UNIVERSIDAD NACIONAL AUTÓNOMA DE MÉXICO POSGRADO EN CIENCIAS BIOLÓGICAS INSTITUTO DE BIOLOGÍA SISTEMÁTICA SISTEMÁTICA Y BIOGEOGRAFÍA DEL COMPLEJO DE ESPECIES CRAUGASTOR PODICIFERUS (ANURA: STRABOMANTIDAE) EN AMÉRICA CENTRAL ÍSTMICA EMPLEANDO ADN MITOCONDRIAL Y NUCLEAR TESIS QUE PARA OPTAR POR EL GRADO DE: DOCTOR EN CIENCIAS PRESENTA: ERICK ARIAS PIEDRA TUTORA PRINCIPAL DE TESIS: DRA. GABRIELA PARRA OLEA INSTITUTO DE BIOLOGÍA, UNAM COMITÉ TUTOR: DR. ADRIÁN NIETO MONTES DE OCA FACULTAD DE CIENCIAS, UNAM DR. JOSÉ MARTÍN GARCÍA VARELA INSTITUTO DE BIOLOGÍA, UNAM Cd. Mx. FEBRERO, 2019 UNAM – Dirección General de Bibliotecas Tesis Digitales Restricciones de uso DERECHOS RESERVADOS © PROHIBIDA SU REPRODUCCIÓN TOTAL O PARCIAL Todo el material contenido en esta tesis esta protegido por la Ley Federal del Derecho de Autor (LFDA) de los Estados Unidos Mexicanos (México). El uso de imágenes, fragmentos de videos, y demás material que sea objeto de protección de los derechos de autor, será exclusivamente para fines educativos e informativos y deberá citar la fuente donde la obtuvo mencionando el autor o autores. Cualquier uso distinto como el lucro, reproducción, edición o modificación, será perseguido y sancionado por el respectivo titular de los Derechos de Autor. UNIVERSIDAD NACIONAL AUTÓNOMA DE MÉXICO POSGRADO EN CIENCIAS BIOLÓGICAS INSTITUTO DE BIOLOGÍA SISTEMÁTICA SISTEMÁTICA Y BIOGEOGRAFÍA DEL COMPLEJO DE ESPECIES CRAUGASTOR PODICIFERUS (ANURA: STRABOMANTIDAE) EN AMÉRICA CENTRAL ÍSTMICA EMPLEANDO ADN MITOCONDRIAL Y NUCLEAR TESIS QUE PARA OPTAR POR EL GRADO DE: DOCTOR EN CIENCIAS PRESENTA: ERICK ARIAS PIEDRA TUTORA PRINCIPAL DE TESIS: DRA. GABRIELA PARRA OLEA INSTITUTO DE BIOLOGÍA, UNAM COMITÉ TUTOR: DR. ADRIÁN NIETO MONTES DE OCA FACULTAD DE CIENCIAS, UNAM DR. JOSÉ MARTÍN GARCÍA VARELA INSTITUTO DE BIOLOGÍA, UNAM MÉXICO, Cd. Mx. FEBRERO, 2019 o M A COORDINACIÓN ns Le q 24 POSGKR/TIDO 2 Ciencias Biológicas OFICIO CPCB/059/2019 Asunto: Oficio de Jurado para Examen de Grado. M. en C. Ivonne Ramírez Wence Directora General de Administración Escolar, UNAM Presente Me permito informar a usted que en la reunión del Subcomité por Campo de Conocimiento de Ecología y Manejo Integral de Ecosistemas del Posgrado en Ciencias Biológicas, celebrada el día 17 de septiembre de 2018, se aprobó el siguiente juraga, para el examen de grado de DOCTOR EN pl CIENCIAS del alumno ARIAS PIEDRA ERICK con número de cuenta 515046843 « con la tesis titulada: «SISTEMÁTICA Y BIOGEOGRAFÍA DEL COMPLEJO DE ESPECIES CRAUGASTOR PODICIFERUS (ANURA: STRABOMANTIDAE) DE AMÉRICA CENTRAL ISTMICA EMPLEANDO ADN MITOCONDRIAL Y NUCLEAR”, realizada bajo la dirección de la DRA. GABRIELA PARRA OLEA: Presidente: DR. JUAN JOSÉ MORRONE LUPI Vocal: DRA. ELLA GLORIA VÁZQUEZ DOMÍNGUEZ Secretario: DR. JOSÉ MARTÍN GARCÍA VARELA Suplente: DR. ALEJANDRO ZALDÍVAR RIVERÓN Suplente DRA. CLAUDIA PATRICIA ORNELAS GARCÍA Sin otro particular, me es grato enviarle un cordial saludo. ATENTAMENTE “POR MI RAZA HABLARA EL ESPIRITU” Cd. Universitaria, Cd. Mx., a 16 de enero de 2019. DR. ADOLFO GERARDO NAVARRO SIGÚENZA COORDINADOR DEL PROGRAMA COORDINACIÓN c.c.p. Expediente del (la) interesado (a) Unidad de Posgrado *Coordinación del Posgrado en Ciencias Biológicas Edificio D, ler. Piso, Circuito de Posgrados Cd. Universitaria Delegación Coyoacán C.P. 04510 Cd. Mx. Tel. 5623 7002 http://pcbiol.posgrado.unam.mx UN M ~ . / SG DO ~ ORDINACi N FI I CB/059/2019 sunto : ficio e r do ra en e rado. . . ne amirez ence ir ctora eneral e dministración scolar, AM r sente e r ilo n ar l d e i n el ubcomité or po e ono miento de cología anejo t gral e cosi t as el sgrado i ncias Biológicas , l r da l i ti bre e 18, r bó l i i nte do ra l en de grado e CTOR I NCIAS el l no I S I A I K n ero e enta 5046843 n sis itul a: "SISTEMÁTICA y OGRAFíA EL MPLEJO PECIES ASTOR DI I US RA: BOMANTIDAE) ÉRICA TRAL S I A L DO N I NDRIAL y CLEAR", l a jo i ci n e RA. ABRIELA RA LEA: r sidente: ocal: ecretario: uplente: uplente R. N SÉ RRONE PI ORA. LA L RIA OUEZ i UEZ R. SÉ ARTIN RCIA ELA R. RO LDiv R I N A. L UDIA TRICIA NELAS ARCIA i tr articular, e s r to viarle rdial l do. TA ENTE " R I ZA BLARA L PIRITU" d . niversitaria, d . x., e ero e 019 . R. LFO ERARD ARRO Ü NZA ORDINADO EL RAMA . .p. xpediente el ) nte do ) COORDINACION nidad e sgrado • ordinación el osgrado iencias iológicas dificio , 1 r. iso, ircuito e osgrados d. niversitaria elegación oyoacán c. . 510 d. x. el. 23 2 tp:/ pcbiol.posgrado.unam.m AGRADECIMIENTOS INSTITUCIONALES Agradezco al Posgrado en Ciencias Biológicas, de la Universidad Nacional Autónoma de México, por la oportunidad de cursar el Doctorado en Ciencias Biológicas, así como el apoyo administrativo y económico. Al consejo Nacional de Ciencia y Tecnología (CONACyT) por la beca (CVU/Becario: 626946/330343) brindada para la realización de mis estudios durante los cuatro años correspondientes al programa de doctorado (2014–2018). A la Dirección General de Asuntos del Personal Académico de la UNAM (DGAPA), dentro del Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica (PAPIIT), que financió los proyectos: “Biodiversidad y conservación de los anfibios de bosques mesófilos de montaña del sur de México”, PAPIIT-DGAPA, IN-209914 y “Sistemática y conservación de anfibios de la zona oriental del eje neovolcánico transversal” PAPIIT-DGAPA, IN-203617. A la Dra. Gabriela Parra Olea, tutora principal, por el apoyo y confianza otorgada durante mi doctorado. Agradezco al Dr. Martín García Varela y al Dr. Adrián Nieto Montes de Oca, que como miembros de mi comité tutoral me acompañaron, asesoraron y motivaron durante el desarrollo de este trabajo. AGRADECIMIENTOS PERSONALES Agradezco a mi asesora Gabriela Parra, por la oportunidad brindada, aún sin conocerme, de realizar mi proyecto en su laboratorio. Gracias por creer en mí, por dejarme crecer e ilusionarme en este el camino de la academia. Pero, sobre todo, gracias por abrirme las puertas de su casa y por su calurosa amabilidad que hicieron que nunca me sintiera extranjero en estas tierras ajenas. Agradezco a la Dra. Ella Vázquez Domínguez y a los doctores Atilano Contreras Ramos, Daniel Piñero Dalmau y Oscar Flores Villela por amablemente evaluar mi examen de candidatura. De la misma manera, agradezco a las doctoras Ella Vázquez Domínguez y Patricia Ornelas García y a los doctores Alejandro Zaldívar Riverón y Juan José Morrone Lupi por amablemente revisar mi tesis y evaluar mi examen de grado. Al Dr. Gerardo Pérez Ponce de León, gracias por su calurosa amabilidad que hizo mucho más placentera mi estancia en México. Por ser un ejemplo a seguir como profesional y como persona. A los compañeros de laboratorio, con quienes compartí actividades académicas y también memorables momentos de diversión. En especial a Mirna García por compartir su conocimiento conmigo y por su amistad, a Aldo López por brindarme su amistad desde el primer momento y por permitirme –a través de sus salidas al campo– conocer lugares hermosos y otros no tanto. A Omar Becerra por compartir su conocimiento de escalada de árboles conmigo. A Delia, Ángela, Fabiola, Ángel, Alejandro y Raquel. A la M. en C. Andrea Jiménez Marín y la M. en C. Laura Márquez Valdelamar por la paciencia y ayuda en mi trabajo de laboratorio. A Lilia Espinosa, por su inmensa paciencia para conmigo, por comprenderme y colaborar en los diversos trámites administrativos que permitieron que mi estancia en la UNAM fuese una realidad. A Rocío Gonzáles por su amabilidad y su apoyo durante mi estancia en el Instituto de Biología. A mi amigo y herpetólogo Brian Kubicki, por compartir su conocimiento sobre los anfibios de Costa Rica conmigo y por siempre tener tiempo para revisar mis desastrosos manuscritos. Gracias por los consejos que me han ayudado a ser una mejor persona y un mejor herpetólogo. Al M.Sc. Gerardo “Cachí” Chaves, por todo su apoyo durante mis estudios en la Universidad de Costa Rica, por motivarme a perseguir mi doctorado, por compartir su conocimiento de la inhóspita Talamanca. Al M.Sc. Federico Bolaños, por inspirarme el amor que ahora siento hacia la herpetofauna, en especial la de Costa Rica. Por compartir conmigo gran cantidad de material que permitió el desarrollo de este proyecto. Muchas gracias Fede, consiente e inconscientemente me formaste y espero algún día ocupar con orgullo y responsabilidad la curación de herpetología en el Museo de Zoología que por tantos años has cumplido con vocación. A mi gran amigo y colega Adrián “Pichi” García, por sus consejos académicos y personales. A mi gran amigo Omar Zúñiga, por compartir todo su conocimiento de la Cordillera de Talamanca conmigo, por acompañarme y guiarme a los lugares más recónditos de Talamanca. A mi amiga Ericka Torres, por revisar y corregir el español de esta tesis. Al Dr. Andrew J. Crawford, por sus invaluables consejos en el desarrollo de este proyecto, por compartir conmigo desinteresadamente muestras que antes consiguió con esfuerzo. Al Dr. Andreas Hertz, por compartir conmigo gran cantidad de material de Panamá sin el cual este proyecto no hubiese sido lo que es. Agradezco a Justo Layam Gabb, Xavier Baltodano y Olmer Cordero por su valioso apoyo durante varias expediciones a Talamanca en busca de mis preciadas ranas y salamandras. Agradezco al Lic. Alejandro Solano Ortiz, ex-Vicecanciller de Relaciones Exterior de Costa Rica y a la Sra. Amelia Hidalgo Zamora, Cónsul de Costa Rica en México, por su amable colaboración diplomática que permitió que la visa para viajar a México fuese posible. A mi tío José Piedra, por ayudarme económicamente cuando más lo necesitaba. A Omar Saltero Terrones, por darme hospedaje y aliento a mi llegada a esta inmensa ciudad, por compartir su Acapulco conmigo y por su amistad. A la Dra. Rubí Meza Lázaro y al Dr. Adam Leaché, por compartir sus conocimientos en secuenciación de nueva generación. Al Ministerio de Ciencia, Tecnología y Telecomunicaciones de Costa Rica (MICITT) por el apoyo económico a través del Programa de Innovación y Capital Humano para la Competitividad PINN-MICITT (PED-0339-15-2). A National Geography Society, que apoyó parcialmente el trabajo de campo en Costa Rica a través del fondo W-346-14. DEDICATORIA A mis padres Belcides Arias y Aurelia Piedra quienes con su esfuerzo incansable me apoyaron y motivaron desde muy pequeño a cumplir cada una de mis metas. Pero sobre todo por darme la educación más valiosa, los valores y la moral. A mi esposa Fanny Hernández, quien ha colmado mis días de amor. Gracias por acompañarme en esta travesía por tierras ajenas, en esta etapa tan maravillosa. Pero sobre todo, por estar en los momentos más difíciles donde muchos se fueron. A mis hermanos, para quienes espero –en mi más humilde intención– ser un buen ejemplo y apoyo. ÍNDICE RESUMEN ...................................................................................................................... 11 ABSTRACT .................................................................................................................... 12 INTRODUCCIÓN GENERAL ........................................................................................ 13 América Central Ístmica .......................................................................................... 13 El grupo de especies Craugastor podiciferus ........................................................... 15 Delimitación de especies ......................................................................................... 19 OBJETIVOS .................................................................................................................... 24 CAPÍTULO I. Relaciones filogenéticas, delimitación y biogeografía .......................... 25 I.I Mitochondrial and nuclear DNA sequence data reveal high cryptic diversity in the Craugastor podiciferus species group (Anura: Craugastoridae) in Isthmian Central America .................................................................................................................. 26 CAPÍTULO II. Revisión taxonómica del grupo de especies Craugastor podiciferus ... 73 II.I. Taxonomic assessment of Craugastor podiciferus (Anura: Craugastoridae) in lower Central America with the description of two new species .............................. 74 II.II. A new species of the Craugastor podiciferus species group (Anura: Craugastoridae) from the premontane forest of southwestern Costa Rica ............... 128 II.III. A new species of Craugastor (Anura: Craugastoridae) from the montane rainforest of the Cordillera de Talamanca, Costa Rica ........................................... 146 DISCUSIÓN Y CONCLUSIONES GENERALES ......................................................... 169 La importancia de las tierras altas de América Central Ístmica .............................. 169 Comentarios taxonómicos al grupo de especies Craugastor podiciferus ................ 171 Delimitación y especies crípticas ........................................................................... 177 Conclusiones ......................................................................................................... 182 LITERATURA CITADA ............................................................................................... 183 APÉNDICES ................................................................................................................. 196 APÉNDICE I. 140 years after William M. Gabb`s climb to Cerro Pico Blanco ..... 197 APÉNDICE II. Biogeografía histórica de los anfibios de América Central: una revisión de hipótesis actuales y perspectivas futuras .............................................. 203 RESUMEN El grupo de especies Craugastor podiciferus (Anura: Craugastoridae) son ranas de desarrollo directo distribuidas desde Honduras hasta el centro de Panamá, y todas las especies descritas confluyen en Costa Rica y el oeste de Panamá –América Central Ístmica. La historia taxonómica del grupo ha sido compleja, especialmente por el alto nivel de polimorfismos fenotípicos a nivel intra- e interpoblacional. Al inicio del presente trabajo se reconocían ocho especies dentro del grupo, aunque se sospechaba la presencia de varias especies no descritas. Algunos trabajos previos validaron la monofilia del grupo de especies C. podiciferus y sugirieron que el origen del mismo se remonta a 20 millones de años. En este trabajo se recolectaron especímenes a lo largo del rango de distribución del grupo de especies C. podiciferus y se obtuvieron secuencias de ADN mitocondrial (16S y COI) y nuclear (SNPs) para reconstruir sus relaciones filogenéticas. Se obtuvieron hipótesis filogenéticas robustas que apoyan la presencia de varias especies no descritas dentro del grupo de especies C. podiciferus. Los análisis de delimitación de especies basados en secuencias de ADN sugieren que el grupo está conformado por al menos 23 linages, muchos de los cuales ameritan revisiones exhaustivas para determinar si deben ser considerados especies distintas. Aunque hubo diferencias entre las topologías obtenidas, los analisis filogenéticos indican que el grupo está representado por cuatro grandes clados. El clado basal, está formado por una especie nueva que se describe en este trabajo. Un segundo clado mayor incluye varios linajes de tierras altas, incluyendo a C. podiciferus sensu stricto. Dos especies fueron descritas dentro de este último clado, así como una que estaba como sinónimo de C. podiciferus Cope (1875). Los dos clados restantes contienen al menos 12 linajes de tierras bajas e intermedias, y dentro de estos clados una especie fue descrita como C. gabbi Arias, Chaves, Crawford, & Parra-Olea 2016. Los resultados sugieren que el grupo de especies C. podiciferus se originó y diversificó en las tierras altas de América Central Ístmica (ACI), ya que sus clados más basales se restrigen a las tierras altas del istmo. Además, los 23 linages identificados están distribuidos en Costa Rica y oeste de Panamá. La evidencia reciente sugiere que ACI emergío durante el Mioceno temprano (~23 millones de años antes), lo cual apoya la hipótesis de que el grupo se diversificó en ACI. Las especies que conforman el grupo de especies C. podiciferus ameritan claras estrategias de conservación, ya que, a pesar de que son relativamente abundantes sus áreas de distribución son restringidas. 11 ABSTRACT The Craugastor podiciferus species group (Anura: Craugastoridae) are direct-developing frogs that are distributed from Honduras to central Panama, with all its described species converging in Costa Rica and western Panama – Isthmian Central America. The taxonomic history of the group has been complex, especially due to the high level of intra- and interpopulation phenotypic polymorphisms. Before this study, eight species were recognized within the group, though the presence of several undescribed species was suspected. Previous studies validated the monophyly of the C. podiciferus species group and suggested that its origin date back to 20 million years ago. In this study, specimens were collected throughout the range of distribution of the C. podiciferus species group and mitochondrial (16S and COI) and nuclear (SNPs) DNA sequences were obtained to infer its phylogenetic relationships. Robust phylogenies were obtained, which support the presence of several unnamed species within of the C. podiciferus species group. DNA sequence-based species delimitation analyses suggest that the group consists of at least 23 lineages, many of which need exhaustive reviews to determine whether they should be considered as different species. Although there were differences among the phylogenies, all phylogenetic analyses indicate that the group consists of four large clades. The basal clade is formed by a new species that is described here. A second major clade includes several highland lineages, including C. podiciferus sensu stricto. Two species were described within this clade and one more that was under synonymy of C. podiciferus Cope (1875). The two remaining clades contain at least 12 lineages of lowlands and midlands; within of these clades one species was described as C. gabbi Arias, Chaves, Crawford, & Parra-Olea 2016. The results suggest that the C. podiciferus species group originated and diversified in the highlands of Isthmian Central America (ICA), by the fact that the most basal clades are restricted to the highlands of the isthmus. Also, the 23 identified linages are distributed in Costa Rica and western Panama. The recent evidence suggests that ICA emerged during the early Miocene (~ 23 million years ago), supports the hypothesis that the study group diversified ICA. The species within of the C. podiciferus species group merit clear conservation strategies, because although they are relatively abundant, their distribution areas are restricted. 12 INTRODUCCIÓN GENERAL América Central Ístmica El Neotrópico es reconocido como una de las regiones más biodiversas del mundo, incluyendo siete de los 25 hotspots de biodiversidad a nivel mundial (Myers et al. 2000). Desde una perspectiva biogeográfica, América Central es el área comprendida entre el Istmo de Tehuantepec, México y la Cordillera de los Andes en Colombia (Savage 1982; Gutiérrez-García & Vázquez-Domínguez 2013). Esta puede ser dividida en dos unidades biogeográficas independientes: 1) América Central Nuclear, el área entre el Istmo de Tehuantepec y el sur de Nicaragua, que alberga las tierras más antiguas de la región (Weber et al. 2007; Martens et al. 2010) y han tenido un papel importante en la diversificación temprana de la biota centroamericana (Savage 2002; Heinicke et al. 2007; Duellman et al. 2016); y 2) América Central Ístmica (ACI), el área que comprende Costa Rica y Panamá e incluye tierras relativamente más jóvenes (Montes et al. 2012a,b, 2015) y que han tenido un papel importante en la conformación actual de la biodiversidad regional permitiendo el Gran Intercambio de la Biota Americana (Bacon et al. 2015, 2016). ACI se caracteriza por su biodiversidad y alto nivel de endemismos (Reid & Miller 1989). Esta región contiene a más especies de anfibios, reptiles, aves, insectos y plantas vasculares por unidad de área que casi cualquier otra parte del mundo (Davis et al. 1997; Anger & Dean 2010; Garrigues & Dean 2014; AmphibiaWeb 2019). Esta gran biodiversidad se atribuye a dos factores, el cierre del Istmo de Panamá y su historia de actividad volcánica y orogénica. El cierre del Istmo de Panamá, permitió el Gran Intercambio de la Biota Americana (GABI por sus siglas en inglés) entre Norteamerica y Sudamerica, permitiendo la confluencia de 13 especies del norte y sur en ACI (Savage 1966; Webb 1991, 2006; Pinto-Sánchez et al. 2012). La actividad volcánica y orogénica en ACI resultó en una alta heterogeneidad climática y geomorfológica (Montes et al. 2012a,b, 2015; Bagley & Johnson 2014). La biodiversidad de ACI se vio reforzada por la diversificación in situ, que generó un alto nivel de endemismos asociados a la alta diversidad de hábitats (Campbell 1999; Savage 2002; Kluge & Kessler 2006; Boza-Oviedo et al. 2012). Desde una perspectiva geológica, ACI se divide en dos grandes bloques: Chorotega y Chocó. Chorotega, que se extiende desde el escarpe de Hess en el sur de Nicaragua hasta la fractura norte de Panamá. Y Chocó, que se extiende desde la fractura norte de Panamá hasta la zona de subducción entre las placas Nazca y sudamericana en Colombia (Gutiérrez-García & Vázquez-Domínguez 2013). Los bloques de Chorotega y Chocó comparten una estrecha relación histórica, por otra parte, el conocimiento sobre su formación ha cambiado aceleradamente. La hipótesis tradicional sobre el origen de la región se denomina “modelo de isla”, ya que propone un origen para ACI reciente en el Mioceno Medio (~15 Ma) a ~3 Ma (Coates & Obando 1996; Coates et al. 2004). Bajo este modelo, en el Mioceno Medio (~15 Ma) ACI estaba conformada por un archipiélago de islas. Este modelo sugiere que el levantamiento regional generó la unión de tierras emergidas, pero desconectadas de Sudamérica, hace ~6.5 Ma y propone el cierre del Istmo de Panamá hace ~3 Ma. Una hipótesis más reciente, denominada “modelo de península”, plantea un levantamiento y un cierre mucho más antiguo (Montes et al. 2012a,b, 2015). Este modelo sugiere que ACI existe como tierra emergida pero desconectada de Sudamérica desde el Oligoceno Tardío al Mioceno Temprano (~25 Ma) y calcula el cierre del Istmo de Panamá entre 15–13 Ma. La hipótesis de Montes et al. (2012a,b, 2015) sobre el cierre temprano del Istmo de Panamá ha sido apoyada por los aportes filogenéticos. Por ejemplo, se conoce que varios grupos se 14 diversificaron en ACI desde el Mioceno Temprano (~23 Ma), incluyendo ranas (Crawford & Smith 2005; Streicher et al. 2009; Duellman et al. 2016), salamandras (Rovito et al. 2015), serpientes (Castoe et al. 2009) y peces dulceacuícolas (Říčan et al. 2013). Aunque los anfibios constituyen uno de los grupos más biodiversos de la región (AmphibiaWeb 2019) y de los mejor estudiados (Taylor 1948, 1952; Savage 1966, 1982, 2002), aún es notable la carencia de estudios biogeográficos sobre su diversificación en ACI, desde la evidencia actual del cierre del Istmo de Panamá durante el Mioceno (Montes et al. 2012a,b, 2015). Este es un tema relativamente nuevo, que plantea una serie de interrogantes con respecto al origen y diversificación de los anfibios en ACI. Por ejemplo, este modelo pone a prueba la importancia de las fluctuaciones climáticas durante el Plioceno–Pleistoceno en la diversificación de taxones de tierras altas (Savage 2002), ya que es factible sugerir que la diversificación es anterior al Pleistoceno (Castoe et al. 2009; Streicher et al. 2009). Las tierras altas de ACI (>750 m) conforman la ecoregión “Talamanca montane forest, TMF” (Olson et al. 2001), caracterizada por cambios altitudinales significativos, alta variación climática y carencia de conectividad con otras tierras altas de América. Las características de TMF se ven enriquecidas por el alto número de anfibios endémicos a TMF (Campbell 1999; García-París et al. 2000; Savage 2002; Boza- Oviedo et al. 2012). El grupo de especies Craugastor podiciferus Uno de los principales retos dentro del conocimiento de los anfibios, es entender las relaciones