ECOLE NATIONALE SUPÉRIEURE AGRONOMIQUE DE MONTPELLIER MONTPELLIER SUPAGRO THÈSE Pour l’obtention du grade de DOCTEUR EN SCIENCES AGRONOMIQUES Ecole Doctorale: Biologie des Systèmes Intégrés, Agronomie – Environnement. Discipline: Ressources Phytogénétiques et Interactions Biologiques. présentée et soutenue publiquement Par John OCAMPO PÉREZ le 16 avril 2007 Titre : Étude la diversité génétique du genre Passiflora L. (Passifloraceae) et de sa distribution en Colombie Study of the genetic diversity of genus Passiflora L. (Passifloraceae) and its distribution in Colombia JURY Dr. Rosemary GILLESPIE, Professeur, Université de Californie, Berkeley, U.S.A. Dr. André CHARRIER, Professeur, Montpellier SupAgro, France Dr. Laure CIVEYREL, Maître de conférence, U. Paul Sabatier, Toulouse, France Dr. Daniel DEBOUCK, Directeur Unité Ressources Génétiques, CIAT, Colombie Dr. Geo COPPENS d’EECKENBRUGGE, CIRAD, Montpellier, France Dr. Philippe FELDMANN, CIRAD, Montpellier, France Examinateur Examinateur Rapporteur Rapporteur Co-Directeur Directeur DEDICATED TO MY FAMILY ACKNOWLEDGEMENTS This study was supported by grants from the Gines-Mera Fellowship Foundation (CBN-CIAT), Bioversity International (formerly IPGRI), the Centre de Cooperation Internationale en Recherche Agronomique pour le Développement. (CIRAD), the International Center for Tropical Agriculture (CIAT), the Colombian Ministry for Environment (MMA), the Research Center of the National Federation of Coffee Growers of Colombia (Cenicafé) and the Instituto Colombiano para el Desarrollo de la Investigación (COLCIENCIAS). I would especially like to express my gratitude to Geo Coppens d’Eeckenbrugge, who has played a critical role in this project from its inception. He introduced me to the Passiflora world, and has been a wonderful advisor throughout the course of this study. His door was always open and he constantly provided me with much needed support and encouragement. Also, Geo and his family provided me with a place to stay and support during visits in Montpellier. I am particularly grateful to Drs. Philippe Feldmann (CIRAD), André Charrier (ENSAM), Ange-Marie Risterucci (CIRAD), Laurence Pascal and Bertrand Schatz (CNRS - CEFE), members of my committee for their scientific advices, comments and corrections of the manuscript. Special thanks to Drs Laure Civeyrel (Université Paul Sabatier), Daniel Debouck (CIAT) and Rosemery Gillespie (Califonia of University) for having accepted to be part of the jury. I am deeply indebted also to Drs. Andrew Jarvis (CIAT) and Xavier Scheldeman (Bioversity International) for their comments and valuable suggestions regarding geographic data analysis. I want to thank Daniel Franco and Mario Ruiz for helping in the installation of living collections (Paraguaicito Experimental Station - Cenicafé), and to German Arroyave and Juan G. Contreras (PASSICOL S.A) for providing the living collection facilities for maracuja. Also, I wish to acknowledge Christian Houel (Passiflora National Collection, Blois, France) and Doyle McKey (CNRS - CEFE) for assistance in obtaining plant material for DNA extraction. I am deeply indebted also to María Restrepo, Felipe Barrera, Lina Farfán and Cristian Olaya of the Caldas University (Colombia) for assistance in collecting field data. Thanks are also due to Mauricio Villegas (Cenicafé), Vicky Barney (Bioversity), Odelio Soto (S.Agricultura, Tenerife), Robinsón Galindo (PN.Catatumbo), Carlos Solarte (Cenicafé), Carolina Alcázar (Proselva), Hernando Criollo (U.Nariño), María Giraldo (Umata, Filandia), José F. Restrepo (U. Caldas), Edgar Díaz (JB. San Jorge, Ibagué), Sergio Ocampo (Aguas de Manizales), Camilo Palacios (PN. Guanenta), Hector Jiménez (Umata, Fredonia), Cesar Londoño (Umata, Marsella), Rafael Izquierdo (Bioversity), Creuci Caetano (U.Nacional), Mike Salazar (CIAT), Alvinxon Castro (U.Chocó), Alvaro Mejia (CIAT) and Gustavo Morales (JB. José Celestino Mutis) for accompanying me on several collecting trips. José O. Velásquez and Segundo Pablo Guaspud (Casa Mutis) was my assistant and traveling companion during my field trip to Mariquita (Tolima) and is a very dear friend. Their ecological knowledge and experience in traveling around Mariquita made for a flawless trip. I am enormously grateful also to Colombian farmers contacted in the fieldwork for their continuous help and availability in localizing a great part of the observed plant material. I would like to thank the following Colombian and international herbaria for providing me access to the specimens: Álvaro Fernández Pérez, Fundación Universitaria de Popayán (AFP); Universidad del Cauca (CUP); Universidad Tecnológica del Chocó (CHOCO); Centro Internacional de Agricultura Tropical (CIAT); Instituto Amazónico de Investigaciones Científicas (SINCHI); Universidad Nacional sede Bogotá (COL); Universidad del Valle (CUVC); Departamento de Recursos Naturales, Universidad de Caldas (FAUC); Federico Medem, Instituto Alexander von Humboldt (FMD); Jardín Botánico Eloy Valenzuela (HEV); Universidad de Antioquia (HUA), Universidad del Quindío (HUQ); Jardín Botánico José Antonio Uribe (JAUM); Gabriel Gutiérrez (MEDEL); Universidad de Nariño (PSO); Universidad Surcolombiana (SURCO); Universidad del Tolima (TOLI); Universidad Industrial de Santander (UIS); José Cuatrecasas Arumi, Universidad Nacional sede Palmira (VALLE); Muséum National d'Histoire Naturelle, Paris, France (P); the Real Jardín Botánico, Madrid, Spain (MA); Royal Botanic Garden, Kew, England, UK (K); Missouri Botanical Garden, USA (MO); The Natural History Museum, Chicago (F); New York Botanical Garden, USA (NY); Marie Selby Botanicals Gardens, USA (SEL). The three latter were visited on line. In the office of Bioversity International (for the Americas) several people have supported me in other aspects through this thesis in the last four years, particularly Ana Luisa Triana, for her collaboration and friendship. Dimary Libreros, Adriana Sánchez, Monica Macias, Angela Cardona, Elcy Lozano, Samir Patiño, Jesus Salcedo, Margarita Baena and Emelda Usma have also shared theirs friendship and congeniality, and helped in the logistics of my work. I want also to express my gratitude to former director, Dr. Ramon Lastra, and the current director, Dr. Marleni Ramírez, who have deposited in me their confidence and support. I am especially grateful to Edgardo Alpizar (Costa Rica), Rommel Montufar (Ecuador), Fabio Parrado and Genny (Bolivia), Gaston Loor (Ecuador), Mauricio Soto (Colombia), and to the Venezolians Elvis Portillo, Climaco Alvarez, José Bustamante, Douglas Rodríguez and Graciela Sepulcre for sharing immemorial moments during my stay in Montpellier. In CIAT, I also wish to express my sincerest thanks to Teresa Sánchez and Rosa González, who have offered me their friendship and advices. Finally, my thanks are due to my family for their love and support. I profoundly thank my wife, Carolina who has been a constant source of love, patience, and encouragement throughout the last four years and to baby who comes in way. My parents have continually believed in me and helped me in innumerable ways in special my mother Rocío. SUMMARY Given its economic importance, the characterization of genus Passiflora is seen as a priority for Andean countries and specific strategies are needed to optimize its use and conservation. The objective of this thesis was to study the distribution and organization of Passifloraceae in Colombia, with a triple aim of exploring the diversity of available genetic resources, evaluating the risk of their erosion and their potential as an indicator group for the conservation of biodiversity, particularly in the coffee growing zone. Colombian Passifloraceae were listed, gathering and georeferencing 3,930 records, for a total of 167 species. Forty-two produce an edible fruit, and nine are commercially cultivated. Most of the 58 endemic species, including 37 narrow endemics, are Andean species of subgenera Tacsonia and Decaloba. Applying the UICN criteria, 70% of the species appear threatened and three extinct. When compared with other regions, the Andes of Colombia and Ecuador constitute the center of Passiflora diversity, whose elevational distribution shows a small peak below 500 m, and two higher ones at 1000-2,000 and 2,500-3,000 m. This pattern corresponds to divergent adaptive trends among infrageneric divisions, subgenus Tacsonia contributing markedly to the highest peak. The climatic data associated with our 3,930 records allowed modeling and summing the distribution of 80 species, so predicting the distribution of species richness. Nine areas with very high richness, but no particular endemism, were identified in the Andean region, three of which correspond to collection gaps as they do not appear on the map of observed diversity. Their striking correspondence with coffee growing zone ecotopes imposes a strategy integrating agricultural and environmental management at the landscape level for preserving this threatened richness as well as a region of particular importance for the country. Both aspects may be combined if Passifloraceae can be used as an indicator of biodiversity in this region, which seems justified by their diversity and characteristics, including multiple ecological interactions with many organisms. Morphological variation was studied in 124 accessions from 60 Passiflora species and eight subgenera with 127 descriptors. Twenty-four quantitative traits showing i high variation among subgenera were selected. The three principal components of variation are associated with (i) flower length; (ii) flower width and bract shape; and (iii) peduncle branching, stem width and leaf length. The projection of accessions in the resulting tridimensional space consistently separates subgenera. A similar selection of 32 qualitative traits, and four categorized quantitative traits, allowed classifying our sample consistently. Most discriminating characters include size of stems and leaves, presence of tendrils, number and distribution of extrafloral nectaries, dimensions and general shape of bracts, width and length of flowers and corona complexity. Our results support seven of the eight Killip’s subgenera of our sample, but no infrasubgeneric classifications. However, the new classification of subgenus Decaloba by Feuillet & MacDougal was partly supported. They converge on many points with previous phylogenetic results obtained with DNA sequences, although the latter group subgenera Tacsonia and Distephana with subgenus Passiflora. The chloroplast and mitochondrial DNA diversity of 213 genotypes belonging to 151 Passiflora species and 15 subgenera recognized by Killip was studied by PCRRFLP, identifying 280 haplotypes for cpDNA and 372 for mtDNA. The principal coordinate analysis on cpDNA data allowed visualizing a strong separation of subgenera Apodogyne, Decaloba, Murucuja, Pseudomurucuja and Psilanthus (constituting the “Decaloba group”), while the neighbor-joining cluster analysis showed three wellsupported clusters within Passiflora, corresponding to the three major divisions of the taxonomy proposed by Feuillet & MacDougal. The first one, named the “Passiflora group”, includes subgenera Calopathanthus, Deidamioides, Distephana, Dysosmia, Dysosmioides, Manicata, Passiflora, Tacsonia, and Tacsonioides, with a very loose substructure and considerable intraspecific variation. The second one includes subgenus Astrophea, and the third is the ‘Decaloba group’. The outgroup species, take an undefined position among the Passiflora clusters. The phenogram from mtDNA data separates four moderately supported clusters. As for cpDNA, a first one corresponds to the ‘Decaloba group’. The other are different, as subgenera Astrophea and Tryphostemmatoides appear integrated within the ‘Passiflora group’, while subgenus Tacsonia forms a uniform distinct cluster, close to another one comprising species of Passiflora series Kermesinae, Simplicifoliae, Lobatae, and Menispermifoliae. The ii analyses of cpDNA and mtDNA give different pictures of the Passiflora diversity, in the position of the outgroup, the relative position of four subgenera, and the relationships between species, which we attribute to different rates of evolution and modes of transmission of the chloroplastic and mitochondrial genomes, combined with reticulate evolution in the genus. Keywords: Passiflora L., Colombia, Andes, coffee growing zone, distribution, biodiversity, endemism, morphological mitochondria. iii descriptors, PCR-RFLP, chloroplast, RÉSUMÉ De par son importance économique, la caractérisation du genre Passiflora est considérée prioritaire par les pays andins, impliquant des stratégies spécifiques pour optimiser son utilisation et sa conservation. L’objectif de cette thèse est d’étudier la distribution et l'organisation des Passifloraceae en Colombie, dans le triple but d'explorer la diversité des ressources génétiques disponibles, d’évaluer le risque de leur érosion ainsi que leur potentiel comme groupe indicateur pour la conservation de la biodiversité, en particulier dans la région caféière. La liste des espèces colombiennes a été établie en rassemblant et localisant 3.930 registres, pour un total de 167 espèces, dont 42 produisent un fruit comestible et neuf sont cultivées commercialement. La plupart des 58 espèces endémiques, y compris 37 endémiques étroites, sont des espèces andines des sous-genres Tacsonia et Decaloba. En appliquant les critères de l'UICN, 70% des espèces semblent menacées et trois éteintes. Par comparaison avec d'autres régions, les Andes de la Colombie et de l'Equateur constituent le centre de diversité de Passiflora, dont la distribution altitudinale montre un petit pic en-deçà de 500 m, et deux plus élevés, vers 1.000-2.000 et 2.5003.000 m. Ce patron reflète des tendances adaptatives divergentes entre divisions infragénériques, le sous-genre Tacsonia contribuant nettement au dernier pic. Les données climatiques associées à nos 3.930 registres ont permis de modéliser la distribution de 80 espèces et de prédire la distribution de la richesse d'espèces. Neuf zones de richesse très élevée, mais sans niveau d’endémisme particulier, ont été identifiées dans la région andine. Trois d’entre elles n’ont pas été prospectées. La correspondance entre ces neuf zones et des ecotopes caféiers impose une stratégie intégrant la gestion agricole et environnementale au niveau du paysage pour préserver en même temps cette richesse biologique et une région d'importance particulière pour le pays. Les deux aspects peuvent être combinés si les Passifloraceae peuvent y être employées comme indicateurs de biodiversité, ce qui semble justifié par leur diversité et leurs caractéristiques, notamment leurs interactions écologiques avec de nombreux organismes. iv La variation morphologique a été étudiée parmi 124 accessions de 60 espèces et huit sous-genres de Passiflora, avec 127 descripteurs. Vingt-quatre traits quantitatifs variant entre sous-genres ont été retenus. Les trois composants principaux de leur variation sont associés à (i) la longueur de fleur ; (ii) la largeur de fleur et la forme des bractées ; et (iii) la bifurcation du pédoncule, la largeur de tige et la longueur de feuille. La projection des accessions dans l'espace tridimensionnel correspondant sépare bien les sous-genres. Un tri semblable de 32 traits qualitatifs, et de quatre traits quantitatifs catégorisés, a permis de classer notre échantillon de façon cohérente. Les caractères distinctifs incluent la taille des tiges et des feuilles, la présence de vrilles, le nombre et la distribution des nectaires extrafloraux, les dimensions et la forme générale des bractées, la largeur et la longueur des fleurs et la complexité de la couronne. Nos résultats appuient sept des huit sous-genres de Killip inclus dans l’échantillon, mais aucune classification infrasubgénérique, sauf, pour partie, la nouvelle classification du sous-genre Decaloba par Feuillet & MacDougal. Ils convergent sur de nombreux points avec des études phylogénétiques à partir de séquences d'ADN, bien que celles-ci unissent les sous-genres Tacsonia et Distephana au sous-genre Passiflora. L’analyse, par PCR-RFLP, de l'ADN chloroplastique et mitochondrial de 213 accessions de 151 espèces de Passiflora et 15 sous-genres a permis d’identifier 280 haplotypes pour le premier et 372 pour le second. L’analyse factorielle des données d’ADNcp sépare fortement les sous-genres Apodogyne, Decaloba, Murucuja, Pseudomurucuja et Psilanthus (constituant le "groupe Decaloba"), tandis que la classification par la méthode du neighbor-joining met en évidence trois groupes majeurs dans Passiflora, correspondant aux trois divisions principales de la taxonomie proposée par Feuillet & MacDougal. Le premier, appelé "groupe Passiflora", inclut les sous-genres Calopathanthus, Deidamioides, Distephana, Dysosmia, Dysosmioides, Manicata, Passiflora, Tacsonia et Tacsonioides, avec une sous-structure très lâche et une variation intraspécifique considérable. Le second inclut le sous-genre Astrophea, et le troisième est le "groupe Decaloba". Les espèces du groupe extérieur prennent une position indéfinie entre les groupes de Passiflora. Le phénogramme résultant de l’analyse de l’ADNmt sépare quatre groupes. Comme pour l’ADNcp, un premier correspond au "groupe v Decaloba". Les autres sont différents, car les sous-genres Astrophea et Tryphostemmatoides sont intégrés avec le ‘groupe Passiflora’, tandis que le sous-genre Tacsonia forme un groupe différencié et homogène, proche d’un groupe plus varié d’espèces des séries Kermesinae, Simplicifoliae, Lobatae et Menispermifoliae du sousgenre Passiflora. Les analyses des deux génomes, chloroplastique et mitochondrial, donnent donc des images divergentes de la diversité de Passiflora, par la position des groupes extérieurs, la position relative de quatre sous-genres, et les relations entre espèces. Ces divergences semblent liées à des différences dans les taux d'évolution et dans leur mode de transmission, en relation avec une évolution réticulée dans le genre. Mots-clés: Passiflora L., Colombie, Andes, région caféière, distribution, biodiversité, endémisme, descripteurs morphologiques, PCR-RFLP, chloroplastes, mitochondries. vi RESUMEN Teniendo en cuenta su importancia económica, la caracterización del género Passiflora es considerada una prioridad por los países andinos y se necesitan estrategias específicas para optimizar su uso y conservación. El objetivo de esta tesis fue estudiar la distribución y organización de las Passifloraceae de Colombia, con un triple objetivo de analizar la diversidad de los recursos genéticos disponibles, evaluar el riesgo de su erosión así como su potencial como grupo indicador para la protección de la biodiversidad, particularmente en la zona cafetera. Las especies colombianas fueron inventoriadas con base en 3.930 registros, para un total de 167 especies, de las cuales 42 producen fruto comestible y nueve son cultivadas comercialmente. La mayoría de las 58 especies endémicas, incluyendo 37 de distribución restringida, son principalmente andinas, de los subgéneros Tacsonia y Decaloba. Aplicando los criterios de la UICN, 70 % de las especies aparecen en peligro y tres son consideradas extintas. En comparación con otras regiones, los Andes de Colombia y Ecuador constituyen el centro de diversidad de Passiflora, cuya distribución altitudinal muestra un leve pico debajo de 500 m, y dos más marcados en los 1.0002.000 y 2.500-3.000 m. Este patrón refleja tendencias adaptativas según divisiones infragenéricas, con particular contribución del subgénero Tacsonia al tercer pico. Los datos climáticos relacionados con los 3.930 registros permitieron modelizar y sumar la distribución de 80 especies, y predecir la distribución de su diversidad. Nueve zonas ultra-diversas fueron identificadas en la región andina, tres de ellas subexploradas. Estas áreas no muestran un nivel de endemismo particular. La correspondencia entre ellas y los ecotopos cafeteros impone una estrategia de conservación que integre la gestión agrícola y el medio ambiente a escala del paisaje para preservar al mismo tiempo esta riqueza biológica y una región de especial importancia para el país. Ambos aspectos pueden ser combinados si se emplean las Passifloraceae como un grupo indicador de biodiversidad en esta región. Esto es justificado por su diversidad y características particulares, incluyendo sus múltiples interacciones con diferentes organismos. vii La variación morfológica fue estudiada en 124 accesiones de 60 especies de Passiflora, y ocho subgéneros de Killip, usando 127 descriptores. Por su diferenciación entre subgéneros, 24 descriptores cuantitativos fueron seleccionados para un análisis de componentes principales. Los tres primeros componentes están relacionados con (i) longitud de la flor; (ii) ancho de la flor y forma de la bráctea; (iii) bifurcación del pedúnculo, diámetro del tallo y longitud de la hoja. La proyección de las accesiones en un espacio tridimensional muestra una clara separación entre subgéneros. Una selección de 32 caracteres cualitativos y cuatro cuantitativos categorizados ha permitido clasificar la muestra de manera coherente. Los caracteres discriminantes incluyen tamaño del tallo y la hoja, presencia de zarcillos, número y distribución de los nectarios extraflorales, dimensiones y forma general de la bráctea, ancho y longitud de flor y complejidad de la corona. Los resultados soportan siete de los ocho subgéneros de Killip representados en la muestra, pero ninguna clasificación infra-subgenérica, con la excepción parcial del subgénero Decaloba sensu Feuillet & MacDougal. El análisis converge en muchos puntos con los estudios filogenéticos realizados a partir de secuencias de ADN, aunque éstos agrupan los subgéneros Tacsonia y Distephana con el subgénero Passiflora. Los genomas cloroplástico y mitocondrial de 213 representantes de 151 especies y 15 subgéneros de Passiflora fueron estudiados por PCR - RFLP, identificando 280 haplotipos para ADNcp y 372 para ADNmt. El análisis factorial de los datos de ADNcp permitió visualizar una fuerte separación de los subgéneros Apodogyne, Decaloba, Murucuja, Pseudomurucuja y Psilanthus (constituyendo el "grupo Decaloba"), mientras que la clasificación por el método de neighbor joining pone en evidencia tres grupos bien soportados dentro de Passiflora, correspondiendo a tres divisiones principales de la clasificación propuesta por Feuillet & MacDougal. El primero, llamado "grupo Passiflora", incluye los subgéneros Calopathanthus, Deidamioides, Distephana, Dysosmia, Dysosmioides, Manicata, Passiflora, Tacsonia, y Tacsonioides, con una débil subestructura y una diferenciación intra-específica considerable. El segundo incluye el subgénero Astrophea, y el tercero el "grupo Decaloba". Las especies del grupo externo toman una posición indeterminada entre los grupos de Passiflora. El fenograma del análisis de ADNmt separa cuatro grupos viii parcialmente soportados, uno de ellos correspondiendo al mismo "grupo Decaloba" del ADNcp. Los subgéneros Astrophea y Tryphostemmatoides aparecen integrados dentro del "grupo Passiflora", mientras que el subgénero Tacsonia constituye un grupo distinto y homogéneo, cerca de otro grupo que incluye especies de las series Kermesinae, Simplicifoliae, Lobatae, y Menispermifoliae del subgénero Passiflora. Así, los análisis de los dos genomas, cloroplástico y mitocondrial, muestran imágenes divergentes de la diversidad de Passiflora, esencialmente por la posición del grupo externo, la posición relativa de los cuatro subgéneros, y las relaciones entre especies. Estas divergencias parecen ligadas a las diferencias en las tasas de evolución y el modo de transmisión de los genomas, así como a una evolución reticulada en el género. Palabras clave: Passiflora L., Colombia, Andes, zona cafetera, distribución, biodiversidad, endemismo, descriptores morfológicos, mitocondria. ix PCR-RFLP, cloroplasto, OBJECTIVES General objectives To study the distribution and organization of the diversity of the Passifloraceae of Colombia, with a triple aim of exploring the diversity of available genetic resources, evaluating the risk of their erosion and their potential as an indicator group for the conservation of biodiversity, particularly in the Colombian coffee growing zone. Specific objectives • To map the distribution and the diversity of the Passifloraceae species of Colombia using a Geographic Information System (GIS). • To characterize the morphologic and molecular diversity at inter- and intraspecific levels. • To evaluate the potential of Passiflora as an indicator group to evaluate the risks of biodiversity erosion and take them into account in the development of strategies for in situ genetic resources management and the conservation of the corresponding natural habitats. x TABLE OF CONTENTS Summary…………………………………………………………………………….. i Résumé……………………………………………………………………………….. iv Resumen……………………………………………………………………………… vii Objectives…………………………………………………………………………….. x Table of contents…………………………………………………………………….. xi List of tables………………………………………………………………………….. xvii List of figures………………………………………………………………………… xix _______CHAPTER I GENERAL INTRODUCTION………...…………………………………………… 1 I.1. Introduction……………………………………………………………………... 2 I.1.1. History…………………………………………………………………………. 2 I.1.2. Taxonomy and general distribution…………………………..……………… 3 I.1.3. Botany………………………………………………………………………….. 4 I.1.4. Biology of reproduction………………………………………………………. 6 I.1.5. Uses…………………………………………………………………………….. 10 I.1.6. Cultivated species……………………………………………………………... 12 I.1.7. Breeding………………………………………………………………………... 13 I.1.8. Passiflora research and prospects, in Colombia and in the Andean region.. 15 I.1.9. Structure of the thesis……………………………………………..………….. 17 _______CHAPTER II DIVERSITY OF COLOMBIAN PASSIFLORACEAE: BIOGEOGRAPHY AND AN UPDATED LIST FOR CONSERVATION…..….................................... 18 II.1. Diversity of Colombian Passifloraceae: biogeography and an updated list for conservation……………………………………………………………………… 19 II.1.1 Abstract…..…………………………………………………………………… 19 xi II.1.2. Introduction…..……………………………………………………………… 20 II.1.3. Materials and methods……….……………………………………………… 24 II.1.3.1. Study area.............................................................................................. 24 II.I.3.2. Herbarium and literature data…….………………………………….. 24 II.I.3.3. Expeditions and samples collected……………………………………. 25 II.I.3.4. Threat status of Passifloraceae…..…..………………………………... 25 II.1.4. Results…………………..…………………………………………………….. 26 II.1.4.1. Data collecting……………………………………………..………… 26 II.1.4.2. Distribution of species richness…………………………..…………... 26 II.1.4.3. New Passifloraceae checklist for Colombia…………….……………. 28 II.1.4.4. Endemism………….………………………………………………….. 32 II.1.4.5. Threatened species………………….………………………………… 34 II.1.5. Discussion……………..……………………………………………………… 35 II.1.6. Conclusions……………………..…………………………………………….. 64 II.1.7. Acknowledgements……….………………………………………………….. 64 II.1.8. Appendix 1. Synonymy …………….………..……………………………….. 65 _________CHAPTER III DISTRIBUTION, DIVERSITY AND IN SITU CONSERVATION OF COLOMBIAN PASSIFLORACEAE……………………......................................... 67 III.1. Distribution, diversity and in situ conservation of Colombian Passifloraceae............................................................................................................... 68 III.1.2. Abstract............................................................................................................ 68 III.1.2. Introduction..................................................................................................... 69 III.1.3. Materials and methods.................................................................................... 73 III.1.3.1. Geography and climate......................................................................... 73 III.1.3.2. Herbarium, field and literature data..................................................... 75 III.1.3.3. Species distribution and richness.......................................................... 75 III.1.3.4. Description of climatic preferences...................................................... 75 III.1.3.5. Potential species distribution................................................................ 76 xii III.1.4. Results and discussion.………….………….………………………………. 77 III.1.4.1. Distribution of observations................................................................. 77 III.1.4.2. Species richness….……..………………………..…………………... 80 III.1.4.3. Species diversity….………………………………………………….. 81 III.1.4.4. Distribution by altitude......................................................................... 81 III.1.4.5. Climatic requirements.…………..…………………………………… 84 III.1.4.6. Areas of distribution - endemic species................................................ 86 III.1.4.7. Potential distribution of species and species assemblages….………. 94 III.1.4.8. Conservation of Passifloraceae and their habitat.………………….. 96 III.1.4.9. Passifloraceae as indicators of biodiversity…….…………………… 98 III.1.5. Conclusions…….……………………………………………………………. 100 III.1.6. Acknowledgements.......................................................................................... 101 _______CHAPTER IV A PHENETIC ANALYSIS OF MORPHOLOGICAL DIVERSITY IN THE GENUS Passiflora L…………………………………………………………………. 102 IV.1. A phenetic analysis of morphological diversity in the genus Passiflora L… 103 IV.1.1. Abstract……………………..……………………………………………...... 103 IV.1.2. Introduction……..………………………………………………………....... 104 IV.1.3. Materials and methods……………………………………………………... 110 IV.1.3.1. Plant materials………………………………………………………... 110 IV.1.3.2. Data collection………………………………………………………... 112 IV.1.3.3. Analyses of quantitative variation…….…..…………………………... 112 IV.1.3.4. Cluster analyses on qualitative data…………………………………. 119 IV.1.4. Results and Discussion…………………………………………………….... 121 IV.1.4.1. Quantitative variation……………………………………………….... 121 IV.1.4.2. Correlations and principal components analysis (PCA)……………... 123 IV.1.4.3. Qualitative variation among and within subgenera………………….. 126 IV.1.4.4. Cluster analysis on the reduced descriptor list……………………….. 129 IV.1.4.5. Cluster analysis on the global descriptor dataset…………………….. 133 xiii IV.1.4.6. The “Passiflora cluster”……………………………………………..... 133 IV.1.4.7. The “Tacsonia cluster”……………………………………………...... 134 IV.1.4.8. The “Decaloba cluster”……………………………………………..... 135 IV.1.4.9. The “Astrophea cluster”…………………………………………….... 137 IV.1.4.10. Morphological and molecular diversity……………………………... 141 IV.1.5. Conclusions………………………………………………………………….. 145 IV.1.6 Acknowledgements………………………………………………………….. 147 _____ __CHAPTER V CHLOROPLAST AND MITOCHONDRIAL DNA VARIATION IN THE GENUS Passiflora L. (PASSIFLORACEAE) AS REVEALED BY PCR-RFLP... 148 V.1. Chloroplast and mitochondrial DNA variation in the genus Passiflora L. (Passifloraceae) as revealed by PCR-RFLP………………………………………... 149 V.1.1. Abstract….......................................................................................................... 149 V.1.2. Introduction…………………………………………………………………... 150 V.1.3. Materials and methods..................................................................................... 157 V.1.3.1. Taxon sampling……………………………………………………… 157 V.1.3.2. DNA Extraction and PCR-RFLP analyses…………………………... 162 V.1.3.3. Data analysis………………………………………………………… 164 V.1.4. Results………………………………………………………………………… 165 V.1.4.1. PCR amplification…………………………………………………… 165 V.1.4.2. Restriction analysis………………………………………………….. 165 V.1.4.3. PCR-RFLP haplotypes………………………………………………. 167 V.1.4.4. Principal co-ordinates analysis……………………………………... 170 V.1.4.5. Cluster analysis……………………………………………………… 170 V.1.5. Discussion…………………………………………………………………….. 181 V.1.5.1. Chloroplast DNA diversity………………………………………….. 181 V.1.5.2. Mitonchondrial DNA diversity………………………………………. 183 V.1.5.3. Divergences in the evolutions of chloroplast and mitochondrial genomes……………………………………………………………………….. xiv 184 V.1.5.4. Diversity of organellar genomes and Passiflora systematics………... 187 V.1.6. Conclusions…………………………………………………………………… 191 V.1.7. Acknowledgements............................................................................................ 192 General discussion…………………………………………………………………… 193 1. Discusion………………………………………………………………………. 194 1.1. Biogeography and conservation……………………………………….. 194 1.2. Morphologycal and molecular diversity……………………………….. 195 1.3. Importance of reticulate evolution in Passiflora………………………. 197 1.4. Phylogeography………………………………………………………... 199 Conclusions and futeres prospect…………………………………………………... 201 1. Conclusions……………………………………………………………... 202 2. Futures prospect……………………………………………………….. 204 Bibliography................................................................................................................. 206 Congress communications.......................................................................................... 230 Annexes......................................................................................................................... 233 Annex 1. Infrageneric classification according to Killip (1938) with emends of Escobar (1988, 1989) and MacDougal (1994)…………………..………… 234 Annex 2. Infrageneric classification according to Feuillet & MacDougal (2003)…………………………………………………………………………. 236 Annex 3. Species cultivated in Colombia………………….…………………. 238 A.3.1. Passiflora edulis Sims……………………………………………... 238 A.3.2. Passiflora ligularis Juss…………………………………………… 239 A.3.3. Passiflora tripartita var. mollissima (Kunth) Holm-Niel. & Jørg.... 240 A.3.4. Passiflora tarminiana Coppens & Barney........................................ 240 A.3.5. Passiflora quadrangularis L……………….…………………........ 242 A.3.6. Passiflora maliformis L………………………………………….... 242 A.3.7. Passiflora alata Curtis...................................................................... 242 A.3.8. Passiflora popenovii Killip………………………………………... 243 Annex 4. Passiflora molecular diversity. Dendrograms obtained in previous studies................................................................................................................. xv 245 A.4a. Phylogenetic tree sensu Muschner et al. (2003)…………………... 245 A.4b. Phylogenetic tree sensu Yockteng (2003).………………………... 246 A.4c. Phylogenetic tree sensu Yockteng & Nadot (2004)………………. 247 A.4d. Phylogenetic tree sensu Hansen et al. (2006)…………………….. 248 Annex 5. List of morphological descriptors in the genus Passiflora L……… 249 xvi LIST OF TABLES ______CHAPTER I Table 1: Breeding objectives in passion fruits…………..………………………….. 15 __CHAPTER II Table 1. Distribution of Passifloraceae by biogeographic region: number of species (bold) and proportion of shared species…………........................................ 28 Table 2. Number of observations and species of Passifloraceae in the 32 Colombian departments……………………………………………………. 31 Table 3. Number of Passifloraceae species in Colombia and the Neotropics………. 32 Table 4. List of 167 Passifloraceae species of Colombia…………………………… 41 _____CHAPTER III Table 1. Number of observations, species, rare and endemic Passifloraceae species by Colombian division. Source for department areas…………………....... 79 Table 2. Factor loadings, eigenvalues and percentages of variance for the first four components, resulting from the PCA analysis on 19 bioclimatic parameters for the 3,930 collection points (Colombian Passifloraceae)…... 85 Table 3. Total number of Passifloraceae present in Colombia. Number of observations, Maximum distance (MaxD) and Circular area (CA) for each species........................................................................................................... 89 _____CHAPTER IV Table 1. List of accessions used in the present study. Taxonomy according to Killip (1938) and emends by Escobar (1988, 1989) and MacDougal (1994)......... 113 Table 2. List of 127 descriptors used in the morphological characterization study.... 119 Table 3. Mean values and coefficients of variation for the whole sample and for the different subgenera………………………………………………………… 122 Table 4. Factor loadings from principal component analysis (varimax normalized rotation) on 24 quantitative descriptors……..…………………………... xvii 125 Table 5. Variation for 32 qualitative and four categorized quantitative descriptors in the different subgenera sampled…..………………………………….. 128 _____CHAPTER V Table 1. List of species used in this study according to classification by Killip (1938), Escobar (1988) and MacDougal (1994)………………………….. 157 Table 2. DNA sequence and type of primer pairs used in the present study by Demesure et al. (1995)……………………………………………………. 163 Table 3. Numbers of haplotypes and fragments for each combination primer/ enzyme……………………………………………………………………. 165 Table 4. Global distribution of the haplotypes among the genera and subgenera studied…………………………………………………………………….. xviii 169 LIST OF FIGURES ___ _CHAPTER I Figure 1. First drawings of Passiflora representing the passion of Christ………… 2 Figure 2. Distribution of genus Passiflora in the world…………………………… 3 Figure 3. Floral elements in the subgenera Passiflora and Tacsonia……………… 5 Figure 4. Passiflora pollinators……………………………………………………. 8 Figure 5. Flower of Passiflora (probably P. ligularis) made in gold by a precolombian goldsmith of Malagana culture in Colombia….……....… 12 _____CHAPTER II Figure 1. Map of distribution of Passifloraceae specimens for 3,930 collections on five biogeographic regions in Colombia. Points on the maps represent sites of collection………………………………………………………. 27 Figure 2. Diagram comparing the similarity in contribution of Passifloraceae species to the floras of the Colombian biogeographic regions (Jaccard distance)………………………………………………………………… 28 Figure 3. Colombian endemic species……………………………………………... 33 Figure 4. Percentual number of the threat status of 165 Passifloraceae native species under the IUCN criteria…………………….………………….. 34 Figure 5. Distribution of species richness of Passifloraceae in American countries. 38 CHAPTER III Figure 1. Colombia’s geopolitical division in 32 departments and biogeographic division in five regions…………………………………………………. 74 Figure 2. Collection localities of Passifloraceae specimens used in this study, among Colombian departments. Points on the maps represent sites of collection………………………………………………………………. 78 Figure 3. Species richness observed for Passifloraceae in 1x1 km grid cells in Colombia (165 species). Points on the maps represent sites of collection……………………………………………………………….. xix 82 Figure 4. Distribution of total species richness (within circles) and species relative diversity in relation to altitude in Colombia (3,930 observations), for Passiflora and five infrageneric groups………………………………... 84 Figure 5. Distribution of Passifloraceae species in the Principal plane for climatic variables, with indication of genera (Ancistrothyrsus and Dilkea) and subgenera of genus Passiflora….............……………………………… 86 Figure 6. Extent of Passifloraceae species distribution in Colombia: circular area (CA50) vs. maximum distance (MaxD)…………………..…………….. 88 Figure 7. Modeled distribution of Colombian Passifloraceae diversity based on data from 80 species presenting more than 10 observations. Ellipses individualize high richness spots mentioned in the text……………….. 95 Figure 8. Distribution of protected areas in Colombia, showing poor correspondence with areas of high Passifloraceae diversity…………… 97 Figure 9. Correspondence between Passifloraceae high richness spots and coffee growing zone ecotopes…………………………………………………. 99 ______CHAPTER IV Figure 1. Schema of a flowering branch of Passiflora vitifolia Kunth ………..…... 105 Figure 2. Variation in shape and color among species from nine of Killip’s subgenera….............................................................................................. 111 Figure 3. Relative variance components for 57 quantitative descriptors. Bold characters are used for traits displaying more than 50% of variance among subgenera………….……………………………………………. 123 Figure 4. Tridimensional plot of the scores of Passiflora accessions for the first three components of quantitative variation. Colors refer to subgeneric classification……………………………………………………............................... Figure 5. Dendrogram obtained with first set of qualitative data............................. Figure 6. Dendrogram obtained on complete set of qualitative data….………….... 126 132 137 Figure 6a. First part of the dendrogram obtained on complet set of qualitative data…….……………………………………………………………….. Figure 6b. Second part of the dendrogram obtained on complet set of qualitative xx 138 139 data……………………………………………………………………... Figure 6c. Third part of the dendrogram obtained on complet set of qualitative data……………………………………………………………….…….. 140 Figure 7. Morphological affinity between typical representatives of P. rubra (a) and P. capsularis (b). Accessions from Colombia (a, Calarcá, Quindío – b, Cartago, Valle del Cauca) ................................................................ 141 _____ _CHAPTER V Figure 1. General schema of the PCR-RFLP markers…………………………….. 164 Figure 2. Interspecific variation for cpDNA (a) and mtDNA (b)…………………. 166 Figure 3. Principal co-ordinates on cpDNA data (PC1-PC2 and TS1-TS2 regions) estimated with 268 CAPS markers…….……………………................. 171 Figure 4. Phenogram derived from on cpDNA (PC1-PC2 and TS1-TS2 regions) data illustrating the distribution of the different Passiflora subgenera studied………………………………….……………………………….. 173 Figure 4a. Cluster analysis on cpDNA data, ‘Decaloba group’, subgenera Astrophea and Tryphostemmatoides, and outgroup genera.………….… 174 Figure 4b. Cluster analysis on cpDNA data, ‘Passiflora group’. Subgenus Tacsonia in clear-brown. ……………………………………...……….. 175 Figure 5. Phenogram derived from on mtDNA (N41-N42 and N1B-N1C regions) data illustrating the distribution of the different Passiflora subgenera studied....................................................................................................... 178 Fugure 5a. Cluster analysis on mtDNA data, ‘Passiflora group’ and subgenera Astrophea and Tryphostemmatoides……………………………………. 179 Figure 5b. Cluster analysis on mtDNA data, ‘Decaloba group’…………………... 180 Figure 6. Probable relationships among main haploid numbers known in Passiflora subgenera and other Passifloraceae genera as proposed by De Melo et al., (2001)…………..………………………………………. 191 ________ _ANNEXES Figure 1. Species cultivated in Colombia: (a) P. edulis f. flavicarpa; (b) P. edulis 241 xxi f. edulis; (c) P. ligularis; (d) P. tripartita var. mollissima; (e) P. tarminiana………………………………………..……………… Figure 2. Species cultivated in Colombia: (a) P. quadrangularis; (b) P. alata; (c) P. maliformis; (d) P. popenovii…………………………….................... 244 All rights reserved Front cover: Passiflora trinervia (Juss.) pollinated by a sword-billed hummingbird (Ensifera ensifera Boissoneau) in the Colombian Andes (drawing by Jesus Salcedo). Photographs by: John Ocampo Passiflora alata Curtis, P. antioquiensis Karst., P. arborea Spreng., P. bogotensis Benth, P. caerulea L., P. capsularis L., P. coriacea Juss., P. edulis f. edulis Sims, P. edulis f. flavicarpa Degener, P. emarginata Humb. & Bonpl., P. erytrophylla Mast., P. flexipes Triana & Planch., P. foetida, L., P. lanata (Juss.) Poir., P. linearistipula Escobar, P. magdalenae Triana & Planch., P. maliformis L., P. manicata (Juss) Pers., P. parritae (Mast.) Bailey, P. popenovii Killip, P. quadrangularis L., P. rubra L., P. sphaerocarpa Triana & Planch., P. tarminiana Coppens & Barney, P. tenerifensis Escobar, P. trinervia (Juss.) Poir., P. tripartita var. mollissima HolmNielsen & Jørgensen, P. vitifolia Kunth. Alvinxon Castro Passiflora arbelaezii Uribe Geo Coppens d’Eeckenbrugge Passiflora ligularis Juss. Gustavo Morales Passiflora longipes Juss. xxii CHAPTER I ____________________________________ General introduction Chapter I: General introduction _______________________________________________________________________ I.1. Introduction I.1.1. History Some plants have received attention from man because they are important to him as sources of food, shelter, medicine, or even narcotics. But the first time passion flowers caught the attention of Europeans, it was for another reason, for the Spaniards who first encountered these plants in the New World in the 16th century, saw in them the elements of the passion of Christ, and a sign that the New World would successfully be converted to Christianity (Uribe 1955a). This religious symbolism gave these plants their common name of Flos Passionis, or “passion flowers”. The Spanish Jesuit Juan Romero presented the first drawing to Pope Paul V (Camollo Borgense) in 1608. A few years later, many similar drawings were made available to a wider audience in Italy and Germany (Kugler & King, 2004). The botanical features in these drawings were transformed to support the religious interpretation very explicitly (Figure 1). Figure 1. First drawings of Passiflora representing the passion of Christ: left, by Dominican monk Simone Parlasca's 1609; right, by Eugenio Petrelli in 1610 (http://www.flwildflowers.com/passiflora.html). 2 Chapter I: General introduction _______________________________________________________________________ I.1.2. Taxonomy and general distribution The family Passifloraceae is placed in the order Malpighiales (Judd et al., 2002), and divided in two tribes, Paropsieae and Passifloreae, with ca. 630 species distributed around the world. Of the 20 currently recognized genera, four are present in America (Ancistrothyrsus, Dilkea, Mitostemma and Passiflora). Passiflora is the largest one, comprising approximately 525 species, mostly distributed in the tropical regions of America, from the coastal zones up to 3,800 m in Andean páramos (Holm-Nielsen et al., 1988). Only 22 species grow in the Old World, in the tropical and sub-tropical regions of Southeast Asia and the Austral Pacific (Figure 2). Figure 2. Distribution of genus Passiflora in the world (by John Ocampo). According to Killip (1938), the genus Passiflora was created by Linnaeus in 1753, who described 24 species in his Species Plantarum, a number increased to 35 by Lamarck (1789). The first extensive monograph of the family was published by Cavanilles in 1790 with 43 species treated. He was followed by authors like Jussieu (1805), De Candolle (1822, 1828), Masters (1872, 1877), and Harms (1898, 1925), who described about 250 species divided into 21 sections (Killip, 1938). In his monograph of 1938, The American Species of Passifloraceae, Killip made the most extensive description of New World species, classifying 355 species into four and 22 subgenera, on the basis of floral 3 Chapter I: General introduction _______________________________________________________________________ morphology (Annex 1). In Colombia, the priest Uribe (1954, 1955a, 1955b, 1957, 1958, 1972) described several new species and Escobar (1986, 1987, 1988a, 1988b, 1989, 1990, 1990 inedited, 1994) revised subgenera Distephana, Manicata (syn. Granadillastrum), Rathea and Tacsonia, including Tacsoniopsis in the latter, and described one additional subgenus, Porphyropathanthus. She passed away in 1993, leaving an inedited document on her revision of subgenus Astrophea. In the last decade, MacDougal and Feuillet have published many papers including the description of about 15 new species, mainly of the subgenera Decaloba and Astrophea (MacDougal, 1992, 1994, 2006; Feuillet, 2002, 2004). Recently, Feuillet & MacDougal (2003; Annex 2) proposed a new infrageneric classification in Passiflora. According to their proposal, based on morphological characters, four subgenera would be recognized: Astrophea and Deidamioides, from South and Central America, Decaloba, from America, Southeast Asia and Australia, and Passiflora, exclusively from America (Ulmer & MacDougal, 2004). Additionally, they downgraded genus Tetrastylis as a section of subgenus Deidamioides. I.1.3. Botany The plants of genus Passiflora are mostly climbers with herbaceous or woody stems and axillary tendrils, or very rarely arbustive or arborescent. Their leaves are alternate, generally simple, entire, lobed or palmate. Stipules are generally present at the base of petioles; the tendrils are also axillary, arising from sterile pedicels. Passion flowers are bisexual or unisexual, regular. Figure 3 shows the different elements of the flowers of two species of subgenera Passiflora and Tacsonia. The large receptacle is often hollowed out like a cup or basin, and bears numerous filamentous or annular appendages between the corolla and stamens, which may be brightly colored and form a conspicuous corona of great diversity. The calyx is composed of 3–5 free or basely connate, imbricate sepals, and the corolla of 3–5 free or basely connate petals, which may be absent in rare cases. The 3–5 (10) stamens are inserted either at the bottom of the perianth, or at the base or top of the gynophore; their filaments are subulate or filiform, free or monoadelphous, sheathing the gynophore; the anthers are versatile, introrse, two-celled, with a 4 Chapter I: General introduction _______________________________________________________________________ longitudinal dehiscence. The ovary is superior, more or less stipitate, very rarely sessile, unilocular, consisting of 3–5 united carpels. The styles are equal in number to the placentae, cohering at the base, distinct at the top, spreading, simple or branched, or separate; stigma are clavate or peltate, sometimes bilobed; The ovules are numerous, anatropous, 1–2 seriate, attached to 3–5 parietal linear placentae by longer or shorter funicles, enlarged into a cupule at the umbilicus. The fruit is a 1-celled, indehiscent berry or a capsule with 3–5 semi-placentiferous valves. Seeds are numerous, with a funicle dilated into a pulpy cupuliform or saccate aril; their testa is crustaceous, foveolate, easily separable from the membranous endopleura, which bears a longitudinal raphe. The straight embryo occupies the axis of a fleshy dotted albumen; the cotyledons are foliaceous and flat, the radicle cylindric, near the hilum, centrifugal. Figure 3. Floral elements in the subgenera Passiflora (left, P. ligularis) and Tacsonia (right, P. tripartita var. mollissima). Photograph by Geo Coppens d’Eeckenbrugge. 5 Chapter I: General introduction _______________________________________________________________________ I.1.4. Biology of reproduction The high variability of flower shape and colors in Passiflora corresponds to several pollination syndromes, showing adaptations to insects, birds and bats (Figure 4). The pollinator behavior may be influenced by morphological as well as chemical floral features (Varassin et al., 2001). Hypanthium length, corolla and corona color, stigma position, and concentrations of sugars or salt in their nectar rewards are fundamental for each pollinator type. Small bees and wasps are common pollinators of small-flowered species, particularly those of subgenera Astrophea, Decaloba and Tryphostemmatoides. Species with flowers of medium or large size are pollinated by large wasps (Xylocopa spp.) mainly in subgenus Passiflora. Insect pollination is generally associated with a well-developed corona, with concentric combination of white, yellow and/or purple colors. The width and depth of the flower, the resistance of the operculum closing the nectary chamber, and the distance between the corona and sexual organs are correlated with the size of the insect. Hummingbird pollination syndrome is the second most prevalent one. It is associated with dominant orange, pink or red colors and a tubular general morphology resulting from a tubular hypanthium and/or an erect corona prolonging the floral tube around the androgynophore. Such syndromes are dominant in particular groups. The highest specialization is presented by Andean species of subgenus Tacsonia and P. trinervia (subgenus Psilanthus), whose very long hypanthia are essentially adapted to the sword-billed hummingbird Ensifera ensifera Boissoneau (Buchert & Mogens, 2001). The same syndrome is displayed in the bright-red flowers of subgenera Distephana and Murucuja, pollinated by smaller hummingbird lowland species, such as Phaethornis superciliosus L. and Trochilus polytmus L. (Snow, 1982). Some species, for example P. kermesina and P. coccinea, are also visited and apparently pollinated by Heliconius butterflies (Benson et al., 1976). Bat-pollination has been described in several species from the Amazon, such as P. mucronata Lam. and P. ovalis Vell. ex Roemer, pollinated by the phyllostomid Glossophaga soricina Pallas, as well as in P. penduliflora Bert, whose most frequent visitor is the Greater Antillean long-tongued bat Monophyllus redmani Leach (Sazima & Sazima, 1978; Kay, 2001). The corresponding pollination syndrome is mainly constituted by nocturnal anthesis, white 6 Chapter I: General introduction _______________________________________________________________________ flowers, the release of an unpleasant odor and a particular arrangement of stigmas and styles. In many species, rapid movements of the stigmas are observed. When the flower opens, the styles are erect; they then bend, bringing the stigma surface close to the stamens, enabling contact with the pollinator; the styles straighten up again before the flower closes. In some flowers, however, this process is incomplete or even absent, precluding natural pollination (Fouqué & Fouqué, 1980; Escobar, 1985; Ruggiero et al., 1976). The female fertility of these flowers is relatively low, some exhibiting incomplete development of the pistil, while male fertility remains unchanged. The proportion of these male functional flowers varies during flowering, being apparently dependent on the resources the plant has previously expended in fruit production (May & Spears, 1988). Most species open their flowers in the morning, until the end of the afternoon, favoring pollen transfer by diurnal pollinators. In bat-pollinated species anthesis logically takes place during the night. In P. penduliflora, it starts in the early evening and ends in the early morning, giving hummingbirds the possibility to contribute to pollination (Kay, 2001). High variability in pollen morphology has been described in more than 200 Passiflora species (Erdtman, 1952; Presting, 1969; García et al., 2002; Barrios et al., 2005). Pollen grains are generally of medium size, prolate to oblate-spherical, reticulated, with luminae of irregular diameter. Passiflora species are very generally allogamous. Cross-pollination is favored by the placement of the anthers below the stigma, large, heavy and sticky pollen grains (Nishida, 1958), and frequent physiological self-incompatibility (Bruckner et al., 1999). The genus Passiflora contains both self-compatible and self-incompatible species (Vasconcellos, 1991). Ho & Shii (1986) observed that the self-incompatibility reaction takes place in the stigma papillae. Their diallel crosses suggested that it is controlled sporophytically by a single multiallelic locus. Bruckner et al. (1995) first confirmed this result, however they later found evidence for a second site of rejection in the style (Rêgo et al., 1998, 2000), 7 Chapter I: General introduction _______________________________________________________________________ which appeared consistent with the action of a second locus acting at the gametophytic level (Suassuma et al., 2003). According to Escobar (1992), self-compatibility is the rule in subgenera Tacsonia and Manicata, whose species set seedy fruits after controlled selfpollination. a b c d Figure 4. Passiflora pollinators: a. honey bee on P. sphaerocarpa; b. sword-billed hummingbird on P. mixta; c. large wasp on P. caerulea; d. bat on P. ovalis. As for self-incompatibility, interspecific compatibility appears to be quite variable among Passiflora species. Spontaneous hybrids have been observed in subgenus Tacsonia and their experimental obtention is relatively easy (Escobar, 1985; Schöniger, 1986). However fertility may decline considerably in subsequent generations, F2 and backcrosses (R1), due to reduced flowering or fruiting, exacerbated by poor seed germination and high mortality (Schöniger, 1986). Nonetheless, interspecific compatibility among the most 8 Chapter I: General introduction _______________________________________________________________________ common species of subgenus Tacsonia is consistent with evidences of interspecific introgression (Primot et al., 2005; Segura et al., 2005). Hybridization has also been attempted among species of subgenus Passiflora to produce new ornamentals of for resistance breeding. Interspecific incompatibility appears more frequent than in subgenus Tacsonia. P. edulis has often been crossed with the closely related P. incarnata, with variable results. Fertile hybrids have been obtained by some researchers (Beal, 1972; Winks et al., 1988) and sterile hybrids by others (Knight, 1991), who resorted to chromosomal doubling to partially restore fertility. Ruberté-Torres & Martín (1974) obtained hybrids from 42 parental combinations of seven passion fruit species. Most hybrids were vigorous with intermediate characteristics of foliage, flowers and fruits. They varied in degree of sterility of pollen and seeds. Payán & Martín (1975) used double pollinations, with compatible and incompatible pollens on different stigmas of the same flowers, and applied hormones to delay flower abscission and overcome incompatibility barriers in 28 crosses involving nine species. More recent reports by Junqueira et al. (2005) appear more optimistic on the feasibility of interspecific crosses, showing positive results from crosses involving 15 species of subgenera Passiflora and Distephana (P. coccinea). Chromosome numbers are now available for more than 150 Passiflora species (Snow & MacDougal, 1993; De Melo et al., 2001). Most species of Passiflora are diploid, with 2n = 12, 18, 20 or 24 chromosomes, although some tetraploids (2n = 24), hexaploids (2n = 36) and octoploids (2n = 72) have been observed (Snow & MacDougal, 1993; De Melo et al. 2001). Several chromosome base numbers (n = 3, 6, 9) have been proposed for the genus, however in the absence of a clear understanding of the phylogenetic relationships among species, there has been no consensus (Storey, 1950; Raven, 1975; Snow & MacDougal, 1993; Yockteng, 2003). According to De Melo et al. (2001), n = 6 is the most probable base number for the genus, whereas n = 9, n = 10 and n = 12 were considered secondary. However, the second most probable base number n = 12, appears to have played an important role in the evolution of the group because it is better represented in other genera of the family (De Melo & Guerra, 2003). Yockteng (2003) and Yockteng & Nadot (2004) presented an alternative hypothesis where n = 12 is the 9 Chapter I: General introduction _______________________________________________________________________ ancestral chromosome number of Passiflora, as genus Adenia (2n = 24) was placed as a sister clade to Passiflora in their phylogenetic study of the ncpGS sequences. In Passiflora, biparental plastid genome transmission was first suspected by Corriveau & Coleman (1988), following epifluorescence microscopy observations on P. edulis. Do et al. (1992) showed that RFLP markers of cpDNA were mostly inherited maternally in crosses between yellow and purple maracuja when the former (P. edulis f. flavicarpa) was used as female parent, and biparental in the reciprocal cross, suggesting asymmetric post-fertilization exclusion processes. Transmission was paternal in the hybrid ‘P. coccinea Aubl x P. edulis f. flavicarpa’. Recently, Mráček (2005) observed biparental transmission between P. menispermifolia Kunth and P. oerstedii Mast., and Muschner et al. (2006) established paternal transmission of cpDNA in four interspecific hybrids of subgenera Passiflora and Dysosmia and maternal transmission in an interspecific hybrid of subgenus Decaloba. All mtDNA were maternally transmitted in the same five hybrids. More recently, Hansen et al. (2007) found paternal or biparental inheritance of cpDNA, in 16 interspecific crosses and maternal inheritance in an intraspecific progeny of P. costaricensis, the only species of subgenus Decaloba in their sample. I.1.5. Uses The discovery of several thousands years old seeds of Passiflora from archaeological sites in Virginia, North America, provides strong evidences of the pre-historic use of the fruits by the Amerindian people (Gremillion, 1989). In Colombia, 200 years before Christ, the Malagana people were inspired by a passion flower for a gold jewel (Figure 5; BRC 2007). Most Passiflora species present an ornamental interest, thanks to their spectacular shapes and colors, others are of pharmaceutical interest for their sedative, antispasmodic, antibacterial and insecticidal properties, but they are even more appreciated for their edible fruits, which are consumed in natura or in preparations such as juices, sherbets and ice cream. 10 Chapter I: General introduction _______________________________________________________________________ Several species of Passiflora have been used extensively in the traditional therapeutics in many countries. P. edulis has been used as a sedative, diuretic, anthelmintic, antidiarrheal, stimulant, tonic and also in the treatment of hypertension, menopausal symptoms, colic of infants in South America (Chopra et al., 1956; Kirtikar & Basu, 1975; Mowrey, 1993). In Nagaland (India), fresh leaves of Passiflora edulis are boiled in a little amount of water and the extract is drunk for the treatment of dysentery and hypertension (Jamir et al., 1999). The extract of P. alata (fragrant granadilla) has been used as an anxiolytic, sedative, diuretic and an analgesic in Brazil (Oga et al., 1984). In the West Indies, Mexico, the Netherlands and South America, the roots of P. caerulea L., have been used as a sedative and vermifuge. In Italy, the plant has been used as an antispasmodic and sedative (Hickey & King, 1988; Dharwan et al., 2004). P. foetida L. leaf infusion has been used to treat hysteria and insomnia in Nigeria (Nwosu, 1999). In India this plant (P. foetida) is widely cultivated and its leaves are applied on the head for giddiness and headache; a decoction is given in cases of biliousness and asthma and the fruit is used as an emetic (Kirtikar & Basu, 1975). In Brazil, the same herb is used in the form of lotions or poultices for erysipelas and skin diseases with inflammation (Chopra et al., 1944). P. incarnata L. is a popular traditional European remedy (Handler, 1962) as well as a homoeopathic medicine (Rawat, 1987) for insomnia, anxiety, and has been used as a sedative tea in North America (Bergner, 1995). The juice of P. maliformis L. is used for intermittent fevers in Brazil. P. quadrangularis (giant granadilla) is used throughout the Caribbean as a sedative and for headaches. Leaf infusion is taken for high blood pressure and diabetes (Seaforth et al., 1983). In Central America, stems of the aerial parts of P. sexflora Juss. and P. vitifolia Kunth have been used against snakebites (Morton, 1981). 11 Chapter I: General introduction _______________________________________________________________________ Figure 5. Flower of Passiflora (probably P. ligularis) made in gold by a precolombian goldsmith of Malagana culture in Colombia, b.C. 200 - 200 a.C (Banco de la República de Colombia, Museo del Oro, Bogotá D.C). I.1.6. Cultivated species More than 80 Passiflora species produce an edible fruit, the most interesting ones belonging to subgenera Passiflora and Tacsonia (Martin & Nakasone, 1970; Coppens d’Eeckenbrugge, 2003). The two botanicals forms of P. edulis Sims, edulis (purple maracuja) and flavicarpa Degener (yellow maracuja), are by far the most important fruit crops in the family, with a world production estimated at ca. 640.000 tons (http://www.passionfruitjuice.com) and a permanent presence on the international market. Other cultivated passion fruits are P. tripartita var. mollissima (Kunth) HolmNielsen & Jørgensen (curuba de Castilla), P. tarminiana Coppens & Barney (curuba India), P. ligularis (sweet granadilla), P. maliformis (granadilla de piedra, cholupa), P. quadrangularis (giant granadilla), P. popenovii Killip (granadilla de Quijos), P. alata (fragrant granadilla) and P. laurifolia L. (golden apple) (Annex 3). These eight species are mainly commercialized in South American for the local and national markets, principally in Colombia and Brazil, with incursions on the international market. Passion fruits are consumed fresh or processed into juices, sherbets, ice cream, and components of industrial pastry and candies. The most important commercial species are susceptible to a large number of pests and diseases, with considerable negative effects on production. Thanks to the high number of species producing and edible fruit of commercial size, the 12 Chapter I: General introduction _______________________________________________________________________ genus Passiflora has high potential for crop diversification and economic development, which induced research institutions in the Andean countries to prioritize their characterization and the evaluation of wild and cultivated populations (Debouck & Libreros, 1995), and develop strategies for conservation and improvement of these genetic resources. I.1.7. Breeding The genetic variability in the genus Passiflora is very wide, both within the genus and within the most cultivated species. This variability can be used by breeders in their efforts to improve traits in the species of subgenera Passiflora and Tacsonia, which are diploid with 2n = 18 chromosomes. In spite of the success of a few flag species, most species with an edible fruit are still poorly known. Few species have diffused out of America and only the maracuja, the sweet maracuja, the granadilla, the stone granadilla and the curubas have been objects of intensive cultivation. Many wild species of Passiflora present characteristics of interest for breeding programs but the hybridization of these species with cultivated species is not always possible. Several levels of passion fruit selection are observed in relation to the coexistence of all domestication stages. Many species, which are rarely cultivated but simply harvested, have never undergone artificial selection. Cultivation of species grown in home gardens has resulted in the creation of some unique types, such as the large-fruited giant granadilla (P. quadrangularis var. macrocarpa), which bears exceptionally large fruit. In commercialized species, such as the maracuja, sweet granadilla and curubas (tacsos), growers carry out phenotypic or mass selection when establishing or renewing new plots, in function of their observations or, sometimes, in function of an ideotype imposed by the local market. Seeds are collected from a small number of good quality fruits plucked from one or two high performing plants. Given the size of the plots, the total population is small and the selection intensity is low, especially if the plantation renewal cycle is long. This practice and seed exchanges maintain considerable variability in the populations. 