RELACIONES BIOGEOGRÁFICAS DE CUATRO ÁREAS DE BOSQUE SECO TROPICAL (COLOMBIA, COSTA RICA Y MÉXICO) UTILIZANDO COMO GRUPO INDICADOR LOS ESCARABAJOS COPRONECRÓFAGOS (COLEOPTERA: SCARABAEIDAE: SCARABAEINAE) TESIS QUE PRESENTA DORA NANCY PADILLA GIL PARA OBTENER EL GRADO DE DOCTOR EN CIENCIAS SISTEMÁTICA Xalapa, Veracruz, México 2006 1 Aprobación final del documento final de tesis de grado: Relaciones biogeográficas de cuatro áreas de bosque seco tropical (Colombia, Costa Rica y México) utilizando como grupo indicador los escarabajos copronecrófagos (Coleoptera: Scarabaeidae: Scarabaeinae) Nombre Firma Director Dr. Gonzalo Halffter Salas Comité Tutorial Dr. Alejandro Espinosa de los Monteros Solís Dr. Juan José Morrone Lupi Dr. Francisco Cabrero Jurado Dr. Pedro Reyes Castillo Dr. Mario Enrique Favila Dra. Ellen Andresen 2 DECLARACIÓN Excepto cuando es explícitamente indicado en el texto, el trabajo de investigación contenido en esta tesis fue efectuado por Dora Nancy Padilla Gil como estudiante de Doctorado en Ciencias Sistemática entre agosto de 2002 y agosto del 2006, bajo la supervisión del Dr. Gonzalo Halffter Salas. Las investigaciones reportadas en esta tesis no han sido utilizadas anteriormente para obtener otros grados académicos, ni serán utilizadas para tales fines en el futuro. Candidato: Dora Nancy Padilla Gil Director de tesis: Dr. Gonzalo Halffter Salas 3 Con amor dedico este trabajo a mis padres Miguel y Ana Delia, a mis hermanos Myriam y Wilson y a mi hijo Nicolas 4 AGRADECIMIENTOS Deseo agradecer al Gonzalo Halffter por la dirección del trabajo de tesis, igualmente a David Edmonds, Alejandro Espinosa de los Monteros y Juan José Morrone, por la orientación y revisión del trabajo. A Federico Escobar, Fernando Vaz de Mello, Mario Zunino, Ángel Solís, Francisco Lorea y Alfonso Díaz, por facilitarme bibliografía y responder mis dudas. A los jurados de tesis: Ellen Andresen, Francisco Cabrero, Pedro Reyes Castillo, Mario Enrique Favila. A Irma Guerrero, Carmen Huerta y Maria Eugenia Rivas por su apoyo. A mis padres Miguel y Ana, por inculcarme el espíritu de superación y a mis hermanos Myriam y Wilson por su ayuda incondicional. Al personal de Postgrado del Instituto de Ecología, por su colaboración y a la Universidad de Nariño, Pasto (Colombia: Nariño) por otorgarme la Comisión de Estudios. 5 CONTENIDO Págs. Presentación 1 CAPITULO 1. Introducción General 3 CAPITULO 2. Biogeography of the areas and Canthonini (Coleoptera: Scarabaeidae) of dry tropical forests in Mesoamerica and Colombia 7 Abstract 8 Inroduction 11 Materials and Methods 13 Results 19 Discussion 29 Literature cited 33 CAPITULO 3. Mesoamerican and Colombian Dry Tropical Forest and Phanaeine Dung Beetles (Coleoptera: Scarabaeidae, Scarabaeine): A Cladistic Analysis of Relationships Abstract 59 Introduction 60 Methods 61 Results 62 Discussion 63 Literature cited 67 CAPITULO 4. Discusión y Conclusiones Literatura citada 82 95 6 ÍNDICE DE FIGURAS, TABLAS Y APÉNDICES Págs. CAPITULO 2 Fig. 1. Study sites 45 Fig. 2. Cladogram of geological areas 46 Fig. 3. Cladogram of the areas of affinity of Canthonini. 47 Table 1. Bibliography consulted for the selection of geomorphological characters 48 Table 2. Geological-geomorphological characters and their states 49 Table 3. Matrix with character states for the cladogram of geomorphological areas 50 Table 4. List of Canthonini species from the study forests 51 Table 5. Data matrix for the cladogram of the Canthonini areas 52 Appendix 1. Scarabaeinae (Canthonini and other tribes) of the forests studied 53 CAPITULO 3 Fig. 1. Study sites 74 Fig. 2. Cladogram of relationships among dry forest sites, using phanaeines as the indicator group 75 Fig. 3. Cladogram of relationships of dry forest sites using Phanaeus as the indicator group 76 Appendix 1. Description of study sites 77 Appendix 2. Phanaeini occurring in the study areas 79 Appendix 3. Cladogram data matrix, study site 80 Appendix 4. Phanaeus, Coprophanaeus and Canthon (Glaphyrocanthon) viridis group 81 CAPITULO 4 Cuadro 1. Canthonini, índice de similitud de Jaccard/ distancia 92 Cuadro 2. Phanaeini índice de similitud de Jaccard/ distancia 93 Cuadro 3. Comparación de la diversidad florística de los bsT 94 7 PRESENTACIÓN Los objetivos de este trabajo son estudiar la biogeografía histórica de los bosques secos tropicales de Mesoamérica1 y Colombia y establecer la relación entre el origen y la distribución de los bosques secos tropicales y la biogeografía de los Scarabaeinae (Canthonini y Phanaeini) presentes en tales bosques secos. Las áreas estudiadas son: en la Costa Pacifica de Mesoamérica, Chamela (México: Jalisco) y Palo Verde (Costa Rica: Guanacaste); en Colombia en la Región Caribe, cuatro sitios: Neguanje, Tierra Bomba, Los Colorados y Zambrano; también en Colombia, en el Valle Superior del Magdalena, el norte del Tolima. Los anteriores bosques secos conforman el grupo interno, las áreas de bosque húmedo tropical constituyen el grupo externo y son Los Tuxtlas (Veracruz, México) y Leticia (Amazonas, Colombia). El capítulo II incluye tanto la biogeografía de las áreas de bosque seco (bsT) como la biogeografía de los Canthonini; el capitulo III comprende la biogeografía de áreas de bsT tomando como indicador a los Phanaeini y el análisis comparativo entre las relaciones de los Canthonini y Phanaeini que comparten los mismos escenarios biogeográficos. La biogeografía de las áreas de bosque seco se abordó a través de un análisis cladístico basado en caracteres geológicos y geomorfológicos de los terrenos donde se ubican los enclaves de bosque seco y su interpretación en el contexto histórico-geográfico, climático y ecológico. Para el análisis biogeográfico de los Canthonini y de los Phanaeini de los bosques secos se utilizó el Análisis de Parsimonia de Endemismos (PAE) enraizando el cladograma con el grupo externo previamente indicado. Para este análisis se tomó en cuenta la información publicada para cada una de las localidades, los estudios faunísticos, ecológicos, biogeográficos, taxonómicos y filogenéticos sobre los Canthonini y Phanaeini, las bases de datos SNIB-CONABIO de México e INBio de Costa Rica. Los resultados revelan la formación de los bosques secos a partir del Pleistoceno y la fijación de sus condiciones ambientales a partir del Holoceno Medio. Considerando los Canthonini, los bosques secos más afines son los de Colombia y Chamela (México). Se plantean tres patrones que explican tanto las relaciones como su presencia en los bosques secos. Los bosques secos de Chamela y Palo Verde más los del Valle Superior de Magdalena y la Región Caribe de Colombia, según los Phanaeini presentan mayor afinidad; se 1 Mesoamérica incluye México y los siete países de América Central (Guatemala, Belice, Honduras, El Salvador, Nicaragua, Costa Rica y Panamá; 7- 18°N y 77- 92°W). Ver distribución geográfica del bosque seco tropical, fig. 1. 1 plantea un patrón biogeográfico que explica los resultados anteriores. El análisis biogeográfico comparativo entre los dos grupos establece la convergencia del patrón de especiación in situ y proporciona elementos para formular varias hipótesis biogeográficas. La biogeografía de los enclaves de bosque tropical seco de Mesoamérica y Colombia y la coincidencia espacio-temporal de los Canthonini y Phanaeini en el mismo escenario permitió proponer elementos tanto para entender el origen y la evolución de los bosques secos como los patrones de distribución de los dos grupos de Scarabaeinae, así como plantear nuevas hipótesis sobre el probable origen espaciotemporal de los Canthonini y Phanaeini de los bosques neotropicales considerados. 2 CAPÍTULO I Introducción General La biogeografía es la disciplina científica consagrada al estudio de los patrones de distribución espacial de los organismos y las causas o procesos históricos y ecológicos que los determinan. En esta definición se subrayan las dos aproximaciones esenciales al estudio de distribución de organismos: la que indaga las causas próximas que operan en el presente (escala ecológica) y la que estudia las causas pretéritas de gran magnitud espacio-temporal (escala histórica) (MartínPiera & Sanmartín 1999). Dentro de la biogeografía histórica contemporánea existen tres enfoques para explicar la distribución de los seres vivos: el dispersalismo, la panbiogeografía y la biogeografía cladística. El más antiguo es la biogeografía dispersalista, que se originó a partir de las ideas de Darwin y Wallace a mediados del siglo XIX (Crisci y Morrone 1992). Trabaja con taxones individuales, en el sentido que son los organismos poblaciones/especies los que se dispersan sobre una geografía estable. Como reacción a la biogeografía dispersalista, a mediados del siglo XX surge la panbiogeografia, la cual fue propuesta originalmente por Croizat (1958). Croizat hizo énfasis en el análisis conjunto de diferentes taxones para buscar patrones comunes de distribución, evitando analizar un sólo taxón como se hacía tradicionalmente. Supone que organismos con distintas capacidades de dispersión pueden compartir similitudes en sus distribuciones, ya que cualquier distribución en plantas de alguna u otra forma tiene su contraparte en los animales (Craw et al. 1999). A partir de la combinación de la biogeografía con la sistemática filogenética, surge la biogeografía cladística en la década de los setenta (Morrone et al. 1996). El Análisis Parsimonioso de Endemicidad (PAE) es considerado como una técnica panbiogeográfíca y ha sido aplicado por diferentes autores (ver Morrone 2004). La panbiogeografía intenta identificar homologías biogeográficas primarias, las cuales suponen una historia biogeográfica común. La biogeografía cladística asume la correspondencia entre las relaciones filogenéticas de los taxones estudiados y las relaciones entre las áreas donde éstos habitan y trata las homologías biogeográficas secundarias, las cuales representan la prueba cladística de la homología biogeográfica primaria antes reconocida. La panbiogeografía y la biogeografía cladística pueden ser combinadas bajo el denominador común de “biogeografía vicariante”, cuyo principal objetivo es la homología biogeográfica (Ebach & Morrone 2005). Las relaciones entre áreas geológicas tratan las áreas como unidades funcionales (reales) y 3 plantean hipótesis similares a las presentadas en sistemática filogenética (Wiley 1988). Craw (1988) implementó esta metodología enraizando el cladograma con base en un área geológica ancestral hipotética codificada en ceros. En la biogeografía histórica la reconstrucción de eventos pasados puede ser realizada desde tres perspectivas diferentes, cada una con objetivos diferentes: 1) reconstrucción de la distribución histórica de grupos individuales (biogeografía del taxón), 2) reconstrucción de las áreas de endemismo (búsqueda de relaciones generales de áreas, biogeografía de área) y 3) reconstrucción de la distribución histórica de biotas (todas las especies que habitan en una región específica y comparten una historia geográfica). Las dos últimas buscan la homología espacial: elementos comunes espacio-temporales compartidos de la biogeografía histórica (Crisci et al. 2006). La base del conocimiento biogeográfico está integrada por los patrones de distribución geográfica de los organismos, las relaciones (genealógicas) entre ellos y el conocimiento acerca de la historia del espacio físico en que se distribuyen. Por tanto, la base explicativa de la biogeografía histórica integra los principios biológicos generales que relacionan espacio y evolución (Andersson 1996). Los patrones de distribución de los grupos de organismos han sido determinados por evolución, adaptaciones fisiológicas y etológicas, mecanismos de dispersión e igualmente por habilidades de dispersión, competencia entre especies, sucesión ecológica, cambios de clima, cambios en el nivel del mar, movimiento de los continentes y recientemente, por los impactos directos e indirectos de los seres humanos (Spellerberg & Sawyer 2000). Los patrones de distribución de los organismos son el resultado de ecología e historia y representan el objetivo común y último hacia la aproximación de la biogeografía ecológica e histórica (Crisci et al. 2006). Los objetivos de este trabajo son estudiar la biogeografía histórica de los bosques secos tropicales de Mesoamérica y Colombia y establecer la relación entre el origen y la distribución de los bosques secos tropicales y la biogeografía de los Scarabaeinae (Canthonini y Phanaeini) presentes en tales bosques secos. El bosque seco tropical (bsT) o bosque tropical caducifolio o selva caducifolia, se define como aquella formación vegetal que presenta una cobertura boscosa continua, con condiciones de lluvia marcadamente estacionales, que se distribuye entre los 0-1000 m de altitud, con temperaturas superiores a los 24 ºC y precipitaciones entre los 700 y 2000 mm anuales (IAvH 1997). Presenta especies arbóreas que pierden sus hojas en la época seca del año durante un lapso variable que en 4 general oscila alrededor de seis meses (Rzedowski 1978). El bsT se encuentra dentro de los ecosistemas más intensamente utilizados, perturbados y menos conservados del Neotrópico. Tanto el origen como la biogeografía histórica de los bosques secos neotropicales siguen siendo imprecisas y poco entendidas (ver Sarmiento 1975, Pennington et al. 2000, Becerra 2005). De acuerdo con Sarmiento (1975), es probable que las formaciones vegetales secas actuales de Centro y Suramérica se hayan originado y evolucionado de manera independiente desde hace 1.8 millones de años, a partir de cuatro grandes comunidades florísticas. La primera incluye los bosques secos tropicales que se extienden desde el sur de México al norte de Sudamérica; la segunda incluye las formaciones de Caatingas del noreste de Brasil; la tercera, los bosques del Chaco, y la cuarta, los arbustales y bosques de clima mediterráneo del centro y sur de Chile. Este trabajo aborda la primera comunidad florística delimitada por Sarmiento (1975). Para tal efecto se eligieron los siguientes bosques secos tropicales: en la Costa Pacifica de Mesoamérica: Chamela (México: Jalisco) y Palo Verde (Costa Rica: Guanacaste); en Colombia en la Región Caribe, cuatro sitios: Neguanje, Tierra Bomba, Los Colorados, Zambrano, y en el Valle Superior del Magdalena: el Norte del Tolima. Los grupos de Scarabaeinae seleccionados son dos tribus, los Canthonini (excepto el género Deltochilum) y los Phanaeini, las cuales presentan distribución neotropical. Estos grupos han sido incluidos en numerosos estudios faunísticos y ecológicos principalmente relacionados con diversidad y abundancia, algunos de éstos en los bosques secos tropicales (ver Andresen 2005). Sin embargo, se desconocen las afinidades de ésta biota en los distintos bsT así como las relaciones con el origen y distribución de los bsT y el porqué de su distribución geográfica. Las relaciones biogeográficas de las áreas de bsT, así como la presencia de los Scarabaeinae en tales áreas se basa en una serie de hipótesis y se explican empleando una perspectiva holística que tiene en cuenta los eventos histórico-geomorfológicos de los terrenos donde se ubican los bosques secos, el origen y eventos climáticos, las afinidades taxonómicas y/o filogenéticas y los factores que influyeron en los procesos de especiación y dispersión de los taxones implicados. En el segundo capítulo se incluye la biogeografía histórica de las áreas de bosque seco que pretende inferir y explicar el origen y relaciones histórico-geomorfológicas de tales enclaves. Las relaciones biogeográficas de los Canthonini y de los Phanaeini (Scarabaeinae) en los bosques secos de Mesoamérica y Colombia se abordarán en sendos capítulos (II y III), los cuales se 5 orientan bajo los siguientes supuestos: 1) El origen y establecimiento de los bosques secos guarda relación con los procesos de dispersión y especiación de los Scarabaeinae examinados. 2) Los Scarabaeinae de los bosques secos muestran relaciones taxonómicas y biogeográficas con los taxones provenientes de Sudamérica. 3) Tanto los Canthonini como los Phanaeini de bsT presentan similaridades en sus patrones biogeográficos. El análisis de sus relaciones aportará argumentos para plantear nuevas hipótesis biogeográficas. El estudio biogeográfico es abordado bajo una concepción histórica y ecológica para permitir una aproximación al entendimiento del origen y de las relaciones de los bosques secos de Mesoamérica y Colombia, la identificación de los patrones biogeográficos de los Scarabaeinae presentes en tales enclaves, y la formulación de hipótesis sobre los procesos que los causan. 6 CAPITULO 2 Biogeography of the areas and Canthonini (Coleoptera: Scarabaeidae) of dry tropical forests in Mesoamerica and Colombia Acta Zoológica Mexicana, vol. 23, 1 (2007) 7 CAPÍTULO II BIOGEOGRAPHY OF THE AREAS AND CANTHONINI (COLEOPTERA: SCARABAEIDAE) OF DRY TROPICAL FORESTS IN MESOAMERICA AND COLOMBIA Dora Nancy PADILLA-GIL1 and Gonzalo HALFFTER 1 1 Instituto de Ecología, A.C., Apdo. Postal 63, 91000 Xalapa, Veracruz, MEXICO [email protected], [email protected] ABSTRACT This biogeographical analysis examines the historical, geological, climatic and ecological processes that have influenced the formation of the dry tropical forests (DTF) of Mesoamerica and Colombia, areas that are the setting for multiple biogeographical stories that in this case are illustrated by the patterns and evolutionary processes of Canthonini. In this study we test the hypothesis that the Canthonini fauna of dry tropical forests has a South American affinity. To this end, we compare extant species from a tract of dry tropical forest in Mexico, a second enclave in Costa Rica, four from the Caribbean region of Colombia and finally one from the north of Tolima in the Upper Magdalena River Valley, Colombia. The geomorphological characteristics of the enclaves of DTF are also compared, as are the geographical distribution and taxonomic affinities of each of the species found in these dry tropical forests. The biogeographical, historical and geological aspects of the enclaves were evaluated using a Parsimony Analysis of Endemicity (PAE), with two tropical rain forests as the outgroup: Leticia (Amazonas, Colombia) and Los Tuxtlas (Veracruz, Mexico). This study reveals the origin and distribution of Neotropical dry forests in the Pleistocene, and the establishment of its dry conditions during the Holocene. It also reveals apparent similarities among the Canthonini of the dry tropical forests of Mexico, Costa Rica and Colombia, with three geographical distribution patterns that correspond to different degrees of expansion towards the north and the diversification of evolutionary lines, and even species of South American origin. The comparison of the cladogram generated for species of Canthonini with that of the geological events that have occurred in the study regions indicates that the distribution of Canthonini in dry tropical forests began during the Pliocene with the re-establishment of the Panamanian 8 connection, with no evidence of previous geomorphological events having any influence. On the other hand, there are few species shared with the tropical rain forests used as the outgroup for the cladograms. Key words: Scarabaeinae, Canthonini, dry tropical forest, Mexico, Costa Rica, Colombia, biogeographical patterns RESUMEN Este análisis biogeográfico examina los procesos históricos: geológicos, climáticos y ecológicos que han influido en la formación de los bosques tropicales secos (DTF) de Mesoamérica y Colombia, bosques que son sitios de múltiples historias biogeográficas como la que ilustran los Canthonini. Sometemos a prueba la hipótesis que la fauna de Canthonini de los bosques tropicales secos tiene una afinidad sudamericana. Para este propósito comparamos las especies que se encuentran en un enclave de bosque tropical seco en México, en un segundo enclave en Costa Rica, en cuatro de la región Caribe de Colombia y finalmente uno más en el norte de Tolima en el valle superior del río Magdalena, Colombia. Las características geomorfológicas de los enclaves son también comparadas, así como la distribución geográfica y las afinidades taxonómicas de cada una de las especies de Canthonini que se encuentran en estos bosques tropicales secos. Los aspectos de historia biogeográfica, geológica y ecológica de los enclaves son evaluados usando un Análisis de Parsimonia de Endemicidad (PAE), utilizando como grupo externo dos lugares de selva siempre-verde: Leticia (Amazonas, Colombia) y Los Tuxtlas (Veracruz, México). Este estudio pone de manifiesto que el origen y distribución de los bosques tropicales secos de la región ocurre durante el Pleistoceno, con una acentuación de las características de sequía durante el Holoceno. También revela similitudes entre los Canthonini de los bosques tropicales secos de México, Costa Rica y Colombia, con tres patrones de distribución que corresponden a diferentes grados de expansión hacia el norte y a la diversificación de líneas evolutivas, e incluso la presencia de especies sudamericanas. La comparación entre el cladograma generado por las especies de Canthonini y el de eventos geológicos, indica que la distribución de los primeros en los bosques tropicales secos comienza en el Plioceno con el restablecimiento de la conexión panameña, sin evidencias de que eventos geomorfológicos previos hayan ejercido influencias. Por otra parte, hay muy pocas especies 9 compartidas con la selva siempreverde usadas como grupo externo en los cladogramas. Palabras clave: Scarabaeinae, Canthonini, bosque tropical seco, México, Costa Rica, Colombia, patrones biogeográficos. 