and their numerous corresponding concepts are still hotly debated, yet relatively little has been discussed regarding supra-specific taxon delimitations (Barkman and Simpson, 2001; De Queiroz, 2007, 2005). Among such higher taxonomic ranges, the genera are important because they inform about discernable trait patterns shared among species groupings (Humphreys and Linder, 2009), and are widely used as biodiversity indicators of biogeographical areas (Gentry, 1986), and even biomes (Ulloa et al., 2017). Generic delimitations are based on several criteria that are often informed by morphological traits, the principle of monophyly, sta- tistical node supports in phylogenies, and even lineage size (i.e. species number). Among these, morphology is perhaps the most common invoked criterion to segregate or subsume species aggregates (Humphreys and Linder, 2009), yet morphological characters are often variable and converge across the angiosperm tree of life (Stull et al., 2018), thus rendering the selection of suitable morphological characters for generic delimitations quite difficult. The orchid family includes about 25,000 species and ca. 750 genera. Its generic classification system is quite dynamic, with hundreds of genera having been subsumed and segregated during the last decade (Chase et al., 2015). Among recalcitrant lineages with complicated generic delim- itations are the Pleurothallidinae, the species-richest subtribe in the Neotropics (5,200 species; (Karremans, 2016; Luer, 2007; Pridgeon et al., 2001)). The high species diversity derived from recent and rapid diversifications and the exceptionally wide spectrum of morphological features have made the classification of this group challenging (Pérez-Escobar et al., 2017a). Previous cladistic and contemporary systematic studies were largely based on morphology (Luer, 1986a; Neyland et al., 1995). Using these studies as a framework, Pridgeon et al. (2001) proposed the first molecular phylogenetic classification of the subtribe by sequencing nuclear and plastid re- gions of 185 selected taxa (3.5% of the species of the Pleurothallidinae). This study laid the foundation for the classification system followed in Genera Orchidacearum (Pridgeon et al., 2005) which divided the subtribe in nine main clades. In the past 10 years, several phylogenetic studies, aimed to increase taxon sampling or add more markers to the previous phylogenetic reconstructions, supported or redefine most of the taxonomic and generic concepts proposed by Pridgeon et al. (2001) and Luer (2006). These phylogenetic re-evaluations covered almost all clades across the subtribe (Abele, 2007; Chiron et al., 2012; Karremans et al., 2013; Karremans et al., 2016; Karremans et al., 2016). One of the few remaining puzzling groups with phylogenetic relationships poorly understood in the Pleurothallidinae is the Lepanthes clade (Bogarín et al., 2018c; Karremans, 2016; Luer, 1986b; Pridgeon et al., 2001) (. 2.1). In its current circumscription, it comprises the genera Ana- thallis Barb.Rodr. (116 spp.), Draconanthes (Luer) Luer (2), Epibator Luer (3), Frondaria Luer (1), Lankesteriana Karremans (21), Lepanthes Sw. (>1200), Lepanthopsis (Cogn.) Ames (44), Trichosalpinx Luer (24) and Zootrophion Luer (26). Moreover, four generic concepts needed to attain monophyly, were recently erected by Bogarín (Bogarín et al., 2018c): Gravendeelia Bog- arín & Karremans (1), Pendusalpinx Karremans & Mel.Fernández (7), Stellamaris Mel.Fernán- dez & Bogarín (1), and Opilionanthe Karremans & Bogarín (1) as well as the reinstatement of 26 Chapter 2 Figure 2.1. Flower morphology of the representatives of the Lepanthes clade: A. Lepanthes. B. Dracon- anthes. C. Pseudolepanthes. D. Stellamaris. E. Frondaria. F. Lepanthopsis. G. Gravendeelia. H. Opil- ionanthe. I. Lankesteriana. J. Pendusalpinx. K. Trichosalpinx. L. Tubella. M. Anathallis. N. Anathallis. O. Zootrophion. P. Zootrophion (Epibator). Photographs A-B, D, F, I, K-O by D.Bogarín, C,G by S. Vieira-Uribe, E by J. Portilla (Ecuagenera), H,J,P by W. Driessen. 27 Generic delimitations in a hypediverse Neotropical orchid lineage Pseudolepanthes (Luer) Archila (10) and Tubella (Luer) Archila (79). The species-richest genus is Lepanthes, which comprises more than 77% of the species of the clade, whereas the remaining genera represent less than 8% of the species diversity each. The Lepanthes clade is widely distributed in the Neotropics ranging from Mexico and Flori- da to southern Brazil and Argentina, including Central America and the Antilles. The species are characterized by infundibular sheaths, also called “lepanthiform sheaths” along the ramicauls of unknown functionality (Luer, 1996b; Pridgeon et al., 2001). These sheaths are unornamented and imbricating in Anathallis, Lankesteriana and Zootrophion, foliaceous with expanded leaf sheaths in Frondaria and sclerotic with ornamentations (spiculate or muriculate) along the ramicauls in the remaining genera (Figs. 2.1-2.2). Regardless of the relative uniformity in plant vegetative characters, flower morphology is highly dissimilar among genera and no single diagnostic floral character distinguishing the group has been recognized. Floral trait variation is most evident in the flower shape (spread, flattened or cupped sepals and petals), color (red, yellow, white, green, purple or maculated), anthesis timing in the inflorescence (simultaneous or successive), shape of sepals, petals and lip (elongated, flattened, ciliated, bilobed), anther position (apical or ventral), pollinaria-associated structures (with or without viscidium), and presence/absence of a synsepal and column foot (Luer, 1986a, 1986b; Pridgeon, 2005) (Fig. 2.1). Previous multi-locus phylogenies strongly supported the monophyly of the Lepanthes clade (Chase et al., 2015; Pridgeon et al., 2005), yet the number of genera to be recognized and their phylogenetic relationships are still unclear. This is likely due to the widespread convergences in reproductive characters in the lineage and the insufficient phylogenetic taxon sampling. Earlier phylogenetic studies in the Pleurothallidinae did not investigated morphological evolutionary patterns, homoplasy and contrasting differences in reproductive traits by combining ancestral state reconstructions (ASR) and a solid phylogenetic framework (Karremans, 2016; Pridgeon et al., 2001). This is essential to test hypotheses of morphological evolution and to disentangle re- calcitrant generic delimitations due to phenotypic similarities. More importantly, theory predicts that synapomorphies or homoplastic characters are attributed to shifts or convergences due to dipteran pollination, but this remains yet to be tested due to the scarce pollination observations across the subtribe. The role of pollinator interactions in the evolution of the Lepanthes clade is currently unknown because only two pollination systems have been reported so far for Lepanthes and Trichosalpinx (Blanco and Barboza, 2005; Bogarín et al., 2018a). Here, we explore the utility of molecular trees and phylogenetic comparative methods to dis- cover suitable morphological characters for generic delimitations. To achieve this, we evaluate the relationships among members of the Lepanthes clade by assessing morphological characters within a phylogenetic framework. We performed ASRs on 18 floral morphological characters using a well resolved phylogenetic inference from nuclear nrITS and plastid matK markers of 122 species covering all recognized genera within the clade (Bogarín et al., 2018c). We want to answer the following questions: (1) which monophyletic genera can be recognized based on a phylogenetic framework? (2) what are the phylogenetically informative characters of each clade based on ASRs? (3) how did such diagnostic morphological characters evolve in the clade? We also provide a detailed generic circumscription of Lepanthes. 