13 Chapter I: General introduction _______________________________________________________________________ Modern breeding has been focused on P. edulis. The first work took place in the developed tropical or subtropical regions where the commercial cultivation of P. edulis was initiated, but with very limited genetic resources and without even knowing the existing variation. These few institutional works concentrated on the clonal propagation of hybrids between yellow and purple maracujas, obtained on a very narrow genetic base, in Australia, Hawai and Florida (Knight, 1992; Vanderplank, 2000; Winks et al., 1988). The main breeding objectives were early flowering, improved yield, resistance to pests and diseases (Fusarium, Phytophthora, Alternaria, Xanthomonas and virus), cold tolerance and fruit quality (Table 1). In spite of the lack of basic knowledge on genetic resources, complex technical approaches have been explored directly, including interspecific hybridizations, in Puerto Rico and Australia (Beal, 1972; Payán & Martín, 1975; Winks et al., 1988), biotechnologies, with in vitro tissue culture (Drew 1991; Cancino & Hodson, 1994), somatic hybridization (Dornelas et al., 1995; Otoni et al., 1995; Barbosa & Vieira, 1997) and genetic transformation with A. tumefaciens (Manders et al., 1994; Silva, 1998; Hall et al., 2000). As the commercial cultivation of P. edulis started in tropical America, their place of origin, with materials repatriated from Hawaii, propagation by seedlings became the rule. This cultural practice, the increase in production areas and the abundance of natural pollinators increased genetic variation, widening the basis for later breeding programs. The most significant ones have taken place in Brazil, first country in production as well as consumption of yellow maracuja and sweet maracuja. The most advanced Brazilian breeding programs have used progeny-testing for the obtention of synthetic populations, aiming at a better productivity, quality and homogeneity of the crop, while maintaining sufficient genetic diversity for efficient cross-pollinations (Coppens d’Eeckenbrugge et al., 2001; Meletti et al., 2005). Several cultivars have been proposed for consumption in natura (e.g. ‘IAC-273’ and ‘IAC-277’; Meletti et al., 2000) and for the agroindustry (e.g. ‘IAC-275’ and ‘CPATU-Casca Fina’; Nascimento et al., 2003). 14 Chapter I: General introduction _______________________________________________________________________ Table 1: Breeding objectives in passion fruits. Objective Productivity Fruit quality • • • • • • • • • • Resistance to fungal diseases Resistance to bacterial diseases • • • • • • Resistance to viral diseases Resistance to nematodes Acclimatization • • • • Selection criteria Vigor, early production, productivity (flowering, fruiting). Self-fertility if clonal selections (out of dated). Adequate and uniform size. Yield of pulp and/or juice. Sugar/acidity ratio (fresh or processed) and pulp aroma. Color of pulp, dark yellow for maracujas, orange to salmon for curubas and giant granadilla. Small seeds. Pericarp colored, turgescent (smooth), durable and resistant during transportation. Alternaria alternata, A. passiflorae and other species for maracujas and curubas. Fusarium oxysporum and F. solani for the purple maracuja and sweet granadilla. Phytophtora nicotianae and P. cinnamoni for maracujas. Septoria passiflorae for the purple maracuja Oidium sp. for curubas. Colletotrichum gloeosporioides (anthracnose) and Cladosporium herborum, especially for curubas. Xanthomonas passiflorae and X. campestris pv. Passiflorae for maracujas, P. alata and giant granadilla. Potyviruses for maracujas (PMV and SMV) and sweet granadilla (ringspot). Tymovirus and closterovirus for maracujas. Meloidogyne incognita and M. javanica for maracujas, giant granadilla, P. alata and curubas Rotylenchulus reniformis for the purple maracuja. Cold resistance for yellow and purple maracujas and frost resistance for curubas. I.1.8. Passiflora research and prospects, in Colombia and in the Andean region There is little factual information available on the Andean passion fruits. This must be generated by comparative tests in the areas where these species are available. Research institutions in Bolivia, Peru, Ecuador, Colombia, and Venezuela could provide information that may be useful as the basis for industrial development. The future of these fruits will depend upon horticultural development. The production of pulp and 15 Chapter I: General introduction _______________________________________________________________________ concentrate has extremely good prospects if commercial-scale production can be established and maintained. In the last ten years, institutions of Andean countries, grouped in a net promoted by Bioversity International (formerly IPGRI), have started a systematic exploration of the genetic resources of Andean Passiflora species, which has allowed the constitution of collections of the most common species. Nowadays, important collections exist in the five countries: Bolivia, Colombia, Ecuador, Peru and Venezuela. Scientists from the region have investigated the genetic variability of the cultivated curubas and related wild species, building the bases for conventional breeding programs. However, similar efforts on the warm climate species have only been carried out in Brazil. The main problems encountered along these research efforts on passion fruit genetic resources have included the difficulty of approaching such a wide morphological and genetic diversity at both intra- and interspecific levels, with related taxonomical problems and inconsistencies, and practical problems for its effective conservation. The specific adaptations and high level of endemism of species originating from a wide range of habitats impose in situ conservation strategies, implying a good knowledge of the distribution of the different species and the repartition of their diversity. In Colombia, several collaborative projects have been focused on Passifloraceae. The Interamerican Development Bank (BID) has supported a regional project, coordinated by Bioversity International in 1994-1997. Colciencias funded in 1999-2001 the national project “Conservación y utilización de los recursos genéticos de pasifloras”, developed by French and Colombian scientists at the IPGRI-Americas office. In 2004, the same group developed a study of diversity of the Passifloraceae and Caricaceae in the Colombian coffee growing zone. All these projects have generated a considerable amount of information on morphology, cytology, palynology, molecular diversity, and biogeography of Passiflora, strengthening the national collections and providing most of the material for the work presented in this thesis. 16 Chapter I: General introduction _______________________________________________________________________ I.1.9. Structure of the thesis Our research is presented in five more chapters. The first two bear on Passiflora geography in Colombia. The first one consists of an inventory of species for Colombia, their distribution among biogeographic regions and administrative divisions, and considers their status of conservation. The second presents a more detailed analysis of the ecogeographic drivers of Passiflora diversity, through distribution modeling in relation to climatic adaptation of the different species, and gives orientations for their in situ conservation. The next two chapters analyze Passiflora diversity, the third chapter concerning morphological diversity and the fourth one genetic diversity. Each of these four chapters is treated independently, with the structure of a scientific article. A wider general discussion is the object of the final chapter. 17 CHAPTER II ____________________________________ Diversity of Colombian Passifloraceae: biogeography and an updated list for conservation Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1 Diversity of Colombian Passifloraceae: biogeography and an updated list for conservation John Ocampo Pérez1, Geo Coppens d’Eeckenbrugge2, María Restrepo1, Andy Jarvis1,3, Mike Salazar1, and Creuci Caetano1,4. 1 Bioversity International (formerly IPGRI), Regional Office for the Americas, A.A. 6713, Cali, Colombia. 2 CIRAD/FLHOR, UPR ‘Gestion des ressources génétiques et dynamiques sociales’, Campus CNRS/Cefe, 1919 route de Mende, 34293 Montpellier, France. 3 International Center for Tropical Agriculture (CIAT), A.A. 6713, Cali, Colombia. 4 Universidad Nacional de Colombia Sede Palmira. Facultad de Ciencias Agropecuarias. Kra. 32 Chapinero, vía Candelaria. Palmira, Valle del Cauca, Colombia. Submitted and accepted in Biota Colombiana II.1.1. Abstract The list of Colombian Passifloraceae was revised, using 3,930 records from literature, herbaria, and field observations. It includes 167 species, 165 of them native, which is equivalent to 27% of the family. Our list brings more details on species distribution and presents 26 species new to Colombia. Passiflora is the most important genus, with 162 species. When compared with other regions, the Andes of Colombia and Ecuador constitute its center of diversity. Within Colombia, the highest diversity is also concentrated in the Andean region, with 81% of the species, particularly in the departments of Antioquia, Valle del Cauca, Cundinamarca, Quindío, Risaralda, and Caldas. The highest number of species is found at elevations between 1000 and 2,000 m. Most common species thrive in disturbed habitats, such as borders of roads, cultivated plots, and secondary forest. Most of the 58 endemic species are high Andean and belong to subgenera Tacsonia and Decaloba. Forty-two species produce an edible fruit, and nine are commercially cultivated. Among the species reported, 70% are threatened to some degree and three are considered extinct. Colombia may still harbor many unknown species in poorly explored departments. A better knowledge of Passiflora diversity and its distribution is needed to develop its economic potential. The urgent task of conserving this threatened richness must target the conservation of these resources as well as their habitat. Both aspects may be combined if Passifloraceae can be used as an indicator of biodiversity in the Andean region, which seems justified by their multiple ecological interactions with many organisms. Keywords: biodiversity, Colombia, Neotropics, Passifloraceae, passionflowers. 19 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1.2. Introduction The Passifloraceae consist of 18 genera and approximately 630 species, distributed throughout the tropics from the coastal zones up to 3,800 m in Andean páramos (HolmNielsen et al., 1988). In America, it is represented by four genera (Ancistrothyrsus, Dilkea, Mitostemma and Passiflora). Passiflora, with about 525 species is numerically and economically the most important genus of the family; it is distributed mainly in the New World (Ulmer & MacDougal, 2004), with only 22 Old World species of subgenus Decaloba (syn. Plectostemma sensu Killip), in the tropical and sub-tropical regions of Southeast Asia and Austral Pacific. Passionflowers are generally perennial lianas or herbaceous vines climbing by tendrils, although some are trees, shrubs, or even annuals. Their wide morphological variation appears to result from the diversity of their habitats as well as their coevolutionary relationships with many organisms, including protective ants (Apple & Feener, 2001), herbivores (particularly Heliconius spp. butterflies; Gilbert, 1982), pollinators, and the plant communities providing them physical support and access to sunlight. Pollination is mainly carried out by insects and birds; several species are batpollinated (Endress, 1994; Büchert & Mogens, 2001) and a few species exhibit elements of the carnivory syndrome (Radhamani et al., 1995). Many species are cultivated for their edible fruit, as ornamentals, or for their medicinal properties (Ulmer & MacDougal, 2004; Coppens d’Eeckenbrugge, 2003; Martin & Nakasone, 1970; Dharwan et al., 2004). P. edulis Sims (maracuja) is by far the best known and economically most important species of the family. When Spanish missionaries arrived in South America in the 16th century, they felt passionflowers were a good omen for their mission. In their unique morphology, they saw the elements of the Passion of Jesus Christ and a sign that the New World would successfully be converted to Christianity (Killip, 1938; Uribe, 1955a). This religious symbolism gave the plant their common name of Flos Passionis, or “passion flowers”. The Latin translation by Pluckenet (1696; cited by Escobar, 1988a) was accepted for the genus Passiflora created by Linnaeus in 1753, who described 24 species in his Species Plantarum (cited by Killip, 1938), a number increased to 35 by Lamarck (1789; cited by 20 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Killip, 1938). The first extensive monograph of the family was published by Cavanilles in 1780 (cited by Killip, 1938), with 43 species treated. They were followed by authors like Jussieu (1805), De Candolle’s (1828), Masters (1872) and Harms (1925), who described about 250 species divided into 21 sections (Killip, 1938). In his 1938 monograph, The American Species of Passifloraceae, Killip made the most extensive description of the New World species, classifying 355 species into 17 genera and 22 subgenera, based on floral morphology (Annex 1). In Colombia, the priest Uribe (1954, 1955a, 1955b, 1957, 1958, 1972) described several new species and Escobar (1986, 1987, 1988a, 1988b, 1989, 1990, 1990 inedited, 1994) revised the subgenera Distephana, Manicata (syn. Granadillastrum), Rathea and Tacsonia, including Tacsoniopsis in the latter, and described one additional subgenus, Porphyropathanthus. She passed away in 1993, leaving an inedited document on her revision of subgenus Astrophea. MacDougal revised Killip’s subgenus Plectostemma in 1994, restoring its ancient name Decaloba. In the last decade, MacDougal and Feuillet have published many papers including the description of about 15 new species, mainly of the subgenera Decaloba and Astrophea (MacDougal, 1992, 1994, 2006; Feuillet, 2002, 2004). Recently, Feuillet & MacDougal (2003; Annex 2) proposed a new infrageneric classification in Passiflora. According to this proposal, only based on morphological characters, four subgenera would be recognized: Astrophea and Deidamioides, from South and Central America, Decaloba, from America, Southeast Asia and Australia, and Passiflora, exclusively from America (Ulmer & MacDougal, 2004). Additionally, they proposed to downgrade genus Tetrastylis as a section of subgenus Deidamioides. Recent molecular analyses partly support the reduction in the number of subgenera (Muschner et al., 2003; Yockteng, 2003; Yockteng & Nadot, 2004; Hansen et al., 2006; see Annex 4a-d), with the existence of at least three major groups, corresponding globally to subgenera Decaloba, Passiflora and Astrophea of the new proposal. On the other hand, molecular data from the different studies are not always consistent on the relative placement of these groups, and their results are less clear at lower levels, with inconsistent grouping of particular species and poor correspondence with some well established morphological divisions. In addition, the monophyly of Passiflora has not been established, and the study of Muschner et al. (2003) even raises some doubts about it. Clearly, more studies, involving more numerous 21 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ species samples, are needed before re-evaluating such a complex and fast evolving group as Passiflora. Colombia’s location and variety of ecosystems places it in second position for biodiversity (MacNeely et al., 1990). The country is divided into five main biogeographic regions: Amazonian, Andean, Caribbean, Orinoquian, and Pacific. The Andean region presents a highly varied topography (100-5,400 m) with three main mountain ranges. Thus, the Eastern, Central and Western Cordilleras separate two large inter-Andean valleys from the Pacific Coast to the West and the Orinoquian ‘Llanos’ to the East. The uplift of the Andes created new habitats and increased local isolation, favoring high speciation rates in many taxa. In Passiflora, a particularly striking example is given by subgenus Tacsonia, whose beautiful and large-flowered species are strictly adapted to high altitudes in cloud forest (2,000-3,800 m) and pollination by the sword-billed hummingbird Ensifera ensifera Lesson, which shows the same distribution (Büchert & Mogens, 2001). As a result of this habitat diversification, the Colombian flora includes one of the world most diverse groups of vascular plants, with 51,220 documented species (May, 1992; UNEP-WCMC, 2004). However, Colombia has undergone recent transformation of large parts of its natural ecosystems, in particular in the Andean region. Seventy percent of the Andes, an area that is vital to the conservation of Colombia’s water supply, have been deforested as a result of both agricultural colonization and human migration (World Press Review, 1993). Destruction of natural habitats has drastically affected many species distributions, often reducing their historical ranges to a set of small, fragmented populations (Brooks et al., 2002). Such habitat alteration is predicted to lead to substantial extinction in the near future. In this context of rapid erosion of biodiversity, Passifloraceae are highly interesting, not only for their fast radiation and spectacular variation in morphology and reproductive biology. Indeed, as stated above, this family is exemplative from the standpoint of coevolution in many respects, such as their particular relationship with specialized herbivores, ants and other nectar feeding insects; most importantly, they are parasites of structure, as they depend on many very different species for their support, from low 22 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ shrubs in disturbed habitats to high trees in primary forests. They are mainly perennials, but their life cycle is much shorter than that of their support. They are sensitive to longterm changes in the ecosystem (dependence on trees) as well as short-to medium-term changes (by their other adaptive traits). Thus, they should constitute an excellent indicator group for the monitoring of biodiversity in Colombia. In addition, Colombia presents a long tradition of diversity in fruit production and consumption, and it is the country with the highest number of marketed passion fruit species, so the study of Passiflora diversity must also be thought in terms of conservation of genetic resources of important or promising fruit crops. The last inventory by Hernández & Bernal (2000) recorded 141 Passifloraceae species distributed in all the biogeographic regions. Forty-eight of them are endemic to Colombia, mostly living in the Andean region. This inventory was based on the study of specimens from five herbaria (COL, HUA, JAUM, MEDEL and MO) and the citations made in publications compiled by several authors that have worked on the family. Several recent, collaborative projects have been focused on Passifloraceae. The Interamerican Development Bank (BID) has supported a regional project, coordinated by IPGRI (currently Bioversity International) in 1994-1997. Colciencias funded in 19992001 the national project “Conservación y utilización de los recursos genéticos de pasifloras”, developed by French and Colombian scientists at the IPGRI Americas office. In 2004, the same group developed a study of diversity of the Passifloraceae and Caricaceae in the Colombian coffee growing zone. All these projects have generated a considerable amount of information on morphology, cytology, palynology, molecular diversity, and biogeography of Passiflora, providing most of the material for the present inventory and allowing us to supplement and update the list of Hernández & Bernal (2000) with new information, such as species new to science or to the country and elements of ethnobotanical information. In addition, the use of a Geographic Information System (GIS) allowed us to re-assess the conservation status of Colombian Passifloraceae. 23 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1.3. Materials and methods II.1.3.1. Study area Colombia is situated in the north of South America, between 12° 26’ 46” N and 4° 13’ 30” S and between 66° 50’ 54” W and 79° 02’ 33” W, covering an area of 1,141,748 km2, with an altitudinal range from the sea level to 5,775 m (http://www.igac.gov.co). The main administrative division defines 32 departments and geographers recognize five biogeographic regions (Hernández et al., 1991). II.1.3.2. Herbarium and literature data The data set consists of information gathered from specimen labels from 18 Colombian herbaria (AFP, CAUP, CDMB, CHOCO, COL, COAH, CUVC, FAUC, FMB, HUA, HUQ, JAUM, MEDEL, PSO, SURCO, TOLI, VALLE, UIS) and five herbaria in other countries (K, MA, MO, NY, P). These collections were gathered between 1750 and 2006. Most specimens were verified or identified, using the keys and descriptions of Killip (1938), and amendments by Holm-Nielsen et al. (1988), Escobar (1988a, 1994), MacDougal (1994) and Tillet (2003). A synonymy list, based on the general list of Feuillet & MacDougal (2003), is given in the Appendix. When possible, voucher label information was used to assign geographic coordinates to specimens, using gazetteers and topographic maps of Colombia (scale 1:50,000 and 1:250,000). The database was supplemented with materials mentioned in species descriptions, essentially those of Killip (1938, 1960), Uribe (1955a), and Escobar (1988a,b, 1989, 1990, 1990 inedited, 1994). Collection records with obviously inaccurate or doubtful data were excluded from the analysis. Coordinates were further checked by plotting all species on a dot map, using the DIVA-GIS 5.2 software (Hijmans et al., 2001). Finally, we followed the infrageneric classification by Killip (1938) with the amendments of Escobar (1988, 1989) and MacDougal (1994). 24 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1.3.3. Expeditions and samples collected The dot map of all geo-referenced specimens was used to plan germplasm collecting trips. The prioritization of explored areas followed three criteria: permission of access (unfortunately not obtained for protected areas), richness of species and collection gaps. The collecting trips were carried out during 2003-2006, covering 555 localities in 17 departments, between 0 and 4,200 m of altitude. The explorations were concentrated in the Andean region, in watersheds, wild forest areas, cultivated fields and road borders. Data were recorded for each collected specimen, including locality names, elevation, geographic coordinates using a hand-held GPS device, status (wild, cultivated or introduced), and ethnobotanical information (if any). These passport data were recorded and tabulated. Finally, the Geographic Information System software DIVA-GIS 5.2 was used to generate a dot map of the distribution of accessions collected/observed during the expedition. II.1.3.4. Threat status of Passifloraceae The distribution area of each native species was characterized by the maximum distance (MaxD) and the circular area (CA50), following the method of Hijmans et al. (2001). This methodology has been applied in a number of studies to provide quantitative assessment of the distribution area required by the Red List criteria, for example by Maxted et al. (2005). MaxD is the largest distance between any pair of observations of one species. CA50 is the total surface within a 50-km radius around all the observations for a same species. These methods were supplemented with historical records of each taxon and subjected to the Red List criteria of the World Conservation Union (IUCN 2003, 2004), involving complex combinations of quantitative observations concerning the size and structure of the population, the range and fragmentation of its distribution (extent of occurrence and area of occupancy), as well as the intensity of their past or foreseeable variation. Along these lines, we considered that CA50 under 20,000 km2, MaxD under 100 km and number of observations under six, as well as the absence of records younger than 100 years, are critical. 25 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1.4. Results II.1.4.1. Data collecting A total of 3,330 herbaria and 45 literature data, concerning 120 species, were gathered and georeferenced when coordinates were not directly available. The highest number of species and specimens were found in the Colombian herbaria COL and HUA, with 1,056 and 976 records, respectively. During the collecting trips, most specimens were observed in forest fragments, gallery forest and forest and road edges, mainly in the watersheds of the coffee growing zone, between 1000 and 2,000 m. In all sites visited during the expeditions, 87 Passifloraceae species were recorded. Five individuals could not be identified. The dot map of Figure 1 shows the spatial distribution of our final dataset of 3,930 records for herbarium (3,330), literature (45) and field collections (555) of Passifloraceae in the different biogeographic regions. II.1.4.2. Distribution of species richness The number of observations and species richness was highest on the Andean slopes (123 species), followed by the Amazonian region with 45 species (Table 1). The Orinoquian was the poorest region, with only 19 species. The Andean and Caribbean regions share the highest number of species (27). By contrast, the Pacific and Caribbean regions only present four species in common. Figure 2 gives a synthetic image of the similarities in species occurrence among regions, confirming a relative similarity between the Amazonian and Orinoquian as well as between the Andean and Caribbean regions. The Pacific Coast Passifloraceae appear relatively divergent. In the Andean region, Antioquia, Valle del Cauca, Cundinamarca and Santander were the departments that displayed the highest richness of specimens and species (Table 2). Considering their area, Quindío, Risaralda and Caldas are even more diverse. The department of San Andrés and Providencia (Caribbean islands) is only represented by P. biflora Lam. and P. pallida L. 26 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Figure 1. Map of distribution of Passifloraceae specimens for 3,930 collections in the five biogeographic regions in Colombia. Points on the map represent sites of collection. 27 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Table 1. Distribution of Passifloraceae by biogeographic region. The diagonal gives their contribution in species number (bold) and contribution to the country’s total. The other cells give the number and proportion of shared species for each pair of regions. Biogeographic region Amazonian Andean Caribbean Orinoquian Pacific amz and car ori pac 45 (28%) 21 (14%) 123 (76%) 9 (12%) 27 (20%) 38 (23%) 15 (31%) 7 (5%) 9 (19%) 19 (12%) 15 (23%) 14 (10%) 4 (6%) 9 (14%) 36 (22%) II.1.4.3. New Passifloraceae checklist for Colombia Table 3 gives the number of species for each genus and subgenus present in Colombia in relation with the number of species present in the Neotropics. The updated inventory of the Colombian species (Table 4) includes a total of 167 Passifloraceae species, representing three genera, Ancistrothyrsus, Dilkea and Passiflora. This is equivalent to 27% of all Passifloraceae. The genus Passiflora is by far the most important with 162 species, representing 11 of Killip’s subgenera, and all the four subgenera defined in the classification proposed by Feuillet & MacDougal (2003). The most abundant species were P. vitifolia Kunth (359 specimens) and P. mixta L. (162 specimens), while 67 species (23%) were represented by a single specimen. Figure 2. Diagram comparing the similarity in contribution of Passifloraceae species to the floras of the Colombian biogeographic regions (Jaccard distance). 28 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ In the expeditions, we found some species that had not been collected in the last decades, such as P. erytrophylla Mast., P. guazumaefolia Juss., and the semi-arborescent P. mariquitensis Mutis ex Uribe. The latter was described in 1783 by José Celestino Mutis during the Botanical Expedition of the “Nuevo Reino de Granada” in Mariquita (Tolima). It was considered extinct by Uribe (1955a) and a synonym of P. pittieri Mast. by Escobar (1990 inedited). However, we could verify that P. mariquitensis still exists, as three specimens that we have collected in highly disturbed forest near Mariquita corresponded very well to the type specimen, while they appeared morphologically distinct from P. pittieri specimens from Costa Rica, Panama, and northwestern Colombia in several traits (e.g. nectar shape, peduncle length, nerve shape). Similarly, we maintained other species that had been considered synonyms by Hernández & Bernal (2000), such as P. mollis H.B.K. (vs. P. cuspidifolia Harms), and P. hahnii Mast. (vs. P. guatemalensis S. Watson), after checking the collected materials against the type specimens. Ours list presents 26 species new to Colombia, from those recognized by Killip (1960), Feuillet & MacDougal (2003) and Ulmer & McDougal (2004) and three inedited from Escobar (1990) and Hernández (2003): Ancistrothyrsus antioquiensis L.K Escobar (ined.), P. alata Curtis, P. andina Killip, P. bucaramangensis Killip, P. candollei Tr. & Planch., P. chocoensis Gerlach & Ulmer, P. cincinnata Mast., P. hahnii (Fourn.) Mast., P. hirtiflora Jørgensen & Holm-Nielsen, P. killipiana Cuatrecasas, P. lyra Planch. & Linden & ex Killip, P. megacoriacea Porter-Utley (ined.), P. mollis Kunth, P. monadelpha Jørgensen & Holm-Nielsen, P. munchiquensis Hernández (ined.), P. occidentalis Hernández (ined.), P. pallida L. (clearly separated from P. suberosa by Porter-Utley, 2003), P. pillosissima Killip, P. popenovii Killip, P. sodiroi Harms, P. tuberosa Jacq., P. rigidifolia Killip, P. tricuspis Mast., P. truxillensis Planch. & Lind. P. caerulea L., recently introduced from Brazil and Argentina and cultivated as an ornamental, was not included in the counts of each department. P. alata was not counted for Quindío and Valle del Cauca either, as the material under cultivation was also introduced from Brazil. Nine more species occur close to the Colombian international border (less than 100 km), and possibly exist also in the country, although they have not been included in this inventory. Another important result is the presence of the genera Ancistrothyrsus and Dilkea in the Andean and Pacific 29 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ regions, the former following the mention of A. antioquiensis by Escobar (1990 ined.). Unfortunately, she passed away before publishing her monograph on arborescent Passifloraceae. Several botanical forms and varieties are mentioned for P. edulis Sims, P. cumbalensis (Karst.) Harms, P. foetida L, P. ligularis Juss., P. longipes Juss., P. rugosa (Mast.) and P. tripartita (Juss.) Poir. A total of 42 species with edible fruit are reported. Nine of them are sold on the international, national and/or local markets, P. edulis f. flavicarpa Degener and P. edulis f. edulis (introduced), P. ligularis, P. tripartita var. mollissima, P. tarminiana Coppens & Barney, P. quadrangularis L., P. maliformis L., P. popenovii Killip, P. nitida Kunth, and P. alata Curtis. Other species, such as P. antioquiensis H. Karst., P. cumbalensis, P. laurifolia L., P. nitida Kunth, P. palenquensis Holm-Niels. & Lawesson P. tiliifolia L., and P. pinnatistipula Cav. are cultivated in home gardens. Some commonly cultivated species seem to depend on human activity for their propagation, which suggests an advanced stage of domestication and/or an incomplete acclimatisation following an ancient introduction. Thus, P. edulis f. flavicarpa, P. ligularis, P. quadrangularis L., P. popenovii P. tripartita var. mollissima, and P. tarminiana, are exceptionally found as feral plants. On the other hand, let us remind that the latter has pullulated as an invasive plant in Hawaii and New Zealand. Another particular case is P. edulis f. edulis, introduced from southern South America, which has naturalized at intermediate to high altitudes, where it is not uncommon in the wild. 30 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Table 2. Number of observations and species of Passifloraceae in the 32 Colombian departments. Department Amazonas Antioquia Arauca Atlántico Bolívar Boyacá Caldas Caquetá Casanare Cauca Cesar Chocó Córdoba Cundinamarca Guainía Guaviare Huila La Guajira Magdalena Meta Nariño Norte de Santander Putumayo Quindío Risaralda S. Andrés y Prov. Santander Sucre Tolima Valle del Cauca Vaupés Vichada Abbreviation Biogeographic region Observation number Species number ama ant ara at bl by cl cq cs cau ce cho cor cun gn gv hu lg ma met na ns pu qu ri sp snt suc to vc va vch amz and car pac and ori Car and car and ori And amz and and ori amz and pac and car and pac and car and ori amz amz and and car car amz and ori and pac and amz and and and pac car and car and and pac amz ori 87 784 10 18 33 145 245 47 4 161 13 211 33 419 16 27 62 21 71 85 170 79 56 150 68 4 203 6 213 420 35 16 19 70 6 7 17 36 36 18 4 42 10 40 9 53 10 14 22 12 31 24 44 36 26 38 24 2 48 3 44 56 20 9 The vernacular names are very diverse for each species. In the Amazonian region, we noted several indigenous names for the species P. foetida var. gossypiifolia Desv. (Iñanaleeg, Murulale), P. holtii Killip (Guachique), P. nitida (Burucuña, Gemarundare, Tuchica, Jino-Gojé), P. serratodigitata L. (Cipo-Cipo), P. vitifolia (Maloca de Fisi). In the Cauca and Nariño departments (south of the Andean region) P. fimbriatistipula Harms and P. ligularis are named Pachuaca and Awapit in the indigenous languages. 31 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Table 3. Number of Passifloraceae species in Colombia and the Neotropics. Genus Ancystrothyrsus Dilkea Mitostemma Passiflora Subgenus Colombia Astrophea Decaloba Dysosmia Distephana Manicata Passiflora Porphyropathanthus Psilanthus Rathea Tacsonia Tryphostemmatoides All Passifloraceae Neotropics 2 3 0 22 52 2 6 1 38 1 3 2 30 4 3 5 3 57 190 20 15 5 156 1 4 3 55 7 167 533 Among the species collected in our expeditions, we found several species growing very commonly in disturbed habitats like road borders, secondary forest margins, and especially riverbanks between 1000 and 2,000 m: P. adenopoda Moc, & Sessé ex DC., P. alnifolia Kunth, P. coriaceae Juss., P. capsularis L., P. rubra L, and P. suberosa L. The latter two can even be considered weeds in the coffee plantations. At higher altitudes (above 2,500 m), P. mixta is also very common in disturbed habitats. II.1.4.4. Endemism Among the 165 native species, 58 (36%) are endemic to the country. The largest concentration of these occurs in the Andean region, principally in the Cordillera Central, in the departments of Antioquia and Tolima. The elevation belt between 1,500 and 2,500 m presents the highest richness of endemic and rare species (≤ 5 observations). Only eight of these were represented with only one specimen (e.g. P. cremastantha Harms), while P. bogotensis Benth and P. antioquiensis were the most collected endemic species, with 23 recorded specimens each. The proportion of endemic species varied considerably among taxonomic groups, especially among the subgenera of Passiflora (Table 4). Thus, Tacsonia (21), Decaloba (13), Passiflora (9) and Astrophea (7) present the highest number of endemic species. Subgenus Tacsonia displays the highest richness 32 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ of endemic species in the Cordillera Central with eight species, mainly of the section Colombiana characterized by a very long peduncle (P. flexipes Triana & Planch., P. linearistipula L.K. Escobar, P. quindiensis Killip and P. tenerifensis L.K. Escobar). Twenty-one species (37%) are restricted to very small areas of one department. These are located mainly in the departments of Antioquia (6), Tolima (4), Santander (3), Cauca (2), while only one such narrow endemic is found for the departments of Bolivar, Boyacá, Chocó, Caldas, Cauca, and Magdalena. Figure 3 shows 15 endemic species from five subgenera. a b c d e f g h i j k l m n o Figure 3. Colombian endemic species (a) P. antioquiensis; (b) P. parritae; (c) P. flexipes; (d) P. linearistipula; (e) P. lanata; (f) P. tenerifensis; (g) P. trinervia; (h) P. longipes; (i) P. erytrophylla; (j) P. bogotensis; (k) P. magdalenae; (l) P. smithii; (m) P. arborea; (n) P sphaerocarpa; (o) P. emarginata. 33 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1.4.5. Threatened species Figure 4 shows the distribution of the 165 Colombian Passifloraceae native species, according to their threat status under the criteria of the IUCN (2003, 2004). Seventy-one percent of them present some threat degree, 10% being critically endangered (CR), 6.1% vulnerable (VU) or endangered (EN). Four of the 16 critically endangered species are endemic. All three extinct species (EX) belong to the Andean subgenus Tacsonia. Unfortunately, the only two species of genus Ancistrothyrsus are included in the category CR. Only 16% of the species were placed in the two categories LC and NT, ‘least concern’ and ‘near threatened’. The species P. alata, P. megacoriacea and P. rigidifolia are placed in the DD category because of deficient data. The 29.3% classified in ‘least concern’, belong mostly to subgenera Decaloba and Passiflora with 18 and 14 species, respectively. Figure 4. Percentual number of the threat status of 165 Passifloraceae native species under the IUCN criteria. 34 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1.5. Discussion Colombia has been subject to many studies focused on inventories of plant species groups (Gentry, 1993; Silverstone-Sopkin & Ramos, 1995; Galeano et al., 1998; Rangel, 1995, 2002). Passifloraceae have been inventoried in taxonomical works by Escobar (1998a, 1989, 1990 inedited) and Hernández & Bernal (2000). As compared to the latter, we have added new information on geographical distribution of each taxon and extended the list to a total of 167 Passifloraceae species, from three genera and the five biogeographic regions, with reports of 26 species new to Colombia. For obvious reasons, the quality of botanical inventories depends on the quality of taxonomical work in this complex family. While the definition of genera and subgenera should not affect significantly studies of the distribution of its diversity across the Colombian territory, such work may be affected to some extent by poor definitions below the subgenus level. Indeed, several morphological groups include species that are very similar, and regularly reported as very difficult to distinguish from each other. In several cases, experts may have underestimated intraspecific variation in widely distributed species, or even intra-individual variation, splitting well known species in several new species only distinguished by a few quantitative or color traits. Among the difficult groups, let us mention particularly subgenus Astrophea, whose species tend to be less well differentiated, at least in sterile specimens, by nectar gland position and number, having only two at the junction of the lamina and petiole, while they may show impressive intraspecific variation in pubescence and intra-individual variation in leaf size and shape according to light exposure and whole tree development (heteroblasty). In subgenus Decaloba also, several morphological groups demand great experience and caution for their identification, even in the most common species such as P. capsularis and P. rubra, which can be found in the same habitats. In the most difficult cases, several species have even changed status several times. For instance, Killip merged P. bauhinifolia H.B.K. with P. andreana Mast. in 1938, and restored it as a distinct species in 1960, while Holm-Nielsen et al. (1988) merged P. bauhinifolia with another close relative, P. alnifolia, a position that we have adopted here. On the other hand, a 35 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ couple of other species may also show very little morphological differentiation, as P. mollis and P. cuspidifolia or P. hahnii and P. guatemalensis, but differ in their altitudinal distribution, which confirms that they are distinct entities. Many new species of subgenus Distephana are also questionable, as one of its two most common species, P. coccinea Aubl., distributed in most of the Amazon, has been split in several species on the basis of bract size, nectar gland numbers, and small variation in numbers and respective colors of the corona series. Concerning Colombia, Vanderplank (2006) underlined that the description of P. coccinea by Escobar (1988) matches perfectly that of P. miniata Vanderplank, so he considered the latter a Colombian species. However, we have not adhered to this opinion for several reasons. Vanderplank described it on material grown in glasshouse and his report does not refer to the examination of Colombian materials. The type and level of the differentiation described between the different new species and P. coccinea is at most of the same order as morphological variation in other common widespread species (e.g. P. vitifolia, P. foetida, P. suberosa, P. alnifolia, P. capsularis, P. mixta, P. cumbalensis, P. maliformis, or P. emarginata). He reported a high level of sexual compatibility with the other common Distephana species, P. vitifolia, which raises the expectation of sexual compatibility with the even closer “true” P. coccinea. Thus we have stuck to the treatment of P. coccinea by Escobar (1988), whose quantitative description is more precise than the original by Aublet (1775) but not fundamentally different. Within subgenus Passiflora, P. maliformis, P. serrulata and P. multiformis constitute other cases of possible overclassification, as they are mostly differentiated by the degree of lobation of their leaves, a trait that is quite variable in many other species, including other Tiliifoliae, such as P. ligularis (Killip, 1938; obs. pers.). A wider problematic group is the series Laurifoliae, with ten species in Colombia, always difficult to identify from incomplete specimens. Although they probably constitute a very young group and they exhibit a high number of common traits, species of subgenus Tacsonia are relatively easy to differentiate. Particularly interesting are the endemics of section Colombiana, from the center of the cordilleras, often characterized by a very long peduncle and linear-lanceolate stipules, and from the northeast and up to the Venezuelan Andes. Several authors have reported easy interspecific hybridization in subgenus Tacsonia, involving cultivated as well as wild materials (Escobar, 1985). This 36 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ phenomenon, by producing spontaneous off-types, may have led to some overclassification in this subgenus. Indeed, of the 30 species reported here for Colombia, five (P. cremastantha Harms, P. formosa Ulmer, P. pamplonensis Planch. & Linden ex Triana & Planch., P. purdiei Killip, P. rigidifolia Killip) are known only from the type material. Whether this is due to high endemism, ancient extinction, or off-types resulting from hybridization cannot be ascertained, unless a second specimen is recorded, as we did for P. linearistipula. It is important to note that P. formosa was described as a new species from the same specimen considered as an off-type of P. lanata (Juss.) by Escobar (1988). Overclassification may be suspected even in better-known species, as P. parritae (Mast.) Bailey, and P. jardinensis L.K. Escobar. Indeed, in populations of the former, we have observed sufficient morphological variation to include the few known specimens of the latter species, which might simply represent a small isolated population. On the other hand, most endemics of subgenus Tacsonia were found in highlands of difficult access, and more species can reasonably be expected to be described from relatively poorly explored areas such as the South of Tolima, Santander and Norte de Santander departments. Our list ranks Colombia as the country with the highest richness of Passifloraceae, followed by Brazil with 127 species. Figure 5 allows comparisons for species richness and relative diversity of passion flowers in the Neotropics, showing the strong influence of latitude (typical of a tropical distribution) and topography on Passiflora diversity. Colombian species richness and diversity is more than twice that of Peru and Venezuela, two countries of similar surface and latitude. Given its much smaller area, Ecuador also presents an impressive diversity. Thus, the northern Andes of Colombia and Ecuador clearly constitute the center of diversity for the genus Passiflora. This is probably due to the greater availability of habitats, especially at high elevations, in these two countries. The presence of three Andean cordilleras in Colombia very probably played a significant role. Indeed, radiation has been very active in the northern Andes, with particular contribution of recent and fast evolving groups, such as subgenera Rathea and Tacsonia, accounting for more than 41 highland species in Colombia and Ecuador. Among them, 21 37 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ (14%) species are endemic to Colombia. Colombian highlands are also rich in representatives of subgenus Decaloba. Figure 5. Distribution of Passifloraceae species richness in American countries, according to information gathered from Killip (1938, 1960), Escobar (1988, 1989, 1990 inedited, 1994), Holm-Nielsen et al. (1988), Jørgensen & León (1999), MacDougal (1994), Vanderplank (2000), Deginiani (2001), Tillet (2003), Ulmer & MacDougal (2004), records of the herbaria cited in this study and many journal articles related with the description of new species present in the America. 38 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ According to Escobar (1988a), 40% of the New World Passifloraceae are found in the Andes. In Colombia, habitats between 1000 and 3,000 m account for only 27% of the land area, yet 81% of the species of Passifloraceae are found there. With 123 species, the Andean region concentrates the highest richness, mainly between 1000 and 2,000 m. The Caribbean region shares the highest proportion of species (27) with the Andean region (Table 1), which is mostly due to the presence of the Sierra Nevada de Santa Marta mountain range in northern Colombia, with a steep gradient of elevation from the Caribbean Sea to summits at 5,775 m. The increase of species richness and endemism with the elevation is generally interpreted as a result of the increasing isolation and decreasing habitat surface in high mountain regions, leading to small, fragmented populations which are prone to speciation (Simpson, 1975; Jørgensen et al., 1995). Another contribution to the particular species richness in Colombia and Ecuador is that of the Pacific Coast region, continuous with the similar highly diverse ecosystems of Central America (Chocó-Darién/Western Ecuador hotspot of Myers et al., 2000), and receiving one of the highest rainfalls in the world, in strong contrast with the conditions prevailing in the westerns Andes and coast of Peru that are arid or semi-arid, or the drier and more contrasted climate of Venezuela. The Passifloraceae species composition of this region appears both diverse and well-differentiated when compared with that of the other biogeographic regions (Figure 2), heightening its interest. This is not surprising, as the Chocó region is recognized as one of the most diverse biotas in the world, with nearly 40% endemism (Gentry, 1986). Until recently, the genera Dilkea and Ancistrothyrsus were only known from the Amazon basin. The description of A. antioquiensis by Escobar (1990 ined.) in the Andes and the observation of Dilkea retusa in the Andes and Pacific regions extend their distribution to other important biota. The distribution of Passifloraceae has been drastically affected by deforestation, principally in the Andean region. Its historical range corresponds to a region with a long history of livestock and agriculture that now supports extensive plantations of coffee, sugar cane, rice, bananas, and potatoes. According to our field observations, very 39 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ common species, such as P. adenopoda, P. alnifolia, P. capsularis, P. coriaceae, P. rubra, P. suberosa, and P. mixta, are mostly species that thrive in secondary forest or disturbed old-growth forest. Human disturbances may even have contributed to extend their distribution, as reported in other plants (Svenning, 1998). According to Myers et al. (2000) and Robbirt et al. (2006), rarity and endemism represent two factors of particular significance in the consideration of risk of decline and extinction. In this context, most Colombian Passifloraceae (70.7%) are under some threat degree according to IUCN criteria. Only 29.3% fall in the ‘least concern’ category (LC), which clearly illustrates the alarming situation for the family (Figure 3). Our results are consistent with the Red List of plants published by the von Humboldt Institute (Calderón, 2005), based on the 141 species listed by Hernández & Bernal (2000), with similar percentages for each category. However, this list only includes P. colombiana L.K Escobar under the category of critically endangered species (CR), while ours places 16 species in this category. Exploration for Passifloraceae was not possible in the protected areas of Colombia that are of essential importance for the conservation of the country’s biodiversity, as we lacked permission of access by the Colombian Ministry for Environment (MMA). Another limiting factor of research for conservation purposes is the conflict situation in many areas (Martin & Szuter, 1999; Dévalos, 2001). Forests in the northern Andes are currently one of the major conservation priorities on a global scale due to their fragility, biological richness, high rates of endemism and multiple anthropogenic threats (Olson & Dinerstein, 1998). As Passifloraceae display very high species richness, endemism and extinction risk in this area, and given their multiple ecological interactions with many organisms, as well as their economic potential, this family should constitute both an important target of conservation efforts and a good indicator of their success. 40 Table 4. List of 167 Passifloraceae species of Colombia. Fifty-eigth endemic species are marked by an asterisk (*); twenty-six species new to Colombia by the abbreviation ‘nr’; nine species probably present in the country are indicated between square brackets. New records, for a given biogeographic region, department (abbreviated as in Tables 1 and 2) or elevation-range are indicated by bold letters. Abbreviations in bold letters in the ‘Notes’ column correspond to the plant habits: shrub (Ab), tree (Ar), and climber (Tr). V.N and I.N. indicate vernacular and indigenous names, respectively. Taxon Biogeographic Region Genus Ancistrothyrsus Harms, 1931 Ancistrothyrsus antioquiensis L.K Escobar (ined), 1988 * nr and [Ancistrothyrsus hirtellus A.H. Gentry, 1992] amz Ancistrothyrsus tessmannii Harms, 1931 amz Genus Dilkea Mast., 1871 Dilkea johannesii Barb. Rodr., 1885 Geopolitical Distribution Elevation Collection for Reference 90-800 Escobar & Roldán 8819 (HUA) - Type 150-350 Gentry & Stein 47114 (MO) Isotype ama pu 50-400 amz va Dilkea parviflora Killip, 1938 amz Dilkea retusa Mast.,1871 amz and pac Bibliographic Reference IUCN Category Notes F.J Roldán (com. pers), Escobar (1990 inedited) Gentry 1992 CR Tr Vester & Matapi 639 (COAH) Holm-Nielsen et al. 1988 CR Tr Reported in the Ecuadorian, Peruvian and Venezuelan Amazon. Tr 100-500 Soejarto 2461 (HUA) Killip 1938 CR Tr ama cq va 100-500 Gentry 64981 (MO) Holm-Nielsen et al. 1988 LC ama ant cho cq gv met pu snt va vc 100-500 López et al. 5947 (COAH) Killip 1938; Uribe 1955b; Holm-Nielsen 1974; HolmNielsen et al.1988 LC Tr V.N.: Canilla de Tente, Tripa de Tente (ama). Edible fruit Tr ant Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Genus Passiflora L., 1753 Subgenus Astrophea (DC.) Masters, 1871 Section Astrophea Passiflora callistema L.K. Escobar, 1994 * car bl 100 E. Forero 487 (COL) - Type Escobar 1990 Inéd., 1994 CR Tr Only known from the type. Section Botryastrophea Passiflora holtii Killip, 1938 amz ama cq gn va 150-500 Jaramillo 7890 (COL) Killip 1938; Escobar 1990 Inéd., 1994 LC/NT Passiflora pyrrhantha Harms, 1926 amz va 400-1000 Shultes & Cabrera 12693 (COL) Killip 1938; Holm-Nielsen et al.1988; Escobar 1990 Inéd., 1994 EN/CR Tr I.N: Guachique, Bejuco (ama). Edible fruit Tr Passiflora securiclata Mast., 1893 amz ori ara by gv va vch 150-500 Betancourt et al. 9753 (COAH) Killip 1960; Escobar 1990 Inéd., 1994 LC Tr Passiflora spicata Mast., 1872 amz gv 150-500 Cuatrecasas 7397 (COL) VU Tr Passiflora spinosa (Poepp. & Endl.) Mast., 1871 amz and ori car ama ant by cq cor cun gn met pu snt va vch 150-500 Zarucchi 4279 (COL) Killip 1938; Holm-Nielsen et al. 1988; Escobar 1990 Inéd, 1994 Killip 1938; Holm-Nielsen 1974; HolmNielsen et al.1988; Escobar 1990 Inéd., 1994 VU Tr V.N.: Cocorella (bl), Bejuco campano (snt) Section Dolichostemma Passiflora citrifolia (Juss.) Mast.,1871 amz va vch 85-500 Barbosa & Zurucchi 2989 (COAH) Killip 1838; Escobar 1990 Inéd. LC Tr 42 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora haughtii Killip, 1938 * and snt 100-700 Haught 1635 (COL) Killip 1938; Escobar 1990 Inéd., 1994 CR Ab Passiflora mariquitensis Mutis ex Uribe, 1954 * and to 420-700 Ocampo et al. 55 (TOLI) Killip 1938; Escobar 1990 Inéd., 1994 CR Ab Formerly considered extinct. Passiflora mutisii Killip, 1938 * and to 600 Mutis 2279 (MA) - Type Killip 1938; Escobar 1990 Inéd., 1994 EX Tr Passiflora pittieri Mast., 1897 pac ant cho 50-1000 Gentry & Aguirre 15318 (COL) Killip 1938; Escobar 1990 Inéd., 1994; Gentry 1976 VU Ab Section Euastrophea Passiflora arborea Spreng., 1826 and car ant bl by cau cl cun hu ma na qu ri to vc 1000-2300 Humboldt & Bonpland 5864 (P) - Type Killip 1938; Pérez 1956; Holm-Nielsen et al. 1988; Escobar 1990 Inéd. NT Passiflora lindeniana Planch. ex Triana & Planch., 1873 and cun ns snt 1000-2700 Linden 1409 (P) - Type Escobar 1994 NT Ar V.N: Cherimoyo (vc), Granadillo arboreo (cun). Edible fruit Ab Passiflora emarginata Humb. & Bonpl., 1813 * and pac cau cl cho na vc 1500-2000 Humboldt & Bonpland (P) - Type Killip 1938; Escobar 1990 Inéd., 1994 LC Passiflora engleriana Harms, 1894 * and ant 1500-2500 Escobar 8853 (COL) Killip 1938; Escobar 1990 Inéd., 1994 VU/EN Passiflora macrophylla Spruce ex Mast., 1883 amz and pac ant cau cho pu na 60-1800 Alcázar & Salgado 1203 (CAUP) Killip 1938; Holm-Nielsen et al. 1988; Escobar 1990 Inéd., 1994 43 LC Ar Edible fruit Ar Ab V.N: Acaba familia (cho) Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora putumayensis Killip, 1938 and pu 1350-2500 Mora-0. 3438 (PSO) Killip 1938; Escobar 1990 Inéd., 1994 EN/CR Ab Passiflora sphaerocarpa Triana & Planch., 1873 * and ant ce cun na ns qu ri snt to vc 400-1700 Schlim 285 (P) - Type Killip 1938; Uribe 1972; Escobar 1990 Inéd., 1994 LC/NT Passiflora tica Gomez-Laur. & L.D. Gómez, 1981 pac ant cho 450-1500 Escobar 2192 (HUA) Escobar 1990 Inéd., 1994 LC/NT Ar V.N: Gulupo de Arbol (cun), Capafraile (to). Edible fruit Ar Section Pseudoastrophea [Passiflora costata Mast., 1872] amz 50-350 Spruce 1670 (K) - Type Killip 1938; Escobar 1990 Inéd., 1994 Passiflora grandis Killip, 1938 * and 1000-2000 Schlim 585 (K) Escobar 1990 Inéd., 1994 EN/CR Tr Reported in the Amazon of Peru, Brazil, Guianas, and Venezuela) (confluence of the rivers Rio Negro and Casiquiare). Ar [Passiflora ovata Martin ex DC., 1828] ori 0-150 Colector n.v. Killip 1938; Escobar 1990 Inéd., 1994 Passiflora phaeocaula Killip, 1927 amz ori gn va vch 150-1100 Madriñan 893 (MO,GH) Killip 1938; Holm-Nielsen 1974; Escobar 1990 Inéd., 1994 LC/NT Tr Ab Passiflora skiantha Huber, 1960 amz gv 150-500 Cuatrecasas 7366 (COL) Killip 1938; Escobar 1990 Inéd. NT/VU Tr ns snt 44 Tr Reported in the Amazon of Venezuela. Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ [Passiflora tessmannii Harms,1926] amz Passiflora venosa Rusby and pac Subgenus Decaloba (DC.) Rchb., 1828 Section Cieca Passiflora apoda Harms,1929 50-500 Tessmann 4385 (N) - Type Killip 1938; Escobar 1990 Inéd., 1994 cho 50-450 Juncosa s.n. (JAUM) n.v. Killip 1938; Escobar 1990 Inéd. VU/EN Tr Reported in the northern Amazon of Peru. Tr and ant cau cl qu na ri to vc 1900-3260 Hazen 9688 (MO) - Isotype Killip 1938; Hernández 2003 LC/NT Tr Passiflora coriacea Juss.,1805 and car pac 250-1500 Uribe 2565 (COL) Croat 1978; Holm-Nielsen et al. 1988 LC Passiflora holosericea L.,1753 car ant by cau cl cho cun hu ma ns qu ri snt to vc at bl ce 0-1400 Cuadros-H 1882 (COL) Killip 1938 LC/NT Tr Passiflora megacoriacea Porter-Utley, 2003 nr car bl 100-200 Killip & Smith 14415 (US) Porter-Utley 2003 DD Tr Passiflora pallidaL., 1753 nr car at bl ma sp 0-200 Dugand & Jaramillo 2844 (COL) Porter-Utley 2003 LC Tr Appel Monkey (sp) Passiflora sodiroi Harms, 1922 nr and cau 1850-2150 Escobar et al. 4368 (PSO) Holm-Nielsen et al. 1988 EN/CR Passiflora suberosa L., 1753 and car ant cau cl cun gv na ns qu snt suc to vc 200-2200 Cuatrecasas 15930 (VALLE) Holm-Nielsen et al. 1988 LC Tr V.N.: Curubita de Monte (ant) amz and car ori pac ama ant bl by cau cl cho cor cq cun gn gv met na ns pu qu snt to va vc vch 0-1500 Killip & Cuatrecasas 58988 (VALLE) Holm-Nielsen 1974; HolmNielsen et al. 1988 LC Tr V.N: Rejito (cun) Section Decaloba Series Auriculatae Passiflora auriculata Kunth, 1817 45 Tr V.N.: Ala de Murcielago Tr Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Series Sexflorae Passiflora sexflora Juss., 1805 and ant hu qu to vc 1700-2300 Zurucchi et al. 5813 (CHOCO) Holm-Nielsen et al. 1988 NT/VU Series Luteae Passiflora filipes Benth., 1843 and qu ri vc 950-1250 Silverstone 7205 (CUCV) Holm-Nielsen et al. 1988 VU Tr Series Miserae Passiflora misera Kunth, 1817 and car ori pac ant at ara bl by cau cl cho cor cun cs lg ma met vc ns 0-1050 E. Forero 9936 (COL) Killip 1938 LC Tr Passiflora tricuspis Mast., 1872 nr and met 1220-2000 Estrada et al. 146 (MA) Killip 1938 CR Tr [Passifora trifasciata Lemaire, 1868] amz ama pu 130-1100 Brandbyge et al.33556 (AAU) Killip 1938; Nielsen et al. 1988 Series Punctatae Passiflora alnifolia Kunth, 1817 and car ant by cau cl cun ma na pu qu ri snt to vc 1400-2500 Hno. Daniel 2803 (MEDEL) Holm-Nielsen et al. 1988 LC Tr Passiflora andreana Mast., 1883 and ant cau cun ma na qu ri 1500-3150 Garcia-B.12949 (COL) Holm-Nielsen et al. 1988 CR Tr Passiflora azeroana L. Uribe, 1955 * and by cun hu snt 2500-3000 Lozano 3718 (COL) Uribe 1957 NT/VU Tr Passiflora biflora Lam., 1789 and car ant at bl ce cl cho cun hu ma met na ns ri sp snt to vc 0-1500 Garcia-B. 11720 (COL) Killip1938; Holm-Nielsen 1974; Croat 1978 46 Tr V.N.: Corvejo (na) Tr Reported in the Amazon of Ecuador, Brazil and Peru. Ornamental (qu) LC Tr V.N.: Peyen Papaya (sp), Desjarretader a (cun) Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora bogotensis Benth., 1845 * and car by cun hu lg ma ns snt vc 2000-3700 Garcia-B. 15291 (COL) Killip 1938; Holm-Nielsen 1974 LC Tr V.N.: Curubo macho (cun) Passiflora bucaramangensis Killip, 1930 * nr and snt 1500-2600 Killip & Smith 16787 (MO) - Isotype Killip 1930, 1938 EN Tr Passiflora candollei Tr. & Planch., 1873 nr amz ama 100 Rudas et al. 2180 (COL) Killip 1938 NT Tr Passiflora chelidonea Mast., 1979 and car pac ant ara cau cho na ns pu ri snt vc 900-3000 Cuatrecasas 12526 (COL) Holm-Nielsen et al. 1988 LC Tr Passiflora cuneata Willd.,1809 and car ant by cho cun ma met ns snt vc 900-3000 Uribe 5973 (COL) Killip 1938; Hno. Daniel 1968; HolmNielsen 1974 LC Passiflora cuspidifolia Harms, 1893 and by cun snt 2000-3200 Prieto 302 (UIS) Holm-Nielsen et al. 1988 LC Tr V.N.: Granadillita de Monte (ant) Tr Passiflora dawei Killip, 1930 * and cun snt 900-1600 Idrobo 2037 (COL) VU/EN Passiflora erytrophylla Mast.,1872 * and by cun 1600-2790 Ocampo et al. 54 (HUA) Killip 1930, 1938; Hernández 2003 Killip 1938; Uribe 1955a Passiflora lyra Planch. & Lind. ex Killip, 1846 nr and ant 400-840 MacDougal 4161 (HUA) Killip 1938 NT/VU Tr Passiflora magdalenae Triana & Planch., 1873 * and cl cun to 200-1200 Uribe 2568 (COL) Killip 1938; Pérez 1956 NT/VU Passiflora micropetala Mast., 1872 amz and 0-710 Perez-A.669 (COL) Holm-Nielsen et al. 1988 Passiflora mollis HBK., 1817 * nr and ama ant by cho cq met pu vc ant cl cho qu lg snt to vc Tr V.N.: Granadillo del Magdalena. Tr 1400-2500 Humboldt & Bonpland (P) - Type Killip 1938; Hno. Daniel 1968 47 EN LC LC/NT Tr Tr Not collected since 1938. Tr Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora monadelpha Jørgensen & Holm-Nielsen, 1987 nr and to vc 2800-3310 Escobar 4859 (HUA) Holm-Nielsen et al. 1988; Hernández 2003 VU/EN Tr Passiflora munchiquensis.Hernández (ined), 2003 * nr and cau vc 1900-3200 Vargas 3909 (HUA) NT/VU Tr Passiflora occidentalis Hernández (ined), 2003 * nr and pac cau cho na pu vc 50-1200 Killip 39025 (COL) LC/NT Tr Passiflora panamensis Killip, 1922 pac car ant bl cho cor 0-500 Zarucchi et al. 5107 (CHOCO) Hernández 2003; A.Hernández (com. pers). Hernández 2003; A.Hernández (com. pers). Killip 1938 Passiflora pilosissima Killip, 1931 * nr and ant vc 1500-2100 Lehmann 7630 (US) Passiflora popayanensis Killip, 1930 * and cau 2400-2900 Passiflora punctata L., 1753 and pac cau cun na vc [Passiflora sandrae J. MacDougal, 2006] pac Passiflora tribolophylla Harms, 1922 * NT Tr V.N.: Gulupa (ant) Killip 1938 CR Tr Lozano 6472 (COL) Killip 1938 VU/EN Tr 20-1750 Romero-C. 3150 (COL) Croat 1978; Holm-Nielsen et al. 1988 LC/NT Tr cho 800-1100 Garwood 1178 (MO) - Type MacDougal 2006 pac ant cau cho 50-1820 Lehmann 5420 (foto, COL) LC/NT Passiflora tuberosa Jacq., 1804 nr and vc 1200 Cuatrecasas 15930 (VALLE) Killip 1938; Hno. Daniel 1968 Killip 1938 Tr Collected in the border of Panama and Colombia (cho) Tr EN Tr Passiflora ursina Killip & Cuatrec., 1960 and ant na vc 2100-3100 Roldán 2345 (HUA) VU/EN Tr Passiflora vespertilio L.,1753 amz and ori ama met na 150-500 Plowman 2425 (COL) Killip 1960; Holm-Nielsen et al. 1988; Hernández 2003 Holm-Nielsen et al. 1988 LC/NT Tr 48 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Section Hahniopathanthus Passiflora guatemalensis S. Watson, 1887 and car ant cl lg ma qu ri to vc 0-1580 Uribe 2532 (COL) Passiflora hahnii (Fourn.) Mast., 1872 nr and car ant cl lg ma to vc 100-1250 Killip & Hazen 8670 (Y) Section Pseudodysosmia Passiflora adenopoda Moc. & Sessé ex DC., 1828 and car ant by cl cun cau ma qu ri to vc 100-2100 Cuatrecasas 15703 (VALLE) Passiflora lobata (Killip) Hutch. ex J.M. MacDougal,1986 pac ant cho 0-1200 Gentry 23791 (COL) Passiflora morifolia Mast., 1872 and na 500-1000 Karsten s.n. (W) n.v. Section Pseudogranadilla Passiflora bicornis Mill., 1768 car ant at bl lg ma 0-500 Saravia 3643 (COL) Passiflora hirtiflora Jørgensen & Holm-Nielsen, 1987 nr and ns 2650 Escobar 3152 (HUA) Passiflora kalbreyeri Mast., 1883 * and car ce ns snt 1100-3100 Killip 20284 (COL) Killip 1938 LC/NT Tr Passiflora menispermacea Triana & Planch., 1873 * and to 1400-3000 Cuatrecasas 9247 (MA) Killip 1938 LC Tr Section Xerogona 49 LC Tr CR Tr V.N.: Granadilla Abroquelada (ant) Holm-Nielsen et al. 1988; MacDougal 1994 LC MacDougal 1994; Ulmer & MacDougal 2004 Killip 1938; MacDougal 1994. NT Tr V.N: Pegajosa (qu), Granadilla Culebra (vc), Gulupo (cun). Edible fruit Tr EN Tr Killip1938; Holm-Nielsen 1974 LC Tr V.N.: Cachito de Venado (bl), Cinco Llagas (at) Tr Killip 1938; Ulmer & MacDougal 2004 Uribe 1955b; Holm-Nielsen 1974; Ulmer & MacDougal 2004 CR Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora capsularis L., 1753 and car pac ant cl cun cho hu lg ma na ns qu snt to vc 100-2000 Uribe 2566 (COL) Passiflora costaricensis Killip, 1922 pac cho 20-1500 Croat 42591 (HUA) Passiflora escobariana J.M. MacDougal, 1992 and ant 1090-1100 MacDougal 3823 (HUA) Isotype Passiflora rubra L., 1753 and car ant cl cau cun hu lg pu na pu qu ri to vc 500-2000 Garcia-B. 17279 (COL) Subgenus Dysosmia (DC.) Killip, 1938 Passiflora foetida var. eliasiiKillip, 1938 car at bl ma 0-500 Penell 12029 (N) amz and car ori pac ama ant ara at bl by cau ce cor cq cs cun cho gn gv hu lg ma met na ns qu snt suc to va vc 0-1500 Schultes 22576 (COL) Passiflora foetida var. gossypiifolia (Desv.) Mast. 1872 50 LC Tr NT Tr VU Tr LC Tr V.N.: Chulupa de Monte (cl) Kiliip 1938 VU Killip 1938; Martin & Nakasone 1970; Romero-C. 1991; Ulmer & MacDougal 2004; Ulmer & Ulmer, 2005 LC Tr V.N.: Flor de la Pasión, Pasionaria (at) Tr V.N.: Granadilla (cho), Flor de la Pasión (at), Gulupo (cun), Bejuco Canastilla (met), Chulupa de Loma (ant hu), Cinco Llagas (cor). I.N.: Iñanaleeg murulale (ama). Edible fruit Killip 1938; Holm-Nielsen 1974; HolmNielsen et al. 1988 Holm-Nielsen 1974; HolmNielsen et al. 1988 MacDougal 1992; Ulmer & MacDougal 2004 Holm-Nielsen et al. 1988 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora foetida var. hispida (DC.) Killip ex Gleason, 1931 and car ant bl cun ns to 0-1500 Killip & Smith 21000 (N) Killip 1938; Ulmer & Ulmer, 2005 LC Passiflora foetida var. isthmina Killip, 1938 and pac na snt vc 0-1200 Killip 5289 (N) Killip 1938 VU Passiflora foetida var. moritziana (Planch.) Killip ex Pull, 1937 car ma 0-500 Killip & Smith 21088 (N) Killip 1938 VU Tr V.N.: Flor de la Pasión (ma) Passiflora foetida var. sanctae-martae Killip, 1938 * nr car ma 0-500 Smith 1532 (P) Kiliip 1938 EN Tr Flor de la Pasión (ma) Passiflora vestita Killip, 1938 amz pu 0-500 Betancourt 5164 (MO) n.v. Killip 1938; Holm-Nielsen et al. 1988 Distephana (Juss.) Killip, 1938 Passiflora coccinea Aubl.,1775 amz ori ama cs gn gv met na va vch 150-1500 Davidse 5321 (COL) Escobar 1988a LC Passiflora involucrata (Mast) A.H. Gentry, 1981 amz ama cq va 150-350 Schultes 6923 (COL) Escobar 1988a LC Tr V.N.: Lluvia Padie, Granadillo de Conga (ama), Granadilla colorada (cs). Edible fruit Tr Passiflora glandulosa Cav., 1790 amz va 150-500 Romero-C. 3668 (AAU) n.v. Killip 1938; Holm-Nielsen 1974 EN Tr Passiflora quadriglandulosa Rodschied, 1796 amz ama gu 150-500 Lozano 604 (COL) Escobar 1988a; Holm-Nielsen et al. 1988 LC/NT Tr 51 VU/EN Tr V.N: flor de la pasión (ma), gulupo (cun) Tr V.N.: Flor de la Pasión (vc) Tr Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora variolata Poepp. & Endl., 1838 amz ama cq va 150-500 Zarucchi 2197 (COL) Escobar 1988a LC/NT Passiflora vitifolia Kunth, 1817 amz and car ori pac ama ant bl by cau ce cl cho cor cq cun lg gv ma met na pu ri snt to va vc vch 0-1800 Cuatrecasas 15740 (VALLE) Killip 1938; Romero C. 1956, 1991; Martin & Nakasone 1970; Holm-Nielsen 1974; HolmNielsen et al. 1988 LC and by cau cl cun na ns qu snt to vc 1400-2700 Richter s.n. (COL) Jussieu 1805; Holm-Nielsen 1974; Escobar 1988a LC Tr V.N.: Tacso (na), Curubo de Monte (qu ns). amz and pac ama cho ns ant 0-1000 Renteria 3542 (COL) Killip 1938; Holm-Nielsen 1974; HolmNielsen et al. 1988; RomeroC. 