10 INTRODUCTION It has been suggested that tribe Canthonini (Scarabaeidae: Scarabaeinae) has an ancient Gondwanian distribution with high species richness in South America (Halffter 1974). In the Americas, the genus with the most species is Canthon, with 174 according to Halffter & Martínez (1977) that are distributed from Argentina to Canada. The origin of Canthon and close genera appears to be northern South America where it reaches its greatest richness in phyletic lines and species, with an expansion of lines toward the periphery of Chile where it is barely represented (Halffter 1974, Rivera-Cervantes & Halffter 1999). From South America, Central and North America were populated by elements of this tribe during two probable great expansion events (Halffter 1964, 1974, 1976, Kohlmann & Halffter 1990). The first could have occurred before or during the Miocene and the second when the connection with South and Central America was reestablished from the Pliocene to the Recent. In the phyletic lines that participated in the first expansion, although it is possible to identify the South American affinities at the level of genus, there was a notable speciation in Mexico and the United States of America, followed in some cases by secondary expansion. For those phyletic lines that participated in the second expansion event, the affinities with northern South America are much more marked, although there was also in situ speciation. The objective of this study is to test this biogeographical hypothesis. To that end, we chose to use the species of Canthonini that inhabit dry tropical forest (DTF). These were selected because they are the least studied Scarabaeinae fauna compared to their tropical rain forest (TRF) counterparts and also because their distribution, now discontinuous, has been much less interrupted in the recent past. This hypothesis regarding the biogeographical history of the Canthonini (common for other groups of Scarabaeinae and, in general, for insects) has been tested in various ways. First, by establishing the similarity of the Canthonini present in a series of enclaves found in DTF in Mexico, Costa Rica and Colombia – places for which we have reliable lists of the fauna. Second, we have analyzed the historical and geological relationships of the regions where these enclaves are located in order to relate the history of the areas with both the phylogeny and the geographical distribution of Canthonini. Finally, the two previous points are integrated in order to recreate the setting of the DTF where the processes of speciation and dispersion of some Canthonini probably took place, and also to explain their geographic distribution. As proposed by Zunino (2005) we 11 intend to compare the distribution and phylogenetic history of the taxa with the geography and geomorphological history of the areas to arrive at a comprehensive interpretation of the entire set of elements. DTF, also known as tropical deciduous forest or low deciduous forest is defined as that formation of vegetation with a continuous woody cover distributed from sea level to 1000 m, with mean temperatures above 24 ºC and precipitation ranging from 700 to 2000 mm per year (Espinal 1985; IAvH 1997a). This vegetation has woody species that lose their leaves during the dry season over a variable period of time that lasts around six months (Rzedowski 1978). According to Rzedowski (1978) DTF is particularly characteristic of the Pacific Slope of Mexico, where it covers extensive areas almost uninterrupted from the south of Sonora and southwest of Chihuahua to Chiapas, continuing southwards in Central America. It penetrates deep into the Balsas and the Santiago River basins, as well as those of the tributaries of these rivers. In the Isthmus of Tehuantepec, DTF passes the watershed and occupies a good part of the Central Depression of Chiapas. On the Atlantic Slope there are at least three large patches of DTF and it is also found on the Yucatan Peninsula (see Trejo 2005). In Mexico, DTF once covered 6 to 14% (270 000 km2) of the area in the country that lies between sea level and 1500 m. DTF has been reduced to 27% of its original cover (Trejo & Dirzo 2000). In Costa Rica almost all of the Nicoya Peninsula and Chira Island in the northeast of the country were covered by this type of forest 100 years ago (Kohlmann et al. 2002). In Central America, DTF once covered 7% (33 600 km2) of the total area between 0 and 1000 m. The remaining tropical dry forest in Central America probably represents less than 2% of the original (Sabogal & Valerio 1998). Colombia has three large regions with DTF and the two largest are the Caribbean Plain, including southern Guajira and the Magdalena River Valley in the Departments of Tolima, Cundinamarca and Huila (IAvH 1997b). In Colombia the status of DTF is critical. It is estimated that only 1% of the original 80 000 km2 of dry to subhumid forests remains (Etter 1993). Figure 1 summarizes the distribution of DTF in the study area. The biogeography of the areas of dry tropical forest is undertaken with cladistic analysis, keeping in mind the past and present geological and geomorphological characteristics of the area. In order to understand the biogeographical relationships between the enclaves of DTF and the species of Canthonini, we used a Parsimony Analysis of Endemicity (PAE) that produces a hierarchical set of the species represented in the different areas, and associates it with the geological or ecological 12 factors, or a combination thereof. PAE can provide the grounds for explaining: 1) the effects of geological events on evolution; 2) the effects of ecological factors on evolution; 3) the influence of geological events on ecological conditions and the evolutionary consequences of these; 4) distribution patterns, i.e. which sets of species can appear in different areas owing solely to events in geological history or the association of the latter with ecological conditions (Rosen 1988). This study has two purposes, the first is to analyze the biogeographical history of the areas of dry tropical forest in Mesoamerica and Colombia based on their geomorphological characteristics and their geological origin; the second is to establish the relationships between these forests using the Canthonini as an indicator group, comparing their distribution with the geomorphological history of those areas and explaining the presence of Canthonini in dry forest. First we carry out the biogeographical analysis of the Neotropical dry forests that are the object of this study, and then the biogeographical analysis of the Canthonini. Some aspects of Canthonini in dry forest are discussed in light of other floristic, faunal, ecological and biogeographical elements that are particularly related to the Canthonini and the geographical limits of some species. MATERIALS AND METHODS Study sites The DTF sites we selected in Mesoamerica and Colombia are enclaves representative of the current distribution of this type of vegetation (Fig. 1). Chamela (Jalisco, Mexico) and Palo Verde (Guanacaste, Costa Rica) are located on the Pacific coast. In the Caribbean region of Colombia we selected four areas: Neguanje, Tierra Bomba, Los Colorados and Zambrano. In the Magdalena River Valley, we selected the DTF located in the north of the Department of Tolima. Two of the sites selected in the Caribbean region, Los Colorados and Neguanje, have the highest floristic richness, basal area and canopy height of the DTF found in the Caribbean and Tolima region (Mendoza 1999). Another factor which led to the selection of these sites was the availability of information on the Scarabaeinae fauna. There are publications for each of the sites, and there are also databases for Costa Rica (INBio) and Mexico (SNIB-CONABIO). The sites are described below together with the most relevant features associated to geological and geomorphologic characters (see Tables 1 & 2). 13 Chamela, Jalisco, Mexico. The Chamela-Cuixmala Biosphere Reserve is located on the Pacific coast (19º 30’ N, 105º 03’ W). The study site is between the San Nicolas River in the north and the Cuixmala River in the south, centred in the surroundings of the Chamela Biological Station (Fig. 1). With altitudes lower than 200 m, the climate has no marked seasonality with respect to temperature. Mean monthly maxima range from 28.8 to 32.2 ºC, and the minima from 15.9 to 22.6 ºC. Mean annual precipitation is 707 mm. The rainy season lasts four months on average, starting at the beginning of July and ending at the beginning of November (Bullock 1988). Chamela belongs to the Jalisco Block (JB), which constitutes a tectono-stratigraphic assemblage from the Late Cretaceous to Early Tertiary (Paleocene), with volcanic and volcaniclastic deposits and marine sedimentary sequences intruded by granitoid plutons. The plutonic and volcanic rocks of the JB are part of the magmatic arc, which is found southeastward along the terrain of the state of Guerrero. On the other hand, Chamela is aseismic and is located on the North American tectonic plate. It is characterized by an undulating landscape, the presence of metals, soil with low permeability and organic matter Palo Verde, Guanacaste, Costa Rica. The Palo Verde Biological Station (Fig. 1) is located in the Palo Verde National Park (10º 20’ N, 85º 18’ W) on the Pacific slope, in the Province of Guanacaste and halfway up the basin of the Tempisque River at an altitude of 10 to 50 m. Mean annual temperature ranges from 24.0 to 27.8 ºC and mean annual precipitation is 1750 mm (Quigley & Platt 2003), with a 6.5 month long dry season each year (SIEPAC 2003). It is characterized by an undulating landscape, with slightly developed soils, and the most abundant minerals are olivine-pyroxene. The local fault is responsible for the seismicity. The Palo Verde region dates from the Paleocene to Early Eocene. This site is characterized by high magnetism, and is located on the Caribbean tectonic plate Colombia, Caribbean Region. The sites selected (Fig. 1) have been described by Mendoza (1999). Located between 50 and 300 m, their mean temperatures are greater than 24 ºC and they receive 700 to 2000 mm of precipitation per year. There are two marked dry seasons per year (IAvH 2000). This region is represented by four sites: a) Zambrano: Forest Reserve, Monterrey. Located in the Departament of Bolivar, Zambrano Municipality (9º 37’ 48” N, 74º 54’ 44” W) at 155 m. Mean annual precipitation is 1048 mm. b) Los Colorados. Located in the Departament of Bolivar, San Juan de Nepomuceno Municipality (9º 51’ 33” N, 75º 06’ 38”W), at 300 m. Mean annual precipitation is 1189 mm. 14 This remnant of DTF belongs to the Los Colorados Flora and Fauna Sanctuary Conservation Unit. c) Tierra Bomba Island. Located in the Department of Bolivar, Cartagena Municipality (10º 21’ 36” N, 75º 34’ 11” W), at 50 m. Mean annual precipitation is 789 mm. d) Neguanje. Located in the Department of Magdalena, in the Santa Marta Municipality (11º 18’05” N, 74º 06’ 11” W) at 300 m. Mean annual precipitation is 1420 mm. Neguanje belongs to the Tayrona National Park. The Caribbean Region is located between the Perijá to the east and the Sierra Nevada. On the surface, fluvial and lacustrine sediments from the Quaternary predominate. It is characterized by a sequence of sandy, slime and clay and feldspar is the most abundant mineral in the sandy fraction of the soil. It is aseismic and located on the Caribbean tectonic plate. In geological terms, Neguanje belongs to the geotectonic province de Santa Marta (land originating in the Mesozoic: Middle Jurassic to Late Cretaceous) and the other three sites belong to the province of Caribbean Plain: Zambrano belongs to the San Jorge-Plato geostructure (the land dates from the Quaternary), Los Colorados to the San Jacinto Belt (the land dates from the Tertiary: Late Eocene) and Cartagena to the Sinú Belt (Tierra Bomba is from the Quaternary). The Sierra Nevada is associated with the Santa Marta fault and the Caribbean Plain is associated with the Romeral fault system in the west. North Tolima. We selected the study area described by Escobar (1997). It is located on the east side bank of the Magdalena River (Fig. 1), 130 km from the city of Ibagué in the jurisdiction of the municipalities of Honda, Armero-Guayabal and Piedras, at (4º 15’ - 5º 10’ N, 74º 45’ - 74º 50’ W) at 250 m. Mean annual precipitation is 1387 mm and the mean annual temperature is 28 ºC, with two well defined dry periods, one from December to March and the other from June to August. North Tolima forms part of the Upper Magdalena River Valley, an inter-Andina valley on the eastern slope of the Cordillera Central. It belongs to the Honda formation dating from the Tertiary: Miocene. There are faults nearby and it is associated with the Plio-quaternary volcanic activity of the Cordillera Central, as reflected in its notable seismicity. Los Tuxtlas, Veracruz, Mexico. The Los Tuxtlas Biological Station is located in the foothills of the San Martin Volcano, 19 km north of the city of Catemaco in the state of Veracruz (Fig. 1; 18º 34’ - 18º 36’ N, 95º 04’ - 95º 09’W) at 150 to 530 m. Mean annual precipitation is 4560 mm and 15 the mean annual temperature is 23.7 ºC (Morón 1979). Los Tuxtlas Reserve belongs to the formation La Laja, which dated Oligocene and is located on the North American tectonic plate. It is characterized by an undulating landscape, igneous rocks of continental origin, nearby mountains, soils with high permeability and organic material, and high seismicity. Leticia, Amazonas, Colombia. The Colombian Amazon has a humid tropical climate. The site is located at 4º 8’ S, 70º 1’ W, and 96 m (Fig. 1). Mean annual precipitation is 3500 mm, with the majority of the rain falling between April and June. Mean annual temperature is 26 ºC, with maximum temperatures in October and November 35 ºC. It is cooler in July with minima from 20 to 25 ºC (Galvis et al. 1979). It is characterized by an undulating landscape, with highly evolved soils, a limited presence of minerals and low levels of natural fertility. Quartz dominated more than 90% of the sandy fraction and potassium was scarce. This geomorph developed over sedimentary rocks with thick grain. The Quaternary deposits are comprised of sand, possibly of eolic origin, with recent terraces and alluvial deposits. This is an aseismic region located on the South America plate. Ingroup and Outgroup Our decision to use a real outgroup is based on the fact that the rain forest provides ecological characteristics that contrast with those of dry forest and take into account previous hypotheses about its probable origin in Neotropical forest. Also, rain forest is supposed to be older than dry forest (see Sarmiento 1975, Platt et al. 1981, Gentry 1982, Hooghiemstra & van der Hammen 2001, Richardson et al. 2001, Hooghiemstra et al. 2002, Graham 2003). The use of an outgroup provides both a geographical context and vegetation formations that contrast with the characteristics of the areas, where Scarabaeinae we are studying are currently found. Geomorphological cladogram Historical and geomorphological relationships were analyzed following Craw (1988). The seven dry forests make up the ingroup, however our analysis differs because for the outgroup we have used real sites of tropical rain forest, instead of coding for the outgroup with zeros. The rain forests selected carrying out the analysis using this outgroup allows the cladogram to be polarized and hence produces a more parsimonious solution than one that uses zeros. In order to select characters and assign character states for each of the sites, more than fifty references were consulted. The list of the main studies consulted is presented by country in table 16 1. The geologic-geomorphological characters provide evidence from different sources of characters: geomorphostructural, geophysical and of edaphic. The geomorphostructural characteristics of the terrain are defined by its geological and morphostructural history (origin and evolution) and by the lithological composition of the materials. Geophysical characters include the geotectonic, magnetic and gravimetric anomalies of the physiographical regions. On the other hand, the origin of the soil is influenced by the relative material, relief, climate and the organisms that are present. Eighteen characters and their states were selected (Tables 2 & 3). Ambiguous characters were discarded as were those for which there is insufficient information for more than two sites. All the characters were given equal weight, with three ordered characters: 8, 9 and 16. These characters were taken as ordered because they are related to chronological sequences of events associated to the scale of geological time. Cladogram of the areas of affinity for the Canthonini The relationships of DTF were analyzed using Parsimony Analysis of Endemicity (PAE), where the taxa are analogous to the sites and the characters to the species. In contrast to the methodology proposed by Rosen (1988) where the cladogram is rooted in a hypothetical ancestral area coded in zeros, here we use an external group, the outgroup, of real sites. The ingroup and outgroup are the same as mentioned for the previous analysis. We chose PAE because it allows for the use of an outgroup in the analysis, the polarization of the cladogram, the addition of taxonomic hierarchies and it has a small data matrix. A cladogram of the areas was generated using lists of the species (Table 4) present in each forest, indicating presence (1) or absence (0) according to the matrix (Table 5). Given that 24 of the 28 species belong to the genus Canthon, the subgenera of this taxon were included. The PAUP 4.0b10 (Swofford 2002) program with Acctran optimization and exhaustive search was used. The results were analyzed based on the majority consensus tree as it offered the greatest resolution. Canthonini Species All the genera of Canthonini present in the forests studied are included in the analysis, with the exception of Deltochilum for which there is insufficient taxonomic and biogeographical information. In order to compile the lists for Palo Verde and the Mexican site, in addition to a review of the publications mentioned below, the following data bases were used: INBio (Costa Rica) and 17 SNIB-CONABIO (Mexico). The following publications were consulted: Chamela - Morón et al. (1988), Andresen (2005); Palo Verde - Kohlmann & Wilkinson (2003); North Tolima - Escobar (1997), Bustos-Gómez & Lopera-Toro (2003); Caribbean Region of Colombia - Escobar (1998, 2000a); species lists for these Colombian sites were also provided by F. Escobar; Leticia – list compiled by Bruce Gill, 12, I, 1997 with material collected by Howden & Nealis (1975), Escobar (2000a), Medina et al. (2001); Los Tuxtlas - Morón (1979), Halffter et al. (1992), Favila & Díaz (1997), Deloya & Morón (1998), Díaz (1998, 2003). Canthon deyrollei was described by Harold in 1868, with the type locality unknown. It has been included for Colombia, without any other data in numerous publications (Vulcano & Pereira 1964, Howden & Young 1981, Solis & Kohlmann 2002). In the publications that we consulted for Colombia this species is either not recorded or is not assigned a locality. Hence, the record for Colombia was not included. Canthon cyanellus and C. indigaceus are included in the analysis with their respective subspecies according to Halffter (1961). Geographical distribution of Canthonini There are many studies that include the geographical distribution of Canthonini in Mexico. Some of those that we consulted for this study are: Halffter 1961, 1964, 1976; Martínez et al.1964; Halffter & Matthews 1966; Barrera 1969; Martínez & Halffter 1972; Morón & Terrón 1984; Morón et al. 1985, 1986; Deloya et al. 1987; Delgado 1989; Kohlmann & Halffter 1990; Palacios-Ríos et al. 1990; Arellano 1992, 2002; Deloya 1992; Capistrán 1992; Estrada et al. 1993; Deloya & Morón 1994; García-Real 1995; Halffter et al. 1995. The most recent and relevant publications on Canthon in Mexico are Rivera-Cervantes & Halffter (1999) and Halffter (1961, 2003). Solis & Kohlmann (2002) review Canthon for Costa Rica, as do Howden & Young (1981) for Panama. In Colombia there are several local revisions (Amézquita et al.1999, Escobar 2000b, Escobar & Chacón de Ulloa 2000, Neita et al. 2003, Pulido et al. 2003). The catalogue prepared by Medina et al. (2001) lists the species for Colombia. Information about the geographic distribution of Canthonini in America is given in Bates (1886-1890), Blackwelder (1944), Vulcano & Pereira (1964), Halffter & Martínez (1966, 1977), as well as in the previously cited articles. 18 RESULTS Geomorphological cladogram The 50% majority consensus tree of four equally parsimonious trees is shown in Figure 2. The cladogram has 43 steps, a consistency index (CI) of 0.60 and a retention index (RI) of 0.56. The Colombia and Palo Verde (Costa Rica) sites are sister groups and these two, in turn are sister group to Chamela (Mexico). Los Tuxtlas (Mexico) represents the first clade banching out, being the sister group to the other sites included in the analysis. There are two well supported clades, the first includes Leticia, Tierra Bomba and Zambrano which have both the era and geological period in common: Quaternary: Pleistocene, Holocene. They are also free of tectonic faults. The second well supported clade grouped all the DTF of Colombia and Leticia. All of the sites in Colombia (except Neguanje) have sedimentary rock, though it originates from different processes. In Leticia, the rock results from dendritic accumulations, terraces and alluvial deposits from the margins of the Amazon River. At Tierra Bomba, it originates in marine deposits. In Zambrano, it results from the fluvial-deltaic action of the Magdalena River, at Los Colorados it comes from rocks corresponding to carbonate facies and in the Upper Magdalena River Valley, it is comprised of carbonacious and clay sediments that are both sandy and conglomerate. The soils of the sites are highly developed and have low concentrations of organic material. Palo Verde and the majority of the sites in Colombia (except Leticia) are characterized by high magnetism and are located on the Caribbean tectonic plate (except Leticia and N. Tolima). Chamela is a sister group to Palo Verde and Colombia, where the soils have optimum concentrations of potassium (except Leticia). On the other hand the forests studied in the Caribbean Region (except Neguanje) and Leticia (the Amazon) are on sites of very recent geological origin (Quaternary) compared to the DTF of North Tolima (C), Costa Rica (Palo Verde) and Mexico (Los Tuxtlas and Chamela) which are of Tertiary origin. The assemblage of the Caribbean Region, Colombia The areas where Los Colorados and Neguanje are found, are part of a mosaic of blocks that later came together to form the Caribbean Region. Los Colorados belongs to the San Jacinto Fold Belt which is the result of the interaction between the oceanic crust of the southwestern Caribbean and the continental crust of northern South America. Forces of tension and compression alternated 19 along the platform’s margin, especially during the pre-Andina period of orogeny (Middle Eocene) which lifted, folded and shaped this belt. The geomorphological features of the Santa Marta Massif (from which the Neguanje is derived) are the result of its location during the Mesozoic-Tertiary at the intersection of faults on the northeastern corner of South America (González et al. 1988). According to Ujueta (2003) this Massif should be considered a northern extension of the Cordillera Central and its relief is a result of mostly vertical tectonic movement accompanied by moderate horizontal movement. The tectonic history of San Jacinto, the location of the Sierra Nevada of Santa Marta and the sedimentation of the Plato and San Jorge basins, is correlated in space and time from the Late Miocene to the Pliocene (Caro & Spratt 2003). The assemblage and the sedimentation of the Caribbean Region of Colombia ended in the Quaternary. It is likely that the climatic and ecological changes were shared by all the sites of the Caribbean Coast of Colombia and so reflect in one way or another the affinities of the biota present in the DTF. The Caribbean Region, together with the north of Venezuela make up the phytogeographic region of Northern South America (Gentry 1995). Geomorphological relationships of the study sites and their relationship to the origin of DTF Three events have exerted a great effect on the evolution of Neotropical flora: the rise of the Andes, the exchange of biota with North America after the formation of the Central American Isthmus and the climatic fluctuations of the Pleistocene (Guariguata & Katan 2002). This, in addition to the culmination of the majority of orogenic processes that occurred on the continental land of Mexico toward the end of the Pliocene (see Challenger 1998) and the sedimentation of some parts of the mid- and lower Balsas River basin in the Pleistocene. The convergence of the DTF studied can be attributed to the climatic and ecological events of recent geological time, probably since the Pleistocene. This confirms the observations of Gentry (1982) who suggests that the high number of species in the dry tropical area of Mexico is probably the result of an active evolutionary diversification in response to the increase in dry climate regimes during the Pliocene and the Pleistocene. This is in accordance with the higher rate of speciation of the genus Bursera (toward the Pliocene) and its concentration in areas of intermediate and low altitudes of the Pacific Slope of Mexico, and in particular the Balsas River Basin (see Becerra 2005, Rzedowski et al. 2005). 