filogenéticas y la diversidad de las ranas con desarrollo directo (superfamilia Brachycephaloidea, Padial et al. 2014). Este grupo es considerado como el menos estudiado de todos los vertebrados (Hedges et al. 2008), contiene poco más de 1100 especies descritas, 15 representando más del 16 % de los anuros descritos en el Mundo (Frost 2019). Por su abundancia relativa, las ranas con desarrollo directo representan componentes importantes dentro de los ecosistemas. Las ranas con desarrollo directo constituyen un grupo importante de depredadores y también gran cantidad de biomasa (Lynch & Duellman 1997; Savage 2002; Pinto-Sánchez et al. 2014). Los vacíos de información dentro de los terraranas (ranas con desarrollo directo) es explicado en parte por la poca cantidad de caracteres morfológicos disponibles para analizar, la plasticidad de los pocos caracteres útiles y el alto nivel de poliformismos morfológicos (Hedges et al. 2008). La tasa de descripción de especies nuevas de ranas con desarrollo directo aumentó a partir de la segunda mitad del siglo XX, alcanzando tasas de hasta 15 especies nuevas por año (Hedges et al. 2008). Sin embargo, por décadas los taxónomos no han alcanzado un consenso sobre las relaciones filogenéticas y la organización de las especies en categorias menores (familias, géneros o subgéneros) que permitieran una mejor clasificación de su diversidad (Lynch 1986, 1993, 2000; Savage 1987, 2002; Hedges 1989; Lynch & Duellman 1997; Hedges et al. 2008). Con el desarrollo de nuevas tecnologías que facilitaron la obtención de datos moleculares a través de la secuenciación y la reconstrucción filogenética, se generaron diversos estudios que evaluaron las relaciones filogenéticas dentro de las ranas con desarrollo directo (Frost et al. 2006; Heinicke et al. 2007, 2009, 2018; Hedges et al. 2008; Pyron & Wiens 2011; Padial et al. 2014; Pie et al. 2017). Hedges et al. (2008) sugirieron el reconocimiento de diversas familias, géneros, subgéneros y grupos de especies. Pyron & Wiens (2011) y Padial et al. (2014) realizaron cambios en la sistemática de las ranas con desarrollo directo. Cinco de estos géneros de ranas con desarrollo directo (Craugastor, Diasporus, Eleutherodactylus, Pristimantis y 16 Strabomantis) habitan ACI, pero Craugastor es el más diverso con casi 100 especies en la región. La monofilia del género Craugastor (Cope) había sido reconocida desde Lynch (1986) basado en análisis de la musculatura de la mandíbula, y su validez como género ha permanecido estable desde Hedges et al (2008). Las relaciones filogenéticas dentro de Craugastor permanecen bajo discusión, actualmente se reconocen tres subgéneros y seis grupos de especies (Padial et al. 2014). Las relaciones filogenéticas dentro de los grupos de especies de Craugastor son poco entendidas (Padial et al. 2014); sin embargo, los estudios empleando taxonomía alfa son de vital importancia, ya que evaluan la diversidad real dentro de un grupo mediante la delimitación de especies. Aunque el género Craugastor representa un porcentaje alto de la fauna anfibia de ACI (Frost 2019), la homoplasia entre las especies y el polimorfismo fenotípico han sido un impedimento para entender la taxonomía y la sistemática de las ranas de desarrollo directo en la región (Lynch 2000; Crawford 2003; Hedges et al. 2008; Streicher et al. 2009). Algunos estudios moleculares con Craugastor de ACI, revelaron la existencia de varias especies no descritas (Crawford 2003; Crawford & Smith 2005; Crawford et al. 2007; Streicher et al. 2009), sugiriendo que la taxonomía de las ranas de desarrollo directo es insuficiente y posiblemente hay una alta proporción de especies crípticas. Sumado al hecho de que los anfibios como grupo contienen mayor proporción de especies crípticas dentro de los vertebrados (Pérez-Ponce de León & Poulin 2016). Que son aquellas especies tratadas bajo un mismo nombre dada su alta similitud morfológica (Bickford et al. 2007). Dentro del género Craugastor (Brachycephaloidea: Craugastoridae), el grupo de especies C. podiciferus es un ejemplo de estas dificultades taxonómicas. El grupo de especies C. podiciferus contiene actualmente nueve especies descritas (Frost 2019). Cope (1875) describió la 17 primera especie del grupo, C. podiciferus y cuatro variedades para ésta. Además, este mismo autor describió otras dos especies, C. muricinus y C. habenatus con especímenes de la misma localidad. Cope (1875) expuso el fenómeno de polimorfismos dentro del grupo al mencionar “los colores de esta especie varían considerablemente, más de lo que yo haya observado en cualquier otra rana.” Cope (1886, 1893) describió posteriormente otras tres especies del grupo, C. bransfordii (Cope, 1886), C. polyptychus (Cope, 1886) y C. stejnegerianus (Cope, 1893), las primeras dos con especímenes de la misma localidad. Tres especies más fueron descritas durante la primera mitad del siglo XX, C. blairi (Barbour, 1928), C. persimilis (Barbour, 1926) y C. underwoodi (Boulenger, 1896). Taylor (1952) realizó la primera revisión exhaustiva para el grupo, describiendo dos especies adicionales, C. rearki y C. costaricensis. Savage & Emerson (1970) basados en análisis morfológicos, redujeron el número de especies dentro del grupo a solamente dos, C. bransfordii y C. podiciferus. Usando datos bioquímicos, Miyamoto (1983) encontró que las poblaciones de C. bransfordii en el pacífico de Costa Rica son altamente divergentes de aquellas en la vertiente del caribe, reasignando el nombre C. stejenegrianus para las poblaciones del pacífico. Estos hallazgos fueron corroborados por diferencias en cariotipos (Chen 2001, 2005). Savage (2002) reasignó los nombres C. persimilis, C. polyptychus y C. underwoodi, dejando a C. rearki y C. costaricensis bajo sinonimia de C. bransfordii y C. muricinus, C. habenatus y C. blairi bajo sinonimia de C. podiciferus. Previo a Savage (2002), dos especies adicionales fueron nombradas, C. jota (Lynch, 1980) y C. lauraster (Savage, McCranie & Espinal, 1996). De esta manera, después de Savage (2002) ocho especies fueron reconocidas como parte del grupo, C. bransfordii, C. jota, C. lauraster, C. persimilis, C. podiciferus, C. polyptychus, C. stejnegerianus y C. underwoodi. 18 Estudios filogenéticos basados en secuencias de ADN han apoyado la monofilia del grupo (Hedges et al. 2008; Pyron & Wiens 2011; Padial et al. 2014), sugiriendo que el grupo contiene especies no nombradas (Crawford 2003; Crawford & Smith 2005; Streicher et al. 2009). Crawford (2003) encontró divergencias genéticas altas dentro del grupo, similares a aquellas encontradas entre géneros de ranas australianas (Anura: Myobatrachidae), aunque según el autor, estas poblaciones no presentan diferencias morfológicas. Crawford & Smith (2005) usando 10 muestras del grupo, encontraron dos especies no nombradas (Craugastor sp. B y Craugastor sp. C) y además demostraron que C. stjenegrianus es parafilético. Posteriormente, Streicher et al. (2009) realizó un estudio filogeográfico utilizando poblaciones de tierras altas (>1000 m), que históricamente fueron asignadas a la especie C. podiciferus. Estos autores encontraron que las poblaciones referidas como C. podiciferus representan un complejo de especies con al menos seis especies no nombradas, y además confirmaron la presencia de Craugastor sp. B (Streicher et al. 2009). Arias et al. (2016) emplearon datos mitocondriales y morfológicos, confirmando los resultados previos (Crawford & Smith 2005) al encontrar que las poblaciones de tierras intermedias del Pacífico Sur de Costa Rica y oeste de Panamá que fueron históricamente asignadas a C. stejnegerianus en realidad pertenecen a una especie distinta, a la que nombraron como C. gabbi Arias, Chaves, Crawford, & Parra-Olea, 2016. Delimitación de especies La delimitación de especies es el proceso por el cual se determinan los límites entre las mismas y se describen nuevas especies (Sites & Marshall 2003). Este proceso está estrechamente relacionado con el concepto de especie, por lo tanto, la reciente adopción de un concepto unificado de especie ha generado una revolución en la delimitación (de Queiroz 2007). De 19 Queiroz (2007) propone separar entre conceptualización de especie y criterios operacionales de especie. El primero, definido como la única propiedad necesaria de la categoría de especie, un concepto teórico, explicado como un linaje que evoluciona independientemente, mientras que los criterios operaciones de especies, son las múltiples líneas de evidencia que son relevantes para la delimitación de estas (de Queiroz 2007). La separación entre la conceptualización de las especies y sus criterios operacionales ha permitido la generación de diversos protocolos que permiten una delimitación reproducible de las especies. De Queiroz (2007) sugiere que las especies deben ser delimitadas utilizando múltiples líneas de evidencia. Aunque se ha sugerido la presencia de varias especies no nombradas dentro del grupo de especies C. podiciferus (Savage 2002; Crawford & Smith 2005; Streicher et al. 2009), solamente una especie, C. gabbi, fue descrita y nombrada posterior a Savage et al. (1996). El descenso en descripción de nuevas especies puede ser explicado por el cambio de paradigma en la sistemática y la necesidad de apoyar las nuevas especies con al menos dos líneas de evidencia, morfológica y molecular (de Queiroz 1998; Dayrat 2005; Will et al. 2005; Padial et al. 2010). En el pasado, los estudios dentro del género carecían de evidencia integradora; por ejemplo, todos los estudios utilizando evidencia molecular no empleaban análisis morfológicos. Sin embargo, la evidencia sugiere que es posible resolver complejos de especies crípticas utilizando un criterio de taxonomía integradora (Padial et al. 2008; Pérez-Ponce de León & Nadler 2010; Arias et al. 2016). Es posible identificar tres clados dentro del grupo de especies C. podiciferus con base en los siguientes caracteres morfológicos: 1) el clado C. bransfordii, conformado por C. bransfordii, C. polyptychus y C. underwoodi, reconocido por tener el tubérculo tenar igual o apenas ligeramente más pequeño que el palmar; 2) el clado C. podiciferus, conformado por C. jota y C. 20 podiciferus, reconocido por tener el vientre liso muy diferente al vientre granular o areolado de las demás especies del grupo; y 3) el clado C. stejnegerianus, representado por C. gabbi, C. lauraster, C. persimilis y C. stejnegerianus, reconocido por tener el tubérculo tenar mucho más pequeño que el palmar. Padial et al. (2014) recuperó los tres clados antes mencionados; sin embargo, el alto grado de similitud morfológica entre las especies del grupo continúa siendo una dificultad para delimitar correctamente las especies y para evaluar la diversidad real dentro de cada grupo de especies. En grupos como Craugastor, reconocidos por altos índices de polimorfismos y homoplasia, la delimitación de especies ha sido particularmente compleja (Lynch & Duellman 1997; Lynch 2000; Savage 2002; Hedges et al. 2008). Algunos autores se han referido a diversas especies de Craugastor como “especies crípticas” (Savage & Emerson 1970; Miyamoto 1983; Savage 2002; Crawford 2003; Streicher et al. 2009). Pérez-Ponce de León & Nadler (2010) distinguieron entre especie críptica sensu stricto (morfológicamente indistinguibles) versus definición funcional de especie críptica, definida por el taxónomo con base en su alto grado de similitud morfológica. Nadler & Pérez-Ponce de León (2011) sugieren que la delimitación de “especies crípticas” se debe basar en prueba de hipótesis desde un enfoque molecular. Si la hipótesis nula de una única especie es rechazada con base en datos moleculares, entonces, un detallado análisis morfológico puede proveer caracteres morfológicos que permitan la separación confiable de las especies y una diagnosis. La delimitación de especies dentro de grupos morfológicamente conservados, se beneficia de los avances en el acceso a secuencias de ADN. La secuenciación Sanger o tradicional ha permito obtener gran cantidad de secuencias, pero están restringidas a secciones específicas del genoma, en especial aquellas del mitogenoma en animales (Hebert et al. 2003). 21 Estas secuencias mitocondriales han sido de gran utilidad para el reconociento de nuevas especies y las propuestas de hipótesis filogenéticas (Vences et al. 2005; Fouquet et al. 2007; Smith et al. 2008), sin embargo, tiene limitaciones para resolver ciertos complejos de especies. El desarrollo tecnológico en la secuenciación de la nueva generación, ha permitido incrementar el acceso a secuencias genómicas de organismos no modelos, ofreciendo una gran oportunidad de superar las dificultades inherentes al uso de los enfoques de secuenciación tradicional (Herrera & Shank 2016). Una de estas metodologías es la secuenciación del ADN asociado a sitios de restricción (RAD-seq), que combina fragmentación enzimática con secuenciación masiva, para la obtención de gran cantidad de marcadores o SNPs (Baird et al. 2008; Peterson et al. 2012). RAD-seq ha mostrado gran utilidad para obtener filogenias y delimitación de especies robustas dentro de grupos taxonómicos de difícil tratamiento (Peterson et al. 2012; Streicher et al. 2014; Leaché et al. 2015). La delimitación molecular de especies es actualmente uno de los campos de mayor desarrollo en sistemática (Sites & Marshall 2003; Flot 2015). Una gran variedad de métodos han sido desarrollados recientemente: ABGD (Puillandre et al. 2012), PTP (Zhang et al. 2013), mPTP (Kapli et al. 2016), BPP (Yang 2015; Yang and Rannala 2010, 2014), GMYC (Pons et al. 2006; Fujisawa & Barraclough 2013), Structurama (Huelsenbeck et al. 2011) y SpedeSTEM (Ence & Carstens 2011), entre otros. La efectividad de los métodos en identificar especies continúa bajo discusión (Sukumaran & Knowles 2017). Sin embargo, el uso de estos métodos puede ser útil como una primera aproximación para reconocer la diversidad dentro de un grupo determinado, en especial para aquellos grupos morfológicamente conservados y polimórficos. Estos métodos pueden fallar en reconocer las especies, aunque pueden servir para identificar 22 especies candidatas que posteriormente deberán ser confirmadas con otras líneas de evidencia como morfología, bioacústica, segregación en nicho u otras. El grupo de especies C. podiciferus actualmente está compuesto por nueve especies de terraranas pobremente estudiadas, abundantes en la hojarasca y ampliamente distribuidas (colectivamente) en ACI (Savage 2002; AmphibiaWeb 2019). Los miembros de este grupo son morfológicamente conservados y con alto nivel de polimorfismos, por lo cual se ha sugerido que contiene varias especies no descritas. El origen del grupo C. podiciferus se ha calculado en ~20 Ma cuando ACI estaba emergida en su posición actual (Crawford & Smith 2005; Streicher et al. 2009). Considerando la distribución y abundancia, el grupo de especies C. podiciferus constituye un excelente modelo para evaluar las relaciones filogenéticas de un grupo de terraranas que diversificó en paralelo con la formación de ACI durante el Mioceno. En el presente trabajo de tesis abordé el estudio del complejo de especies C. podiciferus, incluyendo sus relaciones filogenéticas, su composición taxonómica y además, se describe la importancia de las tierras altas de ACI en la diversificación del grupo. Para responder estas preguntas obtuve secuencias de los genes mitocondriales 16S (16S) y citocromo oxidasa 1 (COI), obtenidos a través de secuenciación Sanger. Además, empleé una técnica de secuenciación de nueva generación (RAD-seq), lo cual permitió analizar cientos de loci y contar con una hipótesis filogenética robusta. 23 OBJETIVOS Objetivo general: Investigar la sistemática y evolución del grupo de especies Craugastor podiciferus con un criterio de taxonomía integradora. Objetivos específicos: 1. Inferir las relaciones filogenéticas dentro del grupo de especies C. podiciferus, usando dos marcadores mitocondriales y secuencias nucleares (SNPs) obtenidas con RAD-seq. 2. Proponer una hipótesis taxonómica para el grupo de especies C. podiciferus, integrando la evidencia molecular y morfológica para la delimitación de las especies. 3. Describir morofológicamente las especies que se han delimitado, incluyendo la descripción de especies nuevas. 4. Proponer una hipótesis del papel de las tierras altas de ACI en la diversificación del grupo de especies C. podiciferus. 24 CAPÍTULO I RELACIONES FILOGENÉTICAS, DELIMITACIÓN Y BIOGEOGRAFÍA 25 Mitochondrial and nuclear DNA reveal high cryptic diversity in the Craugastor podiciferus species group (Anura: Craugastoridae) in Isthmian Central America ERICK ARIAS1,2,3, ANDREW J. CRAWFORD4,5,6, ANDREAS HERTZ7,8, AND GABRIELA PARRA-OLEA1,* 1 Departamento de Zoología, Instituto de Biología, UNAM, AP 70-153 Ciudad Universitaria, CP 04510, Ciudad de México, México. 2 Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México, Av. Ciudad Universitaria 3000, C.P. 04360, Coyoacán, Ciudad de México, México. 3 Escuela de Biología, Universidad de Costa Rica, San Pedro, 11501-2060 San José, Costa Rica. 4 Deparment of Biological Sciences, Universidad de los Andes, A.A. 4976 Bogotá, Colombia. 5 Círculo Herpetológico de Panamá, Apartado 0824-00122, Panama City, Republic of Panama. 6 Smithsonian Tropical Research Institute, Apartado 0843-03092, Panama City, Republic of Panama. 7 Department of Biology, University of Massachusetts Boston, 100 Morrissey Blvd., Boston, MA 02125, USA. 8 Senckenberg Forschungsinstitut und Naturmuseum, Senckenberganlage 25, 60325, Frankfurt am Main, Germany. *Corresponding author: E-mail: gparra@ib.unam.mx Running Title: Systematic of Craugastor podiciferus 26 Abstract The Craugastor podiciferus species group contains ten species of terraranas from Central America with high levels of polymorphism and possibly containing overlooked species masked under the current names. We performed an extensive geographical sampling of the C. podiciferus species group to evaluate its phylogenetic relationships and biogeographic history based on two mitochondrial markers and nuclear ddRAD loci. We also conducted various species delimitation methods to test the presence of unnamed species within the group. The phylogenetic relationships recovered showed that the group contains four major clades, three of them corresponding to previously known clades, and the remaining one recently described. Ours results suggest that the species richness within the group is underestimated; however, the species delimitation was highly discordant between the mitochondrial and nuclear analyses and among methods. We propose that the C. podiciferus species group contains at least 23 lineages, all distributed in Costa Rica and western Panama. A biogeographic analysis shows that the group originated and diversified in the highlands of the Talamancan montane forest ecoregion. Key words: Biogeography, Lagrange, species delimitation, systematic, taxonomy Introduction Species biodiversity is not homogeneously distributed over the globe. The Neotropics are one of the regions with higher biodiversity in the world, including seven of the 25-biodiversity hotspots (Myers et al. 2000). One of these hotspots, the Mesoamerica biodiversity hotspot, includes the Isthmian Central America (ICA) region, which is located in Costa Rica and Panama and it contains an exceptional biodiversity and endemism. This region hosts more species of amphibians, reptiles, birds, insects, and vascular plants per area unit than almost any other place in the world (Davis et al. 1997; Anger & Dean 2010; Garrigues & Dean 2014; AmphibiaWeb 27 2019). The high biodiversity in ICA is explained by two major factors, the closure of the Isthmus of Panama and its long history of volcanic and orogenic activity. The closure of the Isthmus of Panama allowed the Great American Biotic Interchange (GABI) between North and South America, resulting in the confluence of north American and south American species in ICA (Savage 1966; Webb 1991, 2006; Pinto-Sánchez et al. 2012). The long history of volcanic and orogenic activity in ICA resulted in a high geographic and climatic heterogeneity (Bagley & Johnson 2014; Montes et al. 2012a,b, 2015). The diversity in ICA was enhanced by in situ diversification, generating high levels of endemisms associated with a high diversity of habitats (Campbell 1999; Savage 2002; Kluge 2006; Boza-Oviedo et al. 2012). ICA contains the highest species richness relative to area in the World (AmphibiaWeb 2019; Frost 2019) including some northern and southern species and several micro endemic taxa associated to the highlands. Nevertheless, big areas within ICA remain unexplored, and some particular groups (e.g. Craugastor, Diasporus, Pristimantis) have not been well studied yet. Amphibians contain the highest proportion of cryptic species within metazoans (Pérez-Ponce de León & Poulin 2016). This overlooked diversity masks the real diversity of a group and makes it difficult to make biogeographical inferences for the region. Previous molecular studies with direct-developing frogs (Craugastor and Pristimantis) and direct-developing salamanders (Bolitoglossa) revealed extremely high levels of genetic diversity both in highlands (García-París et al. 2000; Wiens et al. 2007; Streicher et al. 2009) and lowlands (Crawford 2003; Crawford et al. 2007; Wang et al. 2008) of the ICA. The Craugastor podiciferus species group (Hedges et al. 2008) is currently composed of ten described species (Arias et al. 2018). This group occurs from eastern Honduras to central Panama, ranging from the sea level to 2700 m a.s.l. in a wide variety of habitats (Savage 2002; AmphibiaWeb 2019). The systematics and taxonomy of the C. podiciferus species group has been poorly studied, though previous molecular studies support the presence of several 28 undescribed species (Savage 2002; Crawford 2003; Crawford & Smith 2005; Streicher et al. 2009). Crawford (2003) found high genetic divergences between populations of C. stejnegerianus in the ICA lowlands. Streicher et al. (2009) used the specimens from the highlands (>1000 m a.s.l.), referred to C. podiciferus, and mentioned that the name C. podiciferus could mask a species complex formed by up to six distinct taxa, and also supported the existence of an undescribed species (Craugastor sp. B) related to C. podiciferus. Based on mitochondrial data, Arias et al. (2016) recently showed that populations formerly considered part of C. stejnegerianus from southwestern Costa Rica and western Panama belong to a different species, which they named C. gabbi. Several studies are concordant with the presence of a great cryptic diversity within amphibians from ICA (García-París et al. 2000; Crawford 2003; Crawford & Smith 2005; Crawford et al. 2007; Streicher et al. 2009; Batista et al. 2016). Nevertheless, these studies were limited to few species and restricted to few mitochondrial or nuclear markers. In addition, these studies were restricted only to species from highlands or lowlands. To our knowledge, no molecular studies have extensively evaluated the phylogenetic relationships and the overlooked diversity of an amphibian species group restricted to ICA, including populations both from highlands and lowlands and using mitochondrial markers and an extensive nuclear dataset. Here we use mitochondrial and genome-scale datasets to: 1) infer the phylogenetic relationships of the C. podiciferus species group, 2) to test the existence of overlooked species within the current names, and 3) to investigate the center of origin for the group and suggest a possible historical framework for its species diversification. 29 FIGURE 1. Geographic distribution of the Craugastor podiciferus species group. 2. Materials and methods 2.1. Taxon sampling Tissue samples were collected from 103 specimens, including all species of the C. podiciferus species group from Honduras, Nicaragua, Costa Rica, and Panama (Fig. 1, Appendix I). The frogs were obtained in the field, euthanized, fixed in a 10% formalin solution and processed over to 70% ethanol for long-term storage. The tissue sample used for the genetic analyses was preserved in 96% ethanol or RNAlater. The vouchers were deposited at the Museo de Zoología, Universidad de Costa Rica (UCR), the Division of Amphibians and Reptiles at the Field Museum of Natural History (FMNH), Circulo Herpetológico, Panama (CH), and Senckenberg Research Institute and Nature Museum, Frankfurt, Germany (SMF). Museum collection 30 acronyms follow Frost (2019), with the addition of AJC refers to Andrew J. Crawford field numbers, AH refers to Andreas Hertz field numbers, and EAP refers to Erick Arias field numbers. 