1991 LC Tr V.N.: Cocorilla (cho). Granadilla, Naracujinha (ama). N.I.: CipoCipo Naracujinha (ama). Subgenus Manicata (Harms) Escobar, 1988 (Syn. Granadillastrum) Passiflora manicata (Juss.) Pers., 1807 Tr V.N.: Granadilla, Oncilla, Parcha de Culebra de Agua (ama) Tr V.N.: Chulupo (cq), Granadilla de Monte (cho), Granadillo (met cq), Gulupa (to). I.N.: Maloca de Fisi (ama). Edible fruit Passiflora (Medik.) Mast., 1871 (Syn. Granadilla) Series Digitatae Passiflora serratodigitata L.,1753 Series Laurifoliae 52 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora ambigua Hemsl. ex Hook., 1902 amz and ori pac ant by cl cho cun hu ma met pu snt vc 0-2000 Fuchs 21744 (COL) Holm-Nielsen et al. 1988 LC Tr Edible fruit Passiflora gleasonii Killip, 1924 ori gn 150-500 Madriñán 1014 (COL) Killip 1938 EN Tr Passiflora guazumaefolia Juss., 1805 and car ce cor bl ma snt 0-500 Uribe 2405 (COL) Killip 1938; Coppens 2003 Passiflora killipiana Cuatrecasas, 1960 nr amz cq 250-500 Schultes 5875 (US) Killip 1960 CR Passiflora laurifolia L., 1753 and amz pac 0-1700 Zarucchi 1824 (COL) Killip 1938 LC Tr Edible fruit Passiflora nitida Kunth, 1817 amz and car ori pac ama cho cq gv hu met snt va ama ant cho cq cun cs gn gv ma met na pu va vc 0-1940 Triana 2931 (P) Killip 1938; Romero-C. 1956, 1991; Holm-Nielsen 1974; García-B. 1975; Croat 1978 LC [Passiflora phellos C. Feuillet, 2004] amz 90-150 Wurdack & Addeley 43479 (NY) - Holotype Feuillet 2004 Tr V.N.: Granadilla (cho met), Granadilla Babosa (na). N.I.: Burucuña, Gemarundare, Tuchica, JinoGojé (va). Edible fruit Tr Reported in the Amazon of Brazil, Peru and Venezuela 53 LC/NT Tr V.N: La Parcha (ce), Cocorilla (ma). Edible fruit Tr Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora popenovii Killip, 1922 nr and cau na vc 1200-2050 Escobar & Escobar 1017 (HUA) Killip 1938; Holm-Nielsen et al. 1988; Romero-C. 1991; Ulmer & MacDougal 2004 Passiflora riparia Mart. ex Mast., 1872 amz cq pu va 300-400 Smith 3157 (US) Killip 1960 LC/NT Passiflora tolimana Harms, 1894 * and ant to vc 820-2000 Echeverry 3627 (TOLI) Killip 1938 NT/VU Series Incarnatae Passiflora cincinnata Mast., 1868 nr and ns 1200 Killip & Smith 20879 (Y) Killip 1938 CR Passiflora edulis f. edulis Sims, 1818 amz and pac ant cl cau cho cun gv met na qu ri snt to vch vc 1100-2750 Idrobo 1637 (COL) Holm-Nielsen et al. 1988; Vanderplank 2000; Ulmer & MacDougal 2004 NE Passiflora edulis f. flavicarpa Degener, 1932 nr amz and car ori pac ant ara bl ce cl cho cun gn hu met pu qu snt ri to vc 0-1800 Silvestone 14399 (CUVC) Killip 1938; Ulmer & MacDougal 2004 NE Series Kermesinae 54 EW Tr V.N: Granadilla de Quijos (na), granadilla caucana, curubejo (cau). Cultivated. Edible fruit Tr Edible fruit Tr Edible fruit Tr Ornamental (qu). Edible fruit Tr Introduced from Brazil in the 1950s. V.N: Curuba Redonda (ant cl ri qu), Gulupa (cun). Cultivated or feral. Edible fruit Tr Introduced from Brazil in the 50s. V.N.: Maracuyá. Cultivated. Edible fruit Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora lehmanni Mast., 1885 * amz and Passiflora smithii Killip, 1930 and car Passiflora trisulca Mast., 1887 * Killip 1938; Holm-Nielsen 1974 Killip 1938; Holm-Nielsen 1974 LC Tr LC Tr V.N.: Curuba Silvestre (to). Marulanda 91 (HUA) Killip 1938; Hno. Daniel 1968 NT Tr 1000-2700 Ocampo 83 (VALLE) Deginani 2001 2450-2500 Cuatrecasas 1808 (COL) Killip 1938 2600 Holm-Nielsen et al. 6200 (AAU) Holm-Nielsen et al. 1988 450 Uribe 1334 (US) ant cau cl cun hu pu qu ri snt vc cun ma qu snt to vc 1000-2000 Uribe 2588 (COL) 500-2000 Killip & Smithii 15015 (MO) Holotype and ant cl vc 1300-1800 Series Lobatae Passiflora caerulea L., 1753 nr and cl cun qu Passiflora gritensis H. Karst., 1859 and by ns [Passiflora montana Holm-Nielsen & Lawesson, 1987] and Passiflora picturata Ker, 1822 nr LC/NT Tr Introduced from Argentina. Ornamental. Edible fruit Tr Killip 1938, 1960 DD Tr Collected on the border of Ecuador and Colombia (na) Tr Passiflora pennellii Killip, 1924 * and ant cun 1200-1600 Uribe 4827 (COL) Killip 1938 NT/VU Tr Passiflora resticulata Mast. & André, 1884 amz and pac cau gv na vc 0-2000 Marulanda & Márquez 1665 (HUA) NT/VU Tr Passiflora semiciliosa Planch & Linden, 1873 * and car ma ns 1850-3000 Garcia-B. 20749 (COL) Killip 1938; Holm-Nielsen 1974;HolmNielsen et al. 1988 Killip 1938 55 VU Tr V.N.: Gulupa, Palcha (ns) Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora subpeltata Ortega, 1798 amz and car ori ant bl cau ce cor cq cs cun ma suc to vc 0-2400 Ramírez et al. 11507 (CAUP) Killip 1938; Holm-Nielsen 1974 LC Tr V.N.: Cocorilla (bl) Series Quadrangulares Passiflora alata Curtis, 1788 nr amz ama 200 Ocampo 82 (VALLE) Killip 1938; Ulmer & MacDougal 2004 DD amz and car ori pac ama ant bl cau cl cho cq cun gn hu ma met na ns qu va ri snt to vc 0-1500 Gentry 15371 (COL) Killip 1938; Romero-C. 1956, 1991; HolmNielsen1974; Holm-Nielsen et al. 1988 LC Tr Introduced (qu vc) from Brazil in the 90s. V.N.: Maracúa. Cutivated. Edible fruit Tr V.N.: Badea (ant cl hu cun met qu ri), corvejo (snt), Granadillo Grande (cau), curuba (vc), Motorro (gn). Cultivated. Edible fruit Series Menispermifoliae Passiflora chocoensis G. Gerlach & T. Ulmer, 2000 * nr pac cho 0-100 Gerlach 434917 (COL) Holotype CR Tr Passiflora menispermifolia Kunth, 1817 amz car and pac ant bl by cho cor cq cun met na ns snt to vc 0-2140 Cuatrecasas 15541 (VALLE) Gerlach & Ulmer, 2000; Ulmer & MacDougal 2004 Croat 1978; Holm-Nielsen et al. 1988 LC Tr V.N.: Chulupe (cq) Series Simplicifoliae Passiflora danielii Killip, 1960 * and ant 1300-2600 Passiflora longipes Juss., 1805 * and cun by qu snt to 2500-3500 Hno. Daniel 1536 (MEDEL) Isotype Sanchez 17 (COL) Passiflora quadrangularis L.,1759 56 Killip 1960; Hno. Daniel 1968 VU/EN Tr Killip 1938 NT Tr Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora longipes var. oxyphylla L. Uribe, 1977 * and by ns snt 2000-2600 Cadena 83 (UIS) Uribe 1977 NT Tr Passiflora oerstedii Mast., 1872 and ori pac ant cau cho cun cau met na qu ri vc 0-2000 Romero-C. 6141 (COL) Killip 1938; Holm-Nielsen 1974; HolmNielsen et al. 1988 LC Tr Series Tiliaefoliae Passiflora ligularis f. lobata (Mast.) Killip, 1938 nr and ant 1800-2000 Archer 1498 (COL) Killip 1938 NT Passiflora ligularis Juss., 1805 and ant cl cun cau by cho hu met na ns pu qu ri snt to vc 1550-2500 Dombey 739 (P) - Type Killip 1938; Romero-C. 1956, 1991; Holm-Nielsen et al. 1988 LC Passiflora magnifica L.K. Escobar, 1990 * and ant 1250-1750 Callejas 6586 (HUA) n.v. VU Passiflora maliformis L.,1753 and car pac ant by cl cau cun cho hu ma na qu snt to vc 0-2200 Humboldt & Bonpland 1804 (P) - Type Escobar 1990; Ulmer & MacDougal 2004 Killip 1938; Romero-C, 1956, 1991; Holm-Nielsen, 1974; García-B. 1975; HolmNielsen et al. 1988 Tr V.N: Granadilla. Cultivated. Edible fruit Tr V.N: Granadilla; Granadilla Pipo (na). N.I.: Awapit (na). Cultivated. Edible fruit Tr Edible fruit Passiflora multiformis Jacq.,1809 and car lg ma ns 0-1300 Romero-C. 8992 (COL) 57 Killip 1938 LC NT/VU Tr V.N: Gulupa, Granadilla de Piedra, o de Hueso (cu, na vc), Gurapa (snt), Chulupa (hu). Cultivated. Edible fruit Tr V.N: Palchita (ns). Edible fruit Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ LC Passiflora palenquensis Holm-Niels. & Lawesson, 1987 pac ant cau cho na vc 0-1200 Espina & Garcia 1951 (COL) Holm-Nielsen & Lawesson 1987; Holm-Nielsen et al. 1988 Passiflora platyloba Killip, 1922 pac cho 0-1050 Gentry & Juncosa 40946 (COL) Gentry 1976 NT/VU Passiflora seemannii Griseb.,1858 amz and pac ori ant by cho cun gn gv met ns snt va vc vch 0-1300 MacDougal 4144 (HUA) Croat 1978 LC Passiflora serrulata Jacq., 1767 car at ma lg 0-500 Bunch 601 (FMB) Killip 1938; Coppens 2003 NT/VU Passiflora tiliifolia L., 1753 and pac ant cau cho cl na qu vc to 1100-2500 González 1411 (CAUP) Killip 1938; Holm-Nielsen 1974; HolmNielsen et al.1988; Coppens 2003 LC/NT Subgenus Porphyropathanthus L.K Escobar, 1989 Passiflora sierrae L.K. Escobar, 1989 * car ma 3000-3700 Cuatrecasas 24375 (COL) Escobar 1989 EN/CR Tr Subgenus Psilanthus (DC.) Killip, 1938 Passiflora bicuspidata (H.Karst.) Mast.,1872 * and by cun ns snt 2500-3500 Rojas 138 (CDMB) Uribe 1972; Killip 1978 VU Tr 58 Tr V.N.:ranadilla (cho), "Camelo" (vc). Cultivated. Edible fruit Tr Edible fruit Tr V.N.: Palcha, Chulupa (met), Granadilla Montañera (cun). Edible fruit Tr V.N.: Guayabita Cimarrona (ma). Edible fruit Tr V.N.: Granadilla, Machimbi (Colombia). Cultivada. Fruto comestible. Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora hyacinthiflora Planch. & Linden, 1873 * and by ma ns 2900-3300 Garcia-B. 20700 (COL) Killip 1938 Passiflora trinervia (Juss.) Poir., 1811 * and cl qu to vc 2500-3700 Cuatrecasas 20241 (VALLE) Subgenus Rathea (Karst.) Killip, 1938 Passiflora andina Killip, 1938 nr and na 2800 Karsten (V) Passiflora colombiana L.K. Escobar, 1986 * and na pu 3000-3600 Mora 6175 (PSO) Paratype and cau 2300-2800 Tryon 6001 (COL) Passiflora cumbalensis var. cumbalensis (H. Karst.) Harms,1894 and na pu 3000-3800 Fernandez 5834 (COL) Passiflora cumbalensis var. goudotiana (Triana & Planch.) L.K. Escobar, 1987 and car ant by cl cq cun hu ma na pu qu ri snt to vc 1800-3300 Uribe 2593 (COL) [Passiflora sanctae-barbarae Holm-Nielsen & Jørgensen, 1987] and 2200-2700 Harling & Andersson 12445 (AAU) Isotype LC/NT Tr Jussieu 1805; Killip 1938 VU Tr Killip 1938; Holm-Nielsen et al.1988 Escobar 1986, 1988 CR Tr CR Tr Escobar 1987, 1988b LC Romero-C. 1956; HolmNielsen 1974; Escobar 1987, 1988; HolmNielsen et al. 1988 Escobar 1987, 1988; HolmNielsen et al. 1988 LC/NT Tr V.N.: Curuba de Monte. Edible fruit Tr V.N.: Curuba Roja, Tauso (na). Edible fruit Subgenus Tacsonia (Juss.) Tr. & Planch, 1873 Section Bracteogama Passiflora cumbalensis var. caucana L.K. Escobar, 1987 * 59 Holm-Nielsen et al. 1988 LC Tr V.N.: Curuba bogotana (cun), Curubo mucura, curuba rosada, Tausa (na). Cultivated. Edible fruit Tr Reported in the northern Andes of Ecuador Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora tripartita var. azuayensis Holm-Nielsen & Jørgensen, 1988 nr and car ant by cun ma ns 2000-2610 Escobar 19999 (HUA) Holm-Nielsen et al. 1988 LC/NT Tr V.N.: Curuba. Edible fruit (by cun) Passiflora tripartita var. mollissima Holm-Nielsen & Jørgensen, 1988 and car ant by cau cl cun ma na ns pu snt vc 2200-3500 Romero-C 8007 (PSO) Holm-Nielsen et al. 1988; Romero-C. 1991; Ulmer & MacDougal 2004 LC Passiflora tarminiana Coppens & Barney, 2001 and ant by cau cl cun hu na qu snt to vc 2000-2900 Coppens 72 (COL) - Type Coppens et al. 2001; Campos 2001 LC Tr V.N.: Curuba de Castilla (ant by cu cl); Tauxso (na). Cultivated. Edible fruit Tr V.N.: Curuba India. Cultivated. Edible fruit. Section Colombiana Series Colombianae Passiflora adulterina L.f.,1781 * and by cun snt to 2600-3600 Barclay 4517 (COL) Escobar 1988a NT Tr Passiflora crispolanata L.Uribe, 1954 * and by cun 2500-3500 Uribe 6773 (COL) Uribe 1954; Escobar 1988a NT Passiflora cuatrecasasii Killip, 1960 * and by cun met snt 2200-3500 Killip 1960; Escobar 1988a VU Passiflora formosa T. Ulmer, 1999 * and by 3000-3100 Cuatrecasas 9479 (foto, MEDEL) Uribe 5945 (COL) Tr V.N.: Curuba Paramera (cun) Tr Ulmer 1999 EN Tr Passiflora lanata (Juss.) Poir., 1811 * and cun by snt to 2200-3500 Uribe 2587 (COL) Jussieu 1805; Holm-Nielsen 1974; Escobar 1988a NT/VU Tr V.N.: Granadilla (cun) Passiflora pamplonensis Planch.& Linden ex Triana & Planch., 1873 * and snt 2000-3000 Funck & Schlim 1385 (foto, VALLE) Escobar 1988a EN/CR Tr Curubita de Piñuela (snt) 60 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora rigidifolia Killip, 1960 * nr and ant 3750 Burke 185 (K) Type Killip 1960 Passiflora rugosa var. rugosa (Mast.) Triana & Planch., 1873 and cun met ns 3000-3500 Peñuela 008 (COL) Escobar 1988a LC/NT Tr Passiflora rugosa var. venezolana L.K. Escobar, 1986 and ns snt 2500-3500 Garcia-B. 20001 (COL) Escobar 1988a LC/NT Tr Passiflora trianae Killip, 1938 * and ns snt 3000-3500 Escobar 569 (COL) Escobar 1988a VU/EN Tr Passiflora truxillensis Planch. & Linden, 1873 nr and ns 1800-3000 V. Barney & G. Coppens (foto), com. Personal Escobar 1988a; Ulmer & Ulmer 2005 EN Tr Series Leptomischae Passiflora antioquiensis H. Karst., 1859 * and ant cau cl cun hu pu qu ri to vc 1800-2700 Escobar 2133 (HUA) Hno. Daniel 1968; Uribe 1972; Garcia- B. 1975; Escobar 1988a LC/NT Passiflora cremastantha Harms, 1922 * and cau 2000-2500 Lehmann 5421 (F) - Type Escobar 1988a EX Passiflora flexipes Triana & Planch., 1873 * and cl qu ri 2500-3380 Vargas 626 (FAUC) Escobar 1988a NT/VU Passiflora leptomischa Harms, 1922 * and ant cau qu vc 2000-2800 Escobar et al. 4421 (PSO) Escobar 1988a LC/NT Passiflora tenerifensis L.K. Escobar, 1988 * and vc 2800-3100 Escobar 4853 (COL) Escobar 1988a, 1989b; Campos 2001 EN/CR 61 DD Tr Known only from the type. Tr V.N.: Granadilla (vc), Curuba Antioqueña (ant). Wild or cultivated in home gardens. Edible fruit Tr Known only from the type. Tr V.N.: Curuba de Monte (cl qu ri). Edible fruit Tr Edible fruit Tr V.N.: Curuba de Monte (vc). Edible fruit Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Series Quindiensae Passiflora linearistipula L.K. Escobar, 1988 * and cl 2650-3170 Ocampo et al. 56 (HUA) Escobar 1988a EN/CR Tr Not collected since 1984. Passiflora quindiensis Killip, 1938 * and to 2900-3100 Uribe 3320 (COL) Escobar 1988; Campos 2001 VU/EN Tr Section Fimbriatistipula Passiflora fimbriatistipula Harms, 1894 * and cau hu 2130-3240 Fernandez et al. 30182 (AFP) Escobar 1988a NT/VU Passiflora uribei L.K. Escobar, 1988 * and na pu 2500-2700 Escobar et al. 2896 (HUA) Uribe 1958; Escobar 1988a EN Tr I.N.: Pachuaca (cau) Tr Section Parritana Passiflora jardinensis L.K. Escobar, 1988 * and ant 2750-3000 Zarucchi 6963 (COL) Escobar 1988b VU/EN Tr Passiflora parritae (Mast.) L.H. Bailey, 1916 * and cl qu ri to 2500-3020 Sánchez 15 (FAUC) Escobar 1988a VU/EN Tr Curuba de Monte (to). Edible fruit Section Poggendorffia Passiflora pinnatistipula Cav.,1799 and ant by cun na ns 2000-3600 Uribe 6643 (COL) Escobar 1988b; Holm-Nielsen 1974; HolmNielsen et al. 1988; Campos 2001 LC/NT Tr V.N.: Curuba Redonda, Gulupa (cun) Cultivated. Edible fruit Passiflora x rosea (H. Karst.) Killip, 1938 and by cun 2500-3500 Uribe 3941 (COL) Escobar 1988a VU Tr Natural hybrid of P. pinnatistipula x P. tripartita var. mollissima. Edible fruit (when fertile) Section Tacsonia 62 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora mixta L. f., 1781 and ant cl by cau cun na ns qu ri snt to vc 1700-3700 Humboldt & Bonpland (P) - Type Escobar 1988a; Holm-Nielsen 1974; HolmNielsen et al. 1988; García-B. 1975 LC Passiflora schlimiana Triana & Planch., 1873 * car ce lg ma 2400-3220 Romero-C. 7407 (COL) Holm-Nielsen 1974; Escobar 1988a; RomeroC. 1991; Coppens 2003 VU/EN Section Tacsoniopsis Passiflora bracteosa Planch. & Linden, 1873 and ns snt 2200-3000 Garcia-B. 20745 (COL) Escobar 1988a EN Passiflora purdiei Killip, 1938 * and cun ma Purdie s.n. (K) n.v. Escobar 1988a EX Subgenus Tryphostemmatoides (Harms) Killip), 1938 Passiflora tryphostemmatoides Harms, 1894 and ant cau hu qu ri vc 1000-2700 Lehmann 5662 (K) - Isotype NT Tr Passiflora gracillima Killip, 1924 and ant cau cl hu na qu to 2000-3150 Penell 9393 (MO) - Isotype Killip 1938; Holm-Nielsen et al. 1988 Killip 1924, 1938 LC Passiflora arbelaezii L. Uribe, 1957 and pac ant cau cho cun na vc 0-2300 Roldán 1162 (COL) Uribe 1957 LC/NT Tr V.N.: Golondrina (cho) Tr Passiflora pacifica L.K. Escobar, 1988 * pac cho na vc 0-1800 Escobar 2143 (HUA) Escobar 1988b LC/NT Tr 63 Tr V.N.: Curuba de Monte (vc), Curubo de Páramo (cun), Palchuaca (cau), Curubito de Indio (cl). Edible fruit Tr V.N.: Curuba. Edible fruit Tr V.N.: Palchoaca (ns, snt) Tr Known only from the type. Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ II.1.6. Conclusions With 167 reported species, Colombia is the country with the highest Passifloraceae richness. This richness is concentrated in the Andean region, particularly in the departments of Antioquia, Valle del Cauca and Cundinamarca. Comparing with other countries indicates that the northern Andes of Colombia and Ecuador constitute the center of diversity for the most important genus, Passiflora. The low level of exploration in parts of the Andes, the Amazonian and the Orinoquian raises expectations that Colombia may still harbor many unknown species. Future studies should encompass new regions, including protected areas and conflict zones. Indeed, a better knowledge of this diversity, and its distribution, is urgent for the in situ conservation of this threatened richness, targeting the conservation of these resources as well as their habitat. Both aspects may even be combined if the genus Passiflora can be used as an indicator of biodiversity in the Andean region, as was the objective of a project in the coffee growing zone. Another important aspect is its direct valorization as a germplasm resource for crop diversification programs, implying the need for a better understanding of its morphological and genetic diversity. II.1.7 Acknowledgements The authors wish to thank the herbaria that provided specimens or collection data, and particularly Francisco J. Roldán (HUA) and Alexandra Hernández (COL), as well as Colciencias, the Colombian Ministry for Environment (MMA) and the Research Center of the Colombian Coffee Grower Federation (Cenicafé) for funding the collecting missions. The first author gratefully acknowledges financial support from the Gines-Mera Fellowship Foundation (CIAT-CBN). We are indebted also to José O. Velásquez (Casa Mutis, Mariquita), Hernando Criollo (U.Nariño), Mauricio Villegas (Cenicafé), Vicky Barney (Bioversity International), Alvinxon Castro (U.Chocó), Robinson Galindo (PNN Catatumbo) and Carolina Alcazar (Proselva) for assistance in obtaining plant data for this study. We are extremely grateful to Colombian farmers contacted in the fieldwork for 64 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ their continuous help and availability in localizing a great part of the observed plant material. II.1.8. Appendix 1. Synonymy = valid name Cieca auriculata M. Roemer, 1846 = Passiflora auriculata Kunth, 1817 Cieca coriacea (Juss) M. Roemer, 1846 = Pasiflora coriacea Juss., 1805 Cieca discolor M. Roemer, 1846 = Passiflora misera Kunth., 1817 Cieca pallida (L.) M. Roemer, 1846 = Passiflora pallida L., 1753 Decaloba alnifolia M. Roemer, 1846 = Pasiflora alnifolia Kunth, 1817 Decaloba biflora (Lam.) M. Roemer., 1846 = Pasiflora biflora Lam., 1789 Decaloba bogotensis (Benth.) M. Roemer, 1846 = Passiflora bogotensis Benth., 1845 Decaloba cuneata M. Roemer., 1846 = Passiflora cuneata Willd., 1809 Decaloba filipes M. Roemer, 1846 = Pasiflora filipes Benth, 1843 Decaloba holosericea M. Roemer, 1846 = Passiflora holosericea L., 1753 Decaloba jacquini M. Roemer, 1846 = Passiflora pulchella Kunth, 1817 Dilkea johannesii var. parvifolia Hoehne, 1915 = Dilkea johannesii Barb.Rodr., 1885 Dilkea acuminata Mast., 1871 = Dilkea retusa Mast., 1871 Dilkea magnifica Steyerm., 1968 = Dilkea retusa Mast., 1871 Dilkea wallisii Mast., 1872 = Dilkea retusa Mast., 1871 Disemma hahnii E. Fourn., 1869 = Passiflora hahnii Mast., 1872 Disemma hahnii Fourn., 1869 = Passiflora hahnii (Fourn.) Mast., 1872 Distephana cuneata M. Roemer, 1846 = Passiflora bicuspidata (H.Karst.) Mast., 1872 Distephana spinosa (Poeep. & Endl.) M. Roemer, 1835 = Passiflora spinosa (Poepp. & Ende.) Mast., 1871 Granadilla rubra Moench, 1802 = Passiflora rubra L., 1753 Grandilla vespertilio Moench, 1802 = Passiflora vespertilio L., 1753 Passiflora erubescens Triana & Planch., 1873 = Passiflora erytrophylla Mast., 1872 Passiflora velata Mast., 1872 = Passiflora serrulata Jacq., 1767 Passiflora williamsii Killip, 1922 = Passiflora platyloba var. williamsii (Killip) A.H. Gentry, 1976 Passiflora adenophylla Mast., 1872 = Passiflora subpeltata Ortega, 1798 Passiflora alba Link & Otto, 1798 = Passiflora subpeltata Ortega, 1798 Passiflora albicans L. Uribe, 1958 = Passiflora uribei L. K. Escobar, 1988 Passiflora angustifolia Swartz, 1788 = Passiflora suberosa L., 1753 Passiflora appendiculata G.F.W. Mey., 1818 = Pasiflora auriculata Kunth, 1817 Passiflora bauhinifolia Kunth, 1817 = Passiflora alnifolia Kunth, 1817 Passiflora bifurca Mast., 1873 = Passiflora cuneata Willd., 1809 Passiflora bilobata Vell., 1827 = Passiflora rubra L., 1735 Passiflora boyacana Killip, 1960 = Passiflora crispolanata L. Uribe, 1954 Passiflora capsularis var. geminifolia DC., 1828 = Passiflora sexflora Juss., 1805 Passiflora caucaense Holm-Niels., 1974 = Passiflora emarginata Humb. & Bonpl., 1813 Passiflora chilensis Miers, 1826 = Passiflora pinnatistipula Cav., 1799 Passiflora cisnana Harms, 1894 = Passiflora rubra L., 1753 Passiflora corumbaensis Barb., 1898 = Passiflora cincinnata Mast., 1868 Passiflora cualiflora Harms, 1906 = Passiflora citrifolia (Juss.) Mast., 1871 Passiflora difformis Kunth, 1817 = Passiflora coriaceae Juss., 1805 Passiflora digitata L., 1763 = Passiflora serratodigitata L., 1753 Passiflora elegans Triana & Planch., 1873 = Passiflora quindiensis Killip, 1938 Passiflora emiliae Sacco, 1966 = Passiflora ambigua Hemsl. ex Hook., 1902 Passiflora eminula Mast., 1883 = Passiflora costata Mast., 1872 Passiflora eriocaula Harms, 1922 = Passiflora rugosa (Mast.) Triana & Planch.var. rugosa,1873 Passiflora erosa Rusby, 1907 = Passiflora morifolia Mast., 1872 Passiflora erosa Rusby, 1906 = Passiflora morifolia Mast., 1872 Passiflora fulgens Wallis ex Morren, 1866 = Passiflora coccinea Aubl., 1775 Passiflora gigantifolia Harms, 1894 = Passiflora macrophylla Spruce ex Mast., 1883 Passiflora glauca Humb. & Bonpl., 1813 = Passiflora arborea Spreng., 1826 65 Chapter II. Biogeography and an updated list for conservation ________________________________________________________________________ Passiflora goudotiana Triana & Planch., 1873 = Passiflora cumbalensis (H. Karst.) Harms var. goudotiana (Triana & Planch.) L. K. Escobar, 1987 Passiflora heydei Killip, 1922 = Passiflora morifolia Mast., 1872 Passiflora hydrophila Barb Rodr., 1891 = Passiflora costata Mast., 1872 Passiflora incana Seemann ex Mast., 1883 = Passiflora seemanni Griseb., 1858 Passiflora inundata Ducke, 1925 = Passiflora costata Mast., 1872 Passiflora laticualis Killip, 1924 = Passiflora misera Kunth., 1817 Passiflora longipes var. retusa Triana & Planch., 1873 = Passiflora longipes Juss., 1805 Passiflora lorifera Mast, & André, 1883 = Passiflora macrophylla Spruce ex Mast., 1883 Passiflora lunata J.E. Smith., 1790 = Passiflora biflora Lam., 1879 Passiflora macrocaropa Mast., 1869 = Passiflora quadrangularis l., 1759 Passiflora micrantha Killip, 1938 = Passiflora erythrophylla Mast., 1872 Passiflora miraflorensis Killip, 1924 = Passiflora sexflora Juss., 1805 Passiflora mollis var. integrifolia Planch. ex Mast., 1872 = Passiflora cuspidifolia Harms, 1893 Passiflora nympheoides Karst., 1859 = Passiflora nitida Kunth, 1817 Passiflora oblongifolia Pulle, 1906 = Passiflora laurifolia L., 1753 Passiflora ocanensis Planch. & Linden, 1873 = Passiflora lindeniana Planch. ex Triana & Planch., 1873 Passiflora ornata Kunth, 1817 = Passiflora maliformis L., 1753 Passiflora pala Planch. & Linden, 1873 = Passiflora bogotensis Benth., 1845 Passiflora paraguayensis Chad., 1899 = Passiflora capsularis L., 1753 Passiflora pennipes Sm., 1819 = Passiflora pinnatistipula Cav., 1799 Passiflora praeacuta Mast., 1887 = Passiflora oerstedii Mast., 1872 Passiflora pubera Planch. & Linden, 1873 = Passiflora sphaerocarpa Triana & Planch., 1873 Passiflora pulchella Kunth,1817 = Passiflora bicornis Mill., 1768 Passiflora quadriglandulosa var. involucrata (Mast.) Killip, 1938 = Passiflora involucrata (Mast.) A.H. Gentry, 1981 Passiflora reticulata Sauv., 1873 = Passiflora holosericea L.,1753 Passiflora salmonea Harms, 1894 = Passiflora parritae (Mast.) Bailey, 1916 Passiflora sanguinea J.E. Smithi, 1819 = Passiflora vitifolia Kunth, 1817 Passiflora schultzei Harms, 1929 = Passiflora arborea Spreng., 1826 Passiflora spherocarpa var. pilosula Mast., 1883 = Passiflora pubera Planch. & Linden, 1873 Passiflora stipulata Aubl., 1858 = Passiflora subpeltata Ortega, 1798 Passiflora suberosa var. pallida (L.) Mast. = Passiflora pallida L., 1753 Passiflora tomentosa Lam. var. mollissima Triana & Planch., 1873 = Passiflora mollissima (Kunth) L.H. Bailey, 1916 Passiflora trisecta Planch. & Linden ex Triana & Planch., 1873 = Passiflora trianae Killip, 1938 Passiflora Van-Volxemii Triana & Planch., 1893 = Passiflora antioquiensis Karst., 1859 Passiflora var. cuellensis Goudot ex Triana & Planch., 1873 = Passiflora menispermifolia Kunth, 1817 Passiflora vesicaria L., 1753 = Passiflora foetida L., 1753 Passiflora vitifolia var. involucrata Mast., 1872 = Passiflora involucrata (Mast.) A.H. Gentry, 1981 Passiflora weberiana André, 1885 = Passiflora morifolia Mast., 1872 Passiiflora acerifolia Schlecht. & Cham., 1830 = Passiflora adenopoda Moc. & Sessé ex DC., 1828 Rathea floribunda Karst., 1859 = Passiflora andina Killip, 1938 Tacsonia adulterina Juss., 1805 = Passiflora adulterina L. f., 1781 Tacsonia bicuspidata H. Karst., 1859 = Passiflora bicuspidata (H. Karst.) Mast., 1872 Tacsonia cumbalensis H. Karst., 1859 = Passiflora cumbalensis var. cumbalensis (H. Karst.) Harms, 1894 Tacsonia cuneata Benth, 1845 = Passiflora bicuspidata (H. Karst.) Mast., 1872 Tacsonia flexipes (Triana & Planch) Mast., 1883 = Passiflora flexipes Triana & Planch., 1873 Tacsonia glandulosa Juss., 1805 = Passiflora glandulosa Cav.,1790 Tacsonia infundibularis Mast., 1883 = Passiflora bracteosa Planch. & Linden, 1873 Tacsonia lanata Juss., 1811= Passiflora lanata (Juss.) Poir., 1811 Tacsonia mixta (L.f.) Juss., 1805 = Passiflora mixta L.f., 1781 Tacsonia mollissima Kunth var. glabrescens Mast.,1872 = Passiflora mollissima (Kunth) L.H. Bailey, 1916 Tacsonia mollissima Kunth, 1817 = Passiflora mollissima (Kunth) L.H. Bailey, 1916 Tacsonia parritae Mast., 1882 = Passiflora parritae (Mast.) L.H. Bailey, 1916 Tacsonia pinnatistipula var. pennipes (Sm.) DC., 1828 = Passiflora pinnatistipula Cav., 1799 Tacsonia pinnatistipula (Cav.) Juss., 1805 = Passiflora pinnatistipula Cav.,1799 Tacsonia quadriglandulosa (Rodschied) DC., 1828 = Passiflora quadriglandulosa Rodschied, 1796 Tacsonia rosea (H. Karst.) Sodiro, 1903 = Passiflora x rosea (H. Karst.) Killip, 1938 Tacsonia rugosa Mast., 1872 = Passiflora rugosa (Mast.) Triana & Planch, 1873 var. rugosa Tacsonia spinescens Klotsch in Schomb., 1848 = Passiflora securiclata Mast., 1893 Tacsonia spinosa Poepp. & Endl., 1835 = Passiflora spinosa (Poepp. & Ende.) Mast., 1871 Tacsonia trinervia Juss., 1805 = Passiflora trinervia (Juss.) Poir., 1811 Tetrastylis lobata Killip, 1926 = Passiflora lobata (Killip) Hutch. ex J.M. MacDougal, 1986 66 CHAPTER III ____________________________________ Distribution, diversity and in situ conservation of Colombian Passifloraceae Chapter III. Diversity and in situ conservation _________________________________________________________________________ III.1. Distribution, diversity and in situ conservation of Colombian Passifloraceae John Ocampo Pérez1, Geo Coppens d'Eeckenbrugge2, Mike Salazar1 and Andy Jarvis1, 3. 1 Bioversity International (formerly IPGRI), Regional Office for the Americas, A.A. 6713, Cali, Colombia. 2 CIRAD/FLHOR, UPR ‘Gestion des ressources génétiques et dynamiques sociales’, Campus CNRS/Cefe, 1919 route de Mende, 34293 Montpellier, France. 3 International Center for Tropical Agriculture (CIAT), A.A. 6713, Cali, Colombia. III.1.1 Abstract Analysis was made of 3,930 records of 165 wild Passifloraceae to assess the distribution of their diversity in Colombia, identify collection gaps, and explore their potential as indicator species. Despite variable collecting density among and within biogeographic regions, the Andean region clearly presents a higher species richness, particularly in the central coffee growing zone and the departments of Antioquia, Cundinamarca and Valle del Cauca. The elevational distribution of diversity shows a small peak below 500 m, and two higher ones between 1000-2,000 and 2,500-3,000 m. This pattern corresponds to divergent adaptive trends among genera and infrageneric divisions. The analysis on 19 climatic variables showed that the two principal variance components, explaining 77% of the total, are respectively associated with temperature and precipitation, without influence of seasonality. Distribution parameters allow recognizing more than 36 narrow endemics. Prediction of species distribution showed nine areas with very high richness (predicted sympatry of 41 to 54 species) in the Andean region, three of which correspond to collection gaps. Endemics were not particularly frequent there, so a prioritization of protected areas based on species richness would not favor their conservation. The sites with high Passiflora diversity are poorly represented in the current system of protected areas. Instead their striking correspondence with ecotopes of the coffee growing zone imposes a conservation strategy integrating agricultural and environmental management at the landscape level. Reciprocally, several traits of Passiflora species make them particularly suited as indicators for any effort of conservation or restoration in this region of importance for the country. Keywords: Andes, Colombia, coffee growing zone, distribution, geographic information systems, Passifloraceae diversity, endemism. 68 Chapter III. Diversity and in situ conservation _________________________________________________________________________ III.1.2. Introduction New World Passifloraceae include four tropical genera (Ancistrothyrsus, Dilkea, Mitostemma and Passiflora) and about 550 species (Ulmer & MacDougal, 2004). The largest genus of the family is Passiflora L., with ca. 525 species distributed in a wide range of habitats, essentially tropical, ranging from humid rain forests to semi-arid subtropics. Most of them are herbaceous or woody vines, usually climbing by tendrils, while a few are trees or shrubs. More than 80 species produce an edible fruit, the most interesting ones belonging to subgenera Passiflora and Tacsonia (Coppens d’Eeckenbrugge, 2003). Among them, is the maracuja, P. edulis Sims, and its yellow-fruited form, P. edulis f. flavicarpa Degener, with a world production estimated at more than 640.000 tons (Passionfruit, 2006). In Colombia, ten more species/forms are cultivated. The fruits of the purple maracuja (P. edulis f. edulis Sims.), the sweet granadilla (P. ligularis Juss.), the giant granadilla (P. quadrangularis L.), the stone granadilla (P. maliformis L.), the granadilla de Quijos (P. popenovii Killip), the sweet maracuja (P. alata Curtis), and the banana passion fruits, ‘curuba de Castilla’ (P. tripartita var. mollissima (Kunth) Holm-Nielsen & Jørgensen), ‘curuba quiteña’ (P. tarminiana Coppens & Barney) are common on the national or regional markets. Those of the ‘granadilla de clima frío’ (P. pinnatistipula Cav.), the ‘curuba roja’ or rosy passion fruit (P. cumbalensis Karst.), the ‘curuba antioqueña’ (P. antioquiensis (Karst) Harms.), P. nitida Kunth and P. palenquensis Holm-Niels. & Lawesson are also cultivated, at a very local scale. These and other Passiflora species may be used for other purposes, such as garden ornamentals or pharmaceuticals (Coppens d’Eeckenbrugge et al., 2001). Killip in 1938 classified the genus Passiflora in 22 subgenera, including 355 American species (Annex 1). This classification has been amended by Escobar (1988, 1989, 1994) who merged subgenera Tacsioniopsis and Tacsonia, and proposed a new subgenus, Porphyropathanthus, and MacDougal (1994), who reinstated the name Decaloba for Killip’s subgenus Plectostemma. More recently, Feuillet & MacDougal (2003; Annex 2) proposed a much deeper revision, reducing the number of subgenera to only four (Astrophea, Decaloba, Deidamioides and Passiflora), and downgrading most of Killip’s divisions to lower levels. This proposal has been partially justified by recent molecular data (Muschner et al., 2003; Yockteng, 2003, Annexe Ib; 69 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Yockteng & Nadot, 2004; Hansen et al., 2006, see Annex 4a-d), however it still needs to be more clearly supported by further studies. Colombia’s geographic location and its variety of ecosystems place it as the second most biodiverse country in the world (McNeely et al., 1990). It is divided into five main biogeographic regions: Amazonian, Andean, Caribbean, Orinoquian, and Pacific (Hernández et al., 1991). The Andean region presents a highly diverse topography (1005,400 m), with three long mountain ranges, the Eastern, Central and Western Cordilleras, which separate two main inter-Andean valleys from the other regions. The uplift of the Andes created new habitats and increased local isolation, favoring high speciation rates in many taxa (Gentry, 1986). The continuously humid climate of the Amazonian and Orinoquian lowlands and the extremely wet climate of the Pacific region contrast with the dryer and more seasonal climate of the Caribbean. As a result, the Colombian flora includes some of the world most diverse groups of vascular plants, with 51,220 documented species (May, 1992; UNEP-WCMC, 2004). It is hoped that most of the floristic richness is located in the protected areas that cover 365,120 km2, approximately 32% of the territory (Parques Naturales de Colombia, 2006), falling under different categories of protection, including Natural National Parks, Flora and Fauna Sanctuary, Natural National Reserves, Unique Natural Area, Park Way and Indigenous Areas. Smaller forest reserves have also been created to protect river basins for water supply. On the other hand, destruction of many natural habitats has drastically affected species distributions, often reducing their historical ranges to a set of small, fragmented populations (Brooks et al., 2002). Such habitat alteration is predicted to lead to substantial extinction in the near future. Many conservation biologists have focused their attention on areas presenting high levels of endemism and diversity, and that are also experiencing a high rate of loss of ecosystems. Such regions, characterized by localized concentrations of biodiversity under threat, and representing priorities for conservation actions, are defined as biodiversity hotspots (Myers et al., 2000; Sechrest et al., 2002). The application of this concept in the case of Colombia implies very wide studies to investigate the distribution of biodiversity and endemism across the country. Complete inventories are not realistic at that scale, so other approaches 70 Chapter III. Diversity and in situ conservation _________________________________________________________________________ have been taken to exploit incomplete biodiversity data, combining remote sensing and field sampling/inventories of indicator taxa at different scales (Villareal et al., 2006). We proposed here to use climatic niche modeling and test the potential of Passifloraceae as an indicator of biodiversity in Colombia, as this family represents several interesting traits in terms of diversity, adaptation and evolution. Indeed, Colombia is particularly rich in Passifloraceae. The family is represented by 167 species (about 27% of the total), grouped in three genera, Ancistrothyrsus, Dilkea, and Passiflora, with greatest diversity in the Andean region (123 species). The country has 58 endemic Passifloraceae species, 95% of them exclusively Andean, implying a high extinction risk, as the Andes region is the most densely populated and hence disturbed of the country (Ocampo et al., 2007). According to the categories and criteria of IUCN Red List of threatened species (IUCN, 2003) and Ocampo et al. (2007) have established that more than 100 Colombian species of Passifloraceae are threatened to some degree, while three species are considered extinct. As vines, most Passiflora species have adapted to many different habitats, particularly for their support. They are medium-lived organisms depending on longer-lived trees and shrubs, which makes them responsive to both medium and long-term changes of their ecosystems. They also show high levels of co-evolution with their herbivores, the bestknown example being that of Heliconius butterflies (Gilbert, 1982), and some species even exhibit elements of the carnivory syndrome (Radhamani et al., 1995). They have developed mutualism with protector insects as nectar-feeding ants (Apple & Feener, 2001), and with a wide range of pollinators, including small and large insects, birds, and even bats (Büchert & Mogens, 2001; Sazima & Sazima, 1978). Finally, given the economic importance of several of its representatives, the genus Passiflora constitutes an important genetic resource, and the characterization and evaluation of wild and cultivated populations is seen as a high priority for Andean countries because of its clear potential for development and crop diversification (Debouck & 71 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Libreros, 1995). Strategies for conservation and improvement are needed to optimize the use and conservation of this resource. Biodiversity data have been traditionally produced through a variety of complementary approaches using field survey and sampling, museum records, botanical collections, and, in recent times, spatial analysis of digital data integrated within a Geographical Information System (GIS). In each area, the combination of geological, edaphic, climatic, ecological, historical and anthropic factors produce a unique range of constraints defining unique patterns of genetic diversity (Maxted et al., 1995). GIS makes it possible to build maps of species richness, potential distribution and endemism, to prioritize areas for conservation based on principles such as complementarity, and to assess the completeness of existing protected areas networks (Peterson, 2001). Applications have been developed in recent years that offer new possibilities for understanding biological diversity. Several of these methods use climatic variables as the principal drivers of herbarium or collecting data, often in combination with spatial environmental information, and are generally acknowledged to generate valuable additional information for diversity studies and conservation actions (Franklin, 1995; Skov, 2000; Lehmann et al., 2002). Such modeling tools have been applied to problems of plant biogeography (Hijmans et al., 2001; Hijmans and Spooner, 2001; Midgley et al., 2002; Vargas et al., 2004; Ferguson et al., 2005), conservation (Jones et al., 1997; Hijmans et al., 2000; Jarvis et al., 2002, 2003, 2005; Kingston & Waldren, 2005), evolutionary ecology (Zaharieva et al., 2004), invasive or endemic species management (Peterson & Robins, 2003; Peterson, 2004; Leimberck et al., 2004), and potential areas for plant collection (Rodríguez et al., 2005). In Passiflora, Segura et al. (2003) reported a first mapping of the potential distribution of five species of subgenus Tacsonia in the Andean countries using the Floramap software of Jones et al. (2002). Their results showed evidence of intra-specific variation in climatic adaptation along the Andes, from Colombia to Peru. Recently, Scheldeman et al. (2006) used the DIVA-GIS computer program to determine the distribution and environmental adaptation of highland papayas (Vasconcellea spp.) based on germplasm and herbarium data. The combination of observed diversity maps and potential diversity maps permitted identifying important collection gaps, mainly in Colombia and Ecuador. 72 Chapter III. Diversity and in situ conservation _________________________________________________________________________ The present study of the distribution and diversity of Passifloraceae in Colombia was conducted through sequentially: (1) assessing the geographic distribution of collections, (2) analyzing it in terms of species richness across the Colombian territory, (3) inferring the potential distribution of each species with predictive distribution models, (4) summing these spatial predictions to produce a map of potential diversity, and (5) locating collecting gaps by detecting those areas where Passifloraceae species are likely to occur but have no yet been collected. Combining these results permits an analysis of the current status of in situ and ex situ conservation of Passifloraceae in Colombia. It also provides elements to evaluate the potential of this group as an indicator for the detection of biodiversity hotspots and monitoring of conservation/restoration efforts. III.1.3. Materials and methods III.1.3.1. Geography and climate Colombia is located in the north of South America, between 12° 26’ 46” N and 4° 13’ 30” S and between 66° 50’ 54” W and 79° 02’ 33” W, covering an area of 1,141,748 km2, with altitudes ranging from the sea level to 5,775 m (IGAC, 2006). It is divided in 32 departments. Its natural habitat diversity is distributed among five main biogeographic regions (Amazonian, Andean, Caribbean, Orinoquian and Pacific). Figure 1 shows the distribution of the 32 departments and the five biogeographic regions of the country. Colombian climates are tropical, with relatively uniform temperatures throughout the year. Precipitations vary greatly, with some of the wettest parts of the world in the Pacific lowlands (average annual rainfall reaching 10,000 mm) contrasting with extremely dry areas in the coast (< 500 mm per year), and show a tendency to increase with altitude in the Andean region. 73 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Figure 1. Colombia’s geopolitical division in 32 departments and biogeographic division in five regions. Amazonas (ama), Antioquia (ant), Arauca (ara), Atlántico (at), Bolívar (bl), Boyacá (by), Cauca (cau), Cesar (ce), Caldas (cl), Córdoba (cor), Caquetá (cq), Casanare (cs), Cundinamarca (cun), Chocó (cho), Guainia (gn), Guaviare (gv), Huila (hu), La Guajira (lg), Magdalena (ma), Meta (met), Nariño (na), Norte de Santander (ns), Putumayo (pu), Quindío (qu), Risaralda (ri), Santander (snt), San Andrés y Providencia (sp Island), Sucre (suc), Tolima (to), Vaupés (va), Valle del Cauca (vc), Vichada (vch). 74 Chapter III. Diversity and in situ conservation _________________________________________________________________________ III.1.3.2. Herbarium, field and literature data The dataset used in this analysis consists of the information gathered and georeferenced by Ocampo et al. (2007), from 3,930 individuals belonging to 167 Passifloraceae species, consisting of 3,330 specimens from 23 herbaria (AFP, CAUP, CDMB, CHOCO, COL, COAH, CUVC, FAUC, FMB, HUA, HUQ, JAUM, K, MA, MEDEL, MO, NY, P, PSO, SURCO, TOLI, VALLE, UIS), 555 records from explorations in localities of 17 departments, and 45 records of specimens mentioned by Killip (1938, 1960), Uribe (1955a,b) and Escobar (1988a,b, 1989, 1994). III.1.3.3. Species distribution and richness The number of observations and their elevation range were tabulated for each species. The distribution of species was plotted on dot-maps using the DIVA-GIS software. To quantify the area over which each species is distributed, the maximum distance (MaxD) and circular area (CAr) were calculated, following the method of Hijmans et al. (2001), and compared. MaxD is the longest distance between any pair of observations of one species. CAr is calculated by assigning a circle of radius r to each observation and calculating the area covered by all circles for each species. In this case, a radius of 50 km was used (CA50). Species richness was calculated as the number of species within a defined area, superimposing species location maps, using the point-to-grid richness analysis tool in DIVA-GIS, with a 0.1 x 0.1° grid (corresponding to 12 x 12 km at the equator). The circular neighborhood option was applied with a 2° radius (Hijmans & Spooner, 2001) to eliminate border effects due to the assignation of the grid origin. Species richness was used as a measure of taxonomic diversity because it is a simple, useful, and widely used parameter (Gaston, 1996). III.1.3.4. Description of climatic preferences In order to highlight the factors that may influence species geographic distribution, and provide an indication of their tolerance to abiotic stress, climate data were extracted for each collection point. DIVA-GIS was used to develop climatic models for predicting the occurrence/diversity of Passifloraceae species in the study area. This package uses 75 Chapter III. Diversity and in situ conservation _________________________________________________________________________ WorldClim climate data (Hijmans et al., 2005), which consists of global surfaces of climate with a 30” grid resolution (corresponding to 1 x 1 km at the equator) derived from a network of over 12,500 meteorological stations across Latin America, 1,479 of which are located in Colombia. For each collection site, 19 bioclimatic variables (derived from 12 monthly means for temperature, rainfall and diurnal temperature range according to Busby (1991) were extracted from the WorldClim data set. Principal components analysis (PCA) was performed on climatic variables from the collection sites where at least one specimen was present. A varimax normalized rotation was applied with the STATISTICA® 6.0 software. To promote readability, the centroid, i.e. the arithmetic average of the factor scores, was used to represent each species general climatic preferences. III.1.3.5. Potential species distribution Maps of potential species distribution were produced using the BioClim method inside DIVA-GIS for those species with more than 10 observations. Predictive distribution modeling (also termed ecological niche modeling) attempts to predict the geographic distribution of a species using sites of known existence to understand the environmental adaptation of the species, and extrapolating this knowledge to other regions with similar conditions, where no collection has been made. BioClim was chosen because it is a robust methodology, requiring presence-only data (Hijmans & Graham, 2006). It was run using the 19 bioclimatic variables from WorldClim (Hijmans et al., 2005) with 30” resolution (equivalent to 1 km at the equator). Eighty-six species with fewer than 10 observations were omitted from the analysis, as the number of points was too low for reliable results. Unfortunately, this excluded many of the endemic and rare species. Whilst soils and habitat are likely to be also important in determining the geographic distribution of these species, they were not included in the model due to a lack of consistent, continuous data at the country level. Finally, an analysis of complementarity (Rebelo, 1994) was applied to identify the lowest number of protected areas needed for the conservation of native Passifloraceae. 76 Chapter III. Diversity and in situ conservation _________________________________________________________________________ III.1.4. Results and discussion III.1.4.1. Distribution of observations Figure 2 and Table 1 show the distribution of collection/observation points for all specimens. The Andean region of Central Colombia is by far the most densely explored, particularly in the central coffee growing zone (Quindío, Caldas and Risaralda; 18.93 to 77.20 observations/1000 km²) and the three large departments of Antioquia, Valle del Cauca and Cundinamarca (12.49 to 19.82 observations/1000 km²). By comparison, the northeastern Andes (Boyacá, Santander, Norte de Santander) and the central department of Tolima appear less well explored (3.59 to 9.39 observation/1000 km²). The situation is more difficult to appreciate in the southern Andes, as the departments of Cauca and Nariño also belong in good part to the Pacific region. However, they show a collection density only slightly superior to that of Chocó, which indicates that they have also been less explored than the central departments of the Andean region. The situation is heterogeneous in the Caribbean region, with only two of its seven departments exhibiting more than 3 observations/1000 km² (excluding the atypical case of the small San Andrés and Providencia islands). Finally, the Amazonian and the Orinoquian are by far the less explored biogeographic regions of the country, although they cover half of its area. The mean number of observations per species also reflects variation in exploration among departments (Table 1), confirming the much denser exploration in the Andes of Antioquia, Cundinamarca and Valle del Cauca (more than 7 observations per species) and in the Pacific region, while this ratio takes much lower values in the other biogeographic regions. However, the relation between exploration density and this indicator is not simple, as the numerous observations in the central coffee growing zone (Caldas, Quindío and Risaralda) are distributed among a very wide diversity of species, so the mean number of observations per species is not as high as could be expected for such densely explored areas. 77 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Figure 2. Collection localities of Passifloraceae specimens used in this study, among Colombian departments. Points on the maps represent sites of collection. 78 Chapter III. Diversity and in situ conservation _________________________________________________________________________ This variation in exploration of the Colombian territory can be related in good part to human geography. The superposition of roads and rivers on the map (not presented here) shows that collections were mostly made along rivers in the Orinoquian and Amazonian regions, and roads in other regions, which is a very common and self-explained observation (Hijmans et al. 2000). This is likely why regions with poor infrastructure (e.g. Amazonian, Orinoquian and Pacific) have been little explored. An aggravating factor is collector safety. Most of these areas also suffer from chronic political conflict. However other areas are affected although better equipped in infrastructure. Table 1. Number of observations, species, rare and endemic Passifloraceae species by Colombian division. Source for department areas: IGAC (2005, http://www.igac.gov.co). Area (km2) Nb. observ. Nb. observ. /1000km2 Total species Total species /1000km2 Total species/Log. area Observ. / species Rare species Endemic species Andean Antioquia Boyacá Caldas Cundinamarca Huila Quindío Norte de Santander Risaralda Santander Tolima 62.869 23.012 7.291 23.942 18.331 1.943 22.007 3.592 30.537 22.672 785 145 245 419 62 150 79 68 207 213 12.49 6.30 33.60 17.50 3.38 77.20 3.59 18.93 6.78 9.39 70 36 36 53 22 38 36 24 48 44 1.11 1.56 4.94 2.21 1.20 19.56 1.64 6.68 1.57 1.94 14.588 7.502 7.502 11.045 4.585 7.919 7.502 5.002 10.003 9.169 11.21 4.03 6.81 7.91 2.82 3.95 2.19 2.83 4.31 4.94 28 14 14 23 18 25 25 20 31 27 6 1 1 0 0 0 0 0 3 4 Andean and Pacific Cauca Nariño Valle del Cauca 30.985 32.046 21.195 161 170 420 5.20 5.30 19.82 42 44 56 1.36 1.40 2.69 8.753 9.170 11.670 3.83 3.79 7.38 24 27 28 1 0 1 Pacific Chocó 46.530 211 4.53 40 0.86 8.336 5.28 23 1 Caribbbean Atlántico Bolívar Cesar Córdoba La Guajira Magdalena S. Andrés y Providencia Sucre 3.319 26.469 22.213 25.020 20.848 22.742 53 10.917 18 33 13 33 21 84 4 6 5.42 1.25 0.59 1.32 1.01 3.69 75.47 0.55 7 15 10 9 12 31 2 3 2.11 0.57 0.45 0.36 0.58 1.36 37.74 0.27 1.459 3.126 2.084 1.876 2.501 6.460 0.417 0.625 2.57 2.20 1.30 3.67 1.75 2.71 2.00 2.00 5 9 9 6 9 19 2 2 0 1 0 0 0 1 0 0 Orinoquian Arauca Casanare Meta Vichada 23.393 44.428 85.286 100.242 10 4 85 16 0.43 0.09 1.00 0.16 6 4 24 9 0.26 0.09 0.28 0.09 1.250 0.834 4.930 1.876 1.67 1.00 3.56 1.78 3 4 14 6 0 0 0 0 Amazonian Amazonas Caquetá Guainía Guaviare Putumayo Vaupés 109.665 91.725 70.691 55.391 24.885 54.135 87 47 16 27 56 35 0.79 0.51 0.23 0.49 2.25 0.65 19 18 10 14 26 20 0.17 0.20 0.14 0.25 1.04 0.37 3.751 3.751 2.084 5.418 2.918 4.168 4.05 2.61 1.60 1.93 2.15 1.75 16 13 9 11 20 10 0 0 0 0 0 0 Predominant region / department 79 Chapter III. Diversity and in situ conservation _________________________________________________________________________ This is particularly obvious in the less explored Andean departments (Tolima, Santander, Norte de Santander and part of Boyacá) and in most departments of the Caribbean. Conversely, most densely populated areas, particularly those around main cities and their universities (Bogotá, Medellín, Cali, central coffee growing zone), have been much more densely explored. This high heterogeneity in exploration must be taken into account to avoid a sampling bias when comparing the situation among departments (Table 1). However, even doing so, Figure 2 and Table 1 show a clear concentration of collection density in the Andean departments. Furthermore, a closer look at the heterogeneity of collection density within these departments, as shown in Figure 1, confirms the association between topography and the density of botanical observations. The most likely explanation is simply that the higher Passiflora diversity in the Andes made their exploration more rewarding and stimulating for its students. The distribution of species richness clearly confirms this view. III.1.4.2. Species richness Departments of the Andean region present a higher species richness (Table 1). The only non-Andean department showing a comparable value for this parameter is Chocó. Among the Andean departments, Antioquia has by far the highest number of species (70 species; 43% of the total), followed by Valle del Cauca and Cundinamarca. These three departments also show the three highest mean numbers of observations per species, which confirms that they are much better explored than the others. Concerning rare species (≤ 5 observations, ≤ 100 km of MaxD and ≤ 20,000 km2 of CA50), Santander (northeast) occupies the first place, with 31 species, followed by Valle del Cauca (28), and Antioquia (28), Nariño (27), and Tolima (27). Thus, there is little doubt that a more thorough exploration north of the Eastern Cordillera (Santander) and south of the Central Cordillera (Tolima) would discover more specimens per observed species and/or more species. This problem is similar in the departments of the Amazonian, Orinoquian and Pacific regions, where species richness is very poor in relation to their surface. The number of rare species is abnormally high as compared to the total number, indicating a sampling bias due to very poor exploration in these regions. In the case of Chocó (Pacific region) collecting efforts have been practically limited to the borders of the main access road. 80 Chapter III. Diversity and in situ conservation _________________________________________________________________________ III.1.4.3. Species diversity When species richness is related to department size, the most diverse area clearly corresponds to the central coffee growing zone, as this ratio appears to be several times higher in Caldas, Risaralda and Quindío than in the other Andean departments. A precise comparison with departments of other regions is only possible if the species are equally sampled, i.e. if the number of observations per species is equivalent. This is the case for Chocó, Amazonas, and Córdoba, all of them showing a much lower diversity. The map of Passifloraceae diversity, as produced by the GIS analysis (Figure 3), confirms the importance of the Andes and the special contribution of the central coffee growing zone. This correspondence is so striking that one can wonder if, in addition to the highly variable topography, the soil fertility of this area has particularly contributed to its biodiversity. III.1.4.4. Distribution by altitude The genera Ancistrothyrsus and Dilkea reach altitudes of 800 m, mostly in the Amazonian region. In contrast, the genus Passiflora is distributed between sea level and 3,700 m. Figure 4 presents the relationship between elevation range and species diversity (species number related to diversity) for genus Passiflora, which appears trimodal, with maximal values below 500 m and in the ranges 1000-1,500 and 2,500-3,000 m. The species number decreases sharply after 3,500 m until the limit of 4,000 m. With the aim of understanding better this very particular altitudinal distribution, we have taken into account the complexity of this genus, gathering the sixteen subgenera of Passiflora present in Colombia into five groups defined on morphological and molecular grounds (cfr. chapters IV and V), and analyzed altitudinal distribution of richness in these species subsets. This grouping is similar to that of the proposal of Feuillet & McDougal (2003), except that Killip’s subgenera Rathea and Tacsonia are maintained as a distinct fifth group, because of its large and elongated flowers, mostly with red or pink color and very reduced crown, specifically adapted to pollination by the sword-hummingbird. The four others correspond to (1) subgenus Astrophea (trees and shrubs), (2) a Decaloba-like group, equivalent to subgenus Decaloba sensu Feuillet & McDougal (subgenera Apodogyne, Decaloba, Murucuja, Porphyropathanthus, Pseudomurucuja and Psilanthus; mostly species with laminar nectaries, small apetalous flowers, pollinated by bees and small insects, and small fruits), 81 Chapter III. Diversity and in situ conservation _________________________________________________________________________ (3) a Deidamioides-like group, equivalent to subgenus Deidamioides sensu Feuillet & MacDougal (subgenera Deidamioides and Tryphostemmatoides), and (4) a Passiflora-like group gathering subgenera Calopathanthus, Distephana, Dysosmia, Dysosmioides, Passiflora, and Manicata, i.e. species with large flowers, pollinated by large bees or hummingbirds, and large fruits. Figure 3. Species richness observed for Passifloraceae in 1x1 km grid cells in Colombia (165 species). Points on the maps represent sites of collection. 