20 One can think about a general process of south-north expansion since, as Rzedowski (1978) has said, the dry tropical forests of Mexico are characterized by a strong predominance of Neotropical elements, and the scarcity or absence of holarctic elements. Although dry forests have their origin in the Pleistocene, it is likely that their current, dry conditions were set during the Middle Holocene in the Caribbean Region of Colombia, as well as on the Pacific coast of Mexico and Costa Rica. During the Middle Holocene, about 5000 to 7000 years B.P. (Steig 1999), the dominant climate system was ITCZ (Intertropical Convergence Zone), given that the ENSO (El Niño-Southern Oscillation) was absent or weak (Fontugne et al. 1999, Cole 2001, Tudhope et al. 2001, Riedinger et al. 2002). Haugh et al. (2001) propose that the driest climatic phase of the Holocene results from the change in the ITCZ to its current position. Cladogram of the areas of affinity of Canthonini The cladogram of the areas of affinity of the Canthonini, with 28 taxa and three subgenera of Canthon (Figure 3) is the result of the majority consensus of three trees, with 40 steps, and CI of 0.77 and an RI of 0.64. The cladogram groups the Colombian DTF as follows: the first group is comprised of the forests of Tierra Bomba, Zambrano and Neguanje (Caribbean Region); the second is North Tolima alone, and the third group is comprised of Los Colorados, also of the Caribbean Region. In the first group the subgenus Glaphyrocanthon is not represented by a single species; in the other two groups this subgenus is represented by C. (Gl.) subhyalinus. The Canthonini of the dry tropical forest of Chamela and Palo Verde are related to those of the forests of Colombia. Those of Palo Verde show an affinity to both those of Chamela and those of Colombia. The canthonines of the two tropical rain forests are totally different from each other and do not share any species. There are three species that characterize the relationships between the dry forests of Colombia: Canthon juvencus, C. cyanellus sallei and C. lituratus (except in Los Colorados). The Amazon (Leticia) only shares C. aequinoctialis with the dry forests of North Tolima and Los Colorados Caribbean Region, Colombia). Chamela has the most exclusive species (Canthon corporali, C. humectus and C. pacificus), and is followed by Palo Verde (Costa Rica): Canthon deyrollei and C. meridionalis, and North Tolima (Colombia) with C. acutus. Eighty-six percent of the species included in the analysis belong to the genus Canthon. To 21 explain these results and explore why such species are found in the forests studied, their distribution patterns are analyzed below (see Table 4 and Appendix 1). Analysis of the distribution of Canthonini in the dry forests studied The cladogram of the species of Canthon for North America (including Mexico) proposed by Kohlmann & Halffter (1990) shows the spread of this genus from South America in two big expansion events: the first towards the Miocene and the second towards the Plio-Pleistocene and continuing to the Recent (see also Halffter 1964, 1976). The phyletic lines that correspond to the first expansion were not influenced by the events that affected the integration and distribution of the dry forests which came afterwards. Their presence in these forests (Chamela) are a result of later colonization. In contrast, the species of the second expansion exhibit a correspondence to the historical and biogeographical conditions that influenced the distribution of DTF: orography that was broadly similar to the current situation, the reestablishment of the connection and exchange of biota between South and North America, and an expansion of the dry conditions in the Holocene. In order to explain the presence of Canthonini species in the enclaves of DTF we have studied both the distribution patterns proposed by Halffter (see Halffter 1964, 1976; Kohlmann & Halffter 1990) and the phyletic lines proposed for taxa of the genus Canthon according to the taxonomic affinities recognized by Halffter & Martínez (1977), and propose a pattern of speciation in situ. Pattern 1 This pattern is observed for species belonging to the phyletic lines that expanded through Mexico, most likely before the Miocene (the High Plains Distribution Pattern of Halffter). These were the earliest species of Neotropical affinity to penetrate the Mexican Transition Zone (MTZ) (Halffter 1976, 1978). Canthon (C.) humectus belongs to this pattern, and represents a phyletically isolated line within the subgenus Canthon according to Halffter & Martínez (1977). Of the eight subspecies (Halffter & Halffter 2003) distributed throughout the High Plains of Mexico, Oaxaca, Chiapas and Guatemala, only C. humectus riverai Halffter & Halffter is found in Chamela, a new record not indicated by Andresen (2005), perhaps because it is outside of the area she studied. This species occurs at the lowest altitude for both the species and Canthonini found on the High Plains. It was previously collected in the lower part of the Manantlán Mountain Range (Jalisco) in DTF at altitudes from 700 to 1000 m. Its presence in Chamela can 22 be considered a result of secondary expansion. In a study of the entire Scarabaeoidea superfamily in Chamela, Morón et al. (1988) found some species with the ancient MTZ dispersal patterns, though the majority have the Neotropical Pattern that developed from the Pliocene onwards (see below). Pattern 2 These species participated in the expansion that probably began in the Pliocene with the reestablishment of the Panamanian connection, and have different degrees of northwards expansion in North and Central America, to the United States of America (and thus together result in the Typical Neotropical Distribution Pattern of Halffter). The affinities of these species with the South American fauna are much better defined than is the case for Pattern 1 (see affinities under each species). The penetration of these species into the MTZ (Pliocene onwards) coul have occurred in distinct stages, depending on the species’ antiquity. The most ancient species have undergone processes of subspeciation in the MTZ. We will examine the distribution of these species, from the oldest (and with the widest distribution) to the most recent. a) Species with a wide distribution that reaches the United States of America. C. (C.) indigaceus LeConte ranges from Panama to Texas and Arizona in the United States. In Mexico it is found in tropical landscapes on the Gulf slope, the Pacific slope and the Yucatan Peninsula. It is comprised of three subspecies that are found under different ecological conditions: Canthon i. chiapas in TRF; C. i. chevrolati in places with greater insolation, with less or no woody vegetation; and C. i. indigaceus in DTF or more xerophyllous vegetation. In Chamela we found C. i. indigaceus; in Guanacaste, C. i. chevrolati; and in the TRF of Los Tuxtlas, C. i. chiapas. Overall, C. i. indigaceus is a good example of a species exhibiting the Typical Neotropical Pattern with a wide distribution in the MTZ. As in the previous case (Canthon humectus), this species constitutes a phyletic line within the subgenus Canthon (Halffter & Martínez 1977) that has only one species. C. (C.) cyanellus LeConte has a similar geographic history and distribution, although the latter is broader. It is found from Peru and Brazil to the southern United States and under a wide variety of ecological conditions though it is usually associated with tropical forests or treed sites along an altitudinal range from sea level to 2600 m. This species is found at all the DTF sites we studied and in Los Tuxtlas, but not in Leticia. In Mexico it is found all along the Pacific coast, from Chiapas to Jalisco and on the Atlantic slope 23 it is found on the Yucatan Peninsula and along the Gulf Coast of Mexico up to Tamaulipas and Nuevo León. In the interior of Mexico it is found in the states of Puebla, Morelos, Hidalgo and San Luis Potosí. In Central America, it is found on both the Pacific and the Atlantic slopes; in northern Panama it is found near the Caribbean Sea (Bocas del Toro) and to the south on the border with Colombia in the Darien region on the Pacific coast. In Colombia, according to our findings this species is found throughout the Caribbean Region and the high valley of the Magdalena River, as well as in the Department of Meta (Amézquita et al. 1999, Medina et al. 2001). In Venezuela this species occurs in Táriba, San Cristóbal, as well as in Arima, Trinidad and in Uaupés, Brazil (the Amazon) according to the records of Solís & Kohlmann (2002). Canthon cyanellus belongs to a phyletic line (the bispinus line) with many species in South America (Halffter & Martínez 1977). Malagoniella (M.) astyanax (Harold). The genus Malagoniella is distributed throughout the Neotropical region, except in Chile and the Antilles. It is richest in species in Argentina, Uruguay, Paraguay and southern Brasil (Halffter & Martínez 1966). The subgenus Malagoniella has a similar distribution. It includes the only taxon found in the north, M. (M.) astyanax yucateca (Harold), that is found in Central America, Neotropical areas of Mexico and at a point on the border of the United States of America. The other subspecies that make up M. astyanax are South American. Citations for Colombia possibly refer to M. (M.) astyanax columbica (Harold), a subspecies for which the distristubion is limited to Colombia (Halffter & Martínez 1966), where it is found in the DTF of Zambrano and North Tolima, as well as in the Departments of Chocó and Magdalena (Medina et al. 2001). This species appears to have a preference for DTF and has not been recorded in Leticia (the Amazon). M. (M.) astyanax yucateca (Harold) has a very wide distribution in Mexico and Central America, but has been found at isolated points that are often separated by great distances. In this case, it is difficult to attribute this to capture inefficiency (at least for many of the places where it has not been found), given that this is a large and very striking species. It has been collected in Brownsville, Texas, on the border with Mexico, and also at Tamazunchale (San Luis Potosí), Quintana Roo, Escárcega (Campeche), Puerto Ángel (Oaxaca), Cacahoatan, Tapachula and Rosario Izapa (Chiapas), Yucatan (Bates 1886-1890; Halffter & Martínez 1966; SNIBCONABIO database), Guatemala, in the west of Nicaragua and Costa Rica. The Chiapas and Central American sites are on the Pacific slope. In Costa Rica it is restricted to the Guanacaste 24 region, to the northwest of the Pacific coast, over an altitudinal range of 10 to 400 m. In the present study, it is recorded for Palo Verde. Except for the capture in Brownsville (Texas), the rest were captured in tropical forest. M. astyanax yucateca appears to be most often associated with seasonal forests (i.e. those that have a dry season). It has not been found in TRF (Los Tuxtlas, Veracruz or Selva Lacandona, Chiapas) within its distribution perimeter. The distribution of M. astyanax appears to reflect an ancient expansion from South America, one that is currently quite fragmented. Pseudocanthon perplexus (LeConte). The genus Pseudocanthon Bates is comprised of eight species, and two of these are found in Mexico. Of the Mexican species, one is found in the Antilles where there are also five endemic species. The other species is South American. Pseudocanthon perplexus is found in Mexico in both TRF and DTF. It is captured sporadically in sites that are distant from each other. Its distribution extends from the United States of America down to South America. It was found in Chamela and Palo Verde of the DTF we studied. In Mexico it is distributed along the coast of the Gulf of Mexico up to the eastern United States, and on the Pacific coast to Sonora. In Costa Rica, it has been recorded in Guanacaste Province and in Alajuela on the Cordillera Central, over an altitudinal range of 10 to 540 m (INBio database). In Panama, it has been found in Balboa in the Canal Zone (Pacific coast). b) Species with limited penetration into Mexico. Canthon (C.) morsei Howden is distributed from Ecuador to the warm areas of Mexico, where it extends along the Pacific slope from Chiapas (Halffter et al. 1992) to Jalisco (García-Real 1995). Along the Atlantic slope it ranges northwards up to Tamaulipas. In Costa Rica it is found in the northeast on the Pacific coast (Solis & Kohlmann 2002). In Colombia, it is found in the Upper Magdalena River Valley (Tolima). This species is associated with DTF and TRF. In the DTF studied, it was recorded in Palo Verde and in Colombia only in North Tolima. In Mexico, it was not found in the DTF of Chamela, although it is has been recorded in deciduous and semideciduous tropical forests in the state of Jalisco relatively close to Chamela. According to Halffter & Martínez (1977) C. morsei belongs to a phyletic line of the subgenus Canthon that includes other South American species. C. (Gl.) subhyalinus Harold is distributed from Bolivia to Mexico and is associated with TRF, especially those sites with monkeys (Halffter 1991, Estrada et al, 1993, Rivera-Cervantes & Halffter, 1999). This species was not found in the DTF of Chamela or Costa Rica. In Colombia it 25 is associated with DTF in Tolima and in Los Colorados (Bolívar); other records for Colombia include the Departments of Antioquia and Cundinamarca (Medina et al. 2001). In Panama, Howden & Young (1981) have recorded it for Darien, Santa Fe, and on the Pacific coast and in the Caribbean on Barro Colorado and in the Canal Zone. In Costa Rica it is found in the foothills of premontane wet forest. In Mexico it is only found in the south in the states of Chiapas, Quintana Roo and Veracruz. Canthon subhyalinus is a beetle associated with TRF (RiveraCervantes & Halffter 1999), and so it is probable that the neighbouring vegetation between TRF and DTF facilitates the dispersal of this species toward DTF. This explains the presence of C. subhyalinus in Los Colorados, where the DTF includes a patch of TRF, a lone remnant in the Caribbean Region of Colombia (Gentry 1995). It is also found in the DTF of North Tolima, which is between two mountain ranges and in the foothills of the mountain where wet montane forest predominates. The northern distribution limit for C. subhyalinus coincides with that of the howler monkey, Alouatta palliata. It is probable that the dispersal of these species coincide from South America up to Los Tuxtlas. This is supported by field observations of the arboreal behavior of these beetles and their attraction to the dung of these monkeys (Howden & Young 1981, Halffter 1991, Estrada et al. 1993, Rivera-Cervantes & Halffter 1999). There is additional support in the origin of these monkeys, Late Miocene to the Pliocene, and their dispersal during the Pleistocene (Cortés-Ortiz et al. 2003), as well as the biogeographic pattern proposed in this study for C. subhyalinus. C. (Gl.) euryscelis Bates belongs to the same group of species of the subgenus Glaphyrocanthon as C. (Gl.) subhyalinus (see Rivera-Cervantes & Halffter 1999). It is distributed in TRF from Mexico to Panama. In the DTF we studied, it was collected at Palo Verde (Guanacaste). Solis & Kohlmann (2002) indicate that it has a broad tolerance of climate, and is found from sites with marked seasonality (altitude < 800 m in the Province of Guanacaste) to very humid ones. In Mexico it has been recorded in the TRF of Los Tuxtlas. Agamopus lampros Bates belongs to a genus (currently under revision by Fernando Vaz de Mello) with four South American species, of which A. lampros extends to Central America and the tropical areas of Mexico. We recorded this species in the DTF of Chamela, Palo Verde (Guanacaste) and Los Colorados (Colombia). 26 c) The species mentioned below belong to South American phyletic lines whose northern geographic distribution only extends to Central America. Canthon (C.) mutabilis Harold, within the subgenus Canthon, belongs to the bispinus line that, as mentioned in the section on C. cyanellus, is comprised of many South American species. Its distribution extends from Argentina to Costa Rica. It was found in Palo Verde (Guanacaste) and in North Tolima (Colombia). Solis & Kohlmann (2002) indicate that in Costa Rica it is found from sites that are very humid to those with marked seasonality and a dry season as long as six months. In Colombia, Medina et al. (2001) have recorded it for the Departments of Bolívar and Meta, associated with dry conditions. Canthon juvencus, C. mutabilis and C. meridionalis are abundant in the driest forests of some of the smallest islands in Lake Gatun and Gamboa (Caribbean region) near the continent (Gill 1991). Canthon (C.) lituratus (Germar) is also widely dispersed from Argentina to Costa Rica. It belongs to a phyletic line within the subgenus Canthon that includes another South American species (Halffter & Martínez 1977). In Colombia it is found in Tolima and in the Caribbean Region in all the DTF studied, except for Los Colorados. In Colombia, Medina et al. (2001) have recorded it for the Departament of Valle. In Costa Rica it is only found to the southwest of the Pacific coast, in the savannahs bordering the Térraba River (Solis & Kohlmann 2002). Its distribution extends southwards from the Pacific, and it has also been recorded for the provinces of Chiriqui and Cocle in Panama at altitudes above 800 m (Howden & Young 1981). Canthon (C.) juvencus Harold is found from Brazil to Costa Rica. In Colombia it is found in the DTF of the Caribbean Region and in the Upper Magdalena River Valley (Tolima) as well as in other Departments in Colombia: Guainia, Guaviare, and Meta (Amézquita et al. 1999, Medina et al. 2001, Escobar 2000 b). In Costa Rica it is restricted to the southern Pacific in TRF located below 500 m (Solis & Kohlmann 2002). It is found in Panama in the pronvinces of Panamá, Colón and the Canal zone (Howden & Young 1981). Canthon septemmaculatus (Latreille) has a broad distribution range from Argentina to Costa Rica. In the DTF studied, it was only recorded in Zambrano Caribbean Region, Colombia). Medina et al. (2001) have recorded it for the Colombian Departments of Bolívar, Chocó, Meta, Caquetá and Nariño. In Panama it has been recorded in the provinces of Los Santos, Coclé, Panamá and the Canal Zone. In Costa Rica it is only found to the southwest of the Pacific coast, in the savannas surrounding the Térraba River (Solis & Kohlmann 2002). 27 Canthon (C.) aequinoctialis Harold occurs from Brazil to Belize. According to Halffter & Martínez (1977) it belongs to a South American phyletic line of the genus Canthon. In Colombia it is found in both the DTF of the Upper Magdalena River Valley (Tolima), in Los Colorados (Caribbean Region) and in the TRF of Leticia (the Amazon). It has also been found in the Departments of Antioquia, Caquetá, Choco, Guainia, Guaviare, Meta, Nariño, Valle, and Cauca (Escobar 2000 a, Medina et al. 2001, Neita et al. 2003, Pulido et al. 2003). In Panama it is found to the north, on the Pacific coast and in the Canal Zone on the Caribbean. In Costa Rica it is spread out along the Caribbean and to the north of the country, but it was not recorded in Palo Verde (Guanacaste). On the Pacific coast, it has been cited for two sectors: to the south on the Osa Peninsula and in the basin of the River Tárcoles, with a marked preference for TRF (Solis & Kohlmann 2002). Pattern 3 These taxa likely reflect the processes of speciation in situ in the DTF of Mexico, Costa Rica and Colombia. Glaphyrocanthon, a subgenus of Canthon with many South American species, has two phyletic lines in Mexico with a clear origin in the north of South America. We have already referred to one of these lines, that of C. (Gl.) subhyalinus. The second line is comprised of an isolated species with South American affinities. In addition there is the viridis group with strong speciation in Mexico and which clearly belongs to the subgenus Glaphyrocanthon although it has no evident relationship to South American species. The viridis group is comprised of 15 species, of which only C. (Gl.) meridionalis Martínez, Halffter & Halffter has a distribution that expands towards Central America, opposite to the dominant direction for the distribution of the Canthonini studied. Several of the species of this group are associated with TRF, many others are found in DTF. Canthon (Glaphyrocanthon) pacificus Rivera & Halffter and C. (Gl.) corporali Balthasar belong to this group and both form part of a set of species that are distributed in Mexico along the Pacific slope, from the state of Nayarit to the state of Oaxaca. Canthon (Gl.) corporali is widely distributed in association with DTF, including the Balsas River Basin and the coast of the Pacific on both sides of its outlet. Canthon (Gl.) pacificus is found in dry deciduous and semideciduous forests of the Pacific coast, from the state of Jalisco to the state of Oaxaca. From the same viridis group in Guanacaste the lone species C. (Gl.) meridionalis Martínez, 28 Halffter & Halffter is found and extends to Central America: Guatemala, El Salvador, Nicaragua, and Costa Rica (Solis & Kohlmann 2002). In Costa Rica it has a wide distribution, on both the Pacific and the Atlantic sides (Solis & Kohlmann 2002). Canthon (C.) deyrollei Harold has a geographic distribution from Guatemala to Costa Rica. It is the only species in its phyletic line (Halffter & Martínez 1977). In Costa Rica it is restricted to the area with the greatest climatic seasonality, i.e. land below 600 m in the province of Guanacaste, in the northwest of Costa Rica (Solis & Kohlmann 2002). Canthon (Gl.) acutus Harold was collected in North Tolima. It has also been found in other Departments in Colombia: Bolívar, Guainia and Meta at elevations of 300 m (Medina et al. 2001). DISCUSSION 1) The geological (Fig. 2) and Canthonini species composition (Fig. 3) cladograms only coincide in their general features. The dry tropical forests from Mexico to Colombia arose during a relatively recent geological stage. Hence, historical geological events prior to the Pliocene do not appear to have affected the distribution of this vegetation type. Its distribution is determined by a combination of temperature and rainfall conditions, particularly by a prolonged and marked dry season (see Rzedowski 1978). 