28 Chapter 2 Figure 2.2. Vegetative and flower morphology of the characters evaluated: A. repent habit in Ana- thallis. B. caespitose habit with longer inflorescences than leaf in Pseudolepanthes. C. prolific ram- icauls in Tubella. D. ornamented lepanthiform bracts in Trichosalpinx. E. laminar, mobile lip (i) of Trichosalpinx. F. bilobed stigma and glenion (g) in Lepanthopsis. G. Appendix (a) at the lip base of Lepanthes. H. Column foot (cf) and ventral anther in Gravendeelia. I. Bilobed lip (b) and apical anther in Lepanthes. J. Ventral anther (an) and stigma (s) in Anathallis. K. Pollinarium with viscidium (v) and caudicles (c) in Lepanthes. L. Pollinarium with caudicles (c) in Trichosalpinx. Photographs A-L by D.Bogarín, b. by S. Vieira-Uribe. 29 Generic delimitations in a hypediverse Neotropical orchid lineage 2.2 Materials and Methods 2.2.1 Taxon sampling We sampled 148 accessions of 120 species from every generic name erected in the group. We included Anathallis (6 spp.), Draconanthes (1 sp.), Frondaria (1 sp.), Gravendeelia (1 sp.), Lankesteriana (5 spp.), Lepanthes (61 spp.), Lepanthopsis (6 spp.), Opilionanthe (1 sp.), Pen- dusalpinx (8 spp.), Pseudolepanthes (2 sp.), Stellamaris (1 sp.), Trichosalpinx (8 spp.), Tubella (14 spp.) and Zootrophion (6 spp.). Members of the Trichosalpinx subgenus Xenia Luer (five spp.) were not sampled due to unavailability of material. Voucher information, NCBI GenBank accessions, and references for each DNA sequence are listed in Appendix S1 (). A total of 88 sequences were newly generated (49 from nrITS and 39 from matK) and complimented with sequences from previous studies (Karremans, 2014; Pérez-Escobar et al., 2017a; Pridgeon et al., 2001). Acianthera cogniauxiana (Schltr.) Pridgeon & M.W. Chase and Acianthera fenestrata (Barb.Rodr.) Pridgeon & M.W.Chase were chosen as outgroups based on Pridgeon et al., (2001). 2.2.2 Phenotypic character selection We scored 18 macro-morphological characters (Table 2.1) which are considered taxonomically informative or ecologically important that have been used to characterize some of the genera. Data were obtained by direct observations from herbarium material (CR, AMES, JBL, K, L, PMA, UCH, W herbaria) and living material collected in the field or cultivated at Lankester Botanical Garden, the Hortus botanicus Leiden or private orchid collections. Observations were complimented with morphological data compiled from monographs on the Pleurothallidinae (Luer, 1986a; b, 1991, 1996a, 1997a, 2004, 2006; Pridgeon, 2005; Luer and Thoerle, 2012) and with digital documentation (photographs and drawings) from JBL databases. We generated additional macro-morphological data with a Scanning Electron Microscope (SEM) using fixed flowers dehydrated in a series of ethanol solutions (70%–96%–≥99.9%) and acetone ≥99.8%. Critical-point drying was performed in an Automated Critical Point Dryer Leica EM CPD300 (Leica Microsystems, Wetzlar, Germany) following the manufacturer’s procedures. Samples were sputter-coated with 20 nm of Pt/Pd in a Quorum Q150TS sputter-coater and observed with a JEOL JSM-7600F (Tokyo, Japan) field emission scanning electron microscope, at an acceler- ating voltage of 10 kV. For macro-photography we used a Nikon® D7100 (Tokyo, Japan) digital camera and a PB-6 Nikon bellows. We edited the images in Adobe Photoshop® CC (Adobe Systems Inc., California, U.S.A). 2.2.3 DNA extraction We extracted total genomic DNA from about 50-100 mg of silica gel dried leaf/flower tissue. Each sample was placed in 2 ml Eppendorf® tube with three glass beads (7 mm) and sterile sand. The tubes were frozen in liquid nitrogen for about 1-2 minutes and powdered in a Retsch MM 300 shaker for 3 minutes. We followed the 2× CTAB (Hexadecyltrimethylammonium bromide) protocol for isolating DNA (Doyle and Doyle, 1987). DNA was quiantified with a Qubit 3.0 Fluorometer (TermoFischer Scientific®). 30 Chapter 2 2.2.4 Amplification, sequencing and alignment The polymerase chain reaction (PCR) mixture, the primers for the nrITS (17SE and 26SE) and plastid matK (2.1aF and 5R) regions and amplification profiles followed Karremans (Karremans et al., 2016). Sanger sequencing of both regions was conducted by BaseClear (https://www.ba- seclear.com) on an ABI 3730xl genetic analyzer (Applied Biosystems, Foster City, California, U.S.A). Sequences were deposited in NCBI GenBank. We used Geneious® R9 (Biomatters Ltd., Auckland, New Zealand (Kearse et al., 2012)) for the editing of chromatograms and pairwise alignment. Sequences were aligned in the online MAFFT platform (Multiple Alignment using Fast Fourier Transform, http://mafft.cbrc.jp/alignment/server/) using default settings. We adjust- ed and trimmed the resulting alignment manually. The concatenated dataset (nrITS +matK) was built with Sequence Matrix v100.0 (Vaidya et al., 2011). When sequences were not available, they were analyzed as missing data. Table 2.1. Characters and scoring of the 18 morphological traits assessed with ancestral character esti- mations and the main references illustrating or discussing these characters. Characters States References Habit (0) caespitose; (1) repent (Luer, 1986a; Pridgeon, 1982; Stern et al., 1985) Ramicauls (0) non-prolific; (1) prolific (Luer, 1986a; Pridgeon, 1982; Stern et al., 1985) Ramicauls’ bracts (0) unornamented; (1) ornamented; (2) foliaceous (Luer, 1991, 1990) Inflorescence (0) simultaneously flowering; (1) successively flowering (Luer, 1986a, 1983) Inflorescence length (0) shorter than leaves; (1) longer than leaves (Luer, 1986a, 1983) Flowers (0) fully opening; (1) bud-like (Luer, 1982) Dorsal sepal concavity (0) concave; (1) flattened (Luer, 1996b; Luer, 2006) Synsepal (0) absent; (1) present (Luer, 1986a; Luer, 1996b; Luer, 1997) Sepal shape (0) oblong-acute; (1) ovate-acuminate (2) ovate- acute (Luer, 1986a; Luer, 1996b; Luer, 2006) Petals shape (0) dissimilar; (1) subsimilar (Luer, 1997; Luer, 2006, 1986a) Lip shape (0) laminar; (1) bilobed (Luer, 1996b; Luer, 2006) Lip mobility (0) mobile; (1) sessile (Bogarín et al., 2018a; Luer, 2006) Glenion of the lip (0) absent; (1) present (Luer, 1991) Appendix of the lip (0) absent; (1) present (Luer, 1996b) Column foot (0) absent; (1) present (Benzing and Pridgeon, 1983; Luer, 1986a) Stigma shape (0) entire; (1) bilobed (Luer, 1991, 1990) Anther position (0) ventral; (1) dorsal (Luer, 1996b) Pollinaria-associated structures (0) with caudicles; (1) with caudicles+viscidium (Karremans et al., 2013; Stenzel, 2000) 31 Generic delimitations in a hypediverse Neotropical orchid lineage 2.2.5 Phylogenetic analyses We analyzed the individual and concatenated datasets of nrITS and matK with Bayesian infer- ence (BI), maximum likelihood (ML) and maximum parsimony (MP) analyses. The model of evolution and the parameters were calculated using the Akaike Information Criterion (AIC) in jModelTest2 v2.1.7 (Darriba et al., 2012). All analyses were run in the CIPRES Science Gateway V. 3.1 (http://www.phylo.org/sub_sections/portal/) (Miller et al., 2010). To evaluate the incongru- ence between plastid and nuclear datasets we followed the pipeline implemented by Pérez-Esco- bar et al. (2017a) using the Procrustean Approach to Cophylogeny (PACo) application (Balbuena et al., 2013) in R (http://data- dryad.org/review?doi=doi:10.5061/dryad.q6s1f). This procedure identifies potential conflicting outliers contributing to incongruent phylogenies. The matK se- quences from the retrieved conflicting terminals were removed and replaced by missing data because inferences derived from plastid markers are usually more in conflict with morphological observations as compared with inferences derived from nuclear markers (Pérez-Escobar et al., 2016a). A new concatenated matrix was re-aligned using the cleaned matK dataset and then analyzed with BI, ML, and MP approaches. These analyses were contrasted with the original inferences from concatenated datasets. We performed the Bayesian inference analyses with MrBayes v.3.2.6 on XSEDE (Huelsen- beck and Ronquist, 2001) with the following parameters: number of