82 Chapter III. Diversity and in situ conservation _________________________________________________________________________ The comparison between partial curves clearly shows three distinct patterns in the adaptive potential of these five groups. Astrophea and the Passiflora-like group present a bimodal distribution with a first cohort of species adapted to tropical lowlands below 500 m, with 16 and 28 species respectively, and a second one adapted to medium elevations (from 1000 to 2,000 m). Very few species can be found at higher altitudes, with only one record of P. lindeniana Planch. ex. Triana & Planch. just below 2,700 m for Astrophea, and seven species for Passiflora. The opposite is true for subgenus Tacsonia, which shows exclusive adaptation to cold highland climates, as it is typically concentrated between 2,500 and 4,000 m, with a peak between 2,500 and 3,000 m. The fast radiation of this subgenus is clearly the cause of the peak of the global curve around 2,500-3,000 m. The third pattern is that of the Decaloba-like group, whose range of adaptation extends spectacularly from 0 to more than 3,000 m, with no lowland peak and a slight peak around 1000-1,500 m, and the small Deidamioides-like group, which shows a similar quite uniform distribution of species richness from 0 to 3,150 m, mostly in the Pacific and Andean regions. An interrogation remains concerning the first inflexion of the global curve and those of Astrophea and Passiflora-like groups in the range 500-1000 m. Interestingly, Jørgensen & León-Yánez (1999) report a bimodal altitudinal distribution of vines in the Ecuadorian flora, with maximal diversity below 500 m and in the 2,000-3,000 m range, and a maximal diversity for Passiflora between 2,500 and 3,000 m. Taking latitudinal variation into account (the same Tacsonia species usually show a higher distribution in Ecuador, with a difference of about 300-500 m), this corresponds very well with our observations in Colombia. Considering all Passifloraceae, the variation in number of Ecuadorian species with altitude (Kessler, 2002) follows the same pattern as in Colombia. The Ecuadorian richness and high endemism level for subgenus Tacsonia is another strong point of convergence with the Colombian situation. According to Jørgensen & León-Yánez (1999) bimodality in altitudinal vine diversity distribution might be due to differential collecting intensity. However, there is no reason to expect a more continuous pattern. Indeed, Kessler (2002) showed that there is no common elevational pattern for diversity, but a wide variety of independent patterns at all taxonomic levels, with maxima at different elevations, and that endemism appeared highest in the narrowest and most fragmented elevational belts. “The degree to which these influences become visible along the elevational gradient are 83 Chapter III. Diversity and in situ conservation _________________________________________________________________________ determined by which combination of species is analysed”. Strikingly, the same conclusion may be drawn within Passifloraceae, and particularly within Passiflora, taking into account infrageneric divisions. This result restricts the potential use of Passifloraceae as an indicator group to the Andean region, where they have developed most of their diversity. Figure 4. Distribution of total species richness (within circles) and species relative diversity in relation to altitude in Colombia (3,930 observations), for Passiflora and five infrageneric groups. III.1.4.5. Climatic requirements A PCA was carried out on the 19 climatic variables for the 3,930 records of our dataset. The first two principal components explain 77% of the total variance (Table 2). The first one accounts for half of the variation (49%) and is strongly correlated (r > 0.97) with variables associated to temperature (maximum, mean and minimal, but not seasonality in temperature). The second one explains 28% of the total variation and is related principally with precipitation in the whole year and in particular seasons (but again, not for seasonality in precipitation). Figure 5 shows the distribution of species in the principal plane. The first axis differentiates Andean species adapted to temperatures below 15 °C (i.e. >2,000 m), on the left side from those growing below 2,000 m, on the right side. Characteristically, these rightmost species originate from the Amazonian and Orinoquian. The second axis separates 84 Chapter III. Diversity and in situ conservation _________________________________________________________________________ the species in respect to precipitation. Thus P. arbelaezii L. Uribe, P. costaricensis Killip, P. chocoensis Gerlach & Ulmer, P. lobata (Killip) Hutch. ex MacDougal, P. occidentalis Hernández, P. pacifica L.K. Escobar, P. palenquensis Holm-Niels. & Lawesson and P. tica Gómez-Laur. & L.D. Gómez logically shows preferences for high precipitation, a predominant condition in the Pacific region, and all are predicted to exist sympatrically. At the other extreme of the second axis, are species adapted to lower precipitation levels, specifically to the marked dry season of the Caribbean region, such as P. bicornis Mill., P. serrulata Jacq., P. guazumaefolia Juss. and P. pallida L. Amazonian species tend to take an intermediate position. Table 2. Factor loadings, eigenvalues and percentages of variance for the first four components, resulting from the PCA analysis on 19 bioclimatic parameters for the 3,930 collection points (Colombian Passifloraceae). Bioclim Parameters Principal component Annual Mean Temperature Mean Monthly Temperature Range Isothermality Temperature Seasonality Max, Temperature of Warmest Month Min, Temperature of Coldest Month Temp, Annual Range Mean Temperature of Wettest Quarter Mean Temperature of Driest Quarter Mean Temperature of Warmest Quarter Mean Temperature of Coldest Quarter Annual Precipitation Precipitation of Wettest Month Precipitation of Driest Month Precipitation Seasonality Precipitation of Wettest Quarter Precipitation of Driest Quarter Precipitation of Warmest Quarter Precipitation of Coldest Quarter Eigenvalue Percentage of variance 1 2 3 4 0.98 0.08 0.00 0.45 0.97 0.98 0.08 0.98 0.98 0.98 0.98 0.24 0.29 0.09 0.23 0.28 0.09 0.10 0.29 9.26 0.17 -0.21 0.06 0.03 0.16 0.20 -0.22 0.17 0.18 0.17 0.17 0.96 0.91 0.91 -0.55 0.91 0.93 0.87 0.89 5.37 0.09 -0.16 -0.95 0.77 0.12 0.06 0.37 0.09 0.10 0.11 0.07 0.04 0.15 -0.28 0.60 0.17 -0.25 -0.20 0.05 1.74 -0.03 -0.96 -0.01 -0.18 -0.12 0.04 -0.89 -0.02 -0.04 -0.04 -0.03 0.10 0.10 0.13 0.00 0.09 0.13 0.12 0.02 1.51 48.74 28.26 9.15 7.94 The species repartition in the principal plane consistently reflects the potential for climatic adaptation of the groups that were defined for the analysis of altitudinal distribution. Thus, subgenus Tacsonia shows a marked adaptation potential to cool conditions, while subgenus 85 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Astrophea and the Passiflora-like group show much higher potential in hot and mild climates. The genera Ancistrothyrsus and Dilkea are even more clearly adapted to lowlands, mainly in the Amazonian region. The Decaloba-like group shows a much broader adaptation range, explaining its quite constant presence across the different biogeographic regions. Figure 5. Distribution of Passifloraceae species in the PCA principal plane for climatic variables, with indication of genera (Ancistrothyrsus and Dilkea) and subgenera of genus Passiflora. III.1.4.6. Areas of distribution - endemic species Table 3 gives MaxD and CA50 for the 165 Passifloraceae native species of Colombia. Figure 6 shows a good correspondence (R² = 0.77) between these distribution parameters, whose comparison also provides information on species dispersion. Species with a high MaxD but relatively low CA50 indicate low densities, as a result of biological rarity and/or under-collection, whereas high CA50 relative to MaxD indicates high density due to 86 Chapter III. Diversity and in situ conservation _________________________________________________________________________ population biology and/or intense collection. The species with the widest distributions in Colombia (more than 1,100 km MaxD) are those that show a very wide distribution including most neotropical regions, such as the very common P. foetida L., P. auriculata, P. quadrangularis and P. laurifolia L. (wild and cultivated) P. suberosa L, and P. serratodigitata L., P. capsularis L., P. rubra L., and P. misera Kunth, and others of still considerable regional distribution, such as P. vitifolia Kunth, P. coccinea Aubl., P. spinosa (Poepp. & Endl.) Mast., P. nitida, P. subpeltata Ortega, P. maliformis, P. menispermifolia Kunth, P. biflora Lam., and Dilkea parviflora Killip. Only P. arborea Spreng (Panamá to Ecuador) and P. cumbalensis (Colombia to Peru) show a more restricted distribution. These high-MaxD species are concentrated at low to medium elevations, the only exception being P. cumbalensis. According to IUCN criteria, they are not threatened (Least Concern category), except for P. arborea (Near Threatened; Ocampo et al., 2007). Between 200 and 1,100 km of MaxD, there are species of regional importance, such as P. mixta L.f., P. ligularis, and endemics with a relatively wide distribution, such as P. sphaerocarpa Triana & Planch. (805 km²), P. lehmanni Mast. (878 km²), P. antioquiensis and P. mollis Kunth. The latter displays a relatively high CA50 in its group, as its 17 observations are quite scattered along the Cordillera Occidental. The position of P. coriacea Juss. in this group of medium dispersion is very surprising, as it is found in all neotropical countries. The 71 species with MaxD values below 225 km include 36 narrow endemics, 21 of which are exclusive to nine departments, particularly Antioquia (six species), Tolima (four) and Santander (three). The 14 others show similar MaxD and CA50 but live across administrative divisions. Only four of these 36 narrow endemics are represented by ten or more observations while ten species are only known from the type collection. Ocampo et al. (2007) considered a MaxD under 100 km as a criterion of rarity and threat, which places 26 endemic species in a very critical situation. The situation of 33 non-endemic species with a MaxD less than 100 km must be examined in relation to their distribution in neighbor countries. P. truxilliensis Planch. & Linden ex Triana & Planch. is shared with Venezuela, but it is a narrow endemic living around the border. The distribution of 14 species extends to farther places in neighboring countries, and 18 species present a wide distribution, extending to non-neighboring countries. For example, P. tricuspis is only reported once, in the Andean foothill, so it has a null MaxD, however its distribution extends south to 87 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Bolivia. Sixteen of these 30 species are adapted to lowland conditions, which suggests that their apparent rarity is in fact due to the poor collecting in the corresponding regions. Figure 6. Extent of Passifloraceae species distribution in Colombia: circular area (CA50) vs. maximum distance (MaxD). 88 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Table 3. Total number of Passifloraceae present in Colombia. Number of observations, maximum distance (MaxD) and circular area (CA) for each species. Endemic species are highlighted by an asterisk (*). RC: species rare for Colombia; Roc: species rare in other countries; Rne: rare narrow endemic, Ne: narrow endemic; Re: rare endemic; Ce: common endemic. Presence in a richness spot is indicated by Rs and the identification number used in Figure 7. Species Ancistrothyrsus antioquiensis L.K. Escobar (ined.)* Ancistrothyrsus tessmannii Harms Dilkea johannesii Barb. Rodr. Dilkea parviflora Killip Dilkea retusa Mast. Passiflora adenopoda Moc. & Sessé ex DC. Passiflora adulterina L.f. * Passiflora alnifolia Kunth Passiflora alata Curtis Passiflora ambigua Hemsl. ex Hook. Passiflora andina Killip Passiflora andreana Mast. Passiflora antioquiensis H. Karst. * Passiflora apoda Harms Passiflora arbelaezii L. Uribe Passiflora arborea Spreng. Passiflora auriculata Kunth Passiflora azeroana L. Uribe * Passiflora bicornis Mill., Passiflora bicuspidata (H.Karst.) Mast. * Passiflora biflora Lam. Passiflora bogotensis Benth. * Passiflora bracteosa Planch. & Linden Passiflora bucaramangensis Killip * Passiflora callistema L.K. Escobar * Passiflora candollei Tr. & Planch. Nb. observ. MaxD (km) CA (km2) Rare species Endemics and distribution 2 1 1 22 5 51 43 121 1 48 1 3 55 43 48 67 128 10 11 16 40 56 7 8 1 4 41 0 0 1,185 952 383 234 1,244 0 929 0 45 667 678 746 1,204 1,635 574 675 438 1,326 1,057 122 70 0 854 11,762 7,814 7,814 40,688 106,159 82,650 39,072 170,761 7,814 137,261 7,814 12,214 99,064 83,615 113,491 144,115 334,952 34,734 52,098 61,674 122,047 89,250 23,180 15,032 7,814 26,294 RC RC / Roc RC / Roc Rne (Antioquia) 89 RC Ce RC RC /Roc RC Colombia and Ecuador Ce Ce Ce RC /Roc RC RC RC Ce Colombia and Venezuela Ne (Santander) Rne (Bolivar) Chapter III. Diversity and in situ conservation _________________________________________________________________________ Passiflora capsularis L. Passiflora chelidonea Mast. Passiflora chocoensis G. Gerlach & T. Ulmer * Passiflora cincinnata Mast. Passiflora citrifolia (Juss.) Mast. Passiflora coccinea Aubl. Passiflora colombiana L.K. Escobar * Passiflora coriacea Juss. Passiflora costaricensis Killip Passiflora cremastantha Harms * Passiflora crispolanata L.Uribe * Passiflora cuatrecasasii Killip * Passiflora cumbalensis (Karst.) Harms Passiflora cuneata Willd. Passiflora cuspidifolia Harms, Passiflora danielii Killip * Passiflora dawei Killip * Passiflora emarginata Humb. & Bonpl. * Passiflora engleriana Harms * Passiflora erytrophylla Mast. * Passiflora escobariana J.M. MacDougal Passiflora filipes Benth. Passiflora fimbriatistipula Harms * Passiflora flexipes Triana & Planch. * Passiflora foetida L. Passiflora formosa T. Ulmer * Passiflora glandulosa Cav. Passiflora gleasonii Killip Passiflora gracillima Killip Passiflora grandis Killip * Passiflora gritensis H. Karst. Passiflora guatemalensis S. Watson Passiflora guazumaefolia Juss. Passiflora hahnii (Fourn.) Mast. Passiflora haughtii Killip * Passiflora hirtiflora Jorgensen & Holm-Nielsen 64 18 1 1 3 21 2 59 1 1 11 9 156 9 33 5 4 46 2 6 2 3 18 24 143 1 1 2 29 2 8 11 8 1 1 1 1,437 1,024 0 0 68 1,285 42 741 0 0 246 181 1,196 877 812 180 208 654 110 225 3 48 198 322 1,830 0 0 3 684 14 346 971 349 0 0 0 90 159,962 94,209 7,814 7,814 14,049 107,128 11,910 136,372 7,814 7,814 29,720 21,312 199,941 50,607 86,640 20,590 23,702 78,393 8,902 27,643 8,136 13,227 33,664 36,121 420,440 7,814 7,814 8,075 74,546 9,161 26,115 59,505 41,192 7,814 7,814 7,814 RC RC RC Rne (Choco) RC Rne RC RC RC RC RC RC / Roc RC / Roc Rne (Cauca) Ce Ne (Rs 2) Rne (Antioquia, Rs 2) Rne (Rs 3) Ce Rne (Antioquia, Rs 2) Ne (Rs 2,3) (Antioquia, Rs 2) México to Ecuador Ne Ce RC RC RC Rne (Boyacá) RC Rne RC RC RC Rne (Santander) Chapter III. Diversity and in situ conservation _________________________________________________________________________ Passiflora holosericea L. Passiflora holtii Killip Passiflora hyacinthiflora Planch. & Linden * Passiflora involucrata (Mast) A.H. Gentry Passiflora jardinensis L.K. Escobar * Passiflora kalbreyeri Mast. * Passiflora killipiana Cuatrecasas Passiflora lanata (Juss.) Poir. * Passiflora laurifolia L. Passiflora lehmanni Mast. * Passiflora leptomischa Harms * Passiflora ligularis Juss. Passiflora lindeniana Planch. ex Triana & Planch. Passiflora linearistipula L.K. Escobar * Passiflora lobata (Killip) Hutch. ex J.M. MacDougal Passiflora longipes Juss. * Passiflora lyra Planch. & Lind. ex Killip Passiflora macrophylla Spruce ex Mast. Passiflora magdalenae Triana & Planch. * Passiflora magnifica L.K. Escobar* Passiflora maliformis L. Passiflora manicata (Juss.) Pers. Passiflora mariquitensis Mutis ex Uribe * Passiflora megacoriacea Porter-Utley (ined.) Passiflora menispermacea Triana & Planch. * Passiflora menispermifolia Kunth Passiflora micropetala Mast. Passiflora misera Kunth Passiflora mixta L. f. Passiflora mollis Kunth Passiflora monadelpha Jorgensen & Holm-Nielsen Passiflora morifolia Mast. Passiflora multiformis Jacq. Passiflora munchiquensis Hernandez (ined.)* Passiflora mutisii Killip * Passiflora nitida Kunth 7 1 3 8 8 19 1 32 11 17 21 101 2 4 3 21 4 20 21 6 122 62 3 1 2 43 11 54 162 17 7 1 4 4 1 72 238 0 305 1,197 35 426 0 284 1,350 805 449 914 395 8 194 334 69 716 129 33 1,208 889 10 0 18 1,410 1,318 1,148 966 554 67 0 147 200 0 1,452 91 25,632 7,814 17,746 48,827 11,335 41,237 7,814 45,476 84,672 91,156 46,331 170,123 15,628 8,695 23,115 45,557 14,716 90,432 31,127 12,215 212,270 114,036 8,436 7,814 9,610 167,659 68,015 145,398 191,787 208,941 33,665 7,814 17,652 22,441 7,814 279,511 RC / Roc RC RC RC / Roc Re Ne (Antioquia) Ce Colombia to Peru Ce Ce Ce RC / Roc RC RC Colombia and Venezuela Rne (Caldas) Ce RC / Roc RC RC RC RC RC / Roc RC RC RC RC Ne Ne (Antioquia, Rs 2) Rne (Tolima) Rne (Tolima, Rs 4) Colombia and Ecuador Rne Rne (Tolima) Chapter III. Diversity and in situ conservation _________________________________________________________________________ Passiflora occidentalis Hernandez (ined.)* Passiflora oerstedii Mast. Passiflora pacifica L.K. Escobar * Passiflora palenquensis Holm-Niels. & Lawesson Passiflora pamplonensis Planch.& Linden ex Tr. & Planch. * Passiflora pallida L. Passiflora panamensis Killip Passiflora parritae (Mast.) L.H. Bailey * Passiflora pennellii Killip * Passiflora phaeocaula Killip Passiflora picturata Ker Passiflora pilosissima Killip * Passiflora pinnatistipula Cav. Passiflora pittieri Mast. Passiflora platyloba Killip Passiflora popayanensis Killip * Passiflora popenovii Killip Passiflora punctata L. Passiflora purdiei Killip * Passiflora putumayensis Killip Passiflora pyrrhantha Harms Passiflora quadrangularis L. Passiflora quadriglandulosa Rodschied Passiflora quindiensis Killip * Passiflora resticulata Mast. & André Passiflora rigidifolia Killip * Passiflora riparia Mart. ex Mast. Passiflora rubra L. Passiflora rugosa (Mast.) Triana & Planch Passiflora schlimiana Triana & Planch. * Passiflora securiclata Mast Passiflora seemannii Griseb. Passiflora semiciliosa Planch & Linden * Passiflora serratodigitata L. Passiflora serrulata Jacq. Passiflora sexflora Juss. 10 41 9 20 1 6 15 14 6 5 1 2 21 1 4 6 17 8 1 1 1 112 4 8 4 1 3 90 12 7 4 40 4 18 10 14 474 728 510 1,181 0 898 295 100 313 498 0 270 750 0 201 64 636 592 0 0 0 1,676 414 225 414 0 716 1,351 421 181 849 1,341 578 1,566 331 353 92 42,350 148,975 39,585 100,769 7,814 50,078 41,614 20,357 24,413 28,305 7,814 15,628 57,114 12,661 16,471 15,078 31,075 40,022 7,814 7,814 7,814 314,317 21,256 24,711 18,938 7,814 23,442 117,934 35,549 27,852 30,708 129,777 26,175 67,105 29,354 43,143 Ce Ce RC Rne (N. de Santander) RC Ne Ce RC / Roc RC Re RC RC RC RC RC / Roc RC / Roc Ne (Cauca) Rne RC Ne (Tolima, Rs 4) RC RC Rne (Antioquia) Roc RC / Roc Ne Colombia and Venezuela RC Re Chapter III. Diversity and in situ conservation _________________________________________________________________________ Passiflora sierrae L.K. Escobar * Passiflora skiantha Huber Passiflora smithii Killip Passiflora sodiroi Harms Passiflora sphaerocarpa Triana & Planch. * Passiflora spicata Mast. Passiflora spinosa (Poepp. & Endl.) Mast. Passiflora suberosa L. Passiflora subpeltata Ortega Passiflora tarminiana Coppens & Barney Passiflora tenerifensis L.K. Escobar * Passiflora tica Gómez-Laur. & L.D. Gómez Passiflora tiliifolia L. Passiflora tolimana Harms * Passiflora trianae Killip * Passiflora tribolophylla Harms * Passiflora tricuspis Mast. Passiflora trinervia (Juss.) Poir.* Passiflora tripartita (Juss.) Poir. Passiflora trisulca Mast. * Passiflora truxillensis Planch. & Linden ex Triana & Planch. Passiflora tryphostemmatoides Harms Passiflora tuberosa Jacq. Passiflora uribei L.K. Escobar * Passiflora ursina Killip & Cuatrec. Passiflora variolata Poepp. & Endl. Passiflora venosa Rusby Passiflora vespertilio L. Passiflora vestita Killip Passiflora vitifolia Kunth Passiflora x rosea (H. Karst.) Killip 2 1 28 1 35 1 20 66 35 28 4 5 48 12 2 1 1 27 56 8 1 25 1 3 2 6 1 3 1 359 7 46 0 827 0 878 0 1,521 1,497 1,344 832 71 319 1,010 426 39 0 0 220 1,210 441 0 557 0 54 7 412 0 292 0 1,729 161 93 12,194 7,814 72,555 7,814 96,244 7,814 118,197 158,860 89,527 103,373 15,195 23,119 97,205 33,711 11,594 7,814 7,814 36,932 145,398 25,258 15,628 77,831 7,814 12,960 8,503 27,059 7,814 20,887 7,814 456,229 20,988 RC RC / Roc Rne (Magdalena) Colombia and Peru RC / Roc Colombia and Ecuador Ce Colombia and Brazil RC / Roc RC RC RC RC RC Rne (Valle del Cauca) Ce Rne Rne Ne (Rs 5) RC / Roc Ce Colombia and Venezuela RC RC RC / Roc Rne Colombia and Ecuador RC / Roc RC RC / Roc Colombia and Ecuador Chapter III. Diversity and in situ conservation _________________________________________________________________________ III.1.4.7. Potential distribution of species and species assemblages For each of the 80 species with more than ten observations, the BioClim model provided a probabilistic prediction of its geographic distribution using climate data as the driver of species distribution. The results cannot be presented here for each individual species, but are available upon request from the authors. The potential distribution of richness was calculated by summing the 80 spatial predictions of species occurrence (Figure 7). The areas of highest predicted species richness (41 to 54 predicted sympatric species) are mostly located in the center of the country, on the slopes of the three cordilleras, between 1000 and 2,000 m of elevation. Collection has been intense in many of them, showing a high diversity, however the correspondence is not perfect between observed and modeled distribution. While the species-rich areas of Antioquia and Caldas, Quindío, Cundinamarca and eastern Boyacá, and even the poorly explored but promising Santander, are well represented on the map (areas 2, 5, 3, 4 and 1 respectively), only very small richness spots are drawn for Valle del Cauca (area 7), or no richness spot indicated for Cauca and southern Huila. Conversely, predicted richness spots 6, 8 and 9 (eastern Tolima-northern Huila- southern Cundinamarca, western Caquetá, Nariño) were not detected in the analysis of observed diversity, indicating potential collecting gaps. The model predicts a very poor representation of Passiflora in the lowlands of the Caribbean and Orinoquian and part of the Pacific, as well as in the Sierra Nevada de Santa Marta, an isolated mountain range on the Caribbean Coast, reputed for its high level of endemism. In both cases, this may be attributed to the poor exploration of these areas (low densities of observation) and to poor representation of their species (few observations per species) resulting in them not having sufficient observations to be used in the predictive modeling. This bias can be corrected by further collecting in these regions. Alternatively, materials of Colombian species collected in border regions of neighboring countries, belonging to the same biogeographic entities (e.g. the Venezuelan Llanos for the Orinoquian, Costa Rican and Ecuadorian Pacific, Brazilian, Ecuadorian and Peruvian Upper Amazonian) might be used to refine these models and increase the number of observations per species under analysis. The biodiversity hotspot concept not only considers species richness and diversity but also endemism. In an analysis of New Zealand fern diversity, Lehmann et al. (2002) observed a 94 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Figure 7. Modeled distribution of Colombian Passifloraceae diversity based on data from 80 species presenting more than 10 observations. Ellipses individualize high richness spots mentioned in the text. poor correlation between both parameters. For Colombian Passifloraceae, we could not establish rigorously their correspondence, as the analysis was not designed for rare species, 95 Chapter III. Diversity and in situ conservation _________________________________________________________________________ however we compared their distributions, distinguishing four categories among the 58 endemics: those with a relatively wide distribution (MaxD > 100 km, 19 species), the narrow endemics (MaxD under 100 km, three species), the rare endemics (less than six observations, three species), and the rare narrow endemics (both criteria, twenty species). Six of the 10 narrow endemics and seven of the 25 rare narrow endemics live in one of the nine hotspots defined by the diversity analysis. None of the three rare endemics are found in those sites. Of the 27 rare and/or narrow endemics living out of those hotspots, three are considered extinct (Ocampo et al., 2007), seven are only adapted to lowlands, which easily explains their absence in Andean hotspots; two species are endemic to the Sierra Nevada de Santa Marta, an area of endemism that could not be sufficiently taken into account for reasons explained previously. Finally, only 11 of the 27 living Andean rare/narrow endemic species, i.e. less than 50%, live in one of the hotspots. This proportion must be compared with more than 54 out of 80 in the case of the non-rare species whose distribution determined those hotspots. Thus, preserving these nine areas should have a less positive impact on the conservation of narrow endemics than on the general Passifloraceae diversity. Using an analysis of complementarity reserve selection developed by Rebelo (1994), 52 sites of 25km x 25km were selected to represent all 165 native species throughout the country. The best five sites, in Caldas, Risaralda, Norte de Santander, southern Antioquia and Boyacá, capture a total of 64 species. In just seven sites, 50% of all species could be conserved, though many of the endemic/rare species are not captured in these sites. III.1.4.8. Conservation of Passifloraceae and their habitat Figure 8 combines the estimated distribution of Passifloraceae diversity with that of protected areas in the Andes of Colombia, showing a general lack of correspondence. Most Andean protected areas are concentrated around the summits, obviously targeting páramo ecosystems. Very few small protected areas harbor a high diversity of Passifloraceae: the watershed forest reserves of Sierra del Peligro (Boyacá, 16.5 km²), Río Nare (Antioquia, 118.8 km²), Río San Francisco, Cuchillas Peñas Blancas, Cerro Quininí (Cundinamarca, 28.8, 16.3 and 18.0 km²), whose responsibility has recently been upgraded to the department level. The Parque Nacional Farallones (Valle del Cauca) is the only reserve of 96 Chapter III. Diversity and in situ conservation _________________________________________________________________________ national importance to protect part of a small Passifloraceae hotspot, on the eastern fringes of the nature reserve. This poor coverage is not good news, neither for a family including 71% threatened species, nor for the habitats where these species have developed numerous interactions with many other organisms. Figure 8. Distribution of protected areas in Colombia, showing poor correspondence with areas of high Passifloraceae diversity. Figure 9 shows a striking general superposition of areas of high Passifloraceae diversity on certain Colombian coffee growing zone ecotopes (Cenicafé, 2005) whose conservation is of 97 Chapter III. Diversity and in situ conservation _________________________________________________________________________ the utmost importance for the country. This is not surprising, as the corresponding elevation belts include or enclose those of major diversity (1000-2,500). It clearly shows that efforts for the conservation of Passifloraceae habitats and genetic resources must be integrated in the more general management of the coffee growing zone environment at the landscape level. The latter can be ensured by coordinating existing actions for watershed protection, management of private and low-level public reserves, as well as recreational areas for public education (rural and peripheral urban parks, arboreta, etc.), creation of environmental corridors, and improvement of agricultural practices, particularly in coffee farms (crop association, shade tree diversity, and guadua bamboo growing for building materials and water management at farm level). III.1.4.9. Passifloraceae as indicators of biodiversity According to Pearson (1994), an ideal indicator taxon should cumulate seven criteria: (i) a well-known and stable taxonomy, (ii) well known natural history, (iii) readily surveyed and manipulated, (iv) higher taxa broadly distributed geographically and over a breadth of habitats, (v) lower taxa specialized and sensitive to habitat changes, (vi) patterns of diversity in other taxa, and (vii) potential economic importance. Passifloraceae clearly fill the fifth and seventh criteria, taking into account that several common species are indicators of more or less perturbated habitats. Concerning the fourth criterion, our analyses have repeatedly underlined that Colombian Passifloraceae distribution is concentrated in the Andean region, so their use as indicators should be restricted to the corresponding elevation belts. Lianas growing in high trees are not always easily surveyed (third criterion), however their typical structures, showy flowers and interesting fruits make them easy to identify as a group, catching the attention of local populations and specialists (e.g. protected area staff), who can thus help in localizing the different species in particular places. The application of molecular techniques should produce important progress in the complex taxonomy of this group and further the understanding of its natural history. The sixth criterion is particularly important. The numerous interactions of Passifloraceae with other organisms (surrounding vegetation, pollinators, and herbivores) constitute a first indication that their diversity is necessarily related to that of other components of the ecosystem. Another indication came from a preliminary study, where we found an excellent correspondence between the 98 Chapter III. Diversity and in situ conservation _________________________________________________________________________ distributions of diversity of Passiflora and Vasconcellea (mountain papayas), another plant group whose radiation is also clearly related to the rise of the Andes (Scheldeman et al., 2006). Similar results must be obtained with more plant taxa before considering unequivocally Passifloraceae as a reliable surrogate for floral diversity in Andean ecosystems. However, given the excellent correspondence between Passifloraceae diversity maps and coffee growing zone ecotope maps, we may already recommend them as useful indicators of habitat degradation or restoration in this environmentally and economically very important region. Figure 9. Correspondence between Passifloraceae high richness spots and coffee growing zone ecotopes. 99 Chapter III. Diversity and in situ conservation _________________________________________________________________________ III.1.5. Conclusions Collections of Passifloraceae have not been uniform as a consequence of difficulty of access and/or chronic social conflict in many areas. It has been much denser in the central coffee growing zone, Antioquia, Valle del Cauca and Cundinamarca. The southern and northeastern Andes, and the Caribbean have been little explored. For the lowland forests of the Pacific, the Orinoquian and the Amazonian, data are so poor that they are misleading. Despite the resulting sampling bias, collecting parameters clearly point to the concentration of observed Passifloraceae diversity in the Andes, and more particularly the central coffee growing zone. This is further highlighted by the elevational pattern of diversity, showing three peaks, a small one for lowlands, and two higher ones at intermediate and high elevations (1000-2,000 and 2,500-3,000 m). A more thorough analysis shows that this trimodal pattern corresponds to different adaptive evolutions among genera and infrageneric divisions of Passiflora. This is also reflected in the analysis of the climatic preferences of these infrageneric groups. The analysis of species distribution areas shows a trend for more extent dispersion of species occurring at low and intermediate elevations. On the contrary, narrow endemics are more frequent among highland species. The modeled species richness map allowed to identify nine richness spots of variable size, three of which, located in the southern and southeastern Andes of Colombia, correspond to collection gaps, as they were not detected in the analysis of observed diversity. Another probable collection gap, not detected by diversity modeling, corresponds to the Sierra Nevada de Santa Marta, an isolated mountain range with both high diversity and endemism. The proportion of endemics living in high richness spots is lower than the proportion of all species used for modeling, confirming the lack of relation between diversity concentration and endemism reported in other studies. If this is further substantiated in different groups of organisms, it could limit the application of the biodiversity hotspot concept, as the best protected areas for diversity would not necessarily provide protection to a high proportion of narrow endemics. 100 Chapter III. Diversity and in situ conservation _________________________________________________________________________ Passifloraceae diversity is not currently conserved by the current network of Colombian protected areas. On the contrary, it is particularly concentrated on certain ecotopes of the coffee growing zone, i.e. highly disturbed habitats, so any conservation effort must be integrated in local management strategies at the landscape level. Passifloraceae may provide an interesting indicator to evaluate the outcome of such efforts. III.1.6. Acknowledgements The first author gratefully acknowledges financial support from the Gines-Mera Fellowship Foundation (CIAT-CBN). Part of this research has been funded by the Colombian Ministry for Environment and the Research Center of the Colombian Coffee Grower Federation (Cenicafé) through of the collaborative project “Estudio de la Diversidad de las Passifloraceae y Caricaceae en la zona cafetera de Colombia”. Finally, we thank Dr. Xavier Scheldeman (Bioversity International) for his scientific advice regarding data analysis. 101 CHAPTER IV ___________________________________ A phenetic analysis of morphological diversity in the genus Passiflora L. Chapter IV. Morphological diversity ______________________________________________________________________ IV.1. A phenetic analysis of morphological diversity in the genus Passiflora L. John Ocampo Pérez1* and Geo Coppens d’Eeckenbrugge2 1 Bioversity International (formerly IPGRI), Regional Office for the Americas, A.A. 6713, Cali, Colombia. 2 CIRAD/FLHOR, UPR ‘Gestion des ressources génétiques et dynamiques sociales’, Campus CNRS/Cefe, 1919 route de Mende, 34293 Montpellier, France. IV.1.1. Abstract Morphological variation was studied in 124 accessions from eight subgenera and 60 species among the most common cultivated and wild species of genus Passiflora, using the analysis of variance components and principal component analysis (PCA) on 43 quantitative traits, and neighbor joining cluster analysis on 84 qualitative traits. The coefficients of variation appear generally superior for subgenus Decaloba, indicating stronger differentiation, as compared with subgenera Passiflora and Tacsonia. Twentyfour quantitative descriptors showing high variation at the subgenus level, were selected for the PCA. The five principal components retained represent 84% of the total variation. The first one (32%) is closely associated with flower length (hypanthium, nectary chamber, androgynophore) and secondarily with the constriction of the floral cup above the nectary chamber. The second one (27%) is associated with flower width (length of bracts and length of corolla and corona elements) and bract shape. The third one (14%) is associated with peduncle branching, stem width and leaf length, which relates it clearly with variation between subgenus Astrophea, and secondarily subgenus Tryphostemmatoides, and all other subgenera. The projection of accessions in the resulting tridimensional space consistently separates subgenera. A selection of 32 qualitative traits and four categorized quantitative variables, whose segregation follows divisions among Killip’s subgenera, allowed classifying our 60-species sample consistently, using a strictly phenetic approach. Most discriminating characters include size of stems and leaves, presence of tendrils, number and distribution of extrafloral nectaries, dimensions and general shape of bracts, width and length of flowers, corona complexity, ovary shape, and, although they could not be systematically analyzed, fruit size and color. Eight of the nine Killip’s subgenera represented in our sample are supported by the morphological analysis, although subgenus Tryphostemmatoides is only supported in the quantitative analysis. In a second analysis, 74 qualitative descriptors were incorporated in the cluster analysis, which increased distances and improved bootstrap values, without affecting the general structure of the dendrogram, neither at the subgeneric nor at the interspecific level. Our results support seven of the eight Killip’s subgenera of our sample, but no infrasubgeneric classifications. However, the new classification of subgenus Decaloba by Feuillet & MacDougal was partly supported. They converge on many points with previous phylogenetic results obtained with DNA sequences, although the latter group subgenera Tacsonia and Distephana with subgenus Passiflora. Key words: Passiflora L, systematics, morphological descriptors, principal components, variation. 103 Chapter IV. Morphological diversity ______________________________________________________________________ IV.1.2. Introduction With more than 525 species, Passiflora L. is the most important genus of the Passifloraceae, a family including nearly 630 species. It is essentially neotropical, with only 22 species native to the Old World, in Southeast Asia, Australia and Oceania. Passionflowers are herbaceous or woody vines, usually climbing by tendrils, but a few are trees or shrubs. Other typical vegetative traits include alternate leaves, axillary stipules, and petiolar and/or laminar nectary glands. In addition, the genus exhibits several unique floral features, such as an androgynophore, a complex corona, constituted of one or several concentric rows of filaments, and a limen-operculum system limiting access to the nectary chamber, with impressive interspecific variation in size, shape and colors (see Figure 3 of Chapter I and Figure 1 of present chapter). Many species exhibit interesting fruits, pharmacological properties (e.g. sedative effect) and/or ornamental potential. More than 80 Passiflora species produce an edible fruit, the most interesting ones belonging to subgenera Passiflora and Tacsonia (Coppens d’Eeckenbrugge, 2003). The two botanicals forms of P. edulis Sims, flavicarpa Degener (yellow maracuja) and edulis (purple maracuja) are by far the most important crops in the family, with a world production estimated at ca. 640.000 tons (http://www.passionfruitjuice.com). Other cultivated passion fruits are P. tripartita var. mollissima (Kunth) Holm-Nielsen & Jørgensen (curuba de Castilla), P. tarminiana Coppens & Barney (curuba India), P. ligularis Juss. (sweet granadilla), P. maliformis L. (granadilla de piedra), P. quadrangularis L. (giant granadilla), P. popenovii Killip (granadilla de Quijos), P. alata Curtis (fragrant granadilla) and P. laurifolia L. (golden apple). These eight species are mainly commercialized on South American local and national markets, principally in Colombia and Brazil, with incursions on the international market. Passion fruits are consumed fresh or processed into juices, sherbets, ice cream, and components of industrial pastry and candies. The most important commercial species are susceptible to a large number of pests and diseases, with considerable negative effects on production. The high potential of Passiflora for crop diversification and economic development induced research institutions in the Andean countries to prioritize their characterization and the evaluation of wild and cultivated populations (Debouck & Libreros, 1995), and develop strategies for conservation and improvement of these genetic resources. 104 Chapter IV. Morphological diversity ______________________________________________________________________ Figure 1. Schema of a flowering branch of Passiflora vitifolia Kunth (drawing by Jesus Salcedo). 105 Chapter IV. Morphological diversity ______________________________________________________________________ The high diversity of combinations of characters related to the nectary glands, stipules, corona, operculum and limen (Figure 1 and 2), have been heavily relied upon for delimiting infrageneric divisions in Passiflora taxonomy, dividing it into subgenera, sections, and series. Although most species appear well delimited, there are many cases, in low level taxa (sections and series), where two or more species are very difficult to distinguish. The reference work is that of Killip (1938), who described 355 American species, plus 20 species in 1960, placing them in 22 subgenera (Annex 1). This classification was amended and completed with the renaming of several subgenera, the suppression of subgenus Tacsoniopsis, and the creation of subgenus Porphyropathantus (Escobar, 1988a,b, 1989; McDougal, 1994). In his description of the general Passifloraceae morphology, Killip (1938) underlined the particular contribution of certain organs to taxonomy. Among those of primary importance, he included the peduncle, the bracts (shape and position), the operculum, the presence or absence of petiolar glands and the grooved or reticulate seed surface, the two latter being almost perfectly correlated within subgenus Decaloba. He also mentioned habit (vines vs. trees), stipule shape and margin, leaf nervation (subgenus Astrophea), shape and arrangement of corona elements, petal absence, and fruit size. However, from the identification keys given by Killip (1938), no clear correlation emerges between main divisions and a particular character hierarchy. The presence of uncinate trichomes on the leaf epidermis was given particular importance by MacDougal (1994) in his revision of section Pseudodysosmia of subgenus Decaloba, which he defined as the “hooked trichome group”. In the same work, MacDougal mentions the taxonomic importance of the posture of apex in relation to the main shoot axis. However, a cernuous shoot tip is found in parts of subgenera Decaloba, Astrophea, and Tacsonia (Ulmer & MacDougal, 2004), which suggests instead that this trait is not determinant for delimiting subgenera. Recently, Feuillet & MacDougal (2003; Annex 2) have proposed a new general classification, taking into account the Old World species and recognizing only four subgenera, Astrophea (unchanged; trees and shrubs, rarely lianas), Decaloba (vines with small flowers and fruits, the latter usually black), Deidamioides (vines, with twoflowered peduncles, not clearly defined as a morphological group) and Passiflora (vines to lianas, with large flowers and fruits). However, this classification has not been fully developed yet, the placement of most species must be deduced from earlier literature, and its morphological basis is not clear. In the book of Ulmer & MacDougal (2004), 106 Chapter IV. Morphological diversity ______________________________________________________________________ where about 200 species are described and assigned a place in the new classification, the revision of Passiflora morphology underlines taxonomic importance for shoot apex posture, extrafloral nectaries, bract size and shape, operculum, and seed coats, but only bract and seed traits are specifically mentioned as markers of subgenera, sections, and series. The new classification was presented as the “last premodern” classification, because it does not take into account the results obtained from molecular biology techniques. However, even on strictly morphogical grounds, it lacks the support of a systematic evaluation of morphological diversity in the genus and its subtending structure. The situation is further complicated as many new species have been described during the elaboration of this classification, most often by splitting well-known species on the basis of slight variations from their type, observed in small populations (e.g. P. miniata Vanderplank, P. longicuspis Vanderplank and P. aimae Anonay & Feuillet, from P. coccinea Aubl., and P. formosa Ulmer, from P. lanata Juss.). On the other hand, the first molecular studies carried out on significant Passiflora species samples have consistently validated the three of the four major subdivisions proposed by Feuillet & MacDougal (2003). The results of Muschner et al. (2003; Annex 4a), on nuclear ribosomal internal transcribed spacers, (ITS-1 and ITS-2) and plastid trnL-trnF intergenic spacer, Yockteng (2003; Annex 4b), on chloroplast matK Yockteng & Nadot (2004; Annex 4c), on chloroplast-expressed glutamine synthetase (ncpGS), and Hansen et al. (2006; Annex 4d), on sequences analysis of the chloroplast (rpoC1 and trnL-trnT), confirm the clear separation of three clades corresponding to the new contours of subgenera Decaloba, Astrophea, and Passiflora. These three major clades correspond to cytogenetic groups, as they appear characterized by chromosome numbers of 2n = 12, 24, and 18, respectively. The results of Hansen et al. (2006) also support the small subgenus Deidamioides, whereas the two species that represent it in the study of Yockteng & Nadot (2004) are split in two widely divergent branches. In addition, the latter study indicated that four other small subgenera, Dysosmia (DC.) Killip, Tryphostemmatoides (Harms) Killip, Polyanthea (DC.) Killip, and Tetrapathea (DC.) Rchb., should also be recognized. More general problematic points, in these three genus-wide studies, are the monophyly of genus Passiflora, the position of subgenus Astrophea, and the low resolution at infrasubgeneric levels, particularly in the Passiflora clade, questioning the validity of supersections and series, where many ambiguities and inconsistencies still persist. In fact, such phylogenetic molecular studies 107 Chapter IV. Morphological diversity ______________________________________________________________________ cannot give us a definitive answer on the relationship between morphologically illdefined species and species groups. Actually, if two or more species within a series are only distinguished by the exact number of filament whorls in the corona, the color of one of these whorls, or the number of nectary glands on the petiole, we must first question the discontinuity in the variation described, at both morphological and genetic levels, and assess objectively the morphological basis of the classification. Despite the impressive morphological diversity described among Passiflora species, few studies have compared intra- and intersubgeneric, and intra- and interspecific variation with statistical tools. A first study was conducted by Villacís et al. (1998) on the most common species of subgenera Tacsonia and Manicata, on Colombian and Ecuadorian accessions. Floral traits were mostly represented in their set of 33 qualitative descriptors, and vegetative traits in the set of 28 quantitative descriptors. The former showed limited intraspecific variation and a consistent picture of interspecific relations, while the latter provided more information on intraspecific variation but a less consistent picture of the differences between species. The descriptor list was corrected, to take into account traits specific to subgenus Tacsonia, and augmented to 62 qualitative and 67 quantitative descriptors, giving a better balance between floral and vegetative traits. The sample was also amplified, to include more species and more origins, including similar materials from Venezuela, Colombia, Ecuador, Peru, and Bolivia (Coppens d’Eeckenbrugge et al., 2002). This study showed wide variation at both intra and interspecific levels, with a geographic component in the species best represented, and the overall picture was consistent with results obtained in the same subgenus with biochemical and molecular data (Segura et al., 2002, 2003, 2005). These convergent data led to the formal description of the cultigen P. tarminiana Coppens & Barney as a distinct species (Coppens d’Eeckenbrugge et al., 2001). The same descriptor list was used to study morphological variation in the three most common cultivated banana passion fruits and their hybrids, showing maternal effects in the hybrid phenotypes and confirmed the spontaneous introgression occurring between the wild P. mixta L. and the cultivated P. tripartita var. mollissima (Primot et al., 2005). With a descriptor list similar to the one used by Villacís et al. (1998), Medina et al. (2000) evaluated 25 accessions of the subgenera Passiflora, Manicata, Tacsonia, Dysosmia, and Decaloba. The qualitative descriptors allowed a clear separation between subgenera Passiflora, Tacsonia and Decaloba, with two groups in the latter. P. foetida 108 Chapter IV. Morphological diversity ______________________________________________________________________ L. (subg. Dysosmia) and P. manicata (Juss.) clustered with subgenera Passiflora and Tacsonia respectively. The quantitative descriptors allowed distinction of subgenera Tacsonia and Manicata, but showed no structure among species of the three other subgenera. A very detailed descriptor list was also used by Porter-Utley (2003) to study section (Killip) or supersection (Feuillet & MacDougal) Cieca of subgenus Decaloba, and more particularly the species complex around P. suberosa L. and P. coriacea Juss. Seventy quantitative traits were measured, 33 to 39 of which could be categorized and gathered with qualitative traits for neighbor joining cluster analyses of the different subsamples. Morphological data appeared consistent with molecular data in confirming monophyly of section/supersection Cieca, recognizing P. tridactylites Hook. f. and P. pallida L. as distinct species, and detecting occasional introgression of the latter with its close relative P. suberosa. On the other hand, there was considerable incongruence between molecular (ITS sequences) and morphological phylogenies, which was mostly attributed to smaller sample and intraspecific variation in the molecular data. A Brazilian Passiflora collection including ten species was characterized by Crochemore et al. (2003) with 22 qualitative and quantitative descriptors. The results showed clear differentiation between the two botanical forms of P. edulis. More recently, De Oliveira et al. (2005) tested a new morphometric method, based on leaf structures in a sample of ten Passiflora species. The method was very accurate in correctly differentiating among species, but two species were not consistently classified, P. foetida, a problematic species in all classifications, and P. miersii Mast. Although its potential must be further assessed on wider samples, the method is very promising for species identification, as it can be applied on sterile specimens. On the other hand, this advantage becomes a limitation for any comparison with classical taxonomy works, where flower structures play an essential role, so this method must be seen as a powerful additional tool. In conclusion, despite the obvious interest of an objective classification in a plant family of such morphological richness and complexity, only a few teams have developed and applied the necessary methodology. Except for the studies targeting specifically subgenus Tacsonia or section/supersection Cieca of subgenus Decaloba, their success has been limited by the size and diversity of their species sample. This limitation is not specific to morphological characterization. However, it tends to be much more severe than in molecular characterization, because of the need of field germplasm collections gathering species with variable climatic adaptations in one or very few places where they can develop until flowering. In the case of wild Passiflora accessions, a possible 109 Chapter IV. Morphological diversity ______________________________________________________________________ solution is in situ characterization, which then must be taken into account to avoid or reduce environmental bias in the analysis. The present study benefited from projects on diversity of Colombian Passifloraceae, including a component of collecting and establishing germplasm in field collections. As Colombia is the country with the highest Passiflora species diversity, both for wild and cultivated materials (Ocampo et al., 2007; Chapter II), a wide species sample could be studied, although practical limitations allowed describing accessions of only 51 species from eight subgenera, out of the 164 reported for the genus in the country. A few nonnative species were added, extending the sample to 60 species. We revised the descriptor lists used by Medina et al. (2000) and Coppens d’Eeckenbrugge et al. (2002), aiming at a better description of variation in subgenera other than Tacsonia (particularly Distephana, Decaloba, Astrophea and Tryphostemmatoides). Our goal was twofold, to test the utility of the revised set of descriptors over a wide range of Passiflora species, and to study morphological divergence among subgenera, species and accessions. IV.1.3. Materials and methods IV.1.3.1. Plant materials The morphological study was carried out in six experimental stations maintained by several institutions in Colombia (Andes region) (University of Caldas, the coffee grower federation research institute - Cenicafé, CIRAD/IPGRI, Passicol S.A). These germplasm collections were situated in La Italia (Caldas; 1,100 m), Tesorito (Caldas; 2,400 m), Paraguacito (Quindío; 1,250 m), El Moral (Valle del Cauca; 2,200 m), Tenerife (Valle del Cauca; 2,750 m) and El Tambo (Cauca; 1,800 m). The germplasm sample consisted of 261 individuals, representing 124 accessions and 60 species belonging to the genus Passiflora and nine Killip’s subgenera (Table 1 and Figure 2). The Colombian accessions were collected by Ocampo et al. (2007) in 2001-2004. Accessions from other countries were obtained from previously established collections. Geographic distribution was taken into account in the selection of accessions of a same species. Narrow endemics (e.g P. trinervia (Juss.) Poir.) are 110 Chapter IV. Morphological diversity ______________________________________________________________________ represented by one accession, and widespread species (e.g. P. edulis f. flavicarpa) by one accession per region. Three plants were grown from seeds for each accession, at a distance of 3 m between rows and 3 to 5 m within rows, according to adult plant size. The taxonomic treatment of Killip (1938), with emends by Escobar (1988a,b, 1989, 1994) and MacDougal (1994) is followed in this text. P. foetida L. P. arborea Spreng. P. coriacea Juss. ASTROPHEA DECALOBA DYSOSMIA P. vitifolia Kunth P. maliformis L. P. manicata (Juss.) Pers. DISTEPHANA PASSIFLORA MANICATA P. arbelaezii L. Uribe TRYPHOSTEMMATOIDES P. trinervia (Juss.) Poir. PSILANTHUS P. tarminiana Coppens & Barney TACSONIA Figure 2. Variation in shape and color among species from nine of Killip’s subgenera. 111 Chapter IV. Morphological diversity ______________________________________________________________________ IV.1.3.2. Data collection The descriptor list (Annex 5) was developed, in various stages, mostly by experts of Bioversity International (formerly IPGRI), CIRAD and CORPOICA. It was adapted to take into account the wide diversity in our collections. It included 43 quantitative and 84 qualitative descriptors, presented synthetically in Table 2. They were assessed on three individuals per accession, and five measures were taken for quantitative characters for each individual. The color characters were recorded with the Royal Colour Chart (Royal Horticultural Society, 2002). Quantitative fruit traits were not taken into account, as they are too often submitted to convergent selection processes, both in the wild and in cultivated species. IV.1.3.3. Analyses of quantitative variation Shape descriptors were computed as ratios of crude ones. Quantitative data were submitted to an analysis of variance to compare variation among and within subgenera, species, accessions, and individuals. To identify characteristics that mostly contributed to the variation among subgenera, we selected traits for which more than half of the variance was caused by variation at this level. When shape descriptors showed similar discriminating power, they were preferred over crude descriptors, to avoid giving too much importance to size components of variation. The selected descriptor set was submitted to a principal component analysis (PCA), carried out with the varimax normalized rotation option using the STATISTICA 6.0 software (Hill & Lewicki, 2006), retaining those factors with an eigenvalue superior to one, and the individuals were projected on the first three PCA axes. 112 Chapter IV. Morphological diversity ______________________________________________________________________ Table 1. List of accessions used in the present study. Taxonomy according to Killip (1938) and emends by Escobar (1988a,b, 1989) and MacDougal (1994). Subgenus / section / serie / specie Subgenus Astrophea (DC.) Masters, 1871 Section Euastrophea (Harms.) Killip Passiflora arborea Spreng. Passiflora emarginata Humb. & Bonpl. Passiflora sphaerocarpa Triana & Planch. Subgenus Decaloba(DC.) Rchb., 1828 Section Cieca (Medic.) Mast Passiflora coriacea Juss. Passiflora suberosa L. Section Decaloba (DC.) Mast Series Auriculatae Passiflora auriculata Kunth Series Lutae Passiflora filipes Benth. Series Miserae Passiflora misera Kunth Passiflora trifasciata Lemaire Series Punctatae Passiflora alnifolia Kunth Code No. ind. Country Locality Status arbCA emaCA emaVA sphVA 2 2 2 2 Colombia Colombia Colombia Colombia Manizales (Caldas) Manizales (Caldas) Yotoco (Valle Del Cauca) Cali (Valle Del Cauca) Wild, edible fruit Wild, edible fruit Wild Wild corCA corVA corVL cotTO subAN subCA subCAL subSA subVA 2 2 2 2 2 2 2 2 3 Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Palestina (Caldas) Cartago (Valle Del Cauca) Palmira (Valle Del Cauca) Ibagué (Tolima) Jerica (Antioquia) Manizales (Caldas) Manizales (Caldas) Barichara (Santander) Palmira (Valle Del Cauca) Wild Wild Wild Wild Wild Wild Wild Wild Wild aucCA 2 Colombia Victoria (Caldas) Wild filRI 2 Colombia Pereira (Risaralda) Wild misVA triEC 2 2 Colombia Ecuador Jamundí (Valle Del Cauca) Misahuallí (Napo) Wild Wild alnCA 2 Colombia Manizales (Caldas) Wild 113 Chapter IV. Morphological diversity ______________________________________________________________________ Passiflora bogotensis Benth. Passiflora biflora Lam. Passiflora cuspidifolia Harms Passiflora erythrophylla Mast. Passiflora magdalenae Triana & Planch. Series Sexflorae Passiflora sexflora Juss. Section Xerogona (Raf.) Killip Passiflora capsularis L. Passiflora rubra L. Section Pseudodysosmia (Harms.) Killip Passiflora adenopoda Moc. & Sessé ex D.C Section Hahniopathanthus (Harms.) Killip Passiflora guatemalensis S. Wats. Subgenus Dysosmia (DC.) Killip, 1938 Passiflora foetida var. gossypiifolia (Desv.) Mast. Passiflora foetida var. hispida (DC.) Killip ex Gleason Subgenus Distephana (Juss.) Killip, 1938 Passiflora vitifolia Kunth Subgenus Manicata (Harms) Escobar, 1988 Passiflora manicata (Juss.) Pers. alnNA alnSA bogCU bifCU bifTO cusBO 2 2 2 2 2 1 Colombia Colombia Colombia Colombia Colombia Colombia Chachagui (Nariño) Tona (Santander) Bogotá (Cundinamarca) Mariquita (Tolima) La Mesa (Cundinamarca) Duitama (Boyacá) Wild Wild Wild Wild Wild Wild eryBO magCO 2 2 Colombia Colombia Duitama (Boyacá) Victoria (Caldas) Wild Wild sexQU 2 Colombia Calarca (Quindío) Wild capAN capSA capVA rubCA rubQU 2 2 2 2 2 Colombia Colombia Colombia Colombia Colombia Jerico (Antioquia) Barichara (Santander) Cartago (Valle Del Cauca) Manizales (Caldas) Buenavista (Quindío) Wild Wild Wild Wild Wild adeQU adeTO 2 2 Colombia Colombia Buenavista (Quindío) Ibagué (Tolima) Wild, edible fruit Wild, edible fruit guaCA 2 Colombia Filadelfia (Caldas) Wild fotCH fotTO 2 3 Colombia Colombia Quibdó (Chocó) Armero (Tolima) Wild, edible fruit Wild, edible fruit vitCA vitTO 2 2 Colombia Colombia Victoria (Caldas) Ibagué (Tolima) Wild, edible fruit Cultivated, edible fruit manEC manSA manSA 1 2 2 Ecuador Colombia Colombia Baños (Tungurahua) Santander Santander Wild Wild Wild 114 Chapter IV. Morphological diversity ______________________________________________________________________ Subgenus Passiflora [= Granadilla (Medic.) Mast. 1871] Series Quadrangulares Passiflora alata Curtis Passiflora quadrangularis L. Series Digitatae Passiflora serratodigitata L. Series Tiliaefoliae Passiflora maliformis L. Passiflora ligularis Juss. Passiflora serrulata Jacq. Passiflora tiliifolia L. Series Laurifoliae Passiflora guazumaefolia Juss. Passiflora nitida Kunth Passiflora popenovii Killip manSA manVA 2 1 Colombia Colombia alaBR alaVA quaHU quaVA 3 3 3 2 Brasil Brasil Colombia Colombia serrBR 2 Brasil malAN malCA malVA malQU malHU malVA malVA malTO ligCA ligEC ligQU ligRI ligQU serrMA tilVA 2 3 2 2 2 3 1 2 3 1 3 1 2 2 2 Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Colombia Fredonia (Antioquia) Palestina (Caldas) Buga (Valle Del Cauca) Calarca (Quindio) Rivera (Huila) Tulua (Valle Del Cauca) La Unión (Valle Del Cauca) Ibagué (Tolima) Anserma (Caldas) Cuenca (Azuay) Salento (Quindio) Santa Rosa (Risaralda) Genova (Quindio) Plato (Magdalena) El Cerrito (Valle Del Cauca) Cultivated, edible fruit Wild, edible fruit Wild, edible fruit Wild, edible fruit Cultivated, edible fruit Wild, edible fruit Wild, edible fruit Wild, edible fruit Cultivated, edible fruit Cultivated, edible fruit Wild, edible fruit Wild, edible fruit Cultivated, edible fruit Wild, edible fruit Wild, edible fruit guzMA nitCH popCA popNA 2 1 2 2 Colombia Colombia Colombia Colombia Plato (Magdalena) Quibdó (Chocó) Timbío (Cauca) Chachagui (Nariño) Wild, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Series Serratifoliae 115 Santander El Cerrito (Valle Del Cauca) Wild Wild Paicol (Huila) Palmira (Valle Del Cauca) Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Wild, edible fruit Chapter IV. Morphological diversity ______________________________________________________________________ Passiflora bahiensis Klotzsch Series Incarnatae Passiflora cincinnata Mast. Passiflora edulis f. edulis Sims Passiflora edulis f. flavicarapa Degener Passiflora incarnata L. Series Kermesinae Passiflora lehmanni Mast. Passiflora smithii Killip Series Lobatae Passiflora caerulea L. Passiflora subpeltata Ortega Passiflora gibertii N.E. Brown Subgenus Psilanthus (DC.) Killip, 1938 Passiflora trinervia (Juss.) Poir. Subgenus Tacsonia (Juss.) Tr. & Planch, 1873 Section Colombiana Series Leptomischae Passiflora antioquiensis Karst. Passiflora flexipes Triana & Planch. Passiflora tenerifensis L.K. Escobar bahBR 3 Brasil cinBR cinBS edeCA edeCAL edeCC edePE edeQU edf-BR edfCA edfHU edfPE edfPR edfVA incUS 3 3 3 2 2 1 1 4 3 3 3 3 3 1 Brasil Brasil Colombia Colombia Colombia Peru Colombia Brasil Colombia Colombia Peru Peru Colombia U.S.A lehQU smiTO 2 2 caeFR supAN gibBR Salvador (Bahia) Cultivated, Edible fruit Anserma (Caldas) Miami (Florida) Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Colombia Colombia Buenavista (Quindío) Ibagué (Tolima) Wild Wild 3 3 2 France Colombia Brasil Montpellier (Herault) Santa Marta (Magdalena) Cultivated, edible fruit Wild Wild triCO 2 Colombia Salento (Quindío) Wild ant-AN antCA fleQU tenVA 2 2 2 2 Colombia Colombia Colombia Colombia Sta. Rosa Osos (Antioquia) Manizales (Caldas) Salento (Quindío) El Cerrito (Valle Del Cauca) Wild, edible fruit Wild, edible fruit Wild, edible fruit Wild, Edible fruit 116 Salamina (Caldas) Manizales (Caldas) Puracé (Cauca) Genova (Quindío) La Unión (Valle Del Cauca) Rivera (Huila) Chapter IV. Morphological diversity ______________________________________________________________________ Series Colombianae Passiflora adulterina L.f. Passiflora lanata (Juss.) Poir. Series Quindiensae Passiflora linearistipula L.K. Escobar Section Bracteogama Passiflora cumbalensis var. cumbalensis (H. Karst.) Harms Passiflora luzmarina Jorgensen Passiflora tarminiana Coppens & Barney Passiflora tripartita var. mollissima Holm-Nielsen & Jørgensen Passiflora tripartita var. tripartita Holm-Nielsen & Jørgensen Section Parritana Passiflora jardinensis L.K. Escobar Passiflora parirtae (Mast.) L.H Bailey Section Poggendorffia Passiflora pinnatistipula Cav. Passiflora x rosea (H.Karst.) Killip Section Tacsonia Passiflora mathewsii (Mast.) Killip tenVA 1 Colombia El Cerrito (Valle Del Cauca) Wild, Edible fruit aduBO lanBO 2 2 Colombia Colombia Duitama (Boyacá) Duitama (Boyacá) Wild Wild linCA 2 Colombia Manizales (Caldas) Wild cumEC cumNA luzEC luzEC tarAR tarBO tarCC tarEC tarPE tarVA tarVE tvmCU tvmEC tvmNA tvmPE tvmVA tvmVE tvtEC 2 1 1 1 3 2 2 4 1 3 3 4 1 2 1 3 1 2 Ecuador Colombia Ecuador Ecuador Argentina Colombia Colombia Ecuador Peru Colombia Venezuela Colombia Ecuador Colombia Peru Colombia Venezuela Ecuador Tulcán (Carchi) Pasto (Nariño) Loja (Loja) Loja (Loja) Castellar Boyacá Silvia (Cauca) Baños (Tungurahua) El Cerrito (Valle Del Cauca) Tachira Ambato (Tungurahua) Wild, edible fruit Wild, edible fruit Wild, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Cultivated, edible fruit Wild, Edible fruit jarAN parTO 2 2 Colombia Colombia Jardin (Antioquia) Herveo (Tolima) Wild Wild, edible fruit pinBO xroBO 1 1 Colombia Ecuador Boyacá Tuta (Boyacá) Wild, edible fruit Wild, Edible fruit matEC 1 Ecuador Cuenca (Azuay) Wild 117 El Cerrito (Valle Del Cauca) Tachira Cundinamarca Baños (Tungurahua) Pasto (Nariño) Chapter IV. Morphological diversity ______________________________________________________________________ Passiflora mixta L.f. Subgenus Trryphostemmatoides (Harms) Killip, 1938 Passiflora arbelaezii L. Uribe Passiflora gracillima Killip mixVA mixVA mixVA mixVA mixVA 2 3 3 1 1 Colombia Colombia Colombia Colombia Colombia El Cerrito (Valle Del Cauca) El Cerrito (Valle Del Cauca) El Cerrito (Valle Del Cauca) El Cerrito (Valle Del Cauca) El Cerrito (Valle Del Cauca) Wild, edible fruit Wild, edible fruit Wild, edible fruit Wild, edible fruit Wild, edible fruit arbCH graQU 1 1 Colombia Colombia Quibdó (Chocó) Salento (Quindío) Wild Wild 118 Chapter IV. Morphological diversity ______________________________________________________________________ IV.1.3.4 Cluster analyses on qualitative data According to the PCA results, certain quantitative variables were selected on the basis of their contribution to the axes, categorized and added to the qualitative dataset, provided that the corresponding information was not yet included in a purely qualitative descriptor. The resulting dataset was treated in two steps. A first set of qualitative variables was selected on the basis of their contribution to differentiation between subgenera, discarding those that show frequent variation at lower levels. A second set included all qualitative descriptors. Both sets were submitted to a neighbor joining cluster analysis (Saitou & Nei, 1987), using the coefficient of dissimilarity of Sokal & Michener and calculating bootstrap values from 100 replicates, with the DARwin 5.2 software (Perrier et al., 2003). This phenetic approach was preferred because of the relatively poor information on morphological evolution in Passiflora. Table 2. List of 127 descriptors used in the morphological characterization study. Organ Stem Tendril Stipule Petiole Leaf Qualitative characters (84) Habit Shape Pubescence Anthocyanin Presence Shape Pubescence Anthocyanin Presence Permanence Color Pubescence Shape Margin Anthocyanin Anthocyanin Pubescence Color Nectary shape Nectary stipe Heterophylly Lobe number Margin Base shape Apex shape Presence of acumen Code STHA STSH STPU STAN TEPR TESH TEPU TEAN SPPR SPPE SPCO SPPU SPSH SPMA SPAN PEAN PEPU PECO PENS PESN LEPO LELN LEMA LEBS LEAS LEPA Pubescence – adaxial Pubescente – abaxial Anthocyanin – lamina Anthocyanine – nerves Color – adaxial Heteroblasty Presence of laminar nectaries LEAX LEPB LEAL LEAN LECA LEPH LENL 119 Quantitative characters (43) Diameters Internode length Code STDI STIN Length Width Terminal arista length SPLE SPWI SPTA Length Distance from base to first gland Nectary number PELE PEDG PENM Margin serration density Angle between lateral lobes Central lobe length Right lobe length Central lobe width Distance between leaf sinus and petiole insertion Nectary number on lamina Nectary number on leaf margin LEMS LEAB LELC LERL LECL LESS LELA LENN Chapter IV. Morphological diversity ______________________________________________________________________ Peduncle Bract Flower Fruit Distribution of laminar nectaries Presence of marginal nectaries Distribution of margin nectaries Nectary shape Pubescence Anthocyanin Bifurcation LEDN LEPN LELM LENS PDPU PDAN PDBN Union Presence Permanence Pubescence Position of pubescence Color Anthocyanin Shape Margin Apex shape Marginal nectaries Corona type Corolla type Orientation Pubescence on corolla Hypanthium pubescence Dominant petal color Chlorophylla on exterior of sepal BRUN BRPR BRPE BRPU BRPP BRCO BRAN BRSH BRMA BRAS BRNM FLCY FLCT FLOR FLPU FLHP FLCP FLCS Keel-shaped sepals Sepal awn Presence of petals Union of sepals Color of filaments at base Color of filaments at apex Distribution of anthers Ovary pubescence Color of ovary Color of style Color distribution on styles Color of stigmas Color of androgynophore Color distribution on androgynophore Pubescence of androgynophore Limen margin Nectary chamber ring Hypantium type Internal color of hypanthium Chlorophylla on exterior of hypanthium Anthocyanin on exterior of hypanthium Nectaries on sepals Dominant sepal color Anthocyanin on exterior of sepals FLKS FLSA FLPP FLUS FLCB FLCA FLDA FLOP FLCO FLCS FLDS FLCG FLCN FLDN Type Shape FLPN FLML FLNR FLHY FLCI FLCE FLAE FLNS FLCP FLAP FRTY FRSH 120 Length Diameter Pedicel length Length to first bifurcation Length to second bifurcation Length Width PDLE PDDI PDPL PDLF PDBS BRLE BRWI Orientation (in degrees to vertical) Petal length Petal width Sepal length Sepal width Diameter of nectary chamber Hypanthium diameter above nectary chamber Hypanthium diameter -distal Flower length Hypanthium length Length of nectary chamber Number of corona series Filament length Staminal filaments length Ovary length Style length Gynophore length FLOG FLPL FLPW FLSL FLSW FLNC FLHD Androgynophore length Operculum length Limen length FLAL FLOP FLLL FLHS FLLE FLHL FLCN FLNS FLFL FLSF FLOL FLSL FLGL Chapter IV. Morphological diversity ______________________________________________________________________ IV.1.4. Results and Discussion IV.1.4.1. Quantitative variation As expected, a very high variability was observed among the 124 accessions in the field. Table 3 gives the mean values and coefficients of variation for the whole sample and for the different subgenera. Coefficients of variation appear generally superior for subgenus Decaloba, as compared with subgenera Passiflora and Tacsonia that have comparable representation in number of species. This higher relative variation is much more obvious for inflorescences (18 out of 20 traits), and shape ratios (10 out of 13 ratios), of higher taxonomical importance, than for vegetative parts (9 of 21 descriptors), suggesting a higher interspecific differentiation in this subgenus. Figure 3 shows the relative variance components for 57 quantitative descriptors. All the descriptors present a residual variance under 25% and then a high repeatability. Many descriptors appear to be efficient in discriminating among subgenera. Thus, the proportion of variance at this level exceeds 50% for 26 of them, including stem diameter, leaf margin indentation, leaf length, numbers of nectar glands on leaf margins and petiole, diameter of peduncle, length of first and second order peduncle segments, dimensions and shape of bracts, length of flower, hypanthium, sepals and petals, nectary chamber, crown longest series, androgynophore, stamens and ovary, relative constriction above nectary chamber, and bract/hypanthium length ratio. At the species level, 28 characters are more important. They are related to dimensions of stipule, lobation (angle between lateral nerves, shape of central lobe, length of lateral lobe, distance between leaf sinus and petiole insertion), number of laminar nectary glands, position of petiolar nectary glands, length of peduncle, diameter of hypanthium, length of gynophore, shape of petals and sepals, androgynophore/hypanthium length ratio (defining protrusion of gynoecium and androecium) and pedicel/peduncle ratio. At lower levels, variance between accessions and plants rarely contributes more than 20% of the total. 