2.2. Mitochondrial data 2.2.1. Amplification and sequencing We extracted total genomic DNA from the preserved tissue samples using the Animal Genomic DNA Kit (BioBasic Canada Inc.), the DNeasy Blood & Tissue Kit (Qiagen), or the phenol- chloroform standard protocol (Sambrook & Russell 2006). We amplified the large subunit ribosomal RNA (16S) and cytochrome oxidase subunit I (COI) mitochondrial genes. The primers 16Sar and 16Sbr (Palumbi et al. 1991) were used for 16S and dgLCO and dgHCO (Meyer 2003) for COI. Amplifications were performed using a total volume of 15 µL, which contained 1 µL DNA template (c. 50 ng µL-1), 0.75 U Taq polymerase (Amplificasa®, Biotecnologias Moleculares), 1 x PCR buffer with 1.5 mM MgCl2, 0.2 mM deoxynucleotide triphosphates (dNTPs), and 0.3-0.5 µM forward and reverse primers. The PCR conditions were as follow: 16S, an initial cycle of 5 min at 94ºC, followed by 35 cycles of 45 s at 94ºC, 30 s at 50 or 55ºC, 45 or 120 s at 72ºC, plus a final cycle of 3 min at 72ºC; COI, an initial cycle of 2 min at 94ºC, followed by 35 cycles of 30 s at 94ºC, 30 s at 50ºC, 45 s at 72ºC, plus a final cycle of 3 min at 72ºC. PCR products were cleaned with ExoSap-IT (USB Corporation) and sequenced in both directions using the original amplification primers and BigDye termination reaction chemistry (Applied Biosystems). The cycle-sequencing products were column-purified with Sephadex G-50 (GE Healthcare) and run on an ABI 3500xL Genetic Analyzer (Applied Biosystems). Consensus sequences for each individual were constructed using SEQUENCHER 5.3 (Genes Codes Corp.). 2.2.2. Phylogenetic analyses 31 We generated 16S and COI sequences for members of the C. podiciferus species group, and retrieved sequences from GenBank for C. gollmeri, which was included as outgroup. A list with the examined material, their localities and DNA voucher and GenBank accession numbers used in this study is provided in Appendix I. Sequence alignments were performed using the MUSCLE 3.7 software (Edgar 2004) with default parameters and trimmed to the point where a majority of the taxa had sequence data. We partitioned the sequence data by gene, and further partitioned COI by codon position. We used PartitionFinder v1.1.1 (Lanfear et al. 2012) and the Bayesian Information Criterion (BIC) to select the best partition scheme and the best model of sequence evolution for each partition. We used a single set of branch-lengths across all partitions (branchlengths=linked), the search of the best partition scheme was using a heuristic search (scheme=greedy). We defined four partitions one for 16S and three for COI according to its codon positions. The partition scheme and the model of sequence evolution for each partition as selected by PartitionFinder (Lanfear et al. 2012) were used in the Bayesian methods (see below). We used maximum likelihood and Bayesian methods to infer a phylogenetic tree from the concatenated loci in the final dataset. We performed the maximum likelihood analysis using RAxML-HPC v8 (Stamatakis 2014) with the GTR + GAMMA model of nucleotide substitution (default model of RAxML) and the –f a option, which searches for the best-scoring tree and performs a rapid bootstraps analysis (i.e. 1000 bootstraps) to estimate node support by resampling the data. A partitioned Bayesian phylogenetic analysis was performed using MrBayes 3.2.6 (Ronquist et al. 2012) with the partition scheme and the models of sequence evolution that was previously selected. Two separate analyses were run, each consisting of 20 million generations, sampling every 1000 generations and using four chains with default heating parameters. We examined a time-series plot of the likelihood scores of the cold chain to check stationarity using the Tracer 1.6 software (Rambaut et al. 2014). We discarded the first 25% of trees as burn-in and used the remaining trees to estimate the consensus tree along with the posterior probabilities for each node and each parameter. 32 We used the program BEAST v1.8.3 (Drummond et al. 2012) to estimate ultrametric phylogenetic tree using an uncorrelated lognormal relaxed clock, a Yule tree prior, and with the partition scheme and the model of sequence evolution for each partition as selected previously. We ran the analysis for 50 million generations, sampling every 1000 generations, and discarding the first 5000 samples as burn-in when estimating a consensus tree. All phylogenetic analyses were run on the CIPRES portal (Miller et al. 2010). Estimates of pairwise evolutionary genetic divergence between groups were computed using MEGA7 (Tamura et al. 2013), assuming uncorrected distances based on the Tamura 3-parameter model (Tamura 1992), with rate variation among sites modeled as a gamma distribution with the shape parameter = 4 as the default of the software. 2.3. Nuclear data 2.3.1. ddRADseq data collection We generated ddRADseq data from 48 samples following the protocol described by Peterson et al. (2012) and modified by Leaché et al. (2015). High-molecular weight DNA was purified with RNase, examined for quality on agarose gels, and quantified with a Qubit 2.0 fluorometer (Thermo Fisher Scientific). We used 1000 ng of genomic DNA for each sample, except for three samples whose account of DNA was of 241–630 ng. We double-digested the genomic DNA with 20 units each of a rare cutter SbfI (restriction site 5´-CCTGCAGG-3´) and a common cutter MspI (restriction site 5´-CCGG-3´) in a single reaction with the manufacturer recommended buffer (New England Biolabs) for 2 h at 37 ºC. Fragments post-digestion were purified with Serapure 1.5X and quantified with a Qubit 2.0 fluorometer (Thermo Fisher Scientific) before ligation of barcoded Illumina adaptors onto the fragments. The oligonucleotide sequences used for barcoding and adding Illumina indexes during library preparation were those employed in Leaché et al. (2015). The barcodes differed by at least two base pairs to reduce the chance of errors caused by inaccurate barcode assignment. 33 Equimolar amounts of each sample were pooled, with each pool containing up to eight unique barcoded samples, in a 96-well plate format. Each pool was purified with Serapure 1.5X, rehydrated in 50 μL and quantified with a Qubit 2.0 fluorometer (Thermo Fisher Scientific) before size selection. The pooled libraries were size-selected (~500 pb) on a e-gel (Invitrogen) according to manufacturer instructions. The two external and internal lines (next to leader line) in the e-gel were not used, in the four lines available only two different libraries were run, with a maximum of 500 ng by line. The size-selected libraries were purified with a Qubit 2.0 fluorometer (Thermo Fisher Scientific) and amplified using PCR with the primers designed by Leaché et al. (2015) and using the Phire Hot Start II polymerase (Thermo Fisher Scientific). The amplified libraries were purified with Serapure 1.5X, quantified with a Qubit 2.0 fluorometer (Thermo Fisher Scientific). The fragment size distribution and concentration of each pool was determined on an Agilent BioAnalyzer and qPCR was performed to determine sequenceable library concentrations before multiplexing equimolar amounts of all 6 pools for sequencing on a single Illumina HiSeq 2000 lane (100 bp, single-end run) at the Vincent J. Coates Genomics Sequencing Laboratory at UC Berkeley. 2.3.2. ddRADseq bioinformatics We processed raw Illumina reads with the software pipeline ipyrad v0.5.15 (Eaton 2014), which consists of seven steps. We first demultiplexed the samples using their unique barcode and adapter sequences. Before of the second step, six samples with <50,000 reads passing the quality filter were excluded from further analyses. In the second step, reads were edited and filtered. The 6-bp restriction site overhang and the 5-bp barcode were removed. Sites with accuracy of the base call under 99% (Phred quality score = 20) were changed into ‘‘N” characters, and reads with >9 N’s (~10 %) were discarded. During the steps 3–6 the reads from each sample were clustered using the program 34 VSEARCH version 1.11.1 (https://github.com/torognes/vsearch) and aligned with MUSCLE version 3.8.31 (Edgar 2004). The first clustering step establishes homology among reads within samples. We determined the optimal value for the clustering parameter using the clustering threshold series approach described by Ilut et al. (2014). This method, which maximizes the account of cluster with two haplotypes, seeks to assemble reads into loci such that false homozygosity (splitting reads from a single locus into two) and false heterozygosity (due to clustering of paralogs) is minimized (i.e., the optimum clustering threshold). We generated a clustering threshold series (sensu Ilut et al. 2014) using similarity thresholds ranging from 0.85 to 0.98 for 19 samples. The optimal clustering threshold was 0.9 (Fig. 2a), which is used within and between sample clustering. After the clustering of reads within samples, we estimated the error rate and heterozygosity from the base counts in each site across all clusters, and these values were used to generate consensus sequences for each cluster. Consensus sequences were then clustered across samples and aligned as described above. Within 3–6 steps, we also discarded loci that had >4 undetermined or heterozygous sites (default ipyrad settings), or >2 haplotypes (to filter paralogs), and used a minimum depth of coverage of 6 for genotype calls. Following Nieto-Montes de Oca et al. (2017), we performed multiples replicates of the seventh step to determine the optimal value for the parameters: maximum numbers of SNPs allowed in a locus, maximum number of insertions/deletions allowed in across-sample clusters, and the maximum proportion of samples allowed to share a heterozygous site. The number of retained loci increased linearly with higher numbers of SNPs allowed until it began to plateau at a maximum of 30 SNPs (Fig. 2b). We used this value for the assembly of the final dataset under the rationale that above this value the small number of additional loci that were retained potentially represented paralogs. The number of retained loci increased with higher numbers of indels allowed, until a maximum of 8 indels (default ipyrad settings) (Fig. 2c). Thus, we permit a maximum of 8 insertions/deletions in across-sample clusters for the assembly of the final dataset. The number of loci retained increased until a value of 0.1 (corresponding to 9 samples), which 35 we again chose for the final value under the rationale that loci exhibiting higher shared heterozygosity potentially represented paralogs (Fig. 2d). We set the minimum number of ingroup samples with data for a given locus to be retained in the final dataset to about 25% of the samples (11). Unless otherwise stated, all phylogenetic analyses were performed with this dataset. FIGURE 2. a) Variation in the proportion of clusters with 1, 2, and >2 alleles retrieved with different similarity thresholds; b) variation in the number of retrieved loci with different maximum numbers of SNPs in a final locus; c) variation in the number of retrieved loci with different maximum numbers of insertions/deletions in across-sample clusters (c); d) variation in the number of retrieved loci with different maximum proportions of samples with a shared heterozygous site. The different line colors in a represent different numbers of alleles/cluster. 2.3.3. Phylogeny reconstruction 36 We used maximum likelihood and Bayesian methods to reconstruct a phylogenetic tree from the concatenated ddRAD loci in the final dataset. We ran RAxML, MrBayes, and BEAST analyses using the same parameters as the mitochondrial dataset, though only the GTR+I+G model was used for these analyses. We also performed RAxML, MrBayes, and BEAST analyses for each of the three clades within of the C. podiciferus species group (i.e. C. branfordii clade, C. podiciferus clade, and C. stejnegerianus clade). To perform the analyses by clade, we replicated the pipeline in ipyrad for each clade maximizing the number of retained loci. We did not attempt to estimate gene trees because the short read lengths cause the individual loci to be minimally phylogenetically informative (Nieto-Montes de Oca et al. 2017). 2.4. Species delimitation We used various methods to address species boundaries in the C. podiciferus species group and evaluated the effect of the tree used. The analyses were similar for datasets, mitochondrial and nuclear data. The species delimitation based on nuclear the dataset was ran using both the complete phylogeny and the partial phylogenies (see section 2.3.3), except for the BPP analysis, where only the partial phylogenies were used. We used three species discovery methods, GMYC, PTP, and mPTP that infer putative species limits on a given phylogenetic tree. The GMYC method (Pons et al. 2006; Fujisawa & Barraclough 2013) infers the transition point between inter- and intra-species branching rates on a time-calibrated ultrametric tree. We ran the GMYC analyses on the web server (http://species.h-its.org/gmyc/) under the single threshold and multiple threshold models using the ultrametric trees from the concatenated BEAST analyses (see above). The PTP method (Zhang et al. 2013) uses the number of substitutions to identify significant changes in the rate of branching in a phylogenetic tree (which may not be ultrametric). We ran PTP in the web server (http://species.h-its.org/ptp/) for 500,000 generations, with thinning = 100 and burn-in = 10%. The bPTP web server runs both the maximum likelihood 37 and Bayesian versions of PTP. Following a conservative approach, we considered as species those clades with a value greater than 0.01 (mitochondrial) or 0.05 (nuclear). The values above the nodes in PTP results represent posterior delimitation probabilities; the posterior probabilities of those taxa form one species. Given that PTP uses any tree completely bifurcated, we used the trees of RAxML, MrBayes, and BEAST to evaluate the effect of the tree as input. All analyses were replicated excluding the out-group. The third method, the mPTP (Kapli et al. 2016), is similar to the PTP because it uses the number of substitutions to identify significant changes in the rate of branching in a phylogenetic tree. However, mPTP incorporates different rates of coalescence within clades, allowing different levels of intraspecific genetic diversity. Similar to PTP, mPTP runs both ML and MCMC analyses. MCMC analyses were run for 100 million generations, sampling every 10,000 and the first 2 million generations were discarded as burn-in. All ML and Bayesian analyses were run both –single and –multi rates of coalescence among species and used a minimum branch length of 0.0001. Following a conservative approach, for the MCMC analyses we consider as species those clades with a value greater than 0.95 (mitochondrial) or 0.99 (nuclear). As in PTP, we used the trees of RAxML, MrBayes and BEAST to evaluate the effect of the tree as input. We used a fourth method, the Automatic Barcode Gap Discovery (ABGD; Puillandre et al. 2012). Unlike the three tree-based methods, it uses an aligned matrix. This method was used only for the mitochondrial dataset. Analyses were performed separately for 16S and COI. The method seeks to quantify the location of the barcode gap that separates intra- from interspecific distances. We used Pmin (0.01) and Pmax (0.1), JC69 and K80 corrected distances and relative gap width 1.5 (default) and 1.0. We also used BPP version 3.1 (Yang 2015; Yang & Rannala 2010, 2014) to jointly perform species delimitation and species tree inference under the multispecies coalescent model. We used the method A10 that evaluate the species delimitation from a guide tree, using a 38 rjMCMC algorithm (Rannala & Yang 2013). The assignment of individuals to putative species is fixed, and the model is to merge different specified groups into a single species, but not to split pre-defined populations into multiple species. We used the topology of BEAST analyses as user- specified guide tree. The analysis was run for a total of 2,500,000 iterations (sampling interval of 5) with a burn-in of 1000. We evaluated the influence of the ancestral population size (θ) and root age (G) considering three different combinations as in Leaché & Fujita (2010). The first combination of priors: θ ∼ G (1, 10) and θ ∼ G (1, 10. The second combination of priors: θ ∼ G (2, 2000) and θ ∼ G (2, 2000). The third combination is a mixture of priors: θ ∼ G (1, 10) and θ ∼ G (2, 2000). All BPP analyses used algorithm 0 with the fine-tuning parameter ε = 15 each with 500,000 generations (each fifth sampled) and a burn-in of 10,000 produced consistent results. With the nuclear dataset, the number of generations was 100,000. Each analysis was run at least twice to confirm consistency between runs. Following a conservative approach, only speciation events simultaneously supported by probabilities superior or equal to 0.99 for all three combinations of priors were considered for species delimitation. We also used the genetic distance for delimitation for the mitochondrial dataset. Although, not considered as a species delimitation method, the genetic distance has been used as indicator of separation interspecific. For amphibians, the 16S gene fragment has been suggested as a DNA barcode marker for amphibian diversity inventories (Vences et al. 2005) to complement the standard COI-5’ marker used in general for animals (Smith et al. 2008). Fouquet et al. (2007) suggested a threshold of 3% in 16S gene as significant for to identify possible distinct species. Vences et al. (2005) suggested a threshold of 10% in COI for identify candidate species. Estimates of pairwise evolutionary genetic divergence between groups were computed using MEGA6 (Tamura et al. 2013), assuming uncorrected distances based on the Tamura 3-parameter model (Tamura 1992), with rate variation among the sites modeled as a gamma distribution with the shape parameter = 4 as the default of the software. We evaluated the species identity with the 39 anterior methods. If the genetic distance between the clades was lower than the threshold the clades were collapsed. 2.5 Geographic range evolution To infer routes of dispersal or vicariance events, the distribution range of the Craugastor podiciferus species group was divided into five areas based on Terrestrial Ecoregions of the World proposed by Olson et al. (2001): (A) Costa Rican seasonal moist forest, (B) Isthmian- Atlantic moist forest, (C) Isthmian-Pacific moist forest, (D) Talamancan montane forest and (E) Choco-Darien moist forest. We used the R package BioGeoBEARS (Matzke 2014) and implemented the LAGRANGE DEC model (Ree & Smith 2008) within a maximum likelihood framework. Furthermore, a founder-event speciation was added to any of these models and estimated as an additional free parameter j. Because no species is distributed over more than four defined areas, we set the maximum number of areas to four. 2.6 Relaxed molecular clock analysis We used the BEAST calibrated tree to infer the ancestral area probability. The lack of C. podiciferus in the fossil record makes dating divergences based on molecular sequence data difficult. We use the dating for the group estimated by Streicher et al. (2009) to make a rough estimate of divergence times within the species group. Following Streicher et al. (2009) and assuming that ICA has emerged ~25 Mya, we restricted the origin of the C. podiciferus species group to 25 Mya. Details for the analyses of BEAST were described above. 40 FIGURE 3. Maximum clade credibility tree from the concatenated BEAST analysis of Craugastor podiciferus species group based on 16S and COI mitochondrial DNA genes. Posterior probabilities from BEAST are shown. The scale bar refers to the estimated substitutions per site. The asterisks represent support of >95. 41 • • ~ C. aenigmGliclIs cla ~ d ~ ' ~_ ! . [=== . [====~~ . ~ ", podicifc:/'IIs c1ade I " . C. slejllegeriolllts c1ade " . • e branifordii clade .71 • 0.51 • • • ,K, Monte Verde Monte Verde : Monte Verde ¡ ~~ ! ~ ~¡j!' t r , Sjola · ~'i~g~BI~~g : ;¡Il c",rdo , JIl crardo n c",rdo : ~,\",l,lCCr;1 , ;hul§Jccra · ,la oSlclla · Fila OSlCll a , F! a OSlc"" ,FIla oSlclla 3. Results 3.1 Phylogeny of the Craugastor podiciferus species group 3.1.1. Mitochondrial phylogenies. The resulting mitochondrial data matrix included 96 sequences with a total sequence length of 1222 bp including gaps (565 bp for 16S and 657 bp for COI). The following substitution models were selected: GTR+I+G for 16S, K80+I+G for COI codon position 1, HKY+I for COI codon position 2, and GTR+G for COI codon position 3. The mitochondrial genetic distances are shown in Table 1. The phylogeny inferred with BEAST (Fig. 3) found the C. podiciferus species group to be composed of four major clades. These clades correspond to the three previously recognized morphological complexes (C. bransfordii, C. podiciferus, and C. stejnegerianus), and an additional basal clade of a group that was described recently. The first basal clade, C. aenigmaticus, is composed of samples from the Cordillera de Talamanca. The second major clade, the C. podiciferus clade, contains two well-supported subclades. The first includes samples from La Nevera, Alturas, and Fortuna from the highlands of southwestern Costa Rica and western Panama, and the second has samples of the type locality of C. podiferus and several populations from the highlands of Costa Rica and western Panama. The C. stejnegerianus clade comprises six groups: C. persimilis, C. gabbi, C. rearki, C. stejnegerianus, Craugastor sp. Neilly, and Craugastor sp. Quepos. This clade ranges from eastern Honduras to western Panama. The fourth major clade, C. bransfordii clade is composed of six groups: C. bransfordii, C. underwoodi, Craugastor sp. Fila Carbon, Craugastor sp. Panama, Craugastor sp. Quebradas, and Craugastor sp. Vereh. 42 FIGURE 4. Bayesian phylogram derived from MrBayes of the Craugastor podiciferus species group based on the 16S and COI mitochondrial DNA gene markers. Posterior probabilities of clades and Bootstraps values are before and after slashes, respectively. The scale bar refers to the estimated substitutions per site. The asterisks represent posterior probability values > 0.95. 43 e gol/",,,"; 0.07 T ab le 1 . M ea n un co rr ec te d ge ne ti c di st an ce s am on g li ne ag es o f C ra ug as to r po di ci fe ru s sp ec ie s gr ou p us in g 16 S ( ab ov e) a nd C O I (b el ow ) m it oc ho nd ri al g en es . G en et ic d is ta nc es w it hi n th e th re e m aj or c la de s ar e hi gh li gh te d. V al ue s le ss th at th e th re sh ol ds ( 3% in 1 6S a nd 1 0% in C O I) a re sh ow n in r ed . ID 16 S/ C O I 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 1 C ra ug as to r ae ni gm at ic us –– 17 .5 13 .3 15 .6 15 .0 14 .2 14 .4 15 .1 14 .8 14 .8 17 .3 –– 14 .2 16 .7 16 .3 16 .2 16 .8 18 .7 20 .3 19 .6 18 .9 19 .6 21 .7 2 C ra ug as to r sa gu i 21 .8 –– 7. 9 7. 4 7. 3 7. 1 7. 4 9. 5 8. 2 8. 4 9. 1 –– 11 .1 17 .5 12 .0 11 .1 12 .1 12 .8 16 .1 16 .2 16 .2 16 .3 18 .0 3 C ra ug as to r zu ni ga i 25 .0 17 .3 –– 3. 4 7. 2 6. 2 7. 3 9. 3 9. 4 9. 6 8. 9 –– 12 .3 17 .0 13 .2 13 .7 13 .1 13 .8 16 .3 16 .4 15 .7 16 .4 19 .4 4 C ra ug as to r bl ai ri 20 .2 20 .3 18 .3 –– 7. 6 7. 4 7. 6 10 .5 8. 3 9. 9 9. 4 –– 13 .3 17 .7 12 .0 12 .5 13 .0 14 .1 17 .0 17 .1 17 .0 17 .2 21 .0 5 C ra ug as to r sp . M on te V er de 19 .5 16 .6 18 .7 19 .9 –– 2. 5 2. 0 3. 7 6. 6 4. 4 4. 5 –– 9. 1 14 .9 9. 9 11 .1 10 .4 11 .5 14 .6 14 .7 14 .6 14 .8 16 .6 6 C . p od ic if er us 22 .2 17 .8 18 .5 19 .8 12 .0 –– 2. 7 4. 1 6. 6 4. 6 5. 1 –– 7. 6 13 .6 10 .1 11 .2 9. 6 11 .5 13 .2 13 .2 13 .1 13 .2 15 .9 7 C ra ug as to r sp . P ic o B la nc o 20 .8 18 .4 20 .7 17 .2 11 .3 10 .0 –– 2. 