2) The flora of dry tropical forests is almost exclusively comprised of species with Neotropical affinities (Rzedowski 1978). The Canthonini fauna has the same affinities. As shown by Trejo (2005), there is a high degree of plant species exchange between different dry tropical forest sites. It is not surprising that the same phenomenon occurs in the Canthonini. The cladograms of the dry tropical forests and the Canthonini exhibit the same affinities found by Gentry (1995) for flora; i.e. a similarity between the dry forests of northern Colombia and Venezuela with that of Chamela, with 15 shared genera, while Chamela and Guanacaste (Costa Rica) only share six genera. 3) The Canthonini species found in the dry tropical forests studied show a clear gradient of affinity associated with the different degrees of expansion of the Canthonini from South America. In spite of the great distance and the ecological barriers that exist between Colombia and Chamela, the affinities are marked and appear to indicate a certain degree of continuity among the dry forests studied. There is little doubt that the expansion of the distribution area of South American Canthonini, 29 and that of many other Scarabaeinae with the same biogeographical history was greatly facilitated by the presence of mammalian megafauna that has since disappeared (see Janzen & Martin 1982). Nevertheless, there are differences among the enclaves of dry tropical forest. In Chamela, the only pre-Pliocene expansion species was collected (Canthon humectus), along with several species of ancient expansion within the Neotropical Pattern, and several species whose distribution in Mexico is more recent, including two belonging to a group (Canthon (Glaphyrocanthon) viridis group) which speciated in Mexico. Palo Verde has a mix of PlioPleistocene and recent expansion species. In Colombia, species with a South American distribution dominate, although it shares species with Palo Verde and Chamela. The biogeographical hypothesis proposed in the Introduction about the origin of Canthonini and their northwards expansion at different times and with different degrees of penetration is supported. The relationships between distinct phyletic lines and even species, with South American fauna clearly demonstrates these biogeographical phenomena. 4) Several of the species found in Colombia, but not in Palo Verde or Chamela, have the northern limit of their distribution in the Térraba River basin in the extreme south of Costa Rica on the Pacific coast. This is the case for C. lituratus, C. juvencus and C. septemmaculatus (see Solis & Kohlmann 2002). As these species do not reach Palo Verde the differences of this site with those of Colombia are accentuated. This northern limit in the distribution of South American species to the extreme south of Costa Rica is interesting. The lowlands of Nicaragua to the north of Lake Nicaragua were proposed by Halffter (1976) as the southern limit of the Mexican Transition Zone. The present study shows the South American affinity of the Canthonini of Costa Rica for dry tropical forest (they share 40% of the species with the DTF of North Tolima), and this seems to apply in general to those species found in other types of vegetation (see Solis & Kohlmann 2002). 5) A good number of the species found in the sites we studied are characteristic of dry tropical forest, and even exclusive to this vegetation type. However, some species are shared with tropical rain forest. The proportion of shared species is much lower in Chamela, moderate in Palo Verde and greatest in the sites of Colombia. The latter is also reflected in the greater affinity of the Canthonini of Colombian DTF with those of Leticia (TRF), than with those of the DTF of Chamela (Fig. 3). Palo Verde shares 30% of its species with Los Tuxtlas, a forest where the DTF and the TRF species of Mesoamerica meet. 30 6) As regards in situ speciation, the most notable example is that of the viridis group, which has no direct relationship with the South American fauna of Canthon (Glaphyrocanthon). This group underwent much diversification in Mexico that appears to be recent and influenced by the process of drying out during the Holocene (many of its species are adapted to dry forest). 7) The enclaves of dry tropical forest in Colombia, in both the Caribbean Region and the Upper Magdalena Valley, are close to tropical rain forest. In the first case, Neguanje is part of the Province of the Sierra Nevada of Santa Marta where the vegetation changes with increasing altitude and in Los Colorados there is still a remanent of rain forest. The Upper Magdalena River Valley is surrounded by rain forest on the skirts of the Cordillera Central and the Cordillera Oriental, and to the north by the Middle Magdalena Valley. This allows them to share species with adjacent ecosystems. 8) The Canthonini fauna of the tropical rain forests (Los Tuxtlas and Leticia) that were used as the outgroup, is very different between these two rain forests and also different from the fauna of the dry tropical forests. Considering all Scarabaeoidea, the similarity in species between Chamela and Los Tuxtlas is only 13 (Sørensen’s QS index), while between Chamela and other dry tropical forest in Mexico the value is 41 (Morón et al. 1988). 9) The quality of the species list is the main limitation of the biogeographical analysis presented here. We have gathered all of the published information as well as that recorded in the databases of CONABIO and INBIO. Even so, these collections cannot be considered exhaustive, especially for Colombia. 10) The predominance of roller species (Canthonini) in dry tropical forest is well defined. On average 33.45% of the Scarabaeinae species found are Canthonini, a value that rises to 35.45% if we exclude Neguanje as the least representative site. In contrast, in the tropical rain forests studied, on average only 28% are rollers. The affinity of the rollers for sunny conditions is even greater in sites that are drier and have very open vegetation. 11) There have been several studies of the changes in the abundance of Scarabaeinae species from the dry season to the rainy season in dry tropical forest: Chamela (Andresen 2005), North Tolima (Escobar 1997), Guanacaste, Costa Rica (Janzen 1983) have produced similar results and emphasize the severity of the dry season. Soil humidity and ambient temperature control the activity and life cycle of Scarabaeinae (Halffter 1991, Martínez & Montes de Oca 1994, BustosGómez & Lopera-Toro 2003, Andresen 2005). On the other hand, the Scarabaeinae of dry 31 tropical forests show a marked tendency for a generalist diet (copro-necrophagous). In Chamela 70% of Scarabaeinae (Andresen 2005) and in the north of Tolima 71% (Bustos-Gómez & Lopera-Toro 2003) have this type of diet. Closing Remarks The biogeographical analysis of areas of dry tropical forest in Mesoamerica and Colombia has provided and integrated geological, geomorphological, climatic and ecological elements that have brought us closer to understanding the creation and relationships of Neotropical dry forests. The use of an outgroup polarized the cladograms, adding a new dimension to the study by establishing an axis for comparison that stood out in all analyses. The analises carried out, along with previous knowledge of the group allowed for an explanation of the spatio-temporal dynamics exhibited by the Canthonini in dry tropical forests. Dry tropical forest represents a setting where the processes of vicariance and dispersal occur. This study reveals Canthonini’s routes of dispersal via the dry forests while the speciation processes of some Canthon, appearing to be synchronous with the establishment of the dry forests of Mesoamerica and Colombia, also indicate the affinities of Canthonini to dry tropical forest, very similar to that of the flora of said ecosystem. Biogeographical and ecological elements were brought together and these allowed us to explain both their presence in and the preference of some Canthonini for dry forest, and/or their exclusive distribution in some dry forests of Colombia and their absence from those of Mesoamerica, as was the case for C. (Gl.) subhyalinus, so clearly demonstrated by the PAE. ACKNOWLEDGEMENTS The first author is grateful to A. Espinosa de los Monteros and J. J. Morrone for their comments on the manuscript; to F. Escobar, M. Zunino, A. Solís, F. Vaz de Mello and A. Díaz, for providing bibliography; to C. Álvarez for providing access to the SNIB-CONABIO data base projects E7, K5, P134, G19, H125, Mexico City and the data base of the National Biodiversity Institute, Costa Rica, Atta: Information System on Costa Rican Biodiversity (http://www.inbio.ac.cr/atta/index.htm). Thanks also to the Universidad de Nariño, Pasto (Colombia: Nariño) for the award of a Commission to Study. We thank two anonymous reviewers for their very useful comments. We are also are grateful to B. Delfosse for translating the manuscript into English. This study represents part of the project 32 “Analysis of the relationships between alpha, beta and gamma diversity on different spatial scales: the historical and ecological processes involved”, Stage V (CONABIO-Mexico). LITERATURE CITED Alfaro, E. A., A. Alvarado, & A. Chavarri. 2001. Cambios edáficos asociados a tres etapas sucesionales de bosque tropical seco en Guanacaste, Costa Rica. Agronomía Costarricense, 21 (1):7-20. Alvarado, G., R. Barquero, I. Boschini, S. Chiesa, & M. Carr. 1986. Relación entre geotectónica y vulcanismo en Costa Rica. Revista CIAF, 11 (1-3):246-264. Andresen, E. 2005. Effects of season and vegetation type on community organization of dung beetles in a tropical dry forest. Biotropica, 37 (2):291-300. Amézquita, S J., A. Forsyth, A. Lopera, & A. Camacho. 1999. Comparación de la composición y riqueza de especies de escarabajos coprófagos (Coleoptera: Scarabaeidae) en remanentes de bosque de la Orinoquia Colombiana. Acta Zool. Mex. (n. s.), 76:113-126. Arellano, L. 1992. Distribución y abundancia de Scarabaeidae y Silphidae (Insecta: Coleoptera) en un transecto altitudinal en el estado de Veracruz. Bachelor of Science Thesis. Facultad de Ciencias, Universidad Nacional Autónoma de México. ________. 2002. Evaluación de la diversidad alfa, beta y gamma de Scarabaeoidea y Silphidae (Insecta: Coleoptera) en la región Cofre de Perote, Veracruz, México. Master of Science Thesis. Facultad de Ciencias, Universidad Nacional Autónoma de México. Aucott, J. W. 1983. The application of moving average analysis to seismic data and its interpretation related to the structure of Ecuador and Colombia. Memorias 10a Conferencia Geológica del Caribe. Cartagena de Indias. INGEOMINAS. Bogotá, Colombia. Bandy, W., V. Kostoglodov, A. Hurtado-Díaz, & M. Mena. 1999. Structure of the southern Jalisco subduction zone, Mexico, as inferred from gravity and seismicity. Geofísica Internacional, 38 (3):127-136. Barrera, A. 1969. Coleoptera en la Colección Nacional. Acta Zool. Mex., 9 (6):1-90. Bates, H. W. 1886-1890. Pectinicornia and Lamellicornia. Biol. Cent. Amer., Zool. Insecta. Coleoptera. Vol. 2. Part. 2. Taylor and Francis, London. Becerra, J. X. 2005. Timing the origin and expansion of the Mexican tropical dry forest. PNAS, 102 (31):10919-10923. 33 Blackwelder, R. 1944. Checklist of the Coleopterous insects of Mexico, Central America, the West Indies and South America. U. S. Natl. Mus. Bull., 185:196-220. Bullock S. H. 1988. Rasgos del ambiente físico y biológico de Chamela, Jalisco, México. Folia Entomol. Mex., 77:5-17 Bustos-Gómez L. F., & A. Lopera-Toro. 2003. Preferencia por cebo de los escarabajos coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) de un remanente de bosque seco tropical al norte del Tolima (Colombia). In G. Onore, P. Reyes-Castillo, & M. Zunino (eds.), Escarabeidos de Latinoamérica: Estado del conocimiento, pp. 59-65. Monografías Tercer Milenio, SEA, Vol. 3, Zaragoza, Spain. Campo-Alves, J. 2003. Disponibilidad y flujos de nutrimentos en una toposecuencia con bosque tropical seco en México. Agrociencia, 37:211-219. Capistrán, H. F. 1992. Los Coleopteros del Parque de Flora y Fauna Silvestre Tropical Pipiapan, Catemaco, Veracruz, México. Bachelor of Science Thesis. Facultad de Biología, Universidad Veracruzana, Xalapa, Veracruz. Caro, M., & D. Spratt. 2003. Tectonic Evolution of the San Jacinto Fold Belt, NW Colombia. CSEG Recorder 36-43. Challenger, A. (1998) Utilización y conservación de los ecosistemas terrestres de México, pasado, presente y futuro. CONABIO. México. Chiesa, S., G. E. Alvarado, M. Pecchio, M. Corella, & A. Zanchi. 1994. Contribution to petrological and stratigraphical understanding of the Cordillera de Guanacaste lava flows, Costa Rica. Rev. Geol. América Central, 17:19-43. Cole, J. 2001. A slow dance for El Niño. Science, 291:1496-1497. CONABIO. 1990. Vegetación potencial de Rzedowski. Escala 1: 400 000. México. Cortés-Ortiz, L., E. Birmingham, C. Rico, E. Rodríguez-Luna, I. Sampaio, & M. RuizGarcía. 2003. Molecular systematics and biogeography of the Neotropical monkey genus, Alouatta. Mol. Phylogenet. Evol., 26:64-81. Craw, R. 1988. Continuing the synthesis between panbiogeography, phylogenetic systematic and geology as illustrated by empirical studies on the biogeography of New Zeland and the Chatham Islands. Syst. Zool., 37 (3):291-310. Cuevas, O. J. L., L. L. A. Díaz, & G. B. Polo. 2003. Mapas generalizados de las anomalías del Caribe Occidental y América Central. V Cuban Geological Congress. Memorias Geomin. La 34 Habana. Cuba. Delgado, L. L. 1989. Fauna de Coleópteros Lamellicornios de Acahuizotla, Guerrero, México. Bachelor of Science Thesis. Facultad de Ciencias, Universidad Nacional Autónoma de México. Deloya, C. 1992. Necrophilous Scarabaeidae and Trogidae Beetles of Tropical Deciduous Forest in Tepexco, Puebla, México. Acta Zool. Mex. (n.s.), 52:1-13. Deloya, C., & M. A. Morón. 1994. Coleópteros Lamellicornios del Distrito de Jojutla, Morelos, México (Melolonthidae, Scarabaeidae, Trogidae y Passalidae). Listados Faunísticos de México V. Universidad Nacional Autónoma de México. Deloya, C., & M. A. Morón. 1998. Scarabaeoidea (Insecta: Coleoptera) necrófagos de Los Tuxtlas, Veracruz y Puerto Angel, Oaxaca, México. Dugesiana, 5 (2):17-28. Deloya, C., G. Ruíz-Lizárraga, & M. A. Morón. 1987. Análisis de la entomofauna necrófila en la región de Jojutla, Morelos, México. Folia Entomol. Mex., 73:157-171. Díaz, R. A. 1998. Ecología y comportamiento de escarabajos rodadores del estiércol (Scarabaeidae: Scarabaeinae) de selvas y pastizales en Los Tuxtlas, Veracruz. Master of Science Thesis. Ecología y Ciencias Ambientales. Facultad de Ciencias, Universidad Nacional Autónoma de México. ________. 2003. Efecto de la fragmentación de selvas en poblaciones de Scarabaeidae y Silphidae (Coleoptera) de Los Tuxtlas, México. Doctoral Thesis. Facultad de Ciencias, Universidad de Alicante. Spain. Escobar, F. 1997. Estudio de la comunidad de coleópteros coprófagos (Scarabaeidae) en un remanente de bosque seco al norte del Tolima, Colombia. Caldasia, 19 (3):419-430. ________. 1998. Análisis regional de la comunidad de escarabajos coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) de los bosques secos de la región Caribe de Colombia. In M. E. Chaves, & N. Arango (eds.), Informe Nacional sobre el Estado de la BiodiversidadColombia, pp. 72-75. Instituto Alexander von Humboldt, PNUMA, Ministerio del Medio Ambiente. Bogotá, Colombia. ________. 2000a. Diversidad y distribución de los escarabajos del estiércol (Coleoptera: Scarabaeidae: Scarabaeinae) de Colombia. In F. Martín-Piera, J. J. Morrone, & A. Melic (eds.), Hacia un proyecto CYTED para el inventario y estimación de la diversidad entomológica en Iberoamérica: PrIBES-2000, pp. 197-210. Monografías Tercer Milenio, 35 SEA, Vol. 1, Zaragoza, Spain. ________. 2000b. Diversidad de coleópteros coprófagos (Scarabaeidae: Scarabaeinae) en un mosaico de habitats en la Reserva Natural Nukak, Guaviare, Colombia. Acta Zool. Mex. (n.s.), 79:103-121. Escobar, F., & P. Chacón de Ulloa. 2000. Distribución espacial y temporal en un gradiente de sucesión de la fauna de coleópteros coprófagos (Scarabaeinae, Aphodiinae) en un bosque tropical montano, Nariño, Colombia. Rev. Biol. Trop., 48 (4):961-975. Espinal, L. S. 1985. Geografía ecológica del departamento de Antioquia. Rev. Fac. Nac. Agr. Medellín, Colombia, 38 (1):24-39. Estrada, A., G. Halffter, R. Coates-Estrada, & D. A. Meritt Jr. 1993. Dung beetles attracted to mammalian herbivore (Alouatta palliata) and omnivore (Nasua narica) dung in the tropical rain forest of Los Tuxtlas, México. J. Trop. Ecol., 9 (1):45-54. Etter, A. 1993. Diversidad ecosistémica en Colombia hoy. In S. Cárdenas, & H. D. Correa (eds.), Nuestra diversidad biológica, pp. 43-61. Fundación Alejandro Escobar, Colección María Restrepo de Angel, CEREC. Santafé de Bogotá. Favila, M. E., & A. Díaz. 1997. Escarabajos coprófagos y necrófagos. In E. González-Soriano, R. Dirzo, & R. Vogt (eds.), Historia Natural de los Tuxtlas, pp. 383-387. Universidad Nacional Autónoma de México. Ferrari, L., & J. Rosas-Elguera. 1999. Late Miocene to Quaternary extension at the northern boundary of the Jalisco block, western Mexico: The Tepic-Zocalo rift revised. In H. DelgadoGranados, G. Aguirre-Díaz, & J. M. Stock (eds.), Cenozoic Tectonics and Volcanism of Mexico, pp. 1-23. Geol. Soc. Amer. Special Paper 334. Ferrari, L., G. Pasquarè, S. Venegas-Salgado, & F. Romero-Ríos. 1999. Geology of the Western Mexican Volcanic Belt and adjacent Sierra Madre Occidental and Jalisco block. In H. Delgado-Granados, G. Aguirre-Díaz, & J. M. Stock (eds.), Cenozoic Tectonics and Volcanism of Mexico, pp. 65-83. Geol. Soc. Amer. Special Paper 334. Fontugne, M., P. Usselmann, M. Lavallée, M. Julien, & C. Hatté. 1999. El Niño variability in the coastal desert of southern Peru during the mid Holocene. Quat. Res., 52:171-179. Galvis, J., G. A. Huguett, & T. P. Ruge. 1979. Geología de la Amazonía Colombiana. Boletín Geológico. INGEOMINAS, Bogotá, Colombia, 22 (3):1-153. García-Real, E. 1995. Abundancia, distribución y estructura de la comunidad de escarabajos 36 coprófagos y necrófagos (Coleoptera: Scarabaeidae) en un gradiente altitudinal de la Sierra de Manantlán, Jalisco-Colima, México. Master of Science Thesis. Colegio de Postgraduados, Montecillo, México. Gentry, A. H. 1982. Neotropical floristic diversity: phytogeographical connections between Central and South America, Pleistocene climatic fluctuations, or an accident of the Andean orogeny? Ann. Missouri Bot. Gard., 69:557-593. ________. 1995. Diversity and floristic composition of Neotropical dry forest. In S. H. Bullock, H. A. Mooney & E. Medina (eds.), Seasonally dry tropical forests, pp. 146-195. Cambridge University Press, Cambridge. Gill, B. D. 1991. Dung beetles in tropical American forests, In I. Hanski, & Y. Cambefort (eds.), Dung beetle ecology, pp. 211-229. Princeton University Press. Princeton, NJ. Giunta, G., E. Navarro, L. Beccaluva, S. Bellia, P. Comin-Chiaramonti, G. Dengo, W. Montero, F. Urbani, & S. Carnemolla. 1996. Geología de las márgenes de la placa del Caribe: generalidades en Guatemala, Costa Rica, La Española y resultados preliminares del análisis de una transversa en la cordillera de la Costa de Venezuela. Rev. Geol. América Central, 19 (20):7-28. Gómez, L. D. 1982. The origin of the Pteridophyte flora of Central America. Ann. Missouri Bot. Gard., 69:548-556. González, H., A. Nuñez, & G. París. 1988. Mapa geológico de Colombia. Escala: 1:1.500.000. INGOMINAS. Bogotá, Colombia. Graham, A. 2003. In the Beginning: Early Events in the Development of Mesoamerica and the Lowland Maya Area. In A. Gómez-Pompa, M. F. Allen, S. L. Fedick, J.J. Jiménez-Osorio (eds.), The Lowland Maya Area three millennia at the human-wildland human-wildland interface, pp. 31-44. Food Products Press an Imprint of the Haworth Press, Inc. New York. Guariguata, M. R., & G. H. Katan. 2002. El bosque Neotropical: pasado y presente (Sección I). Ecología y conservación de bosques Neotropicales. Libro Universitario Regional. Cartago, Costa Rica. Halffter, G. 1961. Monografía de las especies norteamericanas del género Canthon Hoffsg. (Coleoptera, Scarabaeidae) Ciencia, Mex., 20 (9-12):225-320. ________. 1964. La entomofauna mexicana, ideas acerca de su origen y distribución. Folia Entomol. Mex., 6:1-108. 37 ________. 1974. Eléments anciens de l’Entomofaune Néotropicale: Ses implications biogeographiques. Quaest. Entomol., 10:223-262. ________. 1976. Distribución de los insectos en la Zona de Transición Mexicana. Folia Entomol. Mex., 35:1-64 ________. 1978. Un nuevo patrón de dispersión de la Zona de Transición Mexicana: El Mesoamericano de Montaña. Folia Ent. Mex., 39-40:219-222. ________. 1991. Historical and ecological factors determining the geographical distribution of beetles (Coleoptera: Scarabaeidae: Scarabaeinae). Folia Entomol. Mex., 82:195-238. ________. 2003. Tribu Scarabaeini. In M. A. Morón (ed.), Atlas de los escarabajos de México Coleoptera: Lamellicornia, vol 2: Scarabaeidae, Trogidae, Passalidae y Lucanidae, pp. 211229. Argania Editio. Barcelona, Spain. Halffter G, M. E. Favila, & L. Arellano. 1995. Spatial distribution of three groups of Coleoptera along an altitudinal transect in the Mexican Transition Zone and its biogeographical implications. Elytron, 9:151-185. Halffter, G., M. E. Favila, & V. Halffter. 1992. A comparative study of the structure of the scarab guild in Mexican tropical rain forests and derived ecosystems. Folia Entomol. Mex., 84:131-156. Halffter, G., & A. Martínez. 1966. Revisión monográfica de los Canthonina americanos (Coleoptera, Scarabaeidae) Ia Parte. Rev. Soc. Mex. Hist. Nat., 27:89-177. Halffter, G., & A. Martínez. 1977. Revisión monográfica de los Canthonina americanos. Parte IV. Clave para géneros y subgéneros. Folia Entomol. Mex., 38:29-107. Halffter, G., & E. G. Matthews. 1966. The natural history of dung beetles of the subfamily Scarabaeinae (Coleoptera: Scarabaeidae). Folia Entomol. Mex., 12-14:1-312. Halffter, V., & G. Halffter. 2003. Nuevas subespecies de Canthon humectus (Say) (Coleoptera: Scarabaeidae: Scarabaeinae). Folia Entomol. Mex., 42 (3):329-340. Haugh, G., K. Hughen, D. Sigman, L. Peterson, & U. Röhl. 2001. Southward migration of the Intertropical Convergence Zone through the Holocene. Science, 293:1304-1307. Henríquez, C., E. Bornemisza, & F. Bertsch. 1994. Fijación de potasio en vertisoles, inceptisoles, andisoles y ultisoles de Costa Rica. Agronomía Costarricense, 18 (2):133-140. Hernández-Quintero, J. E. 2003. Magnetic satellite anomalies over Mexico. The Geological Society of America, Cordilleran Section, 99th Annual Meeting (April 1-3 2003). Puerto 38 Vallarta. Jalisco. Herrera, L. F., G. Sarmiento, F. Romero, P. J. Botero, & J. Berrio. 2001. Evaluación ambiental de la Depresión Momposina (Colombia) desde el Pleistoceno Tardío a los paisajes actuales. Geología Colombiana, 26:95-121. Hooghiemstra, H & T. van der Hammen. 2001. Desarrollo del bosque húmedo Neotropical en el Neógeno y Cuaternario: la hipótesis de los refugios. In J. Llorente-Bousquets, & J.J. Morrone (eds.), Introducción a la biogeografía en Latinoamérica. Teorías, conceptos, métodos y aplicaciones, pp. 129-136. Facultad de Ciencias. UNAM. Hooghiemstra, H., T. van der Hammen, & A. Cleef. 