generations Ngen=50×106 for the combined and individual datasets, number of runs (nruns=2), number of chains to run (nchains=4), temperature parameter (temp=2) and sampling frequency of 1,000 yielding 50,001 trees per run. The log files from MrBayes were inspected in Tracer v.1.6 to check the con- vergence of independent runs (i.e. with estimated sample size (ESS) > 200). The initial 25% of trees were discarded as burn-in and the resulting trees were used to obtain a 50% majori- ty-rule consensus tree. Maximum likelihood analyses were performed with RAxML-HPC2 on XSEDE (8.2.10) (Stamatakis et al., 2008) choosing the GTRGAMMA model for bootstrapping and 1,000 bootstrap iterations. Parsimony analyses were performed with PAUPRat: Parsimony ratchet searches using PAUP* (Nixon, 1999; Sikes and Lewis, 2001; Swofford, 2002) with 1,000 ratchet repetitions, seed value=0,20% percent of characters to perturb (pct=20), origi- nal weights 1 for all characters (wtmode=uniform) and a tree bisection-reconnection branch swapping algorithm (swap=TBR). The 50% majority rule consensus trees for ML and MP were obtained with PAUP v4.0a152. and observed in FigTree v.1.3.1. The statistical support of the clades was evaluated with the values of posterior probability (PP) for BI reconstruction, boot- strap for ML (MLB) and parsimony bootstrap for MP (MPB). The support values (PP) were added to the branches on the Bayesian 50% majority-rule consensus tree with additional support values shown for ML and MP when the same topology was retrieved. We considered clades with MPB ≥ 70%, MLBS ≥ 70% and PP ≥ 0.95% as well supported. To investigate phylogenetic relationships among genera, we also conducted a network analysis with 3,000 tree replicates of the BI inference of the combined dataset in Splits Tree4 v.4.11.3 (Huson and Bryant, 2006) with a 0.20 cutoff value. Resulting trees were manipulated with R programming language (R Core Team, 2017) under R Studio (Gandrud, 2013) using the packages APE, ggtree and phytools (Paradis et al., 2004; Revell, 2012; Yu et al., 2017). Final trees were edited in Adobe® Illustrator CC (Adobe Systems Inc., California, U.S.A). 32 Chapter 2 To obtain ultrametric trees for the character evolution assessments we estimated the diver- gence times in BEAST v.1.8.2 using the CIPRES Science Gateway (Miller et al., 2010). The clock-likeness of the data was tested by observing the coefficient of variation (CV) of relaxed clock models. Speciation tree model selection was achieved by executing the Bayes factor test on Yule Process (Y), Birth Death-Process (BD) and Birth-Death-Incomplete Sampling (BDIS) models under strict and uncorrelated lognormal molecular clock models. For each model, we assigned a normal prior distribution of 16.45 (±2.5 standard deviations) Ma to the root node of the Lepanthes clade and 12.93 (±2.5 standard deviations) Ma to the node of Zootrophion with the remainder of the members of the Lepanthes clade using the values calculated from the fossil-cal- ibrated chronogram of the Pleurothallidinae by Pérez-Escobar et al. (2017a). We performed two MCMC with 50×106 generations and sampling every 1,000 generations with a Marginal likeli- hood estimation (MLE) of 50 path steps, 10×105 length of chains and log likelihood for every 1000 generations. We inspected the convergence of independent runs size in Tracer v.1.6 as explained above. To compare the divergence time estimates among the speciation models (Y, BD and BDIS) we used Bayes factors calculated with marginal likelihood using stepping stone sampling derived from the MLE path sampling. 2.2.6 Ancestral State Reconstruction (ASRs) Ancestral state reconstructions were assessed with ML, stochastic character mapping (SCM), and BI using phylograms and ultrametric trees. For the ML approach we explored the following models: equal rates (ER), symmetrical (SYM) and all rates different (ARD). We relied on the re-rooting method of Yang et al. (1995) and the function ACE implemented in the R-package phytools. The best-fitting model was selected by comparing the log-likelihoods among these models using likelihood ratio tests. Scaled likelihoods at the root and nodes were plotted in the time-calibrated consensus phylogenetic tree. For the stochastic mapping analyses based on joint sampling we performed 100 replicates on 100 randomly selected trees (10,000 mapped trees) from the best fitting time-calibrated BEAST analysis. The trees were randomly selected using the R function samples.trees (http://coleoguy.blogspot.de/ 2012/09/randomly-sampling-trees. html). Results of transitions and the proportion of time spent in each state were calculated and summarized in phytools with the functions make.simmap and describe.simmap (Bollback, 2006; Revell, 2012). These analysis were performed following the scripts by Portik and Blackburn (2016). ML and BI inferenced were executed in the program BayesTraits V3 (Pagel, 1999, 1994; Pagel and Meade, 2006). To account for phylogenetic uncertainty, ancestral character estimates were calculated using a randomly sampled set of 1000 trees from the post burnin sample of the 50,000 ultrametric trees obtained from the best fitting time-calibrated BEAST analysis as described above. We used the option AddNode for reconstruction of internal nodes of interest comprising every generic group of the Lepanthes clade and the root node. For the ML approach, we used the method Multistate with 10 ML attempts per tree and 20,000 evaluations in order to preliminary assess prior distributions. For the BI, we chose the method Multistate and MCMC parameters of 30,010,000 iterations, sample period of 1,000, burnin of 10,000, auto tune rate deviation and stepping stones 100 10,000. We used the method Reversible-Jump MCMC with hyper-prior exponential to assess the best fitting models in proportion to their posterior probabil- 33 Generic delimitations in a hypediverse Neotropical orchid lineage Figure 2.3. The 14 genera recognized in the Lepanthes clade in the 50% majority-rule consensus tree based on BI analysis of concatenated dataset. Plotted branch values for MPB, MLBS and PP are given for each well-supported clade of interest. Letters represent genera and numbers clades grouping the genera. Photographs A,B, D, F-G, J-N by D.Bogarín, C by S. Vieira-Uribe, E by J. Portilla (Ecuagen- era), H-I by W. Driessen. A B C D E F G H I J K L M N Outgroups Posterior probability / ML boostrap Parsimony boostrap Lepanthes (A) Draconanthes (B) Lepanthopsis (F) Frondaria (E) Stellamaris (D) Pseudolepanthes (C) Gravendeelia (G) Opilionanthe (H) Pendusalpinx (I) Lankesteriana (J) Trichosalpinx (K) Tubella (L) Anathallis (M) Zootrophion (N) Epibator (O) Lineages 1-10=Clades A-I=Genera Tu. dura WD42 Pe. sijimii AK5994 L. latisepala DB11102 Tu. todziae MF540 L. bradei AK5267 T. blaisdellii MF2 L. cuspidata AK6239 Ps. cf. colombiae AK6456 L. sandiorum DB8171 L. mentosa DB11533 Tu. cedralensis FP7049 L. tristis DB11294 L. eximia DB9600 L. maduroi DB2974 A. rabei AK4794 T. blaisdellii DB292 L. cascajalensis DB4836 T. minutipetala MF446 Ls. obliquipetala AK5626 Tu. arbuscula DB8881 Ep. ximeniae AK6502 L. monteverdensis DB11482 L. ribes AP153 L. horichii AK5507 A. pabstii AK4821 La. cuspidata DB9619 Ls. apoda AP126 G. chamaelepanthes DB11881 Z. vulturiceps FP3960 S. pergrata DB12038 L. williamsii DB11292 Pe. sp. WD39 Tu. cedralensis AK6010 L. elata DB10554 L. lindleyana DB8392 T. ringens MF577 A. linearifolia JH2336 A. rabei DB12050 L. droseroides MF945 Ps. cf. colombiae AK6455 Z. gracillenthum AK5282 L. woodburyana JH2931 Tu. nymphalis AK5950 L. rafaeliana DB11658 L. velosa FP6504 L. hermansii FP8611 L. blephariglossa DB9604 L. candida DB11656 L. demissa DB2981 T. blaisdellii AK5308 L. confusa DB3087 T. reflexa DB4075 L. nycteris DB11875 L. disticha AK4589 Pe. cfpatula WD41 L. gustavoromeroi DB11293 L. matamorosii AK5661 Ac. fenestrata JL Orchids s.n. Tu. sp. DB9739 Z. hypodiscus JBL01480 T. blaisdellii K19977412 L. machogaffensis DB10522 T. rotundata