6 6. 9 4. 7 3. 1 –– 9. 8 14 .5 10 .4 11 .7 9. 9 11 .8 15 .2 15 .3 15 .4 15 .3 14 .9 8 C ra ug as to r sp . S an G er ar do 21 .9 18 .4 21 .4 20 .5 12 .8 12 .4 7. 3 –– 7. 8 5. 8 4. 5 –– 9. 8 14 .2 10 .9 11 .7 9. 2 13 .4 15 .3 15 .4 15 .5 15 .4 13 .7 9 C ra ug as to r sp . S io la 24 .0 19 .8 22 .8 19 .6 12 .8 12 .9 11 .4 13 .7 –– 8. 0 9. 0 –– 11 .0 15 .8 9. 2 10 .3 10 .5 11 .8 13 .0 11 .5 11 .5 11 .6 16 .1 10 C ra ug as to r sp . C hu m ac er a 19 .3 19 .3 21 .3 18 .3 12 .4 11 .8 9. 9 12 .8 11 .6 –– 7. 4 –– 11 .2 17 .6 11 .8 12 .1 12 .7 14 .2 17 .8 16 .8 16 .8 16 .9 19 .9 11 C ra ug as to r sp . F ila C os te ña 21 .2 17 .8 23 .4 19 .9 12 .5 12 .3 12 .0 14 .4 13 .8 9. 6 –– –– 12 .0 16 .5 11 .4 12 .7 10 .9 12 .7 14 .2 15 .9 15 .8 16 .0 16 .0 12 C ra ug as to r sp . Q ue br ad as 25 .4 19 .2 22 .0 24 .5 24 .2 22 .6 24 .2 25 .3 26 .3 22 .8 25 .2 –– –– –– –– –– –– –– –– –– –– –– –– 13 C ra ug as to r sp . V er eh 26 .9 20 .2 23 .4 23 .9 19 .7 22 .6 23 .8 23 .7 26 .0 20 .8 22 .6 16 .2 –– 5. 4 6. 0 6. 8 5. 0 9. 5 11 .6 10 .2 10 .2 10 .3 10 .2 14 C . b ra ns fo rd ii 25 .3 18 .8 21 .1 23 .0 21 .1 19 .9 20 .6 22 .3 24 .0 20 .8 21 .2 16 .7 11 .7 –– 10 .4 11 .3 9. 4 14 .4 14 .1 14 .1 14 .1 14 .2 13 .4 15 C . u nd er w oo di 23 .5 19 .1 21 .8 19 .9 19 .4 19 .7 19 .7 22 .4 22 .9 20 .9 21 .2 15 .3 14 .5 13 .2 –– 4. 1 5. 4 8. 5 11 .5 10 .9 10 .9 11 .0 12 .1 16 C ra ug as to r sp . F ila C ar bo n 22 .9 20 .7 23 .4 22 .9 22 .4 23 .3 23 .9 26 .3 24 .9 22 .5 23 .0 17 .3 14 .0 15 .6 9. 5 –– 4. 3 7. 7 12 .2 11 .6 11 .5 11 .6 11 .6 17 C ra ug as to r sp . P an am a 21 .3 18 .2 22 .5 22 .2 21 .1 21 .1 19 .8 23 .4 24 .1 21 .5 22 .1 17 .3 14 .3 13 .2 11 .9 14 .0 –– 7. 6 11 .8 10 .5 10 .4 10 .5 8. 8 18 C . p er si m il is 24 .2 22 .4 26 .5 24 .0 23 .0 23 .8 24 .9 27 .4 24 .1 26 .2 26 .4 23 .8 20 .8 21 .8 19 .1 19 .6 20 .9 –– 7. 6 6. 3 7. 0 6. 5 9. 0 19 C . g ab bi 25 .4 26 .7 27 .6 24 .3 24 .1 23 .5 24 .7 26 .3 26 .7 26 .7 26 .9 22 .0 20 .1 23 .3 20 .6 19 .6 21 .5 16 .7 –– 4. 2 4. 3 4. 3 8. 0 20 C ra ug as to r sp . N ei lly 24 .9 27 .5 27 .4 22 .6 24 .6 27 .2 26 .9 28 .9 29 .3 29 .8 29 .2 24 .2 22 .4 22 .9 20 .5 20 .4 22 .6 16 .3 14 .1 –– 1. 1 0. 2 6. 0 21 C ra ug as to r sp . Q ue po s 23 .0 27 .3 27 .4 22 .6 24 .6 25 .3 26 .0 26 .2 27 .3 28 .4 26 .2 23 .6 23 .4 22 .7 19 .1 20 .1 22 .2 15 .3 15 .1 6. 9 –– 1. 1 6. 1 22 C . s te jn eg er ia nu s 21 .8 27 .5 25 .2 22 .3 25 .1 25 .5 25 .8 27 .0 28 .7 27 .8 27 .2 22 .6 21 .6 21 .6 20 .9 20 .3 21 .9 16 .2 14 .7 5. 6 5. 5 –– 6. 0 23 C . r ea rk i 22 .6 25 .4 24 .9 23 .8 23 .6 28 .0 27 .6 28 .6 28 .3 27 .9 27 .7 23 .4 20 .8 22 .0 22 .2 20 .0 21 .1 17 .2 15 .7 11 .9 11 .6 13 .8 –– 44 The mitochondrial phylogenies inferred with MrBayes (Fig. 4) and RAxML (not shown) were highly discordant with the BEAST phylogeny (Fig 3). Only the Craugastor aenigmaticus clade and the C. stejnegerianus clade were recovered with the same internal relationships. Within of the C. podiciferus clade, the Craugastor zunigai + Craugastor blairi subclade was recovered and the relationships of the subclade that includes the type locality of C. podiciferus was similar to the BEAST topology. Within the C. bransfordii clade, only the subclades C. bransfordii + Craugastor sp. Vereh and C. underwoodi + Craugastor sp. Fila Carbon were recovered. 3.1.1. Nuclear phylogenies. The Illumina HiSeq2500 lane generated 66,364,543 reads demultiplexed to 48 samples. The total number of retained loci in the nuclear data matrix was of 7,770 (697,771 characters; 33% of missing data). The phylogenies inferred with RAxML, MrBayes and BEAST had the same topology (Fig. 5). The C. podiciferus species group was composed of four major clades, which were similar in composition to the one obtained by the mitochondrial data set and the BEAST analysis, only with differences on some internal relationships. The basal clade, Craugastor aenigmaticus, is formed by samples from the montane rainforest of the Cordillera de Talamanca. The C. podiciferus clade is composed of two subclades, one contain to C. blairi, C. sagui, and C. zunigai along the highlands of southwestern Costa Rica and western Panama. A second subclade includes C. podiciferus sensu stricto and other populations from the highlands of ICA (1000– 2700 m a.s.l.). Several populations phylogenetically close to C. podiciferus in the mitochondrial phylogeny were not included in the nuclear analysis. 45 FIGURE 5. Maximum clade credibility tree from the concatenated BEAST analysis of the Craugastor podiciferus species group based on ddRAD dataset. Posterior probabilities from BEAST are shown. The scale bar refers to the estimated substitutions per site. The asterisks represent posterior probability values > 0.95. 46 The C. bransfordii clade was composed of C. bransfordii, C. underwoodi, Craugastor sp. Fila Carbon, Craugastor sp. Panama, and Craugastor sp. Vereh which occur from northern Nicaragua to central Panama. The relationships within the C. bransfordii clade differ from the mitochondrial topology. In the nuclear tree, the lineage Craugastor sp. Panama is sister to the other lineages within the clade (essentially from Costa Rica). The fourth major clade (C. stejnegerianus clade) was highly discordant with the mitochondrial phylogeny. In the nuclear tree, C. stejnegerianus is paraphyletic, with some populations of C. stejnegerianus (Quepos) sister to C. rearki, whereas C. stejnegerianus (sensu stricto) was sister to C. gabbi + other population of C. stejnegeriaus (Neilly). The basal position of C. persimilis from Caribbean Costa Rica was identical in mitochondrial and nuclear phylogenies. 3.2 Species delimitation The results from the species delimitation analyses are shown in Figs 6 and 7. Considerably differences were found between the mitochondrial and nuclear data sets and between the different methods. In the mitochondrial analyses, some methods (i.e. GMYC and PTP) identified > 60 species, though others (i.e. mPTP Bayesian) only identified 13 species. With the mitochondrial data, all methods identified all described and nearly all-putative species except the three lineages within the species C. stejnegerianus that were identified in the nuclear analysis (Craugastor sp. Neilly, Craugastor sp. Quepos, and C. stejnegerianus). The nuclear data, which only had 42 specimens, had several differences among the species delimitation methods, some (i.e. mPTP Bayesian with MrBayes and RAxML trees) only recognized a single species for the entire complex whereas others (i.e. mPTP maximum likelihood) identified up to 36 species. Incongruence as also found between analyses with the 47 complete phylogeny and those with the phylogenies by clade. Within the C. stejnegerianus clade, the GMYC and PTP methods identified more species using the phylogeny by clade than the complete phylogeny. However, within the C. podiciferus clade, the GMYC and PTP methods identified more species using the complete phylogeny than using the phylogeny for the clade. These results highlight the impact of the selection of the initial phylogeny used in the methods GMYC, PTP, and mPTP, which are tree-based methods. FIGURE 6. Comparison of species delimitation results of the Craugastor podiciferus species group based on the 16S and COI mitochondrial DNA gene markers. Each colored bar represents a species delimited by each method tested. 48 FIGURE 7. Comparison of species delimitation results of the Craugastor podiciferus species group based ddRAD dataset. Each colored bar represents a species delimited by each method tested. The nuclear data, which only had 42 specimens, had several differences among the species delimitation methods, some (i.e. mPTP Bayesian with MrBayes and RAxML trees) only recognized a single species for the entire complex whereas others (i.e. mPTP maximum likelihood) identified up to 36 species. Incongruence as also found between analyses with the complete phylogeny and those with the phylogenies by clade. Within the C. stejnegerianus clade, 49 the GMYC and PTP methods identified more species using the phylogeny by clade than the complete phylogeny. However, within the C. podiciferus clade, the GMYC and PTP methods identified more species using the complete phylogeny than using the phylogeny for the clade. These results highlight the impact of the selection of the initial phylogeny used in the methods GMYC, PTP, and mPTP, which are tree-based methods. 3.3 Biogeography The result of the ancestral area reconstruction is presented in combination with the BEAST tree for ddRAD data (Fig. 8). A Talamancan origin for the C. podiciferus species group was inferred during the Middle Miocene. A Talamancan origin for the C. podiciferus clade was also supported, with a dispersal event to Isthmian-Pacific moist forest (lowland) and another dispersal event to Costa Rican seasonal moist forest. The origin of the clade C. stejnegerianus + C. bransfordii was supported as Isthmian-Atlantic moist forest, with five independent dispersal events from the Isthmian-Atlantic moist forest to Talamancan montane forest during the Pleistocene and three dispersal events in Isthmian-Pacific moist forest, explaining the current patterns of distribution. The elevational distribution of the 23 lineages identified is shown in Fig. 9. The two basal clades, Craugastor aenigmaticus clade (in red) and C. podiciferus clade (in blue), are restricted to highlands (1000–2700 m a.s.l.), however, all the 23 lineages have populations above 750 m a.s.l., at the Talamanca montane forest (Fig. 1, 9). The lineages distributed at highlands have narrower elevational ranges (Fig. 9); all lineages distributed under 1000 m a.s.l. have altitudinal ranges higher than 750 m but four lineages restricted to highlands have altitudinal ranges smaller than 250 m. In addition, the highlands of the Pacific versant are more structured; in the Pacific versant eight lineages are distributed exclusively over 1000 m a.s.l. whereas in the Atlantic versant only five lineages are restricted to highlands. 50 FIGURE 8. Summary of results of LAGRANGE biogeographical analysis. 4. Discussion 4.1 Systematics and biogeography of the C. podiciferus species group The highlands of ICA played an important role in the diversification of several groups of vertebrates (Savage 2002; García-París et al. 2000; Castoe et al. 2009; Boza-Oviedo et al. 2012; 51 Duellman et al. 2016; Arias et al. 2018). However, there is controversy in the chronological dating of the processes responsible for this diversification. Based on the evidence of climatic oscillations and assuming a recent uprising of Isthmian Central America in the Pliocene- Pleistocene (~ 5 Mya), Savage (2002) proposed a model of speciation for highland taxa (Montane speciation). This author suggested that the oscillations between glacial and interglacial periods that began in the Pliocene (~ 5 Mya) and continued during the Pleistocene were responsible for the origin and distribution of many highlands species. Nevertheless, phylogeographic analyses of the frog Craugastor podiciferus (Streicher et al. 2009) and Neotropical snakes (Atropoides, Bothriechis, and Cerrophidion; Castoe et al. 2009) have disagreed with these events in the Pliocene as a major factor associated with cladogenesis of these highland species. Castoe et al. (2009) found that the diversity of Middle American highland pitvipers is explained by diversification during the Miocene and Pliocene. The chronological results shown by Castoe et al. (2009) partly overlaps (in the Pliocene) with the time suggested by Savage (2002), however it predates the Pleistocene. Streicher et al. (2009) estimated that the time of origin for the most recent common ancestor (MRCA) of the C. podiciferus species group occurred between 16.82 and 27.84 Mya. The origin of the C. podiciferus species group is explained by the events of dispersion from Nuclear Central America to ICA when the last emerged as a peninsula at its current position (Montes et al. 2015). Our results indicate that the four major clades diverged during the Miocene, supporting Castoe et al.’s (2009) results. In addition, it is possible that climatic fluctuations occurred during the Pliocene-Pleistocene favored the origin of several lineages within the C. podiciferus species group. Streicher et al. (2009) argued that climatic fluctuations during the Pliocene-Pleistocene favored the diversification within the C. podiciferus clade. These 52 climatic fluctuations affected the Pacific slopes mainly (Savage 1966; Crawford et al. 2007); that explains their high genetic structure, with species having narrow elevational and latitudinal ranges (Fig. 1, 9). Also, the Pacific slopes have more lineages at highlands and less at lowlands compared to the Caribbean slope. On the Caribbean slopes the lineages having broad elevational and latitudinal ranges, this distribution could indicate that the Caribbean slope was less impacted by climatic fluctuations in the past. However, the fewer species at intermediate lands of Caribbean slopes can also be explained by the lack of surveys in that region and could increase with more surveys conducted in the future. The Craugastor podiciferus species group is distributed from eastern Honduras to central Panama. The 23 identified lineages have populations in Costa Rica and western Panama in an area smaller than 40,000 km2. The Craugastor aenigmaticus clade is restricted to Montane Rainforest, ranging from 2330–2700 m a.s.l, the highest altitudinal distribution for a species in this group (Arias et al. 2018). Given that Craugastor aenigmaticus is basal within the species group, it is suggested that the ancestor was distributed on the Montane zonation and diversified to lower elevations. The clade C. podiciferus, as shown by Streicher et al. (2009), possibly diversified due to climatic fluctuations that isolated the suitable habitat on peaks in the mountain ranges. The C. bransfordii + C. stejnegerianus clade contains species ranging from sea level to 1600 m a.s.l. The C. bransfordii clade diversified mainly on the Caribbean slopes of Nicaragua, Costa Rica, and Panama, and only one lineage was found on the Pacific slopes of Costa Rica. No obvious barriers separate the species of the C. bransfordii clade, except for paleo-climatic differences. The C. stejnegerianus clade contains more species on the Pacific slope of Costa Rica, with two species distributed on the Caribbean slopes of Nicaragua, Costa Rica, and Panama. Possibly, the driest conditions of the Pleistocene on the pacific slope of Costa Rica 53 (Crawford et al. 2007) marked the diversification of this clade on the pacific slope from isolated ancestors of the species on wet patches. FIGURE 9. Elevational distribution of the lineages within of the Craugastor podiciferus species group, color bars correspond with the four major clades. The gray bar represents the continental divisor. 4.2 Species delimitation and taxonomic comments Seven described and putative species were supported as distinct evolutionary lineages in all analyses. Nevertheless, large differences were found between mitochondrial and nuclear analyses 54 and between methods. There is not a consensus for taxonomic decision from the delimitation results, and our suggestions are based on both mitochondrial and nuclear analyses. We identified 23 lineages, including described species and putative species. This approach aims to minimize taxonomic instability. Below are details the major clades and its species. 4.2.1 The Craugastor aenigmaticus clade. This clade was consistently supported as a distinct evolutionary lineage in all analyses. This species was named recently, is microendemic to the Montane Rainforest of the Cordillera de Talamanca, southwestern Costa Rica, ranging from 2390–2700 m a.s.l. (Arias et al. 2018). This species corresponds with the higher altitudinal distribution for the species group. This species is separated from other members of the C. podiciferus species group by very significant genetic distances, with mean uncorrected genetic distances higher than 13.3 % in 16S and 19.3 % in COI (Table 1). 4.2.2. The Craugastor podiciferus clade The species C. blairi, C. sagui, and C. zunigai each represent separate species from the topotypic C. podificiferus specimens. Several delimitation analyses identified these populations as separate evolutionary lineages. The mitochondrial phylogenies (RAxML and Bayesian) did not recovered these populations clustered with C. podiciferus. Instead, they were placed at the base of the complex. The three species are allopatric, distributed latitudinally in southwestern Costa Rica and western Panama. The clade Craugastor blairi corresponds to the species Craugastor sp. B of Crawford & Smith (2005) and Streicher et al. (2009). These three clades are separated from each other by mean uncorrected genetic distances higher than 3.4 % in 16S and 17.3 % in COI (Table 1). 55 The number of species that form the C. podiciferus subclade, that also includes populations from the type locality of C. podiciferus, is unclear. Several species were consistently identified with mitochondrial sequence data, but only some were included in the nuclear analysis. Streicher et al. (2009) performed a phylogenetic analysis for this clade, suggesting that it was composed of six species. Of the seven putative species in the mitochondrial analysis, only three were included in the nuclear data set, which were supported as separate species in BPP analyses. We identified seven clades within this subclade, which differ morphologically (Arias Unpublished data) and are separated from each other by mean uncorrected genetic distances higher than 2.0 % in 16S and 7.3 % in COI (Table 1). Thus, we suggest that these seven clades may represent separate species. We recognize seven species within the C. podiciferus clade, four of Streicher et al. (2009) and three additional putative species that were no included in any previous work. Streicher et al. (2009) estimated a time to MRCA of the C. podiciferus clade of between 4.70 and 8.18 Ma. We hypothesize that the lack of support for the distinctness of these taxa in some analyses may reflect the fact that they are recently derived, and argue that they are nonetheless on evolutionary independent trajectories and therefore should be recognized as separate species. We included for the first time populations from the type locality of C. podiciferus. This type locality was scope of discussion by Savage (1970) and Arias & Chaves (2014), who later corrected the type locality to Caribbean slopes of Cerro Kamuk. We restrict C. podiciferus to the populations of Cordillera Volcánica Central from Costa Rica and Cordillera de Talamanca on Costa Rica and western Panama. In the Cordillera de Talamanca C. podiciferus is restricted to the Caribbean slopes. This clade corresponds to the clades C and D of Streicher et al. (2009). The clade Craugastor sp. Monte Verde corresponds to the clade A of Streicher et al. (2009), a 56 species restricted to the Cordillera de Tilarán and Volcánica Central. The clade Craugastor sp. San Gerardo corresponds to the clade B of Streicher et al. (2009), a species distributed on Cordillera de Tilarán and Volcánica Central. In Monte Verde both species, Craugastor sp. Monte Verde and Craugastor sp. San Geardo, are in plausible simpatry due to the proximity of both localities. However, the micro-sympatry remains unknown. Also, an occurrence in sympatry of C. podiciferus and Craugastor sp. Monte Verde seems possible. The clade Craugastor sp. Fila Costeña corresponds to the clades E and F of Streicher et al. (2009), a species restricted to South Pacific Costa Rica. The clade Craugastor sp. Pico Blanco is known only by a population on the Valle Central. The clade Craugastor sp. Chumacera is known only by a population on the Pacific slopes of the Cordillera de Talamanca, just as the clade Craugastor sp. Siola that is known from a single population on the Caribbean slope of the Cordillera de Talamanca. The species C. jota was referred to the C. podiciferus species group (Hedges et al. 2008) and assigned to the C. podiciferus clade based in morphology. This species was described with specimens from the Changena River in western Panama, from 760 m a.s.l. collected by Linda Trueb in 1966. In the description, Lynch (1980) defined the type locality on the basis of the map shown by Trueb (1968). However, the map was drawn without GPS coordinates and the rivers name could not match with the now rivers names. Here were included specimens from Changena River, western Panama, very near to the type locality of C. jota (Linda Trueb pers. comm.). These specimens were grouped within C. podiciferus sensu stricto, because we suggest C. jota should be referred as a junior synonym of C. podiciferus. 4.2.3. The Craugastor bransfordii clade The nuclear phylogeny supports the monophyly of the C. bransfordii clade, but it was not 57 supported in some mitochondrial analyses. We suggest that the C. bransfordii clade is composed of six separate species, only five of which were included in the nuclear analysis. These six lineages are separated from each other by mean uncorrected genetic distances higher than 4.1 % in 16S and 9.5 % in COI (Table 1). Craugastor bransfordii included specimens (UCR20559) collected near the type locality, San Juan River on the Costa Rica-Nicaragua border. Craugastor bransfordii is distributed from north Nicaragua to central Caribbean Costa Rica.The nominal species C. polyptychus was described with specimens from the same type locality as C. bransfordii and in the same publication (Cope 1886). It was seen as a synonym of C. bransfordii (Savage & Emerson 1970; Miyamoto 1983) until Savage (2002) tentatively resurrected this name and assigned it to specimens from Caribbean Costa Rica, but pointed out that further taxonomic work is needed to clarify the status of the species. Since only one clade within the C. bransfordii clade was found in the northern Costa Rica and southern Nicaragua we suggest that C. polyptychus should be referred as a junior synonym of C. bransfordii. The clade Craugastor sp. Fila Carbon corresponds –in part– to C. polyptychus of Savage (2002). This clade represents a putative species undescribed distributed in south Caribbean Costa Rica and western Panama. The clade C. underwoodi includes specimens from Cascajal and Cinchona, near the type locality. This species is distributed in the premontane forest of the Cordillera de Guanacaste, Tilarán, Volcánica Central, and the northern edge of the Cordillera de Talamanca. The clade Craugastor sp. Quebradas includes specimens from the only locality known on the Pacific slopes for the C. bransfordii clade. We consider that the clade Craugastor sp. Quebradas represents a separate species due to its allopatric distribution and the large genetic distances to all other samples of 15.3 % or higher in COI (Table 1). The clade Craugastor sp. Vereh includes specimens from two localities in the premontane forest in the central Caribbean of Costa Rica. 58 The phylogenetic position of this clade varies between mitochondrial and nuclear phylogenies. Finally, the clade Craugastor sp. Panama is composed of populations from Panama; these populations were attributed to C. bransfordii (Leenders 2016), but based on the mitochondrial and nuclear phylogenies this clade from Panama is not closely related to C. bransfordii. We suggest that these populations from Panama represent at least one separate species. 