2002. Paleoecología de la flora boscosa. In M. R. Guariguata, & G. H. Katan (eds.), El bosque Neotropical pasado y presente (Sección I) Ecología y conservación de bosques Neotropicales, pp. 43-58. Costa Rica. Howden, H. F., & V. Nealis. 1975. Effects of clearing in a tropical rain forest on the composition of the coprophagous scarab beetle fauna (Coleoptera). Biotropica, 7:77-85. Howden, H. F., & O. P. Young. 1981. Panamanian Scarabaeinae: taxonomy, distribution and habits (Coleoptera Scarabaeidae). Contrib. Amer. Ent. Inst., 18 (1):1-104. (IAvH) Instituto Alexander von Humboldt. 1997a. Caracterización ecológica de cuatro remanentes de bosque seco tropical de la región Caribe colombiana. Grupo de Exploraciones Ecológicas Rápidas, IAvH. Villa de Leyva. Colombia. ________. 1997b. Diversidad Biológica. Informe nacional sobre el estado de la biodiversidad Colombia. IAvH, Ministerio del Medio Ambiente, PNUMA. Santafé de Bogotá, Colombia. ________. 2000. Colombia Megadiversa: cinco años explorando la riqueza de un país biodiverso. Santafé de Bogotá, Colombia. Jaccard, S., M. Münster, P. O. Baumgartner, C. Baumgartner-Mora, & P. Denyer. 2001. Barra Honda (Upper Paleocene-lower Eocene) and El Viejo (Campanian-Maastrichtian) carbonate platforms in the tempisque area (Guanacaste, Costa Rica). Rev. Geol. América Central, 24:9-28. Janzen, D. H. 1983. Seasonal change in abundance of large nocturnal dung beetles (Scarabaeidae) in a Costa Rican deciduous forest and adjacent horse pasture. Oikos, 41:274283. Janzen, D. H., & P. S. Martin. 1982. Neotropical anachronisms: the fruits the Gomphotheres ate. Science, 215:19-27. 39 Kohlmann, B., & G. Halffter. 1990. Reconstruction of a specific example of insect invasion waves: the cladistic analysis of Canthon (Coleoptera: Scarabaeidae) and related genera in North America. Quaest. Entomol., 26 (1):1-28. Kohlmann, B., & J. Wilkinson. 2003. Fronteras Biogeográficas: Coincidencia entre factores climáticos, topográficos, geológicos e históricos. In J. J. Morrone, & J. Llorente-Bousquets (eds.), Una Perspectiva Latinoamericana de la Biogeografía, pp. 221-226. Las Prensa de Ciencias, Facultad de Ciencias, UNAM. México. Kohlmann, B., J. Wilkinson, & K. Lulla. 2002. Costa Rica desde el espacio. Universidad EARTH. San José, Costa Rica. López-Ramos, E. 1980. Geología General y de México. V. 1-3. Publicaciones Secretaría de Educación Pública. México. Malagón, D. (ed.) 1988. Suelos y bosques de Colombia. Instituto Geográfico Agustín Codazzi. Subdirección Agroecológica. Bogotá, Colombia. Martínez, A., & G. Halffter. 1972. New taxa of American Canthonina (Coleoptera, Scarabaeinae). Ent. Arb. Mus. Frey, 20 (1):33-66. Martínez, A., G. Halffter, & V. Halffter. 1964. Notas sobre el género Glaphyrocanthon (Coleoptera: Scarabaeinae: Canthonina). Acta Zool. Mex. (n.s.), 7 (3):1-42. Martínez I., & E. Montes de Oca. 1994. Observaciones sobre algunos factores microambientales y el ciclo biológico de dos especies de escarabajos rodadores (Coleoptera: Scarabaeidae: Canthon). Folia Entomol. Mex., 91:47-59. Medina, C., A. Lopera-Toro, A. Vítolo, & B. Gill. 2001. Escarabajos coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) de Colombia. Biota Colombiana, 2 (2):131-144. Mendoza, C. H. 1999 Estructura y riqueza florística del bosque seco tropical en la región Caribe y el Valle del Río Magdalena, Colombia. Caldasia, 21 (1):70-94. Meschede, M., U. Barckhausen, & H. U. Worm. 2000. Desarrollo del centro de dispersión entre las placas de Cocos y Nazca y los trazos de los puntos calientes. Rev. Geol. América Central, 23:5-16. Molina, L. E. (ed.). 1996. Geomorfología y Aspectos Erosivos del Litoral Caribe Colombiano. INGEOMINAS. Cartagena, Colombia. Montero, W. 2001. Neotectónica de la región central de Costa Rica: frontera oeste de la microplaca de Panamá. Rev. Geol. América Central. 24:29-56. 40 Morón, M. A. 1979. Fauna de Coleópteros Lamellicornios de la Estación Biológica Tropical Los Tuxtlas, Veracruz. An. Inst. Biol. UNAM., ser. Zoología, 1:375-454. Morón, M. A, J. F. Camal, & O. Canul. 1986. Análisis de la entomofauna necrófila del área norte de la Reserva de la Biosfera Sian Ka’an, Quintana Roo, México. Folia Entomol. Mex., 69:83-98. Morón, M. A., C. Deloya, & L. Delgado-Castillo. 1988. Fauna de coleópteros Melolonthidae, Scarabaeidae y Trogidae de la región de Chamela, Jalisco, México. Folia Entomol. Mex., 77:313-378. Morón, M., & R. Terrón. 1984. Distribución altitudinal de los insectos necrófilos en la sierra norte de Hidalgo, México. Acta Zool. Mex. (n.s.), 3:1-47. Morón, M. A., F. J. Villalobos, & C. Deloya. 1985. Fauna de Coleópteros Lamellicornios de Boca de Chajul, Chiapas, México. Folia Entomol. Mex., 66:57-118. Neita, M. J., L. L. Pardo, M. D. Quinto, & D. N. Cuesta. 2003. Los escarabajos copronecrófilos Coleoptera:Scarabaeidae en la parcela permanente de investigación en biodiversidad (PPIB) en Salero, Chocó. In F. G. Cossio, & A. Ramos (eds.), Diversidad Biológica de un Bosque Pluvial Tropical, pp. 79-90. Universidad Tecnológica del Chocó. Instituto de Investigaciones Ambientales del Pacífico. Comunidad de Salero, Unión Panamericana, Chocó. Nelson, C. E., & F. Nietzen. 2000. Metalogenia de oro y cobre en América Central. Rev. Geol. América Central, 23:25-41. Palacios-Ríos, M., V. Rico-Gray, & E. Fuentes. 1990. Inventario preliminar de los Coleoptera Lamellicornia de la zona de Yaxchilán, Chiapas, México. Folia Entomol. Mex., 78:49-60. París, G., & J. A. Romero. 1994. Fallas activas en Colombia. Boletín geológico. INGEOMINAS, Bogotá, Colombia, 34 (2-3):1-53. Platt J., B. Leyden, M. Salgado-Labouriau, W. M. Lewis, Jr., C. Schubert, M. W.Binford, D. G. Frey, D. R. Whitehead, & F. H. Weibezahn. 1981. Late Quaternary Environmental History of Lake Valencia, Venezuela. Science, 214: 1299-1305. Pulido, H. L. A., C. R. A. Riveros, H. F. Gast, & P. von Hildebrand. 2003. Escarabajos coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) del Parque Nacional Natural Serranía de Chiribiquete, Caquetá, Colombia, Parte I. In G. Onore, P. Reyes-Castillo, & M. Zunino (eds.), Escarabeidos de Latinoamérica: Estado del Conocimiento, pp. 51-58. Monografías Tercer 41 Milenio, vol. 3, SEA, Zaragoza, Spain. Quigley, M. F., & W. J. Platt. 2003. Composition and structure of seasonally deciduous forest in the Americas. Ecol. Monogr., 73 (1):87-106. Ramón, J. C., G. Vidal, A. Rosero, P. Gómez, & H. Borja. 2001. Reevaluación del modelo geológico del Campo Tello (Valle Superior del Magdalena) y sus implicaciones en el programa de desarrollo. Geología Colombiana, 26:59-77. Rich, P. V., & T. H. Rich. 1983. The Central American dispersal route: biotic history and paleography. In D. H. Janzen (ed.) Costa Rican Natural History, pp. 12-34. The University of Chicago Press, USA. Richardson, J. E., R. T. Pennington, T. D, Pennington, & P. M. Hollingsworth. 2001. Rapid diversification of a species-rich genus of Neotropical rain forest trees. Science, 293: 22422245. Riedinger, M. A., M. Steinitz-Kannan, W. M. Last, & M. Brenner. 2002. A ~6100 14 Cyr record of El Niño activity from the Galapagos Islands. Journal of Paleolimnology, 27:1-7. Rivera-Cervantes, L. E., & G. Halffter. 1999. Monografía de las especies mexicanas de Canthon del subgénero Glaphyrocanthon (Coleoptera: Scarabaeidae: Scarabaeinae) Acta Zool. Mex. (n.s.), 77:23-150. Rosen, B. R. 1988. From fossil to earth history: applied historical biogeography. In A. A. Myers, & P. S. Giller (eds.), Analytical Biogeography, pp. 437-481. Chapman, & Hall. London, UK. Rubio, L. C., M. F. Mosquera, & B. H. Aguilera. (eds.) 1977. Mapa hidrogeológico de Colombia. Escala: 1:3000 000. INGEOMINAS. Rutz, L. M. 2002. Microsismicidad del noroeste del Bloque Jalisco (México). Aplicación a la sismotectónica y peligrosidad sísmica de la zona. Bachelor of Science Thesis. Universidad de Granada. Facultad de Ciencias. Spain. Rzedowski, J. 1978. Vegetación de México. Editorial Limusa, México. Rzedowski, J., R. Medina Lemos, & G. Calderón de Rzedowski. 2005. Inventario del conocimiento taxonómico, así como de la diversidad y del endemismo regionales de las especies mexicanas de Bursera (Burseraceae). Acta Botánica Mexicana, 70:85-111. Sabogal, C., & L. Valerio. 1998. Forest composition, structure, and regeneration in a dry forest of the Nicaragua Pacific Coast. In F. Dallmeier, & J. A. Comiskey (eds.), Forest Biodiversity in North, Central and South America, and the Caribbean: Research and Monitoring, pp. 187- 42 212. UNESCO, París, & Parthenon Publishing Group. Sánchez, L., W. Martínez, & J. Flórez. 1998. Mapa gravimétrico de Colombia. Anomalia total de Bouguer. Escala: 1: 2.000 000. Instituto Geográfico Agustín Codazzi (IGAC) Bogotá, Colombia. Sánchez-Valbuena, H. A. (ed.) 1992. Atlas de Colombia. Instituto Geográfico Agustín Codazzi. Subdirección Geográfica. Santafé de Bogotá, Colombia. Sarmiento, G. 1975. The dry plant formations of South America and their floristic conections. J. Biogeography, 2: 233-251. Schmidt-Effing, R. 1980. El origen del istmo centroamericano como vínculo de dos continentes. In W. Zeil (ed.), Nuevos Resultados de la Investigación Geocientífica Alemana en Latinoamérica, pp. 21-32. Deutsche Forschungsgemein-schaft (DFG). Bonn, Germany. SIEPAC. 2003. Términos de referencia sobre impactos transfronterizos y regionales del proyecto SIEPAC línea de transmisión en América Central. Comisión Holandesa de Desarrollo Ambiental. (www.iadb.org/regions/re2/siepac/cap6.pdf) Solís, A., & B. Kohlmann. 2002. El género Canthon (Coleoptera: Scarabaeidae) en Costa Rica. G. it. Ent., 10 (50):1-68. Sommer-Cervantes, I., L. Flores-Delgadillo, & M. Gutiérrez-Ruiz. 2003. Caracterización de los suelos de la Estación de Biología Tropical Los Tuxtlas. In J. Alvarez-Sánchez, & E. Naranjo-García (eds.), Ecología del Suelo en la Selva Tropical Húmeda de México, pp. 17-67. Instituto de Ecología, A.C. y Universidad Nacional Autónoma de México. Steig, E. J. 1999. Mid-Holocene climate change. Science, 286:1485-1487 Swofford, D. L. 2002. PAUP. Phylogenetic Analysis Using Parsimony, Version 4. Sinauer Associates, Sunderland, Massachusetts. Taborda, A. B. 1950. Contribución al conocimiento de la geología del Tolima. INGEOMINAS. Laboratorio Nacional de Fomento Minero Ibagué, Tolima, Colombia. Trejo, I. 2005. Análisis de la diversidad de la selva baja caducifolia en México. In G. Halffter, J. Soberón, P. Koleff, & A. Melic (eds.), Sobre Diversidad Biológica: El Significado de las Diversidades Alfa, Beta y Gama, pp. 111-122. Monografías Tercer Milenio, SEA, Vol. 4, Zaragoza, Spain. Trejo, I., & R. Dirzo. 2000. Deforestation of seasonally dry tropical forest: a national and local analysis in Mexico. Biol. Conserv., 94:133-142. 43 Trejo, I., & R. Dirzo. 2002. Floristic diversity of Mexican seasonally dry tropical forest. Biodiversity and Conservation, 11:2063-2048. Tudhope, A. W., C. P. Chilcott, M. T. McCulloch, E. R. Cook, J. Chappell, R. M. Ellam, D. W. Lea, J. M. Lough, & G. B. Shimmieeld. 2001. Variability in the El Niño- Southern Oscillation through a Glacial-Interglacial Cycle. Science, 291:1511-1517. Ujueta, G. 2003. La falla de Santa Marta-Bucaramanga no es una sola falla; son dos fallas diferentes: la falla de Santa Marta y la Falla de Bucaramanga. Geología Colombiana, 28:133153. Vásquez, M. A. 1983. Soils. In D. H. Janzen (ed.), Costa Rican Natural History, pp. 63-65. The University of Chicago Press. U.S.A. Vulcano, M. A., & F. S. Pereira. 1964. Catalogue of Canthonini (Col. Scarab.) inhabiting the western hemisphere. Ent. Arb. Mus. Frey., 15:570-685. Zunino, M. 2005. Filogenia de áreas de distribución: algunas reflexiones teóricas. Acta Zool. Mex. ( n.s.), 21 (1):115-118. 44 Fig.1 Study sites,DTF (1-7 dry forest sites): 1 = Chamela, 2 = Palo Verde, 3 = Neguanje, 4 = Tierra Bomba, 5 = Los Colorados, 6 = Zambrano, 7 = Norte del Tolima; TRF (8-9): 8 = Los Tuxtlas, 9 = Leticia (Amazonas). The dotted area corresponds to the distribution of DFT, based on Gómez (1982), IAvH (1997 b), CONABIO (1990) and Trejo & Dirzo (2002). 45 1(0) 2(0) 10(0) 75/57 4(1) 16(0) Leticia 5(2) 13(0) 14(0) 100/81 T. Bomba 7(1) 8(2) 9(3) 75/54 18(1) Zambrano 12(1) Colombia 100/57 1(2), 14(0) 6(1) 18(1) Los Colorados 16(1) 100/70 3(0) 11(0) 15(0) 4(3) 8(0) 9(0) 11(2) Neguanje 5(0) 10(1) 1(1) 5 (2) 9(2) 17(1) 2(1), 14(0) N. Tolima P. Verde 3(1) 4(2) 16(1) 14(0) Central America Chamela 6(1) 17(1) Mexico 13(0) 3(1) 4(2) 7(1) Los Tuxtlas Fig. 2. Cladogram of geological areas. The numbers above the nodes indicate 50% majority consensus tree frequency / Jackknife frequency with 20% deletion, 500 replicas. In each clade the appearance of derived = and reversed = characters is indicated. 46 T. Bomba 100/85 Zambrano 67/54 Neguanje 100/75 N. Tolima 100/81 Los Colorados 100/67 Leticia Chamela P. Verde Los Tuxtlas Fig. 3. Cladogram of the areas of affinity of Canthonini. The numbers above the nodes indicate 50% majority consensus tree frequency / Jackknife frequency with 20% deletion, 500 replicas. 47 Table 1. Bibliography consulted for the selection of geomorphological characters. Mexico López-Ramos 1980, Bullock 1988, Bandy et al. 1999, Ferrari et al. 1999, Ferrari & Rosas-Elguera 1999, Rutz 2002, Campo-Alves 2003, Hernández-Quintero 2003, Sommer-Cervantes et al. 2003. Costa Schmidt-Effing 1980, Rich & Rich 1983, Vásquez 1983, Alvarado et al. 1986, Rica Chiesa et al. 1994, Henríquez et al. 1994, Meschede et al. 2000, Nelson & Nietzen 2000, Alfaro et al. 2001, Jaccard et al. 2001, Montero 2001, Cuevas et al. 2003. Colombia Taborda 1950, Rubio et al. 1977, Galvis et al. 1979, Aucott 1983, González et al. 1988, Malagón 1988, Sánchez et al. 1998, Sánchez-Valbuena 1992, París & Romero 1994, Giunta et al. 1996, Molina 1996, Herrera et al. 2001, Ramón et al. 2001, Caro & Spratt 2003. 48 Table 2. Geological-geomorphological characters and their states 1. Land profile. undulating (0), valley (1), flat (2) 2. Potassium. low (0), optimal (1) 3. Organic material in the soil. high (1), low (0) 4. Most abundant minerals in the sandy fraction of the soils. feldspar (0), quartz (1), olivinepyroxene (2), amphiboles (3) 5. Tectonic plate. Caribe (0), North America (1), South America (2) 6. Seismicity. high (0), low-absent (1) 7. Faults. present (0), absent (1) 8. Geological era. Mesozoic (0), Tertiary (1), Quaternary (2) 9. Geological period. Middle Jurassic – Late Cretaceous (0), Eogene [Oligocene, Eocene, Paleocene] (1), Neogene [Pliocene, Miocene] (2), Pleistocene, Holocene (3) 10. Magnetism. low (0), high (1) 11. Type of rock. sedimentary (0), igneous (1), metamorphic (2) 12. Nearby volcanic range. present (0), absent (1) 13. Soil permeability. high to moderate (0), low (1) 14. Origin of rock. marine (0), continental (1) 15. Evolution of the soil profile. zoned (highly evolved) (0), not zoned (young or slightly developed) (1) 16. Origin of the plate. Paleozoic (0), Mesozoic (1) 17. Metals. Ag, Pt, Au, Fe. absent (0), presente (1) 18. Bouguer’s anomalies (gravimetric anomalies). positive (0), negative (1) 49 Table 3. Matrix with character states for the cladogram of geomorphological areas. 123456789012345678 Leticia ( C ) 000121123001010000 Neguanje ( C ) 210301000120110100 Zambrano ( C ) 210001123101110101 Los Colorados ( C ) 210001011101100101 T. Bomba ( C ) 210001123101000100 N. Tolima ( C ) 110020012100100010 Chamela ( M ) 010011011010111010 Los Tuxtlas ( M ) 001210111010011000 P. Verde (CR) 011200011110101100 50 Table 4. List of Canthonini species from the study forests. Numbers are those used in the matrix of Table 2. The subgenera of the genus Canthon are Glaphyrocanthon (Gl.), Canthon (C.), Goniocanthon (G.). 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Canthon (C.) aequinoctialis Harold C. (Gl.) acutus Harold C. (Gl.) corporali Balthasar C. (C.) cyanellus cyanellus LeConte C. (C.) cyanellus sallei Harold C. (C.) deyrollei Harold C. (Gl.) euryscelis Bates C. (Gl.) femoralis (Chevrolat) C. (G.) fulgidus Redtenbacher C. (C.) humectus riverai Halffter & Halffter C. (C.) indigaceus indigaceus LeConte C. (C.) indigaceus chiapas Robinson C. (C.) indigaceus chevrolati Harold C. (C.) juvencus Harold C. (C.) lituratus (Germar) C. (Gl.) luteicollis Erichson C. (Gl.) meridionalis (Martínez, Halffter & Halffter) C. (C.) morsei Howden C. (C.) mutabilis Harold C. (Gl.) pacificus Rivera & Halffter C. (Gl.) semiopacus Harold C. septemmaculatus (Latreille) Insertae sedis C. (Gl.) subhyalinus Harold C. (Gl.) vazquezae (Martínez, Halffter & Halffter) Cryptocanthon peckorum Howden Pseudocanthon perplexus (LeConte) Agamopus lampros Bates Malagoniella astyanax (Harold) 51 Table 5. Data matrix for the cladogram of the Canthonini areas 12345678901234567890123456789012 Chamela ( M ) 00110000011000000001000001101001 Los Tuxtlas ( M ) 00010011000100000100001101001001 P. Verde (CR ) 00001110000010001100000001111001 N. Tolima ( C ) 11001000000001100110001000011001 Zambrano ( C ) 00001000000001100000010000011010 Los Colorados ( C ) 10001000000001000000001000101001 T. Bomba ( C ) 00001000000001100000000000001000 Neguanje ( C ) 00001000000001100000000000001000 Leticia ( C ) 10000000100000010000100010001101 52 Appendix 1. Scarabaeinae (Canthonini and other tribes) of the forests studied MEXICO, CHAMELA Agamopus lampros Bates Canthon corporali Balthasar Canthon pacificus Rivera & Halffter Canthon cyanellus cyanellus LeConte Canthon humectus riverai Halffter & Halffter Canthon indigaceus LeConte Pseudocanthon perplexus (LeConte) Deltochilum timidum Howden Deltochilum gibbosum (Fabricius) Phanaeus demon Laporte Phanaeus furiosus Bates Coprophanaeus pluto Harold Digitonthophagus gazella (Fabricius) Onthophagus landolti Harold Onthophagus igualensis Bates Onthophagus hoepfneri Harold Canthidium sp. Dichotomius colonicus (Say) Dichotomius amplicollis (Harold) Ateuchus rodriguezi (Borre) Uroxys sp. Copris lugubris Boheman PALO VERDE,COSTA RICA Agamopus lampros Bates Canthon cyanellus sallei Harold Canthon deyrollei Harold Canthon euryscelis Bates Canthon indigaceus chevrolati Harold Canthon meridionalis (Martinez, Halffter & Halffter) Canthon morsei Howden Canthon mutabilis Harold Pseudocanthon perplexus (LeConte) Deltochilum lobipes Bates Malagoniella astyanax (Harold) Phanaeus wagneri Harold Phanaeus demon Laporte Phanaeus eximius Bates Coprophanaeus telamon (Erichson) Coprophanaeus pluto (Harold) Onthophagus acuminatus Harold Onthophagus championi Bates 53 Onthophagus landolti Harold Onthophagus hoepfneri Harold Onthophagus marginicollis Harold Onthophagus batesi Howden & Cartwright Canthidium laetum Harold Canthidium guanacaste Howden & Gill Dichotomius yucatanus (Bates) Dichotomius centralis (Harold) Dichotomius annae Kohlmann & Solis Ateuchus rodriguezi (Borre) Copris lugubris Boheman COLOMBIA, CARIBE REGION Zambrano Canthon cyanellus sallei Harold Canthon juvencus Harold Canthon lituratus (Germar) Canthon septemmacutatus (Latreille) Malagoniella astyanax (Harold) Coprophanaeus jasius (Olivier) Canthidium sp. Dichotomius belus (Harold) Onthophagus marginicollis Harold Onthophagus lebasi Boucomont Uroxys sp. Eurysternus impressicollis Laporte Los Colorados Agamopus lampros Bates Canthon aequinoctalis Harold Canthon cyanellus sallei Harold Canthon juvencus Harold Canthon subhyallinus Harold Coprophanaeus jasius (Olivier) Diabroctis cadmus Harold Phanaeus hermes Harold Canthidium sp. Dichotomius belus (Harold) Onthophagus marginicollis Harold Onthophagus lebasi Boucomont Onthophagus sp. Uroxys sp. Tierra Bomba Canthon cyanellus sallei Harold Canthon juvencus Harold 54 Canthon lituratus (Germar) Diabroctis cadmus Harold Canthidium sp. Dichotomius belus (Harold) Onthophagus marginicollis Harold Onthophagus lebasi Boucomont Onthophagus landolti Harold Uroxys sp. Eurysternus impressicollis Laporte Neguanje Canthon cyanellus sallei Harold Canthon juvencus Harold Canthon lituratus (Germar) Phanaeus prasinus Harold Canthidium sp. Dichotomius belus (Harold) Dichotomius sp. Dichotomius sp. Onthophagus marginicollis Harold Onthophagus lebasi Boucomont Onthophagus landolti Harold Eurysternus impressicollis Laporte Eurysternus caribaeus (Herbst) COLOMBIA, NORTH TOLIMA Canthon acutus Harold Canthon aequinoctialis Harold Canthon cyanellus sallei Harold Canthon juvencus Harold Canthon lituratus (Germar) Canthon morsei Howden Canthon mutabilis Harold Canthon subhyalinus Harold Malagoniella astyanax (Harold) Phanaeus hermes Harold Dichotomius belus (Harold) Dichotomius sp. 1 Dichotomius sp. 2 Onthophagus landolti (Harold) Onthophagus lebasi (Boucomont) Onthophagus marginicollis (Harold) Onthophagus rubrescens (Blanchard) Canthidium sp. Ateuchus sp. 55 Uroxys sp. Canthidium sp. Eurysternus plebejus (Harold) MEXICO, LOS TUXTLAS (for rainforest, M. Favila pers. com.) Canthon euryscelis Bates Canthon femoralis (Chevrolat) Canthon morsei Howden Canthon vazquezae (Martinez, Halffter & Halffter) Canthon subhyalinus Harold Canthon edmondsi Rivera & Halffter Deltochilum scabriusculum Bates Deltochilum pseudoparile Paulian Delthochilum gibbosum sublaeve Bates Phanaeus endymion Harold Sulcophanaeus chryseicollis (Harold) Onthophagus nasicornis Harold Onthophagus rhinolophus Harold Canthidium aff. ardens Bates Canthidium centrale Boucomont Canthidium aff. perceptibile Howden & Young Scatimus ovatus Harold Uroxys boneti Pereira & Halffter Uroxys bidentis Howden & Young Uroxys transversifrons Howden & Gill Ontherus mexicanus Harold Bdelyropsis newtoni Howden Dichotomius satanas (Harold) Copris laeviceps Harold Eurysternus angustulus (Harold) Eurysternus caribaeus (Herbst) Eurysternus velutinus Bates COLOMBIA, LETICIA Canthon aequinoctialis Harold Canthon fulgidus Redtenbacher Canthon luteicollis Erichson Canthon semiopacus Harold Cryptocanthon peckrorum Howden Canthonella n. sp. Scybalocanthon sp. Deltochilum amazonicum Bates Deltochilum carinatum Westwood Deltochilum sp. 1 Deltochilum sp. 2 56 Deltochilum sp. 3 Coprophanaeus telamon (Erichson) Coprophanaeus n. sp. Oxysternon conspicillatum (Weber) Oxysternon silenus Laporte Phanaeus bispinus Bates Phanaeus cambeforti Arnaud Phanaeus chalcomelas (Perty) Phanaeus meleagris Blanchard Onthophagus haematopus Harold Onthophagus sp. 1 Onthophagus sp. 2 Onthophagus sp. 3 Ontherus pubens Génier Ontherus diabolicus Génier Uroxys sp. 1 Uroxys sp. 2 Dichotomius boreus (Oliver) Dichotomius mamillatus (Felsche) Dichotomius ohausi (Luederwalt) Dichotomius podalirius (Felsche) Dichotomius sp. 1 Dichotomius sp. 2 Dichotomius sp. 3 Dichotomius sp. 4 Canthidium bicolor Bouc. Canthidium gerstaeckeri Harold Canthidium sp. 1 Canthidium sp. 2 Canthidium sp. 3 Canthidium sp. 4 Canthidium sp. 5 Bdelyrus sp. Ateuchus murrayi (Harold) Ateuchus sp. 1 Ateuchus sp. 2 Eurysternus caribaeus Herbst Eurysternus cayennensis Laporte Eurysternus confusus Jessop Eurysternus foedus Guérin Eurysternus hirtellus Dalman Eurysternus inflexus (Germar) Eurysternus velutinus Bates 57 CAPITULO 3 Mesoamerican and Colombian Dry Tropical Forest and Phanaeine Dung Beetles (Coleoptera: Scarabaeidae, Scarabaeine): A Cladistic Analysis of Relationships (Enviado a Biotropica) 58 CAPÍTULO III Mesoamerican and Colombian Dry Tropical Forests and Phanaeini Dung Beetles (Coleoptera: Scarabaeidae, Scarabaeinae): A Cladistic Analysis of Relationships D. N. Padilla-Gil and D. Edmonds ABSTRACT The purpose of this paper is to examine the historical geographical relationship among enclaves of tropical dry forest in Mexico, Costa Rica and Colombia using the phanaeine dung beetles as the indicator taxon, and to compare the results with those of a parallel study of the cantonini dung beetles. Focal dry forest enclaves studied are Chamela (southwestern Mexico), Palo Verde (Costa Rica) and five localities in upper Rio Magdalena valley and Caribbean regions of Colombia. For the cladistic analysis two humid tropical tropical forest served as outgroups: Los Tuxtlas (Mexico) and Leticia (Colombia). Dry forest area relationships were analized with a Parsimony Analysis of Endemicity (PAE) with species of Phanaeus and related phanaeines as characters. Area cladograms obtained indicate closest relationships between dry forest enclaves of Chamela and Palo Verde, and between those Tolima and Los Colorados in the Caribbean region of Colombia. The Mesoamerican diversification of Phanaeini seems to coincide spatiotemporally with the presumed origin of tropical dry forest of the study areas and has resulted in several endemic communities. The history of dry forest phanaeine and cantonini dung beetles coincide in general aspects. In both cases the endemic dry forest taxa are presumed to have arisen in situ. Key words: dry tropical forest, Phanaeini, Canthonini, Mexico, Costa Rica, Colombia, area cladistic. 