Phylogeography of the neotropical epiphytic orchid , genus Dracula

The tropical Andes constitute a natural barrier between the Pacifi c Ocean and the Atlantic; in these mountains, are a great variety of Ecosystems, defi ned by factors such as orography, winds, humidity, temperature, among others. Some of these Ecosystems have different environmental conditions from tropical ones. In them, there is a great Biodiversity, in some cases endemic and associated with relatively small geographic areas. An example of this biodiversity is the orchids of the genus Dracula, about which discussions are currently generated due to the diffi culty in classifying their members. The present work shows a study where DNA was isolated and sequenced from plant samples obtained from 52 species of orchids of the genus Dracula, which were analyzed using the MEGA7 software. Phylogenetic analysis of the DNA sequences showed a well-resolved topology that refl ects a geographical pattern of several major clades of the Pacifi c and Atlantic watersheds. Geophysical conditions of the Andes have generated greater biodiversity of the genus Dracula on the side of the Pacifi c. Although the species Dracula cordobae and alessandroi reported on both sides of the study site belong to the same clade and show limited mobility through the drier area to the South of the mountain range.


INTRODUCTION
The tropical Andes top the list of worldwide hotspots about numbers endemics and endemic species/area ratio for plants and vertebrates (Myers et al. 2000). This is the result of a variety of processes, mainly geological and climatic. The Andes mountain range divides South America from North to South into two large blocks: to the West the Pacifi c Ocean watershed and the East the Atlantic Ocean watershed. In Ecuador, the Andes constitute an impressive meridian mountain barrier whose width generally varies between 100 and 150 km from East to West, the central part being the narrowest; while, in the South near the Peruvian border, the Andean mountainous reliefs prolonged by the Amazonian sub-Andean mountain ranges are characterized by lower annual average rainfall (Maldonado & Solano 2013), a general drop in altitudes to constitute a mountain range, certainly less marked but measuring 180 to 200 km wide (Maldonado & Solano 2013, Winckell 1997. The vegetation of the Amazon is much older than that of central Ecuador. The fl ora of the mountains was formed after the geological survey of the Andes (Patzelt 1996). Besides, climatic fluctuations during the Pleistocene influenced the distribution of plants (Prance 1982). The territory does not have a completely tropical climate, presenting a wide variety depending on the orographic or topographic location. The snowy peaks, steep slopes, deep canyons and isolated valleys of these mountains have resulted in a great diversity of microhabitats that favour speciation. Also, its location between the lowlands of the Amazon, the Chiquitanía and the Gran Chaco to the East and the Chocó, Tumbes-Guayaquil, and the arid systems of the Sechura desert to the West, determine complex dynamics of species exchange and isolation (Herzog & Jørgensen 2011).
Ecuadorian flora exceeds twenty thousand species distributed in a great diversity of habitats, more than two thousand are arboreal and about four thousand are orchids (FAO & MINISTERIO DEL AMBIENTE 2012, Patzelt 1996. In the Guayas river basin, with an approximate area of 34500 km 2 , 12.57% of the territory of Ecuador presents, for instance, more than two hundred species of orchids that are not found in other parts of the country (Patzelt 1996). Orchids are one of the largest, most diverse and widely distributed plant families and can be found in all habitats except the polar regions (Cox 2013, Cribb et al. 2003. A high proportion of them are epiphytes, forming part of the canopy in humid forests where they play a relevant role in the processes of storage and circulation of water (Castellanos Castro & Torres Morales 2018, Pypker et al. 2006. Therefore, they are related to a wide variety of organisms, such as fungi, insects, birds, and even mammals, participating, for example, in the cycling of nutrients (Castellanos Castro & Torres Morales 2018). Also, they support specialized pollinating organisms as well as the abundance and diversity of flowers that generate pollinator biodiversity (Lázaro et al. 2020, Storck-Tonon & Peres 2017.