AK4386a L. kleinii FP7999 L. terborchii DB11876 L. saltatrix WD26 Tu. parsonsii AK3305 Tu. fruticosa JBL11580 L. calodyction DB11872 L. cloesii DB11877 Ls. astrophora AK5766 L. blepharistes FP8720 G. chamaelepanthes AK4815 Z. machaqway AK6505 T. memor DB6462 L. glicensteinii AK5818 Tu. todziae AK3983 Ep. hirtzii AK4848 Pe. descendens AK4866 S. pergrata DB5635 L. caprimulgus DB11874 Tu. parsonsii AK3302 Tu. dirhamphis DB11882 A. lewisiae DB1056 L. elegans DB7606 D. aberrans AK5978 L. turialvae DB2394 Pe. dependens CvdB2011 Pe. sp. WD40 T. reflexa MF195 L. whittenii MF909 L. ferrelliae FP8806 L. dikoensis DB1625 L. sijmii DB11879 L. regularis DB7756 A. burzlaffiana AK4857 T. minutipetala FP7581 L. calliope DB11873 L. mystax DB11446 Tu. alabastra AK5540 T. memor DB8696 La. barbulata AK5750 T. orbicularis MF65b G. chamaelepanthes AP127 La. barbulata AK5447 L. dolabriformis DB10375 Pe. vasquezii AK6496 Ls. prolifera AK5722 A. peroupavae AK5759 L. siboei DB9927 Ls. floripecten AK3006 Ep. hirtzii AK6503 L. ankistra AK6147 Ps. colombiae AK6458 L. cribii FP8711 L. atrata DB11053 F. caulescens CL18778 O. manningii DB11883 T. orbicularis JH1349 T. pringlei AK6706 La. barbulata DB8606 La. duplooyi AK4888 Ls. floripecten CvdB2063 Ac. cogniauxiana AK5879 La. fractiflexa DB8988 Pe. berlineri JH1605 Tu. robledorum AK5491 Tu. pusilla DB11841 Ls. floripecten DB7795 L. olmosii DB3005 T. pringlei AK6463 Ls. ubangui WD L. stenorrhyncha DB11517 L. brunnescens DB2994 Tu. notosibirica AP225 L. queveriensis DB10854 G. chamaelepanthes PL459a Pe. berlineri AK5770 T. orbicularis AR6474 L. variabilis AK6380 L. martinae DB11878 L. gargantua DB11868 La. barbulata AK5187 S. pergrata DB6502 L. spadariae DB11676 L. myiophora FP7971 La. casualis AK6190 L. montisnarae AK6536 0.94 / 58 100 1.0 / 92 100 0.97 /56 100 0.96 /78 0.93 /32 100 1.0/99 100 1.0/100 100 1.0/98 100 1.0/100 100 1.0/100 100 1.0/100 100 1.0/54 100 0.87/30 100 1.0/98 100 1.0/94 100 1.0/99 100 1.0/91 100 0.96/54 100 1.0/97 100 0.98/72 100 1.0/64 100 1.0/97 100 1.0/100 1001 2 3 4 5 7 8 11 13 F G J I K L M O N 0.02 Zootrophion + Epibator Anathallis Tubella Lankesteriana Pendusalpinx Trichosalpinx s.s Lepanthes Opilionanthe Gravendeelia Lepanthopsis Stellamaris Pseudolepanthes Draconanthes Frondaria A B C D E H Outgroup (Acianthera) 1.0/100 100 1.0/100 100 0.98/80 6 9 10 12 14 34 Chapter 2 ities according to the MCMC approach. We chose the hyper-prior approach as recommended by Meade and Pagel (2016) in order to reduce the arbitrariness when choosing priors. Therefore, we selected the option reversible jump hyper-prior exponential with prior distribution set according to the transition ranges obtained from a preliminary ML analysis. The input files for BayesTraits V3 were partially constructed with Wrappers to Automate the Reconstruction of Ancestral Char- acter States (WARACS) (Gruenstaeudl, 2016). The BayesTraits outputs files were analyzed in R with the BayesTraits wrapper (btw) by Randi H Griffin (http://rgriff23.github.io/projects/btw. html) and other functions from btrtools and BTprocessR (https://github.com/hferg). The MCMC stationarity of parameters (ESS values >200) and convergence of chains were checked in Tracer v1.6.0 and plotted in R with the packages coda (Plummer et al., 2006) and the function mc- mcPlots of BTprocessR. We reconstructed the ancestral states for all nodes of the tree and plotted the mean probabilities retrieved at each node with phytools. 2.3 Results Matrix statistics of the 148 accessions from the 120 species (including two outgroup acces- sions) and parsimony information for nrITS, matK and concatenated datasets are summarized in Appendix S2. 2.3.1 Gene trees The inferences of the BI, ML and MP from the nrITS dataset yielded similar topologies and high support for the 14 genera recognized as members of the Lepanthes clade but with some differenc- es in the topology among the relationships of those clades (Appendices S3,S4). Some differences were observed in the placement of Anathallis, Lankesteriana, Pendusalpinx, Trichosalpinx and Tubella and in the position of L. obliquipetala, which was placed outside the clade Lepanthop- sis+Gravendeelia. The relationships among Lepanthes, Draconanthes, Pseudolepanthes, Stella- maris were consistent. In contrast, the inferences from the matK dataset showed several polyto- mies and low support values for most of the clades Appendices S4,S5). 2.3.2 Incongruence between nuclear and plastid datasets A total of 24 terminals were detected as incongruent with ML and 34 with BI. Of those, 20 terminals were retrieved as incongruent by both inferences (Appendix S1; S6). The topology of the BI, MP and ML trees inferred from the concatenated datasets excluding/including the plastid conflicting sequences recognized essentially the same generic clades but showed some differenc- es in the topology and support values in their intergeneric relationships (Appendices S1, S6, S7). 2.3.3 Concatenated approach (nrITS + matK) Consistent with the inferences based on nrITS, the BI, ML and MP analyses from the concatenat- ed dataset converged in the same generic groupings with high support values for all the genera 35 Generic delimitations in a hypediverse Neotropical orchid lineage of the Lepanthes clade (Fig. 2.2 and Appendix S1). The support values slightly increased after removing the potential outliers from the plastid dataset. In contrast, despite the consistent topol- ogies and high support obtained for all genera, the relationships among them differed using the original datasets (as well as the nrITS dataset alone). However, these relationships were higher supported in the analyses after removing the detected potential outliers from the matK dataset and the phylogenetic relationships obtained were topologically most similar among BI, ML and MP (Fig. 2.3, Appendices S6,S7). In addition, we show the support values of the inferences with/ without PACo in Appendix S6. Consistent with the high support values obtained with BI infer- ence, the inferred network did not show phylogenetic uncertainty in the clades of the 14 genera of the Lepanthes clade. 2.3.4 Phylogenetic relationships and generic clades We obtained strong support for recognizing 14 genera within the Lepanthes clade (Figs. 2.3-2.4). Lepanthes (Clade A) was supported as monophyletic in all the analyses (MPB=100, MLBS=100 and PP=1.0) and sister to Draconanthes (Clade B). The clustering of Lepanthes+Draconanthes was well supported in all the analysis (MPB=100%, MLBS=100% and PP=1.0). The accessions of Pseudolepanthes (Clade C) grouped together with high support (MPB=100%, MLBS=100% and PP=1.0) and this genus was sister to Lepanthes+Draconanthes (Clade 1). The accessions of Stellamaris pergrata (Ames) Mel.Fernández & Bogarín (Clade D) were well supported and the group was sister to Lepanthes+Draconanthes+Pseudolepanthes (Clade 2) (MLBS=80% and PP=0.98). When phylogenetic incongruence was not considered, these two genera clustered in a clade with strong support in the MP tree (MPB=100). The genus Frondaria (Clade E) was found to be related to Lepanthes, Draconanthes, Pseudolepanthes, Stellamaris (Clade 3), well supported (MPB=100 and PP=0.97) but lacking support in the ML analysis (MLBS=56%). Clade 4 made up by Clade 3+Frondaria and comprised the species more related to the core of Lepan- thes whereas Lepanthopsis (Clade F) and Gravendeelia (Clade G) both clustered in Clade 6 as its sister group (Clade 5). Most of the nodes of these clades were well supported (MPB>100%, MLBS>72% and PP>0.98) with the only exception being Clade 6 with low support for ML but well supported by MPB>100% and PP>0.98. The genus Opilionanthe was sister to Clade 5 + Clade 6 with high support for MPB=100%, moderately supported by BI (PP=0.94) and low support for ML (MLBS=58%). Topologically, Opilionanthe always clustered apart from the other generic clades discussed here. Related to the groups of Clade 7 (members of the core of Lepanthes and Lepanthopsis) was a group consisting of species related to Trichosalpinx s.s. (Clade K), Pendusalpinx (Clade J) and Lankesteriana (Clade I) all highly supported as genera (MPB=100%, MLBS≥94% and PP=1.0). This topology was retrieved with high to moderate support (MPB=100%, MLBS≥54% and PP≥0.96) after removing incongruences. Tubella (Clade L) and Anathallis (Clade M) were highly supported as genera (MPB=100%, MLBS=100% and PP=1.0). The internal relationships of Clade 12 received low support with ML (MPB≤30%) and BI (PP≤0.87) but high support by MP (MPB=100%). Clade 14 comprising Zootrophion (Clade N) and Epibator (Clade O),was well supported in all the analyses (MPB=100%, MLBS≥98% and PP=1.0). The most constant well supported topologies among all the analyses were the clus- tering of Zootrophion (MPB=100%, MLBS≥99%, PP=1.0), Lankesteriana and Pendusalpinx 36 Chapter 2 (MPB=100%, MLBS≥91%, PP=1.0), Lepanthes+Draconanthes (MPB=100%, MLBS≥88%, PP=1.0) and the clustering of the genera related to the core of Lepanthes (Clade 4) with Lepan- thopsis+Gravendeelia (Clade 6) (MPB=100%, MLBS≥88% and PP=1.0). 