4.2.4. The Craugastor stejnegerianus clade Within the C. stejnegerianus clade large differences were found between the mitochondrial and nuclear phylogenies, mainly in the relationships between C. gabbi and C. stejnegerianus. Arias et al. (2016) recently described C. gabbi and discussed the differences between C. gabbi and C. stejnegerianus. We suggest that the C. stejnegerianus clade is composed of six separate lineages. The clade C. persimilis includes specimens from Suretka, Talamanca, what is the type locality of that species. Craugastor persimilis is distributed on the central and south Caribbean slopes of Costa Rica. The specimens from Honduras and Nicaragua that were initially referred to C. lauraster are closely related to specimens from central Caribbean of Costa Rica. The populations from central Caribbean of Costa Rica correspond to C. rearki (Taylor 1952), synonymized under C. bransfordii by Savage & Emerson (1970). We suggest that the name C. rearki should be resurrected to include populations from Caribbean of Costa Rica, Nicaragua, and Honduras and C. lauraster should be referred as a junior synonym of C. rearki. The populations of the C. stejnegerianus clade on the pacific slope are highly conflictive in the mitochondrial and nuclear phylogenies. Arias et al. (2016) supported the distinctiveness of this species based on their mitochondrial phylogeny, morphology, and ecological preferences. However, in the nuclear phylogeny C. gabbi is sister to C. stejnegerianus sensu stricto. If both 59 species are recognized, it would also be necessary to recognize Craugastor sp. Neilly and Craugastor sp. Quepos as separate species. The clade Craugator sp. Quepos is the sister clade to C. rearki because we suggest that this represent a separate species distributed in the central pacific and Central Valley of Costa Rica. We also recognized the clade Craugastor sp. Neilly as a separate evolutionary lineage restricted to southeast Pacific of Costa Rica. Craugastor stejnegerianus is restricted to south Pacific of Costa Rica. To exclude the three species within the C. stejnegerianus (Craugastor sp. Neilly, Craugastor sp. Quepos, and C. stejnegerianus) that were identified only in the nuclear analysis, these are separated from each other by mean uncorrected genetic distances higher than 4.2 % in 16S and 11.6 % in COI (Table 1). The phylogenetic relationships and the delimitation of species suggest the presence of several undescribed, overlooked species, highlighting the problem of potential cryptic species. The clades herein identified should be morphologically re-evaluated following the phylogenetic relationships. However, a preliminary morphological review of the genetic lineages within the C. bransfordii clade show few morphological characteristics that would allow for a differentiation of those lineages. Similar problems occur within the C. stejnegerianus clade and it is very difficult to identify these lineages using morphological characters with the exception of C. persimilis (Arias et al. 2016). It is very important that extensive studies about the use of habitat, acoustic, behavior, alimentation, and other data will be carried out to better understand the dynamic of these species. 5. Conclusion The C. podiciferus species group represents an ideal model for studies on amphibian overlooked diversity, given its high abundance, wide geographic distribution (collectively), high genetic 60 diversity, high levels of polymorphism, direct development (lack of larval stage), some species are miniaturized (C. persimilis reaches to 23 mm in female adults), and only some vocalization species are known (intraspecific recognition unknown). However, to date there are only few studies on the members of this species group due the lack of clear limits between its species. Following our objectives, we report 1) the first phylogeny for the C. podiciferus species group sampling extensively its entire geographic distribution, 2) although there is disagreement about the species delineation methods that were performed, 23 lineages were consistently recovered, revealing several undescribed species, and 3) the diversification of this group is strongly associated to the highlands of Isthmian Central America, all lineages ranges above the 750 m. a.s.l. Acknowledgments We thank Laura Márquez-Valdelamar and Andrea Jiménez-Marín for their laboratory assistance; Federico Bolaños for the use of specimens from the Museo de Zoología of the Universidad de Costa Rica; Omar Zúñiga, Olmer Cordero, Justo Layam Gabb, and Xavier Baltodano provided valuable assistance in the field during the expeditions. EA thanks the Posgrado en Ciencias Biológicas for its support of this study, the CONACyT for the students grant (CVU/Becario) 626946/330343, and the Programa de Innovación y Capital Humano para la Competitividad PINN-MICITT for the students grant (PED-0339-15-2). Laboratory work was funded by grant from PAPIIT-UNAM (IN203617) to GP-O. Fieldwork was partially supported by the National Geography Society (Grant number W-346-14). We acknowledge the Costa Rican Ministry of Environment and Energy (MINAE) for providing the corresponding scientific collecting permits for this expedition (SINAC-SE-GAS-PI-R 007-2013 and 59-2015). 61 Literature cited AmphibiaWeb (2019) AmphibiaWeb: Information on Amphibian Biology and Conservation [web application]. Berkeley, California. Available from: http://amphibiaweb.org (accessed 10 January 2019). Anger, G.R. & Dean, R. (2010) The birds of Panama: a field guide. Cornell University Press, Nueva York, 456 pp. Arias, E. & Chaves, G. (2014) 140 years after William M. Gabbs's climb to Cerro Pico Blanco. Mesoamerican Herpetology, 1: 176–180. Arias, E., Chaves, G., Crawford, A.J. & Parra-Olea, G. (2016) A new species of the Craugastor podiciferus species group (Anura: Craugastoridae) from the premontane forest of southwestern Costa Rica. Zootaxa, 4132: 347–363. Arias, E., Chaves, G. & Parra-Olea, G. (2018) A new species of Craugastor (Anura: Craugastoridae) from the montane rainforest of the Cordillera de Talamanca, Costa Rica. Phyllomedusa, 17: 211–232. Bagley, J.C. & Johnson, J.B. (2014) Phylogeography and biogeography of the lower Central American Neotropics: diversification between two continents and between two seas. Biological Reviews, 89: 767–790. Batista, A., Köhler, G., Mebert, K., Hertz, A. & Veselý, M. (2016) An integrative approach to reveal speciation and species richness in the genus Diasporus (Amphibia: Anura: Eleutherodactylidae) in eastern Panama. Zoological Journal of the Linnean Society, 178: 267–311. Boza-Oviedo, E., Rovito, S.M., Chaves, G., García-Rodríguez, A., Artavia, L.G., Bolaños, F. & Wake, D.B. (2012) Salamanders from the eastern Cordillera de Talamanca, Costa Rica, with descriptions of five new species (Plethodontidae: Bolitoglossa, Nototriton, and Oedipina) and natural history notes from recent expeditions. Zootaxa, 3309: 36–61. 62 Campbell, J.A. (1999) Distribution patterns of amphibians in Middle America. In: Duellman, W.E. (Ed.), Patterns of Distribution of Amphibians: A Global Perspective. Johns Hopkins University Press, Maryland, pp. 111–210. Castoe, T.A., Daza, J.M., Smith, E.N., Sasa, M.M., Kuch, U., Campbell, J.A., Chippindale, P.T. & Parkinson, C.L. (2009) Comparative phylogeography of pitvipers suggests a consensus of ancient Middle American highland biogeography. Journal of Biogeography, 36: 88–103. Cope, E.D. (1886) Thirteenth contribution to the herpetology of tropical America. Proceedings of the American Philosophical Society, 23: 271–287. Crawford, A.J. (2003) Huge populations and old species of Costa Rican and Panamanian dirt frogs inferred from mitochondrial and nuclear gene sequences. Molecular Ecology, 12: 2525–2540. Crawford, A.J. & Smith, E.N. (2005) Cenozoic biogeography and evolution in direct-developing frogs of Central America (Leptodactylidae: Eleutherodactylus) as inferred from a phylogenetic analysis of nuclear and mitochondrial genes. Molecular Phylogenetics and Evolution, 35: 536–555. Crawford, A.J., Bermingham, E. & Carolina, P.S. (2007) The role of tropical dry forest as a long‐ term barrier to dispersal: a comparative phylogeographical analysis of dry forest tolerant and intolerant frogs. Molecular Ecology, 16: 4789–4807. Davis, S.D., Heywood, V.H., Herrera-MacBryde, O., Villa-Lobos, J. & Hamilton, A. (1997) Centres of Plant Diversity: a Guide and Strategy for Their Conservation, Volume 3: The Americas. IUCN Publications Unit, Cambridge, 562 pp. Drummond, A.J., Suchard, M.A., Xie, D. & Rambaut, A. (2012) Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution, 29: 1969–1973. Duellman, W.E., Marion, A.B. & Hedges, S.B. (2016) Phylogenetics, classification, and biogeography of the treefrogs (Amphibia: Anura: Arboranae). Zootaxa, 4104: 1–109. 63 Eaton, D.A. (2014) PyRAD: assembly of de novo RADseq loci for phylogenetic analyses. Bioinformatics, 30: 1844–1849. Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research, 32: 1792–1797. Fouquet, A., Gilles, A., Vences, M., Marty, C., Blanc, M. & Gemmell, N.J. (2007) Underestimation of species richness in Neotropical frogs revealed by mtDNA analyses. PLoS ONE, 2: e1109. Frost, D.R. 2019. Amphibian Species of the World: an Online Reference. Version 6.0. Electronic Database accessible at: http://research.amnh.org/herpetology/amphibia/index.html. American Museum of Natural History, New York, USA. Captured on 10 January 2019. Fujisawa, T. & Barraclough, T.G. (2013) Delimiting species using single-locus data and the Generalized Mixed Yule Coalescent approach: a revised method and evaluation on simulated data sets. Systematic Biology, 62: 707–724. García-París, M., Good, D.A., Parra-Olea, G. & Wake, D.B. (2000) Biodiversity of Costa Rican salamanders: implications of high levels of genetic differentiation and phylogeographic structure for species formation. Proceedings of the National Academy of Sciences, 97: 1640–1647. Garrigues, R. & Dean, R. (2014) The birds of Costa Rica: A field Guide. Cornell University Press, Nueva York, 425 pp. Hedges, S.B., Duellman, W.E. & Heinicke, M.P. (2008) New World direct-developing frogs (Anura: Terrarana): molecular phylogeny, classification, biogeography, and conservation. Zootaxa, 1737: 1–182. Ilut, D.C., Nydam, M.L. & Hare, M.P. (2014) Defining loci in restriction-based reduced representation genomic data from nonmodel species: sources of bias and diagnostics for optimal clustering. BioMed Research International, 2014. 64 Kapli, P., Lutteropp, S., Zhang, J., Kobert, K., Pavlidis, P., Stamatakis, A. & Flouri, T. (2016) Multi-rate Poisson Tree Processes for single-locus species delimitation under Maximum Likelihood and Markov Chain Monte Carlo. Bioinformatics, 063875. Kluge, J. & Kessler, M. (2006) Fern endemism and its correlates: contribution from an elevational transect in Costa Rica. Diversity and Distributions, 12: 535–545. Lanfear, R., Calcott, B., Ho, S.Y. & Guindon, S. (2012) PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Molecular Biology and Evolution, 29: 1695–1701. Leaché, A.D. & Fujita, M.K. (2010) Bayesian species delimitation in West African forest geckos (Hemidactylus fasciatus). Proceedings of the Royal Society of London B: Biological Sciences, 277: 3071–3077. Leaché, A.D., Banbury, B.L., Felsenstein, J., Nieto-Montes de Oca, A. & Stamatakis, A. (2015) Short tree, long tree, right tree, wrong tree: new acquisition bias corrections for inferring SNP phylogenies. Systematic Biology, 64: 1032–1047. Leenders T. (2016) Amphibians of Costa Rica: a field guide. Comstock Publishing Associates, Ithaca, New York, United States, 531 pp. Lynch, J.D. (1980) A new frog of the genus Eleutherodactylus from western Panama. Transactions of the Kansas Academy of Science, 83: 101–105. Matzke, N.J. (2014) Package ‘BioGeoBEARS’. BioGeography with Bayesian (and Likelihood) Evolutionary Analysis in R scripts. Version 0.2. 1. 360pp. Meyer, C.P. (2003) Molecular systematics of cowries (Gastropoda: Cypraeidae) and diversification patterns in the tropics. Biological Journal of the Linnean Society, 79: 401– 459. Miller M.A., Pfeiffer, W. & Schwartz, T. (2010) Creating the CIPRES Science Gateway for inference of large phylogenetic trees. Proceedings of the Gateway Computing Environments Workshop (GCE), 14 November 2010: 1–8. 65 Miyamoto, M.M. (1983) Biochemical variation in the frog Eleutherodactylus bransfordii: geographic patterns and cryptic species. Systematic Zoology, 32: 43‒51. Montes, C., Bayona, G., Cardona, A., Buchs, D.M., Silva, C A., Morón, S., Hoyos, N., Ramírez, D.A, Jaramillo, C.A. & Valencia, V. (2012)a Arc‐ continent collision and orocline formation: Closing of the Central American seaway. Journal of Geophysical Research: Solid Earth, 117: B04105. Montes, C., Cardona, A., McFadden, R., Morón, S.E., Silva, C.A., Restrepo-Moreno, S., Ramírez, D.A., Hoyos, N., Wilson, J., Farris, D., Bayona, G.A., Jaramillos, C.A., Valencia, V., Bryan. J. & Flores, J.A. (2012)b Evidence for middle Eocene and younger land emergence in central Panama: implications for Isthmus closure. Geological Society of America Bulletin, 124: 780–799. Montes, C., Cardona, A., Jaramillo, C., Pardo, A., Silva, J. C., Valencia, V., Ayala, C., Pérez- Angel, L.C., Rodríguez-Parra, L.A., Ramírez, V. & Niño, H. (2015) Middle miocene closure of the Central American seaway. Science, 348: 226–229. Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature, 403: 853–858. Nieto-Montes de Oca, A., Barley, A.J., Meza-Lázaro, R.N., García-Vázquez, U.O., Zamora- Abrego, J. G., Thomson, R.C. & Leaché, A.D. (2017) Phylogenomics and species delimitation in the knob-scaled lizards of the genus Xenosaurus (Squamata: Xenosauridae) using ddRADseq data reveal a substantial underestimation of diversity. Molecular Phylogenetics and Evolution, 106: 241–253. Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V., Underwood, E.C. & Loucks, C.J. (2001) Terrestrial Ecoregions of the World: A New Map of Life on Earth: A new global map of terrestrial ecoregions provides an innovative tool for conserving biodiversity. BioScience, 51: 933–938. 66 Palumbi, S., Martin, A., Romano, S., McMillan, W.O., Stice, L. & Grabowski, G. (1991) The simple fool’s guide to PCR, version 2.0. Special Publication, Department of Zoology and Kewalo Marine Laboratory, University of Hawaii, Honolulu, Hawaii, USA. Pérez-Ponce de León, G. & Poulin, R. (2016) Taxonomic distribution of cryptic diversity among metazoans: not so homogeneous after all. Biology Letters, 12: 20160371. Peterson, B.K., Weber, J.N., Kay, E.H., Fisher, H.S. & Hoekstra, H.E. (2012) Double digest RADseq: an inexpensive method for de novo SNP discovery and genotyping in model and non-model species. PLoS ONE, 7: e37135. Pinto-Sánchez, N.R., Ibáñez, R., Madriñán, S., Sanjur, O.I., Bermingham, E. & Crawford, A.J. (2012) The great American biotic interchange in frogs: multiple and early colonization of Central America by the South American genus Pristimantis (Anura: Craugastoridae). Molecular Phylogenetics and Evolution, 62: 954–972. Pons, J., Barraclough, T.G., Gomez-Zurita, J., Cardoso, A., Duran, D.P., Hazell, S., Kamoun, S., Sumlin, W.D. & Vogler, A.P. (2006) Sequence-based species delimitation for the DNA taxonomy of undescribed insects. Systematic Biology, 55: 595–609. Puillandre, N., Lambert, A., Brouillet, S. & Achaz, G. (2012) ABGD, Automatic Barcode Gap Discovery for primary species delimitation. Molecular Ecology, 21: 1864–1877. Rannala, B. & Yang, Z. (2013) Improved reversible jump algorithms for Bayesian species delimitation. Genetics, 194: 245–253. Rambaut, A., Suchard, M.A., Xie, W. & Drummond, A.J. (2014) Tracer v1.6.1. Available from http://beast.bio.ed.ac.uk/Tracer. Ree, R.H. & Smith, S.A. (2008) Lagrange: software for likelihood analysis of geographic range evolution. Systematic Biology, 57: 4–14. Ronquist, F., Teslenko, M., Van Der Mark, P., Ayres, D. L., Darling, A., Höhna, S., Larget, B., Liu, L., Suchard, M. & Huelsenbeck, J.P. (2012) MrBayes 3.2: efficient Bayesian 67 phylogenetic inference and model choice across a large model space. Systematic Biology, 61: 539–542. Sambrook, J. & Russell, D.W. (2006) Purification of nucleic acids by extraction with phenol: chloroform. Cold Spring Harbor Protocols, 2006: pdb-prot4455. Savage, J.M. (1966) The origins and history of the Central American herpetofauna. Copeia, 1966: 719–766. Savage, J.M. (1970) On The trail of the Golden Frog: with W. Arszewicz and Gabb in Central America. Proceedings of the California Academy of Sciences: Festschrift for George Sprague Myers, 38: 273–288. Savage, J.M. (2002) The Amphibians and Reptiles of Costa Rica: A Herpetofauna between two Continents, between two Seas. The University of Chicago Press, Illinois, 934 pp. Savage, J.M. & Emerson, S.B. (1970) Central American frogs allied to Eleutherodactylus bransfordii (Cope): a problem of polymorphism. Copeia, 1970: 623–644. Smith, M.A., Poyarkov Jr., N.A. & Hebert, P.D.N. (2008) CO1 DNA barcoding amphibians: Take the chance, meet the challenge. Molecular Ecology Resources, 8: 235–246. Stamatakis, A. (2014) RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics, 30: 1312–1313. Streicher, J.W., Crawford, A.J. & Edwards, C.W. (2009) Multilocus molecular phylogenetic analysis of the montane Craugastor podiciferus species complex (Anura: Craugastoridae) in Isthmian Central America. Molecular Phylogenetics and Evolution, 53: 620–630. Tamura, K. (1992). Estimation of the number of nucleotide substitutions when there are strong transition–transversion and G + C–content biases. Molecular Biology and Evolution, 9: 678–687. Tamura, K., Stecher, G., Peterson, D., Filipski, A. & Kumar, S. (2013) MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Molecular Biology and Evolution, 30: 2725– 2729. 68 Taylor, E.H (1952) The salamanders and caecilians of Costa Rica. University of Kansas Science Bulletin, 34: 695–791. Trueb, L. (1968) Variation in the tree frog Hyla lancasteri. Copeia, 1968: 285–299. Vences, M., Thomas, M., Bonett, R.M. & Vieites, D.R. (2005) Deciphering amphibian diversity through DNA barcoding: chances and challenges. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 360: 1859–1868. Wang, I.J., Crawford, A.J. & Bermingham, E. (2008) Phylogeography of the pygmy rain frog (Pristimantis ridens) across the lowland wet forest of isthmian Central America. Molecular Phylogenetics and Evolution, 47: 992–1004. Webb, S.D. (1991) Ecogeography and the Great American Interchange. Paleobiology, 17: 266– 280. Webb, S.D. (2006) The Great American Biotic Interchange: patterns and processes. Annals of the Missouri Botanical Garden, 93: 245–257. Wiens, J.J., Parra-Olea, G., García-París, M. & Wake, D.B. (2007) Phylogenetic history underlies elevational biodiversity patterns in tropical salamanders. Proceedings of the Royal Society Biological Sciences Series B, 274: 919–928. Yang, Z. (2015) The BPP program for species tree estimation and species delimitation. Current Zoology, 61: 854–865. Yang, Z. & Rannala, B. (2010) Bayesian species delimitation using multilocus sequence data. Proceedings of National Academy of Sciences USA, 107: 9264–9269. Yang, Z. & Rannala, B. (2014) Unguided species delimitation using DNA sequence data from multiple loci. Molecular Biology and Evolution, 31: 3125–3135. Zhang, J., Kapli, P., Pavlidis, P. & Stamatakis, A. (2013) A general species delimitation method with applications to phylogenetic placements. Bioinformatics, 29: 2869–2876. 69 Appendix 1. Institutional voucher numbers and locality information for the specimens used in the molecular phylogenetic analyses. Museum collection acronyms follow Frost (2019), with the addition of AJC refers to Andrew J. Crawford field numbers, EAP refers to Erick Arias field numbers, and CRARC refers to Costa Rica Amphibian Research Center private collection. Species Institutional vouchers Collection locality Elevation (m) Geographic coordinates Lat Lon Craugastor aenigmaticus clade Craugastor aenigmaticus SMF: 104020 Changuinola, Bocas del Toro, PA 2388 8.9139 -82.7088 Craugastor aenigmaticus UCR: 21951 Telire, Talamanca, CR 2700 9.3488 -83.1750 Craugastor aenigmaticus UCR: 22737 Buenos Aires, Puntarenas, CR 2660 9.3224 -83.2028 Craugastor podiciferus clade Craugastor sagui SMF: 104014 Nole Duima, Ngöbe Buglé, PA 1762 8.5571 -81.8245 Craugastor sagui SMF: 104015 Nole Duima, Ngöbe Buglé, PA 1700 8.4997 -81.7724 Craugastor zunigai UCR: 20389 Buenos Aires, Puntarenas, CR 1500 9.1112 -83.1006 Craugastor zunigai UCR: 22709 Coto Brus, Puntarenas, CR 1980 8.9751 -82.8243 Craugastor blairi FMNH: 257562 Gualaca, Chiriquí, PA 1000 8.7500 -82.217 Craugastor blairi SMF: 104023 Gualaca, Chiriquí, PA 1280 8.6781 -82.2101 Craugastor blairi SMF: 102024 Gualaca, Chiriquí, PA 1730 8.6775 -82.1980 Craugastor blairi SMF: 104027 Bugaba, Chiriquí, CR 2134 8.8494 -82.5154 Craugastor podiciferus CRARC: 0012 Turrialba, Cartago, CR 2250 10.0192 -83.7132 Craugastor podiciferus EAP: 0536 El Guarco, Cartago, CR 2395 9.7126 -83.9488 Craugastor podiciferus EAP: 0810 Talamanca, Limón, CR 1860 9.3659 -83.0417 Craugastor podiciferus SMF: 104005 Changuinola, Bocas del Toro, PA 1766 8.9908 -82.6716 Craugastor podiciferus UCR: 19853 Telire, Talamanca, CR 1817 9.3580 -83.2294 Craugastor podiciferus UCR: 19856 Telire, Talamanca, CR 1817 9.3580 -83.2294 Craugastor podiciferus UCR: 19860 Telire, Talamanca, CR 2108 9.3645 -83.2164 Craugastor podiciferus UCR: 19862 Telire, Talamanca, CR 2108 9.3645 -83.2164 Craugastor podiciferus UCR: 20992 Alfaro Ruiz, Alajuela, CR 2143 10.2272 -84.3482 Craugastor podiciferus UCR: 22146 Vázquez de Coronado, San José, CR 1700 10.0263 -83.9448 Craugastor podiciferus UCR: 22201 Dota, San José, CR 2395 9.7126 -83.9488 Craugastor sp. Chumacera UCR: 22120 Buenos Aires, Puntarenas, CR 1821 9.3218 -83.4546 Craugastor sp. Chumacera UCR: 22690 Pérez Zeledón, San José, CR 1793 9.3267 -83.4706 Craugastor sp. Fila Costeña EAP: 0509 Golfito, Puntarenas, CR 1546 8.7878 -83.0306 Craugastor sp. Fila Costeña EAP: 0519 Pérez Zeledón, San José, CR 1350 9.4415 -83.6848 Craugastor sp. Fila Costeña FMNH: 257651 Coto Brus, Puntarenas, CR 1350 8.7833 -82.9833 Craugastor sp. Fila Costeña UCR: 16585 Dota, San José, CR 1400 9.5353 -83.8580 Craugastor sp. Fila Costeña UCR: 22091 Pérez Zeledón, San José, CR 1488 9.4410 -83.6830 Craugastor sp. Monte Verde FMNH: 257669 Monte Verde, Puntarenas, CR 1500 10.2773 -84.5891 Craugastor sp. Monte Verde FMNH: 257673 Monte Verde, Puntarenas, CR 1500 10.2773 -84.5891 Craugastor sp. Monte Verde UCR: 16361 Alfaro Ruiz, Alajuela, CR 1930 10.2176 -84.3671 Craugastor sp. Monte Verde UCR: 22675 Puntarenas, Puntarenas, CR 1726 10.3202 -84.7987 Craugastor sp. San Gerardo CRARC: 0247 Tilarán, Guanacaste, CR 1470 10.3600 -84.8000 Craugastor sp. San Gerardo FMNH257671 Monte Verde, Puntarenas, CR 1500 10.2773 -84.5891 Craugastor sp. San Gerardo UCR: 16353 Sarapiquí, Heredia, CR 1500 10.2022 -84.1625 Craugastor sp. Pico Blanco UCR: 22226 Escazú, San José, CR 2242 9.8646 -84.1429 Craugastor sp. Pico Blanco UCR: 22228 Escazú, San José, CR 2242 9.8646 -84.1429 Craugastor sp. Siola EAP: 0817 Talamanca, Limón, CR 1300 9.3987 -83.0200 Craugastor stejnegerianus Craugastor gabbi EAP: 0626 Coto Brus, Puntarenas, CR 1541 8.9515 -82.8346 Craugastor gabbi UCR: 21863 Coto Brus, Puntarenas, CR 1200 8.7889 -82.9583 Craugastor gabbi UCR: 21864 Coto Brus, Puntarenas, CR 1200 8.7889 -82.9583 70 Appendix 1. Continued. Species Institutional vouchers Collection locality Elevation (m) Geographic coordinates Lat Lon Craugastor persimilis EAP: 0586 Talamanca. Limón, CR 121 9.5773 -82.9343 Craugastor persimilis FMNH: 257567 Turrialba, Cartago, CR 550 9.8917 -83.6500 Craugastor persimilis FMNH: 257571 Turrialba, Cartago, CR 550 9.8917 -83.6500 Craugastor persimilis UCR: 22211 Paraíso, Cartago, CR 1050 9.7841 -83.7517 Craugastor persimilis UCR: 22212 Paraíso, Cartago, CR 1050 9.7841 -83.7517 Craugastor rearki EAP: 0554 Upala, Guanacaste, CR 764 10.7109 -85.0406 Craugastor rearki EAP: 0555 Upala, Guanacaste, CR 764 10.7109 -85.0406 Craugastor rearki EAP: 0572 Siquírres, Limón, CR 537 10.0595 -83.5452 Craugastor rearki MVZ: 263735 Altagracia, Altagracia, NI 466 11.4687 -85.5069 Craugastor rearki SMF: 79759 Matagalpa, Matagalpa, NI 1300 12.9993 -85.9092 Craugastor rearki UCR: 20600 Los Chiles, Alajuela, CR 45 11.0513 -84.7393 Craugastor rearki UCR: 16343 Bagaces, Guanacaste, CR 640 10.7072 -85.0844 Craugastor rearki UCR: 22240 Pococí, Limón, CR 228 10.2257 -83.7712 Craugastor rearki UCR: 21149 Limón, Limón, CR 400 9.9260 -83.1880 Craugastor rearki UCR: 21152 Limón, Limón, CR 400 9.9260 -83.1880 Craugastor rearki USNM: 559393 Puerto Lempira, Gracias a Dios, HN 190 14.9275 -84.5339 Craugastor stejnegerianus EAP: 0508 Golfito, Puntarenas, CR 28 8.6906 -83.4815 Craugastor stejnegerianus EAP: 0512 Buenos Aires, Puntarenas, CR 812 9.0905 -83.1247 Craugastor stejnegerianus EAP: 0514 Osa, Puntarenas, CR 45 