The Orchidaceae form one of the largest families of angiosperms, as well as one of the most fascinating by reason of their diversity and specialization in floral. While many workers have described new genera and species of orchids, there has been very little monographic work, and we may safely say that the family has really been very little studied, considering its size and complexity (Dressler & Calaway 2009). A characteristic of orchids is their production of up to six million seeds/fruits in powder form that are dispersed in the wind over considerable distances (Trapnell et al. 2019), is surprisingly poorly known in a fossil state. The shortage of orchid fossil data is discussed mainly in the context of limitations in evolutionary and phylogenetic studies of this group of plants The first species of Dracula to be described was Masdevallia chimaera by Professor Reichenbach in 1872. Following his example, every species of Dracula continued to be described in Masdevallia until 1978. Luer (1993 included in Dracula the species with carinated leaves, internally pubescent sepals, widened and bilaminated or bivalved petals apically; and, without the tooth that Masdevallia has at the label margin or the base, a generally sacciform labellum with radial veins and divided into a hypochile and epichile. Vegetative differences allow the Dracula genus to be easily distinguished from other genera of Pleurothallidinae, but generally, it does not allow the differentiation between Dracula species when plants are not blooming (Meyer & Cameron 2009). Some complexes of species or variable species, that cause much confusion and debate, are minimally distinguishable from other species due to the difficulty to differentiate them from other species of the same genus (group). Natural variations can be found within a species complex (Karremans 2016, Solano-Gómez et al. 2008. It is even suspected that some taxa collected in nature are natural hybrids, such as D. hawleyi and D. radiosyndactyla (Luer 1993). Currently, 120 species of the Dracula genus are recognized for the neotropics from Central America to South America; 55 species in the genus have been described in the territory of the Republic of Ecuador and recently several new species have been described (Baquero & Meyer 2014, Luer 1993, Peláez et al. 2009). They mainly grow in cloud forests, where they can be found at an altitude between 300 and 2800 meters (Baquero & Moncayo 2017, Luer 1993.
The advent of molecular techniques has dramatically advanced our understanding of the phylogenetic relationships in family Orchidaceae. The Internal Transcribed Spacer (ITS) region of nrDNA possesses moderate interspecific variation and has been the primary source of characters for phylogenetic analysis at lower taxonomic levels (Baldwin et al. 1995, Hu et al. 2016. Several papers on nucleotide sequences in the Orchidaceae family of the ITS nuclear genome region in molecular phylogenetics are presented in the literature (Batista et al. 2013, Whitten et al. 2012 Phylogeography examines the correspondence between genetic characteristics and geographic distribution of different species (Avise et al. 1987). The genetic structure of the population is as much a product of history as of current migration patterns and isolation of populations (Avise et al. 1987, Bermingham & Mortiz 1998. In recent decades molecular studies have established classical biogeographic patterns, associated with past geological and bioclimatic changes that led to genetic divergences in many groups of species. (Schmitt 2007). The species diversification process is often associated with niche changes in the newly emerged lineages. However, an opposing force known as niche conservatism means that related species tend to resemble each other in their niche requirements (Šmíd et al. 2019). One of those possible components is the species' climatic niche (Šmíd et al. 2019). Phylogenetic studies used cp DNA sequences as matK from the ITS region to test models of historical biogeography (Fritsch & Cruz 2012), showing a well-resolved topology that reflects a geographical pattern of several clades (Holderegger & Abbott 2003, Loera et al. 2012, Romaschenko et al. 2014. Microclimatic variability in Tropical Andes plays a key role in shaping species distributions and their ability to cope with environmental change (Ayala-Izurieta et al. 2017, Montejo-Kovacevich et al. 2020). The present work shows a study where DNA was isolated and sequenced from plant samples obtained from 52 species of orchids of the genus Dracula, which were analyzed using the MEGA7 software. Phylogenetic analysis of the DNA sequences obtained show a geographical pattern of the main clades of the Pacific and Atlantic watersheds. As greater biodiversity on the Pacific Ocean watershed generated by the geological uplift of the Andes is seen, the trade winds, which in the Andean range of Ecuador always go from east to west, as well as the effect of low temperatures, gave rise to the development of different intraspecific lineages in the genus Dracula. In the south of the country, the mountain range presents lower rainfall and heights than the study area; the species D. cordobae and D. alssandroi have been reported on both the Pacific and Atlantic watersheds, suggesting the presence of a genetic exchange corridor for orchids of the Dracula genus. To fulfill this task, the ITS1 region of the rDNA was analyzed from 52 samples of species of the genus Dracula out of 55 described in Ecuador. For comparison, some taxa of the genus Masdevallia were included.