2.3.5 Character evolution The ASR were based on the one-rate model ER which was consistently better than the SYM and ARD models (Appendices S8,S10). These estimations were obtained using phylograms from MrBayes and ultrametric trees from BEAST calculated under the BD as the best model spe- ciation model according to Bayes Factors test (Appendix S11). Estimations based on the Re- versible-Jump MCMC model yielded similar results compared to the rates obtained with SCM (Appendices S12, S14). For the MCMC approach with BayesTraits V3 the best results were obtained with the hyperprior adjusted to the previously obtained ML transition rates (from 0 to 0.03). The ACE, SIMMAP and re-rooting methods yielded identical scaled-likelihoods at the root state and the estimations with MCMC revealed essentially the same results obtained with 0.01 Pendulsapinx Lankesteriana Trichosalpinx Tubella Stellamaris Lepanthes Draconanthes Pseudolepanthes Frondaria Lepanthopsis Gravendeelia Opilionanthe Zootrophion +Epibator Outgroup Anathallis Figure 2.4. Split network showing the 14 genera of the Lepanthes clade inferred from 3,000 tree replicates of the BI inference. The network shows well supported groups without uncertainty in the relationships. 37 Generic delimitations in a hypediverse Neotropical orchid lineage ACE and SIMMAP with ambiguous estimations for the characters of inflorescence length and synsepal (Table 2.2). Characters states of the common ancestor suggest that plesiomorphic fea- tures are a caespitose habit with non-prolific, unornamented ramicauls, simultaneously flowering inflorescences, fully opening flowers with concave, ovate-acute dorsal sepals, dissimilar petals, the presence of a column foot, a laminar, mobile lip without glenion and a ventral anther with entire stigma (Table 2.3). The most common character state transitions are: a caespitose to repent/pendent habit, orna- mented to unornamented bracts, non-prolific to prolific ramicauls, simultaneously flowering to successively flowering inflorescences, shortening of inflorescences, fully opening flowers to bud- like flowers, ovate-acute to ovate-acuminate/oblong-acute sepals, concave to flattened dorsal sepals, dissimilar to subsimilar petals, loss of a column foot and synsepal, movable to sessile lip, entire to bilobed stigma, ventral to dorsal anther and pollinarium with naked caudicles to caud- icles with a viscidium (Figs. 2.5-2.6). Probabilities favoring reversal transitions from prolific to non-prolific ramicauls, foliaceous to ornamented/unornamented bracts, repent to caespitose hab- it, bud-like to opening flowers, subsimilar to dissimilar petals, oblong-acute to ovate-acuminate/ ovate-acute sepals, presence of a glenion to absence, sessile to mobile lip, absence of a column foot to presence, dorsal to apical anther, bilobed to entire stigma and pollinarium with caudicles and a viscidium to lack of a viscidium, were found to be unlikely. Lip shape from laminar to bi- lobed and vice-versa showed a similar probability (Figs. 2.5-2.6). Twelve homoplastic characters and seven synapomorphic characters were detected (Table 2.3). The combination of a sessile lip, absence of a column foot, dorsal anther and pollinarium with caudicles and viscidium are fea- tures only observed in Lepanthes, Draconanthes, Pseudolepanthes and Lepanthopsis, whereas mobile lips, a column foot, ventral anther and pollinarium with caudicles are observed in all other genera investigated. 2.4 Discussion 2.4.1 Phylogenetics of the Lepanthes clade In this section, we discuss the nomenclatural changes needed to redefine the Lepanthes clade as proposed by Bogarín (Bogarín et al., 2018) as well as the relationships among these genera based on the phylogenetic insights and morphological evolution of key characters as presented in this study. Basically, the Lepanthes clade comprises four main clades: Zootrophion, Anathallis, Trichosalpinx and Lepanthes. Zootrophion is, with the inclusion of Epibator, confirmed as one of the early diverging clades. The next early diverging clade is Anathallis, which is here related to Tubella (see further discussion on Trichosalpinx s.l). Anathallis was initially re-established for the species of Pleurothallis subgenus Acuminatia sect. Alata and Pleurothallis subgenus Speck- linia sect. Muscosae (Luer, 1999). Anathallis is confirmed monophyletic with the inclusion of Panmorphia Luer, and the exclusion of members of Pleurothallis subgenus Acuminatia sect. Acuminatae, that belong to Stelis s.l. (Karremans et al., 2013). In addition, Karremans (2014) established the genus Lankesteriana because its members were closely related to Pendusalpinx rather than to Anathalllis s.s. as suggested by Pridgeon et al. (2001). Trichosalpinx as previously 38 Chapter 2 circumscribed (Luer, 1997; Pridgeon et al., 2005) is confirmed as polyphyletic, and therefore recircumscribed. The species belonging to Pendusalpinx and Tubella are confirmed to be un- related to Trichosalpinx and therefore excluded, while the genera Gravendeelia, Opilionanthe, Pseudolepanthes and Stellamaris are placed for the first time in a phylogenetic framework and recognized as distinct (Bogarín et al., 2018). The polyphyly of Trichosalpinx was suggested in previous studies but the former genera were not evaluated or the sampling was too incomplete to allow a redefinition of these groups (Karremans, 2016). The relationships recovered here also suggest that only Pendusalpinx and Lankesteriana are closely related to Trichosalpinx s.s. As suggested by Pridgeon et al. (2001), members of Tubella are isolated from Trichosalpinx and Pendusalpinx (P. berlineri) but the relationships of this genus were not clearly established due to low support (Pérez-Escobar et al., 2017a) recovered Tubella as sister to Lankesteriana, Pen- dusalpinx and Trichosalpinx s.s. However, here with the inclusion of members of the clade not previously evaluated (Gravendeelia, Opilionanthe, Pseudolepanthes and Stellamaris), this rela- tionship changed and Tubella is now recovered as sister to Anathallis. The most recently diverging clade of the Lepanthes clade consists of Lepanthes and its al- lies: the genera Draconanthes, Gravendeelia, Lepanthopsis, Opilionanthe, Pseudolepanthes and Stellamaris. With the exception of Draconanthes and Lepanthopsis, these genera were formerly all treated under Trichosalpinx s.l. However, we confirm here that they are closely related to Lepanthes and Lepanthopsis rather than to Trichosalpinx s.s. In addition, Lepanthopsis is found monophyletic with the inclusion of Expedicula. In the next sections we discuss the morphologi- cal characters supporting the new classification of the Lepanthes clade proposed here. 2.4.2 Morphological evolution Our character reconstructions improved the understanding of the evolution of phenotypic traits used to classify the genera of the Lepanthes clade. We identified homoplastic