121 Chapter IV. Morphological diversity ______________________________________________________________________ Table 3. Mean values and coefficients of variation for the whole sample and for the different subgenera. LELC LERL LECL LESS LELA LENN PELE PEDG PENM BRLE BRWI PDLE PDDI PDPL PDLF PDBS FLNS FLOG FLPL FLPW FLSL FLSW FLNC FLHD DHS FLHS FLLE FLCN FLFL FLSF FLOL FLSL FLGL FLAL FLOP FLLL PEDG/PELE LECL/LELC BRWI/BRLE FLPW/FLPL FLSW/FLSL FLNC/FLHD FLLE/FLHL FLCN/FLHL FLAL/FLHL PDDI/PDPL PDPL/PDLE BRLE/FLHL FLGL/FLAL LECL/LECL 340 30.00 0.00 285.54 0.52 8.69 0.55 7.48 0.87 2.72 0.94 6.81 0.29 105.42 0.40 96.95 0.21 83.91 0.41 37.59 0.39 21.72 0.55 0.00 0.00 22.57 0.34 9.25 0.57 6.09 0.37 36.30 0.29 15.70 0.36 86.71 1.43 2.50 0.00 8.93 0.40 0.00 0.00 1.40 0.80 10.99 2.02 44.28 0.25 19.39 0.23 46.91 0.24 18.65 0.28 15.07 0.17 10.21 0.23 12.00 0.21 103.36 0.17 71.01 0.28 10.17 0.25 2.10 0.80 14.86 0.32 11.60 0.17 13.61 0.24 4.37 0.67 82.69 0.21 6.61 0.24 1.53 0.36 0.41 0.45 0.39 0.30 0.43 0.25 0.44 0.16 0.41 0.23 1.53 0.23 1.54 0.26 0.15 0.34 1.23 0.32 0.08 0.70 0.29 0.77 0.54 0.39 0.06 1.49 0.39 0.30 310 10.81 0.24 98.50 1.05 5.34 0.47 2.46 1.88 0.18 3.40 0.21 3.50 80.38 0.39 59.80 0.38 65.31 0.35 33.50 0.53 45.76 0.38 4.09 1.10 0.00 25.26 0.69 15.25 1.18 0.81 1.22 2.63 1.68 1.28 2.73 19.74 0.63 1.51 0.02 4.42 0.71 0.00 0.00 2.00 1.10 72.73 0.96 7.98 0.76 2.76 0.93 14.62 0.38 5.62 0.39 8.55 0.34 8.73 0.36 9.59 0.36 11.52 0.31 6.32 2.29 3.64 0.34 7.68 0.40 4.26 0.34 3.45 0.40 4.62 0.31 0.93 0.26 6.25 0.36 2.12 0.49 0.87 0.70 0.60 0.67 0.58 0.44 0.18 1.41 0.24 0.82 0.42 0.38 0.99 0.08 5.85 0.71 1.89 0.78 3.36 0.79 0.11 0.65 0.40 1.36 0.92 1.28 0.17 0.41 0.59 0.37 50 230.00 0.25 54.67 0.32 4.45 0.16 2.86 0.29 0.00 0.00 0.00 247.27 0.26 0.00 123.49 0.31 0.00 0.00 0.00 26.20 0.36 26.38 0.37 2.00 0.00 2.10 0.10 1.30 0.00 17.88 0.35 2.92 0.11 13.44 0.42 23.30 0.21 23.30 0.21 1.00 0.00 135.00 0.20 24.08 0.19 8.44 0.35 24.47 0.16 9.36 0.27 7.59 0.17 7.55 0.17 8.28 0.21 23.24 0.18 5.82 0.25 4.92 0.11 13.39 0.19 6.36 0.14 4.98 0.14 6.76 0.33 1.00 0.00 12.40 0.30 1.84 0.17 0.60 0.34 1.01 0.03 0.50 0.12 0.62 0.07 0.34 0.23 0.38 0.25 1.01 0.03 4.25 0.30 0.89 0.24 2.21 0.31 0.18 0.25 0.76 0.43 0.38 0.21 0.09 0.39 0.76 0.26 35 30.00 0.00 321.63 0.31 11.65 0.12 18.64 0.12 3.87 0.19 8.63 0.06 116.74 0.07 112.82 0.24 117.99 0.27 52.39 0.23 41.35 0.22 0.00 0.00 28.63 0.21 7.89 0.36 9.20 0.29 42.97 0.10 26.04 0.09 65.70 0.25 2.50 0.00 13.70 0.12 0.00 0.00 5.00 0.00 158.00 0.14 42.27 0.08 14.65 0.09 44.36 0.08 14.17 0.13 15.85 0.09 10.93 0.07 14.64 0.11 42.44 0.07 13.03 0.24 8.58 0.08 5.82 0.23 12.19 0.12 8.20 0.10 12.14 0.12 2.70 0.19 29.11 0.09 10.37 0.08 1.25 0.16 0.28 0.32 0.47 0.14 0.61 0.07 0.35 0.10 0.32 0.16 1.46 0.10 3.42 0.21 0.69 0.22 2.35 0.24 0.04 0.27 0.22 0.27 3.52 0.28 0.09 0.23 0.50 0.12 25 15.00 0.00 40.04 0.14 2.70 0.26 5.35 0.10 6.22 0.04 5.00 0.00 92.24 0.03 73.22 0.04 45.47 0.04 42.16 0.05 41.05 0.08 0.00 0.00 21.92 0.16 0.00 0.00 20.92 0.11 16.20 0.27 49.39 0.16 1.50 0.00 2.60 0.00 0.00 0.00 4.20 0.84 142.00 0.23 18.90 0.14 7.92 0.13 19.47 0.11 7.22 0.12 8.32 0.05 8.40 0.05 8.40 0.05 12.88 0.08 8.35 0.10 2.48 0.06 11.20 0.33 4.60 0.13 3.71 0.20 6.44 0.09 1.00 0.02 8.46 0.11 2.32 0.07 1.40 0.07 0.00 0.58 0.04 0.78 0.28 0.42 0.08 0.37 0.09 0.99 0.03 1.56 0.17 0.30 0.10 1.03 0.19 0.03 0.16 0.05 0.16 2.55 0.22 0.12 0.12 0.58 0.04 122 20 35.00 0.00 61.30 0.14 6.20 0.17 1.00 0.00 0.00 6.10 0.05 77.90 0.10 115.10 0.09 96.30 0.13 49.30 0.11 44.00 0.02 0.00 7.00 0.15 30.85 0.18 0.50 0.00 2.00 0.00 26.70 0.19 8.70 0.18 46.85 0.13 2.50 0.00 4.92 0.07 0.00 0.00 2.00 0.00 110.00 0.42 56.25 0.06 13.58 0.20 59.60 0.05 16.60 0.07 13.90 0.13 11.90 0.15 12.75 0.06 55.75 0.07 12.07 0.09 6.10 0.26 24.10 0.05 14.20 0.12 10.76 0.06 13.20 0.06 2.72 0.09 42.08 0.08 10.00 0.06 3.55 0.14 0.02 0.22 0.43 0.11 0.34 0.32 0.24 0.14 0.28 0.11 1.17 0.05 4.64 0.08 0.50 0.22 3.50 0.08 0.05 0.13 0.11 0.15 2.25 0.27 0.06 0.06 0.47 0.14 10 5.00 0.00 22.91 0.10 0.46 0.61 0.25 0.21 0.00 0.00 0.00 32.11 0.38 0.00 22.38 0.15 0.00 0.00 0.00 19.09 0.14 19.09 0.14 2.00 0.00 1.00 0.00 0.50 0.00 23.86 0.10 0.80 0.00 8.24 0.05 26.50 0.06 16.50 0.10 2.00 0.00 0.00 13.54 0.19 3.56 0.27 14.51 0.29 5.51 0.34 7.10 0.14 6.81 0.14 6.81 0.14 8.30 0.11 4.96 0.16 3.80 0.18 5.83 0.11 2.46 0.15 2.85 0.07 3.63 0.07 0.51 0.06 4.86 0.04 0.84 0.06 0.65 0.08 1.00 0.00 0.76 0.26 0.50 0.00 0.26 0.16 0.38 0.17 1.05 0.13 1.70 0.16 0.77 0.12 1.00 0.12 0.03 0.11 0.35 0.10 0.21 0.15 0.11 0.08 0.16 0.08 10 9.60 0.05 50.80 0.04 8.70 0.08 1.00 0.00 0.00 0.00 33.10 0.08 116.50 0.01 92.70 0.03 18.70 0.08 89.40 0.01 6.00 0.00 0.00 24.70 0.05 0.00 0.00 8.60 0.10 2.70 0.18 99.30 0.01 1.11 0.05 3.30 0.15 0.00 0.00 1.00 0.00 172.50 0.03 11.20 0.08 4.90 0.06 39.10 0.02 7.00 0.13 10.00 0.08 8.70 0.06 7.50 0.07 162.00 0.02 112.90 0.02 29.30 0.07 2.90 0.25 8.30 0.06 11.90 0.06 5.10 0.14 9.20 0.11 150.50 0.01 4.20 0.10 0.00 0.00 0.16 0.08 0.32 0.24 0.44 0.11 0.18 0.13 1.16 0.14 1.44 0.03 0.26 0.08 1.33 0.02 0.01 0.05 0.03 0.15 0.08 0.10 0.06 0.12 0.43 0.11 1304 37.96 1.11 235.09 0.76 9.82 0.83 5.59 1.22 1.03 1.96 3.89 0.92 75.40 0.75 103.05 0.46 63.42 0.72 54.96 0.69 25.74 0.83 1.02 2.79 0.11 8.10 31.99 0.66 16.65 0.98 3.27 0.85 25.31 0.75 13.93 0.81 49.56 1.41 2.30 0.24 7.81 0.54 1.10 4.64 1.02 4.68 3.10 0.91 73.71 0.94 30.28 0.55 10.86 0.66 33.51 0.46 13.11 0.50 12.33 0.32 10.82 0.34 13.30 0.39 44.56 0.88 26.47 1.16 6.48 0.57 17.37 1.02 10.13 0.50 7.81 0.47 9.92 0.43 2.34 0.94 30.69 1.14 3.88 0.67 1.38 0.65 0.52 0.64 0.52 0.42 0.47 0.53 0.33 0.42 0.40 0.28 1.17 0.26 3.07 0.87 0.74 1.32 1.75 0.95 0.09 0.75 0.30 1.10 1.67 0.92 0.14 0.80 0.52 0.42 Organ Stem 504 44.01 0.15 321.40 0.55 14.45 0.73 5.98 1.24 0.13 5.77 4.26 0.81 58.09 1.17 120.92 0.26 51.21 1.06 75.49 0.56 15.37 1.16 0.00 0.00 44.24 0.58 23.91 0.75 2.90 0.58 33.91 0.50 21.50 0.47 44.50 0.45 2.66 0.12 8.63 0.41 0.00 0.00 5.00 0.00 114.94 0.45 34.58 0.28 10.36 0.43 36.15 0.26 14.71 0.34 13.32 0.26 13.04 0.30 17.39 0.29 27.00 0.23 11.97 0.31 5.56 0.32 35.39 0.44 11.08 0.25 8.30 0.30 11.11 0.25 1.91 0.58 11.36 0.29 2.75 0.46 1.65 0.64 0.56 0.53 0.59 0.37 0.64 0.22 0.30 0.24 0.41 0.24 1.04 0.13 2.33 0.21 0.47 0.22 1.01 0.35 0.08 0.77 0.23 0.63 2.89 0.47 0.18 0.68 0.58 0.44 Leaf LEMS LEAB Passiflora Tacsonia Decaloba Astrophea Manicata Dysosmia Distephana Thryphostemmatoides Psilanthus Total sample Bract SPWI SPTA n Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Mean CV Flower STIN SPLE Level Shape relation Factor Descriptors STEDI Chapter IV. Morphological diversity ______________________________________________________________________ Figure 3. Relative variance components for 57 quantitative descriptors. Bold characters are used for traits displaying more than 50% of variance among subgenera. IV.1.4.2. Correlations and principal components analysis (PCA) A Pearson correlation (r) matrix shows high coefficients, between 0.71 and 0.97, among dimensions of certain floral components, such as petals and sepals, hypanthium, nectary chamber, operculum, androgynophore, stamens and ovary. The length of the corona is correlated with the upper hypanthium diameter (r = 0.75). Bract dimensions are moderately correlated with these traits (0.33 to 0.65, but 0.75 with ovary length). Stem diameter is correlated with peduncle branching (r = 0.73), which is due to the association of these traits in representatives of subgenus Astrophea. From the 26 quantitative descriptors showing high variation at the subgenus level, 24 were selected for the PCA, discarding two of them to avoid redundancy between shape ratios and the original traits. Five principal components were retained, that represent 84% of total variation (Table 4). The first one (32%) is very clearly associated with flower length (hypanthium, nectary chamber, androgynophore) and secondarily with the 123 Chapter IV. Morphological diversity ______________________________________________________________________ constriction of the floral cup above the nectary chamber. The second one (27%) is associated with flower width (length of bracts and length of corolla and corona elements) and bract shape. The third one (14%) is associated with peduncle branching, stem width and leaf length, which relates it clearly with variation between subgenus Astrophea, and secondarily subgenus Tryphostemmatoides, and all other subgenera. The fourth one (5%) is associated with the number of nectaries on leaf margin, which essentially relates it to the differentiation of subgenus Distephana. The last one (5%) is correlated only with leaf serration. Figure 4 presents the accessions in the three first axes, showing a clear grouping by subgenus. The representatives of subgenus Tacsonia are placed on the right along the first axis, in relation to their long and wide flowers. A few accessions, with shorter flowers (P. luzmarina Jørgensen and P. pinnatistipula Cav.), appear closer to the origin of this axis, together with P. manicata individuals. P. trinervia, of subgenus Psilanthus, is placed even further on the right, thanks to its very long floral tube, however it is clearly separated on the second axis by its much narrower flowers and minute setaceous bracts. On the left side, subgenera Passiflora and Decaloba are not differentiated by the flower length axis, but by the second, flower width related, axis. At the extremes of this second axis, we find the accessions of the large-flowered P. alata Curtis and P. quadrangularis (section Quadrangulares of subgenus Passiflora), on one side, and those of the small-flowered P. arbelaezii and P. gracillima Killip of subgenus Tryphostemmatoides, on the other side. As expected, the third axis clearly differentiates subgenera Astrophea and Tryphostemmatoides. On the whole, subgenera Passiflora, Tacsonia, Decaloba, Psilanthus, Astrophea and Tryphostemmatoides are clearly separated in the main tridimensional space. As expected, P. manicata (subgenus Manicata) takes an intermediate position between subgenera Passiflora and Tacsonia. This species not only combines morphological traits typical of both subgenera, but intermediate ecoclimatic requirements as well, as it may be found at lower elevations than tacsos, but higher elevations than representatives of subgenus Passiflora. The representatives of subgenus Tacsonia that come closest to P. manicata are P. pinnatistipula and P. luzmarina, two tacsos with relatively shorter floral tubes. The former is also differentiated by a filamentous corona, instead of the typical reduced tacso coronas. Another species taking a particular position is P. foetida, of Killip’s subgenus Dysosmia, placed near both subgenera Passiflora and Decaloba, but closer to 124 Chapter IV. Morphological diversity ______________________________________________________________________ the former. Subgenus Dysosmia is considered a supersection of subgenus Passiflora in the Feuillet & MacDougal classification. Table 4. Factor loadings from principal component analysis (varimax normalized rotation) on 24 quantitative descriptors. Principal components Descriptors STDI LEMS LELC LENN PENM PDDI PDLF PDBS BRLR FLPL FLPW FLSL FLSW FLLE FLHL FLCN FLFL FLSF FLOL FLSL FLAL FLOP BRWI/BRLE FLNC/FLHD Expl.Var Prp.Totl 1 2 3 4 5 -0.088 0.353 0.019 -0.017 0.648 0.135 -0.054 -0.054 0.296 0.501 0.713 0.577 0.520 0.960 0.904 0.824 -0.521 0.529 0.677 0.454 0.964 0.711 -0.160 0.700 0.215 0.365 0.571 0.040 0.407 0.847 -0.136 -0.113 0.723 0.774 0.595 0.716 0.754 0.179 0.063 0.208 0.723 0.651 0.658 0.693 0.064 0.236 0.718 0.056 -0.892 0.152 -0.686 0.011 0.013 -0.287 -0.943 -0.973 0.186 -0.021 -0.024 0.036 0.034 0.045 0.086 0.026 0.034 0.082 0.076 0.067 0.047 0.099 -0.221 0.042 -0.023 0.041 0.009 0.962 -0.055 0.010 -0.007 -0.006 -0.091 0.235 0.095 0.243 0.074 0.017 -0.100 -0.080 -0.007 0.097 0.061 0.073 0.033 0.363 -0.131 0.043 0.041 0.759 -0.195 0.014 0.158 0.145 -0.044 -0.046 0.240 0.042 0.028 -0.028 -0.108 0.090 0.119 -0.137 -0.329 0.171 -0.049 0.263 0.099 0.214 0.241 0.368 7.610 0.317 6.496 0.271 3.336 0.139 1.256 0.052 1.238 0.052 125 Chapter IV. Morphological diversity ______________________________________________________________________ Figure 4. Tridimensional plot of the scores of Passiflora accessions for the first three components of quantitative variation. Colors refer to subgeneric classification. IV.1.4.3. Qualitative variation among and within subgenera Our first attempt to reduce the number of qualitative descriptors led us to retain 32 of them on the basis of their potential to discriminate among subgenera. The criterion was that the descriptor appears monomorphic or shows a highly dominant condition at least in one subgenus, while polymorphic among subgenera. Four quantitative descriptors were categorized and added because of their high correlations with the principal components of quantitative variation. Thus the first component was represented by androgynophore length, the second one by sepal length, the third one by stem diameter and leaf length. The fourth and fifth ones were not included to avoid redundancy with very similar qualitative descriptors. Table 5 synthesizes the observations for these descriptors. The species of subgenus Astrophea exhibit the highest number of unique/rare traits, including tree habit, wide stems of irregular section, very long leaves, absence of tendrils, short triangular stipules, dorsal scar-like nectaries (appressed against or near petiole), branched peduncles, bright-yellow sickle-sword-shaped corona filaments, and tricostate ovaries. Unique and rare traits of the two species of subgenus 126 Chapter IV. Morphological diversity ______________________________________________________________________ Tryphostemmatoides include peduncle branching (shared with species of subgenus Astrophea and P. sexflora Juss. of subgenus Decaloba), the presence of tendrils at the axil of the peduncles, and the retuse leaf apex (unique in our sample, although this trait can be observed in individuals of P. emarginata Humb. & Bonpl.). P. vitifolia Kunth (subgenus Distephana) is differentiated by its tubular corona, formed by the partial fusion of its elements and conspicuous nectary glands on leaf sinus and bracts, P. foetida (subgenus Dysosmia) only by its pinnatisect bracts, and P. trinervia (subgenus Psilanthus) by the absence of a limen. Most of these traits are typical for each of these subgenera, ensuring that they will not bias the cluster analysis in terms of subgeneric classification. Subgenera Decaloba, Passiflora, Tacsonia and Manicata do not show unique traits, however they can be separated by clear segregations in nonexclusive traits. Thus, in subgenus Decaloba, the presence of nectary glands in the lamina is only shared with subgenus Psilanthus, the flat hypanthium with subgenus Tryphostemmatoides, the relatively small flower size with subgenera Astrophea, Dysosmia, and Tryphostemmatoides. On the other side, subgenera Passiflora, Distephana, Manicata and Tacsonia share wide flowers and the general presence of petiolar nectaries. The last three and subgenus Psilanthus also share large red or pink corollas and long tubular flowers (long androgynophores), typical of hummingbirdpollinated species. Floral tube length reaches extreme values in Psilanthus and Tacsonia, with the exceptions of P. pinnatistipula and P. luzmarina, in relation to their specific pollinator Ensifera ensifera Boissoneau, the sword-billed hummingbird. In addition they present reduced coronas of short filaments or tubercles, generally in one row only, while two-row coronas are most common in subgenera Decaloba, Tryphostemmatoides, Astrophea, and Distephana, and highly complex coronas (more than three rows) are typical in subgenera Passiflora, Manicata and Dysosmia. Bracts are foliaceous in subgenera Passiflora and Tacsonia. Fruit shape is generally globose to short ovate in subgenera Astrophea, Decaloba, Tryphostemmatoides, Dysosmia, Passiflora and Distephana, and oval to fusiform in subgenera Tacsonia, Manicata, and Psilanthus. Fruit color would also be interesting, with a particular frequency of blackish fruits in subgenus Decaloba, however this descriptor could not be observed in all species. 127 Chapter IV. Morphological diversity ______________________________________________________________________ Table 5. Variation for 32 qualitative and four categorized quantitative descriptors in the different subgenera sampled. Astrophea Feuillet & MacDougal (2003) Killip (1938), Escobar (1988), MacDougal (1994) Species/accession number Main pollinators Decaloba Astrophea n= 3/4 medium (honey) bees Decaloba n= 17/31 small to large bees - wasps 2n = 24 1.000 - 2.200 m Chromosome number Altitudinal range Deidamioides Passiflora Passiflora n= 19/44 large (carpenter) bees 2n = 12, 22, 24, 36 Psilanthus n= 1/1 sword-billed hummingbird 2n = 12 Distephana n= 1/2 hummingbirds Dysosmia n= 1/2 bees 60 - 2.700 m 2.600 - 3.200 Manicata n= 1/1 hummingbirds 2n = 18, 20, 22 Tacsonia n= 16/36 sword-billed hummingbird 2n = 18 2n = 18 2n = 18 20 - 2.400 m 50 - 1.200 m Tryphostemmatoides n= 2/2 bees 2n = 18 2n = 12 30 - 1.200 m 2.100 - 3.700 m 1.900 - 2.500 m 50 - 2.000 m Descriptors Habit Stem section Secondery xylema Tendril position Stipule Stipule nectaries (conspicuous) Leaf lobation Leaf base tree irregular present absent short triangular absent one cuneate-rounded Leaf apex Leaf margin Laminar nectaries Leaf margin nectaries (conspicuous) obtuse/acute entire absent absent Petiolar nectaries absent Dorsal nectaries Peduncle branching Bract shape present present linear Nectary on bract Flower orientation Corolla shape Dominant corolla color Corona type Corona filaments Number of corona series absent erect reflex white filamentous sickle-sword-shaped free uniseriate Corona color-clear (longest row) Corona color-darkest (longest row) white yellow Petals present Sepal awn Nectar chamber ring Hypanthium Limen Ovary shape Fruit shape absent present campanulate present tricostate globose Categorized (from quantitatives traits) Stem diameter Leaf length vine terete/angular absent axillary setaceous/linear/foliaceous-aristate absent three cuneate-rounded/cordate/ peltate (P. coriacea and P. guatemalensis ) rounded/obtuse/acute entire/serrate (P. adenopoda ) present/absent (five species)b absent/sinus/ all margin (P. adenopoda ) absent/orbicular (P. adenopoda )/ cylindrical (P. coriaceae and P. suberosa ) /auriculate (P. auriculata ) absent absent/present (P. sexflora ) setaceous/linear/ foliaceous (P. adenopoda , P. guatemalensis ) /absent (four species)* absent erect/intermediate/pendular intermediate/campanulate white filamentous linear free biseriate/ uniseriate (P. guatemalensis and P. adenopoda )/ triseriate (P. filipes and P. magadalenae ) white/purple/ yellow (P. guatemalensis ) present/ absent (P. coriaceae and P. suberosa ) absent/present absent flat present globose globose/ elongate (P. rubra and P. capsularis ) ≥ 120 mm Androgynophore length b ≤ 20 mm < 23 mm vine terete absent axillary linear present three cordate vine terete absent axillary foliaceous-aristate present three cordate vine terete/angular absent axillary setaceous/linear/foliaceous-aristate absent one/three cuneate-rounded/cordate vine angular absent axillary foliaceous-aristate absent three cordate vine terete absent axillary and peduncle setaceous absent one rounded acute/very acute entire present absent acute serrate absent sinus acute serrate absent absent acute/very acute serrate absent absent/sinus (P. jardinensis ) obtuse to very acute serrate absent absent retuse entire absent leaf base absent rounded to very acute entire/serrate absent absent/sinus lateral lobe (P. edulis f. edulis ) all except auriculate orbicular absent linear/orbicular linear absent absent absent linear absent absent foliaceous absent absent linear absent absent pinnatisect absent absent foliaceous absent absent foliaceous absent present linear absent pendular campanulate red (pink) filamentous linear free uniseriate absent/present erect/intermediate/pendular intermediate/reflex white/red filamentous linear free pentaseriate present erect reflex red filamentous linear fused at base biseriate absent erect intermediate/reflex white filamentous linear free 3- to 5-seriate absent erect/intermediate/pendular campanulate/intermediate/reflex red (including pink or orange) tuberculous/ filamentous, linear free uni- to 5-seriate absent erect reflex red filamentous linear free pentaseriate absent erect reflex white filamentous linear free biseriate white white white purple white (P. guazumaefolia ) present red red white purple purple purple white white present present white/purple purple white (P. jardinensis ) present present present absent absent tubular absent globose elongate present present campalunate present globose globose/ elongate (P. quadrangularis ) present present tubular present globose globose present present campanulate present globose globose present present tubular present globose elongate/ globose (P. pinnatistipula ) present present tubular present globose elongate absent absent flat present globose globose ≤ 20 mm ≥ 30 mm < 60 mm < 190 mm ≥ 30 mm < 60 mm < 190 mm ≤ 20 mm ≥ 30 mm < 60 mm < 190 mm ≥ 30 mm < 60 mm < 190 mm ≤ 20 mm > 30 mm ≤ 30 mm (six species)ª < 23 mm > 30 mm ≤ 30 mm > 30 mm ≤ 30 mm > 25 mm < 23 mm > 30 mm ≤ 30 mm (P. luzmarina and P. mathewsii ) > 25 mm < 190 mm > 30 mm < 23 mm b vine terete/angular/winged absent axillary setaceous/linear/foliaceous-aristate absent/present (P. maliformis ) one/three/more cuneate-rounded/cordate present > 215 mm < 190 mm < 190 mm (P. sphaerocarpa ) ≤ 30 mm ≤ 30 mm Sepal length vine angular absent axillary setaceous absent three cordate > 25 mm b b b a < 190 mm a > 25 mm a < 23 mm P. capsularis *, P. coriacea*, P. rubra *, P. suberosa *, P. sexflora , P. guatemelensis , P. bahiensis , P. edulis f. edulis , P. guazumaefolia , P. smithii , P. subpeltata , P. serrulataa. 128 a < 190 mm a Chapter IV. Morphological diversity ______________________________________________________________________ Certain species show unusual trait combinations in their subgenus. This is particularly true in subgenus Decaloba. Thus, P. adenopoda Moc. & Sessé ex DC. shows foliaceous bracts, serrate leaf margins, with conspicuous nectary glands, orbicular petiolar glands, and an uniseriate corona. P. guatemalensis S. Watson also shows foliaceous bracts and glandless leaves, plus peltate leaves and a yellow uniseriate corona. P. sexflora shows multiple peduncles. P. coriacea, P. suberosa, P. capsularis L. and P. rubra L. lack bracts; in addition, the first two present petiolar nectaries, while the last two lack such glands in all their parts and produce an elongate fruit. IV.1.4.4. Cluster analysis on the reduced descriptor list Figure 5 presents the dendrogram obtained from these observations on the first set of descriptors. The four best-represented subgenera, Passiflora, Tacsonia, Astrophea, and Decaloba, are supported by the analysis. Their placement on the dendrogram shows a polarization of the latter according to several traits. On one side, we find subgenera Passiflora, Distephana, Tacsonia, and Manicata, i.e. species producing large flowers and fruits and very generally bearing petiolar nectaries, with a base chromosome number of n = 9. They also share foliaceous bracts, with the relative exception of P. vitifolia, whose long bracts appear linear. They are further divided between the carpenter bee-pollinated species (subgenus Passiflora) and the hummingbird-pollinated species of subgenus Tacsonia, P. vitifolia, and P. manicata. The consistency of this subclassification compensates for the low associated bootstrap values. The placement of P. manicata in the Tacsonia cluster supports the gathering of the Andean subgenera Tacsonia and Manicata in a same infrageneric taxon, as in the classification of Feuillet & MacDougal (2003). On the other hand, P. vitifolia is a good representative of the uniform subgenus Distephana, so its position does not support its downgrading to a supersection of subgenus Passiflora, also proposed by these authors. On the other side of the tree, we find subgenera whose species produce small to medium flowers and fruits, with relatively simple coronas of generally two rows of filaments (rarely one or three), where petiolar nectaries are rare, with a base chromosome number of 12 for tree species and 6 for the others. As could be expected from the number of their rare traits, the subgenus Astrophea species of our sample appear very uniform, and well separated in a very distant cluster. P. trinervia (subgenus Psilanthus) is placed on another long branch, inserted at the same position. A third, much larger, cluster is constituted by all the species of subgenera Tryphostemmatoides and Decaloba, but P. adenopoda. This 129 Chapter IV. Morphological diversity ______________________________________________________________________ species is placed on a well-separated branch, inserted in an intermediate position between the Passiflora-Distephana-Tacsonia-Manicata clusters and the AstropheaDecaloba-Tryphostemmatoides clusters. Although the branch bearing the representatives of subgenus Tryphostemmatoides, P. gracillima and P. arbelaezii, is relatively long, it is clearly inserted within subgenus Decaloba, suggesting that the qualitative morphological differentiation of subgenus Tryphostemmatoides is fragile, which is consistent with the very low number of traits supporting it (Table 5), but contrasting with the PCA results on quantitative traits. The position of P. adenopoda may look surprising, as it is not consistent with either classification (Decaloba section Pseudosysosmia for Killip, supersection Bryonioides for Feuillet & MacDougal), however, this species showed several unusual features as compared to Decaloba as a group. P. foetida (subgenus Dysosmia), takes a very similar position, that is clearly more consistent with its classification in subgenus Dysosmia by Killip (1938) than with its inclusion in subgenus Passiflora by Feuillet & MacDougal (2003). Interestingly, these two problematic species materialize the separation between the two cytogenetic groups in our tree. Indeed, chromosome counts for P. adenopoda give 2n = 12 (MacDougal, 1994), as in most species of subgenus Decaloba, while those for P. foetida vary between 2n = 18, 20, and 22 (Yockteng & Nadot, 2004; De Melo et al., 2001). According to de Melo et al. (2001) and De Melo & Guerra (2003), P. foetida appears cytologically quite isolated, but closer to the n = 9 group, its smaller chromosomes and areticulate interphase nuclei being similar to species with n = 6, while its chromosome number, higher karyotype symmetry, CMA staining properties, and the number of 45S rDNA sites make it similar to species of subgenus Passiflora. The first set of qualitative data also allows distinguishing some structure within clusters corresponding to subgenera. Thus, within the Passiflora cluster, one main branch corresponds to P. edulis f. edulis and medium-flowered species as P. bahiensis Kl. and P. guazumaefolia Juss., one to large-flowered species of series Incarnatae, i.e. P. edulis f. flavicarpa, P. incarnata, and P. cincinnata Mast., one to series Quadrangulares and the most typical representatives of series Tiliifoliae, one to typical representatives of series Laurifoliae, one to accessions of P. maliformis, and one to species of series Kermesinae. Series Lobatae is shared between the Incarnatae subcluster (including P. caerulea L. and P. gibertii Brown) and the Kermesinae subcluster (P. subpeltata Ortega), which shows the fragility of these subclusters, as P. gibertii and P. subpeltata 130 Chapter IV. Morphological diversity ______________________________________________________________________ are morphologically very similar. Within the Tacsonia-Manicata cluster, there are three main branches. A first one includes the common and widely dispersed P. mixta, P. tripartita, P. tarminiana, P. cumbalensis and endemic species related to one of them, as P. mathewsii (Mast.) Killip and P. luzmarina. A second one includes species of Section Colombiana, plus two accessions of P. tripartita var. mollissima, and the last one includes the relatively short-tubed species P. pinnatistipula and P. manicata. Within the Decaloba-Tryphostemmatoides cluster, one branch corresponds to section Cieca (P. coriacea and P. suberosa) and one to section Xerogona (P. capsularis and P. rubra). P. guatemalensis, the only representative of section Hahniopathantus, is placed apart. Section Decaloba is split between three distinct branches, one for the closely related P. alnifolia Kunth and P. bogotensis Benth., one for P. auriculata Kunth, and one for all its other representatives. 131 Chapter IV. Morphological diversity ______________________________________________________________________ n = 12 n = 10 n=6 n=9 n = ?? Figure 5. Dendrogram obtained with first set of qualitative data. Distances of Sokal & Michener. 132 Chapter IV. Morphological diversity ______________________________________________________________________ IV.1.4.5. Cluster analysis on the global descriptor dataset In a second analysis, all qualitative descriptors (74) were incorporated in the cluster analysis, except those related to pubescence and anthocyanins because of a likely higher sensitivity to environmental conditions, and the difficulty to control redundancy of information (e.g. pubescence of different organs is often, but not always, correlated). Removing these descriptors did not affect the structure of the dendrogram, neither at the subgeneric nor at the interspecific level, however it increased distances and improved bootstrap values. The general structure of the tree based on most descriptors (Figure 6) is highly similar to the one obtained with the reduced descriptor set. Branches support is improved, as bootstrap values increase to more than 50% between subgenera, indicating that descriptors showing high polymorphism within subgenera, and sometimes even at the intraspecific level, still contribute to differentiation among subgenera. Distances and bootstrap values between species, and even accessions of a same species, are much higher, reflecting high polymorphism at the lowest levels. However, this does not affect the tree general consistency, as subdivisions among species can be even more easily interpreted. Below the species level, individuals are most often grouped by accessions, but these accessions do not cluster according to their geographic origin. IV.1.4.6. The “Passiflora cluster” Within subgenus Passiflora, all species are consistently separated from each other. Several subclusters are constituted by species of obvious morphological affinity, but these fail to reflect Killip’s series. Every time a series is represented by three species, one of them diverges. A first example is the subcluster associating the two forms of P. edulis and P. incarnata (series Incarnatae), but not P. cincinnata. A second one is the P. popenovii and P. nitida subcluster (series Laurifoliae), which does not include the less typical representative P. guazumaefolia, with small to medium flowers and a white hirsute corona, placed closer to another relatively small-flowered species, P. bahiensis. Third is the subcluster associating P. tiliifolia and P. ligularis (series Tiliifoliae), which does not include P. maliformis. The latter is highly variable and forms a cluster of its own, only including the very similar P. serrulata, which essentially differs by its serrated and bi-/trilobed leaves and smaller flowers and fruits. Fourth, the series Lobatae is split between the loose subcluster of P. subpeltata and P. gibertii and a diverging P. caerulea branch. In two cases, there is no discordance, but only two species are represented, i.e. P. quadrangularis and P. alata (series Quadrangulares), 133 Chapter IV. Morphological diversity ______________________________________________________________________ and P. lehmannii and P. smithii (series Kermesinae). The subdivisions observed within the Passiflora cluster reinforce the general impression that the creation of series have essentially resulted from the too obvious morphological similarity within small groups of species of subgenus Passiflora, which have provided the core of the taxonomist’s series or sections. This is the case for Tiliifoliae, where P. tiliifolia, P. ligularis, P. palenquensis, P. platyloba are so closely related that their descriptions are often accompanied by a detailed comparison. Another case is that of series Quadrangulares, with P. alata, P. quadrangularis and P. trialata, and series Kermesinae, with P. lehmanni, P. smithii and P. trisulca, all species that are differentiated by very few traits, such as nectary shape. The most difficult case is that of Laurifoliae, where many species can only be distinguished with considerable difficulties, as very few traits of unknown genetic determinism (exact position of nectaries on petiole, number and relative length of corona whorls), are used to differentiate them. When, in a second step, other, better-differentiated species, are aggregated to these uniform groups, problems and inconsistencies arise, and the solutions depend on which traits are prioritized among the dazzling quantity of polymorphic traits. Another option is to widen the infrageneric groups. Thus, in their classification, Feuillet & MacDougal (2003) have gathered series Laurifoliae, Quadrangulares, and Tiliifoliae in a supersection Laurifolia. This view is not supported by our results either. Indeed, in our tree, the link between P. popenovii, P. nitida Kunth, P. quadrangularis, P. ligularis, and P. tiliifolia, is not supported, and the accessions of P. maliformis (also classified in supersection-Laurifolia /seriesLaurifoliae by Feuillet & MacDougal) are placed in a distinct subcluster. In conclusion, our analysis of the Passiflora cluster can be conciliated neither with the infrasubgeneric classification of Killip, nor with that of Feuillet & MacDougal. Its essentially radial structure points to the risk of overclassification instead, which is probably the reason why Killip did not use the section level in this subgenus. IV.1.4.7. The “Tacsonia cluster” The structure of the branch corresponding to subgenera Manicata and Tacsonia shows the close relationship between the most common species of the latter, P. mixta, P. tripartita, P. tarminiana, and P. cumbalensis, forming a cluster onto which are grafted the less common P. mathewsii (very similar to P. mixta) and P. luzmarina (very similar to P. cumbalensis). The second subcluster, constituted by Colombian endemics, indicates greater genetic distances. It gathers species from sections Parritana 134 Chapter IV. Morphological diversity ______________________________________________________________________ (P. parritae (Mast.) L.H. Bailey and P. jardinensis L.K. Escobar) and Colombiana of Escobar. The latter is represented by series Colombianae (P. lanata and P. adulterina L. f.), Quindiensae (P. linearistipula) and Leptomischae (P. antioquiensis H. Karst.., P. flexipes Triana & Planch. and P. tenerifensis L.K. Escobar). Escobar’s classification is not supported by our analysis. Among the common species, there appears no division between sections Tacsonia and Bracteogama. Previous morphological and isozyme studies (Villacís et al., 1998, Segura et al., 2002, 2005), as well as hybridization experiments (Schöniger, 1986), have clearly shown closer similarity of P. tripartita (section Bracteogama) with P. mixta (section Tacsonia) than with P. cumbalensis (section Bracteogama). In the second subcluster, there appears to be a slight differentiation of species with “normal” peduncle from those with extremely long peduncles (series Leptomischae of section Colombiana). In the former group, all remaining sections and series are mixed. Only the two species of series Colombianae, P. lanata and P. adulterina, show a very close morphological similarity, with distances remaining well under the order of intraspecific variation. These distances would have been only slightly higher if pubescence had better been taken into account. In this respect, it must be noted that the two specimens of P. lanata were typical for all characters, and they exhibited a lanate ovary. Killip (1938) and Uribe (1955a) mention a glabrous ovary for this species, a trait not mentioned in the original description by Jussieu (1805). Escobar describes the plant as pubescent, except for leaf upper face, with lanate flowers and fruits, which implies necessarily that the ovary is lanate too. In herbarium material, we have found both glabrous and lanate ovaries for this species. The confuse situation between P. lanata, P. adulterina, P. cuatrecasasii Killip, and the recently proposed P. formosa (whose author mentions a lanate ovary as an important distinctive feature - in fact the only qualitative one - in the comparison with P. lanata), justifies a multivariate analysis to separate intra- and interspecific variation in what could be a species complex or simply a lower number of variable species. IV.1.4.8. The “Decaloba cluster” Differentiation appears higher in subgenus Decaloba than in subgenera Passiflora and Tacsonia, with the formation of six subclusters. The largest one is formed by accessions of section Decaloba, including P. alnifolia, P. bogotensis, P. biflora Lam., P. cuspidifolia Harms, P. erythrophylla Mast., P. magdalenae Triana & Planch. (series Punctatae), P. filipes Benth (series Lutae, although very similar to P. magdalenae), 135 Chapter IV. Morphological diversity ______________________________________________________________________ P. misera Kunth and P. trifasciata Lemaire (series Miserae). The substructure of this large cluster does not support Killip’s subdivisions. Instead, it agrees with the grouping of all these species in a uniform section Decaloba of supersection Decaloba, as proposed by Feuillet & MacDougal. P. auriculata is placed at the base of this large cluster, at a respectable distance, which may justify its classification in a distinct series (Killip’s Auriculatae) or supersection (Feuillet & MacDougal supersection Auriculata). P. sexflora (section Decaloba series Sexflorae) is even more isolated from the other species of section Decaloba. Its placement in our tree questions its inclusion within Killip’s section Decaloba, as well as in the section Decaloba sensu Feuillet & MacDougal. P. guatemalensis, is also individualized, which is consistent with its classification in a distinct section Hahniopathanthus in both classifications. The sixth branch of the Decaloba cluster is subdivided in two consistent branches, one gathering P. suberosa and P. coriacea, and the other uniting P. rubra and P. capsularis (section Xerogona in both classifications). P. suberosa and P. coriacea are classified in a same section Cieca in both classifications, however this section is placed in supersection Decaloba in the proposal of Feuillet & MacDougal, which is not supported by our results. As could be expected in such common and widely distributed species, the accessions of P. suberosa, P. capsularis and P. rubra present a high polymorphism. However, the two last species could not be separated on the basis of their morphology. Their accessions cluster two by two according to the population of origin, sometimes with relatively wide intrapopulational variation, but not following interspecific boundaries. Indeed, all authors insist on the difficulty to distinguish them. Killip (1938) and Holm-Nielsen et al. (1988) state that this is practically impossible on sterile specimens. Their descriptions give particular importance to fruit shape and ovary pubescence, although they do not agree on the distribution of ovary pubescence between the two species. Ulmer & MacDougal (2004) insist on the color of the corona base. Our identification was based on these three criteria, prioritizing fruit shape and corona color (see Figure 7), but then we had to admit three different levels in ovary pubescence (glabrous, puberulent, hirsute) in P. capsularis. Changing the priority of criteria would just give a different but also inconsistent identification. Thus it appears that these traits can combine in different ways, so they are not reliably discriminant, and the analysis for all other traits, including or not pubescence, only confirms that the status of these two species should be revised, downgrading them to botanical varieties, unless elements of reproductive biology contradict this view. But even this is improbable, as self136 Chapter IV. Morphological diversity ______________________________________________________________________ incompatibility and self-compatibility coexist within P. capsularis (Ulmer and MacDougal, 2004). IV.1.4.9. The “Astrophea cluster” Subgenus Astrophea appears very uniform, which could be expected as only one section, Euastrophea, is represented in our small sample of this subgenus. The accessions share most of their qualitative traits, and the divergence of one population of P. emarginata can only be attributed to differences in leaf shape descriptors. This accession also differs slightly in the pubescence of the ovary, but this trait did not interfere as it was not used in the cluster analysis. The sample was too limited to draw any conclusion on intra-and interspecific boundaries between the species represented. n = 10 n = 12 n=6 I n = ?? n=9 II III Figure 6. Dendrogram obtained on complete set of qualitative data. Distances of Sokal & Michener. 137 Chapter IV. Morphological diversity ______________________________________________________________________ I Figure 6a. First part of the dendrogram obtained on the complete set of qualitative data. 138 Chapter IV. Morphological diversity ______________________________________________________________________ II Figure 6b. Second part of the dendrogram obtained on the complete set of qualitative data. 139 Chapter IV. Morphological diversity ______________________________________________________________________ III Figure 6c. Third part of the dendrogram obtained on the complete set of qualitative data. 140 Chapter IV. Morphological diversity ______________________________________________________________________ a b Figure 7. Morphological affinity between typical representatives of P. rubra (a) and P. capsularis (b). Accessions from Colombia (a, Calarcá, Quindío – b, Cartago, Valle del Cauca). IV.1.4.10. Morphological and molecular diversity To appreciate the reliability of the morphological approach to Passiflora diversity, we have compared some of the interspecific associations or divergences with similar results obtained in phenetic studies with biochemical and molecular markers on samples including some of our species. Indeed, a first series of genetic studies were carried out on smaller species samples, mostly from Colombia too. In the trees obtained with RAPD and cpDNA RFLP markers by Fajardo et al. (1998) and Sánchez et al. (1999), the species of subgenus Tacsonia only constitute one subcluster within a large cluster gathering them with species of subgenus Passiflora. Subgenus Decaloba is represented by P. coriacea and P. adenopoda, both species strongly diverging from this Passiflora- 141 Chapter IV. Morphological diversity ______________________________________________________________________ Tacsonia cluster, but also between themselves, which appears consistent with our results. In the RAPD study, subgenera Distephana and Astrophea are represented respectively by P. vitifolia and P. spinosa (Poepp. & Endl.) Mast., and both species are placed at considerable distance from the Passiflora-Tacsonia cluster, their divergence being intermediate between that of P. adenopoda and that of P. coriacea. Within the Tacsonia subcluster, the distances between the species, P. tripartita var. mollissima, P. cumbalensis, P. pinnatistipula, and P. antioquiensis, follow the same order as in the corresponding morphological cluster. This is still true when the comparison is extended to subgenus Manicata, considering the results obtained by Segura et al. (2002, 2003) with AFLP markers (P. tenerifensis and P. parritae also included), and with isozymes, although P. antioquiensis is placed closer to the most common tacsos than P. pinnatistipula in the isozyme study. Another convergence between morphological and AFLP markers is the clear separation of P. maliformis from the typical species of series Tiliifoliae of subgenus Passiflora (Ocampo et al., 2004). Genetic relationships between subgenera and between particular species can also be deduced from subsequent phylogenetic studies carried out on wider species samples by Muschner et al. (2003) with ITS, trnL-trnF and rps4 sequences, by Yockteng (2003) with chloroplastic matK sequences, Yockteng & Nadot (2004) with sequences of the nuclear chloroplast-expressed glutamine synthetase gene (ncpGS), and Hansen et al. (2006) with trnL/trnT sequences. All these studies support the existence of three major clades, one corresponding to subgenus Astrophea, one formed around subgenus Decaloba, and one formed around subgenus Passiflora. When included in these studies (i.e. all studies except that of Muschner et al., 2003), species of subgenus Tacsonia form a subclade within the Passiflora clade. P. vitifolia, and other representatives of subgenus Distephana in the Yockteng’s studies, are also included in this large Passiflora clade, although they are not grouped consistently in the matK tree (Yockteng, 2003). The relative position of the three major clades differs among studies. Subgenus Astrophea takes an intermediate position in the ITS tree, it is closer to the Passiflora clade in the other trees of Muschner et al. (trnL-trnF and rps4), but closer to the Decaloba clade in the trnL-trnT tree of Hansen et al., while the Decaloba clade appears closer to the Passiflora clade in the sample of Yockteng (ncpGS and matK trees). P. foetida is clearly placed within the Passiflora clade according to rps4 and trnL-trnF sequences, but more distant, although basal to this clade, in the trees obtained with ITS, 142 Chapter IV. Morphological diversity ______________________________________________________________________ ncpGS, and matK sequences. P. adenopoda is basal to the Decaloba clade in the ncpGS tree, as well as in the ITS study of Krosnick & Freudenstein (2005). P. morifolia Mast., another species of section Pseudodysosmia, is also basal to the Decaloba clade, on a branch between Decaloba and Astrophea, in the ITS and trnL-trnF trees. In the ncpGS tree of Yockteng & Nadot (2004) and the ITS study of Krosnick & Freudenstein (2005), subgenus Tryphostemmatoides, represented by P. tryphostemmatoides Harms and by P. arbelaezii L. Uribe respectively, is placed close to subgenus Astrophea. Strangely, in the study of Yockteng & Nadot (ncpGS tree), P. sanguinolenta Mast., a representative of subgenus Psilanthus, is placed in a Passiflora subclade dominated by representatives of series Tiliifoliae. Within the Decaloba clade, we can recognize the higher level of divergence between subclusters, the association we have observed between representatives of series Punctatae and Miserae in the widest subclade (trnL-trnF tree of Muschner et al., 2003), the relative separation of P. sexflora from this group, and the even more distant position of P. coriacea and of the couple formed by P. capsularis and P. rubra (ITS, trnL-trnF, and ncpGS sequences). However, the two last species are closely associated with P. sexflora in the ncpGS tree. As in our morphological study, the Passiflora clade identified in phylogenetic studies generally shows loose relations between species. The interpretation of the poorly supported subclades is very uneasy, with the partial exception of the ncpGS tree presented by Yockteng & Nadot (2004), where branches are better defined, although not easier to interpret, given, for example, the dispersion of species belonging to the series Incarnatae and Laurifoliae. Among the close associations documented by our morphological study, we can only recognize those of P. quadrangularis with P. alata and P. incarnata with P. edulis (ITS and ncpGS trees). As a first major point of conclusion on this comparison between morphological and molecular diversity, we can underline that the major morphological divisions observed in our study find support in the genetic studies. The cytological groups are always validated, with the clear separation of subgenera Astrophea (n = 12), Tryphostemmatoides and Decaloba (n = 6) between themselves and from subgenera Passiflora, Tacsonia, and Distephana 143 (n = 9). Concerning subgenus Chapter IV. Morphological diversity ______________________________________________________________________ Tryphostemmatoides, the consistency between morphological and genetic studies is clear only when considering our quantitative analysis, where it is associated with subgenus Astrophea mostly on peduncle traits (third principal component). This trait is represented also in the qualitative descriptors, however its effect is blurred by the high number of traits shared with subgenus Decaloba. While the comparison is difficult for subgenus Tryphostemmatoides, it is impossible for subgenus Psilanthus, because of insufficient data and the unlikely placement of P. sanguinolenta in the Passiflora clade in the ncpGS study. The two species, P. adenopoda and P. foetida, that take an intermediate position in the general “morpho-cytological” pattern, or their close relatives, are consistently placed in intermediate positions, in most phylogenetic studies, P. adenopoda or P. morifolia (section Pseudodysosmia of subgenus Decaloba) appearing basal to a general Decaloba clade and P. foetida (subgenus Dysosmia) basal to the general Passiflora clade. The comparison becomes more difficult at lower, infrasubgeneric, levels. Subgenus Decaloba appears better structured than the other numerous subgenera, and shows similarities in morphological and molecular diversity patterns, with the grouping of Killip’s sections Punctatae and Miserae, and the differentiation of species of sections Xerogona, Cieca, and series Auriculata and, less clearly, Sexflorae. The placement of P. adenopoda in the different trees questions the inclusion of section Pseudodysosmia, while the structure observed among representatives of several sections provides support to some simplification, but not for as many fusions as those operated in the new morphological classification of Feuillet & MacDougal. In any case, more species should be gathered in a same phenetic study before revising objectively the morphological classification. Within the n = 9 group, molecular data and morphological diverge partially, as studies of DNA sequences allow the distinction of a Tacsonia-Manicata group and fail to separate clearly subgenus Distephana, placing both of them within a Passiflora clade, while morphological analysis supports these three subgenera at the same level of differentiation. The fact that species of subgenera Distephana, Tacsonia and Manicata have developed ornithophyly is obviously related to their strong morphological differentiation, which does not minor the importance of their separation from subgenus Passiflora. Whether their probable evolution from a “Passiflora-like” common ancestor 144 Chapter IV. Morphological diversity ______________________________________________________________________ justifies their inclusion in the bee-pollinated Passiflora subgenus, as proposed in the new classification, is just the same classical question about considering birds as dinosaurs. In the end, it seems a problem of putting more emphasis on the adaptative forces commanding evolution or more emphasis on the genetic structure that subtend them. Concerning subgenus Passiflora sensu Killip, no clear structure appears at the interspecific level that could result in clear subdivisions into series. The study of sequence variation for the ncpGS gene provides the only tree with reasonably well supported structure at this level, however several obvious abnormalities question the robustness of the information. Our morphological observations only confirm closer associations between the most typical representatives of some series, however the number of contradictions with the classification and the lack of a clear hierarchy in the branch structure point to the difficulty of the work and the risk of under- or overclassification, leading to chose between a limited number of poorly supported series or a great number of poorly represented series. Similarly, the structure of the TacsoniaManicata branch does not support clearly sections and series in subgenus Tacsonia, however it allows differentiation between two groups of tacsos, one corresponding to common species that probably have their center of diversity in Ecuador, as is obvious for P. cumbalensis, P. luzmarina and P. matthewsii, and very likely for P. mixta, P. tripartita and P. tarminiana (Segura et al., 2005), and another cluster only including species endemic to Colombia, with a slight but clear differentiation related to extreme variation for peduncle length. IV.1.5. Conclusions In the absence of a clear set of morphological criteria for discriminating at the different hierarchic levels of the infrageneric classification of Passiflora, we have used a quite exhaustive list of 43 quantitative and 84 qualitative descriptors. A shorter list of 32 qualitative traits, selected after analyzing variation among Killip’s subgenera, allowed to classify our 60-species sample consistently, using a strictly phenetic approach. Most discriminant characters include size of stems and leaves, presence of tendrils, number and distribution of extrafloral nectaries, dimensions and general shape of bracts, width and length of flowers, corona complexity, and, although they could not be systematically analyzed, fruit size and color. Eight of the nine Killip’s subgenera 145 Chapter IV. Morphological diversity ______________________________________________________________________ represented in our sample are supported by the morphological analysis, although subgenus Tryphostemmatoides is only supported in the quantitative analysis. By contrast, the simplification proposed by Feuillet & MacDougal is not clearly supported in our analyses, except for the possible inclusion of P. manicata in subgenus Tacsonia, as this species is intermediate with subgenus Passiflora for quantitative traits but very similar to tacsos for most qualitative traits. Furthermore, the placement of P. adenopoda and its close relatives, in our analyses as well as in molecular ones, questions their treatment as a division of subgenus Decaloba. More generally, as compared to molecular results, our phenetic approach provides a better resolved vision of relations among passifloras. Beyond the differentiation between cytological groups (n = 6, 12 or n = 9; de Melo et al., 2001) underlined in most analyses, appear other features of considerable importance for their evolution. The division between the two cytological groups is particularly paralleled by a division on the presence and position of extrafloral nectaries and the complexity of the corona, showing the importance of coevolution. In the same line, Yockteng (2003) underlined the differentiation appearing in the spectrum of cyanogenic components developed against herbivores in the two cytological groups. Even clearer appears the coevolution with pollinators, causing the main line of floral divergence between subgenera Passiflora, pollinated by large bees, Tacsonia, pollinated only by the sword-billed hummingbird, and Distephana, pollinated by other hummingbirds. In the n = 6/12 group, where small to medium size insects dominate as pollinators, the morphological divergence of P. trinervia (subgenus Psilanthus), showing exactly the same adaptation to Andean highlands and pollination by the same bird species as subgenus Tacsonia, is not less significant. This convergence is logically expressed in the evolution of the corona. However, the minute corona of P. trinervia, and other typical representatives of subgenus Psilanthus, probably results from the transformation of the two-ranked corona common in subgenera Astrophea and Decaloba, while the reduced corona of most tacsos has evolved from the complex corona observed in all species of subgenus Passiflora at the same time as their specialized hypanthium. Remnants of these complex coronas can still be observed in species with a less developed hypanthium, as P. manicata and P. antioquiensis, with one or two external whorls, plus a few other series of thinner filaments under the floral tube throat, and even in typical long-tubed tacso flowers. Thus, a more or less developed second whorl is not rare in P. mixta, while loose whorls of very thin white filaments can be observed far in the hypanthium of P. tripartita var. mollissima. The 146 Chapter IV. Morphological diversity ______________________________________________________________________ evolution has probably been different for subgenus Distephana, and its differentiation within the n = 9 group earlier in the history of Passiflora, as its species show no such signs of an ancestral complex corona. Instead, they are constituted by two or three rows, which can be compared with numbers observed in subgenera Decaloba and Astrophea. The fusion of the elements into a prolongation of the floral tube is paralleled in certain representatives of subgenus Decaloba, such as P. tulae Urban (Murucuja), forming another striking case of convergence between species from very different evolutive backgrounds. IV.1.6. Acknowledgements This research has been funded by Colciencias and the Colombian Ministry for Environment, with support of the Research Center of the Colombian Coffee Grower Federation (Cenicafé) through the projects: ‘Conservación y utilización de los recursos genéticos de pasifloras’ and ‘Estudio de la diversidad de las Passifloraceae y Caricaceae de la zona cafetera de Colombia’. The authors are indebted to María Restrepo, Felipe Barrera, Cristián Olaya and Lina Farfán for assistance in collecting field data, Daniel Franco and Mario Ruiz for help in installation works of living collections (Paraguacito Experimental Station – Cenicafé), and to German Arroyave and Juan G. Contreras (PASSICOL S.A) for providing living collection facilities for maracuja. The first author gratefully acknowledges financial support from the Gines-Mera Fellowship Foundation (CIAT - CBN). Finally, they are also thankful to Dr. Philippe Feldmann (CIRAD) for comments and suggestions. 147 CHAPTER V ____________________________________ Chloroplast and mitochondrial DNA variation in the genus Passiflora L. (Passifloraceae) as revealed by PCR-RFLP Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1. Chloroplast and mitochondrial DNA variation in the genus Passiflora L. (Passifloraceae) as revealed by PCR-RFLP John Ocampo Pérez1,2*, Geo Coppens d’Eeckenbrugge2, Ange-Marie Risterucci3. 1 Bioversity International (formerly IPGRI), Regional Office for the Americas, A.A. 6713, Cali, Colombia; CIRAD, UPR ‘Gestion des ressources génétiques et dynamiques sociales’, Campus CNRS/Cefe, 1919 route de Mende, 34293 Montpellier, France; 3CIRAD-AMIS, UMR 1096, Avenue Agropolis, TA 40/03, 34398 Montpellier, cedex 1, France. 2 V.1.1. Abstract The chloroplast and mitochondrial DNA diversity of 213 accessions belonging to 151 Passiflora species of 15 subgenera recognized by Killip (1938) was studied by PCR-RFLP analysis of two non-coding cpDNA regions (psbC- trnS and trnS - trnfM) and two non-coding mtDNA region (nad4-1/2 and nad1-B/C). This sample set was supplemented with six accessions from three African Passifloraceae genera Adenia, Barteria, and Smeathmannia, as outgroup species. The PCR-amplified cpDNA regions were digested with six endonucleases. A total of 614 fragments were scored, of which 93% were found to be polymorphic in the sample. Two-hundred-eighty haplotypes were found for the chloroplast and 372 for the mitochondria. A higher level of interspecific variation was detected in the mtDNA regions than in the cpDNA regions. The first two axes of the principal co-ordinates analysis accounted for 59% of the total variation on cpDNA data. They allowed visualizing a strong structure, as the genera Adenia, Barteria, and the Passiflora subgenera Astrophea, Calopathanthus, Dysosmia, Distephana, Manicata, Passiflora, Tacsonia, Tacsonioides and Tryphostemmatoides, and P. deidamioides occupy the left half of the principal plane, while the species of subgenera Apodogyne, Decaloba, Murucuja, Pseudomurucuja and Psilanthus form a very well separated group, placed in a quite extreme position on the right, only the accessions of Smeathmannia and P. lancetillensis (subgenus Deidamioides) taking intermediate positions. The phenogram obtained by the neighbor-joining method on cpDNA data is more coherent with the major divisions of the taxonomy proposed recently by Feuillet & MacDougal than the corresponding mtDNA tree. The cpDNA tree shows three major, well supported clusters within Passiflora. The first one, named the “Passiflora group”, includes subgenera Calopathanthus, Deidamioides, Distephana, Dysosmia, Dysosmioides, Manicata, Passiflora, Tacsonia, and Tacsonioides, with a very loose substructure and considerable intraspecific variation. The second one includes subgenus Astrophea, and the third one, named the “Decaloba group”, comprises most species of subgenera Apodogyne, Decaloba, Murucuja, Pseudomurucuja and Psilanthus. P. gracillima (subg.enus Tryphostemmatoides) appears in basal position, forming a fourth cluster of its own. The position of P. lancetillensis (subgenus Deidamioides) is undefined, as it is placed on a long branch between the outgroup and the “Decaloba group”, the outgroup itself taking an undefined position among the three major Passiflora clusters. The phenogram obtained with mtDNA data separates four main clusters. As for cpDNA, a first large, well supported, cluster corresponds to the “Decaloba group”, where accessions are grouped by species, with the only exception of P. adenopoda. The numerous accessions of section Decaloba series Punctatae and Miserae tend to form a subcluster, while Killip’s subgenera Apodogyne, Murucuja, Pseudomurucuja and Psilanthus cannot be recognized. The other major clusters are different from those evidenced by cpDNA data, as subgenera Astrophea and Tryphostemmatoides appear integrated within the “Passiflora group”, while subgenus Tacsonia forms a uniform distinct cluster, close to another one comprising species of series Kermesinae, Simplicifoliae, Lobatae, and Menispermifoliae. The analyses of chloroplastic and mitochondrial fragments gave very different pictures on the genetic structure of genus Passiflora. Differences appear at all levels, in the position of the outgroup, the relative position of four subgenera, and the relationships between species. The divergence in the information obtained from chloroplast and mitochondrial genomes are attributed to differences in their rate of evolution and mode of transmission and to reticulate evolution in the genus. Key words: Passiflora, PCR-RFLP, chloroplast and mitochondrial DNA, variation, evolution. 