59 INTRODUCTION Tropical dry forest (deciduous tropical forest) is characterized as that tropical plant formation presenting the following suite of features: a) a continuous woody coverage of species that loose their leaves during a dry season of variable length; b) an average annual temperature above 24ºC; c) seasonal rains of 700 to 2000 mm alternating with a marked dry season; d) an altitudinal range from 0 to 1000 m (Espinal, 1985; IavH, 1997a; Rzedowski, 1978). According to Rzedowski (1978), this formation is especially characteristic of the Pacific slopes of Mexico, where it covers large areas practically uninterrupted from southern Sonora and southwestern Chihuahua through Chiapas into Central America (Fig. 1). It penetrates deeply inland along the Santiago and Balsas Rivers and their tributaries in southern Mexico and the central valleys of Chiapas and several areas along the Gulf coast into Yucatan. Tropical dry forest at one time covered about 270,000 km2 (6-14%) of Mexican territory below 1500 m; in modern times coverage by intact forest has been reduced to about 27% of that original area (Trejo and Dirzo, 2000). In Costa Rica about 100 years ago, tropical dry forest covered almost the entire peninsula of Nicoya and the island of Chira (Kohlmann et al. 2002). While at one time dry deciduous forest covered as much as 7% (33,600 km2) of the landmass of Central America below 1000 m, its present day coverage is probably less than 2% of this original area (Sabogal and Valerio, 1998). There are three large areas of dry tropical forest in Colombia. The two largest of these occur in the Caribbean plains (including the southern portion of the Department of Guajira) and the valley of the Magdalena River in the departments of Tolima, Cundinamarca and Huila (IavH, 1997b). These forests are severely threatened in Colombia, occupying roughly 1% of their original area of approximately 80,000 km2 (Etter, 1993). Dry tropical forests originated in the Pleistocene and, in middle Holocene, became entrenched in the caribbean region of Colombia as well as the Pacific coast of Mexico and Costa Rica (Padilla-Gil and Halffter, 2007). The relationships among dry tropical forests of Mesoamerica and Colombia and the tribe Canthonini has been considered by Padilla-Gil and Halffter (2007). This parallel study of Colombian and Mesoamerican dry forests is based on an analysis of the dung beetle tribe Phanaeini, a New World group comprising about 150 species. Twelve genera are currently included in the tribe, of which five are known to occur in dry tropical forest habitats in the study areas (Phanaeus, Coprophanaeus, Sulcophanaeus, Oxysternon and Diabroctis. Remaining genera are Dendropaemon, Homalotarsus, Tetrameira, Megatharsis, Gromphas, Oruscatus and 60 Bolbites). Several works treat the taxonomy and distribution of the tribe, among them Olsoufieff (1924), Edmonds (1972), and Arnaud (2002). Philips et al. (2004) presents the only recent view of the phylogeny of the tribe. Three genera are the subjects of recent revisions, Phanaeus (Edmonds, 1994), Sulcophanaeus (Edmonds, 2000) and Oxysternon (Edmonds and Zidek, 2004); and a revision of Coprophanaeus is a work in preparation by Edmonds and Zidek (pers. comm.). The objectives of this study are as follows: a) to compare the fauna of phanaeines in several enclaves of Mesoamerican and Colombian dry forests; b) to relate faunal composition of these areas to the historico-geomorphological characteristics of the regions in which the enclaves occur; c) to consider other factors that might explain the modern distributions of dry forest phanaeines; d) to develop a global hypothesis to explain the biogeography of the phanaeines in Mesoamerican and Colombian dry forests; and e) to integrate, insofar as possible, biogeographic hypotheses concerning the tribes Phanaeini and Canthonini. METHODS Study Sites: The study sites utilized in this study (Fig. 1) are the same representative enclaves of dry forest habitat used in the analysis of Canthonini by Padilla-Gil and Halffter (2007). Chamela (Mexico, Jalisco) and Palo Verde (Guanacaste, Costa Rica) lie in the Pacific coastal belt of Mesoamerica. Four sites (Neguanje, Tierra Bomba, Los Colorados and Zambrano) lie in the Caribbean region of Colombia; the fifth Colombian site, the northern portion of Tolima, occupies the upper valley of the Magdalena River. The ecogeographical characters of the study sites are detailed in the Appendix 1. In addition to faunal lists and other publications that treat phanaeine distribution, geographical data were extracted from the databases maintained by INBio (Costa Rica) and SNIB-CONABIO (Mexico). Area Cladogram: The area relationships were analyzed using Parsimony Analysis of Endemicity (PAE) as modified by Padilla-Gil and Halffter (2007), where taxa are analogs of localities and characters analogs of species using an outgroup. The ingroup is defined as the collection of seven study sites; the outgroup is two enclaves of tropical humid forest (Los Tuxtlas in Veracruz, Mexico, and Leticia in Amazonas, Colombia). The outgroup provides an important point of view where contrasting ecological and taxonomic elements can be brought to bear on the current discussion. Species lists used for the analysis comprise Appendix 2 (see below); Appendix 3 is a 61 data matrix where presence (1) or absence (0) of a given taxon is noted by study and by outgroup sites. Analysis utilized PAUP 4.0b10 (Swofford, 2002) with Acctran optimization and exhaustive search. Species Distribution: In addition to the databases of InBio and SNIB-CONABIO, the species distribution comprising Appendix 2 were derived from the following literature: Morón et al., 1988 and Andresen, 2005 (Chamela); Escobar, 1997; Bustos-Gómez and Lopera-Toro, 2003 (northern Tolima); Escobar, 1998, 2000a (caribbean Colombia); Howden and Nealis, 1975 (Leticia); Morón, 1979; Halffter etal., 1992; Favila and Díaz, 1997; Deloya and Morón, 1998 and Díaz, 1998, 2003 (Los Tuxtlas). In addition, unpublished data were provided by Federico Escobar (Caribbean Colombia) and Bruce Gill (Leticia). Geographic Distribution: Distributional data were compiled from the following sources (in addition to those listed above): Edmonds (1994, 2000, 2003); Morón and Terrón, 1984; Morón et al., 1985; Morón et al., 1986; Deloya et al., 1987; Deloya, 1992; Halffter et al., 1992; Estrada et al., 1993; Deloya and Morón, 1994; Garcia-Real, 1995; Halffter et al., 1995; Medina et al,. 2001; Amézquita et al., 1999; Escobar 2000a, 2000b; Neita et al., 2003; Pulido et al., 2003; Howden and Young, 1981; Martinez and Clavijo, 1990; and Gámez, 2004; Bates 1887; Blackwelder, 1944 and Barrera, 1969. RESULTS Two area cladograms were generated: one based on all phanaeines from all sites, based on the majority consensus tree as it offered the greatest resolution; another based only on species of Phanaeus, the largest phanaeine genus, from all sites except Zambrano and Tierra Bomba, where the genus is not known to occur. Cladogram of the areas of affinity of Phanaeini (Fig. 2): The most closely related dry tropical forests considering all 23 phanaeine species are Chamela (Mexico) and Palo Verde (Costa Rica). Shared species unique to these sites are Coprophanaeus pluto and Phanaeus demon. Species unique to a single site are as follows: Chamela: Phanaeus furiosus Bates; Palo Verde: P. eximius Bates, P. wagneri Harold. Species unique to outgroup sites are: Los Tuxtlas: Coprophanaeus gilli Arnaud, Phanaeus endymion Harold, P. mexicanus Harold, P. sallei Harold, P. tridens Laporte y 62 Sulcophanaeus chryseicollis (Harold); Leticia: P. meleagris Blanchard, P. chalcomelas (Perty), P. cambeforti Arnaud, P. bispinus Bates, Oxysternon silenus Laporte, O. conspicillatum Weber and Coprophanaeus parvulus (Olsoufieff). Cladogram of the areas of affinity of Phanaeus (Fig. 3): The most closely related dry tropical forests considering only those 14 species of Phanaeus is a tie between Chamela and Palo Verde on one hand, and Tolima and Los Colorados on the other. In the first case, the shared species is Phanaeus demon; in the second, P. hermes. DISCUSSION Biogeographic Setting of Dry Tropical Forests: The geomorphological cladogram based on the tribe Canthonini produced by Padilla-Gil and Halffter (2007) links as sister groups the forest enclaves of Colombia ((((((Leticia, T. Bomba) Zambrano) Los Colorados) Neguanje) Tolima) Palo Verde, Chamela), while that resulting from this study groups Palo Verde and Chamela (Mexico). The geologic cladogram and that based on phanaeine taxonomy agree only in general terms. The dry tropical forests of Mexico and Colombia are of relatively recent origin such that pre-Pliocene events seem to have had negligible influence on dry forest phanaeine distribution. The appropriate geological setting to examine phanaeine evolution in dry forests is Pleistocene (see Padilla-Gil and Halffter, 2007). Biogeographic Patterns of Tropical Dry Forest Phanaeines: Philips et al. (2004) postulate that the tribe Phanaeini is a monophyletic group within which Phanaeus and Oxysternon are sister groups (with Sulcophanaeus nearest outgroup) with co-origin in South America. Edmonds (1972, 1994) concludes that the origin and initial diversification of Phanaeus, sen. str, are Mesoamerican and occurred at a much earlier date than the relatively recent penetration of Sulcophanaeus, Coprophanaeus and Oxysternon into that area. The diversification of all four genera in Mesoamerica seems to coincide spatio-temporally with the presumed origin of Mesoamerican tropical dry forests. In situ speciation is dry forest enclaves has resulted in the following endemic phanaeine communities. Dry forests endemics of the Pacific region of Mexico (Coprophanaeus pluto nogueirai; Phanaeus demon demon; P. furiosus; P. nimrod; P. zapotecus): The known distribution of Coprophanaeus pluto nogueirai extends from Chamela northward along the Pacific plains into 63 southern Arizona, and southward along the Pacific seaboard to Colima. Phanaeus demon extends north from Chamela into Sinaloa and southward to Costa Rica (as P. demon excelsus); P. furiosus is restricted to the dry forests of southwest Mexico as far north as Sonora; P. nimrod is endemic to the Valley of Oaxaca, and P. zapotecus occurs in the dry forests of the Sierra Madre del Sur in southern Oaxaca. The Pacific Coastal Region of Mexico, especially Chamela and the Balsas River valley, is a speciation center for the phanerogam plants (see Becerra, 2005; Gentry, 1995; Rzedowski, 1991; and Rzedowiski et al. 2005). Likewise, the dry forests of Oaxaca possess a high rate of plant endemism (Salas-Morales et al. 2003; Gallardo-Cruz et al. 2005). Dry forests endemics of Central America (Phanaeus eximius; P. demon excelsus Bates; P. wagneri wagneri): P. eximius occurs in the Pacific dry forests of Guanacaste northward along the coast to southern Guatemala; it also occurs in a restricted area near the Caribbean coast of Honduras. Phanaeus demon excelsus is restricted to Costa Rica Pacific dry forests. P. wagneri ranges from the dry forests of Costa Rica to the central valley of Chiapas, and from Honduras northward into the Yucatan Peninsula. Dry forests endemics of Colombia (Phanaeus hermes; P. prasinus): P. hermes occurs in northern Tolima, the Colombian Caribbean and Norte de Santander. It has also successfully adapted to the premontane humid forests on the slopes of the Sierra del Perijá, as well as in Panama and Costa Rica. Phanaeus prasinus occurs in the Colombian Caribbean in Neguanje and Norte de Santander. In Venezuela it occurs in the arid region of Falcon, the Maracaibo depression and the Cordillera de Merida. These species are closely related, the only two members of the P. hermes group (Edmonds, 1994); both also enter the sub Andean selva. Diabroctis cadmus (probably endemic) occurs in the dry forest of Tierra Bomba and Colorados (Colombian caribean) and northern Venezuela. Several other phanaeines also occur in the Colombian Caribbean and northern Venezuela but are absent from the forest enclaves under study here: Sulcophanaeus steinheili (an endemic) and Coprophanaeus gamezi (probably endemic). Other dry forest phanaeines: most widely distributed phanaeine genera are Phanaeus and Coprophanaeus, each with species as far south as Argentina and as far north as Texas. Except for limited penetration and Mesoamerican diversification of Sulcophanaeus, and Oxysternon, phanaeines remain a largely South American group. In the case of Phanaeus, the subgenus 64 Notiophanaeus is South American with limited penetration into Mesoamerica, while Phanaeus, sen. str., originated and diversified almost exclusively in Mesoamerica. Only the genera Phanaeus, Coprophanaeus, Sulcophanaeus and Diabroctis include dry tropical forest species, only the first two in Mesoamerica. In South America, S. steinheili occurs north Colombian and Venezuelan dry forest habitats; its closest relative, S. imperator, inhabits dry forests of northern Argentina and Paraguay (Chacoan regions). Other South American dry forest Phanaeus include P. achilles in Ecuador and Peru, and the species pair P. palaeno – P. kirbyi in the cerrados and similar formations of central portion of the continent. Most (if not all) members of the subgenera Megaphanaeus and Metallophanaeus of Coprophanaeus inhabit thorn and other dry forests from Brazil to Argentina. Biogeography of the Genus Phanaeus: If we consider dry forest species of Phanaeus in the context of the biogeography of the genus as a whole, the follow observations emerge: 1) Pacific dry forests are areas of high endemism; in the case of Chamela, no species is shared with other dry forests of the Mexican Gulf Region (Halffter and Arellano, 2002; Halffter et al., 1995; see also the case of plant endemism reported by Rzedowski, 1978); the same is true for 2) no species occurs both in the dry forests and in humid forests examined; in each case, resident species are highly stenotopic; 3) in Mesoamerica, several species have broad eco-geographic ranges that include xeric habitats, among them P. amithaon, P. tridens, P. daphnis, P. mexicanus (Edmonds, 1994; García-Real, 1995; Halffter et al., 1995; Halffter and Arellano, 2002). Comparative Biogeography of the tribes Phanaeini and Canthonini: 1) The north-south distributional limit for dry forest species (and perhaps other species) in both tribes lies in Costa Rica (southern limit for Coprophanaeus p. pluto, Phanaeus demon excelsus, P. eximius and P. w. wagneri; northern limit for P. hermes). The principal physical barrier appears to be the Cordillera de Talamanca, which was uplifted in late Miocene-early Pliocene (8-5 mya) during closing of the Panamanian portal (Coates & Obando, 1996) and the Rio Grande de Térraba. In the case of Canthonini (Padilla-Gil and Halffter, 2007), the Talamanca zone is the northern limit for the dry forest taxa Canthon lituratus (Germar), C. juvencus Harold y C. septemmaculatus (Latreille). The interplay of the climatological effects of the Pacific current and physiographic complexity of the region has made the southern Pacific coast of Costa Rica and Panama, which is wetter than coastal areas to the north, is an important barrier to other groups as well (see Kohlmann et al., 65 2002, for a discussion of climatic differences between these areas). 2) Our best guess is that the diversification of both Phanaeini and Canthonini in Mesoamerica and northern ColombiaVenezuela began in the Pleistocene (see Padilla-Gil and Halffter, 2007). 3) Dry forest canthonines clearly demonstrate a wave of diversification from South America. Chamela is home to species that have followed all three distribution patterns recognized by Padilla-Gil and Halffter (2007); while the other dry forest sites examined in Costa Rica and South America have species only of one pattern, recent invaders/endemics. In contrast, the phanaeine faunas of these areas are endemics resulting from in situ speciation. 4) The Mexican Pacific coast fauna of both groups has colonized adjacent dry forests to the north, for example, the Jojutla District of Morelos, where half the species of Phanaeini and Canthon are shared with Chamela (see Deloya and Moron, 1994). The Jojutla district of the state of Morelos separates two major physiogeographic regions, the Balsas Depression and the Transverse Volcanic Axis; and the fauna of the two dung beetle groups reflects this transition. 5) The Canthon (Glaphyrocanthon) viridis species group (RiveraCervantes and Halffter, 1999) and the genus Phanaeus (see Edmonds, 1994) share two Mesoamerican speciation centers; viz., Mexican Transition Zone mountain systems (below 2000 m for the viridis group, 2700 m for Phanaeus) and the Pacific region dry forests. The diversification of both groups in both centers was undoubtedly synchronous and association with the later stages of the Pliocene-Pleistocene orogenic development of the Mexican Transition Zone (see Appendix 4). 6) In the case of Canthon, the evolutionary scenario developed on the basis of the interrelationships among the dry forest faunas and their relationships with humid forest faunas confirm previous hypotheses of a South American origin and diversification. It does not exhibit, as does Phanaeus, two major centers of extensive radiation, Mesoamerica and South America. The tropical dry forests of Mesoamerica and Colombia are important centers of diversification for not only plants and other animals but also for dung beetles. In both the case of the canthonines (Padilla-Gil and Halffter, 2007) and that of phanaeines analysis of available data permit the establishment of primary biogeographic hypotheses (sensu Morrone, 2004). These in turn will permit future comparative studies based on other Scarabaeinae groups and, perhaps a global hypothesis that explains the current distribution and evolution of this remarkable group of beetles. 66 ACKNOWLEDGEMENTS Our sincerest thanks to Gonzalo Halffter, A. Espinosa de los Monteros and J.J. Morrone for their careful review of the manuscript; to F. Escobar and F. Vaz de Mello for assistance with bibliography; and to C. Álvarez for permission to access the SNIB-CONABIO database (Projects E7, K5, P 134, G 19, H125); and to the Universidad de Nariño, Pasto (Colombia) for awarding NP-G a “Comisión de Estudios.” 67 LITERATURE CITED ANDRESEN, E. 2005. Effects of season and vegetation type on community organization of dung beetles in a tropical dry forest. Biotropica 37(2): 291-300. AMÉZQUITA, S.J., A. FORSYTH, A. LOPERA, AND A. CAMACHO. 1999. Comparación de la composición y riqueza de especies de escarabajos coprófagos (Coleoptera: Scarabaeidae) en remanentes de bosque de la Orinoquia Colombiana. Acta Zoologica Mexicana n.s. 76: 113-126. ARNAUD, P. 2002. Phanaeini. Les Coléoptères du Monde. 28. Hillside Books. Canterbury. BARRERA, A. 1969. Coleoptera en la Colección Nacional. Acta Zoologica Mexicana 9 (6): 1-90. BATES, H. W. 1886-1890. Pectinicornia and Lamellicornia. Biol. Cent. Amer., Zool. Insecta. Coleoptera. Vol. 2. Part. 2. London. BLACKWELDER, R. 1944. Checklist of the coleopterous insects of Mexico Central America the West Indies and South America. United States National Museum. Smithsonian Institution Washington Bulletin 185: 196-220. BECERRA, J.X. 2005. Timing the origin and expansion of the Mexican tropical dry forest. PNAS. 102 (31): 10919-10923. BUSTOS-GÓMEZ, L.F., AND A. LOPERA-TORO. 2003. Preferencia por cebo de los escarabajos coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) de un remanente de bosque seco tropical al norte del Tolima (Colombia). In: Onore G, Reyes-Castillo P, Zunino M, eds. Escarabeidos de Latinoamérica: Estado del conocimiento. Monografías Tercer Milenio, SEA, Vol. 3, Zaragoza, Spain, 59-65. COATES, A.G., AND J.A. OBANDO. 1996. The Geologic Evolution of the Central American Isthmus. In: Jackson JBC, Budd AF, Coates AG, eds. Evolution and Environment in Tropical America. The University of Chicago Press, Chicago, 21-56 DELOYA, C. 1992. Necrophilous Scarabaeidae and Trogidae Beetles of Tropical Deciduous Forest in Tepexco, Puebla, México. Acta Zoologica Mexicana n.s. 52: 1-13. DELOYA, C., AND M.A. MORÓN. 1994. Coleópteros Lamelicornios del Distrito de Jojutla, Morelos, México (Melolonthidae, Scarabaeidae, Trogidae y Passalidae). Listados Faunísticos de México V. UNAM. México. DELOYA, C., AND M.A. MORÓN. 1998. Scarabaeoidea (Insecta: Coleoptera) necrófagos de “Los Tuxtlas”, Veracruz y Puerto Angel, Oaxaca, México. Dugesiana 5(2):17-28. DELOYA, C., G. RUÌZ -LIZÁRRAGA, AND M.A. MORÓN. 1987. Análisis de la entomofauna necrófila 68 en la región de Jojutla, Morelos, México. Folia Entomológica Mexicana 73: 157-171. DÍAZ, R. A. 1998. Ecología y comportamiento de escarabajos rodadores del estiércol (Scarabaeidae: Scarabaeinae) de selvas y pastizales en Los Tuxtlas, Veracruz. Master of Science Thesis. Faculty of Science, UNAM. México. DÍAZ, R. A. 2003. Efecto de la fragmentación de selvas en poblaciones de Scarabaeidae y Silphidae (Coleoptera) de Los Tuxtlas, México. Doctoral Thesis. Faculty of Science, Universidad de Alicante. Spain. EDMONDS, W.D. 1972. Comparative skeletal morphology, systematics and evolution of the Phanaeine dung beetles (Coleoptera: Scabaeidae). University of Kansas Science Bulletin, 49: 731-874. EDMONDS, W.D. 1994. Revision of Phanaeus Macleay, a new world genus of Scarabaeine dung beetles (Coleoptera: Scabaeidae, Scabaeinae). Contributions in Science, Natural History Museum of Los Angeles County 44: 1-105. EDMONDS, W.D. 2000. Revision of the Neotropical dung beetle genus Sulcophanaeus (Coleoptera: Scarabaeidae: Scarabaeinae). Folia Heyrovskyana, supplementum 6: 1-60. EDMONDS, W.D. 2003. Tribu Phanaeini. In: Morón MA, ed. Atlas de los escarabajos de México Coleoptera: Lamellicornia vol 2: Scarabaeidae, trogidae, Passalidae y Lucanidae Barcelona, Spain, 58-65. EDMONDS, W.D., AND J. ZÍDEK. 