characters, that are not suitable for generic circumscriptions, as well as synapomorphies that are useful to base classifications on (Table 2.3). Plant habit (caespitose or repent) evolved several times with a higher transition frequency from caespitose to repent. This was found for other groups within the Pleurothallidinae as well, possibly as an adaptation to different environments. Prolific ramicauls evolved from nonprolific ones independently in four clades. The lack of ornamentation of the ramicauls confused taxonomists as the close relationship of Zootrophion, Anathallis and Lank- esteriana with Lepanthes, Lepanthopsis and Trichosalpinx s.l. was not recognized previously. In addition, a combination of plesiomorphic and homoplastic characters in Trichosalpinx s.l., such as the ornamentation of the ramicauls, concave dorsal sepals, ovate-acuminate, caudate petals, mobile, laminar lips with a column foot and ventral anthers caused misclassifications of the now separated genera Gravendeelia, Pendusalpinx, Opilionanthe and Stellamaris. Assessment of oth- er potential diagnostic traits was needed for these genera in order to avoid a classification based on homoplastic characters. For example, the synapomorphic sub-similar petals in Opilionanthe are a diagnostic feature of the genus, showing a very low probability of transition back to the ancestral state, dissimilar petals. Inflorescence type and length are also variable characters in the Pleurothallidinae (Luer, 1986a). Although groups show trends towards the presence of one of the states only, there are 39 Generic delimitations in a hypediverse Neotropical orchid lineage always exceptions to the rule. For example, all the species of Lepanthes studied here have inflo- rescences shorter than the leaves but some species (not studied here) have inflorescences longer than the leaf. The opposite is observed in Trichosalpinx (Luer, 1997). The ancestral traits re- covered for the anther position, presence of a column foot, pollinarium type, and lip mobility suggest that these are associated with the pollination mechanism. In general, flowers with a column foot, movable lips, a dorsal anther and a pollinarium without viscidium are pollinated by insects that enter the flower using the laminar lip. When trying to move in reverse to depart from the flower, the dorsal part of the insect scrapes the dorsal anther off the column in the area of the caudicles and removes the pollinarium (Bogarín et al., 2018a; Borba et al., 2002; Karremans et al., 2015b; Pansarin et al., 2016). This mechanism predominates in Zootrophion, Tubella, Anath- allis, Trichosalpinx, Lankesteriana, Pendusalpinx, Opilionanthe, Gravendeelia, Frondaria and Characters ML(ACE) SCM (SIMMAP) BI (RevJump) 0 1 2 0 1 2 0 1 2 Habit: (0) caespitose; (1) repent 0.99 0.01 - 0.99 0.01 - 0.99 0.01 - Ramicaul growth: (0) non-prolific; (1) prolific 1 0 - 1 0 - 1 0 - Bracts of ramicauls: (0) unornamented; (1) ornamented; (2) foliaceous 0.78 0.21 0 0.82 0.18 0 0.73 0.19 0.08 Inflorescence: (0) simultaneously flower- ing; (1) successive 0.95 0.05 - 0.97 0.03 - 0.98 0.02 - Inflorescence length: (0) shorter; (1) longer (than leaves) 0.43 0.57 - 0.46 0.54 - 0.07 0.93 - Flower appearance: (0) fully opening; (1) bud-like 1 0 - 1 0 - 1 0 - Dorsal sepal concavity: (0) concave; (1) flattened 1 0 - 1 0 - 1 0 - Synsepal: (0) absent; (1) present 0.07 0.93 - 0.06 0.94 - 0.47 0.53 - Sepals shape: (0) oblong-acute; (1) ovate-acuminate (2) ovate-acute 0.01 0.01 0.98 0.01 0.01 0.98 0.29 0.08 0.63 Petals: (0) dissimilar; (1) subsimilar 1 0 - 1 0 - 1 0 - Lip shape: (0) laminar; (1) bilobed 1 0 - 1 0 - 1 0 - Lip mobility: (0) mobile; (1) sessile 1 0 - 1 0 - 1 0 - Glenion: (0) absent; (1) present 1 0 - 1 0 - 1 0.06 - Appendix: (0) absent; (1) present 1 0 - 1 0 - 1 0 - Column foot: (0) absent; (1) present 0 1 - 0 1 - 0 1 - Stigma: (0) entire; (1) bilobed 1 0 - 1 0 - 1 0 - Anther: (0) ventral; (1) dorsal 1 0 - 1 0 - 1 0 - Pollinarium: (0) caudicles; (1) caudi- cles+viscidium 1 0 - 1 0 - 1 0 - Table 2.2. Marginal probability of the root state as estimated with ACE, SCM and BI. 40 Chapter 2 Stellamaris. The recent discovery of biting midges of the genus Forcipomyia (Ceratopogonidae) as pollinators of two species of Trichosalpinx highlights the importance of the mobile, papillose, ciliate lip for the pollination of this group (Bogarín et al., 2018a). Additional micromorphological observations of the flowers of these three genera, such as the papillose surface of the lip with striated cuticles and secretions of proteins as possible rewards support a hypothesis of floral con- vergence (Bogarín et al., 2018a). The flowers of some species of Anathallis, Tubella and Opil- ionanthe are similar to other pleurothallids, such as the white flowered Specklinia calyptrostele, which is visited by biting midges of the genus Atrichopogon (Ceratopogonidae) (Karremans et al., 2016), suggesting that floral similarities are prone to homoplasy due to the adaptations to similar pollination mechanisms. The predominance of an ancestral morphology adapted to pollination by biting midges makes these characters unsuitable for generic classification. The combination of a sessile lip, absence of a column foot, dorsal anther and pollinarium with caudicles and viscidium is only observed in Lepanthes, Draconanthes, Pseudolepanthes and Lepanthopsis, whereas mobile lips, a column foot, ventral anther and pollinarium with caudicles are observed in all other genera (Luer, 1996b; Luer, 1997). Synapomorphic characters of Lepanthes, such as an appendix, in combination with a viscidium and a sessile lip are key features for a pollination system by sexual deception (Blanco and Barboza, 2005). Even though pollination observations are documented only for a handful of species of this genus, the floral synapomorphies indicate that a pseudocopulation strategy is Characters Plesiomorphy Synapomorphy Homoplasy Habit caespitose - repent Ramicauls non-prolific - prolific Ramicauls’ bracts unornamented foliaceous (Frondaria) ornamented Inflorescence simultaneous - successively flowering Inflorescence length * - shorter/longer than leaves Flower appearance fully opening bud-like (Zootrophion) - Dorsal sepal concavity concave - flattened Synsepal * - absent/present Sepal shape ovate-acute - oblong-acute/ovate-acuminate Petals shape dissimilar subsimilar (Opilionanthe) - Lip shape laminar bilobed (Lepanthes) - Lip mobility mobile - sessile Glenion of the lip absent present (Lepanthopsis) - Appendix of the lip absent present (Lepanthes) - Column foot present - absent Stigma shape entire bilobed (Lepanthopsis) - Anther position ventral - dorsal Pollinarium with caudicles - caudicles+viscidium Table 2.3. Cladistic classification of the 18 morphological characters assessed. Plesiomorphic characters detected with marginal probability at the root state (Table 2) *=ambiguous character at the root state. Synapomorphic and homoplastic characters based on SCM calculations. 