149 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1.2. Introduction The family Passifloraceae is divided into two tribes, Pariopsieae and Passifloreae, and includes more than 650 species (Escobar, 1988; Ulmer & MacDougal, 2004) distributed throughout the tropics. The base number of chromosome varies among x = 6 and x = 9 (De Melo & Guerra, 2003). With approximately 525 species, including several cultivated ones, Passiflora is numerically and economically the most important genus of the family. Passionflowers are generally vines, although some representatives are shrubs or trees. P. edulis Sims (yellow maracuja) is by far the best-known and economically most important species of the family. In the last extensive revision of the genus, Killip (1938) classified 355 Passiflora species into 22 subgenera (Annex 1). For long, while the list of species was considerably extended, his views were only amended or supplemented. In Colombia, Escobar (1988a,b, 1989, 1990 inedited, 1994) reviewed subgenera Astrophea, Distephana, Manicata, Rathea and Tacsonia, merging subgenera Tacsoniopsis and Tacsonia, subdividing them into sections and series, and proposing one additional subgenus, Porphyropathanthus. Recently, Feuillet & MacDougal (2003; Annex 2) have proposed a new infrageneric classification of Passiflora recognizing only four subgenera, further divided into 16 supersections. Three of their subgenera are strictly American: Astrophea, Deidamioides, and Passiflora, with 57, 17 and 234 species respectively, with an essentially tropical distribution. Decaloba, with its 204 species, is mainly distributed in America, but it is also represented in Southeast Asia and Australia (Ulmer & MacDougal, 2004). In this new proposal, the genus Tetrastylis is placed as a section of subgenus Deidamioides, and the subgenus Tetrapathea from New Zealand is excluded from the genus Passiflora (Hutchinson, 1967; Green, 1972; Yockteng & Nadot, 2004). Both classifications of Killip (1938) and Feuillet & MacDougal (2003) are based on the extreme morphological richness and complexity of Passiflora, whose species present numerous particular traits, including a wide variation in leaf shape, even within species and within individuals (heterophylly and heteroblasty), the presence of extrafloral nectaries in different parts (on leaf petiole, lamina or margins, on bracts, on sepals), floral 150 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ traits showing a high level of coevolution with particular pollinators, particularly in corolla color, the variable development of the hypanthium and that of the corona. However, no clear hierarchy emerges in the relative contributions of these traits to taxonomy of the genus. Ocampo & Coppens d’Eeckenbrugge (in preparation; cfr. Chapter IV), have used a list of 127 morphological descriptors in a phenetic approach of morphological differentiation among 60 Passiflora species from nine of the Killip’s subgenera. The quantitative descriptors clearly separated subgenera Astrophea, Decaloba, Passiflora, Psilanthus, Tacsonia, and Tryphostemmatoides. Subgenus Dysosmia appeared intermediate between Passiflora and Decaloba, while the representatives of subgenera Manicata and Distephana showed affinity with subgenera Tacsonia and Passiflora respectively. Qualitative trait analysis showed a major distinction between species with 2n = 12 or 24 chromosomes, and species with 2n = 18 chromosomes, and discriminated more clearly among Killip’s subgenera Astrophea, Psilanthus, Decaloba, in the former cytological group, and Distephana, Dysosmia, Passiflora and Tacsonia, in the second cytological group. Subgenera Manicata and Tryphostemmatoides could not be distinguished from subgenera Tacsonia and Decaloba respectively. While these main divisions could be easily interpreted following Killip’s subgenera, this was not true at lower levels. No clear subdivisions were observed within subgenus Passiflora and only a slight geographical structure was detected in subgenus Tacsonia. Differentiation was higher in subgenus Decaloba, where subclusters partly supported the simplification proposed by Feuillet & MacDougal. The most obvious contradiction with both classifications was the placement of P. adenopoda (subgenus Decaloba section Pseudodysosmia) in an intermediate position, between subgenera Passiflora and Decaloba. In many plant groups, molecular genetics have provided powerful new tools to understand the structure and evolution of species diversity. The first studies of genetic variation in Passiflora were based on RAPD, RFLP, AFLP and isozyme markers, and initiated in a breeding perspective, focusing on cultivated species and their wild relatives in subgenera Passiflora and Tacsonia (Fajardo et al., 1998; Sánchez et al., 1999, Segura et al., 2002; Segura et al., 2003a; Ocampo et al., 2004; Segura et al., 2005). As discussed 151 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ in the preceding chapter, their results are consistent with our morphological analyses, but their narrow samples limited the reach of comparisons. The application of molecular tools to phylogenetic analysis of wider samples is more recent. The first genus-wide molecular analysis of Passiflora, using the nuclear ribosomal internal transcribed spacers, (ITS-1 and ITS-2), plastid trnL-F intergenic spacer and the rps4 plastid gene, was reported by Muschner et al. (2003; Annex 4a). Eleven subgenera were represented by 61 species. The ITS alignment appeared highly variable, with shorter sequences in subgenera with 2n = 18 or 2n = 20 chromosomes than in subgenera with 2n = 12 or 24 chromosomes. The three phylograms consistently showed three major clades, corresponding with three of the subgenera proposed by Feuillet & MacDougal (2003), i.e. the ‘Passiflora clade’, composed of 2n = 18/20 species, the ‘Decaloba clade’ composed of 2n = 12 species, and one composed of 2n = 24 species (i.e. subgenus Astrophea). The representatives of subgenus Deidamioides contributed to the Decaloba clade in the ITS phylogram, while they appeared independent from the three major clades in the trnL-trnF and rps4 phylograms. The substructure of the two most numerous clades, ‘Passiflora’ and ‘Decaloba’, were not consistent with either taxonomical classification. Within the ‘Passiflora clade’, bootstrap values are low and there is no clear structure that could correspond to sections or series, although a few associations are recognized between closely related species, such as P. alata and P. quadrangularis, P. edulis and P. incarnata, in the ITS tree. These associations are not expressed in the trnL-trnF tree, where even the two species of the uniform subgenus Distephana do not appear related. The ‘Decaloba clade’ shows a stronger substructure, with an association between species of sections Decaloba, Miserae and Punctatae of subgenus Decaloba, an smaller subclusters representing sections Cieca and Xerogona, while P. morifolia (section Pseudodysosmia) takes a relatively basal position in both phenograms. However, in the trnL-trnF tree, P. rubra strongly diverges from the Decaloba clade, which is not the case for P. capsularis, a species that is so similar that our morphological analysis led us to consider the two as synonyms (see Chapter IV). An even stronger inconsistency concerns the placement of the outgroups in both phenograms. Adenia keramanthus is positionned as sister to the ‘Passiflora clade’, with a support value of 97%, while Mitostemma 152 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ brevifilis is placed as sister to the ‘Decaloba’ and Astrophea clades in the ITS tree, both species diverging from Passiflora species in the trnL-trnF tree. In addition, Mitostemma brevifilis also appears in an unexpected position at the base of the ‘Passiflora’ and Astrophea clades in the rps4 tree. Muschner et al. (2003) concluded that the monophyly of the genus Passiflora was not supported by any statistical or phylogenetic method, “and in several trees there was even support of a non-monophyletic Passiflora”. One year after Muschner et al. (2003), Yockteng & Nadot (2004; Annex 4c) published a phylogenetic study of a 91 species-sample representing 17 of the 23 subgenera recognized by Killip (1938) and Escobar (1988a), based on the sequences of the chloroplast-expressed glutamine synthtase gene (ncpGS). Their phenogram showed the same three major clades, with significant differences in length of introns. These results provided support to three of the four subgenera defined by Feuillet & MacDougal (2003), but indicated that three additional subgenera, Polyanthea (DC.) Killip, Dysosmia (DC.) Killip, and Tetrapathea (DC.) Rchb., should also be recognized. The Astrophea clade is sister to the other species, except P. cirrhiflora. P. tryphostemmatoides (subgenus Tryphostemmatoides) appears closely related to it. Within the Passiflora clade, appears a Tacsonia subclade that also includes P. manicata (subgenus Manicata) and P. racemosa (subgenus Calopathantus). Two other subclades are more difficult to interpret. One of them includes a small third-order cluster constituted by species of subgenus Distephana, joined surprisingly by P. cincinnata (subgenus Passiflora series Incarnatae). Several third-order subclusters appear dominated by particular morphological groups, but too many exceptions hamper a consistent interpretation. The dispersion of species of the uniform series Laurifoliae provides the best example of this situation. The inclusion of P. sanguinolenta (subgenus Psilanthus) is the strangest abnormality in the Passiflora clade. Like in the study of Muschner et al. (2003), the Decaloba clade shows a stronger substructure, with similar associations, and relatively basal position of the species of section Pseudodysosmia. Very surprisingly, this Pseudosysosmia cluster included P. alnifolia, thus isolated from its closest relatives. An even more surprising placement is that of P. candida (subgenus Astrophea) close to P. auriculata. The same large sample was studied for sequences of the chloroplastic gene matK (Yockteng, 2003; Annex 4b). 153 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ The results were generally consistent in defining the same three clades, despite some divergence in their relative positions. Tryphostemmatoides appeared fully included in the Astrophea clade. Resolution was poor at the section and series level, however second order subdivisions were also consistent with ITS and ncpGS results in the ‘Decaloba clade’, and a Tacsonia-Manicata-Calopathanthus subclade was individualized within the ‘Passiflora clade’. P. candida was placed more logically in the Astrophea clade, which discards the hypothesis of a misidentification invoked in the interpretation of the ncpGS study. Krosnick & Freudenstein (2005) showed that the 22 species of Old World Passiflora form a monophyletic group and supported their placement by Feuillet & MacDougal in supersection Disemma, within subgenus Decaloba, using plastid (trnL-F) and nuclear (ITS) DNA sequences. More recently, Hansen et al. (2006; Annex 4d) analyzing chloroplast sequences rpoC1 and trnL/trnT from 136 species of genus Passiflora, also obtained the three major clades, with problems of resolution and consistency at section and series level, and a few striking cases of inconsistent positioning. The most spectacular example came from P. microstipula, a species with 2n = 18 chromosomes, two accessions being placed in the Passiflora clade, close to the very different P. nitida, and a third one at the base of the Decaloba clade. As stated by the authors, the most likely explanation is an extreme case of chloroplast capture. Hansen et al. (2006) sustain that their data unequivocally support the subgeneric classification of Feuillet & MacDougal, however the position and composition of subgenus Deidamioides is not clearly defined, and the phylogram presented places Dilkea and Tetrapathea as sister group to a Decaloba-Deidamioides clade, which throws additional doubts on Passiflora monophyly. This revision of the phylogenetic studies consistently shows the existence of three major clades in the genus Passiflora, with a stronger structure in the Decaloba clade than in the Passiflora clade. When tacsos (species of subgenus Tacsonia) are included in the sample, they are placed in a particular subclade of the latter. Beyond this general pattern, several points of conflict are obvious too. First are the inconsistencies in the relative placement 154 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ of the three major clades. Such variation affects more markedly the placement of several morphologically atypical species. Also, there are cases of highly improbable placement concerning a few species. The variable placement of the outgroup species is even more problematic, generating doubts on the monophyly of the genus. A comparison of all phenograms also shows that chloroplastic sequence data tend to produce trees with less consistent and less supported structure than the nuclear ncpGS and ITS trees. In the last decade, the genetic information present in the plant chloroplastic genome has been particularly used for phylogenetic purposes (Clegg et al., 1994; Vekemans et al., 1998), because of its non-Mendelian mode of inheritance and low rate of evolution as compared to that of the nuclear genome, which makes it very useful in studying the variation at levels higher than the species, considering that the probability of detecting intraspecific variation is low. This trend has been furthered accentuated by the development of efficient techniques to study this variation, among them PCR-RFLP or CAPS (cleaved amplified polymorphic sequence), which employs universal primers for amplification of specific noncoding DNA, followed by their restriction. This technique soon appeared to be more efficient than traditional RFLP to reveal polymorphism. Its successful application to a number of crop and forest species for which extensive nucleotide information is now available (Mes et al., 1997; Lakshmi et al., 2000; Parani et al., 2000; Duval et al., 2003; Van Droogenbroeck et al., 2004; Kyndt et al., 2005), has changed the common perception of the variability of chloroplast genome. Reviewing cases of cpDNA variation at the intraspecific level, and even within individuals, Harris & Ingram (1991) have underlined that hybridization, introgression and lineage sorting all may influence a taxon’s position in cpDNA phylogenies. They have also underlined that the occurrence of widespread biparental plastid transmission in the Angiosperms makes the implicit assumption that hybridization can be ignored not always tenable. The mitochondrial genome has been less used because of its high degree of intramolecular recombination and low rate of base substitution. In Passiflora, biparental plastid transmission was first suspected by Corriveau & Coleman (1988), following epifluorescence microscopy observations on P. edulis. Do et al. (1992) showed that RFLP markers of cpDNA were mostly inherited maternally in crosses between yellow 155 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ and purple maracuja when the former (P. edulis f. flavicarpa) was used as female parent, and biparental in the reciprocal cross, suggesting asymmetric post-fertilization exclusion processes. Transmission was paternal in the hybrid ‘P. coccinea x P. edulis f. flavicarpa’. More recently, Mráček (2005) observed biparental transmission between P. menispermifolia and P. oerstedii and heteroplasmy in resulting hybrids. Muschner et al. (2006) established paternal transmission of cpDNA in four interspecific hybrids of subgenera Passiflora and Dysosmia and maternal transmission in an interspecific hybrid of subgenus Decaloba. All mtDNA were maternally transmitted in these five hybrids. Hansen et al. (2007) studied 17 crosses and found paternal or biparental inheritance of cpDNA, except in their intraspecific crosses, where it was predominantly maternal. Hansen et al. (2007) evoked the possible effect of genetic divergence between the parents, however it should be noted that the progeny studied was from P. costaricensis, the only species of subgenus Decaloba in their study. Hence, an alternative and simpler explanation, consistent with the results of Muschner et al. (2006), could be differences in cpDNA transmission between subgenera. The study presented here was planned before the publication of phylogenetic studies in Passiflora. The implementation of projects including extensive collecting of Passiflora germplasm in Colombia, the country with the widest Passiflora diversity (Chapter II), provided a unique opportunity to study a very wide sample of the genus, taking into account possible variation at the infraspecific level. The species sample was significantly widened thanks to the contribution of the French National collection. The PCR-RFLP technique was chosen because of its relative simplicity, rapidity, and cost efficiency. At that time, sequence analyses would have been more expensive, not allowing the study of the genomes of both organelles on so many individuals. Two cpDNA spacer regions, (psbC-trnS and trnS-trnfM), and two mtDNA introns (nad4-1/2 and nad1-B/C) were analyzed for variation at both intra-and intergeneric levels in the 151 Passiflora species of our sample and four species of the Old World genera Adenia, Barteria and Smeathmannia. The objective was to study the general structure of genetic diversity in the genus, with particular attention to the cultivated species and their close relatives. 156 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1.3. Materials and methods V.1.3.1. Taxon sampling The germplasm sample consisted of 213 individuals from 151 species, representing 15 of the subgenera recognized by Killip (1938), Escobar (1988) and MacDougal (1994), and eight outgroup accessions of Adenia, and Barteria, and Smeathmannia, from Africa (Table 1). This sample included several representatives of the principal cultivated species. Most of the materials were collected in Colombia, from the wild, home gardens, and farms (see Chapter II). Voucher specimens were deposited at COL and VALLE herbaria. Other samples were obtained from the living National Collection in Blois (France), mainly collected in French Guiana and Brazil, and maintained by Christian Houel. Specimen identification was based on the morphological descriptions of Killip (1938), Holm-Nielsen et al. (1988), Escobar (1988a,b, 1994) and MacDougal (1992, 1994), and the comparison with herbarium material (mostly COL, HUA, MEDEL, COAH, K, MA, MO, NY, P, PSO). Infrageneric taxonomy follows the same references, unless specified otherwise. Table 1. List of species used in this study according to classification by Killip (1938), Escobar (1988a,b) and MacDougal (1994). Infrageneric clasification Collection data Status Passifloraceae Juss. ex Kunth. Tribe Paropsieae D.C., 1828 Genus Barteria Hooker, J.D., 1860 B. fistulosa Mast. Cameroon, Kandara (Central Province) Wild B. fistulosa Mast. Cameroon, Ebodjé (Southwest Province) Wild B. nigritiana Hook. f. Cameroon, Ebodjé (Southwest Province) Wild B. nigritiana Hook. f. Gabon Wild B. solida F. J. Breteler Cameroon, Dikome Balue (Southwest Province) Wild Genus Smeathmannia Solander ex R. Brown, 1821 S. pubescens Sol. ex R. Br. Cameroon, Ebodje (Southwest Province) Wild S. pubescens Sol. ex R. Br. Cameroon, Mamalles (Southwest Province) Wild National Collection - Blois - France Wild Tribe Passifloreae D.C., 1828 Genus Adenia Forsskal, 1775 Adenia glauca Schinz Genus Passiflora L., 1753 Subgenus Astrophea (DC.) Mast., 1871 Section Dolichostemma 157 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ P. mariquitensis Mutis ex L. Uribe Colombia, Tolima, Mariquita Wild Section Euastrophea P. arborea Spreng. Colombia, Caldas, Manizales Wild P. emarginata Humb. & Bonpl. Colombia, Caldas, Manizales Wild P. emarginata Humb. & Bonpl. Colombia, Valle del Cauca, Yotoco Wild P. sphaerocarpa Triana & Planch. Colombia, Valle del Cauca, Cali Wild P. sphaerocarpa Triana & Planch. Colombia, Tolima, Ibagué Wild P. macrophylla Spruce ex Mast. Ecuador, National Collection - Blois - France Wild Section Pseudoastrophea P. citrifolia (Juss.) Mast. Guiana French, Belgian National Botanic Garden - Meise Wild P. citrifolia (Juss.) Mast. National Collection - Blois - France Wild P. haematostigma Mart. ex Mast. Guiana French, National Collection - Blois - France Wild P. kawensis Feuillet National Collection - Blois - France Wild Subgenus Calopathanthus (Harms) Killip, 1938 P. racemosa Brot. National Collection - Blois - France Wild Subgenus Decaloba (DC.) Rchb. Section Cieca P. apoda Harms Colombia, Caldas, Manizales Wild P. coriacea Juss. Colombia, Caldas, Palestina Wild P. coriacea Juss. Colombia, Tolima, Ibagué Wild P. coriacea Juss. Colombia, Valle del Cauca, Palmira Wild P. exoperculata Mast. Ecuador, Tunguragua, Ceballos Wild P. gracilis J. Jacq. ex Link USA, National Collection - Blois - France Wild P. holosericea L. National Collection - Blois - France Wild P. monadelpha P. Jorg. & Holm-Niels. Colombia, Valle del Cauca, El Cerrito Wild P. mutiflora L. National Collection - Blois - France Wild P. suberosa L. Colombia, Antioquia, Jerico Wild P. suberosa L. Colombia, Caldas, Manizales Wild Series Apetalae P. apetala Killip National Collection - Blois - France Wild Series Auriculatae P. auriculata Kunth Colombia, Caldas, Victoria Wild P. auriculata Kunth Guiana French, National Collection - Blois - France Wild P. jatunsachensis Schwerdtfeger National Collection - Blois - France Wild Series Luteae P. filipes Benth. Colombia, Risaralda, Pereira Wild Series Miserae P. amalocarpa Barb. Rodr. National Collection - Blois - France Wild P. misera Kunth Colombia, Valle del Cauca, Jamundi Wild P. misera Kunth National Collection - Blois - France Wild P. trifasciata Lem Colombia, Quindio, Buenavista - (introduced) Wild Section Decaloba P. tricuspis Mast. Bolivia, Santa Cruz Wild P. tricuspis Mast. Brasil, National Collection - Blois - France Wild Series Punctatae P. alnifolia Kunth Colombia, Valle del Cauca, El Cerrito Wild P. alnifolia Kunth Colombia, Caldas, Manizales Wild P. alnifolia Kunth Colombia, Nariño, Pasto Wild P. biflora Lam. Colombia, Caldas, Victoria Wild P. biflora Lam. Colombia, Tolima, Mariquita Wild P. boenderi MacDougal National Collection - Blois - France Wild P. bogotensis Benth. Colombia, Cundinamarca, Bogotá Wild P. colinvauxii Wiggins Ecuador, National Collection - Blois - France Wild 158 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ P. cuneata Wild. Colombia, Boyacá, Duitama Wild P. cuneata Wild. National Collection - Blois - France Wild Wild P. cupraea L. National Collection - Blois - France P. erytrophylla Mast. Colombia, Boyacá, Duitama Wild P. gilbertiana MacDougal National Collection - Blois - France Wild P. occidentalis Colombia, Valle del Cauca, Buenaventura Wild P. pohii Mast. National Collection - Blois - France Wild P. magdalenae Triana & Planch. Colombia, Caldas, Victoria Wild P. magdalenae Triana & Planch. Colombia, Cundinamarca, San Juan de Rio Seco Wild P. micropetala Mast. National Collection - Blois - France Wild P. vespertilio L. Guiana French, National Collection - Blois - France Wild P. yucatanensis Killip Mexico, National Collection - Blois - France Wild Series Sexflorae P. allantophylla Mast. National Collection - Blois - France Wild Section Eudecaloba P. aurantia G. Forst. Australia, National Collection - Blois - France Wild P. herbertiana Ker-Gawl. National Collection - Blois - France Wild Section Hahniopathanthus P. guatemalensis S. Watson Colombia, Caldas, Filadelfia Wild Section Organenses P. ornithoura Mast. National Collection - Blois - France Wild Section Pseudodysosmia P. adenopoda Moc. & Sessé ex DC. Colombia, Caldas, Manizales Wild P. adenopoda Moc. & Sessé ex DC. National Collection - Blois - France Wild P. karwinskii Mast. National Collection - Blois - France Wild P. lobata (Killip) Hutchinson ex MacDougal National Collection - Blois - France Wild Section Pseudogranadilla P. bicornis Mill. National Collection - Blois - France Wild P. bicornis Mill. National Collection - Blois - France Wild P. indecora Kunth National Collection - Blois - France Wild P. telesiphe Knapp & Mallet National Collection - Blois - France Wild Section Xerogona P. rubra L. Colombia, Quindio, Calarca Wild Subgenus Deidamioides (Harms) Killip, 1938 P. deidamioides Harms National Collection - Blois - France Wild P. lancetillensis MacDougal & Meerman National Collection - Blois - France Wild Subgenus Distephana (Juss,) Killip, 1938 P. aimae Annonay & Feuillet Guiana French, National Collection - Blois - France Wild P. coccinea Aubl. Guiana French, Belgian National Botanic Garden - Meise Wild P. quadriglandulosa Rodschied National Collection - Blois - France Wild P. speciosa Gardn. National Collection - Blois - France Wild P. variolata Poep & Endl. National Collection - Blois - France Wild P. vitifolia (Harv.) Harms Colombia, Chocó, Quibdo Home garden P. vitifolia (Harv.) Harms Colombia, Tolima, Ibagué Wild Subgenus Disosmioides Killip, 1938 P. campanulata Mast. Brasil, National Collection - Blois - France Wild P. setulosa Killip National Collection - Blois - France Wild Subgenus Dysosmia DC., 1938 P. foetida var. moritziana (Planch.) Killip ex Pull Guiana French, National Collection - Blois - France Wild P. foetida var. gossypiifolia (Desv.) Mast. P. arida (Mast. & Rose) Killip Colombia, Chocó, Quibdo National Collection - Blois - France Wild Subgenus Manicata (Harms) Escobar, 1988 P. macropoda Killip National Collection - Blois - France Wild 159 Wild Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ P. manicata (Juss.) Pers. Colombia, Quindio, Salento Wild P. manicata (Juss.) Pers. Ecuador, Tunguragua, Ambato Wild P. trisecta Mast. National Collection - Blois - France Wild Subgenus Murucuja (Medic.) Mast., 1871 P. murucuja L. National Collection - Blois - France Wild P. tulae Urban National Collection - Blois - France Wild Subgenus Pseudomurucuja (Harms.) Killip P. perfoliata L. National Collection - Blois - France Wild Subgenus Psilanthus (DC.) Killip, 1938 P. trinervia (Juss.) Poir. Colombia, Quindio, Salento Wild National Collection - Blois - France Wild Subgenus Passiflora Series Digitatae Killip P. serrato-digitata L. Series Incarnatae P. cincinnata Mast. Brasil Wild P. edulis f. edulis Sims Colombia, Cauca, Timbio Home garden P. edulis f. edulis Sims Colombia, Antioquia, Fredonia Wild P. edulis f. edulis Sims Colombia, Cundinamarca, San Juan de Río Seco Home garden P. edulis f. flavicarpa Degener Brasil, Araguari Cultivated P. edulis f. flavicarpa Degener Colombia, Amazonas, Leticia Home garden P. edulis f. flavicarpa Degener Ecuador, Guayas, Guayaquil Cultivated P. edulis f. flavicarpa Degener Perú, Santa Vilma Cultivated P. edulis f. flavicarpa Degener Colombia, Chocó, Quibdo Cultivated P. edulis f. flavicarpa Degener Colombia, Valle del Cauca, La Unión Cultivated P. edulis f. flavicarpa Degener Brasil, National Collection - Blois - France Wild P. incarnata L. USA, Florida, Miami Home garden P. incarnata L. USA, National Collection - Blois - France Wild Series Kermesinae P. edmundoi Sacco Brasil, National Collection - Blois - France Wild P. kermesina Link & Otto National Collection - Blois - France Wild P. lehmanni Mast Colombia, Quindio, Calarca Wild P. lehmanni Mast Colombia, Caldas, Manizales Wild Wild P. loefgrenii Vitta Brasil, National Collection - Blois - France P. miersii Mast. National Collection - Blois - France Wild P. smithii Killip Colombia, Tolima, Fresno Wild P. smithii Killip Colombia, Tolima, Ibagué Wild P. trisulca Mast. National Collection - Blois - France Wild Series Laurifoliae P. acuminata DC. National Collection - Blois - France Wild P. ambigua Hermsl. Colombia, Belgian National Botanic Garden - Meise Wild P. carinata National Collection - Blois - France Wild P. crenata Feuillet & Cremers National Collection - Blois - France Wild P. fernandezii L. K. Escobar Bolivia, National Collection - Blois - France Wild Wild P. gabrielliana Vanderplank Guiana French, National Collection - Blois - France P. laurifolia L. Guiana French, National Collection - Blois - France Wild P. nigradenia Rusby National Collection - Blois - France Wild P. nitida Kunth National Collection - Blois - France Wild P. nitida Kunth Colombia, Chocó, Quibdo Home garden P. odontophylla Harms ex Glaz. Brasil, National Collection - Blois - France Wild P. popenovii Killip Colombia, Cauca, El Tambo Home garden P. popenovii Killip Colombia, Nariño, Chachagui Home garden P. riparia Mart. ex Mast. National Collection - Blois - France Wild P. rufostipulata Feuillet Guiana French, Belgian National Botanic Garden - Meise Wild Series Lobatae 160 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ P. amethystina Mikan National Collection - Blois - France Wild P. caerulea L. France, Herault, Montpellier - (introduced) Home garden P. caerulea L. Brasil, National Collection - Blois - France Wild P. eichleriana Mast. National Collection - Blois - France Wild P. elegans Mast. National Collection - Blois - France Wild P. garckei Mast. Guiana French, Belgian National Botanic Garden - Meise Wild Wild P. garckei Mast. Guiana French, National Collection - Blois - France P. gibertii N. E. Brown National Collection - Blois - France Wild P. gritensis Karts. National Collection - Blois - France Wild P. mooreana Hook. National Collection - Blois - France Wild P. pallens Poepp. ex Mast. National Collection - Blois - France Wild P. sprucei Mast. National Collection - Blois - France Wild P. stipulata Aubl. National Collection - Blois - France Wild P. subpeltata Ortega National Collection - Blois - France Wild P. tucumanensis Hook. National Collection - Blois - France Wild P. urubicensis Cervi National Collection - Blois - France Wild Series Marginatae P. marginata Mast. National Collection - Blois - France Wild Series Menispermifolia P. crassifolia Killip National Collection - Blois - France Wild P. menispermifolia Kunth Colombia, Tolima, Ibagué Wild P. menispermifolia Kunth National Collection - Blois - France Wild P. nephrodes Mast. Bolivia, National Collection - Blois - France Wild P. reitzii Sacco National Collection - Blois - France Wild Series Quadrangulares P. alata Curtis Brasil Cultivated P. alata Curtis Brasil, National Collection - Blois - France Wild P. quadrangularis L. Colombia, Huila, Paicol Cultivated Series Serratifoliae P. bahiensis Klotzsch Brasil, Bahia, Salvador Wild P. serratifoliaL. National Collection - Blois - France Wild Series Simplicifoliae P. dispar Killip National Collection - Blois - France Wild P. galbana Mast. Guiana French, National Collection - Blois - France Wild P. mapiriensis Harms Guiana French, National Collection - Blois - France Wild P. oerstedii Mast. Colombia, Tolima, Ibagué Wild P. oerstedii Mast. Colombia, Cauca, Popayán Wild P. oerstedii var. choconiana (S. Watson) Killip National Collection - Blois - France Wild P. subrotunda Mast. National Collection - Blois - France Wild Series Tiliifoliae P. ligularis Juss. Colombia, Nariño, Ipiales Cultivated P. ligularis Juss. Colombia, Risaralda, Marsella Cultivated P. ligularis Juss. Colombia, Caldas, Manizales Wild P. ligularis Juss. Colombia, Quindio, Calarca Home garden P. maliformis L. Colombia, Santander, Barichara Home garden P. maliformis L. Colombia, Antioquia, Fredonia Home garden P. maliformis L. Colombia, Huila, La Plata Cultivated P. multiformis Jacq. Colombia, Norte de Santander, Ocaña Home garden P. palenquensis Holm-Niels. & Lawesson Colombia, Chocó, Quibdo Cultivated P. seemannii Griseb National Collection - Blois - France Wild P. serrulata Jacq. Colombia, Magdalena, Plato Wild P. tiliifolia L. Colombia, Quindio, Filandia Wild P. tiliifolia L. Colombia, Valle del Cauca, El Cerrito Wild Subgenus Tacsonia (Juss.) Tr. & Planch, 1873 161 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Section Colombiana Series Colombiana P. adulterina L. f. Colombia, Boyacá, Duitama P. adulterina L. f. Colombia, National Collection - Blois - France Wild P. lanata (Juss.) Colombia, Boyacá, Duitama Wild Series Leptomischae P. ampullacea (Mast.) Harms National Collection - Blois - France Wild P. antioquiensis Karst. Colombia, Caldas, Villamaria Wild P. antioquiensis Karst. Colombia, Antioquia, Santa Rosa de Osos Wild P. antioquiensis Karst. Colombia, Caldas, Manizales Wild P. flexipes Triana & Planch Colombia, Quindio, Salento Wild P. tenerifensis L. K. Escobar Colombia, Valle del Cauca, El Cerrito Wild Series Quindensae P. linearistipula L. K. Escobar Colombia, Caldas, Manizales Wild Section Bracteogamma P. cumbalensis var. cumbalensis (Karst.) Harms Ecuador P. cumbalensis var. goudotiana (Tr. & Planch.) L.K. Escobar Colombia, Boyacá, Duitama Wild Wild Wild P. luzmarina Jorgensen Ecuador, Loja, Loja Wild P. tarminiana Coppens & Barney Colombia, Caldas, Villamaria Home garden P. tarminiana Coppens & Barney Perú, Huancavelica Home garden P. tarminiana Coppens & Barney Venezuela, Táchira Home garden P. tarminiana Coppens & Barney Venezuela, Tachira Home garden P. tarminiana Coppens & Barney Colombia, Cauca, Slivia Wild P. tripartita var. mollissima (Kunth) Holm-Niel & Jorg Colombia, Boyacá, Nuevo Colón Cultivated P. tripartita var. mollissima (Kunth) Holm-Niel & Jorg Colombia, Cundinamarca, Tequendama Wild P. tripartita var. mollissima (Kunth) Holm-Niel & Jorg Venezuela, Táchira, Villa Paez, Betania Home garden P. parritae (Mast.) L. H. Bailey Colombia, Tolima, Herveo Wild Section Poggendorffia P. pinnatistipula Cav. Ecuador, Tunguragua, Ambato Wild Section Parritana Escobar Section Tacsonia P. matthewssii (Mast.) Killip Ecuador Wild P. mixta L. f. Colombia, Valle del Cauca, El Cerrito Wild P. mixta L. f. Colombia, Cauca, Totoró Wild Subgenus Tacsonioides (DC.) Killip, 1938 P. mendoncaei Harms National Collection - Blois - France Wild P. reflexiflora Cav. National Collection - Blois - France Wild P. umbilicata (Griseb.) Harms National Collection - Blois - France Wild Colombia, Salento, Quindío Wild Subgenus Tryphostemmatoides (Harms) Killip, 1938 P. gracillima Killip V.1.3.2. DNA Extraction and PCR-RFLP analyses Total genomic DNA was extracted from frozen leaves following the protocol of Doyle & Doyle (1987) with minor modifications and purified on anion exchange micro-columns (Qiagen). All samples were qualified on a 0.8% agarose gel, in 1x TBE buffer and electrophoresed at 110 V during 1 hour and visualized under UV light. 162 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Two cpDNA regions, intergeneric spacer (psbC- trnS and trnS - trnfM), and two mtDNA introns (nad4-1/2 and nad1-B/C), were amplified using PCR and the universal primer pairs indicated by Demesure et al. (1995) and evaluated in the subgenus Passiflora by Vargas (2000). The PCR products were subsequently digested with six endonucleases selected by Varón (2000) (Hae III, Hha I, Hinf I, Hpa II, Taq I, and Rsa I - BioLabs©; Table 2). The PCR reaction mix contained 25 ng of template DNA, 10 mM Tris-HCl, 50 mM KCL, 1.5 mM MgCl2, 0.15 µM of each primer, 2.5 µM of each dNTP and 10 units of Taq polymerase, in a total volume of 50 µl. PCR-amplification was performed at 95°C for 4 min for initial denaturation, followed by 35 cycles at 94°C for 30 s, 57°C for 1 min, 72°C for 3 min, and was terminated by 10 min at 72°C. To confirm successful amplification and to determine the size of amplified fragments, 8 µl of PCR products were separated by electrophoresis in a 1% agarose gel, in 1 x TBE buffer. The DNA fragments were visualized by UV fluorescence after staining with ethidium bromide. The approximate product size was calculated by comparison of the migration distance of the PCR product with a DNA marker (1-kb ladder). Ten microliters of each PCR product was restricted in a volume of 20 µl containing 5 units of restriction enzyme for 3 hours at 37°C, according to the manufacturer’s procedures. Restriction fragments were separated on 1.2% agarose gels containing ethidium bromide in 1x Trisborate EDTA (TBE), run at 110 V for 3 hours, and visualized under UV light. Figure 1 shows the general schema of the PCR-RFLP markers. Table 2. DNA sequence and type of primer pairs used in the present study by Demesure et al. (1995). Primers psbC [psII 44 kd protein] trnS [tRNA-Ser (UGA)] trnS [tRNA – Ser (UGA)] trnfM [tRNA- fMet (CAU)] nad4 exon 1 nad4 exon 2 nad1 exon B nad1 exon C Code PC1 – PC2 Sequence (5’- 3’) GGTCGTGACCAAGAAACCAC GGTTCGAATCCCTCTCTCTC GAGAGAGAGGCATTCGAACC CATAACCTTGAGGTCACGGG CAGTGGGTTGGTCTGGTATG TCATATGGGCTACTGAGGAG GCATTACGATCTGCAGCTCA GGAGCTCGATTAGTTTCTGC TS1 – TS2 N41 – N42 N1B – N1C 163 Organelle - DNA type Chloroplast Chloroplast Mitochondrial Mitochondrial Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1.3.3. Data analysis Most restriction patterns were impossible to interpret in terms of mutations, because of the high number of bands. Instead, these were scored for presence (1) or absence (0) and used to identify chloroplast and mitochondria haplotypes, generating a binary data matrix for the subsequent phenetic analyses. The Sokal & Michener (1958) coefficient of genetic similarity was calculated for each pair of haplotypes. The similarity matrix was employed in a principal co-ordinate analysis (PCO) and the construction of a phenogram by the neighbor-joining method (Saitou & Nei, 1987), calculating bootstrap values from 1000 replicates. These analyses were performed using the DARwin 5.0 software (Perrier et al., 2003). A/A R R B/B A/B R R R R R R DNA PCR + Digestion with R + gel electrophoresis Figure 1. General schema of the PCR-RFLP markers (CAPS). 164 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1.4. Results V.1.4.1. PCR amplification All four primer pairs used in the present study successfully amplified the corresponding cpDNA and mtDNA regions in 151 species of genus Passiflora and four species of Adenia, Barteria and Smeathmannia, producing fragment whose length varied from 820 to 2,040 bp (Table 3). Polymorphism for chloroplast fragment (PC1-PC2) distinguishes P. tripartita var. mollissima, P. antioquiensis, P. pinnatistipula and P. murucuja from all other species. Four different fragment lengths were observed for TS1-TS2, two for subgenus Decaloba, with a difference between Australian and American species, one for three species of subgenus Passiflora series Laurifoliae (P. popenovii, P.nitida and P. laurifolia), and one for all the other species. Eight fragment lengths are observed for the mitochondrial N1B-N1C, discriminating P. coriacae, P. multiflora, P. perfoliata and all the species of subgenus Tacsonia. For the N41-N42 region, only one fragment was detected with a weak degree of amplification. Table 3. Numbers of haplotypes and fragments for each combination primer/ enzyme. Primers Size of amplification products (bp) Degree amplification Hinf I Rsa I Hpa II Hae III Hha I Taq I Total number haplotypes / fragments PC1 – PC2 TS1 – TS2 N41 – N42 N1B – N1C 1875-1896-1930-2000 820-1200-1250-1345-14001470-1500-1550-1650-16931735-1780-1810 2036 1500-1580-1605-1690-18851910-2000-2040 strong 13 / 17 2 / 4* 19 / 18 29 / 34 27 / 34 15 / 16 strong 60 / 28 34 / 26 32 / 46 10 / 11 13 / 16 30 /22 weak 27 /22 34 / 32 48 / 59 36 / 31 50 / 47 22 / 29* good 26 / 29 17 / 21 27 / 22 20 / 22 16 / 16 12 / 31 * not interpretable. V.1.4.2. Restriction analysis All twenty-four fragment/enzyme combinations revealed polymorphism, but restriction profiles could not be read in two cases [(N41-N42/TaqI) and (PC1PC2/RsaI)]. A total of 614 fragments were scored, of which 93% were found to be polymorphic in the sample (Table 3). All the polymorphism was due to insertion-deletion (indel) mutations. Because of the high number of fragments in relation to the sample size (213 individuals), banding 165 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ patterns could not be interpreted in terms of particular mutations. Figure 2 presents examples of the interspecific variation for cp/mtDNA. The PCR-RFLP analysis of the chloroplast regions showed polymorphism with all enzymes, resulting in 268 interpretable polymorphic fragments for the 11 fragment/enzyme combinations. The 11 mitochondrial fragment/enzyme combinations exhibited a high polymorphism too, with 307 interpretable fragments, but the resolution of the fragments was less clear. The combination N41-N42/HpaII was by far the most Ladder 1kb P. oerstedii P. maliformis P. smithii P. lehmannii P. alata P. ambigua P. cincinnata P. garckei P. popenovii P. trisulca P. caerulea P. quadrangularis P. citrifolia P. coriacea P. filipes P. erytrophylla P. suberosa P. cuneata P. magdalenae P. linearistipula P. antioquiensis P. parritae P. tricuspis P. alnifolia P. luzmarina P. flexipes P. mixta P. tarminana Ladder 1kb polymorphic one, with 59 fragments differentiating among genera and subgenera. 12.216 bp. 3.054 bp. 2.036 bp. 1.636 bp. 1.018 bp. 506 bp. 396 bp. 344 bp. 298 bp. 220 bp. a TS1II)) TS1-TS2 (Hpa (HpaII 12.216 bp. 3.054 bp. 2.036 bp. 1.636 bp. 1.018 bp. 506 bp. 396 bp. 344 bp. 298 bp. 220 bp. b N1B1II)) N1B1-N1C (Hpa (HpaII Figure 2. Interspecific variation for cpDNA (a) and mtDNA (b) among different Passifora species. 166 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1.4.3. PCR-RFLP haplotypes A total of 280 chloroplast haplotypes and 372 mitochondrial haplotypes were found (Table 3). Their global patterns of distribution among genera and subgenera are summarized in Table 4. For the chloroplast, the TS1-TS2 region displayed more variation (177 haplotypes) than the PC1-PC2 region (103 haplotypes). Chlorotypes clearly distinguished subgenus Decaloba from the rest of the sample, as its species exhibited two or three specific haplotype patterns in all fragment/enzyme combinations. Subgenus Astrophea showed specific chlorotypes after the restriction of the TS1-TS2 region with Rsa I and Taq I. Subgenus Distephana was differentiated by the TS1-TS2/Taq I combination. Subgenera Dysosmia, Manicata and Tacsonia shared haplotypes for many combinations, so they cannot be clearly distinguished from each other. Subgenus Tacsonia appeared highly uniform for most combinations. Within subgenus Passiflora, most species of series Laurifoliae presented specific restriction patterns of the TS1-TS2 fragment (Hha I, Rsa I, and Taq I. Several combinations (Hinf I and Taq I) allowed distinguishing another morphological group, the typical representatives of series Tiliifoliae (P. ligularis, P. palenquensis, P. tiliifolia, P. seemannii). The outgroup (Adenia, Barteria and Smeathmannia) showed specific chlorotypes in most combinations. When this was not the case, they usually showed similarities with species of subgenus Passiflora. A higher number of mtDNA haplotypes were observed, 245 for the N41-N42 region and 127 for the N1B-N1C region. Subgenus Decaloba again displayed many specific haplotypes, however it shared several mitotypes with Barteria and Smeathmannia. All accessions of subgenus Tacsonia but four were clearly differentiated by highly specific and uniform mitotypes for the N1B-N1C regions. The four exceptions were Ecuadorian accessions of P. pinnatistipula, P. matthewsii, P. luzmarina and P. cumbalensis, sharing most restriction patterns with subgenus Passiflora. Subgenus Murucuja was differentiated by the N1B-B1C/Hha I combination. 167 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ In most species represented by several individuals, intraspecific variation was observed for both cpDNA and mtDNA. Although represented by two accessions only, P. foetida and P. vitifolia displayed the highest diversity of haplotypes. Other remarkable cases are those of P. ligularis, P. maliformis, P. alata, P. caerulea, and P. sphaerocarpa. 168 Table 4. Global distribution of the haplotypes among the genera and subgenera studied. Primers Hinf I Rsa I PC1 – PC2 Decaloba two groups - Adenia - Barteria Smeathmannia - other subgenera Impossible to interpret TS1 – TS2 Decaloba two groups - Passiflora four groups - (Barteria Smeathmannia) other subgenera N41 – N42 N1B – N1C Hpa II Hae III Hha I Taq I Decaloba three groups Decaloba - Adenia - other subgenera Smeathmannia - other subgenera Decaloba three groups Decaloba - other subgenera - Passiflora two groups - Tacsonia two groups - other subgenera Decaloba three groups - Passiflora four groups - Astrophea other subgenera Decaloba two groups - Passiflora three groups - Adenia (Barteria Smeathmannia) other subgenera Decaloba three groups - Passiflora two groups - other subgenera Decaloba two groups - Passiflora two groups - other subgenera Decaloba three groups - Passiflora two groups - Distephana Astrophea - Adenia (Barteria Smeathmannia) other subgenera Decaloba two groups - Passiflora Astrophea - (Barteria - Smeathmannia) other subgenera Decaloba three groups - Passiflora two groups - Adenia other subgenera Decaloba two groups Passiflora four groups - Tacsonia two groups - Barteria two groups - other subgenera Decaloba two Passiflora five groups - Adenia - other subgenera Decaloba two groups - Passiflora three groups - Barteria two groups - other subgenera Impossible to interpret Decaloba three groups - Passiflora two groups - Tacsonia two groups - (Barteria Smeathmannia) Adenia - other subgenera Decaloba two groups - Passiflora two groups - Tacsonia (Barteria Smeathmannia) Adenia - other subgenera Decaloba two groups - Passiflora two groups - Tacsonia two groups - (Barteria Smeathmannia) other subgenera Decaloba two groups - Passiflora two groups - Tacsonia two groups - other subgenera Decaloba two groups Decaloba two groups - Passiflora two - Tacsonia - Barteria groups - Tacsonia two two groups - Adenia groups - Murucuja Adenia - other subgenera Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1.4.4. Principal co-ordinates analysis The first two axes of the PCO on cpDNA data accounted for 59% of the total variation (Figure 3). They allowed visualizing a strong structure in the diversity under study, as the genera Adenia, Barteria, and Passiflora subgenera Astrophea, Calopathanthus, Dysosmia, Distephana, Manicata, Passiflora, Tacsonia, Tacsonioides and Tryphostemmatoides, and P. deidamioides occupy the left half of the principal plan, while the species of Passiflora subgenera Apodogyne, Decaloba, Murucuja, Pseudomurucuja and Psilanthus form a very well separated group, placed in a quite extreme position on the right, only the accessions of Smeathmannia and P. lancetillensis (subgenus Deidamioides) taking intermediate positions. Within genus Passiflora, only subgenus Astrophea is individualized. The subgenera on the right side form a wide group that we shall call the “Decaloba group”, by analogy with the “Decaloba clade” obtained in previous molecular work (Muschner et al., 2003; Yockteng, 2003; Yockteng & Nadot, 2004, Hansen et al., 2006) and with the subgenus Decaloba of the new classification (Feuillet & MacDougal, 2003). In the same way, the major group on the left will be called the “Passiflora group”. In the latter, only a small group dominated by species of subgenus Passiflora series Laurifoliae are separated from a bulk comprising species of subgenera Tacsonia, Manicata, Calopathanthus, Dysosmia, Distephana and Tacsonioides. P. gracillima (subgenus Tryphostemmatoides) is positioned between the ‘Passiflora group’ and subgenus Astrophea. V.1.4.5. Cluster analysis The Neighbor-Joining tree obtained with cpDNA data shows three major, well-supported clusters, within genus Passiflora (Figure 4). The first one corresponds to the ‘Passiflora group’, i.e. subgenera Calopathanthus, Deidamioides, Distephana, Dysosmia, Dysosmioides, Manicata, Passiflora, Tacsonia, and Tacsonioides, the second one to subgenus Astrophea, the third one to the ‘Decaloba group’, i.e. most species of subgenera Apodogyne, Decaloba, Murucuja, Pseudomurucuja and Psilanthus. P. gracillima (subgenus Tryphostemmatoides) appears in basal position, forming a fourth cluster of its own. The position of P. lancetillensis (subgenus Deidamioides) is undefined, as it is 170 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ placed on a long branch between the outgroup and the ‘Decaloba group’ clusters, the outgroup itself taking an undefined position among the three Passiflora major clusters. Figure 3. Principal co-ordinates on cpDNA data (PC1-PC2 and TS1-TS2 regions) estimated with 268 CAPS marker. The Astrophea cluster shows no particular substructure, however accessions are logically grouped by species. Within the ‘Decaloba group’, P. karwinskii (section Pseudodysosmia) and the Australian species (section Eudecaloba) take a basal position. After their separation, the ‘Decaloba group’ is split in two loose subclusters. The first one contains small branches corresponding to particular sections as Pseudodysosmia and Cieca or to series Auriculatae of section Decaloba (P. auriculata and P. jutasanchensis). The second subcluster gathers species of subgenera Murucuja, Pseudomurucuja, Psilanthus, section Pseudogranadilla of subgenus Decaloba and series Apetalae, Miserae, Punctatae, Sexflorae, and Luteae of section Decaloba. This weak structure 171 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ related to such low-level infrageneric taxa contrasts with the complete lack of differentiation among higher-level divisions of Killip, i.e. the subgenera constituting the ‘Decaloba group’. Thus, P. trinervia (subgenus Psilanthus), P. murucuja (subgenus Murucuja), P. tulae (subgenus Murucuja), and P. perfoliata (subgenus Pseudomurucuja) are well integrated among species of subgenus Decaloba. In addition, different accessions from a same species do not cluster close together, as in P. bicornis, P. magdalenae, P. biflora and P. tricuspis. Within the ‘Passiflora group’, resolution is very poor, and the substructure very weak. Branches are short and not supported. The only clear subcluster roughly corresponds to the small division of the ‘Passiflora group’ in the PCO principal plane. Species from a same subgenus do not tend to form even loose subclusters, with the partial exception of tacsos that show a higher uniformity. Even intraspecific variation can be compared in many cases to interspecific variation. 172 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Figure 4. Phenogram derived from on cpDNA (PC1-PC2 and TS1-TS2 regions) data illustrating the distribution of the different Passiflora subgenera studied. Boostrap values above 20% are indicated under branches. 173 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Figure 4a. Cluster analysis on cpDNA data, ‘Decaloba group’, subgenera Astrophea and Tryphostemmatoides, and outgroup genera. 174 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Figure 4b. Cluster analysis on cpDNA data, ‘Passiflora group’. Subgenus Tacsonia in clear-brown. 175 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ The dendrogram obtained with mtDNA data separates four poorly to moderately supported clusters (Figure 5). As for cpDNA, a first large cluster corresponds to the ‘Decaloba group’. Its accessions are grouped by species, with the only exception of P. adenopoda. Again the Australian P. aurantia/P. herbertiana are paired, as is the case for P. suberosa and P. coriacea. The numerous accessions of section Decaloba series Punctatae and Miserae tend to form a subcluster, while species of subgenera Apodogyne, Murucuja, Pseudomurucuja and Psilanthus are diluted in subgenus Decaloba. The other major clusters are different from those evidenced by cpDNA data. The ‘Passiflora group’ is split among three clusters. The largest one includes a subcluster composed of all species of subgenus Astrophea, but two. P. gracillima (subgenus Tryphostemmatoides) is basal to this Astrophea subgroup. Within this wide PassifloraAstrophea cluster, accessions tend to cluster by species, although strong divergences may appear within particular species, as is the case for P. alata and P. vitifolia. Some loose subclusters can be interpreted, as one dominated by part of the series Laurifoliae, a branch grouping the most typical species of series Tiliifoliae (P. ligularis, P. palenquensis, P. seemannii and P. tiliifolia), a branch grouping four species of subgenus Distephana, and a branch bearing three Ecuadorian and one southern Colombian accessions of subgenus Tacsonia. On the other hand, the most typical representatives of series Incarnatae, i.e. the two forms of P. edulis and P. incarnata, are clearly separated on the tree. Another surprising splitting is that of the morphologically uniform series Quandrangulares, with the divergence of P. alata and P. quadrangularis. A third important cluster is consistently composed by the majority of species from series Menispermifoliae, Kermesinae, and Lobatae. However, at lower levels, this morphological consistency is lost, as indicated by the separation of the very similar P. lehmannii, P. trisulca and P. smithii. Even more surprising is the separation of the morphologically very close P. gibertii and P. subpeltata, the latter on a smaller, distinct cluster, together with two other representatives of series Lobatae. 176 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ The fourth subcluster is very uniform, as it includes all species of subgenus Tacsonia, except the four Ecuadorian accessions mentioned above and an accession of P. pinnatistipula, also from Ecuador, which appears associated with the morphologically very similar P. macropoda in the Passiflora-Astrophea cluster. A few species constitute a cluster of their own. Among them, we find again P. deidamioides, (subgenus Deidamioides), but this time at the base of the Decaloba cluster, P. umbilicata (subgenus Tacsonioides), P. serratifolia and P. multiformis (subgenus Passiflora), and P. haematostigma (subgenus Astrophea), adding to an already difficult interpretation of the mitochondrial tree. Two other surprising informations are the positioning of P. lancetillensis within the Decaloba cluster and the splitting of the two varieties of P. foetida, var. hispida appearing in the Passiflora-Astrophea cluster and var. gossypifolia appearing in the Lobatae-Menispermifoliae-Kermesinae cluster. Last but not least, the African Passifloraceae of the sample again take an abnormal position for an outgroup. 177 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Figure 5. Phenogram derived from on mtDNA (N41-N42 and N1B-N1C regions) data illustrating the distribution of the different Passiflora subgenera studied. Boostrap values above 20% are indicated under branches. 178 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Fugure 5a. Cluster analysis on mtDNA data, ‘Passiflora group’ and subgenera Astrophea and Tryphostemmatoides. 179 Figure 5b. Cluster analysis on mtDNA data, ‘Decaloba group’. Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ V.1.5. Discussion V.1.5.1. Chloroplast DNA diversity Concerning cpDNA diversity, our results are globally consistent with those previously obtained by Muschner et al. (2003) on trnL-F sequences, by Yockteng (2003) with matK sequences (for the best resolved branches of the tree), and Hansen et al. (2006) with trnLtrnT sequences, as we observed three major groups, clearly corresponding to the three major clades of these phylogenetic studies and to the three major subgenera proposed by Feuillet & MacDougal (2003). The clusters/clades corresponding to subgenus Astrophea appear well differentiated and uniform in all studies. Within the ‘Decaloba group’, there appears no structure corresponding to Killip’s subgenera Apodogyne, Decaloba, Murucuja, Pseudomurucuja and Psilanthus, so their fusion into a unique subgenus is again justified by our cpDNA data. The marginal position of the Australian species can be interpreted by an ancient isolation. The two loose subclusters observed after their separation, are partly paralleled by subclades in the trnL-F and trnL-T trees. In the first one, we can recognize the association between P. coriacea and P. suberosa and, more distantly, with P. capsularis and P. morifolia; while subclades including species of series Organenses, Miserae and Punctatae of subgenus Decaloba and/or subgenera Murucuja and Pseudomuruja are observed in both trees. This subdivision within the Decaloba group is not very reliable, mostly because of the intraspecific variation that sometimes results in the placement of accessions of the same species in different subclusters or branches (e.g. P. adenopoda and P. bicornis). However it appears much more consistent with the classification of subgenus Decaloba proposed by Feuillet & MacDougal (2003) than with Killip’s, as all the species grouped in the second subcluster are classified in supersection Decaloba - section Decaloba by the former authors. In the three cpDNA studies, the ‘Passiflora group/clade’ is the one showing the loosest structure and the highest number of inconsistencies between morphological and genetic similarities. The subgenera that compose it are not differentiated as their species tend to be widely dispersed among poorly supported subclusters/subclades, with the exception of P. foetida (subgenus Dysosmia), whose accessions appear as a well supported sister 181 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ group of the ‘Passiflora clade’ in the matK tree of Yockteng (2003), while the representatives of this very particular subgenus are not individualized, neither in the trnLF tree of Muschner et al. (2003) nor in our study. In our tree, the three accessions of P. foetida and P. arida even appear very dispersed within the ‘Passiflora group’. This observation can be generalized to most species represented by several accessions, which underlines that intraspecific variation is of the same order as interspecific and even intersubgeneric variation. Only two small clusters are supported by bootstrap values exceeding 50% and minimal taxonomic interpretation. One includes all species of series Laurifoliae of subgenus Passiflora, except P. fernandezii and P. odontophylla, plus three species of other series, P. crassifolia (series Menispermifoliae) P. kermesina (series Kermesinae), P. oerstedii var. choconiana (series Simplicifoliae). The other one includes three accessions of P. tripartita var. mollissima and one of P. antioquiensis. Killip’s subgenus Tryphostemmatoides is supported by the placement of P. gracillima in our tree, in an independent branch, but slightly closer to subgenus Astrophea. Its close relative, P. tryphostemmatoides, was placed among the arborescent and semi-arborescent species of subgenus Astrophea in the matK study of Yockteng (2003) and as a sister group of subgenus Astrophea in the nuclear ncpGS tree of Yockteng & Nadot (2004). Thus, the three datasets converge in providing reasons to maintain these small herbaceous vines in a distinct subgenus, surprisingly associated with the arborescent or semiarborescent species of subgenus Astrophea. Feuillet & MacDougal classified them as a section of their fourth subgenus, Deidamioides. However, the type species of this subgenus, i.e. P. deidamioides, is placed in the Passiflora group in our tree, which contradicts this view. P. lancetillensis and P. microstipula, two species of subgenus Deidamioides sensu Killip, were transferred to a supersection Pterosperma of subgenus Decaloba in the new classification. In our study, P. lancetillensis is placed at the base of the Decaloba group, but in a distant position, as in the phylogenetic trees of Muschner et al. (2003) and Hansen et al. (2006). In the latter, P. microstipula was very strangely dispersed, with one accession close to P. lancetillensis, and two accessions in the ‘Passiflora clade’, a fact that was later attributed to a case of heteroplasmy (Hansen et al. 2007). In any case, neither the contours of Killip’s subgenus Deidamioides, nor its 182 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ splitting between subgenus Deidamioides and supersection Pterosperma by Feuillet & MacDougal (2003) are clearly supported by cpDNA data. V.1.5.2. Mitonchondrial DNA diversity The results on mtDNA diversity are consistent with those from cpDNA on two main points at the intersubgeneric level. These are the clear separation of the ‘Decaloba group’ and the unexpected placement of the outgroup species, at a comparable distance. Within the ‘Decaloba group’ cluster, we recognize several structural traits observed on cpDNA, such as a better grouping of accessions by species, as compared to the situation in the ‘Passiflora group’, the integration of subgenera Apodogyne, Decaloba, Murucuja, Pseudomurucuja and Psilanthus, the formation of a subcluster supporting the section Decaloba of subgenus Decaloba in the new classification, and associations between particular species as P. suberosa and P. coriacea or the two Australian species. A third important convergence at the subgeneric level is the grouping of P. gracillima (subgenus Tryphostemmatoides) with most species of subgenus Astrophea. The splitting of the ‘Passiflora group’ and the insertion of the AstropheaTryphostemmatoides cluster in its largest division are the two most striking divergences with all chloroplastic and nuclear DNA, as well as morphological data (except for the separation of tacsos), and cannot be explained on the basis of any of the two subgeneric classifications. The largest cluster resulting from the division of the ‘Passiflora group’ shows some loose clustering for species from subgenus Distephana (four of its six representatives), series Tiliifoliae, or simply for accessions from a same species, whereas other clear morphological groups are split, as series Quadrangulares and Incarnatae. The most uniform cluster dissident from the ‘Passiflora group’ is constituted by the bulk of subgenus Tacsonia species. The five other accessions of this subgenus, placed in the largest ‘Passiflora group’ subcluster, share a common recent Ecuadorian or southernColombian origin, constituting a particular case of geographic structure in the data. Interestingly, P. umbillicata, of the subgenus Tacsonioides, also characterized by a relatively long hypanthium, is placed at the base of the Tacsonia subcluster, while P. manicata is placed in the larger Passiflora-Astrophea cluster. The third ‘Passiflora 183 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ group’ cluster seems less easy to interpret, however it integrates remarkably the series Kermesinae, Simplicifoliae, Lobatae, and Menispermifoliae. All these series are differentiated in Killip’s key by their foliaceous, semi-ovate to semi-oblong stipules, “attached on one side above base, hence often appearing reniform”. As for the Tacsonia cluster, some representatives of these series are placed in the larger Passiflora-Astrophea cluster: P. galbana, P. mapiriensis and P. subrotunda (Simplicifoliae), P. elegans, P. tucumanensis (Lobatae), P. reitzii (Menispermifoliae). In addition, three species of series Lobatae, P. mooreana, P. pallens and P. subpeltata form a small independent cluster. In conclusion, the series Kermesinae, Simplicifoliae, Lobatae, and Menispermifoliae can be compared to those of subgenus Tacsonia for their relative divergence from the bulk of the ‘Passiflora group’, however they show a much higher differentiation, both between divergent clusters and within their two specific clusters. The presence of P. racemosa, the unique species of Killip’s subgenus Calopathanthus, among the species of these four series, is not really surprising, considering the shape of its stipules. In contrast, the presence of one of the accessions of P. foetida in this same cluster is very surprising. As in the cpDNA tree, the relative distances between genus Passiflora, or its main divisions, and the outgroup, as well as the placement of the latter, do not provide any support to the monophyly of the genus. V.1.5.3. Divergences in the evolutions of chloroplast and mitochondrial genomes The analyses of chloroplastic and mitochondrial fragments gave very different pictures on the genetic structure of genus Passiflora. Differences appear at all levels, in the position of the outgroup, the relative position of four subgenera, and the relationships between species. According to cpDNA data, the three genera of the outgroup are placed near the Decaloba cluster, giving more emphasis to its separation from the two other subgenera. In the mtDNA tree, they are placed between divisions of the ‘Passiflora group’. In addition, 184 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ while Adenia appears more closely related to Smeathmannia for the chloroplast sequences, it appears more closely related to Barteria for the mitochondrial sequences. The divergences between both datasets have opposite consequences for Astrophea and Tacsonia, the former being clearly differentiated by cpDNA but not so for mtDNA, and the latter being clearly differentiated by mtDNA and diluted in the ‘Passiflora group’ in the cpDNA tree. Subgenus Tryphostemmatoides appears associated with subgenus Astrophea in both datasets. Its independence in the cpDNA tree is blurred in the mtDNA tree as both get integrated in the ‘Passiflora group’. The fourth case concerns the small subgenus Deidamioides. It is more complex as its two representatives are separated in both analyses, P. lancetillensis appearing in basal position close to the ‘Decaloba group’, while its type species, P. deidamioides, is placed in the ‘Passiflora group’ in the cpDNA tree, but at the base of the ‘Decaloba group’ in the mtDNA tree. At the infrasubgeneric level, clustering of chlorotypes and haplotypes appear globally consistent for the ‘Decaloba group’, while we found replicas of the intersubgeneric discrepancies within the ‘Passiflora group’. Thus series Laurifoliae and Incarnatae tend to differentiate in the cpDNA tree, but not in the mtDNA tree, while the reverse is true, and clearer, for series Tiliifoliae and the Kermesinae-Lobatae-MenispermifoliaeSimplicifoliae series. The divergence in the information obtained from chloroplast and mitochondrial genomes could be due to differences in their rate of evolution and mode of transmission. While most phylogenetic studies are based on the chloroplast genome because of its maternal transmission and slow evolution, as compared to the nuclear and mitochondrial genomes, which makes them less sensitive to hybridization events, this assumption has been contradicted in a significant number of cases. In their review of the question, Harris and Ingram had already shown in 1991 that mutation and hybridization/introgression may generate considerable intraspecific variation, constituting a potential problem for cpDNA-based phylogenetic reconstructions. According to their revision, 27% of the families and 21% of the genera show potential for biparental inheritance, and 23% of the 185 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ species show only biparental inheritance, suggesting the existence of a continuum rather than an alternative mode of transmission. A few studies have also shown the possibility of intraindividual variation (heteroplasmy), resulting from biparental transmission, as in Pelargonium (Metzlaff et al., 1981), Oryza sativa (Moon et al., 1987), Medicago (Lee et al., 1988), Musa (Fauré et al., 1994), Cucumis (Harvey et al., 1998) and Actinidia (Chat et al., 2004). In Actinidia, where transmission of chloroplasts is paternal while that of mitochondria is maternal, chloroplastic phylogenetic information provides clear evidences of conflicts with morphological classification and striking incongruences have been observed between maternal and paternal phylogenies, evidencing reticulate evolution related to frequent hybridization events, sometimes involving distant species (Chat et al., 2004). Passiflora probably constitutes a very similar case, where reticulate evolution could account for the fast radiation in the genus, as well as the lack of clear morphological discontinuities at subgeneric levels, while variable patterns in transmission and evolution of organellar genomes explain striking incongruences between different datasets, reflecting multiple origins of species. Indeed, since the observations of Corriveau & Coleman (1988) led them to suspect biparental plastid transmission in Passiflora, several molecular studies have pointed to a high frequency of paternal or biparental inheritance of the chloroplast genome in the genus. Studying a case of genome-plastome incompatibility in Passiflora hybrids, Mráček (2005) showed that biparental transmission of chloroplasts resulted in heteroplasmy of the whole plant, down to the single leaf level, and perhaps even the single cell level. The case of heteroplasmy reported by Hansen et al. (2007) involves two “extremely divergent chloroplast types” in a single individual. All species where biparental and/or paternal transmission was detected belong to the ‘Passiflora group’ (including subgenus Dysosmia), while the only cases studied in subgenus Decaloba evidenced maternal transmission (Do et al., 1992; Mráček, 2005; Muschner et al., 2006) or predominantly maternal transmission (Hansen et al., 2007), which would then explain lower intraspecific variation in the latter. More studies are needed to ascertain whether this was coincidence or one more fundamental difference between the ‘Decaloba group’ 186 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ and the ‘Passiflora group’. Concerning mitochondrial inheritance, mtDNA was maternally transmitted in the five hybrids studied by Muschner et al. (2006). For Passifloraceae phylogenetic studies, the consequences of paternal/biparental transmission of chloroplasts would include all the most severe problems anticipated by Harris & Ingram (1991), i.e. high intraspecific and intraindividual variation, random survivorship of lineages from a polymorphic progenitor (lineage sorting), reticulate evolution of the chloroplast genome through hybridization and introgression, all factors that, interfering with interspecific genetic differentiation, may explain the confuse situation, particularly in the ‘Passiflora group’. The differentiation among the major group/clades (‘Passiflora’, ‘Decaloba’, Astrophea) would limit genetic exchanges between them, explaining the distances observed in most studies. If the maternal transmission of mtDNA observed by Muschner et al. (2006) were further confirmed, this genome would become essential in the elucidation of the relations between Passiflora species. However, the numerous cases of wide intraspecific variation for mtDNA in our data suggest that there is no such clearcut difference in the transmission of the two organellar genomes. V.1.5.4. Diversity of organellar genomes and Passiflora systematics Obviously, the central expectation in the recent molecular studies of Passiflora was that they would give key elements for the understanding of its evolution and the definition of its subgeneric divisions, compensating for the pitfalls in the morphological diversity analysis. What they have produced instead is a series of divergent pictures of Passiflora diversity, one for each genome, plus one from morphological diversity, while providing strong evidence for reticulate evolution in the genus and, very likely, the whole family. The divergences between corresponding datasets are such that they disqualify any intent of deriving a consensual picture by simply piling up these datasets in a global analysis. In other words, there is no basis for a consensus tree. All we can do is trying to draw the clearest lessons from the different analyses. 