8.9655 -83.4411 Craugastor stejnegerianus UCR: 20346 Buenos Aires, Puntarenas, CR 900 9.0860 -83.1110 Craugastor stejnegerianus UCR: 20352 Buenos Aires, Puntarenas, CR 900 9.0863 -83.1105 Craugastor stejnegerianus UCR: 21494 Golfito, Puntarenas, CR 100 8.4046 -83.1197 Craugastor stejnegerianus UCR: 22070 Buenos Aires, Puntarenas, CR 416 9.1520 -83.4256 Craugastor stejnegerianus UCR: 22101 Pérez Zeledón, San José, CR 740 9.3009 -83.7714 Craugastor stejnegerianus UCR: 22280 Osa, Puntarenas, CR 30 9.1966 -83.7870 Craugastor sp. Neilly EAP: 0506 Golfito, Puntarenas, CR 182 8.6985 -82.0489 Craugastor sp. Quepos CRARC: 0148 San Ramón, Alajuela, CR 1140 10.2021 -84.4843 Craugastor sp. Quepos CRARC: 0266 Montes de Oro, Puntarenas, CR 1325 10.1800 -84.6700 Craugastor sp. Quepos EAP: 0524 Aguirre, Puntarenas, CR 12 9.3245 -83.9498 Craugastor sp. Quepos EAP: 0527 Aguirre, Puntarenas, CR 180 9.4776 -84.0473 Craugastor sp. Quepos UCR: 20907 Aguirre, Puntarenas, CR 200 9.4619 -84.0631 Craugastor sp. Quepos UCR: 21007 Aguirre, Puntarenas, CR 200 9.4620 -84.0630 Craugastor sp. Quepos UCR: 22208 Montes de Oca, San José, CR 1210 9.9374 -84.0495 Craugastor bransfordii clade Craugastor bransfordii CRARC: 0176 Siquírres, Limón, CR 755 10.0600 -83.6300 Craugastor bransfordii EAP: 0558 Siquírres, Limón, CR 537 10.0595 -83.5452 Craugastor bransfordii UCR: 20559 Los Chiles, Alajuela, CR 45 11.0513 -84.7393 Craugastor bransfordii UCR: 20951 San Ramón, Alajuela, CR 1095 10.1862 -84.5075 Craugastor bransfordii UCR: 22269 Alajuela, Alajuela, CR 466 10.3121 -84.1778 Craugastor underwoodi EAP: 0534 Paraíso, Cartago, CR 1412 9.7518 -83.7792 Craugastor underwoodi EAP: 0540 Vázquez de Coronado, San José, CR 1708 10.0254 -83.9456 Craugastor underwoodi EAP: 0593 Puntarenas, Puntarenas, CR 1566 10.3160 -84.8061 Craugastor underwoodi CRARC: 0245 Tilarán, Guanacaste, CR 1470 10.3600 -84.8000 Craugastor sp. Fila Carbon EAP: 0577 Talamanca, Limón, CR 94 9.7116 -82.8344 Craugastor sp. Fila Carbon EAP: 0583 Talamanca, Limón, CR 198 9.6064 -82.9115 Craugastor sp. Fila Carbon EAP: 0585 Talamanca, Limón, CR 198 9.6064 -82.9115 Craugastor sp. Fila Carbon EAP: 0797 Talamanca, Limón, CR 1500 9.3773 -83.0371 Craugastor sp. Fila Carbon UCR: 20050 Talamanca, Limón, CR 900 9.6178 -83.2681 Craugastor sp. Fila Carbon UCR: 20052 Talamanca, Limón, CR 900 9.6178 -83.2681 Craugastor sp. Fila Carbon UCR: 20150 Limón, Limón, CR 500 9.8550 -83.1517 Craugastor sp. Fila Carbon UCR: 21122 Limón, Limón, CR 400 9.9260 -83.1880 71 Appendix 1. Continued. Species Institutional vouchers Collection locality Elevation (m) Geographic coordinates Lat Lon Craugastor sp. Fila Carbon UCR: 22533 Limón, Limón, CR 1123 9.8679 -83.2406 Craugastor sp. Quebradas UCR: 16326 Pérez Zeledón, San José, CR 1313 9.4375 -83.6869 Craugastor sp. Panama AJC: 1921 Panamá, Panamá, PA 810 9.3198 -79.2889 Craugastor sp. Panama AJC: 1960 Panamá, Panamá, PA 624 9.2908 -79.3027 Craugastor sp. Panama CH: 6808 Panamá, Panamá, PA 871 9.2744 -79.3178 Craugastor sp. Panama FMNH: 257677 Gualaca, Chiriquí, PA 1000 8.7500 -82.2167 Craugastor sp. Panama FMNH: 257698 Kuna Yala, Kuna Yala, PA 420 9.3167 -78.9833 Craugastor sp. Panama MVUP: 1803 La Pintada, Coclé, PA 800 8.6670 -80.5920 Craugastor sp. Panama MVUP: 1841 La Pintada, Coclé, PA 800 8.6670 -80.5920 Craugastor sp. Panama USNM: 572220 La Pintada, Coclé, PA 800 8.6670 -80.5920 Craugastor sp. Panama USNM: 572221 La Pintada, Coclé, PA 800 8.6670 -80.5920 Craugastor sp. Panama USNM: 572222 La Pintada, Coclé, PA 800 8.6670 -80.5920 Craugastor sp. Panama USNM: 572223 La Pintada, Coclé, PA 800 8.6670 -80.5920 Craugastor sp. Panama SMF: 104010 Changuinola, Bocas del Toro, PA 1258 9.0090 -82.6644 Craugastor sp. Vereh CRARC: 0047 Turrialba, Cartago, CR 1350 9.8681 -83.4621 Craugastor sp. Vereh CRARC: 0048 Turrialba, Cartago, CR 1350 9.8681 -83.4621 Craugastor sp. Vereh CRARC: 0228 Turrialba, Cartago, CR 1650 9.7500 -83.5500 Craugastor sp. Vereh EAP: 0754 Turrialba, Cartago, CR 1650 9.7500 -83.5500 72 CAPÍTULO II REVISIÓN TAXONÓMICA DEL GRUPO DE ESPECIES CRAUGASTOR PODICIFERUS 73 CAPÍTULO II.I TAXONOMIC ASSESSMENT OF CRAUGASTOR PODICIFERUS (ANURA: CRAUGASTORIDAE) IN LOWER CENTRAL AMERICA WITH THE DESCRIPTION OF TWO NEW SPECIES Autores: Erick Arias, Andreas Hertz y Gabriela Parra-Olea En revisión: Amphibian & Reptile Conservation Fecha de envío: Enero de 2019 74 Taxonomic assessment of Craugastor podiciferus (Anura: Craugastoridae) in lower Central America with the description of two new species 1,2,3Erick Arias, 4,5Andreas Hertz, and 1Gabriela Parra-Olea 1 Departamento de Zoología, Instituto de Biología, UNAM, AP 70-153 Ciudad Universitaria, CP 04510, Ciudad de México, México. E-mail: eapiedra@gmail.com 2 Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México, Av. Ciudad Universitaria 3000, C.P. 04360, Coyoacán, Ciudad de México, México. 3 Escuela de Biología, Universidad de Costa Rica, San Pedro, 11501-2060 San José, Costa Rica. 4 Department of Biology, University of Massachusetts Boston, 100 Morrissey Blvd., Boston, MA 02125, USA. E-mail: andreas.hertz@umb.edu 5 Senckenberg Forschungsinstitut und Naturmuseum, Senckenberganlage 25, 60325, Frankfurt am Main, Germany. * Corresponding author: gparra@ib.unam.mx 75 Abstract.—The systematics and taxonomy of the polytypic species Craugastor podiciferus are poorly understood due to the high level of phenotypic polymorphism between and within species and the lack of molecular data from topotypic specimens. We performed a well-sampled study including all known species of the C. podiciferus species group, several localities from highlands in Costa Rica and western Panama, and for the first time, samples from the type locality of C. podiciferus. We performed a phylogenetic analysis based on the DNA sequences of the mitochondrial 16S rRNA (16S) and cytochrome oxidase 1 (COI) genes and a morphometric analysis. Based on our results, we restrict C. podiciferus to the populations from the Cordillera Volcánica Central and Cordillera de Talamanca in Costa Rica and western Panama. Craugastor podiciferus sensu stricto and six additional clades from the highlands of Costa Rica constitute the well-supported C. podiciferus sensu lato clade. Our analyses support the existence of three additional species from the Pacific slope of southwestern Costa Rica and western Panama. Herein, we describe two lineages as new species and provide revised descriptions for C. podiciferus and C. blairi. The name C. blairi is resurrected and used for populations from the Cordillera de Talamanca and Cordillera Central in western Panama. Two additional species are named, one from Cordillera Central in western Panama. This species is easily differentiated by the presence of nuptial pads in adult males, a smooth venter, and flat subarticular tubercles. The second species is named for populations from southwestern Costa Rica. This species is recognized by its coarsely areolate venter, projecting subarticular tubercles, and heel without a projecting tubercle. The recognition of these three species from the lower montane rainforest highlights the role of the highlands on the pacific slope of Costa Rica and Panama in the diversification of the C. podiciferus species group. Keywords. Brachycephaloidea, Central America, cryptic species, DNA barcoding, Talamanca, Terrarana. 76 Introduction The direct-developing frogs of the Craugastor podiciferus species group (Hedges et al. 2008) are found from eastern Honduras to Central Panama (AmphibiaWeb 2019; Savage 2002) with most of the species (nine out of ten) restricted to Isthmian Central America. The distribution of this group ranges from sea level to 2700 m a.s.l. in a wide variety of habitats, from tropical rain forest, cloud forest, to montane forest (Savage 2002). The morphological delimitation between members of the C. podiciferus species group is difficult due to the extremely conserved morphological characters and the high level of phenotypic polymorphism within species and populations. The systematics and taxonomy of the C. podiciferus species group has been poorly studied; however, previous molecular studies have suggested the existence of several unnamed species that are masked under the current names (Savage 2002; Crawford 2003; Crawford and Smith 2005; Streicher et al. 2009). For example, Crawford and Smith (2005) used 10 samples (seven species) of the C. podiciferus species group and found the presence of two unnamed species (Craugastor sp. B and Craugastor sp. C); they also found C. stejnegerianus (Cope, 1893) to be paraphyletic. Recently, Arias et al. (2016) showed that populations formerly considered to be part of the C. stejnegerianus from southwestern Costa Rica and western Panama belonged to a different species and described them as C. gabbi Arias, Chaves, Crawford, and Parra-Olea, 2016. Craugastor podiciferus (Cope, 1875) is the most complicated taxon within the C. podiciferus species group. Its morphological polymorphism has been recognized since its description by Cope (1875), who described it based on four varieties plus two additional species (C. muricinus and C. habenatus) in the same paper, with specimens from the same locality. In addition, Cope (1875) stated that “the colors of this species vary remarkably, more than I have observed to be the case in any other frog”. Subsequently, Barbour (1928) described C. blairi using specimens from western Panama, which Taylor (1952) later synonymized together with C. 77 muricinus and C. habenatus under C. podiciferus. An additional species, C. jota (Lynch, 1980), was named using specimens from western Panama, and it was suggested to be related to C. podiciferus. More recently, the name C. podiciferus has been used for specimens from several highlands populations (1089–2650 m a.s.l.) on both slopes of the mountain ranges in Costa Rica and western Panama (Savage 2002). Streicher et al. (2009) used specimens from several populations referred to as C. podiciferus to perform a well-sampled phylogenetic study of this complex. They found that C. podiciferus is represented by six clades, each likely representing distinct species. In addition, the populations from western Panama were not grouped with the main C. podiciferus clade, which they called Craugastor sp. B. Although the molecular and geographical evidence shown by Streicher et al. (2009) supported the presence of several species under the name C. podiciferus, taxonomic changes have not yet been implemented, mainly due to the lack of topotipic specimens of C. podiciferus. The correct type locality has been the subject of discussion by Savage (1970), and later, Arias and Chaves (2014) concluded that it is actually on the Caribbean slopes of Cerro Kamuk. Here, we included specimens of Craugastor podiciferus from the type locality for the first time together with several additional localities from Costa Rica and western Panama. We used mitochondrial sequences to address the following goals: 1) to identify C. podiciferus sensu stricto in a phylogenetic context, 2) to evaluate the phylogenetic relationships of the C. podiciferus species complex, and 3) to evaluate the taxonomic status of the populations from the Pacific slopes of southwestern Costa Rica and western Panama. The result is a comprehensive revision of the C. podiciferus species complex, in which we describe two new species from southwestern Costa Rica and western Panama. Additionally, we resurrect an old name for a third species in mountainous western Panama. 78 Materials and Methods Species criterion: We follow the general metapopulation lineage species concept (Simpson 1951; Wiley 1978; de Queiroz 2007). Since we adhere to this concept, we recognize a species when there is evidence of the separation of metapopulation lineages preferably based on multiple lines of evidence following a consensus protocol for integrative taxonomy (Dayrat et al. 2005; Padial et al. 2010). Taxon sampling: The frogs were collected in the field, euthanized, fixed in a 10 % formalin solution and processed in 70 % ethanol for long-term storage. A tissue sample was preserved in 96 % ethanol or RNAlater and used for genetic analysis. Museum collection acronyms follow Frost (2019) with the addition of AH referring to Andreas Hertz field numbers. EAP refers to Erick Arias field numbers, and CRARC refers to the Costa Rica Amphibian Research Center private collection. Amplification and sequencing: We determined partial sequences of the large subunit ribosomal RNA (16S) mitochondrial gene for six specimens of Craugastor sp. 1 (Arias et al. 2018), Craugastor sp. 2 (Arias et al. 2018), and Craugastor sp. B (Crawford and Smith 2005) from the pacific slopes of southwestern Costa Rica and western Panama (Fig. 1). We compared the sequences obtained herein with those available in GenBank for the C. podiciferus species group. The protocols for DNA extraction, amplification, sequencing, and editing follow those of Arias et al. (2018). The list of vouchers and GenBank accession numbers used in this study are provided in Appendix I. 79 Fig. 1. Map showing the known populations of Craugastor blairi, C. sagui sp. nov., and C. zunigai sp. nov. from the lower montane rainforest in Southwestern Costa Rica and western Panama, and populations of other species of the C. podiciferus species group inhabiting the highlands of Costa Rica and western Panama. The arrow indicates the type locality of C. podiciferus. Phylogenetic analyses: Sequence alignments were performed using the MAFFT software (Katoh et al. 2017) under the “auto” strategy, default parameters and trimmed to the point where a majority of the taxa had sequence data. We partitioned the sequence data by gene and further partitioned the COI by codon position. We used PartitionFinder v1.1.1 (Lanfear et al. 2012) and the Bayesian information criterion (BIC) to select the best partition scheme and the best model of sequence evolution for each partition. We used a single set of branch-lengths across all partitions 80 (branchlengths=linked), and the search of the best partition scheme was implemented using a heuristic search (scheme=greedy). We defined, a priori, four partitions, one for 16S and three for COI (one for each codon). Phylogenetic analyses were performed using both the maximum likelihood (ML) and Bayesian inference (BI) methods. We performed the maximum likelihood analysis using Garli 2.01 (Zwickl 2006), with 10 search replicates with the following default setting values: streefname=random, attachmentspertaxon=24, genthreshfortopoterm=100000, significanttopochange=0.00001. For bootstrapping, we ran 1000 replicates with the previous settings with the following changes: genthreshfortopoterm=10000, significanttopochange=0.01, treerejectionthreshold=20, as suggested in the Garli manual, to accelerate the bootstrapping. The bootstrap consensus tree was performed using Sumtrees (Sukumaran and Holder 2010b) from the DendroPy package version 4.4.0 (Sukumaran and Holder 2010a). Bayesian phylogenetic analysis was performed using MrBayes 3.2.6 (Ronquist et al. 2012) with the partition scheme and the model of sequence evolution for each partition as selected previously. Two separate analyses were run, each consisting of 20 million generations, sampled every 1000 generations, and four chains with default heating parameters. We examined a time-series plot of the likelihood scores of the cold chain to check the stationarity using Tracer 1.6 software (Rambaut et al. 2014). We discarded the first 25 % of trees as burn-in and used the remaining trees to estimate the consensus tree along with the posterior probabilities for each node and each parameter. Maximum likelihood and Bayesian analyses were run on the CIPRES portal (Miller et al. 2010). Genetic distances (uncorrected p-distances) were computed using MEGA6 (Tamura et al. 2013). Morphometric analyses: We performed a morphometric analysis comparing the three populations from the highlands of Southwestern Costa Rica and western Panama. We examined 19 specimens of Craugastor sp. 1, 7 specimens of Craugastor sp. 2, and 25 specimens of 81 Craugastor sp. B (Appendix II). The specimens were deposited at the Museo de Zoología (UCR), San José, Costa Rica and the Senckenberg Research Institute and Nature Museum, Frankfurt, Germany (SMF). The following morphological measurements were recorded as described by Savage (2002), Duellman and Lehr (2009), and Arias et al. (2016): snout-vent length (SVL), head length (HL), head width (HW), inter orbital distance (IOD), width of the upper eyelid (EW), eye-nostril distance (EN), eye diameter (ED), and tympanum diameter (TY). Measurements were performed using dial calipers and were rounded to the nearest 0.1 mm. To avoid allometric effects relative to differences in size and shape between species and between individuals, we transformed the data using the method of Lleonart et al. (2000). Additional proportions reported herein include the following: EW/IOD, IOD/HW, TY/ED, EN/ED, ED/HL, HL/HW, and EN/HL. The sex of the individuals was determined by the gonadal morphology; the specimens with opaque seminal vesicles were assumed to be adult males, and those with developed oviducts were assumed to be adult females. The general terminology for the morphological characteristics follows Duellman and Lehr (2009). We followed Savage (2002) for the term “supernumerary tubercles”, which we use to refer to the tubercles on the phalanges (between subarticular tubercles); this is different from the tubercles referred to herein as accessory palmar or plantar tubercles. We calculated the mean, standard deviation, and range for each morphometric variable without correction. We used all variables to perform a linear discriminant analysis to determine whether the morphometric variables were effective to predict the species. We validated the proportion of correctly classified individuals using jackknife accuracy (Manly 1994). All analyses were performed using R v3.3.3 (R Core Team 2017). 82 Results Molecular: The resulting mitochondrial data matrix included 59 sequences with a total sequence length of 1222 bp including gaps; 565 bp for 16S and 657 bp for COI. The best strategy partition contains four partitions, one for 16S and one for each codon in COI. The following substitution models were selected: GTR+G for 16S, K80+I+G for COI codon position 1, HKY+I+G for COI codon position 2, and GTR+I+G for COI codon position 3. The mitochondrial genetic distances are shown in Table 1. Genetic distances between Craugastor sp. 1 and all other members of the Craugastor podiciferus species group are 4.9–15.2 % for 16S and 13.5–21.6 % for COI. Craugastor sp. 2 is separated by an uncorrected genetic distance to other members of the Craugastor podiciferus species group of 2.9–15.7 % for 16S and 14.3–21.3 % for COI. Genetic distances between Craugastor sp. B and other members of the Craugastor podiciferus species group are 2.9–16.2 % for 16S and 14.3–19.6 % for COI. The ML and Bayesian trees were similar in topology and show six well-supported clades (Fig. 2). The first is formed by specimens from La Nevera and Cerro Saguí from western Panama. A second and third clade are formed by specimens from southwestern Costa Rica and western Panama, respectively. The fourth clade comprises C. aenigmaticus from the montane rainforest of the Cordillera de Talamanca. A fifth major clade includes seven species of the C. podiciferus species group that mainly occur from the lowlands to mid-elevations from eastern Honduras to central Panama. Finally, a sixth major clade contains specimens from the type locality of C. podiciferus and several other localities from the highlands of Costa Rica and western Panama tentatively referred to this species. The main difference between the ML and the Bayesian topology is the position of the clade that contains the samples from La Nevera and Cerro Saguí. In the ML tree, this group was the sister to all other members of the C. podiciferus species group (although with very low support), while in the Bayesian tree (not shown), this was the sister clade to a clade containing the samples from Las Alturas and Potrero Grande and C. blairi. 83 84 ,--------- CrallgaslOJ" f.!.ollmerii AJCl926 ... SMF104015L.aNevera I 76/9 • 19/- SMFlO4017 La Ne\lera eral/gastor sagui sp. nov. SMF10401 4 Ccrro Sagui UCR22709 Las Alturas UCR22703 Las Alturas e " " UCR20429 Potrero Grande ratlgastor :Untgw sp. nov. UCR20389 Potrero Grande SM Fl04034 Guayabito SM F104027 Volcán Barú SMF I04037 La Estrella USNM563039 Palo Alto .. FMN H257689 La Fortuna Craugaslor blGfn SM F1 04023 Fortuna SM F1 04033 Fortuna SMF102024 Fortuna '" SMFI04020 Jurulungo I ,---"-'1 UCR22737 Cerro Hakú C. aenigmalicus UCR21 951 Cerro Utyum ,-..:'q UCR22643 Guaya.cán le. bransl'ordii UCR22269 San MIguel 'J l '" '" "'/ UCR226688ribri .Ie polyptychus UC R20050 Dabagn • 61 /78 64/8 "'/.* UCR22625 Casca~a' l c. underwoodi UCR22619 Tapantl */ VeR222ll Tausito le persimilis UCR22671 Suretka • ... UCR21864 San vi,ole gabbi UC R21863 San Vilo • 39/6 UCR20352 Potrero Gralldele s feinegerianus EAP0514 PalmarNone . '.J */ SMF79759 Selva Negra le law'as fer USNM559393 Tapalwas . FMN H257673 Monte Verde 501i)( UCRl6361 Tapesco e d" ifi " M "d " UCR22675 Mome Verde . pO I CI erus onte ver e F MNH257669 Mome Verde e d" ifi I UCR22146 Cascajal . po ICI erus s . . , n,. CRARCOl2 Volcán Turrialba " is3 I ~ ;~ ' ~ " :¡' UCR20992 Nectandra 5'1/79 26/7 27151 UCR22201 El Empalme SM F I 04005 Changena UCR23175 Alto Uren . , . UCRI9862 Lorí UCR I9856 Lori UCR 19860 Lori UCRI9853 Lorí e. podiciferus sensu stric fO Ir-- UCR231~9 Alto Urenle. podiciferus "Siola" UCR22228 p~co Blancolc. podiclifierus " Pico Blanco" UCR22226 PICO Blanco 61 CRARC0247 Monte Verde I % FMNH257671 Mame Verde e. podiciferus "San Gerardo" 3019 UC RI 6353 MontaftaAzul UC R22690 C h um~era le podiclifierus "Chumacera" UCR22120 AhamlT3 . 74' · 15/48- UCR 16585 Copey 811 ' EAP0509FilaCostena e d" 'ifi " F'¡ e t - " FMN H25765 1 FilaCostefta • po l el erus 1 a OS ena UCR22091 Quebradas Fig. 2. Maximum likelihood phylogram of the Craugastor podiciferus species group based on the 16S and COI mitochondrial DNA gene markers. Bootstrap proportions and posterior probability (multiplied by 100) values were obtained with MrBayes before and after the slash, respectively. The scale bar refers to the estimated substitutions per site. Asterisks represent support > 95. The blue clade corresponds to C. podiciferus sensu stricto. Morphometry: Morphometric variation and comparisons among the species are shown in Table 2. The proportion of specimens that could be correctly assigned to the species was 82 %, showing a clear morphological separation between the specimens of the three populations of southwestern Costa Rica and western Panama (Fig. 3). Fig. 3. Linear discriminant analysis showing the morphological separation among the three species from the lower montane rainforest in Southwestern Costa Rica and Western Panama. 85 T ab le 1 . M ea n un co rr ec te d ge ne ti c di st an ce s am on g lin ea ge s of t he C ra ug as to r po di ci fe ru s sp ec ie s gr ou p us in g th e C O I (a bo ve ) an d 16 S ( be lo w ) m ito ch on dr ia l g en es . 16 S/ C O I 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1 C . s ag ui s p. n ov . –– 14 .2 5 16 .5 1 18 .0 2 16 .2 8 17 .6 0 17 .1 3 18 .8 5 21 .3 4 21 .5 0 19 .6 3 14 .4 3 15 .0 9 16 .6 7 15 .5 8 15 .5 0 15 .8 9 15 .1 9 2 C . z un ig ai s p. n ov . 6. 05 –– 14 .6 2 19 .3 7 18 .3 0 18 .5 4 18 .2 2 20 .7 9 21 .3 4 19 .9 4 19 .2 4 15 .7 1 15 .4 4 18 .6 1 16 .5 9 17 .4 5 17 .2 1 18 .6 1 3 C . b la ir i 6. 54 3. 19 –– 16 .3 4 18 .4 1 18 .6 9 16 .8 5 19 .2 9 19 .4 2 18 .3 5 19 .1 1 16 .1 0 16 .0 4 16 .5 1 14 .4 9 16 .8 5 15 .2 1 16 .7 4 4 C . a en ig m at ic us 13 .7 3 10 .9 5 12 .1 3 –– 19 .4 7 18 .3 8 19 .2 1 19 .2 1 20 .2 5 18 .3 0 18 .6 4 16 .4 1 18 .3 1 19 .4 2 17 .3 4 18 .3 0 16 .5 1 17 .5 8 5 C . b ra ns fo rd ii 12 .5 0 12 .0 1 11 .2 7 13 .9 7 –– 13 .0 8 11 .2 1 17 .2 5 17 .8 3 18 .3 0 18 .4 6 17 .5 0 17 .7 5 19 .7 8 17 .1 3 18 .2 2 17 .6 8 18 .1 1 6 C . p ol yp ty ch us 8. 91 10 .0 5 9. 80 13 .7 3 8. 33 –– 8. 72 16 .4 3 15 .8 9 16 .6 7 15 .8 9 18 .5 1 19 .5 0 20 .0 2 19 .3 9 20 .9 5 19 .0 0 18 .9 3 7 C . u nd er w oo di 8. 