41 Generic delimitations in a hypediverse Neotropical orchid lineage Figure 2.5. Ancestral state reconstructions of selected morphological characters from stochastic map- ping analyses based on joint sampling (10,000 mapped trees). Arrows represent transitions between states and numbers represent the estimated number of evolutionary changes with proportion in parenthe- sis and the time spent in each state. Posterior probabilities (pie charts) are mapped in a random stochastic character map. External subdivided ring represents the 14 recognized genera. A. Habit. B. Ramicauls. C. Bracts of ramicauls. D. Inflorescence. E. Inflorescence length. F. Flowers. G. Dorsal sepal concavity. H. Synsepal. I. Sepal shape. cespitose repent A. Habit 6.0 (92.0) 0.5 (8.0) (0.95) (0.05) non−prolific prolific B. Ramicaul growth 4.0 (97.5) 0.1 (2.5) (0.87) (0.13) shorter than leaves longer than leaves E. Inflorescence length 2.2 (21.2) 8.1 (78.8)(0.49) (0.51) simultaneous successive D. Inflorescence 5.1 (81.9) 1.1(18.1) (0.48) (0.52) fully opening bud-like 1.0 (99.3) 0.0 (0.7) (0.95) (0.05) F. Flower appearance absent present H. Synsepal 1.9 (31.3) 4.1 (68.7) (0.28) (0.72) C. Ramicauls’ bracts ornamented foliaceous 1.0 (20.8) 0.04 (0.9)(0.71) (0.01) unornamented (0.28) 2.24 (46.0) 1.5 (30.1) 0.06 (1.2) 0.05 (1.0) G. Dorsal sepal concavity concave flattened 4.2 (72.5) 1.6 (27.5) (0.65) (0.34) I. Sepals shape ovate−acute oblong−acute 6.0 (53.6) 0.26 (2.4) (0.95) (0.05) ovate−acuminate (0.95) 4.0 (35.6) 0.21 (1.8) 0.38 (3.4) 0.36 (3.2) 42 Chapter 2 likely to be predominant in the group. Lepanthes-like flowers are also found in species of the former Lepanthes subgenus Brachycladium Luer, today known to belong to the unrelated genus Andinia (Wilson et al., 2017). The floral convergence is probably a result of pollinator selective pressure as suggested by Wilson et al. (2017) on the basis of pollination observations by Álvarez (2011). In Lepanthopsis, autapomorphic characters such as a glenion and bilobed stigma suggest an adaptation to different, yet unknown pollinators as compared to Lepanthes and Trichosalpinx (Blanco and Barboza, 2005; Bogarín et al., 2018a). Lepanthopsis and Gravendeelia are grouped in the same clade and the need for recognition of Gravendeelia is supported by autapomorphic characters of Lepanthopsis such as the presence of a glenion and bilobed stigma. As transitions of these characters to the ancestral state are unlikely, it seems that floral evolution in Lepanthopsis and Gravendeelia took a different path. Floral morphology of Lepanthopsis resembles that of Platystele Schltr. and the autapomorphic characters such as the presence of a glenion and bilobed stigma suggest an adaptation to different, yet unknown pollinators. In contrast, Gravendeelia has a floral morphology oriented towards a pollination system that likely involves the forward and reverse behavior of insects entering and leaving flowers as in Trichosalpinx s.s (Bogarín et al., 2018a, Chapter 6). Ambiguous results obtained for inflorescence length and the formation of a synsepal at the root state, as well as the higher frequency of transitions between different states indicates that these traits evolved independently in several groups within the Pleurothallidinae (Pridgeon, 2005). The synsepal is made up of fused lateral sepals and this condition can be either absent or intermediate, varying between partial to complete fusion. A possible correlation between sexual mimicry and successive flowering in Lepanthes suggests that all flowers opening at the same time might not be an optimal strategy to fool male fungus gnats (Sciaridae), because the presence of several female-mimicking flowers together may accelerate alerting males from being tricked (Anderson and Johnson, 2006; Anderson et al., 2017). In contrast, the meagre rewards for female biting midges in Trichosalpinx flowers suggest that several flowers opening at the same time might be more advantageous for attracting pollinators (Bogarín et al., 2018a). 2.4.3 Circumscription of the genera in the Lepanthes clade Lepanthes: it has been consistently supported as a monophyletic group by previous studies (Bo- garín et al., 2018c; Pérez-Escobar et al., 2017a; Pridgeon et al., 2001). Species of the genus are known for their caespitose habit with lepanthiform sheaths of the ramicaul. Amongst its close rel- atives, the transversely bilobed petals, the bilobed lip with a basal appendix, the elongated column with apical anther, and the pollinarium with a viscidium are diagnostic for the genus. Several earli- er proposed subgeneric divisions of Lepanthes (Luer, 1996a) were not supported by our molecular phylogenetic analyses and will require re-evaluation when a broader sampling becomes available. Draconanthes: based on the former Lepanthes subgenus Draconanthes (Luer, 1996a), is cur- rently made up of two species known only from high elevations in the Andes. It forms a clade that is sister to Lepanthes in the strict sense. Draconanthes and Lepanthes are morphologically similar but the former may be distinguished by the rigid sepals, linear elongated, unlobed petals and a fleshy lip with a rather rudimentary appendix-like structure in contrast with the elaborate appendixes of Lepanthes. 43 Generic delimitations in a hypediverse Neotropical orchid lineage Figure 2.6. Ancestral state reconstructions of selected morphological characters: A. Petals shape. B. Lip shape. C. Lip mobility. D. Glenion of the lip. E. Appendix of the lip. F. Column foot. G. Stigma shape. H. Anther position. I. Pollinarium. absent present D. Glenion of the lip 1.0 (93.0) 0.0 (7.0) (0.93) (0.07) mobile sessile C. Lip mobility 2.1 (86.2) 0.3 (13.8) (0.66) (0.34) absent present E. Appendix of the lip 1.0 (98.2) 0.0 (1.8) (0.77) (0.23) caudicles caudicles+viscidiumentire bilobed I. PollinariumG. Stigma shape 2.1 (86.1) 0.3 (13.9) 1.1 (93.0) 0.1 (7.0) (0.93) (0.07) (0.66) (0.34) ventral dorsal H. Anther position 2.2 (92.5) 0.2 (7.5) (0.69) (0.31) absent present F. Column foot 0.2 (7.7) 2.2 (92.3) (0.31) (0.69) A. Petals shape B. Lip shape disimilar sub−similar 1.0 (99.5) 0.0 (0.5) (0.99) (0.01) laminar bilobed 1.1 (51.0) 1.0 (49.0) (0.75) (0.24) 44 Chapter 2 Pseudolepanthes: the group is sister to a clade that includes Lepanthes and Draconanthes, rather than being related to Trichosalpinx as previously assumed. Pseudolepanthes resembles species of the latter genus in plant architecture, however, its species are immediately set aside by the spreading, linear to narrowly ovate petals, and the laminar, appendix-free lip with a prominent warty callus, which suggest a different pollination strategy as compared to pseudocopulation recorded in Lepanthes (Luer, 1997). Stellamaris: currently includes a single species, Stellamaris pergrata, previously believed to belong to Trichosalpinx. It is sister to a clade including Lepanthes, Draconanthes and Pseudo- lepanthes instead . With the latter it shares the caespitose, non-prolific habit, but it can be dis- tinguished by a very short, few-flowered inflorescence, long-caudate sepals, a callose lip, an elongate column with an incumbent anther and a prominent column foot, and pollinia lacking a viscidium (Bogarín et al., 2018c; Luer, 1997). Frondaria: it can be distinguished by the synapomorphic conspicuous foliaceous sheaths along the stems. Contrary to the terminal leaf, the smaller leafy bracts do not have an abscission layer which is consistent with them being overgrown, green bracts rather than true leaves. Frondaria produces elongate inflorescences with simultaneously opening, white flowers with spreading, acuminate sepals that are virtually indistinguishable from those of the unrelated genera Anath- allis and Tubella. Lepanthopsis: it forms a clade, together with Gravendeelia, that is sister to Lepanthes, Dracon- anthes, Pseudolepanthes, Stellamaris and Frondaria. Species of the genus are recognized by the inflorescences with simultaneously opening, flattened flowers, provided with a fleshy, simple lip with a glenion at the base and a short column with a bilobed stigma (Luer, 1991). A few excep- tions to this scheme are found in Lepanthopsis subgen. Microlepanthes Luer (Luer, 1991). Gravendeelia: it is a monotypic genus sister to Lepanthopsis. Gravendeelia chamaelepanthes (Rchb.f.) Bogarín & Karremans, the only species currently recognized in the genus, undoubtedly represents a species complex in need of further revision. It is morphologically different from Lepanthopsis by the long-prolific, pendent habit, the few-flowered inflorescences with tubular flowers, with elongated sepals, an elongate lip without a glenion and the elongate column with a distinct foot and unlobed stigma (Bogarín et al., 2018c; Luer, 1997). Both plants and flowers of Gravendeelia are so different from Lepanthopsis that their close phylogenetic clade is one of the most unexpected results of this study. The flowers resemble those of the unrelated genera Anathallis, Stellamaris and Tubella. Opilionanthe: it was formerly placed in Trichosalpinx it is sister to a clade that includes Lepan- thes, Draconanthes, Pseudolepanthes, Stellamaris, Frondaria, Lepanthopsis and Gravendeelia. The lepanthiform bracts, caespitose habit and more or less tubular white flowers are reminiscent of Tubella, thus the isolated phylogenetic placement of this species was unexpected. However, O. manningii (Luer) Karremans & Bogarín is immediately distinguished from species belonging to other genera by the sub-orbicular leaves and the long-caudate petals, which are subsimilar to the sepals (Bogarín et al., 2018c). 