Plants
The Dracula genus consists of 120 species including those recently described in Central and South America (Baquero & Meyer 2014, Peláez et al. 2009), they grow preferably in cloud forests where they can be found at an altitude between 300 and 2800 meters (Baquero & Moncayo 2017, Luer 1993). For the present study, samples of plant tissue were collected from private and public collections located in the province of Pichincha: "Orquideario de Sarina", "Jardín Botánico de Quito", " Alberto Guerrero's Collection", and in the province of Carchi canton Huaca: "Digna Chacón's Collection" and "Guadalupe Utreras' Collection", representing a total of 104 samples of leaf tissue from 52 of the 55 species of the genus Dracula described in Ecuador (Baquero & Meyer 2014, Peláez et al. 2009) which corresponds to 44% of the total number of known species of the genus Dracula in the Neotropics (Valencia & Pitman 2000). The fresh leaf samples of each of the specimens were transported in kraft paper envelopes and in hermetically sealed plastic bags with silica gel. In the laboratory they were kept at a temperature of -20 C. The samples were identified by Luis Baquero, curator of the orchid collection of the Botanical Garden of Quito. The living samples are kept in the Botanical Garden "Orquídeas de Sarina", with the collection code from Marco Cerna's field book, Project "Identification molecular of the genus Dracula Orchidaceae 2018".

Geographical location of species and wind
The geographical location of the species under study was obtained from the Red Book of Endemic Plants of Ecuador, Dracula-species and Tropics (Figure 1). Ecuador's wind system is dominated by tropical trade winds with strong easterly winds throughout the year (Beck et al. 2008), effect that has the greatest intensity on the peaks of the Andes. However, local climatic conditions may differ due to the microclimatic influence caused by the topography and the presence of the Pacific Ocean to the west (Jantz et al. 2013).
Of the fifty-five species of Dracula reported for Ecuador (Baquero & Moncayo 2017), only the species D. Felix, D. Trinympharum and D. Soennemarkii were not included in this study, fifty-two species were evaluated, of which fortyone species have been reported for the Pacific Ocean watershed, twelve species have been found in the Amazon River basin, two species have been identified on both watersheds at the south of the study area, and one species have not yet been reported in the wild in the study area, see Table I. As a functional group, species of the genera Masdevallia yungasensis, Masdevallia picturata, and Diodonopsis erinacea were used, whose sequences were downloaded from GENBANK.

DNA extraction
Genomic DNA from leaves samples was extracted using the chemical protocol described by Doyle & Doyle (1987). A sample of approximately 30 mg was sprayed in liquid nitrogen and placed in a eppendorf tube, to which 600 µL of lysis buffer was added, 94 µL SDS at 10%, 118 µL Potassium Acetate (5M). The mixture was stirred and then the supernatant was taken and added to a new tube. Next, 600 µL of a phenolchloroform mixture (1: 1 v / v) was added, stirred and centrifuged to induce the formation of two phases, 250 µL of the upper phase were taken and transferred to a new one microtube, adding 360 µL of cold isopropanol, mixing by inversion. The obtained mixture was stored at -20 °C overnight to obtain the precipitated DNA. After this time, 3 washes were performed, adding 400 µL of 70% ethanol and centrifuging each time a wash was performed. The DNA pellet obtained were resuspended in 30 µL of TE Buffer and stored at -20 °C until use.

Internal transcribed spacer region amplification
Using the Polymerase Chain Reaction technique (PCR) in a conventional thermal cycler type MultiGene™ OptiMax, the amplification of the ITS region was performed using the ITS-1 primer (F-TCCGTAGGTGAACCTGCGG). The PCR products were visualized by horizontal electrophoresis in 1% agarose gels at 90 V for 40 minutes and stained with SYBR Safe, to be visualized under UV light using a MicroBis DNR Bio-Imaging Systems.
The PCR amplified product was sequenced (forward) using the Sanger method (Bernardos et al. 2006  It is observed that each species is located on one side of the mountain range (Pacific slope and Atlantic slope) between 300 and 3000 meters above sea level, with the exception of Dracula alessandroi and Dracula cordobae species highlighted with yellow circles, which they are found on both sides of the mountain range (The numbers identify the species according to the cladogram, the white line represents the highest part of the mountain range).