99 9. 31 9. 07 13 .5 6 7. 11 3. 19 –– 15 .5 0 16 .5 9 16 .9 0 17 .6 0 16 .3 3 16 .8 8 18 .3 0 16 .4 3 18 .3 8 17 .5 2 17 .6 0 8 C . p er si m il is 11 .5 2 11 .5 2 11 .3 4 15 .6 9 11 .2 7 7. 84 8. 09 –– 14 .0 2 13 .7 1 14 .8 8 18 .9 3 19 .2 8 19 .4 7 19 .9 4 21 .0 3 20 .7 2 21 .1 8 9 C . g ab bi 13 .7 3 13 .7 3 14 .0 9 17 .3 2 12 .9 9 11 .0 3 10 .2 9 7. 11 –– 12 .9 3 13 .2 4 19 .0 6 18 .8 2 20 .7 2 19 .4 7 20 .4 8 20 .8 7 21 .1 8 10 C . s te jn eg er ia nu s 14 .5 4 14 .2 2 14 .0 9 16 .3 4 12 .9 9 11 .5 2 10 .7 8 6. 62 4. 41 –– 10 .2 0 19 .6 5 19 .9 5 21 .8 8 19 .7 8 20 .9 5 21 .3 4 21 .3 4 11 C . l au ra st er 14 .8 7 15 .6 9 15 .8 1 17 .1 6 12 .0 1 11 .0 3 10 .7 8 7. 60 7. 84 5. 39 –– 18 .4 8 20 .7 5 21 .5 0 20 .3 3 21 .3 4 20 .8 7 20 .7 2 12 C . p od ic if er us " M on te V er de " 6. 59 6. 21 7. 07 12 .2 5 11 .0 3 9. 23 8. 01 10 .5 4 12 .7 5 13 .4 0 13 .7 3 –– 10 .2 0 10 .8 0 9. 76 10 .9 3 10 .2 3 10 .5 1 13 C . p od ic if er us s en su s tr ic to 5. 69 5. 29 5. 54 10 .6 2 10 .2 0 8. 92 7. 75 9. 34 12 .0 1 11 .9 1 12 .9 9 3. 09 –– 11 .4 0 9. 11 11 .0 6 10 .2 2 10 .8 8 14 C . p od ic if er us " S io la " 7. 84 7. 35 6. 99 12 .9 1 12 .0 1 9. 31 8. 33 11 .0 3 12 .2 5 11 .7 6 14 .7 1 6. 37 5. 20 –– 10 .4 4 11 .9 2 10 .2 8 11 .8 4 15 C . p od ic if er us " P ic o B la nc o" 6. 37 5. 88 6. 13 11 .1 1 10 .0 5 9. 56 8. 33 10 .0 5 13 .2 4 13 .2 4 12 .2 5 2. 45 1. 96 5. 88 –– 6. 93 8. 88 10 .7 5 16 C . p od ic if er us " Sa n G er ar do " 8. 50 8. 66 8. 78 12 .7 5 11 .1 9 10 .3 8 9. 15 11 .3 6 13 .5 6 13 .5 6 12 .5 8 3. 65 4. 85 7. 84 2. 94 –– 10 .9 8 12 .4 6 17 C . p od ic if er us " C hu m ac er a" 7. 35 7. 84 7. 78 12 .3 4 12 .0 1 10 .0 5 9. 31 11 .5 2 13 .7 3 13 .2 4 15 .2 0 4. 58 4. 00 7. 35 4. 17 5. 07 –– 8. 72 18 C . p od ic if er us " F ila C os te ña " 7. 52 7. 03 7. 11 12 .9 1 11 .3 6 10 .0 5 8. 82 10 .5 4 12 .4 2 13 .4 0 13 .0 7 4. 25 3. 63 7. 19 2. 45 4. 96 5. 64 –– 86 Systematics Redefinition of Craugastor podiciferus (Cope, 1875) and C. blairi (Barbour, 1928) Craugastor podiciferus: The precise type locality of Craugastor podiciferus has been a matter of some uncertainty. The taxon was described by Cope (1875) based on material collected by W.M. Gabb from “slope of Cerro Pico Blanco” in 1874. Savage (1974) corrected it to Cerro Utyum, Cantón de Talamanca, Provincia de Limón, 1524–2134 m a.s.l. but collected no additional specimens because he did not reach that site. Recently, Arias and Chaves (2014) corrected the type locality to a place between Cerro Pat and the headwaters of the Río Lari, elev. 1520–2135, Provincia de Limón, Caribbean slope of Cerro Kamuk. The type locality is a remote site within the Parque Internacional La Amistad, in the Cordillera de Talamanca. It is only accessible on foot and requires three days of hiking from the last village (Amubri, Talamanca). We collected specimens in the surroundings of the type locality according to Arias and Chaves (2014) (Fig. 1 and 4a). Based on the phylogenetic relationships, we propose to restrict the taxon Craugastor podiciferus to the populations from Cordillera Volcánica Central from Costa Rica and Cordillera de Talamanca (Caribbean slopes) in Costa Rica and western Panama (Fig. 1). Craugastor jota: This species was described by Lynch (1980) based on specimens from the Changena river in western Panama, at 760 m a.s.l., collected by Linda Trueb in 1966. This species was placed in the C. podiciferus species group based on morphology (Hedges et al. 2008). In this study, we included specimens from Changena river, western Panama, very near the type locality of C. jota (Linda Trueb pers. comm.). In our phylogenetic analyses these specimens fall within the C. podiciferus sensu stricto clade. As a result, we suggest that C. jota should be referred to as a junior synonym of C. podiciferus. 87 Craugastor blairi: The precise type locality of Craugastor blairi is unknown. Barbour (1928) indicated the type locality as “from Gutierrez, Bocas del Toro Province, Panama (near Costa Rican frontier)”. The type series was collected by Emmett R. Dunn and Chester Duryea in the summer of 1923. Their expedition followed a trail from Chiriquicito at the Laguna de Chiriquí (today in the Corregimiento of Miramar, Bocas del Toro) to Boquete, Province of Chiriquí. Most of their trail climbs up the Atlantic slopes of the Cordillera Central (Savage 1970) before descending into the high valley of Boquete. On the same route, Dunn and Duryea collected several other new species and identified the type localities as either “Gutierrez” or “La Loma”. Species with the type locality “La Loma” are Dermophis parviceps (Dunn, 1924), Hyloscirtus colymba (Dunn, 1931), Pristimantis pardalis (Barbour 1928), and Pristimantis caryophyllaceus (Barbour 1928). All these species have a vertical distribution range that enters the lowlands to almost sea level (Köhler 2011). Dunn gives the elevation of La Loma as 2000 feet (610 m) in the description of D. parviceps and as 1500 feet (460 m) in the description of H. colymba. In comparison to “La Loma”, a species with the type locality “Gutierrez” is Craugastor obesus (Barbour 1928). Dunn (1940) also mentioned that he collected C. monnichorum at “Gutierrez”, and both species had a more premontane to montane distribution. Therefore, we believe that “Gutierrez” is further uphill than “La Loma” and thus closer to the continental divide and much closer to Boquete than to the Caribbean lowlands. We collected several specimens at different sites in the vicinity of Boquete on the foothills of Barú Volcano that form a separate clade with specimens collected at Cerro La Estrella, Cerro Guayaba, Guayabito, and La Fortuna (Fig. 1 and 4b). For the above reasons, we resurrect the name C. blairi for this clade. Restricting the taxon Craugastor podiciferus to populations inhabiting the Caribbean slopes of Cordillera de Talamanca of Costa Rica and extreme western Panama and C. blairi to populations inhabiting the Cordillera Central from Barú Volcano to the east, results in two allopatric lineages 88 on the Pacific slopes of Cordillera de Talamanca, Costa Rica and those from Cordillera Central, Panama without assignment to an existing taxon. With no available name present in the synonymy of C. podiciferus for either of these populations, here we describe these two lineages as new species and provide redescriptions for C. podiciferus and C. blairi. Fig. 4. In life photographs of (a) Craugastor podiciferus (UCR 23169) from the Caribbean slope of Cerro Kamuk, Costa Rica, (b) C. blairi (SMF 104032) from Fortuna, Panama, (c) C. sagui sp. nov. (SMF 104018) from La Nevera, Panama, and (d) C. zunigai sp. nov. (UCR 20389) from Potrero Grande, Costa Rica. Photo A by E. Arias, B-C by A. Hertz, and D by E. Boza-Oviedo. 89 Fig. 5. Variation of the ventral views of the right hands. (a) Craugastor podiciferus (UCR 23175), (b) C. blairi (SMF 104032), (c) C. sagui sp. nov. holotype (SMF 104018), and (d) C. zunigai sp. nov. holotype (UCR 22703). Photos A and D by E. Arias, B-C by G. Köhler. Craugastor podiciferus (Cope, 1875) Common name: Polymorphic Dirt Frog (Figs. 4a and 6) Syntypes: USNM 30662, USNM 30664–75, and MCZ 11841. All specimens from “ 5000 to 7000 feet (elevation), on the Caribbean slopes of Cerro Pico Blanco”, collected by William M. Gabb on 1874. 90 Genetic reference specimen: UCR 23175 (EAP 0810), an adult female from Costa Rica: Provincia de Limón: Cantón de Talamanca: Distrito de Telire: Parque Internacional La Amistad, (9.366º, -83.042º; 1860 m a.s.l.), collected by Erick Arias and Omar Zúñiga on 27 October 2016. Referred specimens: UCR 23155 (EAP 0803), adult female, same data as the genetic reference specimen. UCR 23145 (EAP 0792), an adult male from Costa Rica: Provincia de Limón: Cantón de Talamanca: Distrito de Telire: Cerro Pat, Parque Internacional La Amistad, (9.393º, -83.025º; 1450 m a.s.l.), collected by Erick Arias and Omar Zúñiga on 26 October 2016. Assignment to group: Assigned to Craugastor based on molecular analysis and on the following characters: cranial crest absents and Toe III larger than Toe V. Diagnosis: The combination of the following characteristics can be used to distinguish Craugastor podiciferus (Fig. 5a–6) from other described species in the genus: 1) skin on the dorsum is smooth to scattered tubercles; 2) skin on the venter is smooth, at least in the midline; 2) vocal slits in adult males; 3) nuptial pads absent; 4) unwebbed toes; 5) heel with a projecting tubercle; 6) accessory palmar and plantar tubercles absent, usually no supernumerary tubercles under the digits; and 7) subarticular tubercles flat in form. Craugastor (Craugastor) podiciferus is a small species with the following characteristics: (1) skin on the dorsum smooth to scattered tubercles; head smooth; venter smooth; flanks smooth with scattered tubercles to warty; posterior surface of hind limbs surrounding cloaca strongly areolate; some specimens with a pair of scapular, dorsolateral or lateral folds; discoidal fold complete laterally and posteriorly; (2) tympanic membrane round, heavily pigmented; prominent in males, evident in females; annulus evident through the skin; (TY/ED = 44.9–100 %); usually with a pair of supratympanic folds; (3) snout subovoid in the dorsal view, rounded in profile; 91 loreal region concave; canthus rostralis usually rounded; (4) eyelid granular, with several low tubercles forming a more or less distinct ridge on the outer edge of the eyelid continuous with the supratympanic fold (EW/IOD = 40.5–71.2 %); cranial crests absent; (5) vomerine teeth in two transverse fasciculi, behind the choanae; choanae smaller than the dentigerous; (6) vocal slits and large single vocal sac in adult males; nuptial pads absent; (7) Fingers I and II subequal; discs absent, some specimens with terminal transverse grooves on fingers, especially in males; t ips symmetric, usually rounded but pointed in Fingers III-IV in some specimens; pads ovoid to triangular; (8) fingers lack lateral fringes; webbing absent; thenar and palmar tubercles low, ovoid, similar in size; supernumerary tubercles absent; accessory palmar tubercles usually absent but 1-2 low barely distinct tubercles visible in some specimens; subarticular tubercles round in basal outline, flat in form and globular in profile; (9) ulnar fold absent but tubercles sometimes visible; (10) heel with a projecting tubercle; inner edge tarsal with an indistinct short ridge, outer smooth or with tubercles; (11) toes lacking lateral fringes; inner metatarsal tubercle elongate, outer rounded, much smaller than inner, inner and outer metatarsal tubercles projecting; supernumerary and plantar tubercles absent; subarticular tubercles rounded to ovoid in the basal outline, flat in form and obtuse in profile; (12) Toe III larger than Toe V; discs and terminal transverse grooves present on all fingers; tips symmetrical, disc covers spadate; pads triangular; webbing absent; (13) coloration very variable; dorsum tan to light or dark brown, nearly uniform or suffused with black or reddish pigment; frequently paired suprascapular dark spots, area between dorsal folds usually contrasting with flank color; several specimens with a dark mask from the snout continuing above the tympanum and often bordered above by a narrow light line; venter yellow, grayish, or reddish, uniform or with light or dark spots; usually forelimbs and hind limbs with dark bars, some specimens with paired dark spots on the anterior surface of the hind limbs; the upper lip has dark bars with white pigment in the form of faded bars; and (14) SVL in males 21–28 mm; SVL in females 23–40 mm. 92 Fig. 6. Dorsal and ventral views in ethanol of Craugastor podiciferus (a,b) syntype (USNM 30672) and (c,d) molecular specimen reference (UCR 23175). Photos A-B by E.M. Langan, C-D by E. Arias. Comparison: Craugastor podiciferus differs from all other craugastorids of Isthmian Central America, excluding those in the C. podiciferus species group, by having unwebbed toes and a narrow head (HW 37.0–43.3 % of SVL). Craugastor podiciferus differs from other members of the C. podiciferus species group by having the following characteristics (condition for C. podiciferus in parentheses). Craugastor bransfordii (Cope, 1886), C. gabbi, C. lauraster (Savage, McCranie, and Espinal, 1996), C. persimilis (Barbour, 1926), C. polyptychus (Cope, 1886), C. stejnegerianus 93 (Cope, 1893), and C. underwoodi (Boulenger, 1896) differ from C. podiciferus by the following features: a) dorsum usually granular or warty (dorsum smooth to scattered tubercles); b) subarticular tubercles projecting (subarticular tubercles flat); and c) venter coarsely areolate, including the midline (venter smooth, at least in midline). Craugastor aenigmaticus Arias, Chaves, and Parra-Olea, 2018 differ from C. podiciferus by the following features: a) absence of a prominent calcar tubercle on the heel, although some specimens can have one to three small tubercles (a prominent calcar tubercle on the heel); b) venter violet-brown with white blotches (venter yellow, orange, grayish or olive in adults); and c) white prominent folds between subarticular tubercles on the hands of adults (absence of white folds between subarticular tubercles on the hands of adults). Craugastor blairi (Barbour, 1928) differs from C. podiciferus by the following features: a) skin on the venter coarsely areolate (venter smooth, at least in midline); b) absence of a prominent calcar tubercle on the heel (a prominent calcar tubercle on the heel); c) having accessory palmar tubercles (accessory palmar tubercles absent); and d) having subarticular tubercles projecting (subarticular tubercles flat). Craugastor sagui differs from C. podiciferus by the following features: a) nuptial pads in adult males (nuptial pads absent); b) absence of a prominent calcar tubercle on the heel (a prominent calcar tubercle on the heel); and c) absence of vocal slits in adult males (vocal slits in adult males). Craugastor zunigai differs from C. podiciferus by the following features: a) skin on the venter coarsely areolate (venter smooth, at least in midline); b) absence of a prominent calcar tubercle on the heel (a prominent calcar tubercle on the heel); c) accessory palmar tubercles (accessory palmar tubercles absent), and d) subarticular tubercles projecting (subarticular tubercles flat). Natural history: Craugastor podiciferus inhabits the lower montane rainforest (Holdridge 1967; Bolaños et al. 2005), which is characterized by a very short dry season (one to three months), annual precipitation ranging from 3600 to 7500 mm and annual temperature from 12 to 17 ºC. 94 Very little is known about the natural history of C. podiciferus; however, it is important to note that the species was abundant during the months of fieldwork. The specimens were always found on the floor jumping during the active search. Schlaepfer and Figeroa-Sandí (1998) described the call of C. podiciferus from Las Cruces, herein referred to as C. podiciferus “Fila Costeña”. The advertising call of C. podiciferus sensu stricto is unknown, although it is known to vocalize. Distribution: Craugastor podiciferus sensu stricto is restricted to the highlands of the Cordillera Volcánica Central in Costa Rica and Caribbean slopes of the Cordillera de Talamanca in Costa Rica and western Panama (Fig. 1). The altitudinal range of C. podiciferus is 1700–2700 m a.s.l. All populations of C. podiciferus are found in primary forests, and several localities are within protected areas, (i.e., Parque Internacional La Amistad, Parque Nacional Tapantí-Macizo de la Muerte, Parque Nacional Braulio Carrillo, and Parque Nacional Juan Castro Blanco). More fieldwork is necessary to clarify the distribution of C. podiciferus, especially on the northern end of the Cordillera de Talamanca and in the adjacent zone between the Cordillera Volcánica Central and the Cordillera de Tilarán. Remarks: We tentatively assign several populations from the highlands of Costa Rica to Craugastor podiciferus. However, these populations are phylogenetically structured and show uncorrected p-distances in the 16S rRNA gene between 2.45 and 6.37 % (Table 1), and some differ morphologically from typical C. podiciferus and thus may represent as many as six additional unnamed species. These six populations are as follows: 1) the C. podiciferus “Monte Verde” clade, which corresponds to clade A of Streicher et al. (2009), a clade restricted to Cordillera de Tilarán and Volcánica Central, at 1500–1931 m a.s.l. The specimens in this clade are morphologically very similar to C. podiciferus sensu stricto. 2) The C. podiciferus “San Gerardo” clade, which corresponds to clade B of Streicher et al. (2009), a clade restricted to 95 Cordillera de Tilarán and Cordillera Volcánica Central, at 1470–1500 m a.s.l. The specimens in this clade differ from C. podiciferus s.s. in having projecting subarticular tubercles. 3) The C. podiciferus “Pico Blanco” clade, which contains a single population from the northern end of the Cordillera de Talamanca in the Central valley, at 2242 m a.s.l. The specimens in this clade differ from C. podiciferus s.s. in having an areolate venter and projecting subarticular tubercles. 4) The C. podiciferus “Fila Costeña” clade, which corresponds to clades E and F of Streicher et al. (2009), a clade restricted to South Pacific Costa Rica, at 1350–1550 m a.s.l. The specimens in this clade differ from C. podiciferus s.s. by having an areolate venter and accessory palmar tubercles. 5) The C. podiciferus “Chumacera” clade, which is known for only one population on the pacific slopes of Cordillera de Talamanca, at 1750–1850 m a.s.l. The specimens in this clade differ from C. podiciferus s.s. by having accessory palmar tubercles. 6) The C. podiciferus “Siola”, which is known for only one population on the Caribbean slope of the Cordillera de Talamanca, at 1300–1350 m a.s.l. The specimens in this clade differ from C. podiciferus s.s. by having an areolate venter and by the prominent pungent calcar tubercle on the heel. Despite the molecular results and morphological differences between some of these clades, there is little to distinguish some populations in particular characters, especially C. podiciferus sensu stricto, C. podiciferus “Monte Verde”, and C. podiciferus “Chumacera”. In addition, this group forms a monophyletic group, thus agreeing with the definition of the taxon by Savage (2002), Leenders (2016), and Cossel and Kubicki (2018). For these reasons and until new morphological evidence support the distinctiveness, we refrain from raising these clades to the species level. 96 Table 2. Mean, standard deviation (S.D.), and range for morphometric variables by species. Variable Craugastor podiciferus Craugastor blairi Craugastor sagui sp. nov. Craugastor zunigai sp. nov. Mean ± S.D. Range Mean ± S.D. Range Mean ± S.D. Range Mean ± S.D. Range SVL 25.6±4.3 16.9–33.9 23.2±4.8 13.4–30.6 21.5±6.5 13.0–30.1 20.8±3.9 13.8–26.5 HL 10.4±1.5 7.0–12.9 8.3±1.5 5.6–11.0 8.1±2.5 4.7–10.9 8.4±1.5 5.7–10.3 HW 10.2±1.6 6.5–13.3 8.9±1.8 5.5–12.0 8.0±2.4 4.9–11.5 8.2±1.6 5.6–10.7 ED 3.0±0.5 2.0–3.9 2.9±0.6 1.9–3.8 2.9±0.7 2.0–4.0 2.5±0.3 1.9–2.9 TY 1.8±0.4 1.0–2.7 1.8±0.4 0.8–2.6 1.7±0.6 0.8–2.6 2.0±0.3 1.6–2.9 EW 1.9±0.3 1.2–2.4 1.6±0.3 0.9–2.1 1.5±0.4 0.9–2.0 1.6±0.3 1.2–2.1 IOD 3.3±0.5 2.5–4.2 3.3±0.6 1.8–4.3 2.9±0.8 1.7–3.9 2.7±0.5 1.8–3.7 EN 2.6±0.4 1.9–3.2 2.2±0.4 1.3–3.0 2.1±0.8 1.1–3.2 2.0±0.4 1.3–2.6 EW/IOD 2.8±0.3 2.3–3.6 0.50±0.09 0.34–0.73 0.52±0.06 0.46–0.61 0.59±0.08 0.44–0.76 IOD/HW 0.6±0.1 0.4–0.7 0.37±0.04 0.30–0.45 0.37±0.03 0.34–0.40 0.34±0.02 0.30–0.38 TY/ED 0.6±0.1 0.4–1.0 0.66±0.23 0.40–1.09 0.58±0.18 0.40–0.81 0.81±0.17 0.62–1.34 EN/ED 0.9±0.1 0.7–1.0 0.77±0.08 0.63–0.91 0.71±0.11 0.55–0.88 0.81±0.10 0.63–0.98 ED/HL 0.3±0.1 0.2–0.3 0.35±0.03 0.30–0.43 0.37±0.04 0.31–0.43 0.29±0.02 0.26–0.33 HL/HW 1.0±0.1 0.9–1.1 0.94±0.05 0.84–1.05 1.01±0.06 0.95–1.08 1.03±0.05 0.95–1.14 EN/HL 0.2±0.1 0.2–0.3 0.27±0.02 0.23–0.30 0.26±0.02 0.23–0.29 0.24±0.02 0.20–0.27 Craugastor blairi comb. new. (Barbour, 1928) Common name: Blair’s Dirt Frog (Figs. 4b and 7) Holotype: MCZ 13036 from Gutierrez, Bocas del Toro Province, Panama (near Costa Rican border), collected by E. R. Dunn and Chester Duryea on summer 1925. Genetic reference specimen: SMF 104032 (AH 379), an adult male from Panama: Provincia de Chiriquí: Distrito de Gualaca: La Fortuna, western slope of Cerro Pata de Macho (8.679º, - 82.193º; 1793 m a.s.l.), collected by Andreas Hertz and Sebastian Lotzkat on 20 May 2010. Referred specimens: SMF 104030 (AH 377) and SMF 104031 (AH 378), adult males; same date as the genetic reference specimen. SMF 104025 (AH 196), adult female from Panama: Provincia de Chiriquí: Distrito de Gualaca: La Fortuna, (8.672º, -82.200º; 1400 m a.s.l.), collected by Andreas Hertz and Sebastian Lotzkat on 20 March 2009. SMF 104026 (AH 238) 97 and SMF 104027 (AH 239), adult females from Panama: Provincia de Chiriquí: Distrito de Bugaba: Volcán Barú, Sendero Los Quetzales (8.849º, -82.515º; 2134 m a.s.l.), collected by Andreas Hertz and Sebastian Lotzkat on 8 April 2009. Fig. 7. Dorsal and ventral views in ethanol of Craugastor blairi (a,b) holotype (MCZ A-13036) and (c,d) molecular specimen reference (SMF 104032). Photos A-B by the Museum of Comparative Zoology, Harvard University, C-D by G. Köhler. Assignment to group: Assigned to the genus Craugastor based on molecular data and on the following characters: cranial crest absents and Toe III larger than Toe V. Assigned to the C. podiciferus species group based on the following features: narrow head (HW/SVL = 34.6–42.8 %), dorsum smooth to scattered tubercles, unwebbed toes, vocal slits in adult males, and nuptial pads absent. 