45 Generic delimitations in a hypediverse Neotropical orchid lineage Lankesteriana, Pendusalpinx and Trichosalpinx: the three genera are florally similar as they share purplish flowers with a mobile, ciliate lip, attached to a column foot, and an ventral anther and stigma (Bogarín et al., 2018c; Karremans, 2014; Luer, 1997). The vegetative morphology, however, is quite distinct. Species of Lankesteriana can be easily distinguished from Tricho- salpinx and Pendusalpinx by the extremely small habit with ramicauls that lack ornamented lepanthiform bracts that are shorter than the leaves and the successively flowering inflorescences (Karremans, 2014). Trichosalpinx and Pendusalpinx are vegetatively similar to each other, with a large habit with long ramicauls and simultaneously flowered inflorescences. Pendusalpinx differs in having a pendent habit with large, whitish lepanthiform bracts and glaucous leaves (Bogarín et al., 2018c). Based on vegetative morphology alone it is rather unexpected that Lankesteriana and Pendusalpinx are sister to each other. However, these findings are congruent with those of previous studies (Chiron et al., 2012; Karremans, 2014). On the other hand, contrary to what was found by these authors, Lankesteriana and Pendusalpinx are here found to be sister to the genus Trichosalpinx as previously supported by Pérez-Escobar et al. (2017a). Due to the contradictory inferences, the relatively long branches of the Lankesteriana accessions, and the highly diverging morphologies, we remain cautious as to the true phylogenetic relationships between these three genera. It is possible that the similar floral morphology was caused by convergent evolution due to a similar pollination strategy rather than a shared evolutionary history (Bogarín et al., 2018c). Anathallis and Tubella: they are related but with moderate to low support in the BI and ML analyses, therefore, their more detailed relationship remains unresolved. Both taxa received high support as separate genera though. Anathallis is distinguished by the non-lepanthiform sheaths, non-proliferating ramicauls, and the free, star-shaped perianth (Karremans, 2014; Luer, 2006). Some species have purple flowers with mobile lips whereas others share similar micromorpho- logical characters with Lankesteriana, Pendusalpinx and Trichosalpinx s.s. such as the striated cuticles and secretion of proteins (Bogarín et al., 2018a). Members of Pleurothallis subgenus Acuminatia sect. Acuminatia are phylogenetically related to Stelis s.l. and should therefore not be considered as part of Anathallis ( Karremans et al., 2013). Allied to Lepanthes, Lepanthopsis, Trichosalpinx and their allies (named here Clade 8) are members of Tubella, a group traditionally recognized as a subgenus of Trichosalpinx (Luer, 1997; Luer, 1983). It comprises mostly slender plants, with proliferating ramicauls, simultaneously flowering inflorescences with whitish flow- ers and elongated sepals. Zootrophion: it was recovered sister to all other members of the Lepanthes clade. It can be distinguished by the partial opening of the flowers due to the apical fusion on the sepals. As a consequence, the flowers have a single opening on each side, giving them a unique appearance. This feature, present in all species of Zootrophion, is not present in the other members of the Lepanthes clade; however, it is present in other unrelated genera of the Pleurothallidinae. The synsepal is thick and verrucose, the lip is minute. The bracts are large, unornamented and loose. 46 Chapter 2 2.5 Conclusions Generic delimitations based on morphological characters are daunting because of overwhelming homoplasy of the characters traditionally used for circumscriptions. The Lepanthes clade chal- lenged systematists and taxonomists for centuries due to the floral homoplasy untangled here which is possibly resulting from similar pollination systems. We provide evidence for recog- nizing 14 well supported genera as members of the clade based on a combination of molecular phylogenetics and a solid morphological assessment identifying both synapomorphies and ho- moplastic characters. Future research should focus on members of Trichosalpinx subgenus Xenia which are extremely rare but need to be phylogenetically evaluated in order to obtain a com- plete evolutionary scenario for the Lepanthes clade. Based on morphology, we suspect that some members might be related to Lepanthopsis and allies but this hypothesis needs further evaluation. In addition, it is desirable to increase sampling in other groups such as Lepanthopsis (mainly the Antillean species) and Tubella because of floral similarities. Our phylogenetic framework and methodological approach enables the discovery of useful traits for generic classifications, and paves the way for more comprehensive assessments on generic delimitations of similar recal- citrant lineages based on DNA sequences and morphological characters to further improve the systematics of the subtribe. 47 Generic delimitations in a hypediverse Neotropical orchid lineage 48 Chapter 3 Genus-level taxonomical changes in the Lepanthes affinity (Orchidaceae: Pleuro- thallidinae) Diego Bogarín, Adam P. Karremans and Melania Fernández Phytotaxa 340, 128–136. 2018. Abstract. We propose a new classification of the Lepanthes affinity based on phylogenetic re-eval- uation of the Pleurothallidinae. Fourteen genera are recognized as belonging to the affinity. They are found highly supported in a DNA-based phylogenetic inference of combined plastid (matK) and nuclear (nrITS) datasets. The necessary changes, including four novel generic concepts, needed to reorganize the Lepanthes affinity, are proposed here to insure monophyly. The integral discussion on the phylogenetics and biogeography of the group, together with morphological characterization of each clade is presented in Chapter 2. 49 3.1 Introduction With about 5,200 known species, Pleurothallidinae is currently the most species-rich subtribe in the Neotropics, and one of the richest in the Orchidaceae. After the first phylogenetic study of the subtribe by (Pridgeon et al., 2001), and the subsequent proposal to recircumscribe most of its genera (Pridgeon and Chase, 2001), numerous studies aimed to refine or redefine generic concepts in the different clades of Pleurothallidinae have been published. Among the nine great- er clades within the subtribe, the Lepanthes Sw. affinity (Karremans, 2016) is one of the most species-rich, encompassing more than 1,400 species. The currently recognized genera that are members of this clade are Anathallis Barbosa Rodrigues [116], Draconanthes (Luer) Luer [2], Frondaria Luer [1], Lankesteriana Karremans [21], Lepanthes [1122], Lepanthopsis (Cogn.) Ames [45], Trichosalpinx Luer [124] and Zootrophion Luer [26] (Chase et al., 2015; Karremans, 2016). The polyphyletic nature of some of these genera, especially Anathallis and Trichosal- pinx, was suggested by several independent DNA-based phylogenetic analyses and supported by morphological observations (Chiron et al., 2012; Karremans, 2014; Luer, 1997; Luer, 2006; Pérez-Escobar et al., 2017a; Pridgeon and Chase, 2001; Rykaczewski et al., 2017). Nevertheless, no integrate, corrective, classification system was proposed, most likely due to the difficulty of adequately inferring relatedness on the basis of morphology on its own, the availabilit