2004. Revision of the neotropical dung beetle genus Oxysternon (Scarabaeidae: Scarabaeinae: Phanaeini) Folia Heyrovskyana, supplementum 11: 1-58. ESCOBAR, F. 1997. Estudio de la comunidad de coleópteros coprófagos (Scarabaeidae) en un remanente de bosque seco al norte del Tolima, Colombia. Caldasia 19 (3): 419-430. ESCOBAR, F. 1998. Análisis regional de la comunidad de escarabajos coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) de los bosques secos de la región Caribe de Colombia. In: Chaves ME, Arango N, eds. Informe Nacional sobre el Estado de la Biodiversidad-Colombia Instituto Alexander von Humboldt, PNUMA, Ministerio del Medio Ambiente, 72-75. ESCOBAR, F. 2000a. Diversidad y distribución de los escarabajos del estiércol (Coleoptera: Scarabaeidae: Scarabaeinae) de Colombia. In: Martín-Piera F, Morrone JJ, Melic A, eds. Hacia un proyecto CYTED para el inventario y estimación de la diversidad entomológica en Iberoamérica: PrIBES- Monografías Tercer milenio, SEA, Vol. 1, Zaragoza, Spain, 197-210. ESCOBAR, F. 2000b. Diversidad de coleopteros coprófagos (Scarabaeidae: Scarabaeinae) en un 69 mosaico de habitats en la Reserva natural Nukak, Guaviare, Colombia. Acta Zoologica Mexicana n.s. 79: 103-121. ESPINAL, L.S. 1985. Geografía ecológica del departamento de Antioquia. Revista de la Facultad Nacional de Ciencias Agrarias (Medellín) 38 (1): 24-39. ESTRADA, A., G. HALFFTER, R. COATES-ESTRADA, AND D.A. MERITT, JR. 1993. Dung beetles attracted to mammalian herbivore (Alouatta palliata) and omnivore (Nasua narica) dung in the tropical rain forest of Los Tuxtlas, México. Journal of Tropical Ecology 9 (1): 45-54. ETTER, A. 1993. Diversidad ecosistémica en Colombia hoy. In: Cárdenas S, Correa HD, eds. Nuestra diversidad biológica. Fundación Alejandro Escobar, Colección María Restrepo de Angel, CEREC. Santafé de Bogotá, 43-61. FAVILA, M.E., AND A. DÍAZ. 1997. Escarabajos coprófagos y necrófagos. In: Gonzalez-Soriano E, Dirzo R, Vogt R, eds. Historia Natural de los Tuxtlas. Mexico. Universidad Nacional Autónoma de México, 383-387. GALLARDO-CRUZ, J.A., J.A. MEAVE, AND E.A. PÉREZ-GARCÍA. 2005. Estructura, composición y diversidad de la selva baja caducifolia del Cerro Verde, Nizanda (Oaxaca), México. Boletín de la Sociedad Botánica de México 76: 19-35. GÁMEZ, J. 2004. Phanaeini (Coleoptera: Scarabaeinae) de la cordillera de los Andes, depresión de Maracaibo y llanos de Venezuela. Memoria de la Fundación La Salle de Ciencias Naturales 158: 43-60. GARCÍA-REAL, E. 1995. Abundancia, distribución y estructura de la comunidad de escarabajos coprófagos y necrófagos (Coleoptera: Scarabaeidae), en un gradiente altitudinal de la Sierra de Manantlán, Jalisco-Colima, México. Master of Science Thesis. Colegio de Postgraduados, Montecillo. México. GENTRY, A.H. 1995. Diversity and floristic composition of neotropical dry forest. In: Bullock S, Mooney HA, Medina E., eds. Seasonally dry Tropical Forest. Cambridge University Press. Cambridge, Massachussets, 146-194. HALFFTER, G., AND L. ARELLANO. 2002. Response of dung beetle diversity to human-induced changes in a tropical landscape. Biotropica 34 (1):144-154. HALFFTER, G., M.E. FAVILA, AND V. HALFFTER. 1992. A comparative study of the structure of the scarab guild in Mexican tropical rain forests and derived ecosystems. Folia Entomológica Mexicana 84: 131-156. 70 HALFFTER, G., M.E. FAVILA, AND L. ARELLANO. 1995. Spatial distribution of three groups of Coleoptera along an altitudinal transect in the Mexican Transition Zone and its biogeographical implications. Elytron 9: 151-185. HOWDEN, H.F., AND V. NEALIS. 1975. Effects of clearing in a tropical rain forest on the composition of the coprophagous scarab beetle fauna (Coleoptera). Biotropica 7: 77-85. HOWDEN, H.F. AND O.P. YOUNG. 1981. Panamanian Scarabaeinae: taxonomy, distribution, and habits (Coleoptera Scarabaeidae). Contributions of the American Entomological Institute 18 (1): 1-104. INSTITUTO DE INVESTIGACIÓN DE RECURSOS BIOLÒGICOS A LEXANDER VON H UMBOLDT (IAVH). 1997a. Caracterización ecológica de cuatro remanentes de bsT de la región Caribe colombiana. Grupo de Exploraciones Ecológicas Rápidas, IAvH. Villa de Leyva. IAVH. 1997b. Diversidad Biológica. Informe nacional sobre el estado de la biodiversidad Colombia. IAvH, Ministerio del Medio Ambiente, PNUMA. KOHLMANN, B., J. WILKINSON, K. LULLA. 2002. Costa Rica desde el espacio. Universidad AND EARTH. San José, Costa Rica. MARTINEZ A., AND J. CLAVIJO. 1990. Notas sobre Phanaeina venezolanos, con descripción de una nueva subespecie de Diabroctis (Coleoptera, Scarabaeidae, Coprini). Biología Entomologia Venezolana N. S. 5 (20):147-157. MEDINA, C., A. LOPERA-TORO, A. VÍTOLO, AND B. GILL. 2001. Escarabajos Coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) de Colombia. Biota Colombiana 2 (2): 131-144. MORÓN, M.A. 1979. Fauna de coleópteros lamellicornios de la Estación Biológica Tropical “Los Tuxtlas”, Veracruz. Annales Instituto Biología Universidad Autónoma de México, serie zoología 50 (1): 375-454. MORÓN, M.A., AND R. TERRÓN. 1984. Distribución altitudinal de los insectos necrófilos en la sierra norte de Hidalgo, México. Acta Zoologica Mexicana n.s. 3: 1-47. MORÓN, M.A., J.F. CAMAL, AND O. CANUL. 1986. Análisis de la entomofauna necrófila del área norte de la Reserva de la Biosfera “Sian Ka’an”, Quintana Roo, México. Folia Entomológica Mexicana 69: 83-98. MORÓN, M.A, C. DELOYA, AND L. DELGADO-CASTILLO. 1988. Fauna de coleópteros Melolonthidae, Scarabaeidae y Trogidae de la región de Chamela, Jalisco, México. Folia Entomológica Mexicana 77: 313-378. 71 MORÓN, M.A., F.J. VILLALOBOS, AND C. DELOYA. 1985. Fauna de coleópteros Lamelicornios de Boca de Chajul, Chiapas, México. Folia Entomológica Mexicana 66: 57-118. MORRONE J.J. 2004. Homología biogeográfica: las coordenadas espaciales de la vida. Cuadernos del Instituto de Biología 37, Instituto de biología, UNAM, México, DF. NEITA, M.J., L.L. PARDO, M.D. QUINTO, AND D.N. CUESTA. 2003. Los escarabajos copronecrófilos (Coleóptera: Scarabaeidae) en la parcela permanente de investigación en biodiversidad (PPIB) en Salero, Unión Panamericana, Chocó. In: Cossio FG, Ramos A, eds. Diversidad Biológica de un bosque pluvial tropical. Universidad Tecnológica del Chocó, Instituto de Investigaciones Ambientales del Pacífico. Chocó, Colombia, 79-90. OLSOUFIEFF, G. D’ 1924. Les Phanaeides (Coleoptera-Lamellicornia), Famille Scarabaeidae-Tr. Coprini. Insecta. Revue Illustrée d’ Entomologie 13: 4-172. PADILLA-G IL, D.N., AND G. HALFFTER. 2007. Biogeography of the areas and Canthonini (Coleoptera: Scarabaeidae) of dry tropical forest in Mesoamérica and Colombia. Acta Zoológca Mexicana n.s. (in press.) PHILIPS, T.K., W.D. EDMONDS, AND C.H. SCHOLTZ. 2004. A phylogenetic analysis of the New World Phanaeini (Coleoptera: Scarabaeidae: Scarabaeinae): Hypotheses on relationships and origins. Insect Systematics and Evolution 35: 43-63. PULIDO, H.L.A., C.R.A. RIVEROS, H.F. GAST, AND P. VON HILDEBRAND. 2003. Escarabajos coprófagos (Coleoptera: Scarabaeidae: Scarabaeinae) del Parque Nacional Natural “Serranía de Chiribiquete”, Caquetá, Colombia (Parte I). In: Onore G, Reyes-Castillo P, Zunino M, eds. Escarabeidos de Latinoamérica: Estado del conocimiento Monografías Tercer Milenio, vol 3, SEA, Zaragoza, Spain, 51-58. RIVERA-CERVANTES, L.E., AND G. HALFFTER. 1999. Monografía de las especies mexicanas de Canthon del subgénero Glaphyrocanthon Coleoptera: Scarabaeidae: Scarabaeinae. Acta Zoológca Mexicana n.s. 77: 23-150. RZEDOWSKI, J. 1978. Vegetación de México. Ed. Limusa, México. RZEDOWSKI, J. 1991. El endemismo en la flora fanerogámica mexicana: una apreciación analítica preliminar. Acta Botánica Mexicana 15: 47-64. RZEDOWSKI, J., R. MEDINA LEMOS, AND G. CALDERÒN D E R ZEDOWSKI. 2005. Inventario del conocimiento taxonómico, así como de la diversidad y del endemismo regionales de las especies mexicanas de Bursera (Burseraceae). Acta Botánica Mexicana 70: 85-111. 72 SABOGAL, C., AND L. VALERIO. 1998. Forest composition, structure, and regeneration in a dry forest of the Nicaragua Pacific Coast. In: Dallmeier F, Comiskey JA, eds. Forest biodiversity in North, Central and South America, and the Caribbean Research and Monitoring. UNESCO. Paris & Parthenon Publishing Group, 187-212. SALAS-MORALES, S.H., A. SAYNES-VÁSQUEZ, AND L., SCHIBLI. 2003. Flora de la Costa de Oaxaca, México: Lista florística de la Región de Zimatán. Boletín de la Sociedad Botánica de México 72: 21-58. SWOFFORD, D.L. 2002. PAUP. Phylogenetic Analysis Using Parsimony, Versión 4. Sinauer Associates, Sunderland, Massachuseetts. TREJO, I., AND R. DIRZO. 2000. Deforestation of seasonally dry tropical forest: a national and local analysis in Mexico. Biological Conservation 94: 133-142. 73 Fig.1 Study sites,DTF (1-7 dry forest sites): 1 = Chamela, 2 = Palo Verde, 3 = Neguanje, 4 = Tierra Bomba, 5 = Los Colorados, 6 = Zambrano, 7 = Norte del Tolima; (8-9 humid forest sites): 8 = Los Tuxtlas, 9 = Leticia (Amazonas). The dotted area indicates the distribution of dry tropical forest (alter Padilla & Halffter, 2007) 74 100/83 Chamela P. Verde N. Tolima 67/49 Zambrano 67/49 Los Colorados T. Bomba Neguanje Leticia LosTuxtlas Fig. 2. Cladogram of relationships among dry forest sites, using phanaeines as the indicator group. Tree length = 26, CI= 0.88 y RI= 0.57. Numbers above nodes indicated frequencies in the majority rule consensus tree (of 27 trees)/ Jackknife frequency 20% deletion, 500 replicas. 75 62 Chamela P. Verde 63 N. Tolima Los Colorados Neguanje Leticia Los Tuxtlas Fig.3. Cladogram of relationships of dry forest sites using Phanaeus as the indicator group Single tree length 14, CI= 1.0 y RI= 1.0. Indicated values are Bootstrap, with 500 replicas. 76 APPENDIX 1. Description of study sites (see Fig. 1 and Padilla & Halffter, 2007; 1 – 4 dry tropical forest sites; 5 – 6 wet tropical forest sites) 1) Chamela, México, Jalisco. Biosphere Reserve Chamela-Cuixmala (19º 30’ N and 105º 03’ W) on the SW Pacific coast of Mexico between the Río San Nicolás to the north and the Río Cuitzmala to the south and in the vicinity of Chamela Biological Station; altitude below 200 m, average maximum monthly temperatures 28.8 to 32.2 ºC, minima 15.9 to 22.6 ºC; pronounced rainy season July – November, average precipitation 700 mm. 2) Palo Verde, Costa Rica, Guanacaste. Palo Verde Biological Station is located with the Palo Verde National Park (10º 20’ N and 85º 18’ W) on the Pacific coast in the valley of Río Tempisque; altitude 10 to 50 m; average annual temperature 24.0 to 27.8 ºC; average annual precipitation 1750 mm with a pronounced 6.5 month dry season (December-May). 3) Colombia, Caribbean Region. a) Zambrano: Finca Forestal Monterrey. Department of Bolívar, Municipality of Zambrano, 9º 37’ 48” N, 74º 54’ 44”W, 155 m altitude, annual average precipitation 1048 mm. b) Los Colorados. Department of Bolívar, Municipality of San Juan de Nepomuceno, 9º 51’ 33” N, 75º 06’ 38”W, 300 m altitude, average annual precipitation 1189 mm. This remnant of dry forest lies in the Santuario de Flora y Fauna Los Colorados. c) Isla de Tierra Bomba. Department of Bolívar, Municipality of Cartagena, 10º 21’ 36” N, 75º 34’ 11”W, altitude 50 m, average annual precipitation 789 mm. d) Neguanje. Department of Magdalena, Municipality of Santa Marta, within Parque Nacional Tayrona, 11º 18’05” N and 74º 06’ 11” W’, altitude 300 m, annual average precipitation 1420 m 4) Norte del Tolima. A region along the Río Magdalena 130 km north of Ibagué, municipalities of Honda, Armero-Guayabal and Piedras, 4º 15’ - 5º 10’ N and 74º 45’ - 74º 50’ W; altitude 250 m, average annual temperature 28 ºC, annual precipitation 1387 mm with two well defined dry seasons (December-March and June-August). 77 5) Los Tuxtlas, Veracruz, México. Tropical Biology Station “Los Tuxtlas” on the slopes of Volcán de San Martín, 18 km north of Catemaco, Veracruz, 18º 34’ - 18º 36’ N and 95º 04’ - 95º 09’W, altitude 150 to 530 m, average annual precipitation 4560 mm, average annual temperature 23,7 ºC. 6) Leticia, Amazonas, Colombia. City on the Colombian-Brazilian border, 4º 8’ S 70º 1’ W, altitude 96 m. Average annual precipitation 3500 mm, with heaviest rain from April to June; average annual temperature 26 ºC. 78 APPENDIX 2. Phanaeini occurring in the study areas; numbering is the same as used in Appendix 3. 1. Coprophanaeus jasius (Olivier) 2. C. gilli Arnaud 3. C. parvulus (Olsoufieff) 4. C. pluto (Harold) (including P. p. pluto and P. pluto noguerai) 5. C. telamon (Erichson) 6. Oxysternon silenus Laporte 7. O. conspicillatum Weber 8. Diabroctis cadmus (Harold) 9. Phanaeus bispinus Bates 10. P. cambeforti Arnaud 11. P. chalcomelas (Perty) 12. P. demon Laporte (including P. d. demon and P. demon excelsus) 13. P. endymion Harold 14. P. eximius Bates 15. P. furiosus Bates 16. P. hermes Harold 17. P. meleagris Blanchard 18. P. mexicanus Harold 19. P. prasinus Harold 20. P. sallei Harold 21. P. tridens Laporte 22. P. wagneri Harold 23. Sulcophanaeus chryseicollis (Harold) 79 APPENDIX 3. Cladogram data matrix, study site X taxon (M = Mesoamerica; CR = Costa Rica; C = Colombia) 1 2 123456789 0123456789 0123 Chamela ( M ) 000100000 0010010000 0000 Tuxtlas ( M ) 010010000 0001000010 1101 P. Verde (CR ) 000110000 0010100000 0010 N. Tolima ( C ) 000000000 0000001000 0000 Zambrano ( C ) 100000000 0000000000 0000 Los Colorados ( C ) 100000010 0000001000 0000 T. Bomba ( C ) 000000010 0000000000 0000 Neguanje ( C ) 000000000 0000000001 0000 Leticia ( C ) 001011101 1100000100 0000 80 APPENDIX 4. Phanaeus, Coprophanaeus and Canthon (Glaphyrocanthon) viridis group confined to (a) the mountainous area of the Mexican Transition Zone and (b) the Pacific coastal dry forests, respectively: Phanaeus, Coprophanaeus: (a) P. daphnis Harold, P. tridens Laporte, P. endymion Harold, P. halffterorum Edmonds, P. adonis Harold, P. scutifer Bates, P. flohri Nevinson, P. mexicanus Harold, P. amithaon Harold, P. palliatus Sturm, P. quadridens (Say) y P. damocles Harold; (b) Coprophanaeus nogueirai Arnaud, OR C. pluto (Harold), Phanaeus demon Laporte, P. furiosus Bates, P. nimrod Harold, P. zapotecus Edmonds Canthon (Gl.) viridis group: (a) C. circulatus Harold, C. leechi (Martínez, Halffter & Halffter), C. vazquezae (Martínez, Halffter & Halffter), C. viridis (Palisot de Beauvois), C. antoniomartinezi Rivera-Cervantes & Halffter, C. montanus Rivera-Cervantes & Halffter, C. manantlanensis Rivera-Cervantes & Halffter; (b) C. corporali Balthasar, C. delgadoi Rivera-Cervantes & Halffter, C. pacificus Rivera-Cervantes & Halffter, C. deloyai Rivera-Cervantes & Halffter. 81 CAPÍTULO IV Discusión y Conclusiones Metodología utilizada El grupo externo se elige tomando en cuenta un grupo real o un taxón hipotético que el investigador considera como primitivo. Sin embargo, en los análisis del PAE y del cladograma geológico, tradicionalmente se ha venido empleando como grupo externo una localidad hipotética que no tiene taxa, ante la imposibilidad de definir un área como “área primitiva” (ver Rosen 1988). En este trabajo se eligió un grupo externo que corresponde a una comunidad con características ecológicas contrastantes con el bosque tropical seco: el bosque tropical húmedo. Se cuenta con hipótesis previas sobre el posible origen de los bosques neotropicales objeto de estudio: Sarmiento (1975) propone la evolución independiente de las formaciones vegetales secas de Sudamérica en el Cuaternario. Hooghiemstra & van der Hammen (2001) plantean el desarrollo del bosque húmedo tropical sudamericano desde el Mioceno hasta el Cuaternario. Respecto a la región de los Tuxtlas, tanto el vulcanismo Plio-Cuaternario, como los registros palinológicos de Paraje Solo de Veracruz, sugieren que la selva alta perennifolia tiene un origen Plio-Pleistoceno (ver Ferrari et al. 2002, Graham 2003). Presuponemos que éstos bosques húmedos son más antiguos que los bsT. Por otra parte los bosques neotropicales elegidos hacen parte del contexto geográfico y ecológico donde se distribuyen los grupos indicadores (Canthonini y Phanaeini). Los análisis utilizando el grupo externo permitieron la polarización de los cladogramas, la adición de caracteres y particularmente en el cladograma geomorfológico, una dimensión espacio-temporal con el ordenamiento de algunos caracteres, logrando así una solución más parsimoniosa que la que se hubiera obtenido enraizando con un área hipotética codificada en ceros. En suma, el grupo externo permitió establecer un eje comparativo que trascendió en todo el trabajo. En el PAE se adicionaron jerarquías taxonómicas (vg. subgéneros), que en el caso de Phanaeini no afectaron la estructura del árbol y en los Canthonini aportaron nuevos elementos al análisis. Los resultados del trabajo se ven apoyados al observar la similitud faunística de los Canthonini y Phanaeini entre las diferentes localidades a través del índice de Jaccard (Cuadros 1 y 2). Sin embargo, los dendrogramas obtenidos utilizando el índice anterior, muestran grupos formados por ausencia de taxones, sin significado en términos biogeográficos. 82 El origen del bsT de México Según Becerra (2005) el origen del bsT de México ocurrió sincrónicamente con la diversificación del género Bursera hace 20 a 30 Ma (millones de años). Lo cual es discutible, al reinterpretar las gráficas presentadas por Becerra, considerar otros trabajos florísticos (vg. Rzedowski et al., 2005) y confrontarlos con los análisis obtenidos en este trabajo. Los resultados obtenidos por Becerra nos cuentan una historia de extinción y diversificación de las distintas especies del género, al parecer sincrónicos, con varios eventos geológicos y climáticos iniciados probablemente a fines del Cretácico (~ 70 Ma) y que dejaron su huella en los fósiles que datan del Terciario temprano en Colorado-Wyoming (USA). En el Mioceno, las gráficas indican dos procesos, uno probablemente de extinción (~ 15 Ma), sincrónico con la etapa final del levantamiento de la Sierra Madre Occidental y otro de diversificación situado geográficamente en los estados de Nayarit y Sinaloa. En el Plioceno se presentan las más altas tasas de diversificación, cuya área corresponde a la propuesta por Rzedowski et al. (2005) en altitudes medias y bajas de la vertiente pacífica de México (Jalisco a Michoacán). En el Pleistoceno a Reciente en los bsT de Oaxaca continúan su proceso de diversificación. Estos últimos coinciden con el área de bsT objeto de estudio y por tanto son tomados como apoyo a la propuesta planteada por Padilla-Gil & Halffter (2007) sobre el origen de los bsT en el Pleistoceno una vez que ocurrieron los eventos orogénicos que culminaron a finales del Plioceno: el ensamble geológico de parte de la cuenca del río Balsas (en el Pleistoceno) y el probable establecimiento de éstas y otras especies vegetales en tal paisaje. Rzedowski et al. (2005) confirma la existencia de Bursera en todos los estados de México, con excepción de Tlaxcala; igualmente indican el estado de Guerrero como el territorio más rico en componentes de Bursera, seguido por Michoacán, Oaxaca y a mayor distancia Jalisco y Puebla. En el estado de Guerrero se encuentra la parte baja de la cuenca del río Balsas, la cual presenta las más altas tasas de diversificación del género. La evolución más reciente y máxima diversificación en la cuenca oeste del río Balsas se presenta en el límite oeste de los estados de Guerrero y Michoacán. Las especies de Bursera en Sudamérica están limitadas a bosque seco espinoso principalmente del sur de Ecuador y Perú. Este género no es característico ni de los bosques secos de Colombia ni de Las Caatingas en Brasil (ver Sarmiento 1975, Gentry 1995, Pennington et al. 2000, Mendoza 1999, Albesiano & Rangel-Ch 2003). 83 Los puntos anteriores indican que las especies del género Bursera han sido testigo de los cambios histórico-geológico y climáticos acontecidos desde el Cretácico hasta el presente en los ecosistemas secos (bosque espinoso, matorral xerófilo, bsT) en el subtrópico y trópico de Mesoamérica. Sin embargo, la probable biogeografía histórica de un sólo género de plantas no es suficiente para explicar el origen de este ecosistema en México. Se plantea el origen de todo el conjunto de los bosques secos de Mesoamérica y Colombia, tomando en cuenta las características de los bosques secos de México y sus afinidades florísticas: 1) composición florística donde predominan las familias Fabaceae, Euphorbiaceae, Burseraceae (Trejo 2005); 2) Endemismo de las especies de los géneros Bursera, Acacia, Euphorbia e Ipomea (Rzedowski 1991); 3) clima, con condiciones de lluvia marcadamente estacionales; 4) la vertiente pacifica con un clima más cálido que el del Golfo de México debido tanto a su ubicación latitudinal como al levantamiento de los sistemas montañosos Eje Volcánico Transversal y Sierra Madre del Sur (ver Challenger 1998); 5) las afinidades florísticas del bosque seco de Chamela con los de la región fitogeográfica del norte de Colombia y Venezuela (Gentry 1995). Origen de los bsT de Mesoamérica y Colombia El origen de los bosques secos neotropicales de los enclaves estudiados es consecuencia de varios eventos geográficos y geomorfológicos que confluyen espacio-temporalmente hacia fines del Plioceno y Pleistoceno, como: la ubicación latitudinal y la influencia de la formación del Istmo de Panamá; el ensamble geológico de tales sitios: ensamble de la región Caribe de Colombia y el ensamble de los terrenos en la cuenca del río Balsas (Pleistoceno) y los procesos orogénicos de México continental hacia finales del Plioceno. Por otra parte la característica más sobresaliente del bsT es la estacionalidad, relacionada con la desigual distribución de la precipitación a lo largo del año. El principal generador de lluvias en el trópico es la ZCIT: Zona de Convergencia Intertropical (ver para México, Challenger 1998; Costa Rica, Kohlmann et al. 2002; región Caribe de Colombia, Molina 1996). La ZCIT se ubica al sur del ecuador terrestre. El norte de Sudamérica y el mar Caribe reciben relativamente poca precipitación, en comparación con la que reciben la Amazonia central y sur (Hooghiemstra et al. 2002). Otro factor que modifica el clima tropical, especialmente en la vertiente del pacífico es La Oscilación del Sur o del Niño (ENSO), trayendo como consecuencia perturbación atmosférica y redistribución de las lluvias. Al parecer la posición actual de la ZCIT se produjo durante el 84 Holoceno y el ENSO fue ausente o débil durante este mismo período. Con base en estos argumentos Padilla & Halffter (2007) proponen que las actuales condiciones secas se fijaron durante el Holoceno Medio en la Región Caribe de Colombia, así como en la Costa Pacifica de México y Costa Rica. El análisis geomorfológico de los enclaves de bsT de Mesoamérica y Colombia pone en evidencia la evolución espacio-temporal de los bosques secos desde el Pleistoceno, así mismo indica la influencia de los factores climáticos en el origen y evolución de esta comunidad. Esta hipótesis está de acuerdo con: 1) la similaridad en la composición y estructura florística de los bosques secos estudiados (ver Cuadro 3). En todos predominan las Fabaceae, Bignoniaceae y los géneros más diversos en cada bosque son: en Chamela Ipomea, en Guanacaste Cassia y en los bsT de Colombia Capparis (ver Gentry 1995, Mendoza 1999, Quigley & Platt 2003); 2) con la escasez de elementos neárticos en la flora de los bsT (Rzedowski 1978, Quigley & Platt 2003); 3) con las ideas de Sarmiento (1975), quien a partir del análisis de símilitud de