Data analysis
The sequences were individually verified using the free software Finch Tv DNA Sequencing for Windows Geospiza (2012). The GenBank BLAST tool was used to make a comparison of the amplified regions, determining that 24 species of those analyzed are already in the database in the ITS region. Nucleotide sequences were aligned using the MUSCLE software implemented in the Molecular Evolutionary Genetics Analysis (MEGA) version 7.0 program (Tamura et al. 2013). MEGA is computer software originally created at Pennsylvania State University, its latest version of MEGA7 software updated in January 2016 produces relative divergence times for nodes, which are useful in determining the order and spacing of divergence events in species and family trees (Kumar et al. 2016). The evolutionary history was inferred using the Maximum Probability method based on the Tamura-Nei model. The model is based on rates of two types of transition substitutions between purines and pyrimidines and the rate of crosssubstitutions are considered separately, taking into account the unequal frequencies of four nucleotides (Tao et al. 2020). As a result, the tree with the highest recording probability (-19876.71) was obtained (Figure 2). The initial tree for the heuristic search was obtained automatically by applying the Neighbor-joining and BioNJ algorithms to a distance matrix in nucleotide pairs estimated using the Maximum Composite Likelihood (MCL), and then selecting the topology with a higher recording probability value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 52 nucleotide sequences. All positions containing gaps and missing data were eliminated. The evolutionary analyses were performed in MEGA7.

RESULTS AND DISCUSSION
The present work shows a study where DNA was isolated and sequenced from plant samples obtained from 52 species of orchids of the genus Dracula. The species evaluated in this study that have sequences in the NCBI GenBank are 24, and the others 28 species are not registered yet (Table II). Phylogenetic analyses of the genus Dracula showed, in general, a well-resolved topology reflecting a geographical pattern of several major clades of the Pacific and Atlantic watersheds. When analyzing the tree (Figure 2), with the location of the reports of wild species of the genus Dracula in the territory of the Republic of Ecuador. It is found that in clade 92 ( Figure  1), the sixteen species have been reported to the Pacific slope, fifteen of the eighteen species have been reported to the North, only D. pinasensis has been found in the southern part of the Pacific watershed and the species D. chistineana distribution is unknown in wild form in the study area. Clade 76 the five species have been reported in the south, D. fafnir, D. deltoridea, D. lotax and D. mendozae, have been reported in the wild on the Atlantic watershed. Nevertheless, the species D. mopsus has been reported along the Pacific watershed ( Figure 1).
In clade 88, fourteen species have been reported in the northern part of the Pacific slope. In the clade 70, three of the five species have been reported in the southern part Atlantic watersheds where the height of the mountain range decreases showing the presence of a natural corridor in the south of the Andean mountain range in the study area between the Pacific and Atlantic watersheds ( Figure 1). The trade winds from the East throughout the year stimulate mobility from East to West. However, the influence of the trade winds decreases with the height of the mountain range. Therefore, it seems possible the diffusion of pollen and seeds from the West to the East through this biological corridor.
The Andes and the trade winds surround the hotspot to the Northwest of Ecuador limiting the possibility of the diffusion of the genus Dracula species towards the Amazon basin. The only two species reported in the wild on both the Pacific and Atlantic watersheds are D. cordobae and D. dalessandroi from the same main clade 90 ( Figure 1). The species in the study area, which have been reported only to the north of the Pacific watershed, do not share the main clade in any case with species reported to the north of the Atlantic watershed where the Andean mountain range maintains high altitudes.

CONCLUSIONS
The DNA samples obtained of orchids of the genus Dracula show that about 54 % of species analyzed not registered in the NCBI GenBank. Phylogenetic analysis of the DNA sequences obtained showed a well-resolved topology that reflects a geographical pattern of several major clades of the Pacific and Atlantic watersheds. Geophysical conditions of the Andes have generated greater biodiversity of the genus Dracula on the side of the Pacific. Although the species Dracula cordobae and alessandroi reported on both sides of the study area belong to the same clade and show limited mobility through the drier site to the South of the mountain range. Microclimatic variability in the Tropical Andes, the trade winds, among other phenomena plays a crucial role in shaping species distributions created one of the most important hotspots in the world. Although, at the same time, they limit the mobility of species of the Dracula genus to the extensive Amazon River basin.