187 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ While all inferior divisions of the two competing taxonomical treatments only appear occasionally supported by molecular and morphological data, revealing our general tendency to overclassification, the three main divisions of the new classification proposed by Feuillet & MacDougal appear clearly in all molecular studies. The situation for smaller clades is less clear. The fourth subgenus, Deidamioides, is not supported, while our data confirm the statement of Yockteng & Nadot (2004) supporting the persistence of the small subgenus Tryphostemmatoides. All studies on cpDNA diversity have indicated a higher differentiation within the ‘Decaloba group/clade’, and a better relation between genetic and morphological diversities. In addition, our data show a better relation between chlorotype and mitotype diversities in this group. This may be attributed to one or several of three causes. A first hypothesis is that this taxon is more ancient than the other major groups/clades, which would be consistent with the fact that it is present also in Asia and Australia, while the others are of strictly American origin. A second one would be that these mostly herbaceous vines present a shorter generation time, contributing to a faster evolution (Yockteng, 2003). A third likely explanation is that maternal transmission of cpDNA is the most common case in the ‘Decaloba group’. In addition, if both genomes are mostly transmitted in similar ways, it also explains why mtDNA and cpDNA trees are relatively congruent only in this group. The ‘Passiflora group/clade’, as defined by cpDNA and nuclear DNA data, appears as the group where reticulation events have the wider impact on radiation and evolution. Its division in three clusters in the mtDNA tree is the most difficult point to interpret. In this division, the only logics we can see follow ecological adaptation and morphological differentiation. The best-differentiated and most uniform cluster corresponds to subgenus Tacsonia. This taxon is composed of species specifically adapted to the cool conditions of Andean highlands, between 2,500 and 4,000 m.a.s.l, and pollination by the long-billed hummingbird Ensifera ensifera. They have developed a particularly long hypanthium, so the presence of the long-tubed P. umbillicata in a neighbor branch of the tree is not surprising. This young group clearly constitutes the most efficient adaptative answer of 188 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Passiflora to the rise of the Andes, as they are responsible for a species diversity peak along the elevational gradient (see Chapter III). Their recent evolution explains their mtDNA and morphological uniformity (see Chapter IV). The formation of a geographic subcluster by four accessions of Tacsonia from Ecuador and southern Colombia within the largest cluster of Passiflora, close to lowland species, indicates the possibility of capture of mitochondrial genetic material through relatively wide hybridization, although this appears much less frequent than exchanges of chloroplast material. The relative separation of series Kermesinae, Simplicifoliae, Lobatae, and Menispermifoliae shows a similar correspondence with morphological traits (stipule shape) and ecoclimatic adaptation, although the climatic trend is less marked. Most species represented in this cluster are vines adapted to the mild climates of Andean hillsides, between 1,200 and 2,000 m, or of subtropical regions of southern South America (northern Argentina to Brazilian Minas Gerais). By contrast, the Passiflora-Astrophea-Tryphostemmatoides cluster is a composite group, including the trees and treelets of subgenus Astrophea and all vigorous lianas of subgenera Passiflora and Distephana, mostly originating from tropical lowlands or Andean foothills, plus their close relatives from other climates. Thus, in the series Incarnatae, P. incarnata is adapted to the temperate or subtropical conditions of the southern USA, while P. edulis f. edulis is adapted to the mild conditions of southern South America. In the series Tiliifoliae we find two high hillside species, P. tiliifolia and P. ligularis (1,500-2,500m). Interestingly, we also find species of the reniform stipule cluster and species that are intermediate between subgenera Passiflora and Tacsonia, as P. manicata (subgenus Manicata), P. pinnatistipula (a tacso with a shorter tube, less reduced corona and round fruit) and P. macropoda (very similar to the latter, with an even shorter tube and a more complex corona). P. pinnastipula is known to cross spontaneously with typical tacsos, although most of the resulting P. x rosea hybrids are sterile. 189 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Subgenera Astrophea and Tryphostemmatoides are much less divergent from the ‘Passiflora group’ than the ‘Decaloba group’. They are only slightly differentiated by mtDNA, while their restriction profiles only diverge for the cpDNA fragment (trnStrnfM regions), which suggests that their separation was more recent than that of the ‘Decaloba group’. As their chromosome number is 2n = 24, this is consistent with the scheme of genome evolution proposed by De Melo et al. (2001). According to these authors, the 2n = 20 (subgenus Dysosmia) and 2n = 18 (‘Passiflora group’) evolved from the 2n = 12 of the ‘Decaloba group’ via a duplication followed by descendent disploidy (Figure 6). The tree proposed by De Melo et al. (2001) would be more parsimonious if we accept that the divergence of the ‘Decaloba group’ is more ancient than that of Adenia (2n = 12), or, in other words, if we accept that genus Passiflora is not monophyletic. This latter possibility is consistent with the cpDNA studies of Muschner et al. (2003), Hansen et al. (2006) and our cpDNA data, as well as the geographic distributions of these taxa. The question of Passiflora monophyly unavoidably imposes to remember that Decaloba and Astrophea were once generic names and the possibility of reconsidering the relative status of the three major clades. Morphology, biogeography, cytogenetics and possibly differences in inheritance of plastid genomes converge with molecular data in differentiating the ‘Decaloba group/clade’ and, to a lesser degree, clades corresponding to the current subgenera Astrophea and Tryphostemmatoides. The future of lower-level divisions, as sections and series, will obviously depend on the answer to the main question. In a hypothetical simplified genus Passiflora, many problems would be solved spontaneously. For example, Tacsonia or Distephana would very logically recover a subgeneric status, reducing the need for lower taxonomic levels. On the other hand, whatever the levels of the divisions in the classification, some species will remain problematic, such as P. manicata and P. pinnatistipula, which show morphological affinities with both Killip’s subgenera Tacsonia and Passiflora, nuclear DNA affinities with the former (Segura et al., 2002; Yockteng & Nadot, 2004) and mtDNA affinity with the latter. 190 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ Figure 6. Probable relationships among main haploid numbers known in Passiflora subgenera and other Passifloraceae genera, as proposed by De Melo et al. (2001). V.1.6. Conclusions The clear conflicts observed between classifications based on morphological, chloroplast DNA and mitochondrial DNA data can be explained by a combination of reticulate evolution and biparental/paternal transmission of organellar genomes in Passiflora and probably the whole family. This situation appears further complicated by differences in the transmission between the chloroplastic and the mitochondrial genomes, the latter being probably more frequently maternal. There is even a possibility that organellar transmission differs among subgenera. In particular, the situation in subgenus Decaloba sensu Feuillet & MacDougal should be further assessed. This situation imposes particular caution in using molecular data for the resolution of the many problems in Passiflora systematics and evolution. Only major divisions have been confirmed by the molecular studies. On one hand, these tend to confirm the necessary 191 Chapter V. Chloroplast and mitochondrial DNA variation ________________________________________________________________________ simplification of Passiflora taxonomy and the strong reduction in the number of subgenera, as undertaken by Feuillet & MacDougal (2003). On the other hand, the monophyly of the genus is questioned by the data, a problem which must be solved before any significant taxonomic modification. The possibility of restoring genera Decaloba, as established by Roemer in 1846, and Astrophea, by Reichenbach in 1828 (Killip, 1938), should be considered. Interspecific relations appear particularly confusing in the large-fruited species of the ‘Passiflora clade’, which includes all the passion fruits presenting economic importance or breeding potential. In the absence of a strong structure, this entire group can be considered as genetic resources for developing new passion fruit species and cultivars. Breeders should not suspect interspecific barriers a priori. On the contrary, the ‘Decaloba clade’ and subgenus Astrophea should not be attributed the same potential for fruit crop development. V.1.7. Acknowledgements The authors wish to acknowledge Christian Houel (Passiflora National Collection, Blois, France) and Doyle McKey (CNRS, France) for assistance in obtaining plant material. A great part of this research has been funded by the Ministerio del Medio Ambiente from Colombia and Cenicafé through the project CEN-303-2003 “Estudio de la diversidad de las Passifloraceae y Caricaceae en la zona cafetera de Colombia”. The first author is grateful to the Gines-Mera Foundation (CIAT) for financial support. 192 General discussion General discussion _______________________________________________________________________ 1. Discussion 1.1. Biogeography and conservation Colombia has been subject to many studies focused on inventories of plant species groups (Gentry, 1993; Silverstone-Sopkin & Ramos, 1995; Galeano et al., 1998; Rangel, 1995, 2002). Passifloraceae have been inventoried in taxonomical works by Escobar (1998a,b, 1989, 1990 inedited) and Hernández & Bernal (2000). As compared to the latter, we have added new information on geographical distribution of each taxon and extended the list to a total of 167 Passifloraceae species, from three genera and the five biogeographic regions, with reports of 26 species new to Colombia. Our list ranks Colombia as the country with the highest richness of Passifloraceae, followed by Brazil with 127 species. Colombian species richness and diversity is more than twice that of Peru and Venezuela, two countries of similar surface and latitude. Given its much smaller area, Ecuador also presents an impressive diversity. Thus, the northern Andes of Colombia and Ecuador clearly constitute the center of diversity for the genus Passiflora. Escobar (1988a) had already underlined that 40% of the New World Passifloraceae are found in the Andes. Indeed, radiation has been very active in the northern Andes, with particular contribution of recent and fast evolving groups, such as subgenus Tacsonia, particularly adapted to highlands in comparison to other infrageneric groups. Not surprisingly, among the more than 41 tacso species in Colombia and Ecuador, 21 (14%) are endemics. Colombian highlands are also rich in representatives of subgenus Decaloba. On the whole, habitats between 1000 and 3,000 m account for only 27% of the Colombian land area, but 81% of the species of Passifloraceae are found there. With 123 species, the Andean region concentrates the highest richness, mainly between 1000 and 2,000 m. The increase of species richness and endemism with the elevation is generally interpreted as a result of the increasing isolation and decreasing habitat surface in high mountain regions, leading to small, fragmented populations which are prone to speciation (Simpson, 1975; Jørgensen et al., 1995). Another contribution to the particular species richness in Colombia and Ecuador is that of the Pacific Coast region, continuous with the similar highly diverse ecosystems of Central America (Chocó-Darién/Western Ecuador 194 General discussion _______________________________________________________________________ hotspot of Myers et al., 2000), and receiving one of the highest rainfalls in the world, in strong contrast with the conditions prevailing in the westerns Andes and coast of Peru that are arid or semi-arid, or the drier and more contrasted climate of Venezuela. The distribution of Passifloraceae has been drastically affected by deforestation, principally in the Andean region, as their favorite habitats correspond to areas with a long history of livestock and agriculture that now supports extensive plantations of coffee, sugar cane, rice, bananas, and potatoes. Forests in the northern Andes are currently one of the major conservation priorities on a global scale due to their fragility, biological richness, high rates of endemism and multiple anthropogenic threats (Olson & Dinerstein, 1998). As Passifloraceae display very high species richness, endemism and extinction risk in this area, and given their multiple ecological interactions with many organisms, as well as their economic potential, this family should constitute both an important target of conservation efforts and a good indicator of their success. This appears particularly important as we have shown that the conservation of Passifloraceae and their habitat is not taken into account by the current system of protected areas in the Colombian Andes, due to a general lack of correspondence between the modeled distribution of Passifloraceae diversity and these areas, as presented in Figure 8 (Chapter III). Instead, the striking superposition of areas of high Passifloraceae diversity on certain Colombian coffee growing zone ecotopes (Figure 9; Chapter III) points to the need of developing a completely different approach, integrating agriculture, tourism, watershed management, and nature conservancy at the landscape level, in a region whose conservation is of utmost importance for the country. 1.2. Morphological and molecular diversity As a first major point, we can underline that the major morphological divisions observed in our study find support in the genetic studies. The cytological groups are always validated, with the clear separation of subgenera Astrophea (n = 12), Tryphostemmatoides and Decaloba (n = 6) between themselves and from subgenera Passiflora, Tacsonia, and Distephana (n = 9). Concerning subgenus Tryphostemmatoides, the consistency between morphological and genetic studies is clear 195 General discussion _______________________________________________________________________ only when considering our quantitative analysis, where it is associated with subgenus Astrophea mostly on peduncle traits (third principal component). This trait is represented also in the qualitative descriptors, however its effect is blurred by the high number of traits shared with subgenus Decaloba. While the comparison is difficult for subgenus Tryphostemmatoides, it is impossible for subgenus Psilanthus, because of insufficient data and the unlikely placement of P. sanguinolenta in the ‘Passiflora clade’ in the ncpGS study. The two species, P. adenopoda and P. foetida, that take an intermediate position in the general “morpho-cytological” pattern, or their close relatives, are consistently placed in intermediate positions, in most phylogenetic studies, P. adenopoda or P. morifolia (section Pseudodysosmia of subgenus Decaloba) appearing basal to a general ‘Decaloba clade’ and P. foetida (subgenus Dysosmia) basal to the general ‘Passiflora clade’. The comparison becomes more difficult at lower, infra subgeneric, levels. Subgenus Decaloba appears better structured than the other subgenera, and shows similarities in morphological and molecular diversity patterns, with the grouping of Killip’s sections Punctatae and Miserae, and the differentiation of species of sections Xerogona, Cieca, and series Auriculata and, less clearly, Sexflorae. The placement of P. adenopoda in the different trees questions the inclusion of section Pseudodysosmia, while the structure observed among representatives of several sections provides support to some simplification, but not for as many fusions as those operated in the new morphological classification of Feuillet & MacDougal (2003). In any case, more species should be gathered in a same phenetic study before revising objectively the morphological classification. Within the n = 9 group, molecular and morphological data diverge partially, as studies of DNA sequences allow the distinction of a Tacsonia-Manicata group and fail to separate clearly subgenus Distephana, placing both of them within a Passiflora clade, while morphological analysis supports these three subgenera at the same level of differentiation. The fact that species of subgenera Distephana, Tacsonia and Manicata have developed ornithophyly is obviously related to their strong morphological 196 General discussion _______________________________________________________________________ differentiation, which does not minor the importance of their morphological separation from subgenus Passiflora. Whether their probable evolution from a “Passiflora-like” common ancestor justifies their inclusion in the bee-pollinated subgenus Passiflora, as proposed in the new classification, is just the same classical question about considering birds as dinosaurs. In the end, it seems a problem of putting more emphasis on the adaptative forces commanding evolution or more emphasis on the genetic structure that subtend them. Concerning subgenus Passiflora sensu Killip, no structure appears at the interspecific level that could result in clear subdivisions into series. The study of sequence variation for the ncpGS gene provides the only tree with reasonably wellsupported structure at this level, however several obvious abnormalities question the robustness of the information. Our morphological observations only confirm closer associations between the most typical representatives of some series, however the number of contradictions with the classification and the lack of a clear hierarchy in the branch structure point to the difficulty of the work and the risk of under- or overclassification, leading to chose between a limited number of poorly supported series or a great number of poorly represented series. Similarly, the structure of the Tacsonia-Manicata branch does not support clearly sections and series in subgenus Tacsonia, however it allows differentiation between two groups of tacsos, one corresponding to common species that probably have their center of diversity in Ecuador, as is obvious for P. cumbalensis, P. luzmarina and P. matthewsii, and very likely for P. mixta, P. tripartita and P. tarminiana (Segura et al., 2005), and another cluster only including species endemic to Colombia, with a slight but clear differentiation related to extreme variation for peduncle length. 1.3. Importance of reticulate evolution in Passiflora Many Passifloraceae species are cross-compatible (Vanderplank, 2000; Ulmer & MacDougal, 2004). Based on morphological observations, Escobar (1980) stated that also in nature these interspecific hybridizations occasionally occur in areas where species distributions overlap. In the molecular analyses presented here, several discrepancies were observed between chloroplast and mitochondrial DNA phenograms, suggesting both recent and ancient hybridization and introgression events and corroborating the 197 General discussion _______________________________________________________________________ observations of Escobar (1980). These data clearly suggest that evolution in Passiflora is reticulate rather than exclusively dichotomous and branching. Especially the origin of P. foetida provides an excellent example of reticulate evolution and once more illustrates the importance of hybridization in speciation within Passiflora. Such phenomena are well-documented and considered fairly common in the evolution of plants (Van Droogenbroeck et al., 2004; Chat et al., 2004; Vriesendorp & Bakker, 2005). The combination of significant reticulate evolution, with a biparental/paternal inheritance of the choloroplast genome and, although to a lesser extent, the mitochondrial genome, particularly in the ‘Passiflora group’, is the most likely explanation to the observations of wide intraspecific, and even intra-individual, organelle DNA variation, or conversely morphologically distant species sharing very similar organellar genomes, indicating captures of chloroplasts and mitochondria, and the resulting striking inconsistencies between the ITS, matK, ncpGS and TrnL/trnT phylogenies published so far (Muschner et al., 2003; Yockteng, 2003; Yockteng & Nadot, 2004; Hansen et al., 2006; see Annex 4ad). Taking into account these elements, the molecular phenograms presented in the present thesis can serve as a base for further exploration of the genetic diversity and interspecific relationships, phenotypic evolution, historical ecology and phylogeography of the wide Passiflora diversity. At another level, the results presented in this work are also valuable for the development of better in situ and/or ex situ conservation strategies of Passiflora diversity. The extent of natural hybridization occuring in the different areas of sympatry should be considered when dealing with Passiflora genetic resources. Although hybridization has long been recognized as an important factor in the evolution of plant species, the harmful effects of hybridization have also led to the extinction of many populations and species (Allendorf et al., 2001; Barton, 2001). Such effects of hybridization can be most problematic in the situation when a rare Passiflora species comes into contact with a more abundant one. Such cases could get more common as the abundance of several Passiflora species has been reported to decline steadily. Thus, in Colombia, we have determined that three Passifloraceae species can be considered extinct and 70% of the species are threatened 198 General discussion _______________________________________________________________________ (Chapter II). Conservation programs, not only in Colombia, should be aware of such effects in preserving the genetic integrity of Passifloraceae species in the long term. 1.4. Phylogeography As discussed in the Chapter III, South America and especially the biodiversity hotspot described as ‘the Tropical Andes’ have higher plant species diversity than any other habitat on the planet. How this diversity arose is unexplained. One theory suggested that species richness in the tropics is the result of the gradual accumulation of species over a long geological period, with low rates of extinction in stable equatorial climates (Stebbins, 1974). However, more recent discoveries suggest that Neotropical climates were unstable over the past 2 million years during the Pleistocene (Whitmore & Prance, 1987). Cyclical glacial events led to periods of cooler and/or drier climate in which forest species may have withdrawn to small refugial pockets. According to this view, the present species diversity could be more recent, resulting from speciation through allopatric differentiation of populations in separate refugia (Haffer, 1982). Other recent geological phenomena that have been suggested as driving neotropical speciation are the uplift of the northwestern Andes from about 5 million years ago and the bridging of the Isthmus of Panama some 3.5 million years ago (Simpson & Todzia, 1990). Gentry (1989) speculated that nearly half of the Neotropical flora might be accounted for by explosive speciation. This rapid diversification is characteristic of plant evolution on the SouthAmerican continent since its isolation from Africa, Meso-America and AntarcticaAustralia during the late Cretaceous and early Tertiary Period, about 60 to 70 million years ago (Burnham & Graham, 1999; Dino et al., 1999). These geographic connections allowed high exchange of flora, according to fossil registers of many families (Taylor & Taylor, 1993; Taylor, 1995). In the genus Passiflora this exchange is evident with the presence of the subgenus Tetrapathea and the section and series Hollrungiella/Eudocaloba (of subgenus Decaloba) in Oceania and South Asia. Taking into account all these elements, it is most likely that isolation and adaptive radiation into ecologically extreme habitats at the time of the Andean uplift also have led to rapid diversification and differentiation in the genus Passiflora. The relatively wide 199 General discussion _______________________________________________________________________ size and restricted distribution of the genus Passiflora and paleobotanical data from 1720 millions of years (Miocene) provide support to this hypothesis (Dorofeev, 1963). As illustrated by Richardson et al. (2001), molecular evidence can be useful in assessing the validity of phylogeographical hypotheses. The sequence analyses reported by Muschner et al. (2003), Yockteng (2003), Yockteng & Nadot (2004), Hansen et al. (2006) and our results with cp/mtDNA, revealed only low levels of sequence divergence among Passiflora taxa, suggesting that these have diversified recently. In particular all molecular data place the species of the highland subgenus Tacsonia (with a cylindrical elongate floral tube) within the ‘Passiflora group’. However, in the mtDNA phenogram, this subgenus forms a well-supported independent group. This stronger differentiation of mtDNA appears consistent with a faster evolution of the mitochondrial genome, as compared to the chloroplast genome, the particular altitudinal distribution of subgenus Tacsonia and the hypothesis of a rapid diversification and differentiation in the genus Passiflora in relation to the uplift of the Andes. 200 Conclusions & future prospects Conclusions and prospects _______________________________________________________________________ 1. Conclusions With 167 reported species, Colombia is the country with the highest Passifloraceae richness. This richness is concentrated in the Andean region, particularly in the departments of Antioquia, Valle del Cauca and Cundinamarca. Comparisons with other countries indicate that the northern Andes of Colombia and Ecuador constitute the center of diversity for the most important genus Passiflora. Collections of Passifloraceae have not been uniform as a consequence of difficulty of access and/or chronic social conflict in many areas. They have been much denser in the central coffee growing zone, Antioquia, Valle del Cauca and Cundinamarca. The southern and northeastern Andes, and the Caribbean have been little explored. For the lowland forests of the Pacific, the Orinoquian and the Amazonian, data are so poor that they are misleading. Despite the resulting sampling bias, collecting parameters clearly point to the concentration of observed Passifloraceae diversity in the Andes, and more particularly the central coffee growing zone. The analysis of species distribution areas shows a trend for a wider dispersion of species occurring at low and intermediate elevations. On the contrary, narrow endemics are more frequent among highland species. The modeled species richness map allowed identifying nine richness spots of variable size, three of which, located in the southern and southeastern Andes of Colombia, correspond to collection gaps, as they were not detected in the analysis of observed diversity. Another probable collection gap, not detected by diversity modeling, corresponds to the Sierra Nevada de Santa Marta, an isolated mountain range with both high diversity and endemism. The proportion of endemics living in high richness spots is lower than the proportion of all species used for modeling, confirming the lack of relation between diversity concentration and endemism reported in other studies. If this is further substantiated in different groups of organisms, it could limit the application of the 202 Conclusions and prospects _______________________________________________________________________ biodiversity hotspot concept, as the best-protected areas for diversity would not necessarily provide protection to a high proportion of narrow endemics. Passifloraceae diversity is not conserved by the current network of Colombian protected areas. On the contrary, it is particularly concentrated on certain ecotopes of the coffee growing zone, i.e. highly disturbed habitats, so any conservation effort must be integrated in local management strategies at the landscape level. Passifloraceae may provide an interesting indicator to evaluate the outcome of such efforts. In the absence of a clear set of morphological criteria for discriminating at the different hierarchic levels of the infrageneric classification of Passiflora, we have used a quite exhaustive list of 43 quantitative and 83 qualitative descriptors. A shorter list of 32 qualitative traits, selected after analyzing variation among Killip’s subgenera, allowed classifying our 60-species sample consistently, using a strictly phenetic approach. Eight of the nine Killip’s subgenera represented in our sample are supported by the morphological analysis, although subgenus Tryphostemmatoides is only supported in the quantitative analysis. By contrast, the simplification proposed by Feuillet & MacDougal is not clearly supported in our analyses, except for the possible inclusion of P. manicata in subgenus Tacsonia, as this species is intermediate with subgenus Passiflora for quantitative traits but very similar to tacsos for most qualitative traits. Chloroplast and mitochondrial molecular trees provide clear evidence of conflicts with morphological classifications. This suggests that the infrageneric classifications that have been established in the past by Killip (1938), Escobar (1988a,b, 1989) and Feuillet & MacDougal (2003) do not reflect the molecular evolutionary history of the genus Passiflora, at least with respect to the supersections and series. Moreover, the frequent occurrence of reticulation events in the evolution of Passiflora could explain the lack of morphological discontinuities at subgeneric levels, while variable patterns in transmission and evolution of organellar genomes explain striking incongruences between different datasets, reflecting multiple origins of species. 203 Conclusions and prospects _______________________________________________________________________ Breeding programs aimed at producing interspecific hybrids involving the cultivated species of Passiflora should therefore focus on the species belonging to the same clade as subgenus Passiflora. According to our results, the ‘Decaloba clade’ and subgenus Astrophea do not constitute interesting genetic resources for passion fruit breeding. These results constitute potentially crucial inputs for the development of a coherent strategy for the conservation and use of these genetic resources. Studies of Passiflora diversity in the Andean countries, and the maps presented here, will be used in future prospecting and identifying sites for in situ conservation, and more generally guiding government conservation strategies. 2. Future prospects The low level of exploration in parts of the Andes, the Amazonian and the Orinoquian raises expectations that Colombia may still harbor many unknown species. Future studies should encompass new regions, including protected areas and current conflict zones. Indeed, a better knowledge of this diversity, and its distribution, is urgent for the in situ conservation of this threatened richness, targeting the conservation of these resources as well as their habitat. Both aspects may even be combined if the genus Passiflora can be used as an indicator of biodiversity in the Andean region, as was the objective of a project in the coffee growing zone. Another important aspect is its direct valorization as a germplasm resource for crop diversification programs, implying the need for a better understanding of its morphological and genetic diversity. Another direction that should be considered is the sequencing of regions different from those analyzed in this study. By doing so, the value of the phenograms obtained in this work can be assessed and further complemented. A promising source is the nuclear ribosomal internal transcribe spacers (ITS). This region has been widely used in plant phylogenetics because of their high rate of nucleotide substitutions and their power in elucidating infrageneric relations (Taberlet et al., 1991; Alvarez & Wendel, 2003) Also, 204 Conclusions and prospects _______________________________________________________________________ the ITS have already been used in Passiflora species by Muschner et al. (2003) with success, although the size and composition of the sample limited the interpretation at the level of subgenera. It would also be necessary to increase the number of species and in particular to include Asian members of Passiflora as well as species of subgenera not represented in our analysis. Given the importance of hybridization and introgression in the evolution of the genus Passiflora, more studies are also needed to investigate and document hybridization in Passiflora, and its potential implications in the taxonomic problems concerning the genus and the whole family. Detailed studies should focus on the morphologically variable taxa, along with their close relatives in areas where they co-occur. A large number of different methods can provide valuable information regarding hybridization, such as: • • Artificial hybridizations in many species. • artificially produced hybrids. Investigation and characterization of pollen viability and seed germination in Molecular techniques to confirm the hybrid origin of some taxa, combined with cytogenetic techniques, including chromosome counts, flow cytometry, fluorescence in situ hybridization (FISH) and genomic in situ hybridization (GISH). 205 Bibliography Bibliography ________________________________________________________________________ Allendorf F.W., Leary R.F., Spruell P., and Wenburg J.K. (2001) The problems with hybrids: setting conservation guidelines. Trends in Ecology and Evolution 16:613-622. Alvarez I. and Wendel J.F. (2003) Ribosomal ITS sequences and plant phylogenetic inference. Molecular Phylogenetics and Evolution 29(3): 417-434. Apple J. and Feener D. (2001) Ant visitation of extrafloral nectaries of Passiflora: the effects of nectary attributes and ant behavior on patterns in facultative antplant mutualisms. Oecologia 127(3): 409-416. Aublet F. (1775) Histoire des plantes de la Guiane française. Paris : P.F. Didot, 2: 828324. Barbosa L.V. and Vieira A.M. (1997) Meiotic behavior of passion fruit somatic hybrids, Passiflora edulis f. flavicarpa Degener + P. amethystina Mikan. Euphytica 98: 121-127. Barrios L., Caetano C.M., Cardoso C.I., Coppens d´Eeckenbrugge G., and Arroyave J.A., Olaya C. (2005) Caracterización del polen de especies de los géneros Passiflora y Dilkea. Acta Agronómica 54(3): 1-8. Barton N.H. (2001) The role of hybridization in evolution. Molecular Ecology 10: 551 568. Beal P.R. (1972) Two new interspecific hybrids in the genus Passiflora. Sabrao Newsletter 4(2): 113 115. Beal P.R. (1975) Hybridization of Passiflora edulis Sims and P. edulis Sims f. flavicarpa Degener. Queensland Journal of Institute of Agriculture and Animal Sciences 32 (1): 101-111. Benson W.W., Brown K.S., and Gilbert L.E. (1976) Coevolution of plants and herbivores: Passion vine butterflies. Evolution 29: 659-680. Bergner P. (1995) Passionflower. Medical Herbalism 7: 13–14. BRC (Banco de la República de Colombia) (2007) Museo de Oro de Bogotá D.C., Cultura Malagana. http://www.banrep.gov.co/museo/esp/s_calima.htm. Cited 25 January 2007. Brooks T.M., Mittermier R. A., Mittermier C.G., da Fonseca G.A.B., Rylands A.B., Konstant W.R., Flick P., Pilgrim J., Oldfield S., Magin G., and Hilton-Taylor C. (2002) Habitat loss and extinction in hotspot of biodiversity. Conservation Biology 16: 909-923. 207 Bibliography ________________________________________________________________________ Bruckner C.H. and Otoni W.C. (1999) Hibridação em maracujá. In: Borém, A. (Ed.), Hibridação artificial de plantas. UFV, Viçosa, pp. 379–399. Büchert A. and Mogens J. (2001) The fragility of extreme specialization: Passiflora mixta and its pollinating hummingbird Ensifera ensifera. Journal of Tropical Ecology 17: 323-329. Burnham R.J. and Graham A. (1999) The history of neotropical vegetation: New developments and status. Annals of Missouri Botanical Garden 86:546-589. Busby J.R. (1991) BIOCLIM - a bioclimate prediction system. In: Margules CR, Austin MP (eds.) Nature conservation: cost effective biological surveys and data analysis. CSIRO, Melbourne, Australia, pp. 4-68. Campos T. (2001) La Curuba: Su cultivo. Bogotá, Colombia, IICA. 30pp. Calderón E. (2005) Familia Passifloraceae. Programa de biología de la conservación, proyecto flora amenazada. Instituto Alexander von Humboldt. http://www.humboldt.org/ListasRojas/PASSIFLORACAE. Cited 3 abril de 2005. Cancino O. and Hodson E. (1994) Cultivo de tejidos y micropropagación en "maracuyá" Passiflora edulis var. flavicarpa Degener. Tablero, Revista del Convenio Andrés Bello, 18: 81-83. Cavanilles A.J. (1790) Monadelphia classis dissertation decen. Diss. 10: Decima dissertatio botanica de Passiflora. Madrid: Typographia regia. 439-463. Cenicafé (2005) Biodiversidad en la zona cafetera. http://www.cenicafe.org. Cited 25 July 2005. Chat J., Urregui B., Petir R.J., and Nadot S. (2004) Reticulate evolution in kiwifruit (Actinidia, actinidiaceae) identified by comparing their maternal and paternal phylogenies. American Journal of Botany 9(15): 736-747. Chopra R.N., Badhwar R.L., and Ghosh S. (1944) Poisonous Plants of India. Public Service Commission, Govt of West Bengal, Calcutta, pp. 469–472. Chopra R.N., Nayar S.L., and Chopra I.C. (1956) Glossary of Indian Medicinal Plants. CSIR, New Delhi, India, pp. 186–187. Coppens d’Eeckenbrugge G. (2003) Promesas de las pasifloras. Memorias del X Seminario Nacional y IV Internacional sobre Especies Promisorias, Medellín Octubre 29-31 de 2003. CD. 208 Bibliography ________________________________________________________________________ Coppens d`Eeckenbrugge G., Barney V., Møller-Jørgensen P., and MacDougal J.M. (2001) Passiflora tarminiana, a New Cultivated Species of Passiflora subgenus Tacsonia (Passifloraceae). Novon 11: 8-15. Coppens d’Eeckenbrugge G., Segura S.D., Hodson de Jaramillo E. and Gongora G.A. (2001) Passion Fruits. In: Charrier et al. A. (eds.), Tropical Plant Breeding. Montpellier, France, Cirad; Enfield, U.S.A., Science Publishers, pp. 381-401. Coppens d’Eeckenbrugge, G., Restrepo, M.T., Ocampo, J.A., Garcin, F., Olaya, C.A., Farfán, L.C., Vega, J., Rioux, V., Barrera, F., Petersen, J., Guzmán, L., Primot, S., Quispe, P., Caetano, C.M. (2002) Diversidad morfológica en Passiflora subgéneros Tacsonia y Manicata. VIII Congreso Latinoamericano de Botánica & II Congreso Colombiano de Botánica, Cartagena (Colombia), October 13-18, Resúmenes, 417. Corriveau J.L, and Coleman AW. (1988) Rapid screening method to detect potential biparental inheritance of plastid DNA and results for over 200 angiosperm species. American Journal of Botany 75: 1443-1458. Clegg M.T., Gaut B.S., Learn G.H., and Morton B.R. (1994) Rates and patterns of chloroplast DNA evolution. Proceedings of the National Academy of Sciences of the United States of America 91(15): 6795-6801. Croat T. (1978) Flora of Barro Colorado Island. Stanford University Press, Stanford. 943pp. Crochemore M., Molinare H., and Colauto N. (2003) Caracterizao agromorfologica do Maracujazeiro (Passiflora spp.). Revista Brasilera de Fruticultura 25 (1): 510. De Candolle A.P. (1822). Passifloraceae. Prodromus Systematis Naturalis Regni Vegetabilis 3 : 321-338. Paris : Treuttel & WÜrtz. De Candolle A.P. (1828) Mémoires de la Société de Physique et d’Histoire Naturelle de Geneve 1 : 434-436. De Melo F., Cervi A., and Guerra M. (2001) Karyology and cytotaxonomy of the genus Passiflora L. (Passifloraceae). Plant Systematics and Evolution 226: 69-84. De Melo, F. and Guerra M. (2003) Variability of 5S and 45S rDNA sites in Passiflora L. species with distinct base chromosome numbers. Annals of Botany 92: 309316. De Oliviera R., Falvo M., Gomes J., Bernacci L., Carneiro M., Oliveira G., and Martinez O. (2005) Leaf shape analysis using the multiscale Minkonwski fractial 209 Bibliography ________________________________________________________________________ dimension, a new morphometric method: a study with Passiflora (Passifloraceae). Canadian Journal of Botany 83: 287-301. De Wilde W.J.J.O. (1972) The indigenous Old World Passifloras. Blumea 20: 227-250. Debouck D.G. and Libreros Ferla D. (1995) Neotropical montane forests: A fragile home of genetic resources of wild relatives of New World crops. In Biodiversity and Conservation of Neotropical Montane Forests, Proceedings of a Symposium, New York Botanical Garden, 21-26 June 1993 (S.P. Churchill, H. Balslev, E. Forero and J.L. Luteyn, eds.). New York Botanical Garden, New York, pp. 561-577. Deginani N.B. (2001) Las especies argentinas del género Passiflora (Passifloraceae). Darwiniana 39: 43-129. Demesure B., Sodzi N., and Petit R.J. (1995) A set of universal primers for amplification of polymorphic non-coding regions of mitochondrial and chloroplast DNA in plants. Molecular Ecology 4: 129-131. Dévalos L.M. (2001) The San Lucas mountain range in Colombia: how much conservation is owed to the violence?. Biodiversity and Conservation 10: 6978. Dharwan K., Dharwan S., and Sharma A. (2004) Passiflora: a review update. Journal Ethno-Pharmacology 94: 1-23. Dino R., Pocknall D.T., and Dettman M.E. (1999) Morphology and ultrastructure of elater-bearing-pollen from the Albian to Cenomanian of Brazil and Ecuador: implications for botanical affinity. Review of Paleobotany and Palynology 105: 201-235. Do Y.Y., Shii C.T., and Huang P.L. (1992) Restriction patterns and inheritance of chloroplast DNA in Passiflora edulis Sims and its related species. P. 10-13. In the impact of biological research on agricultural productivity (eds.). Taipei, Taiwan. Dornelas M.C., Tavares J.C. de O., and Vieira M.L. (1995) Plant regeneration from protoplast fusion in Passiflora spp. Plant Cell Reports 15: 106-110. Dorofeev P.I. (1963) Primulaceae. In A. L. Takhtajan (ed.), Oznovij Paleontologii, 517– 518. Akademia Nauk, Moscow, Russia. Doyle J.J., and Doyle J.L. (1987) A rapid DNA isolation procedure from small quantities of fresh leaf tissues. Phytochemical Bull 19:11-15. 210 Bibliography ________________________________________________________________________ Drew R.A. (1991) In vitro culture of adult and juvenile bud explants of Passiflora spp. Plant Cell Tissue and Organ Culture 26: 23-28. Duval M.F., Buso G.S.C., Ferreira F.R., Noyer J.L., Coppens d'Eeckenbrugge G., Hamon P., and Ferreira M.E. (2003) Relationships in Ananas and other related genera using chloroplast DNA restriction site variation. Genome 46: 990-1004. Endress P.K. (1994) Diversity and evolutionary biology of tropical flowers. Cambridge, England: Cambridge University Press. Erdtman G. (1952) Pollen morphology and plant taxonomy: Angiosperms (An introduction to Palinology). Leiden: E.J. Brill. (Reprinted in 1986). 553p. Escobar L.K. (1980) Interrelationships of the edible species of Passiflora centering around Passiflora mollissima (H.B.K.) Bailey subgenus Tacsonia. Ph.D Dissertation, University of Texas. Escobar, L.K. (1981) Experimentos preliminares en la hibridación de especies comestibles de Passiflora. Actualidades Biológicas 10: 103-111. Escobar L.K. (1985) Biología reproductiva de Passiflora manicata e hibridación con la curuba Passiflora mollissima (H.B.K.) Bailey. Actualidades. Bioliologicas. 14 (54): 111-121. Escobar L.K. (1986) New species and varieties of Passiflora (Passifloraceae) from the Andes of South America. Systematic Botany 11(1): 88-97. Escobar L.K. (1987) A taxonomic revision of the varieties of Passiflora cumbalensis (Passifloraceae). Systematic Botany 12 (2): 238-250. Escobar, L. K. (1988a) Monografía No. 10. Passifloraceae. Passiflora. Subgéneros: Tacsonia, Rathea, Manicata and Distephana. Bogotá, D.E., Colombia, Universidad Nacional de Colombia. Escobar L.K. (1988b) Passifloraceae. Flora de Colombia 10. Instituto de Ciencias Naturales, Universidad Nacional de Colombia, 138pp. Escobar L.K. (1988c) Novedades en Passiflora (Passifloraceae) de Colombia Mutisia 71:1-8. Escobar L. K. (1989) A new subgenus and five new species in Passiflora (Passifloraceae) from South America. Annals of the Missouri Botanical Garden 76: 877-885. Escobar L.K. (1990) Two new species of Passiflora (Passifloraceae) from northern South America. Phytologia 69 (5): 364-365. 211 Bibliography ________________________________________________________________________ Escobar L.K. (1990, inedited). Una revisión taxonómica de Passiflora subg. Astrophea (Passifloraceae). 289 pp. Escobar L.K. (1992) La sistemática y evolución de las Pasifloras. In: Primer simposio internacional de Pasifloras. Universidad Nacional de Colombia sede Palmira, Colombia, pp. 51-54. Escobar L.K. (1994) Two new species and a key to Passiflora subg. Astrophea. Systematic Botany 19 (2): 203-210. Fajardo D., Angel F., Grum M., Tohmé J., Lobo M., Roca W., and Sánchez I. (1998) Genetic variation analysis of the genus Passiflora L. using RAPD markers. Euphytica 101: 341-347. Faure S., Noyer J., Carreel F., Horry J.P., Bakry F., and Lanaud C. (1994) Maternal inheritance of chloroplast genome and paternal inheritance of mitochondrial genome in banana (Musa acuminata). Current Genetics 25: 265-269. Ferguson M.E., Jarvis F., Stalker H.T., Williams D.E., Guarino L., Valls J.M., Pittman R.N., Simpson C.E., and Bramel P.J. (2005) Biogeography of wild Arachis (Leguminosae): distribution and environmental characterization. Biodiversity and Conservation 14: 1777- 1798. Feuillet C. (2002) A new series and three new species of Passiflora subgenus Astrophea from Guianas. Brittonia 54: 18-29. Feuillet C. (2004) Passiflora phellos, a new species in subgenus Passiflora (Passifloraceae). Novon 14: 285-287. Feuillet C. and MacDougal J.M. (2002) Checklist of recognized species names of passion flowers. Passiflora 12 (2): 41-43. Feuillet C. and MacDougal J.M. (2003) A new infrageneric classification of Passiflora L. (Passifloraceae). Passiflora 13 (2): 34-38. Fouqué A. and Fouqué R. (1980) Quelques notes sur la granadille jaune. Passiflora edulis Sims var. flavicarpa. Fruits 35: 309-312. Franklin J. (1995) Predictive vegetation mapping: geographic modeling of biospatial patterns in relation to environmental gradients. Progress in Physical Geography 19: 474-499. Galeano G., Suarez S., and Balslev H. (1998) Vascular plant species count in a wet forest in the Chocó area on the Pacific coast of Colombia. Biodiversity and Conservation 7: 1563-1572. 212 Bibliography ________________________________________________________________________ García M.T.A., Galati B.G., and Anton A.M. (2002) Microsporogenesis, microgametogenesis and pollen morphology of Passiflora spp. (Passifloraceae). Botanical Journal of the Linnean Society 139:383-394. García-Barriga H. (1975) Flora medicinal de Colombia. Tomo II, Instituto de Ciencias Naturales, Universidad Nacional – Bogotá, D.E. Colombia. 253-264. Gaston K.J. (1996) Species richness: measure and measurement. In: Gaston K.L. (ed.) Biodiversity: A biology of numbers and difference. Blackwell Science, Oxford, UK, pp. 77-113. Gentry A.H. (1976) Additional Panamanian Passifloraceae. Annals of the Missouri Botanical Garden 63: 341-345. Gentry A.H. (1982) Neotropical floristic diversity: phytogeographical connections between Central and South America, Pleistocene climatic fluctuations, or an accident of the Andean orogeny?. Annals of Missouri Botanical Garden 69:557-593. Gentry A.H. (1986) Endemism in tropical versus temperate plant communities. In Soulé M.E. (ed), Conservation biology. The science of scarcity and diversity. Sinauer Associates, Sunderland, Massachussets, pp. 153-181. Gentry A.H. (1986) Species richness and floristic composition of Choco´ Region plant communities. Caldasia 15: 71–75. Gentry A.H. (1989) Northwest South America (Colombia, Ecuador and Peru). In Campbell D.G., and Hammond H.D. (eds), Floristic inventory of tropical countries. New York Botanical Garden, Bronx. Pp. 391-400. Gentry A.H. (1992) New species of woody plants from Amazonian Peru. Novon 2(4): 333-338. Gentry A.H. (1993) “Riqueza de especies y composición florística.” In: P. LEYVA (Ed.). Colombia Pacífico, Vol. 1. Fondo Protección del Medio Ambiente José Celestino Mutis, Publicaciones Financiera Eléctrica Nacional (FEN), Bogotá. Gerlach G. and Ulmer T. (2000) Una nueva especie de Passiflora, subgénero Passiflora (Passifloraceae) del oeste de Colombia. Caldasia 22 (2): 231-235. Gilbert L.E. (1982) The evolution of a butterfly and a vine, Heliconius butterflies. Scientific American 247: 110-121. Green P.S. (1972) Passiflora in Australasia and the Pacific. Kew Bulletin 26(3): 539-558. 213 Bibliography ________________________________________________________________________ Gremillion K.J. (1989) The development of a mutualistic relationship between humans and maypops (P. incarnata L.) in the Southeastern United States. Journal of Ethnobiology 9: 135–158. Haffer J. (1982) General Aspects of the Refuge Theory In: Prance GT (Ed) Biological diversification in the Tropics. Columbian University Press, New York, USA, pp 6-24. Hall R. M., Drew R.A., Higgins C.M., Dietzgen R.G. (2000) Efficient organogenesis of an Australian passion fruit hybrid (Passiflora edulis x Passiflora edulis var. flavicarpa) suitable for gene delivery. Austin Journal Botany 48:673–680. Handler N. (1962) Psychomimetic Medicine, The First Hahnemann Symposium. Lea and Febiger, Philadelphia. Hansen A.K., Lawrence G., Simpson B.B., Downie S.R., Stephen R., Cervi A.C., and Jansen R.K. (2006) Phylogenetic relationships and chromosome number evolution in Passiflora. Systematic Botany 31(1): 138-150. Hansen A.K., Escobar L.K., Lawrence E., Gilbert E., and Jansen K. (2007) Paternal, maternal and biparental inheritance of chloroplast genome in Passiflora (Passifloraceae): American Journal of Botany 94 (1): 42-46. Harms H. (1898) Zür morphologie der ranken und blütenstände bei den Passifloraceae. Botanische Jahrbücher für Systematik 24: 163-178. Harms H. (1925) Passifloraceae. In Die naturlichen Pflanzenfamilien 21, Germany: W. Engelmann, 470-507. Harris S.A. and Ingram R. (1991) Chloroplast DNA and biosystematics: the effects of intraspecific diversity and plastid transmission. Taxon 40:393–412. Havey M.J., McCreight J.D., Rhodes B., and Taurick G. (1998) Differential transmission of the Cucumis organellar genomes. Theoretical and Applied Genetics 97: 122-128. Hernández J., Hurtado A., Ortiz R., and Walschburger T. (1991) Unidades Biogeográficas de Colombia En: Hernández J., Ortiz R., Walshburger T., y Hurtado A. (Eds.) Estado de la Biodiversidad en Colombia Informe Final Santafé de Bogotá, Instituto Colombiano para el Desarrollo de la Ciencia y la Tecnología “Francisco José de Caldas” - Colciencias. Hernández A. (2003) Revisión taxonómica de Passiflora, subgénero Decaloba (Passifloraceae) en Colombia. Tesis, Facultad de Ciencias, Departamento de Biología, Universidad Nacional de Colombia sede Bogotá. 138pp. 214 Bibliography ________________________________________________________________________ Hernández A. and Bernal R. (2000) Lista de especies de Passifloraceae de Colombia. Biota Colombiana 1 (3): 320- 335. Hernández A.. and García N. (2006) Las pasifloras (familia Passifloraceae). En: Libro rojo de plantas de Colombia. Las bromelias, las labiadas y las pasifloras. Garcia N., and Galeano G. (eds.). volumen 3: 553-663. Hickey M. and King C. (1988) 100 Families of Flowering Plants. Cambridge University Press, Cambridge, pp. 130–133. Hijmans R.J., Garrett K.A., Huamán Z., Zhang D.P., Schreuder M., and Bonierbale M. (2000) Assessing the geographic representativeness of genebank collections: the case of Bolivian wild potatoes. Conservation Biology 14(6): 1755-1765. Hijmans R.J., Schreuder M., De la Cruz M., and Rojas E. (2001) Computer tools for spatial analysis of plant genetic resources data: DIVA-GIS. Plant Genetic Resources Newsletter 27: 15-19. Hijmans R.J., Guarino L., Bussink C., Mathur P., Cruz M., Barrentes I., and Rojas E. (2005) DIVA-GIS. A geographic information system for the analysis of species distribution data. Manual, version 5.2. Hijmans R.J. and Graham C.H. (2006) The ability of climate envelope models to predict the effect of climate change on species distributions. Global Change Biology 12(12): 2272-2281. Hijmans R.J. and Spooner D.M. (2001) Geography of wild potato species. American Journal of Botany 88: 2101-2112. Hill T. and Lewicki P. (2006) STATISTICS Methods and Applications. StatSoft, Tulsa, UK. Hno. Daniel (J. González Patiño) (1968) Curiosidades de una flor y de una familia botánica (La Flor de la Pasión). Boletín del Instituto de la Salle, Bogotá. 208: 261-270. Ho W.F., and Shii C.T. (1986) Incompatibility system in passion fruit (Passiflora edulis Sims). Acta Horticulturae 194: 31–38. Holm-Nielsen L.B. (1974) Notes on central Andean Passifloraceae. Botaniska Notiser 127:338-351. Holm-Nielsen L.B., Jørgensen P.M., and Lawesson J.E. (1988) Passifloraceae. In: Harling & L. Andersson (eds.), Flora of Ecuador 31: 124pp. 215 Bibliography ________________________________________________________________________ Howell C.W. (1976) Edible fruited Passiflora adapted to South Florida growing conditions. Proceeding Florida State Horticultural Society 89: 236-238. Hutchinson J. (1967) Passifloraceae. The general of florewing Plants. Vol. 2. Oxford: Clarendon Press. 364-374. IGAC (Instituto Geográfico Agustín Codazzi) (2006) Mapas departamentales. http://www.igac.gov.co. Cited 18 November 2005. IUCN (World Institute for Conservation and Environment) (2003) Guidelines for application of IUCN red list criteria at regional levels. Version 3.0. IUCN Species Survival Commission. IUCN, Gland, Switzerland and Cambridge, UK. IUCN (World Institute for Conservation and Environment) (2004) Guidelines for using the IUCN Red List categories and criteria. Standards and petitions subcommittee of the IUCN SSC Red List programme committee. IUCN, Gland, Switzerland and Cambridge, U.K. Jamir T.T., Sharma H.K., and Dolui A.K. (1999) Folklore medicinal plants of Nagaland, India. Fitoterapia 70, 395–401. Jarvis A., Guarino L., Williams D., Williams K., & Hyman G. (2002) Spatial analysis of wild peanut distributions and the implications for plant genetic resource conservation. Plant Genetic Resources Newsletter 131: 29-35. Jarvis A., Ferguson M.E., Williams D.E., Guarino L., Jones P.G., Stalker H.T., Valls J.F.M., Pittman R.N., Simpson C.E., and Bramel P. (2003) Biogeography of wild Arachis: assessing conservation status and setting future priorities. Crop Science. 43(3): 1100-1108. Jarvis A., Williams D., Guarino L., Caballero P.J. and Mottram G. (2005) Use of GIS for optimizing a collecting mission for a rare wild pepper (Capsicum flexuosum Sendtn.) in Paraguay. Genetic Resources and Crop Evolution 52 (6): 671-682. Jones P.G., Beebe S.E., Tohme J., and Galwey N.W. (1997) The use of geographical information systems in biodiversity exploration and conservation. Biodiversity and Conservation 6: 947-958. Jones, P., Guarino L., and Jarvis A. (2002) Computer tools for spatial analysis of plant genetic resources data: 2. FloraMap. Plant Genetic Resources Newsletter 130:1–6. Jørgensen P.M., Lawesson J.E.., and Holm-Nielsen L.B. (1984) A guide to collecting passion flowers. Annals of the Missouri Botanical Garden 71(4): 1172-1174. 216 Bibliography ________________________________________________________________________ Jørgensen P.M., Ulloa C., Madsen J.E., and Valencia R. (1995) A floristic analysis of the high Andes of Ecuador. Churchill S.P., Balslev H., Forero E., and Luteyn J.L.(eds.). Biodiversity and Conservation of Neotropical Montane Forests, Proceedings of the Neotropical Montane Forest Biodiversity and Conservation Symposium, The New York Botanical Garden, Bronx, NY: 221-237. Jørgensen P.M., and León-Yánez S. (1999) Results. In: P.M. Jørgensen and S. LeónYánez (eds.): Catalogue of the vascular plants of Ecuador. Monograph Systematic Botonay, Missouri Botanical Garden 75: 42-108. Judd W.S., Campbell C.S., Kellogg, E.A., Stevens P.F., and Donoghue M.J. (2002) Plant systematics: a phylogenetic approach. Sinauer Associates, Inc., Sunderland, Massachusetts, USA. Junqueira N.T.V., Bragal M.F., Faleiro F.G., Peixoto J.R., e Bernacci L.C. (2005) Potencial de espécies silvestres de maracujazeiro como fonte de resistência a doenças. 81-108p. En: Maracujá: germoplasma melhoramento genético. (Eds.) Fábio Gelape Faleiro, Nilton Tadeu Vilela Junqueira, Marcelo Fideles Braga. Planaltina, DF : Embrapa Cerrados. 670 p.l. Jussieu A.L. de. (1805) Second Mémoire sur la famille des Passiflorées, et particulièrement sur quelques espèces nouvelles du genre Tacsonia. Annales du Muséum d’Histoire Naturelle 6: 388- 396. Kay E. (2001) Observations on the pollination of Passiflora penduliflora. Biotropica 33:709–713. Kessler M. (2002) The elevational gradient of Andean plant endemism: varying influences of taxon-specific traits and topography at different taxonomic levels. Journal of Biogeography 29: 1159-1165. Killip E.P., Meggers W.F., and Hewett D.F. (1924) New species of Passiflora from Tropical America. Journal of the Washington Academy of Sciences 14 (5): 108-112. Killip E.P. (1930) Ten new species of Passiflora, mainly from Colombia and Peru. Journal of the Washington Academy of Sciences 20 (15): 374-381. Killip E.P. (1938) The American Species of Passifloraceae. Field Museum of Natural History Publication. Botanical Series 19: 613pp. Killip E.P. (1960) Supplemental notes to the American species of Passiflora with descriptions of new species. Contributions from the U.S. National Herbarium 35: 2 (Tomo I). 