géneros de plantas de formaciones vegetales secas de Sudamérica hipotétiza sobre la posible conformación de la comunidad de bsT del sur de México hasta el norte de Sudamérica, con posible origen en el Pleistoceno; 4) con el proceso de especiación de Bursera desde el Plioceno al Reciente (ver primer punto de la discusión); 5) con las afinidades florísticas y de Canthonini y Phanaeini presentes en los bsT de Mesoamérica y Colombia. Relación entre la distancia de los bosques y la similitud faunística Las distancias entre los enclaves de los bosques escogidos influyen ligeramente en la similitud faunística de los Canthonini y los Phanaeini. La correlación (p < 0,05) es negativa, pero estadísticamente no es significativa (ver Cuadro 1 y 2). En los Canthonini, el 30% y en los Phanaeini el 10 %, de la variación de la distancia entre los bosques se relaciona con la variación en la similitud de los taxones. Las afinidades faunísticas de cada uno de los grupos con los bosques están relacionadas con los procesos evolutivos y biogeográficos considerados en este trabajo. Biogeografía de los Canthonini y Phanaeini de los bosques neotropicales estudiados Las relaciones de los Canthonini y en especial del género Canthon en los bosques secos de Mesoamérica y Colombia se explican mediante tres patrones biogeográficos, lo que evidencia la 85 dinámica espacio/temporal de este grupo en los bosques secos. Por otra parte, se confirma tanto el origen como la dispersión de la mayor parte de las especies a partir de Sudamérica. Las especies de Canthon (Gl.) grupo viridis presentan una explosión de especiación en México. Los escenarios en que ocurre el fenómeno corresponden a zonas de altitud baja y media en las montañas de la ZTM y a los bosques tropicales de baja altitud. Las especies de los bsT son más numerosas que las de los bhT. La diversificación se inicia desde finales del Plioceno al Reciente. Lo cual coincide tanto con el fin de los procesos orogénicos de México continental, como con la formación de los bosques secos de México. De acuerdo con los resultados de este trabajo es probable que las especies de C. (Gl.) grupo viridis de los bosques secos de la vertiente pacífica de México, así como las especies de Canthon de especiación in situ de Costa Rica y de Colombia se hayan especiado en cada uno de estos bosques, desde el Pleistoceno, concomitante con la evolución y desarrollo de la comunidad de bosque seco tropical. Los Phanaeini de los bsT de México y Costa Rica presentan mayor afinidad como los de los bosques secos de la región Caribe y de los valles interandinos de Colombia. Tanto por las afinidades taxonómicas como geográficas, los Phanaeini de los bosques secos estudiados constituyen comunidades endémicas que probablemente evolucionaron sincrónicamente con los bsT desde el Pleistoceno. El estudio comparativo de los patrones de distribución de Canthonini y de Phanaeini en los bsT, revela la especiación mucho más reciente de los Phanaeini y evidencia la posible diversificación sincrónica espacio/temporal del género Phanaeus y del subgénero Canthon (Glaphyrocanthon) grupo viridis en México. Los Phanaeini y los Canthon de los bsT de la vertiente Pacífica de México presentan una composición característica que los diferencia tanto de los bsT de la vertiente del Golfo de México, como de las especies que se distribuyen en las montañas de la ZTM. Los bsT de Mesoamérica y los del norte de Sudamérica están constituidos principalmente por elementos neotropicales y presentan afinidades florísticas entre sí como lo señaló Gentry (1995) el bsT de Chamela comparte 15 géneros con la región fitogeográfica del norte de Colombia y Venezuela. Son reconocidas las especies provenientes tanto de Centroamérica como de Sudamérica en los bosques secos de Oaxaca (Salas-Morales et al. 2003). Este trabajo muestra que los Canthonini presentan las mismas afinidades en los bsT de Mesoamérica y Colombia 86 (evidencia apoyada por el índice de Jaccard, Cuadro 1) y al igual que ocurre con las plantas tanto los Canthonini como los Phanaeini en la vertiente pacífica de México presentan las tasas más altas de especiación in situ. Sin duda, numerosas plantas y animales como los Canthonini y Phanaeini, existieron antes de integrarse a la comunidad de bsT. La comunidad que conocemos hoy como bsT probablemente inicio su evolución a partir del Plioceno y se integra como conjunto a partir del Pleistoceno cuando ocurrieron todos los procesos geomorfológicos descritos anteriormente en cada uno de los bsT, así como los procesos de dispersión y especiación tanto florísticos como faunísticos. Los bosques secos de Mesoamérica y Colombia constituyen un escenario interesante para los procesos evolutivos y de dispersión de múltiples especies tanto de la flora fanerogámica como de la entomofauna. Es destacable la vertiente seca del Pacífico y la cuenca del río Balsas en México, como centro de especiación de algunos Canthon y de Phanaeus, así como el corredor formado por los bosques secos desde el sur de México hasta Costa Rica, ámbito de distribución de parte de las especies de Canthonini y Phanaeini. Los bsT de la vertiente pacífica de México son heterogéneos. Los bosques secos de Jalisco a Michoacán y los de Oaxaca, particularmente presentan las tasas más altas tanto de diversidad como de endemismo. Esto se ve reflejado por la especiación diferencial espacio/temporal de las especies de Bursera (véase primer punto de la discusión); por la alta riqueza florística de los bsT de Michoacán (Caleta e Infiernillo) equiparable con la de los bosques de Oaxaca (Copalita) y la alta diversidad beta (ver Trejo 2005); por la estructura y composición característica de los bosques secos de Oaxaca (Salas-Morales et al. 2003, Gallardo-Cruz et al. 2005) y por las especies de Phanaeus y de Canthon circunscritas a los bosques secos de Oaxaca son P. nimrod, P. zapotecus y C. (Gl.) delgadoi. Las otras especies se encuentran ampliamente distribuidas en la vertiente pacífica de México. Análisis de los Canthonini y Phanaeini de los bsT y bhT La mitad de las especies de Canthon presentes en la selva de Los Tuxtlas (ver capitulo 2, apéndice 1) mantienen su rango de distribución geográfico hasta Sudamérica. C. (Gl.) subhyalinus, C. (C.) morsei y C. (Gl.) edmondsi; están asociadas a bhT y probablemente han colonizado de forma secundaria algunos lugares de bsT: C. morsei se encuentra en los bosques secos del norte de Tolima y en Palo Verde; C. subhyalinus en algunos bosques secos de la región 87 Caribe de Colombia y en norte del Tolima. Es interesante anotar que ninguna de estas tres especies fue registrada en Leticia. Es muy probable que tanto los Canthonini incluidos en el segundo patrón de distribución de los bsT, así como las especies citadas anteriormente de la selva de los Tuxtlas, se hayan diversificado en el norte de Sudamérica, una vez finalizó el levantamiento del norte de los Andes (Mioceno Medio) y se establecieron las cuencas actuales de los ríos Magdalena, Orinoco y Amazonas. Esta explicación esta de acuerdo con: 1) la afinidad taxonómica de las especies planteadas en este patrón de distribución, 2) su asociación a los bosques neotropicales (secos y húmedos de baja altitud), 3) la afinidad neotropical de los Canthonini de los bosques tropicales de tierras bajas y bosques tropicales subandinos de los Andes del norte de Sudamérica, 4) la disimilitud con los bsT del este de Bolivia y otras formaciones vegetales de ambiente seco del sur de Sudamérica (Cerrado y Chaco), 5) la época de cierre del Istmo de Panamá, 6) la ausencia de estas especies en las Antillas, 7) los diferentes rangos de distribución alcanzados por las especies, en sentido SurNorte, 8) la hipótesis también está de acuerdo con la probable expansión posterior desde la mitad más austral de Sudamérica de algunas especies como Canthon lituratus, Canthon mutabilis, Canthon septemmaculatus. Ninguna especie de Canthonini de Leticia es compartida ni por los bsT de Palo Verde y Chamela ni por Los Tuxtlas (ver Cuadro 1). Por otra parte C. aequinoctialis presente en Leticia probablemente ha colonizado los bosques secos del norte del Tolima y los Colorados en la región Caribe de Colombia. La Amazonía parece ser el escenario para las especies que evolucionaron según la hipótesis antes propuesta. Esto esta de acuerdo con la expansión hacia el norte de C. aequinoctialis que llega hasta Belice y con C. angustatus, presente en Leticia pero ausente en los otros bosques estudiados. Esta especie asociada a bhT, se distribuye hasta el sur de México, Selva Lacandona (Halffter com. pers.), así como de especies con especiación in situ, más recientes y con rango geográfico restringido a la Amazonía. Los Canthonini tanto de los bsT como de los bhT analizados comparten probablemente la misma historia biogeográfica. Sin embargo el número mayor de especies asociadas a los bsT, así como la mayor especiación en este ecosistema permiten reconocer a los Canthonini como característicos de los bosques secos neotropicales. Por otra parte el origen y diversificación de los primates neotropicales y especialmente del 88 género Alouatta (ver Schneider, 2000; Cortés-Ortiz et al., 2003) coincide espacio-temporalmente con el posible origen de los Canthonini planteado en la hipótesis anterior. Esto en parte explicaría la asociación de algunas de estas especies de Canthon (vg. C. subhyalinus) con A. palliata y posiblemente su distribución geográfica (ver Padilla-Gil & Halffter 2007). En la selva de Los Tuxtlas, Phanaeus endymion y Sulcophanaeus chryseicollis están asociadas a ecosistemas de bhT y muestran rango geográfico restringido a Mesoamérica la primera y al sureste de México la segunda. Las especies de Phanaeus presentes en Leticia están circunscritas a la región de la Amazonia (ver Cuadro 2); las de Oxysternon presentan rango geográfico hasta Centroamérica siempre asociadas a bhT. La asociación de algunas especies de Phanaeini a ecosistemas secos tanto de Mesoamérica como de Sudamérica es interesante desde el punto de vista ecológico y geográfico; al parecer evolucionaron sincrónicamente con la aparición de dichos ecosistemas y presentan rangos geográficos restringidos. Las especies de Phanaeini de los bhT Los Tuxtlas y de Leticia presentan la misma tendencia biogeográfica, especiación in situ, probablemente a partir del Pleistoceno. Al igual que en las plantas, la integración de los escarabajos estudiados al bsT es un fenómeno geológicamente reciente, como lo evidencia la dinámica espacio temporal de los Canthonini y la especiación de los Phanaeini. Sin lugar a dudas, tanto los Canthonini como muchas plantas de origen sudamericano han experimentado procesos similares espaciotemporalmente desde el Plioceno para llegar integrar la comunidad que hoy forma el bsT de Mesoamérica y Colombia. Gradiente de distribución en Costa Rica De acuerdo con la pluviosidad el oeste de Costa Rica queda dividido: al norte predomina el bsT y el promedio anual de lluvias es menor a 2000 mm (provincia de Guanacaste). Al sur con predominio del bhT, con promedio de lluvia mayor a 2000 mm anuales, delimitada al norte por el río Grande de Tárcoles y al sur por el río Grande de Térraba y las cordilleras Central y de Talamanca. Debido a este contraste climático y de vegetación la dispersión de algunas especies en ambas direcciones norte-sur se ve restringido. Algunos Phanaeini con afinidad por los bsT de México penetran hasta Guanacaste y algunas especies de Canthon de Sudamérica no llegan a Guanacaste, alcanzando en ambos casos diferentes grados de penetración. Al parecer hay otras especies tanto 89 de animales como de vegetales que presentan también sus límites geográficos en esta área. Consideraciones finales Es la primera vez que se analiza la biogeografía de los Canthonini y Phanaeini considerando las especies de Colombia y Mesoamérica así como su asociación a un tipo de vegetación, en este caso, al bosque tropical seco. El trabajo revela el predominio de los Canthonini como un fenómeno biogeográfico característico de los bosques secos y la integración espacio-temporal de estos Scarabaeinae en la reconstrucción de la biogeografía histórica del bsT. El análisis geomorfológico de los enclaves de bosque seco y la convergencia de los elementos que regulan el clima en la zona neotropical y su posible localización y acción en tiempo geológico reciente permitieron integrar, sintetizar y plantear la hipótesis sobre el origen de los bosques secos de Mesoamérica y Colombia. Se proporcionan varios elementos que contribuyen a la reconstrucción de la biogeografía histórica de la comunidad de bosque seco de Mesoamérica y Colombia, tales como el análisis geomorfológico de los enclaves de bsT, el planteamiento de la hipótesis sobre su origen, la comparación tanto con las hipótesis previas sobre su origen como de afinidades de los bsT basados en la vegetación y los patrones de distribución de Canthonini y Phanaeini. Las plantas, los Canthonini y los Phanaeini de bsT han experimentado un proceso de evolución para llegar a integrar la comunidad que es el bsT. Las plantas y los Canthonini estudiados coinciden en presentar un componente sudamericano de afinidad tropical y las plantas junto con los Canthonini y Phanaeini una alta evolución in situ, principalmente en los bsT de la vertiente pacífica de México. Se contribuye al conocimiento de la biogeografía de los Canthonini y Phanaeini con base en la reunión e integración de varios elementos tales como: el rango geográfico de las especies; las relaciones de afinidad de los grupos taxonómicos en los bsT y/o la conformación de comunidades endémicas de bsT; las relaciones filogenéticas y/o taxonómicas de cada grupo; el origen de los bosques secos (escenario donde se presentan). La integración de los anteriores elementos permitieron el planteamiento de los patrones de distribución y ubicarlos espaciotemporalmente, así como la comparación entre los géneros Canthon y Phanaeus circunscritos a patrón de especiación in situ en Mesoamérica. Algunas especies de Bursera, de Canthon (vg. especies de C. (Gl.) grupo viridis) y los 90 Phanaeini de bsT, al parecer se encuentran en procesos de especiación y es posible que el mismo proceso este ocurriendo en los bhT Los Tuxtlas y Leticia. El proceso de expansión de las especies de Canthon desde Sudamérica hacia Mesoamérica continúa y es posible que futuras investigaciones refuercen este patrón de distribución. La generalización del segundo patrón de distribución de los Canthonini con relación a las especies de los bhT, teniendo en cuenta las afinidades taxonómicas y el contexto biogeográfico de las especies de este grupo en Sudamérica permitieron formular la hipótesis sobre el posible origen espacio-temporal de las especies que integran este patrón de distribución. La realización de estudios similares con otros grupos de insectos u otros organismos en los bsT de Mesoamérica y Colombia permitirán la adición de más elementos en la reconstrucción de la biogeografía histórica del bosque seco. 91 Cuadro 1. Canthonini, índice de similitud de Jaccard/ distancia (km) aproximada de los bosques tropicales objeto de estudio. Chamela Chamela Los Tuxtlas Palo Verde Tolima Zambrano Los Colorados Tierra Bomba Neguanje Tuxtlas 0.25/1177 PVerde Tolima Zamb LCol TBom Neguanje Leticia 0.29/2544 0.32/4214 0.32/4088 0.18/4080 0.18/4082 0.11/4230 0.11/5218 0.15/1418 0.14/3017 0/2930 0/2910 0.09/2972 0/3151 0/4111 0.21/1598 0.13/1540 0/1508 0.08/1515 0/1705 0/2680 0.2/578 0.17/655 0.17/701 0.09/767 0.09/1111 0.4/103 0.4/166 0.33/217 0.33/1593 0.6/163 0.6/1722 0.25/101 0.33/171 0.33/1778 1/1785 Leticia 92 Cuadro 2. Phanaeini, índice de similitud de Jaccard/ distancia (km) aproximada de los bosques tropicales objeto de estudio. Chamela Chamela Los Tuxtlas Palo Verde Tolima Zambrano Los Colorados Tierra Bomba Neguanje Tuxtlas 0.3/1177 PVerde Tolima Zamb LCol TBom 0.3/2544 0.22/4214 0.04/4088 0.04/4080 0.13/4082 0.04/4230 0.04/5218 0/1418 0.33/3017 0/2930 0/2910 0/2972 0/3151 0/4111 0.09/1598 0/1540 0/1508 0/1515 0/1705 0/2680 0/578 0/655 0/701 0/767 0/1111 0/103 0.33/166 0/217 0/1593 0.33/101 Neguanje Leticia 0/163 0/1722 0.33/171 0/1778 0/1785 Leticia 93 Cuadro 3. Comparación de la diversidad florística de los bsT , muestreos de 0.1 ha (DAP 2.5cm) (Gentry 1995, Mendoza 1999, Quigley & Platt 2003) Localidad individuos Precip. Temperatura meses Área basal Alt. del dosel N. de familias N. de especies N. de mm/año °C secos m2/0.1 ha m Zambrano Los Colorados Tierra Bomba Neguanje Norte del Tolima Chamela (Tierras altas 1) Chamela (Tierras altas2) Chamela (Arroyo) Guanacaste 1048 1189 789 1420 1387 748 748 748 24 28 28 27 28 25 25 25 6-7 6-7 6-7 11.4 7 6-8 6-8 6-8 2.31 6.15 2.77 3.68 3.94 2.63 2.18 5.10 7-9 22-25 7-9 22-25 10-18 15 24 41 26 31 31 37 34 46 55 121 56 67 62 91 89 103 471 534 556 337 515 399 506 453 (bosque de galería) 1600 25 6 4.16 20 35 63 95 individuos 94 Literatura citada Albesiano, S. & J. O. Rangel-Ch. 2003. La vegetación del río Chicamocha (Santander, Colombia). Caldasia 25 (1): 73-99. Andersson, L. 1996. An ontological dilemma: epistemology and methodology of historical biogeography. J. Biogeography 23: 269-277 Andresen, E. 2005. Effects of season and vegetation type on community organization of dung beetles in a tropical dry forest. Biotropica 37 (2): 291-300. Becerra, J. X. 2005. Timing the origin and expansion of the Mexican tropical dry forest. PNAS. 102 (31): 10919-10923. Challenger, A. 1998. Utilización y conservación de los ecosistemas terrestres de México, pasado, presente y futuro. CONABIO. México. Cortés-Ortiz, L., E. Birmingham, C. Rico, E. Rodríguez-Luna, I. Sampaio, & M. Ruiz-García. 2003. Molecular systematics and biogeography of the Neotropical monkey genus, Alouatta. Mol. Phylogenet. Evol. 26: 64-81. Craw, R. 1988. Continuing the syntesis between panbiogeography, phylogenetic systematic and geology as illustrated by empirical studies on the biogeography of New Zeland and the Catham Islands. Syst. Zool. 37 (3): 291-310. Craw, R. C, J. R. Grehna & M. J. Heads. 1999. Panbiogeografía: Tracking the history of life. Oxford University Press, Nueva York. Crisci, J. V. & J. J. Morrone. 1992. Panbiogeografia y biogeografía cladistica: Paradigmas actuales de la biogeografía histórica. Ciencias 6: 87-97. Crisci, J. V., O. E. Sala, L. Katinas & P. Posadas. 2006. Bridging historical and ecological approaches in biogeography. Australian Systematic Botany 19: 1-10. Croizat, L. 1958. Panbiogeography. Published by the author, Caracas, Venezuela. Ebach, M. C. & J. J. Morrone. 2005. Forum on historical biogeography: what is cladístic biogeography? J. Biogeography 32: 2179-2187. Ferrari, L., T., Orozco, T. Tagami, M. Eguchi, C. Petrone, & J. J.Albarrán. 2002. La “Provincia alcalina oriental” revisitada I: Geología regional. Geos., Unión Geofísica Mexicana A. C., 152. Gallardo-Cruz, J. A., Meave, J. A. & Pérez-García, E. A. (2005) Estructura, composición y diversidad de la selva baja caducifolia del Cerro Verde, Nizanda (Oaxaca), México. Boletín 95 de la Sociedad Botánica de México 76: 19-35. Gentry A. H. 1995. Diversity and floristic composition of neotropical dry forest. In: Bullock S, Mooney H. A, Medina E., eds. Seasonally dry Tropical Forest. Cambridge University Press. Cambridge, Massachussets, 146-194. Graham, A. 2003. In the Beginning: Early Events in the Development ofMesoamerica and the Lowland Maya Area. In A. Gómez-Pompa, M. F. Allen, S. L. Fedick, J.J. Jiménez-Osorio (eds.), The Lowland Maya Area three millennia at the human-wildland human-wildland interface, pp. 31-44. Food Products Press an Imprint of the Haworth Press, Inc. New York. Hooghiemstra H & T. van der Hammen. 2001. Desarrollo del bosque húmedo neotropical en el Neógeno y Cuaternario: la hipótesis de los refugios. 2001, pp. 129-136. In LlorenteBousquets J y Morrone (eds.), Introducción a la biogeografía en Latinoamérica. Teorías, conceptos, métodos y aplicaciones. Facultad de Ciencias. UNAM. Hooghiemstra H, van der Hammen, T., Cleef, A. 2002. Paleoecología de la flora boscosa. pp. 4358 In M. R. Guariguata and G. H. Katan (eds.), El bosque neotropical pasado y presente (Sección I) Ecología y conservación de bosques neotropicales Costa Rica. Instituto Alexander von Humboldt (IAvH). 1997. Caracterización ecológica de cuatro remanentes de bsT de la región Caribe colombiana. Grupo de Exploraciones Ecológicas Rápidas, IAvH. Villa de Leyva. Kohlmann, B., Wilkinson, J. & K. Lulla. 2002. Costa Rica desde el espacio. Universidad EARTH. San José, Costa Rica. Martín-Piera, F. & I. Sanmartín. 1999. Biogeografía de áreas y biogeografía de Artrópodos Holárticos y Mediterráneos, pp. 535-560. In A. Melic, J. J. De Haro, M. Mendez & I. Ribera (eds.), Evolución y Filogenia de Artrópoda. Boletín de la SEA 26. Zaragoza, España. Mendoza, C. H. 1999. Estructura y riqueza florística del bsT en la región Caribe y el Valle del Río Magdalena, Colombia. Caldasia 21(1): 70-94. Molina, L. E. (ed.). 1996. Geomorfología y Aspectos Erosivos del Litoral Caribe Colombiano. INGEOMINAS. Cartagena, Colombia. Morrone, J. J., D. Espinosa & J. Llorente. 1996. Manual de biogeografía histórica. UNAM, México. Morrone J. J. 2004. Homología biogeográfica: las coordenadas espaciales de la vida. Cuadernos del Instituto de Biología 37, Instituto de biología, UNAM, México, DF. 96 Padilla-Gil DN & G. Halffter. 2007. Biogeography of the areas and Canthonini (Coleoptera: Scarabaeidae) of dry tropical forest in Mesoamérica and Colombia. Acta Zoológca Mexicana n.s. (in press.) Pennington, R. T., D. E. Prado & C. A. Pendry. 2000. Neotropical seasonally dry forests and Quaternary vegetation changes. J. Biogeography 27: 261-273. Quigley, M. F., & W. J. Platt. 2003. Composition and structure of seasonally deciduous forest in the Americas. Ecol. Monogr. 73 (1):87-106. Rosen, B. R. 1988. From fossil to earth history: applied historical biogeography. In A. A. Myers, & P. S. Giller (eds.), Analytical Biogeography, pp. 437-481. Chapman & Hall. London, UK. Rzedowski, J. 1978. Vegetación de México. Ed. Limusa, México. Rzedowski, J. 1991. El endemismo en la flora fanerogámica mexicana: una apreciación analítica preliminar. Acta Botánica Mexicana 15: 47-64. Rzedowski, J., R. Medina Lemos & G. Calderón de Rzedowski. 2005. Inventario del conocimiento taxonómico, así como de la diversidad y del endemismo regionales de las especies mexicanas de Bursera (Burseraceae). Acta Botánica Mexicana 70: 85-111. Salas-Morales, S. H., Saynes-Vásquez, A. & Schibli, L. (2003) Flora de la Costa de Oaxaca, México: Lista florística de la Región de Zimatán. Boletín de la. Sociedad Botánica de México 72: 21-58. Sarmiento, G. 1975. The dry plant formations of South America and their floristic conections. J. Biogeography 2: 233-251. Schneider, H. 2000. The current status of the New World monkey phylogeny. An. Acad. Bras. Ciênc. 72 (2): 165-172. Spellerberg, I. F. & Sawyer, J. W. D. 2000. An introduction to applied biogeography. University Press, Cambridge. United Kingdom. Trejo, I. 2005. Análisis de la diversidad de la selva baja caducifolia en México, pp. 111-122. In G. Halffter, J. Soberón, P. Koleff & A. Melic (eds.), Sobre Diversidad biológica: el significado de las Diversidades Alfa, Beta y Gama. Monografías Tercer Milenio, SEA, Vol. 4, Zaragoza. Wiley, E. O. 1988. Vicariance biogeography. Ann Rev Ecol Syst 19: 513-542. 97