217 Bibliography ________________________________________________________________________ Kingston N. and Waldren S. (2005) A conservation appraisal of the rare and endemic vascular plants of Pitcairn Island. Biodiversity and Conservation 14: 781-800. Kirtikar K.R. and Basu, B.D. (1975) Indian Medicinal Plants. Periodical Experts, Dehradun, India, p1103. Knight R.J. and Winters H.F. (1962) Pollination and fruit set of yellow passion fruit in Southern Florida. Proceeding of the Florida State Horticultural Society 75: 412-418. Knight R.J. (1991) Development of tetraploid hybrid passion fruit clones with potential for the north temperate zone. HortScience 26: 1541-1543. Knight R.J. (1992) Characters needed for commercially successful passion fruit. Proceeding Florida State Horticultural Society 105: 280-282. Krosnick S.E. and Freudestein J.V. (2005) Monophyly and floral character homology of Old Word Passiflora (Subgeneus Decaloba: supersection Disemma). Systematic Botany 20(1): 139-152. Kugler E.E. and King L.A. (2004) A brief history of the Passionflower. 15-26 pp. In Ulmer T., and MacDougal J.M. (eds). Passiflora: passionflowers of the word. Timber Press Portland, Oregon. 430 pp. Kyndt T., Romeijn-Peeters E., Van Droogenbroeck B., Romero-Motochi J.P., Gheysen G., and Goetghebeur P. (2005) Species relationships in the genus Vasconcellea (Caricaceae) based on molecular and morphological evidence. American Journal of Botany 92: 1033–1044. Lakshmi M., Senthilkumar P., Parani M., Jithesh M.N., and Parida A. (2000) PCR-RFLP analysis of chloroplast gene regions in Cajanus (Leguminosae) and allied genera. Euphytica 116: 243–250. Lee D.J., Blake T.K., and Smith S.E. (1988) Biparental inheritance of chloroplast DNA and the existence of heteroplasmic cells in Alfalfa. Theoretical and Applied Genetics 76:545–549. Lehmann A., Leathwick J.R., and Overton J.M. (2002) Assessing New Zealand fern diversity from spatial predictions of species assemblages. Biodiversity and Conservation 11: 2217-2238. Leimberck R., Valencia R., and Balslev H. (2004) Landscape diversity patterns and endemism of Araceae in Ecuador. Biodiversity and Conservation 13 (9): 1755-1779. 218 Bibliography ________________________________________________________________________ Lemarck J.B. Monenet de. (1789) Encyclopédie Méthodique, Botanique 3(1). Paris: Panckoucke. Linneaus C. (1753) Species Plantarum. Vol. 2. Stockholm: L. Salvius. MacDougal J.M. (1992) New species of Passiflora subg. Plectostemma (Passifloaceae). Novon 2(4): 358-367. MacDougal J.M. (1994) Revision of Passiflora section Decaloba, Pseudodysosmia (Passifloraceae). Systematic Botany Monographs 14: 146pp. MacDougal J.M. (2006) Passiflora sandrae (Passifloraceae), a new species from Panama. Novon 16: 85-88. MacMillan R.T. and Graves W.R. (1992) Susceptibility of Passiflora spp. to Alternaria passiflorae. In: Memorias Primer Simposio Internacional de Passifloras, Universidad Nacional de Colombia, Palmira, 123-124. MacNeely J.A., Miller K.R.., Reid N.A., Mittemer R.A., and Wainer T.B.. (1990) Conserving the world’s biological diversity World Conservation Union, World Resources Institute, World Wildlife Fund - U.S. World Bank, Washigton, D.C. Manders G., Otoni W.C., d'Utra-Vaz F.B., Blackhall N.W., Power J.B., and Davey M.R. (1994) Transformation of passionfruit (Passiflora edulis f. flavicarpa Degener) using Agrobacterium tumefaciens. Plant Cell Reports 13: 697-702. Martin F.W. and Nakasone H.Y. (1970) The edible species of Passiflora. Economic Botany 24 (3): 333-343. Martin P.S. and Szuter C.R. (1999) War zones and game sinks in Lewis and Clark’s West. Conservation Biology 13: 36-45. Master M.T. (1872) Passifloraceae. In Flora Brasiliensis 13 (1). Ed. K. Mrtius and A. Eichler. 529-628. Master M.T. (1877) A classified synonymic list of all the species of Passiflorae cultivated in Europe gardens, with references to the works in which they are figured. Journal of the Royal Horticultural Society 4: 125-149. Maxted N., van Slageren M.W., and Rihan, J.R. (1995) Ecogeographic surveys. In ‘Collecting plant genetic diversity. Technical guidelines’. (eds L. Guarino, R.V. Ramantha & R.Reid). Chapter 14. CAB International. Maxted N., Mabuza-Dlamini P., Moss H., Padulosi S., Jarvis A., and Guarino L. (2004) African Vigna : An Ecogeographic Study, IPGRI, Rome, Italy. 219 Bibliography ________________________________________________________________________ May P.G. and Spears E.E. (1988) Andromonoecy and variation I in phenotypic gender of Passiflora incarnata (Passifloraceae). American Journal of Botany 75: 18301841. May R.M. (1992) "How many species inhabit the Earth?". Scientific American 267(4): 18-24. McNeely J.A., Miller K.R., Reid N.A., Mittemer R.A., and Wainer T.B. (1990) Conserving the world’s biological diversity World Conservation Union, World Resources Institute, World Wildlife Fund – U.S. World Bank, Washigton, D.C. Medina, C. I., Lobo, M., and Correa, R. 2000. Caracterización morfológica y química de Pasifloras Andinas como apoyo al desarrollo de estas especies. III Seminario de Frutas de Clima Frío Moderado. Manizales, Colombia. 13-18. Meletti L.M.M., Santos R.R., and Minami K. (2000) Melhoramento do maracujazeiroamarelo: obtenção do Composto IAC-27. Scientia Agricola, Piracicaba, v. 56, n. 3, p. 17-24. Meletti L.M.M., Soares-Scott D., Bernacci L.C., e Passos I.R.da S. (2005) Melhoramento genético do maracujá: passado e futuro. 55-78pp. En: Maracujá: germoplasma melhoramento genético. (Eds.) Fábio Gelape Faleiro, Nilton Tadeu Vilela Junqueira, Marcelo Fideles Braga. Planaltina, DF : Embrapa Cerrados. 670 p.l. Mes T. H. M., Friesen N., Fritsch R. M., Klass M., and Konrad B. (1997) Criteria for sampling in Allium based on chloroplast DNA PCR-RFLP''s. Systematic Botany 22: 701–712. Metzlaff M., Borner T., and Hagemann R. (1981) Variations of chloroplast DNAs in the genus Pelargonium and their biparental inheritance. Theoretical Applied Genetics 60: 37-41. Midgley G.F., Hannah L., Millar D., Rutherford M.C., and Powrie L.W. (2002) Assessing the vulnerability of species richness to anthropogenic climate change in a biodiversity hotspot. Global Ecology Biogeography 11:6 445. Moon E., Kao T-H., and Wu R. (1987) Rice chloroplast DNA molecules are heterogeneous as revealed by DNA sequences of a cluster of genes. Nucleic Acids Research 15:611–630. Morton J.F. (1981) Atlas of Medicinal Plants of Middle America. Spring- field, IL, p. 1281. 220 Bibliography ________________________________________________________________________ Mowrey D. (1993) Herbal Tonic Therapies. Keats Publishing Incorporation, New Canaan, CT. Mráček J. (2005) Investigation of interspecific genome-plastome incompatibility in Oenothera and Passiflora. Ph.D Dissertation. Munich University, Germany. 110pp. Muschner V., Lorenz-Lemke A., Cervi A.C., Bonatto S., Souza-Chies T., Salzano F., and Freitas L. (2003) A first molecular phylogenetic analysis of Passiflora (Passifloraceae). American Journal of Botany 90: 1229-1238. Muschner V., Lorenz-Lemke A., Vecchia M., Bonatto S., Salzano F., and Freitas L. (2006) Differential organellar inheritance in Passiflora’s (Passifloraceae) subgenera. Genetica 128 (1-3): 449-453. Myers J.E. (2000) The biodiversity challenge: Expanded hot spots analysis. The Environmentalist 10: 243–256. Myers N., Mittermeier R.A., Mittermeier C.G., da Fonseca G.A.B., and Kent J. (2000) Biodiversity hotspots for conservation priorities. Nature 403: 853-858. Nakasone H.Y., Hirono R., and Ito P. (1967) Preliminary observations on the inheritance of several factors in the passionfruit (Passiflora edulis and forma flavicarpa). Tech. Prog. Rpt. 161. U. Hawaii, Agr. Exper. Sta., Honolulu. Nascimento W.M.O., Tomé A.T., Oliviera M.do S.P.de., Muller C H., e Caravalho J.E.U. (2003) Seleção de progênies de maracujazeiro-amarelo (Passiflora edulis f. flavicarpa) quanto à qualidade de frutos. Revista Brasileira Fruticultura 25(1): 186-188. Negreiros J.R. da Silva., Bruckner C.H., Cruz C.D. De Siqueira D.L., and Pimentel L.D. (2004) Selection of yellow passion fruit vines for vigor and resistance to scab (Cladosporium cladosporioides). Revista Brasilera de Fruticultura 26(2): 272275. Nishida T. (1958) Pollination of the passion fruit in Hawaii. Journal of Economic Entobotanic 51: 146–149. Nwosu M.O. (1999) Herbs for mental disorders. Fitoterapia 70: 58–63. Ocampo J., Coppens d’Eeckenbrugge G., Olano C., and Schnell R. (2004) AFLP analysis for the study of genetic relationships among cultivated Passiflora species of the subgenera Passiflora and Tacsonia. Proceedings of the Interamerican Society for Tropical Horticulture 47: 72-76. 221 Bibliography ________________________________________________________________________ Ocampo J., Coppens d’Eeckenbrugge G., Restrepo M., Salazar M., Jarvis A., and Caetano C. (2007) Diversity of Colombian Passifloraceae: biogeography an updated list for conservation. Biota Colombiana (in press). Oga S., de Freitas P.C.D., Gomes da Silva A.C., and Hanada S. (1984) Pharmacological trials of crude extracts of Passiflora alata. Planta Medica 51: 303–306. Oliveira J.C. de., Ferreira F.R.., Ruggiero C., Nakamura K. (1987) Caracterização e avaliação de germoplasma de Passiflora edulis. In: Congresso Brasileiro Fruticultura, 9. Campinas. Anais. Campinas: SBF, v.2, p. 591-596. Oliveira J.C. and Ferreira F.R. (1991) Melhoramento genético do maracujazeiro. In: "A cultura do maracuja no Brasil", São José, A.R., Ferreira, F.R. and Vaz, R.L. (eds.), Jaboticabal, FUNEP, pp. 187-200. Olson D.M. and Dinerstein E. (1998) “The Global 200: A representation approach to conserving the earth’s most biologically valuable ecoregions.” Conservation Biology 12: 502–515. Otoni V.C., Blackhall N.W., D'Utra Vaz F.B., Casali V.W., Power J.B., and Davey M.R. (1995) Somatic hybridization of the Passiflora species, P. edulis var. flavicarpa Degener and P. incarnata L. Journal of Experimental Botany 46: 777-785. Parani M., Lakshmi M., Ziegenhagen B., Fladung M., Senthilkumar P., and Parida A. (2000) Molecular Phylogeny of Mangroves VII. PCR-RFLP of trnS-psbC and rbcL gene regions in 24 mangrove and mangrove associate species. Theoretical Applied and Genetics 100: 454-460. Parques Naturales de Colombia (2006) http://www.parquesnacionalesdecolombia.gov.co. Cited 10 Sept 2006. Passionfruit (2006) Supply and demand. http://www.passionfruitjuice.com. Cited 3 October 2006. Payán F.R. and Martín F.W. (1975) Barriers to the hybridization of Passiflora species. Euphytica 24: 709-716. Pearson D.L. (1994) Selecting indicator taxa for the quantitative assessment of biodiversity. Philosophical Transactions of the Royal Society of London 345: 75–79. Pérez E. (1956) Plantas útiles de Colombia. Tercera redacción muy corregida y aumentada, con XLV laminas en negro, otras en color y 752 figuras en el texto. Librería Colombiana - Camacho Roldán (Cia, Ltda.) - Bogotá. 611614. 222 Bibliography ________________________________________________________________________ Perrier X., Flori A., and Bonnot, F. (2003) Data analysis methods. In: Hamon, P., Seguin, M., Perrier, X., Glaszmann, J. C. Ed., Genetic diversity of cultivated tropical plants. Enfield, Science Publishers. Montpellier. pp 43 - 76. Peterson A.T. (2001) Predicting species geographic distribution based on ecological niche modeling. Condor 10: 599-605. Peterson A.T. (2004) Predictability of the geography of species invasions via ecological niche modeling. Quarterly Review of Biology 78: 419-433. Peterson A.T. and Robins C.R. (2003) When endangered meets invasive: ecological niche modeling predicts double trouble for spotted owls, Strix occidentalis. Conservation Biology 17: 1161-1165. Porter-Utley K. (2003) Revision of Passiflora subgenus Decaloba supersection Cieca (Passifloraceae). Ph.D dissertation, University of Florida. 444p. Presting D. (1969) Zür morphology der pollenkorner der passifloraceen. Pollen et Spores (2): 193-247. Primot S., Coppens d’Eeckenbrugge G., Rioux V., Ocampo J., and Garcin F. (2005) Variación morfológica de tres especies de curubas (Passiflora tripartita var. mollissima, P. tarminiana y P. mixta) y sus híbridos en el Valle del Cauca (Colombia). Revista Brasilera de Fruticultura 27(3): 467-471. Radhamani T.R., Sudarshana L., and Krishnan R. (1995) Defence and carnivory: dual roles of bracts in Passiflora foetida L. Journal of Biosciences 20: 657-664. Rangel J.O. (1995) La diversidad florística en el espacio andino de Colombia. En S.P. Churchill et al., eds. Churchill, S. P., H. Balslev, E. Forero & J. L. Luteyn (eds.). 1995. Biodiversity and Conservation of Neotropical Montane Forests, Proceedings of the Neotropical Montane Forest Biodiversity and Conservation Symposium, The New York Botanical Garden, Bronx, NY: 187-205. Rangel J.O. (2002) El estado actual del conocimiento de la flora de Colombia. Pág. 570 en: Rangel J.O., J. Aguirre-C & M.G. Andrade-C. (eds), Libro de resúmenes octavo congreso latinoamericano y segundo Colombiano de botánica Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogotá. Raven P.H. (1975) The bases of angiosperm phylogeny: cytology. Annals of Missouri Botanical Garden 62:734-764. Rawat P.S. (1987) Select Your Dose and Potency. B. Jain Publishers (P) Ltd., New Delhi, pp. 481–482. 223 Bibliography ________________________________________________________________________ Rebelo A.G. (1994) Iterative selection procedures: Centers of endemism and optimal placement of reserves. Strelitzia 1:231–257. Rêgo M.M, Bruckner C.H., Finger F.L., Siqueira D.L., and Fernandes A.A. (1998) Selfincompatibility: evidence for two-locus genetic control in yellow passion fruit. Theoretical and Applied Genetics 98:564–568. Rêgo M.M., Rêgo E.R., Bruckner C.H., Da Silva E.A.M., Finger F.L., and Pereira K.J.C. (2000) Pollen tube behavior in yellow passion fruit following compatible and incompatible crosses. Theoretical and Applied Genetics 101:685–689. Reichenbach H.G.I. (1828) Passifloraceae. Conspectus regni vegetabilis. Leipzig, Germany: Carl Cnobloch. 132. Richardson J.E., Pennington R.T., Pennington T.D., and Hollingsworth P.M. (2001) Rapid diversification of a species-rich genus of neotropical rain forest trees. Science 293: 2242-2245. Robbirt K.M., Roberts D.L., and Hawkins J.A. (2006) Comparing IUCN and probabilistic assessments of threat: do IUCN red list criteria conflate rarity and threat?. Biodiversity and Conservation 15: 1903-1912. Rodríguez D., Marín C., Quecán H., and Ortiz R. (2005) Áreas potenciales para colectas del género Vasconcellea Badillo en Venezuela. Bioagro 17 (1): 3-10. Roemer M.J. (1846) Passifloracea. Familiarum naturalium regni vegetabilis synopses monographicae 2. Weimar, Germany. 15-207. Romero-C.R. (1956) Plantas de valor comercial del género Passiflora: Granadilla, curuba, badea, y otras. Agricultura Tropical 12 (6): 403-407. Romero-C.R. (1991) Frutas silvestres de Colombia. Segunda edición actualizada, Volumen con 207 ilustraciones. Instituto Colombiano de Cultura Hispánica. 423-429 p. Royal Horticultural Society. (2002) Royal Horticultural Society Colour Chart. London. Ruberté-Torres R. and Martín F.W. (1974) First generation hybrids of edible passion fruit species. Euphytica 23: 61-70. Ruggiero C., Banzatto DA., and Lam-Sanchez A. (1976) Studies on natural and controlled pollination in yellow passion fruit (Passiflora edulis f. flavicarpa Deg.). Acta Horticulturae 57:121–124. 224 Bibliography ________________________________________________________________________ Saitou N. and Nei M. (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4(4):406425. Sánchez I., Fajardo A., Grum M., Duque M., Lobo M., Tohme J., and Roca W. (1999) Variability of chloroplast DNA in the genus Passiflora L. Euphytica 106 : 1526. Sazima M. and Sazima I. (1978) Bat pollination of the passion flower, Passiflora mucronata, in southeastern Brazil. Biotropica 10: 100-109. Scheldeman X., Willemen L., Coppens d'Eeckenbrugge G., Romeijn-Peeters E., Restrepo M.T., Romero Motoche J., Jimenez D., Lobo M., Medina C.I., Reyes C., Rodriguez D., Ocampo J.A., Van Damme P., and Goetgebeur P. (2006) Distribution, diversity and crop ecology of highland papayas (Vasconcellea spp.) and papaya (Carica papaya L.) in Latin America. Biodiversity and Conservation. DOI 10.1007/s10531-006-9086-x. Schöniger G. (1986) La curuba: técnicas para el mejoramiento de su cultivo. Bogotá: Ed. Guadalupe. 256p. Seaforth C.E., Adams C.D., and Sylvester Y. (1983) A Guide for the Medicinal Plants of Trinidad & Tobago. Commonwealth Secretariat, Marlborough House, Pall Mall, London. Sechrest W., Brooks T.M., da Fonseca G.A.B., Konstant W.R., Mittermeier R.A., Purvis A., Rylands A.B., and Gittleman J.L. (2002) Hotspots and the conservation of evolutionary history. Proceeding National Academy Science 99(4): 20672071. Segura S., Coppens d'Eeckenbrugge G., Ocampo C., and Ollitrault P. (2002) Isozyme variation in Passiflora subgenera Tacsonia and Manicata. Relationships between cultivated and wild species. Genetic Resources and Crop Evolution 50: 417-427. Segura S., Coppens d’Eeckenbrugge G., Bohórquez A., Ollitrault P., and Tohmé J. (2002) An AFLP study of the genus Passiflora focusing on subgenus Tacsonia. Genetic Resources and Crop Evolution 49: 111-123. Segura S., Coppens d’Eeckenbrugge G., López L., Grum M., and Guarino L. (2003) Mapping the potential distribution of five species of Passiflora in Andean countries. Genetic Resources and Crop Evolution 50: 555-566. Segura S., Coppens d'Eeckenbrugge G., Ocampo C-H., and Ollitrault P. (2005) Isozyme variation in Passiflora subgenus Tacsonia: geographic and interspecific differentiation among the three most common species. Genetic Resources and Crop Evolution 52: 455-463. 225 Bibliography ________________________________________________________________________ Silva M.B. (1998) Transformação genética de maracujá-amarelo (Passiflora edulis f. flavicarpa Degener) mediada por Agrobacterium tumefaciens. M.Sc. thesis, Universidade Federal de Vicosa, Vicosa, Brazil. Silverstone-Sopkin P.A. and Ramos-Pérez J.E. (1995) Floristic exploration and phytogeography of the cerro del Torrá, Chocó, Colombia. Biodiversity of Conservation of Neotropical Montane Forests: 169-86. In Biodiversity and Conservation of Neotropical Montane Forests, Proceedings of a Symposium, New York Botanical Garden, 21-26 June 1993 (S.P. Churchill, H. Balslev, E. Forero and J.L. Luteyn, eds.). New York Botanical Garden, New York. Simpson B.B. (1975) Pleistocene changes in the flora of high tropical Andes. Paleobiology 1:273-294. Simpson B.B. and Todzia C.A. (1990) Patterns and processes in the development of the high Andean flora. American Journal of Botany 77:1419-1432. Skov F. (2000) Potential plant distribution mapping based on climatic similarity. Taxon 49: 503-515. Snow A. (1982) Pollination intensity and potential seed set in Passiflora vitifolia. Oecologia 55: 231-237. Snow N., and MacDougal J.M. (1993) New chromosome reports in Passiflora (Passifloraceae). Systematic Botany 18 (2): 261-273. Sokal R.R. and Michener C.D. (1958) A statistical method for evaluating systematic relationships. U. Kansas Sci. Bull .38:1409-1438. Stebbins G.L. (1974) Evolution above the species level. Harvard University Press, Cambridge, USA, 399 p. Storey W.B. (1950) Chromosome numbers of some species of Passiflora occurring in Hawaii. Pacific Science 4: 37-42. Suassuna T.M.F., Bruckner C.H., Carvalho C., and Borém A. (2003) Selfincompatibility in passion fruit: evidence of gametophytic– sporophytic control. Theoretical and Applied Genetics, 106: 298–302. Svenning J.C. (1998) The effect of land-use on the local distribution of palm species in an Andean rain forest fragment in northwestern Ecuador. Biodiversity and Conservation 7: 1529–1537. 226 Bibliography ________________________________________________________________________ Taberlet, P., Gielly L., Patou G., and Bouvet J. (1991) Universal primers for amplification of three noncoding regions of chloroplast DNA. Plant Molecular Biology 17: 1105.1109. Taylor T. N., & Taylor E.L. (1993) The Biology and Evolution of Fossil Plants. Prentice Hall, NJ, USA. 982pp. Taylor D.W. (1995) Cretaceous to tertiary geologic and angiosperm paleobiogeographic history of the Andes. 3-22pp. In Biodiversity and Conservation of Neotropical Montane Forests, Proceedings of a Symposium, New York Botanical Garden, 21-26 June 1993 (S.P. Churchill, H. Balslev, E. Forero and J.L. Luteyn, eds.). New York Botanical Garden, New York, pp. 561-577. Tillet S. (2003) Passifloraceae. In: Flora of the Venezuelan Guayana Volume 7: Myrtaceae-Plumbaginaceae by Julian A. Steyermark, Paul E. Berry, Kay Yatskievych, and Bruce K. Holtz (eds.): 625-667. Ulmer T. (1999) Passiflora formosa sp. nov., a hitherto misunderstood taxon in Passiflora subgenus Tacsonia (Passifloraceae) from Colombia. Edinburgh Journal of Botany 56 (2): 195-198. Ulmer T. and MacDougal J.M. (2004) Passiflora: passionflowers of the word. Timber Press Portland, Oregon. 430 pp. UNEP-WCMC. (2004) World Conservation Monitoring Centre of the United Nations Environment Programme. Species Data (unpublished, September 2004). Web site at: http://www.unep-wcmc.org. Cambridge, England: UNEP-WCMC. Uribe L. (1954) Dos nuevas Passifloraceae colombianas. Mutisia 21: 1-5. Uribe L. (1955a) Pasifloráceas y Begoniáceas de la Real Expedición Botánica del Nuevo Reino de Granada. Ediciones Cultura Hispánica. Madrid 26:1-98. Uribe L. (1955b) Sertula Florae Colombiana 2. Caldasia 7(32):159-165. Uribe L. (1957) Una nueva e interesante Passiflora de Colombia. Caldasia 7(35):335338. Uribe L. (1958) Sertula Florae Colombiae 4. Caldasia 8 (37):127-130. Uribe L. (1972) Catalogo ilustrado de las plantas de Cundinamarca: Passifloraceae, Begoniaceae, Melastomataceae. Instituto de Ciencias Naturales, Facultad de Ciencias, Universidad Nacional. (5): 11-41. Uribe L. (1977) Sertula Florae Colombiae 14. Caldasia 12 (56): 13-18. 227 Bibliography ________________________________________________________________________ Van Droogenbroeck B., Kyndt T., Maertens I., Romeijn-Peeters E., Scheldeman X., Romero-Motochi J., Van Damme P., Goetghebeur P., and Gheysen G. (2004) Phylogenetic analysis of the highland papayas (Vasconcellea) and allied genera (Caricaceae) using PCR-RFLP. Theoretical Applied and Genetics 108: 1473-1486. Vanderplank J. (2000) Passion Flowers. 3nd ed. The MIT Press. Cambridge, Massachusetts. 224 pp. Vanderplank J. (2006) Passiflora miniata: Passifloraceae. Curtis’s Botanical Magazine 23 (3): 223-230. Varassin I.G., Trigo J.R., and Sazima M. (2001) The role of nectar production, flower pigments and odour in the pollination of four species of Passiflora (Passifloraceae) in south-eastern Brazil. Botanical Journal of the Linnean Society 136: 139–152. Vargas J.D. (2000) Relaciones filogenéticas de especies del genero Passiflora (Passifloraceae) con énfasis en el subgénero Tacsonia a partir de RFLP y PCR-RFLP. Tesis (Biólogo), Facultad de Ciencias, Pontificia Universidad Javeriana, Santa fe de Bogota - Colombia. 168pp. Vargas J.H., Consiglio T., Jørgensen P.M., and Croat T.B. (2004) Modeling distribution patterns in a species-rich plant genus, Anthurium (Araceae), in Ecuador. Diversity and Distribution 10(3): 211-216. Varón M.H. (2000) Determinación de polimorfismo en regiones amplificadas de cpDNA, mtDNA y rDNA de Passiflora utilizando enzimas de restricción Tesis (Bióloga), Facultad de Ciencias, Pontificia Universidad Javeriana, Santa fe de Bogota - Colombia. 150pp. Vasconcellos M. (1991) Biología floral do maracujá doce Passiflora alata Dryan nas condiçöes de Botucatu-SP. Tesis Faculdade de Ciencias Agrarias. Universidade Estatudual Paulista. Botucatu. 98p. Vekemans X., Hardy O., Berken B., Fofana B., and Baudoin J-P. (1998) Use of PCRRFLP on chloroplast DNA to investigate phylogenetic relationships in the genus Phaseolus. Biotechnology, Agronomy, Society and Environment 2 (2): 128–134. Whitmore T.C., and Prance G.T. (1987) Biogeography and Quaternary history in Tropical America. Oxford Science Publications, Oxford, UK, 214 p. Villacís L., Grum. M., and Coppens d'Eeckenbrugge G. (1998) Morphological characterization of Andean Passifloras (Passiflora spp.) from Ecuador. Plant Genetic Resources Newsletter 115: 51-55. 228 Bibliography ________________________________________________________________________ Villareal H., Álvarez M., Córdoba S., Escobar F., Fagua G., Gast F., Mendoza H., Ospina M., y Umaña A.M. (2006) Manual de métodos para el desarrollo de inventarios de biodiversidad. Programa de Inventarios de Biodiversidad. Instituto de Investigación de Recursos Biológicos Alexander von Humboldt. Bogotá, Colombia. 2d. edition. 236 pp. Vina A.P. (2001) Correlações e parâmetros genéticos e populações de maracujazeiroamarelo (Passiflora edulis f. flavicarpa) e diversidade molecular no gênero Passiflora. 98p. Tese (Doutorado em Produção Vegetal) - Universidade Estadual do Norte Fluminense, Campos dos Goytacazes. Winks C.W., Menzel C.M., and Simpson D.R. (1988) Passionfruit in Queensland. 2. Botany and cultivars. Queensland Agriculture Journal 114: 217-224. World Press Review. (1993) Colombia's vanishing forests, World Press Review, June 1993, Vol. 40 (6): 43pp. Vriesendorp B., & Bakker F.T. (2005). Reconstructing patterns of reticulate evolution in Angiosperms: what can we do? Taxon 54(3): 593-604. Yockteng R. (2003) Phylogénies: outils pour l’étude de l’histoire évolutive des organismes. L’exemple des pensées (genre Viola L.) et des passiflores (genre Passiflora L.). Université Paris XI – Orsay. Docteur en Sciences. 142pp. Yockteng R. and Nadot S. (2004) Phylogenetic relationships among Passiflora species based on the glutamine synthetase nuclear gene expressed in chloroplast (ncpGS). Molecular Phylogenetics and Evolution 31: 379-396. Zaharieva M., Prosperi J. M., and Monneveux P. (2004) Ecological distribution and species diversity of Aegilops L. genus in Bulgaria. Biodiversity and Conservation 13: 2319-2337. 229 Congress communications Congress communications ________________________________________________________________________ Ocampo Pérez John, Coppens d’Eeckenbrugge Geo, Salazar Mike, Jarvis Andy. (2007). La familia Passifloraceae en Colombia: diversidad y conservación. IV Congreso Colombiano de Botánica. Medellín, abril 2007. Resumen aprobado. Salazar Mike, Ocampo Pérez John, Restrepo María, Coppens d’Eeckenbrugge Geo, Caetano Creuci, Jarvis Andrew, Villegas Mauricio. (2006). First results from diversity mapping of Passiflora (Passifloraceae) and Vasconcellea (Caricaceae) in the Colombian coffee growing zone. First DIVERSITAS Open Science Conference: Integrating biodiversity science for human well-being. Oaxaca, Mexico. November 9-12, 2006. Abstracts (CD). Ocampo Pérez John, Coppens d’Eeckenbrugge Geo, Risterucci Ange-Marie. (2006). Relaciones genéticas en el género Passiflora (Passifloraceae) usando marcadores CAPS sobre el ADN citoplasmático. IX Congreso Latinoamericano de Botánica. Republica Dominicana, Santo Domingo, junio de 2006, p 325. Ocampo Pérez John, Salazar Mike, Coppens d’Eeckenbrugge Geo, Restrepo María, Jarvis Andrew. (2006). Distribución y riqueza de especies de la familia Passifloraceae en Colombia. IX Congreso Latinoamericano de Botánica. Republica Dominicana, Santo Domingo, junio de 2006. p280. Ocampo Pérez John. (2005). Estudios biogeográficos, morfológicos y moleculares de la Passifloraceae en Colombia. Seminario Internacional sobre agrobiodiversidad y producciún de frutales andinos de clima frío y frío moderado. Pasto, Colombia Universidad de Nariño, noviembre 2005. Resúmenes (CD). Ocampo Pérez John, Olano Cecile, Coppens d’Eeckenbrugge Geo, Schnell Raymond. (2004). AFLP analysis for the study genetic relationships among cultivated Passiflora species of the subgenera Passiflora and Tacsonia. 50th Meeting of the Interamerican Society for Tropical Horticulture, Limón, Costa Rica, October 2004. Program and abstract, 44. 231 Congress communications ________________________________________________________________________ Primot Sophie, Coppens d’Eeckenbrugge Geo, Rioux Vincent, Ocampo Pérez John, Garcin Francois. (2003). Variación morfológica de tres especies de curubas (Passiflora tripartita var. mollissima, P. tarminiana y P. mixta) y sus híbridos en el Valle del Cauca (Colombia). Sexto Simpósio Brasileiro sobre a Cultura do Maracujazeiro. November 24 27, Campos de Goytacazes (Brazil). Resumo 96 (Compact Disc). Ocampo Pérez John, Restrepo María, Coppens d’Eeckenbrugge Geo, Caetano Creuci, Villegas Andrés, Salazar Mike. (2004). Diversidad de Passifloraceae en la zona cafetera de Colombia. III Congreso Colombiano de Botánica. Popayán, noviembre 2004. p251. 232 ANNEXES Annexes ______________________________________________________________ Annex 1. Infrageneric classification according to Killip (1938) with emends of Escobar (1988, 1989) and MacDougal (1994). Genus Passiflora L. 507 species 1234- Subgenus Adenosepala Killip Subgenus Apodogyne Killip Subgenus Astephia Killip Subgenus Astrophea (DC.) Masters Section Botryastrophea (Harms) Killip Section Cirrhipes Killip Section Dolichostemma Killip Section Euastrophea (Harms) Killip Section Leptopoda Killip Section Pseudoastrophea (Harms) Killip 5 - Subgenus Calopathanthus (Harms) Killip 6 - Subgenus Chloropathanthus (Harms) Killip 7 - Subgenus Deidamioides (Harms) Killip 8 - Subgenus Distephana (Juss.) Killip 9 - Subgenus Dysosmia (DC.) Killip 10 - Subgenus Dysosmioides Killip 11 - Subgenus Granadilla (Medic.) Masters Series Quadrangulares Killip Series Digitatae Killip Series Tiliafoliae Killip Series Marginatae Killip Series Laurifoliae Killip Series Serratifoliae Killip Series Setaceae Killip Series Pedatae Killip Series Incarnatae Killip Series Palmatisectae Killip Series Kermesinae Killip Series Imbricatae Killip Series Simplicifoliae Killip Series Lobatae Killip Series Menispermifoliae Killip 12 - Subgenus Manicata (Harms) Escobar 13 - Subgenus Murucuja (Medic.) Masters 14 - Subgenus Plectostemma Masters Section Cieca (Medic.) Mast. Section Mayapathanthus Killip Section Decaloba (DC.) Mast. Series Auriculatae Killip Series Heterophyllae Killip Series Sexflorae Killip Series Apetalae Killip Series Luteae Killip Series Organenses Killip Series Miserae Killip Series Punctatae Killip Series Eudecaloba** Section Hahniopathanthus (Harms) Killip Section Hollrungiella** Section Mayapathanthus (Harms) Killip Section Pseudodysosmia (Harms) Killip 234 01 01 01 46 11 01 02 15 01 16 01 01 02 11 11 04 97 02 01 10 01 13 04 01 01 06 01 08 02 14 29 04 05 04 154 25 01 91 02 02 04 02 03 07 05 47 19 03 01 01 18 Annexes ______________________________________________________________ Section Pseudogranadilla (Harms) Killip Section Xerogona (Raf.) Killip 15 - Subgenus Polyanthea (DC.) Killip 16 - Subgenus Porphyropathantus Escobar 17 - Subgenus Pseudomurucuja (Harms) Killip 18 - Subgenus Psilanthus, Killip 19 - Subgenus Rathea (Karsten) Killip 20 - Subgenus Tacsonia (Juss.) Triana & Planchon Section Ampullacea Escobar Section Boliviana Escobar Section Bracteogama DC Prodr. Section Colombiana Escobar Series Leptomischae Escobar Series Colombianae Escobar Series Quindiensae Escobar Section Fimbriatistipula Escobar Section Parritana Escobar Section Poggendorffia Triana & Planchon Section Tacsonia Escobar Section Tacsoniopsis Triana & Planchon Section Trifoliata Escobar 21 - Subgenus Tacsonioides (DC.) Killip 22 - Subgenus Tacsoniopsis (Tr. & Planch) Killip 23 - Subgenus Tetrapathea* 24 - Subgenus Tryphostemmatoides (Harms) Killip (Harms, 1925; de Candolle, 1822; Hutchinson, 1967; Green, 1972)* and (de Wilde, 1972)**. 235 06 08 01 01 04 04 02 48 01 01 12 19 08 09 02 02 02 01 05 03 02 05 01 01 03 Annexes ______________________________________________________________ Annex 2. Infrageneric classification according to Feuillet & MacDougal (2003). Genus Passiflora L. 520 species 1- Subgenus Astrophea (DC.) Masters Supersection Astrophea Section Astrophea Section Capreolata MacDougal & Feuillet Section Leptopoda Killip ex Feuillet & Cremers Supersection Pseudoastrophea (Harms) Feuillet & MacDougal Section Pseudoastrophea (Harms) Killip Section Botryastrophea (Harms) Killip Series Botryastrophea (Harms) MacDougal & Feuillet Series Carnae Feuillet 2- Subgenus Deidamioides (Harms) Killip Section Polyanthea DC. Section Deidamioides (Harms) Feuillet & MacDougal Section Tetrastylis (Barb. Rodr.) Harms Section Mayapathanthus MacDougal & Feuillet Section Tryphostemmatoides Harms 3- Subgenus Decaloba (DC.) Rchb. Supersection Pterosperma Gilbert & MacDougal Supersection Hahniopathanthus (Harms) MacDougal & Feuillet Supersection Disemma (Labill.) MacDougal & Feuillet Section Octandranthus Harms Section Disemma (Labill.) MacDougal & Feuillet Section Hollrungiella Harms Supersection Multiflora (Small) MacDougal & Feuillet Supersection Auriculata MacDougal & Feuillet Supersection Cieca (Medic.) MacDougal & Feuillet Supersection Bryonioides (Harms) MacDougal & Feuillet Supersection Decaloba (DC.) MacDougal & Feuillet Section Decaloba DC. Section Xerogona (Raf.) Killip 4- Subgenus Passiflora Supersection Passiflora Series Passiflora Series Palmatisectae Feuillet & MacDougal Series Pedatae Killip ex Cervi Series Setaceae Killip ex Cervi Supersection Stipulata Feuillet & MacDougal Section Granadillastrum Triana & Planch. Section Calopathanthus Harms Section Tacsonioides DC. Section Kermesinae (Cervi) Feuillet & MacDougal Section Dysosmia DC. Supersection Laurifolia (Cervi) Feuillet & MacDougal Series Laurifoliae Killip ex Cervi Series Quadrangulares Feuillet & MacDougal Series Tiliifolia Feuillet & MacDougal Series Marginatae Killip ex Cervi Supersection Coccinea Feuillet & MacDougal Supersection Distephana (DC.) Feuillet & MacDougal Supersection Tacsonia (Juss.) Feuillet & MacDougal Section Rathea (Karst.) Harms Section Insignes Feuillet & MacDougal 236 57 27 10 15 02 30 17 13 06 07 13 01 01 02 02 07 214 04 05 21 17 03 01 09 08 18 20 119 106 13 236 19 13 01 01 04 95 66 01 04 04 20 42 21 06 14 01 14 05 61 03 05 Annexes ______________________________________________________________ Section Colombiana Escobar Series Colombianae Escobar Series Leptomischae Escobar Series Quidiensae Escobar Section Parritana Escobar Section Fimbriatistipula Escobar Section Tacsoniopsis Triana & Planch. Section Elkea Feuillet & MacDougal Section Tacsonia (Juss.) Harms Section Boliviana (Harms) Feuillet & MacDougal Section Trifoliata Feuillet & MacDougal Section Manicata (Harms) Feuillet & MacDougal 237 19 09 08 02 02 02 02 15 05 02 01 05 Annexes ______________________________________________________________ Annex 3. Species cultivated in Colombia. A.3.1. Passiflora edulis Sims Plant essentially glabrous throughout (except ovary); trilobate leaves (5-25 x 5-20 cm.); sepals white inside and green outside. This species presents two botanical form flavicarpa Degener and edulis, originally differentiated by the color (yellow and purple) and size of the fruit (8-12 x 5-10 cm and 5-8 x 4-6 cm respectively). The yellow maracuja, P. edulis f. flavicarpa is very probably native of Brazil and is the most important passion fruit in the hot tropical areas, under the names of passion fruit, yellow granadilla, yellow maracuja, maracuyá, parchita maracuyá. The yellow maracuja requires high temperatures, between 20 and 34 ˚C, and develops better at lower altitudes. The principal producer is Brazil with 450.000 t.y-1. Colombia reports 5.000 ha located from 0 to 1.200 m, with a production of 17 to 20 t.ha-1.y-1, mainly in the departments of Huila and Valle del Cauca. Between 60 and 70% of the Colombian production is processed into frozen juice, exported to the European market and the remainder is consumed in natura. The purple maracuja, P. edulis f. edulis is native of southern Brazil to northern Argentina and Paraguay. It is now cultivated in most tropical areas with a mild climate (subtropics and tropical highlands). In Colombia, it grows up to 2.500 m. This form is known under the names of purple maracuja, gulupa, curuba redonda, maracuyá rojo, palchita. Colombia, Australia, USA, Kenya, and Zimbabwe are the principal producers with approximately 20.000 t, although the total area is unknown. In Colombia, about 100 ha area cultivated between 1.700 and 2.500 m, with reported yields between 10 and 12 t.ha-1 .y-1. According to Vanderplank (2000), the presently cultivated material originated from several fruits, found in a London market, whose seeds were sent to Argentina. In 1915, the progeny from these seeds passed through the USDA (United States Department of Agriculture) into the United States, which redistributed them to Australia and New Zealand. Yellow maracuja breeding has most often involved direct mass selection of this material with a narrow genetic base and its hybridization with genotypes of the purple form obtained from materials exported to Australia and Hawaii more than one hundred years ago. 238 Annexes ______________________________________________________________ The purple and yellow forms have frequently been crossed and spontaneous hybrids are found in Hawaii and Australia. The direction of crossing determines its success: P. edulis f. flavicarpa should be used as the male (Beal, 1975). The F1 hybrids are intermediate, normal and vigorous (Nakasone et al., 1967). Meiosis is normal with the formation of nine bivalents, but the chiasma frequency is lower in the hybrid than in the parents, which suggests slightly reduced chromosomal homology. In the F2, 6 to 8% of the plants were found to be abnormal (Beal, 1975). These data, as well as differences in the phenology and acclimatization of the two forms, indicate the beginning of differentiation, tempered by the fact that equally important divergences may be observed within the purple form in Brazil (Oliveira et al., 1987). A.3.2. Passiflora ligularis Juss. Vernacular names: sweet granadilla, granadilla. Origin: Andes of Colombia to Peru. Distribution: Mexico to Peru. Plant glabrous with broad heart-shaped leaves (rarely trilobate), flowers white or pinkish white, fruit round to ovoid, 5 to 9 cm long, 4 to 7 cm in diameter, tapering towards the peduncle, with a thin, hard and brittle pericarp, light brown to orange, sometimes tinged of violet, with small light spots or stripes. Its light grey pulp is aromatic, slightly tangy and relished by consumers in raw form. This species is usually cultivated at altitudes of 1.400 to 2.200 m. near the Equator, with extremes of 800 and 3.000 m. It grows at average temperatures of 14 to 22oC and a relative humidity of 70%. It can tolerate short and very light frosts. It is said to be tolerant to parasites and diseases of root and collar, but is susceptible to withering due to Nectria haematococca (anamorphic Fusarium solani) in poorly drained soils. A plantation is productive for four to eight years. The sweet granadilla commences flowering in the ninth month and production starts 75 to 80 days later. Colombia is the principal producer with 2.661 ha (2004) at a density of 400 plants per hectare and yields reach 45 to 50 t ha-1 y-1. Currently, Colombia is exporting this fruit to Europe. There are no commercial varieties of the sweet granadilla. Although some superior types may be propagated by cuttings or grafting, multiplication is usually by seed and the plant allogamy maintains considerable variability in plantations. 239 Annexes ______________________________________________________________ A.3.3. Passiflora tripartita var. mollissima (Kunth) Holm-Nielsen & Jørgensen Vernacular names: Banana passion fruit, curuba de Castilla, tacso, tumbo. Origin: Andes of Venezuela to Bolivia (2.000 - 3.500 m). Distribution: Venezuela to Argentina (mostly cultivated), New Zealand. Plant pubescent, leaves deeply trilobate, 5-10 x 6-12 cm; flowers pendant, corolla variable in size, with pink to deep red sepals and petals, 2.5 to 4 cm long. The crown is reduced to a whorl of white tubercles on a red background. The floral tube is 6 to 11 cm long and 1.5 cm wide; fruit oblong, 6-5 cm x 3-5 cm. with rounded ends, weighing 50 to 150 g (average 80 g). The pericarp is pale yellow, occasionally green, somewhat pubescent, thin and supple but coriaceous. It is rich in pectin and can be processed if in a perfect state, without anthracnose spots. The pulp, constituting 60% of the fruit, is salmon-pink to dark orange, slightly acidic, very pleasantly scented but generally astringent. The plant grows at altitudes between 2.000 and 3.000 m, sometimes up to 3.600 m, at average temperatures of 12 to 15oC and relative humidity of 70 to 80%. It cannot adapt to warmer climates yet it is sensitive to extended periods of frost. Production begins at about 18 months. Commercial cultivation began in the 1950s. In Colombia, 1.795 ha are cultivated, with yield ranging from 7 t.ha-1.y-1 to 45 t, the latter under the most suitable conditions. The major phytosanitary problems are anthracnose, which spoils the fruit, and nematodes of the genus Meloidogyne. Being an allogamous species, its repeated self-fertilization results in a serious loss of vigor (Schöniger, 1986). A.3.4. Passiflora tarminiana Coppens & Barney Vernacular names: curuba india, tacso, tumbo Distribution: Venezuela to Argentina, as a cultigen, naturalized in the USA (Hawaii; invasive species), New Zealand, Philippines, Ceylon, Papua New Guinea, tropical highlands of East Africa. Plant glabrous; leaves trilobate, 7-9 x 12-18 cm; flowers pendent, petals bright pink to light pink; fruit fusiform 10-14 x 3.5-4.5 cm, pericap yellow to orange with dots except along the main vascular bundles. P. tarminiana is adapted to a wide range of elevations as compared to other species of the subgenus Tacsonia and have been introduced into numerous cool tropical and tropical montane areas other than its original Andean range, in some cases becoming an invasive weed. The hybrids with P. mixta and P. tripartita var. mollissima are fertile and show intermediate phenotypes (Primot et al., 2005). This species is tolerant to Alternaria and 240 Annexes ______________________________________________________________ Colletrotrichum. In Colombia, it is cultivated between 2.000 and 2.700 m with yields around 20-25 t.ha-1.y-1, although the cultivated area is not known, as it is frequently confused with P. tripartita var. mollissiima. a b c d e Figure 1. Species cultivated in Colombia: (a) P. edulis f. flavicarpa; (b) P. edulis f. edulis; (c) P. ligularis; (d) P. tripartita var. mollissima; (e) P. tarminiana. 241 Annexes ______________________________________________________________ A.3.5. Passiflora quadrangularis L. Vernacular names: giant granadilla, badea, barbadine. Distribution: West Indies to South America Plant glabrous, stem stout, quadrangular, reaching 50 m in length; leaves simple, oval or oblanceolate, 10 to 25 cm long and 8 to 18 cm broad; the flowers may reach 12 cm in diameter: the inner face of the sepals and petals is white, pink, red or violet; the crown may reach 6 cm in length; the fruit is yellowish-green, sometimes with a pinkish tint, ovoid to oblong. It is 20 to 30 cm long, 10 to 18 cm broad, and weighs 2.8 kg on average, even reaching 4 kg. It develops in 62 to 85 days. The fruit mesocarp is 2 to 3 cm thick, soft and edible but bland. The pulp is transparent, white to orange, sweet and slightly tangy, the taste variable but always less pungent than that of maracuja. Manual pollination is often recommended. This species is tolerant to Alternaria passiflorae (McMillan & Graves 1992) and resistant to withering but highly susceptible to nematodes and Xanthomonas spp. (Oliveira & Ferreira, 1991; Vanderplank 2000). In Colombia 60 ha are cultivated between 0 to 1.000 m, with a yield of 16 to 18 t. ha-1 y-1. A.3.6. Passiflora maliformis L. Vernacular names: stone granadilla, conch apple, granadilla de piedra, chulupa, coque en fer. Origin: West Indies to Ecuador. Distribution: West Indies to Ecuador, North Brazil. Plants glabrous or finely pilosulous with leaves ovate, ovate-lanceolate, or sometimes orbicular ovate; flowers petals densely mottled with dark red-purple; fruits globose, 4 to 6 cm in diameter, green or orange green and rarely purple, pericarp hard to extremely hard. The juice has excellent flavor, is very refreshing and valuable due its high content of ascorbic acid. This plant occurs naturally at mid-elevations (0 to 1.600 m), but in Colombia it is cultivated at 600 to 1.000 m on 97 ha, with yields of 11 to 14 t ha-1 y-1, mainly in the Huila department where there is a local market. A.3.7. Passiflora alata Curtis Vernacular names: Fragrant granadilla, maracuja doce, Maracua. Origin: Brazilian and Peruvian Amazon, Brazilian Planalto in forest galleries. Distribution: Brazil, Peru and Colombia. 242 Annexes ______________________________________________________________ Plant glabrous with stem stout and quadrangular; leaves ovate or ovate-oblong 8-15 x 7-10 cm; flowers petals red inside and green outside; fruit ovoid or pyriform, 8 to 10 cm. long, yellow. In many regions of South America, P. alata is cultivated because of its edible fruit. Brazil is the principal producer with 50 t.ha-1. y-1 and an area of 120 ha. In Colombia, it has been cultivated traditionally in the Amazon and commercially in the North of the Valle del Cauca and Quindío departments from 1999, but its yield statistics are unknown. This species has given a number of hybrids, mainly with the closely related P. quadrangularis. A.3.8. Passiflora popenovii Killip Vernacular names: Granadilla de Quijos, granadilla caucana, curubejo. Distribution: South of Colombia and Ecuador. Plant glabrous, except the ovary and the outer surface of the calyx tube; leaves unlobed, oblong-ovate, 8.5 to16 cm long; fragrant flowers with white petals; fruit ovoid, 6 to 10 cm in diameter, epicarp thin and yellow. The sweet pulp and juice are highly regarded and is enjoyed for its rich aroma and taste. This species is cultivated in Colombia (Cauca and Nariño) and Ecuador (Tungurahua, Napo, Loja) between 1.400 and 2.100 m. and its fruits are offered in local markets in the months of April and May. This plant is regularly propagated by cutting. In Colombia, yields of 140 kg/plant are reported, but the cultivated area is not well known. 243 Annexes ______________________________________________________________ a b c d Figure 2. Species cultivated in Colombia: (a) P. quadrangularis; (b) P. alata; (c) P. maliformis; (d) P. popenovii. 244 Annexes ______________________________________________________________ Annex 4. Passiflora molecular diversity. Dendrograms obtained in previous studies. A.a. Phylogenetic tree sensu Muschner et al. (2003). Maximum-likelihood tree for the ITS spacer in Passiflora and outgroups. Numbers above branches are bootstrap support values (when higher than 50%) based on 1000 replicates. Abbreviations indicate the subgenera as follows: PASS 5 Passiflora; DIST 5 Distephana; TACS 5 Tacsonioides; DYSO 5 Dysosmioides; DYSA 5 Dysosmia; ASTR 5 Astrophea; DECA 5 Decaloba; MURU 5 Murucuja; DEID 5 Deidamioides; PSMU 5 Pseudomurucuja; ADOP 5 Adopogyne. 245 Annexes ______________________________________________________________ A.b. Phylogenetic tree sensu Yockteng (2003). Strict consensus tree of 3,536 most parsimonious trees based on matk) sequences. Bootstrap support values (BS) greater than the alignment of chloroplast-expressed mutarase K (m 70% and decay indices greater than 4 are indicated above branches. Thick lines indicate BS values of 100%, dots indicate nodes supported by 100% in the Bayesian tree. The subdivisions according to Killip (1938) are indicated for each taxon. The sections and series names are abbreviated as follows: Cie, Cieca; Dec, Decaloba; Auri, Auriculatae; Sexf, Sexflorae; Orga, Organenses; Mise, Miserae; Punc, Punctatae; Eud, Eudecaloba; Xero, Xerogona; Psedy, Pseudodysosmia; Psegr, Pseudogranadilla; Hah, Hahniopathanthus; Lept, Leptomischae; Col, Colombianae; Bract, Bracteogama; Tac, Tacsonia; Quad, Quadrangulares; Digi, Digitatae; Tili, Tiliafoliae; Laur, Laurifoliae; Inca, Incarnatae; Simp, Simplicifoliae; Loba, Lobatae; Meni, Menispermifoliae; Doli, Dolichostemma; Euas, Euastrophea; Pseas, Pseudoastrophea. The three clades emerging from the analysis are indicated on the right. 246 Annexes ______________________________________________________________ A.c. Phylogenetic tree sensu Yockteng & Nadot (2004). Strict consensus tree of 32,573 most parsimonious trees ncpGS) sequences. Bootstrap support based on the alignment of chloroplast-expressed glutamine synthetase (n values (BS) greater than 70% and decay indices greater than 4 are indicated above branches. Thick lines indicate BS values of 100%, dots indicate nodes supported by 100% in the Bayesian tree. The subdivisions according to Killip (1938) are indicated for each taxon. The sections and series names are abbreviated as follows: Cie, Cieca; Dec, Decaloba; Auri, Auriculatae; Sexf, Sexflorae; Orga, Organenses; Mise, Miserae; Punc, Punctatae; Eud, Eudecaloba; Xero, Xerogona; Psedy, Pseudodysosmia; Psegr, Pseudogranadilla; Hah, Hahniopathanthus; Lept, Leptomischae; Col, Colombianae; Bract, Bracteogama; Tac, Tacsonia; Quad, Quadrangulares; Digi, Digitatae; Tili, Tiliafoliae; Laur, Laurifoliae; Inca, Incarnatae; Simp, Simplicifoliae; Loba, Lobatae; Meni, Menispermifoliae; Doli, Dolichostemma; Euas, Euastrophea; Pseas, Pseudoastrophea. The three clades emerging from the analysis are indicated on the right. 247 Annexes ______________________________________________________________ A.d. Phylogenetic tree sensu Hansen et al. (2006). Phylogram of 1 of 200,000 trees from analysis A of trnL/trnT sequence data illustrating the distribution of chromosome numbers in the genus. Branch lengths represent character state changes and the number of changes is listed above each branch. 248 Annexes ______________________________________________________________ Annex 5. List of morphological descriptors in the genus Passiflora L. PASSPORT INSTITUTE CODE: ______________________________________________ ACCESSION NUMBER: ______________________________________________ ACCESSION NAME: ______________________________________________ STATUS OF SAMPLE: ______________________________________________ SCIENTIFIC NAME: ______________________________________________ VERNACULAR NAME: ______________________________________________ CULTIVAR ORIGIN: 1. Open pollination: 2. Artificial pollination: 3. Clonal selection: 4. Seedling selection: ______________________________________________ ______________________________________________ ______________________________________________ ______________________________________________ COLLECTING DESCRIPTION: 1. Country of origin: 2. Province / state / Deparment: 3. Country 4. Collection site: 5. Latitude: 6. Longitude: 7. Altitude: ______________________________________________ ______________________________________________ ______________________________________________ ______________________________________________ ______________________________________________ ______________________________________________ ______________________________________________ ENTHNOBOTANICAL DATA: ______________________________________________ ______________________________________________ COLLECTOR: ______________________________________________ DATE: ______________________________________________ NOTES: ______________________________________________ ______________________________________________ ______________________________________________ 249 Annexes ______________________________________________________________ 1. STEM_________________________________________________________ 1.1 HABIT (STHA) 1. Liana (eg. P. incarnata) 2. Tree (eg. P. emarginata) 3. Shrub (eg. P. macrophylla) 1.2 EXTERNAL SHAPE (STSH) 1. Round 2. Striate 3. Subangulate 4. Angulate (e.g. P. quadrangularis) 1 2 3 4 5. Other (Specify) 1.3 ANTHOCYAN (STAN) 1. Absent 2. Medium (< 80%) 3. High (> 80%) 1.4 PUBESCENCE (STPU) 1. Glabrous 2. Few density 3. Tomentoso 4. Villous 5. Pilose 1 2 3 4 1.5 INTERNODE LENGTH (STIN) mm mm mm mm mm mm mm 1.6 DIAMETERS (STDI) mm mm mm 2. TENDRIL______________________________________________________ 2.1 TENDRILS (TEPRE) 1. Absent 2. Present 2.2 SPIRAL SHAPE (TESH) 1. Cylindrical 2. Conic 3. Compound 1 250 2 Annexes ______________________________________________________________ 4. Linear 5. Indefinite 2.3 PUBESCENCE (TEPU) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 1 2 3 4 5 2.4 ANTHOCYANIN (TEAN) 1. Absent 2. Medium (< 80%) 3. High (> 80%) 3. STIPULE______________________________________________________ 3.1 STIPULES (TEPR) 1. Present 2. Absent 3.2 DURATION (SPPE) 1. Permanent 2. Deciduous 3.3 COLOR (SPCO) 3.4 PUBESCENCE (SPPU) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 3.5 SHAPE (SPSH) 1. Setaceous 2. Linear 3. Pinnatisect 4. Lobed 5. Lanceolate 6. Oblate 7. Ovate 8. Reniform (or auricular) 1 2 3 251 4 5 6 7 8 Annexes ______________________________________________________________ 9. Other (specify) 3.6 MARGIN (SPMA) 1. Entire 2. Serrate 3. Serrulate 4. Dentate 5. Doubly dentate 6. Crenate 7. Other (specify) 1 2 3 4 5 6 3.7 ANTHOCYANIN (SPAN) 1. Absent 2. Medium (< 80%) 3. Heigth (> 80%) 3.8 LENGTH (SPLE) mm mm mm mm mm 3.9 WIDTH (SPWI) (including the arist) mm mm mm mm mm 3.10 TERMINAL ARIST LENGTH (SPTA) mm mm mm mm mm 4. PETIOLE______________________________________________________ 4.1 ANTHOCYANIN (PEAN) 1. Absent 2. Medium (< 80%) 3. High (> 80%) 4.2 PUBESCENCE (PEPU) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 4.3 COLOR (PECO) 252 Annexes ______________________________________________________________ 4.4 NECTARY SHAPE (PENS) 1. Discoid (e.g. P. edulis f. flavicarpa) 2. Extended (e.g. P. ligularis) 3. Spherical (e.g. P. adenopoda) 4. Other 1 4.5 STIPE OF NECTARIES (PESN) 1. Absent 2. Intermediate 3. Present 2 1 3 2 3 4.6 LENGHT (PELE) mm mm mm mm mm 4.7 DISTANCE FROM BASE TO FIRST GLAND (PEDG) mm mm mm mm mm 4.8 NECTARY GLAND NUMBER (PENM) 5. LEAF_______________ ____ ____________________________________ 5.1 HETEROPHYLLY (LEPO) 1. Absent 2. Present 5.2 LOBE NUMBER (LELN) 5.3 MARGIN (LEMA) 1. Entire 2. Serrate 3. Serrulate 4. Dentate 5. Doubly dentate 6. Crenate 1 253 2 3 4 5 6 Annexes ______________________________________________________________ 7. Other (specify) 5.4 BASE SHAPE (LEBS) 1. Cuneate 2. Rounded 3. Truncate 4. Cordate (Heart-shaped) 5. Deeply cordate 6. Articulate 7. Other (specify) 1 2 5.5 APEX SHAPE (LEAS) 1. Rounded 2. Obtuse (>90o) 3. Acute 4. Acute (<45o) 5. Other (specify) 3 1 2 5.6 ACUMEN (LEPA) 1. Absent 2. Presence 2 5.7 PUBESCENCE ADAXIAL (LEAX) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 5.8 PUBESCENCE ABAXIAL (LEPB) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 5.9 ANTHOCYANIN - LAMINA (LEAL) 1. Absent 2. Medium (< 80%) 3. High (> 80%) 5.10 ANTHOCYANIN - NERVES (LEAN) 1. Absent 2. Medium (< 80%) 3. High (> 80%) 5.11 COLOR – ADAXIAL (LECA) 254 4 5 3 6 4 Annexes ______________________________________________________________ 5.12 MARGIN SERRATION DENSITY (number on 2 cm) (LEMS) mm mm mm mm mm 5.13 ANGLE BETWEEN LATERAL LOBES (LEAB) mm mm mm mm mm 5.14 CENTRAL LOBE LENGTH (LELC) mm mm mm mm mm 5.15 RIGHT LOBE LENGTH (LERL) mm mm mm mm mm 5.16 CENTRAL LOBE WIDTH (LECL) mm mm mm mm mm 5.17 DISTANCE BETWEEN LEAF SINUS AND PETIOLE INSERTION (LESS) mm mm mm mm mm 5.18 HETEROBLASTY (LEPH) 0. Absent 1. Present 5.19 LAMINAR NECTARIES (LENL) 1. Absent 2. Present 5.20 NECTAR GLAND NUMBER ON LAMINA (LELA) 5.21 DISTRIBUTION OF LAMINAR NECTARIES (LEDN) 1. Along central lobe 255 Annexes ______________________________________________________________ 2. On lateral lobes 3. Around lamina 4. Around apex 5. At base 5.22 PRESENCE OF MARGINAL NECTARIES (LEPN) 1. Absent 2. Present 5.23 NECTAR NUMBER ON LEAF MARGIN (LENN) 5.24 DISTRIBUTION OF LAMINAR MARGIN NECTARIES (LELM) 1. Base 2. Base and sinus 3. Margin 4. Other (specify) 5.25 NECTARY SHAPE (LENS) 1. Discoid (squashed) 2. Extended 3. Spherical 4. Other 1 2 3 6. PEDUNCLE____________________________________________________ 6.1 PUBESCENCE (PDPU) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 6.2 ANTHOCYANIN (PDAN) 1. Absent 2. Medium (< 80%) 3. High (> 80%) 6.3 BIFURCATION (PDBN) 1. Present 2. Absent 6.4 LENGTH (PDLE) 256 Annexes ______________________________________________________________ mm mm mm mm mm mm mm 6.5 DIAMETER (PDDI) mm mm mm 6.6 PEDICEL LENGTH (PDPL) mm mm mm mm mm 6.7 LENGTH TO FIRST BIFURCATION (PDLF) mm mm mm mm mm 6.8 LENGTH TO SECOND BIFURCATION (PDBS) mm mm mm mm mm 7. BRACT________________________________________________________ 7.1 UNION (BRUN) 1. Free 2. United at less than half 3. United at half 4. United at more than half 7.2 BRACT (BRPR) 1. Present 2. Absent 7.3 PERMANENCE (BRPE) 1. Permanent 2. Deciduous 7.4 PUBESCENCE (BRPU) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 257 Annexes ______________________________________________________________ 7.5 POSITION OF PUBESCENCE (BRPP) 1. Absent 2. Adaxial 3. Abaxial 4. Both sides 7.6 COLOR (BRCO) 7.7 ANTHOCYANIN (BRAN) 1. Absent 2. Medium (< 80%) 3. High (> 80%) 7.8 SHAPE (BRSH) 1. Setaceous 2. Linear 3. Pinnatisect 4. Lobate 5. Lanceolate 6. Oblate 7. Ovate 9. Other (specify) 1 2 3 4 5 6 7.9 MARGIN (BRMA) 1. Entire 2. Serrate 3. Serrulate 4. Dentate 5. Doubly dentate 6. Crenate 7. Other (specify) 7.10 APEX SHAPE (BRAS) 1. Round 2. Obtuse (>90 o) 3. Acute 4. Very acute (<45 o) 5. Other 1 2 7.11 MARGIN NECTARIES (BRNM) 1. Present 2. Absent 7.12 LENGTH (BRLE) mm mm mm mm mm 258 3 4 7 Annexes ______________________________________________________________ 7.13 WIDTH (BRWI) mm mm mm mm mm 8. FLOWER_______________________________________________________________ 8.1 CORONA TYPE (FLCY) 1. Tuberculous 2. Filamentous 8.2 COROLLA TYPE (FLCT) 1. Campanulate (e.g. P. mixta) 2. Intermediate (eg. P. manicata) 3. Reflex (eg. P. tarminiana) 4. Other (specify) 1 8.3 ORIENTATION (FLOF) 1. Pendular (eg. P. antioquiensis) 2. Intermediate (eg. P. mixta) 3. Erect (eg. P. manicata) 8.4 ORIENTATION (in Degrees to vertical) (FLOG) 8.5 PUBESCENCE ON COROLLA (FLPU) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 8.6 HYPHANTHIUM PUBESCENCE (FLHP) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 8.7 DOMINANT PETAL COLOR (FLCP) 259 2 3 Annexes ______________________________________________________________ 8.8 CHLOROPHYLL ON EXTERIOR OF SEPAL (FLCS) 1. Absent 2. Present 8.9 KEEL-SHAPED SEPALS (FLKS) 1. Absent 2. Present 8.10 SEPAL AWN (CAS) 1. Absent 2. Present 8.11 UNION OF SEPALS (FLUS) 1. Free 2. United 8.12 PETALS (FLPP) 1. Present 2. Absent 8.13 PETAL LENGTH (FLPL) mm mm mm mm mm 8.14 PETAL WIDTH (FLPW) mm mm mm mm mm 8.15 SEPAL LENGTH (FLSL) mm mm mm mm mm 8.16 SEPALS WIDTH (FLSW) mm mm mm mm mm 8.17 DIAMETER OF NECTARY CHAMBER (FLNC) mm mm mm mm mm 8.18 HYPANTHIUM DIAMETER ABOVE NECTARY CHAMBER (FLHD) mm mm mm mm mm 260 Annexes ______________________________________________________________ 8.19 HYPANTHIUM DIAMETER – DISTAL (FLHS) mm mm mm mm mm 8.20 FLOWER LENGTH (FLLE) mm mm mm mm mm 8.21 HYPHANTIUM LENGHT (FLHL) mm mm mm mm mm 8.22 LENGHT OF NECTARY CHAMBER (FLCN) mm mm mm mm mm 8.23 NUMBER OF CORONA SERIES (FLNS) 8.24 COLOR OF FILAMENTS AT BASE (FLCB) 8.25 COLOR OF FILAMENTS AT APEX (FLCA) 8.26 FILAMENT LENGTH (FLFL) mm mm mm mm mm 8.27 DISTRIBUTION OF ANTHERS (FLDA) 1. Symmetry radial 2. Symmetry bilateral 3. Other 8.28 COLOR OF STAMINAL FILAMENTS (CFE) 261 Annexes ______________________________________________________________ 8.29 STAMINAL FILAMENTS LENGTH (FLSF) mm mm mm mm mm 8.30 OVARY PUBESCENCE (FLOP) 1. Glabrous 2. Low density 3. Tomentose 4. Villous 5. Pilose 8.31 COLOR OF OVARY (FLCO) 8.32 OVARY LENGTH (FLOL) mm mm mm mm mm 8.33 COLOR OF STYLE (FLCS) 8.34 STYLE LENGHT (FLSL) mm mm mm mm mm 8.35 COLOR DISTRIBUTION ON STYLES (FLDS) 1. Uniform 2. Specked 3. Apex 4. Base 5. Other 8.36 COLOR OF STIGMA (FLCG) 8.37 COLOR OF ANDROGYNOPHORE (FLCN) 262 Annexes ______________________________________________________________ 8.38 COLOR DISTRIBUTION ON ANDROGYNOPHORE (FLDN) 1. Uniform 2. Speckled 3. Other 8.39 PUBESCENCE OF ANDROGYNOPHORE (FLPN) 1. Glabrous 2. Few density 3. Tomentose 4. Villous 5. Pilose 8.40 GYNOPHORE LENGTH (FLGL) mm mm mm mm mm 8.41 ANDROGYNOPHORE LENGTH (FLAL) mm mm mm mm mm 8.42 OPERCULUM LENGTH (FLOP) mm mm mm mm mm 8.43 LIMEN LENGTH (FLLL) mm mm mm mm mm 8.44 LIMEN MARGIN (FLML) 1. Flat-entire 2. Flat-serrate 3. Wavy-entire 4. Wavy-serrate 8.45 NECTARY CHAMBER RING (FLNR) 1. Absent 2. Present 8.46 HYPANTHIUM TYPE (FLHY) 1. Flat (eg. P. suberosa) 2. Campanulate (P. ligularis) 3. Tubular (P. tarminiana) 263 Annexes ______________________________________________________________ 8.47 INTERNAL COLOR OF HYPHANTIUM (FLCI) 8.48 CLOROPHYLLA ON EXTERIOR OF HYPANTHIUM (FLCE) 1. Absent 2. Partial 3. Global 8.49 ANTHOCYANIN ON EXTERIOR OF HYPANTHIUM (FLAE) 1. Absent 2. Partial 3. Global 8.50 STYLES NUMBER PER FLOWER (NEF) 8.50 NECTARS ON SEPALS (FLNS) 1. Absent 2. Present 8.51 DOMINANT SEPAL COLOR (FLCP) 8.52 ANTHOCYANIN ON EXTERIOR OF SEPALS (FRSH) 1. Absent 2. Partial 3. Global 9. FRUIT________________________________________________________ 9.1 TYPE (FRTY) 1. Berry 2. Capsule 9.2 PIGMENTATION OF INMATURE FRUITS (FPFIM) 1. Uniform 2. Dotted 3. Lined 4. Other 9.3 DOMINANT COLOR OF THE RIPE FRUIT (FRCF) 264 Annexes ______________________________________________________________ 9.4 SHAPE (FRSH) 1. Spherical / Round 2. Ovoid 3. Oblate 4. Oblong 5. Ellipsoid 6. Fusiform 7. Other 1 2 3 4 5 9.5 PUBESCENCE (FRPU) 1. Absent 2. Present 9.6 WEIGHT (FRWH) g g g g g 9.7 LENGTH (FRLE) mm mm mm mm mm mm mm 9.8 DIAMETER (FRDM) mm mm mm 9.9 TRANSVERSAL SECTION (FRTS) 1. Round 2. Hexagonal 3. Triangular 4. Other (specify) 1 9. PERICARP DIAMETERS (FREM) mm mm mm mm mm 9.11 MESOCARP TEXTURE (FRMT) 1. Hard 2. Soft rough 3. Soft 4. Other 265 2 3 6 Annexes ______________________________________________________________ 9.12 ARIL TASTE (FRAT) 1. Acid 2. Acid sweet 3. Sweet 4. Insipid 5. Other 9.13 SKIN WEIGHT (FRSW) g g g g g 9.14 SEEDS WEIGHT (FRSW) g g g g g 9.15 WEIGHT OF JUICE AND PULP (FRJP) g g g g g 9.16 SKIN TEXTURE (FRST) 1. Soft 2. Resistant to the pressure 3. Brittle 9.17 JUICE PH (FRPH) 9.18 TITRABLE ACIDITY (FRAT) mg/100ml 9.19 TOTAL SOLUBLE SOLIDS (FRTS) 9.20 ASCORBIC ACID (FRVC) mg/100ml 9.21 JUICE AROMA (FRAJ) 1. Weak 2. Intermediate 3. Strong 266 Annexes ______________________________________________________________ 10. SEED_________________________________________________________ 10.1 SURFACE TYPE (SEST) 1. Smooth 2. Reticulate 3. Other 10.2 ARIL COLOR (SEAC) 10.3 COAT COLOR (SECC) 10.4 SEED BRIGHTNESS (SEBS) 1. Mate 2. Intermediate 3. Brilliant 10.5 SEED SHAPE (SESS) 1. Obovate 2. Cordate 3. Pyramidal 4. Obcordate 5. Cuneate 6. Other (specify) 1 2 10.6 WEIGHT (100 seeds) (SEWH) mm mm mm mm mm mm mm mm mm 10.7 LENGTH (SELE)) mm mm mm 10.8 WIIDTH (SEWI) mm mm mm 10.9 NUMBER OF SEEDS PER FRUIT (SENF) g g g g g 267 3 4 5 Annexes ______________________________________________________________ 11. POLLINATION SYNDROME____________________________________ 1. 2. 3. 4. 5. 6. 7. Hummingbirds (e.g. P. mixta, P. vitifolia). Bats (e.g. P. mucronata, P. lobata). Hummingbirds and bats (e.g. P. penduliflora). Bees (e.g. P. alnifolia, P. filipes). Wasps (e.g. P. edulis f. flavicarpa, P. quadrangularis). Bees and Wasps (e.g. P. foetida, P. sphaerocarpa). Others (specify). 12. PEST AND DISEASE SUSCEPTIBILITY___________________________ 12.1 PEST ___________________________________________________________________________ ___________________________________________________________________________ ___________________________________________________________________________ 12.2. FUNGI _____________________________________________________________________________________________ _____________________________________________________________________________________________ _____________________________________________________________________________________________ 12.3 BACTER _____________________________________________________________________________________________ _____________________________________________________________________________________________ _____________________________________________________________________________________________ 12.4 VIRUS _____________________________________________________________________________________________ _____________________________________________________________________________________________ _____________________________________________________________________________________________ 12.5 ABIOTIC STRESS SUSCEPTIBILITY _____________________________________________________________________________________________ _____________________________________________________________________________________________ _____________________________________________________________________________________________ 12.6 OTHERS DISORDERS _____________________________________________________________________________________________ _____________________________________________________________________________________________ _____________________________________________________________________________________________ 268 View publication stats