98 Diagnosis: The combination of the following characteristics can be used to distinguish Craugastor blairi (Fig. 5b and 7) from other described species of the genus: 1) skin on the dorsum is smooth or has scattered tubercles; 2) skin on the venter is coarsely areolate; 3) vocal slits and vocal sac present in adult males; 4) nuptial pads absent; 5) unwebbed toes; 6) heel without an enlarged calcar tubercle, although one to three small tubercles or granules can be present on the heel; 7) accessory palmar and plantar tubercles present, and absence of supernumerary tubercles; and 8) subarticular tubercles projecting. Craugastor (Craugastor) blairi is a small species with the following characteristics: (1) skin on dorsum smooth or has scattered tubercles; head smooth; venter coarsely areolate: flanks smooth with scattered tubercles to warty; posterior surface of hind limbs surrounding the cloaca strongly areolate; some specimens with a pair of scapular, dorsolateral or lateral folds; discoidal fold complete laterally and posteriorly; (2) tympanic membrane round, heavily pigmented in females, translucent in adult males; prominent in males, evident in females; annulus evident through the skin; (TY/ED = 40–109 %); usually with a pair of supratympanic folds; (3) snout subovoid in the dorsal view, rounded in profile; loreal region concave; canthus rostralis usually rounded; (4) eyelid granular, with an evident supraocular tubercle and a more or less distinct ridge on outer edge of eyelid continuous with the supratympanic fold (EW/IOD = 37.2–72.7 %); cranial crests absent; (5) vomerine teeth in two transverse fasciculi, behind the choanae; choanae smaller than the dentigerous; (6) vocal slits and a single vocal sac that is large in adult males; nuptial pads absent; (7) Finger I and II subequal; discs absent, some specimens with terminal transverse grooves on fingers, especially in males; tips symmetric, usually rounded; pads ovoid to triangular; (8) fingers lacking lateral fringes; webbing absent; thenar and palmar tubercles low, ovoid, similar in size; supernumerary tubercles absent; 1–2 accessory palmar tubercles; subarticular tubercles round in the basal outline, projecting in form and globular in the profile; (9) ulnar fold absent but tubercles visible; (10) heel without an enlarged calcar tubercle, although 99 one to three small tubercles or granules can be present on the heel; inner edge tarsal smooth, outer with an incomplete fold and/or tubercles; (11) toes lacking lateral fringes; inner metatarsal tubercle elongate, outer rounded, much smaller than inner, inner and outer metatarsal tubercles projecting; supernumerary tubercles absent; several low plantar tubercles; subarticular tubercles rounded to ovoid in the basal outline, projecting in form and obtuse in profile; (12) Toe III larger than Toe V; discs and terminal transverse grooves present on all fingers; tips symmetrical, disc covers palmate to spadate; pads triangular; webbing absent; (13) coloration very variable; dorsum tan to light or dark brown, nearly uniform or suffused with black or reddish pigment; frequently paired suprascapular dark spots; some specimens with a dark mask from the snout continuing above the tympanum and downward behind the axilla, often bordered above by a narrow light line; venter yellow, grayish, or reddish, uniform or with light or dark spots; usually forelimbs and hind limbs with dark bars, some specimens with paired dark spots on the anterior surface of the hind limbs; upper lip with dark bars with white pigment in the form of faded bars; and (14) SVL in males 15.9–21 mm; SVL in females 13.4–30.6 mm. Comparisons: Craugastor blairi differs from all the other Isthmian Central America craugastorids (except for those in the C. podiciferus species group) by having unwebbed toes and a narrow head (HW 34.6–42.8 % of SVL). Craugastor blairi differs from other members of the C. podiciferus species group by having the following characteristics (condition for C. blairi in parentheses). Craugastor bransfordii, C. gabbi, C. lauraster, C. persimilis, C. polyptychus, C. stejnegerianus, and C. underwoodi differ from C. blairi by the following features: a) dorsum usually granular or warty (dorsum smooth to having scattered tubercles); b) subarticular tubercles obtuse to pointed, at least the distal subarticular tubercles under Toe III and IV (subarticular tubercles projecting, globular); and c) altitudinal range, 0–1600 m a.s.l. (altitudinal range is 1280–2134 m a.s.l. for C. blairi). Craugastor podiciferus differs from C. blairi by the 100 following features: a) a prominent calcar tubercle on the heel (absence of a prominent calcar tubercle on the heel); b) subarticular tubercles flat (subarticular tubercles projecting); and c) venter smooth, at least in the midline (venter coarsely areolate, including the midline). Craugastor aenigmaticus differs from C. blairi by the following features: a) venter smooth (venter coarsely areolate); b) subarticular tubercles flat (subarticular tubercles projecting, globular); and c) prominent white folds between subarticular tubercles (absence of white folds between subarticular tubercles). Craugastor sagui differs from C. blairi by the following features: a) venter smooth (venter coarsely areolate); b) nuptial pads in adult males (nuptial pads absent); c) absence of vocal slits (vocal slits in adult males); and d) subarticular tubercles flat (subarticular tubercles projecting). Craugastor zunigai differs from C. blairi by the absence of an evident supraocular tubercle (eyelid with an evident supraocular tubercle). Natural history: Craugastor blairi inhabits the lower montane rainforest (Holdridge 1967; Bolaños et al. 2005), which is characterized by a very short dry season (one to three months). Annual precipitations range from 3600 to 7500 mm and annual temperature from 12 to 17 ºC. Very little is known about the natural history of C. blairi; however, it is important to note that the species was abundant during the months of fieldwork. The specimens were always found on the floor in leaf litter. It is possibly active during the day. Males call during periods of lower light due to dark clouds or in the evening hours. Calling activity is greater during the rain. Vocalization: Vocalizations of four male specimens (AH 375, 1430 m, dusk, 19.5 °C, 100 % RH, AH 377-379, 1793 m, 18.5 °C, 100 % RH) have been recorded on the slopes of Cerro Pata de Macho. All specimens were calling at dusk, between 18:00 h and 19:00 h after a moderate rain. Calling sites were elevated positions only a few centimeters above the ground such as low vegetation, twigs, or roots. Calling stopped completely after dark. One male (AH375) was recorded at 19.5 °C and 100 % relative humidity. The other three males were recorded at 18.5 °C 101 and 100 % relative humidity. Two very different call types could be distinguished. The first that was emitted by all four males was a “squeak” of 0.035–0.065 (0.050±0.007) seconds in length, given in frequent intervals after 7.938–29.124 (11.974±4.825) seconds, resulting in a call rate of 4.66–9.63 (6.67±2.41) calls per minute. The call has two harmonics, the fundamental of which also contain the dominant frequency at 6844–8063 (7306±313) Hz. The dominant frequency is reached in the middle or towards the end of the call, 0.009–0.053 (0.030±0.012) seconds after ignition of the call. The frequency range of the fundamental harmonic is between 4705–6674 (5680±634) Hz low frequencies and 7181–9343(7965±467) Hz high frequencies. The higher harmonic has a frequency range of 9566–14057 (12057±1159) Hz low frequencies and 14182– 16617 (15175±650) Hz high frequencies, with the highest acoustic intensity at 13313–15656 (14147±626) Hz. The second call type could only be recorded in two of the four males. It was identified as a trilling, i.e., strongly pulsed “chirp” composed of mostly two, or less often one or three, notes. The notes are 0.266–0.692 (0.450±0.116) seconds in length, and the total call duration is 0.745– 1.551 (0.914±0.206), depending on how many notes are contained in the call. Calls are repeated after 5.911–18.265 (10.196±3.522) seconds. A single note contains 12–25 (19±4) pulses that are emitted at a pulse rate of 36–46 (43±2) pulses per second. The dominant frequency of 8813– 10313 (9654±236) Hz is reached in the middle or towards the end of the note at pulse 6–21 (13±5). Calls are frequency modulated with the lowest frequencies of 6044–7640 (6617±508) Hz in the first pulses gradually rising towards 10298–12002 (10848±466) Hz high frequencies in the last pulses of each note. Distribution: Craugastor blairi is restricted to western Panama, on the Cordillera Central from Barú Volcano to the east over the La Fortuna depression into the Serranía de Tabasará. The specimens were collected around Barú Volcano, Cerro La Estrella, Cerro Guayaba, Guayabito, and La Fortuna (Fig. 1). The altitudinal range of the species is 1280–2134 m a.s.l. To our current 102 knowledge, the continuous distribution of C. blairi is interrupted between La Fortuna and Guayabito. Both sites are separated by ~80 airline kilometers. Within this gap, we find another distinct clade that seems to be restricted to the surroundings of Cerro Saguí and Cerro Santiago and that we describe below as C. sagui. More fieldwork will be necessary to clarify the distribution of C. blairi with respect to C. sagui, especially in the area between La Fortuna and Guayabito. It would also be interesting to explore the eastern distribution limits of C. blairi that could be found as far east as Santa Fé, Veraguas. Craugastor sagui sp. nov. Common name: Sagui Dirt Frog (Figs. 4c and 8) Holotype: SMF 104018 (AH481), an adult male from Panama: Comarca Ngöbe-Buglé: Distrito de Nole Duima: southeastern slope of Cerro Saguí (8.563º, -81.821º; 1991 m a.s.l.), collected by Andreas Hertz and Sebastian Lotzkat on 9 September 2010. Paratypes: SMF 104014 (AH483), adult female; same date as the holotype. SMF 104017 (AH342), adult male and SMF 104015 (AH045), adult female from Panama: Comarca Ngöbe- Buglé: Distrito de Nole Duima: La Nevera, southern slopes of Cerro Santiago (8.500º, -81.772º; 1700 m a.s.l.), collected by Andreas Hertz and Sebastian Lotzkat on 13 November 2009. Assignment to group: Assigned to the genus Craugastor based on our molecular data and on the following morphological characters: cranial crests absent, and Toe III longer than Toe V. Assigned to the C. podiciferus species group based on the following features: a narrow head (HW/SVL = 35.8–40.9 %), length of Finger I equal to Finger II, dorsum smooth, toes unwebbed, and nuptial pads in adult males. 103 Fig. 8. Dorsal and ventral views in ethanol of the holotype (SMF 104018) of Craugastor sagui sp. nov. Photographs taken by G. Köhler. Diagnosis: The combination of the following characteristics can be used to distinguish Craugastor (Craugastor) sagui (Fig. 4c, 5c and 8) from the other described species in the genus: 1) skin on the dorsum smooth; 2) skin on the venter smooth; 3) vocal slits absent; 4) nuptial pads in adult males; 5) unwebbed toes; 6) heel without an enlarged calcar tubercle, although one to three small tubercles or granules may be present on the heel; 7) accessory palmar, plantar, and supernumerary tubercles absent; and 8) subarticular tubercles flat. Comparison: Craugastor sagui differs from all other Isthmian Central America craugastorids (except for those in the C. podiciferus species group) by having unwebbed toes and a narrow head (HW 35.8–40.9 % of SVL). Craugastor sagui differs from other members of the C. podiciferus species group by having the following characteristics (condition for C. sagui in parentheses). Craugastor bransfordii, C. gabbi, C. lauraster, C. persimilis, C. polyptychus, C. stejnegerianus, and C. underwoodi differ from C. sagui by the following features: a) dorsum usually granular or warty (dorsum smooth to shagreen with scattered tubercles); b) subarticular tubercles projecting (subarticular tubercles flat); and c) venter coarsely areolate, including the midline (venter smooth, at least in midline). Craugastor podiciferus differs from C. sagui by the 104 following features: a) a prominent calcar tubercle on the heel (Fig. 6) (calcar tubercle absent, although some specimens can have one to three small tubercles); b) vocal slits present in adult males (vocal slits absent); and c) absence of nuptial pads (nuptial pads present in adult males). Craugastor aenigmaticus differs from C. sagui by the following features: a) venter coloration of violet-brown with white blotches (venter yellow, orange, grayish or olive in adults); b) white prominent folds between subarticular tubercles on the hands of adults (absence of white folds between subarticular tubercles on the hands of adults); and c) absence of nuptial pads (nuptial pads present in adult males). Craugastor blairi differs from C. sagui by the following features: a) venter coarsely areolate (venter smooth); b) absence of nuptial pads (nuptial pads present in adult males); c) vocal slits present in adult males (vocal slits absent in adult males); and d) projecting subarticular tubercles (flat subarticular tubercles). Craugastor zunigai differs from C. sagui by the following features: a) venter coarsely areolate (venter smooth); b) subarticular tubercles projecting (subarticular tubercles flat); c) absence of nuptial pads (nuptial pads present in adult males); and d) vocal slits in adult males (vocal slits absent). Description of the holotype: Adult male have an SVL of 18.7 mm (Fig. 4–5). Head relatively narrow, HW = 35.8 % of SVL; snout subovoid in the dorsal view, rounded in profile; snout relatively long (HL = 7.0 mm, 37.4 % of SL), with nostrils directed laterally; in the ventral view, tip of the snout protruding slightly beyond the edge of the lower lip. Internarial area convex; canthus rostralis rounded; intercanthal area flat; loreal region slightly concave; vocal slits absent. Eye moderate (EN/ED = 74.07 %), not protruding beyond the dorsal and ventral outline of the head, directed laterally. Tympanic membrane distinct, covered in skin; tympanic annulus prominent, round, relatively large (77.77 % of ED). Skin on all dorsal and lateral surfaces of the head smooth to shagreen. Upper eyelid granular, without superciliar or supraocular tubercles but with a more or less discernible ridge on the outer edge of eyelid continuous with the 105 supratympanic fold and downward behind the axilla. Postrictal tubercles fused to form a short ridge postero-ventral to the tympanum. Skin on the dorsum and limbs smooth to shagreen. Skin of the chest and throat smooth, venter smooth; ventral surfaces of the thighs smooth; skin of the groin nearly smooth. Flanks shagreen, especially along the antero-ventral flank region. Discoidal fold complete. Forelimb relatively short and slim; fingers moderately long and slim without lateral fringes. Discs absent; fingers with grooves; tips of fingers unexpanded, rounded in the dorsal view; pads ovoid. Supernumerary tubercles absent; accessory palmar tubercles absent; subarticular tubercles rounded in the basal outline, flat in form and globular in profile; thenar and palmar tubercles elongate and flat; thenar and palmar tubercles similar in size. Ulnar fold absent but several tubercles visible. Fingers not webbed. Legs relatively long and slim; heel granular, lacking enlarged tubercles. Discs and grooves in all toes, palmate in Toe IV, spadate in others; pads triangular in Toe IV, ovoid in others. Supernumerary and plantar tubercles absent; subarticular tubercles rounded in the basal outline, flat in form and obtuse in profile; inner metatarsal tubercle elongate, globular; outer metatarsal tubercle rounded, globular; outer much smaller than inner; outer edge tarsal with several tubercles, inner smooth. Cloacal opening directed posteriorly at the mid-level of the thighs. Coloration of the holotype in life: Dorsal background color dark brown-orangish with dark blotches, and a dark interorbital bar. Dorsal surfaces of legs and arms with dark bars, anterior surface of legs with white spots. Upper lip brown-orange with dark bars and scattered white spots. Flanks with a similar dorsal coloration but with a paler orange, with fine white-bluish mottling. Ventral surface of the body and legs dark brown-reddish with white-bluish pigment- forming blotches; ventral surface of the throat similar to the venter but with fewer pale spots. The sole of the feet and hands dark brown with cream-colored tubercles. 106 Coloration of the holotype in ethanol: After eight years in ethanol (70 %), the overall dark brown-orangish on the dorsum faded to pale cream-pinkish with dark brown blotches. The dark brown-orangish on the ventral surface of the body and legs has faded to dark brown. Measurements of the holotype (mm): SL 18.7; HL 7.0; HW 6.7; IOD 2.7; EW 1.4; EN 2.0; ED 2.7; TY 2.1. Measurements in related percentages: EW/IOD 51.85 %; IOD/HW 40.3 %; TY/ED 77.78 %; EN/ED 74.07 %; ED/HL 38.57 %; HL/HW 104.48 %; EN/HL 28.57 %. Etymology: The species was discovered on Cerro Saguí, in western Panama. The scientific name is a noun in the apposition. Natural history: Craugastor sagui inhabits the lower montane rainforest (Holdridge 1967; Bolaños et al. 2005), which is characterized by a very short dry season (one to three months), annual precipitation ranging from 3600 to 7500 mm and annual temperature from 12 to 17 ºC. Very little is known about the natural history of C. sagui; however, it is important to note that the species was relatively abundant during the months of fieldwork. The specimens were exclusively found on the floor in the leaf litter, jumping during the active search. We never recorded a vocalization that could be attributed to C. sagui; however, it is possible that the species does vocalize. Distribution: Craugastor sagui is restricted to western Panama, on the pacific slopes of Cordillera Central. The specimens were collected on the surrounding slopes of Cerro Saguí and Cerro Santiago (Fig. 1). The known altitudinal range of the new species is 1700–1991 m a.s.l. 107 Craugastor zunigai sp. nov. Common name: Zúñiga’s Dirt Frog (Figs. 4d and 9) Holotype: UCR 22703 (EAP 0618), adult male from Costa Rica: Provincia de Puntarenas: Cantón de Coto Brus: Distrito de Sabalito: Finca Las Alturas, Zona Protectora Las Tablas, (8.976º, -82.834; 1732 m a.s.l.), collected by Erick Arias, Gerardo Chaves, and Omar Zúñiga on 15 September 2015. Paratypes: UCR 23014 (EAP 0725), UCR 23016 (EAP 0727), UCR 23017 (EAP 0728), and UCR 23018 (EAP 0729), adult females from Costa Rica: Provincia de Puntarenas: Cantón de Buenos Aires: Distrito de Potrero Grande: Tres Colinas, Parque Internacional La Amistad, (9.123º, -83.066º; 1846 m a.s.l.), collected by Erick Arias, Fanny Hernández, and Omar Zúñiga on September 10, 2016. UCR 23176 (EAP 0831), adult male and UCR 23177 (EAP 0832), adult female from Costa Rica: Provincia de Puntarenas: Cantón de Coto Brus: Distrito de Pittier: Santa María de Pittier, Parque Internacional La Amistad, (9.031º, -82.962; 1920 m a.s.l.), collected by Erick Arias and Omar Zúñiga on 27 March 2018. Assignment to group: Assigned to the genus Craugastor based on molecular data and on the following characters: cranial crests absent and Toe III larger than Toe V. Assigned to the C. podiciferus species group based on our phylogeny and on the following characters: narrow head (HW/SVL = 35.8–41.9 %), dorsum smooth to shagreen with scattered tubercles, unwebbed toes, vocal slits in adult males, and absence of nuptial pads. 108 Fig. 9. Dorsal and ventral views in ethanol of the holotype (UCR 22703) of Craugastor zunigai sp. nov. Photographs taken by E. Arias. Diagnosis: The combination of the following characteristics can be used to distinguish Craugastor (Craugastor) zunigai (Fig. 4d, 5d and 9) from other described species in the genus: 1) skin on the dorsum smooth to shagreen with scattered tubercles; 2) skin on the venter coarsely areolate; 3) vocal slits present in adult males; 4) nuptial pads absent; 5) unwebbed toes; 6) heel without a projecting tubercle, although one to three low tubercles or granules can be present; 7) accessory palmar and plantar tubercles present, usually with no supernumerary tubercles under the digits; and 8) subarticular tubercles projecting. Comparison: Craugastor zunigai differs from all other Isthmian Central America craugastorids (except for those in the C. podiciferus species group) by having unwebbed toes and a narrow head (HW 35.8–41.9 % of SVL). Craugastor zunigai differs from other members of the C. podiciferus species group by having the following characteristics (condition for C. zunigai in parentheses). Craugastor bransfordii, C. gabbi, C. lauraster, C. persimilis, C. polyptychus, C. stejnegerianus, and C. underwoodi differ from C. zunigai by the following features: a) dorsum usually granular or warty (dorsum is smooth to shagreen with scattered tubercles); b) subarticular tubercles obtuse to pointed, at least distal subarticular tubercles under Toe III and IV 109 (subarticular tubercles projecting, globular); and c) altitudinal range, 0–1600 m a.s.l. (altitudinal range is 1500–2100 m a.s.l. for C. zunigai). Craugastor podiciferus differs from C. zunigai by the following features: a) a prominent calcar tubercle on the heel (Fig. 6) (calcar tubercle absent, although some specimens can have one to three small tubercles in C. zunigai); b) venter smooth (venter coarsely areolate); c) subarticular tubercles flat (subarticular tubercles projecting); and d) absence of accessory palmar tubercles (accessory palmar tubercles present). Craugastor aenigmaticus differs from C. zunigai by the following features: a) venter smooth (venter coarsely areolate); b) subarticular tubercles flat (projecting subarticular tubercles, globular); and c) prominent white folds between subarticular tubercles (absence of white folds between subarticular tubercles). Craugastor blairi differs from C. zunigai by the presence of an evident supraocular tubercle (eyelid smooth to granular but without an evident supraocular tubercle). Craugastor sagui differs from C. zunigai by the following features: a) nuptial pads in adult males (nuptial pads absent); b) absence of vocal slits (vocal slits in adult males); c) venter smooth (venter coarsely areolate); and d) flat subarticular tubercles (projecting subarticular tubercles, globular). Description of the holotype: Adult male having an SVL of 19.8 mm (Fig. 9). Head relatively narrow, HW = 35.86 % of SVL; snout subovoid in the dorsal view, rounded in profile; snout relatively long (HL = 7.6 mm, 38.38 % of SL), with nostrils directed laterally; in the ventral view, tip of the snout protruding markedly beyond the edge of the lower lip. Internarial area convex (IN 2.2 mm); canthus rostralis rounded; intercanthal area flat (IC = 3.6 mm); loreal region slightly concave; vomerine teeth transverse, in two fascicles behind the choanae. Tongue round in shape, lacking a distinct posterior notch; teeth absent; choanae moderately large, rounded on the posterior half but flat on the anterior half, hemispherical; paired elongate vocal slits under the posterolateral margins of the tongue and a single internal subgular vocal sac. Eye 110 moderate (EN/ED = 84.1 %), protruding beyond the dorsal outline of head in the ventral view, directed laterally. Tympanic membrane prominent, translucent, and slightly pigmented; tympanic annulus prominent, round, large (134 % of ED). Skin on the dorsal and lateral surfaces of the head smooth. Upper eyelid smooth, without superciliar or supraocular tubercles but with a more or less discernible ridge on the outer edge of the eyelid continuous with the supratympanic fold and downward behind the axilla. Elongate and projecting postrictal tubercle, postero-ventral to the tympanum. Skin on the dorsum and limbs smooth to shagreen with a pair of incomplete dorso-lateral folds, extending from the axillary to the inguinal level. Skin of the chest and throat smooth, venter coarsely areolate with low granules; ventral surfaces of the thighs areolate; skin of the groin smooth. Flanks shagreen with scattered tubercles to areolate, especially along the antero-ventral flank region. Discoidal fold complete. Forelimb relatively short and robust; fingers moderately long and slim without lateral fringes. Discs absent; fingers III-IV with grooves; tips of fingers unexpanded, rounded in dorsal view; pads ovoid. Supernumerary tubercles absent; four small and rounded accessory palmar tubercles; subarticular tubercles rounded in the basal outline, slightly projecting in form, and globular in profile; thenar tubercle elongate and palmar tubercle rounded, flat, similar in size. Ulnar fold absent but some tubercles visible. Fingers not webbed. Legs relatively long and slim (TL = 55.3 % SVL); heel smooth with two barely visible low granules, lacking a projecting tubercle. Discs and grooves in all toes, palmate in Toe IV, spadate in others; pads triangular in Toe IV, ovoid in others. Supernumerary tubercles absent; plantar tubercles small and rounded; subarticular tubercles ovoid in the basal outline, projecting in form, and obtuse in profile; inner metatarsal tubercle elongate, projecting; outer metatarsal tubercle rounded, globular; outer metatarsal tubercle much smaller than inner; outer edge tarsal with an indistinct short ridge and low granules, inner smooth. Cloacal opening directed posteriorly at the mid-level of the thighs. 111 Coloration of the holotype in life: Dorsal background color dark brown suffused laterally with pale brown, head dark brown uniform. Dorsal surfaces of legs and arms with dark bars, posterior surface of legs cream suffused with red. Cloacal opening darker than the posterior surface of the legs. Flanks pale brown, groin suffused with red. Ventral surface of the body and legs cream- yellowish with dark pigment; throat cream with dark pigment. Absence of bars on the lips, mask, and blotches on the dorsal surface. Coloration of the holotype in ethanol: After three years in ethanol (70 %), the overall dark brown on dorsum has remained very similar to that in life. The cream-yellowish color on the ventral surface of the body and legs has faded to pale brown. Measurements of the holotype (mm): SL 25.9; HL 10.3; HW 10.3; IOD 3.7; EW 1.7; EN 2.5; ED 2.9; TY 1.9. Measurements in related percentages: EW/IOD 68.82 %; IOD/HW 28.18 %; TY/ED 54.12 %; EN/ED 81.12 %; ED/HL 27.24 %; HL/HW 107 %; EN/HL 24.34 %. Variation: The morphometric variation is summarized in Table 2. Craugastor zunigai shows a relatively high level of intraspecific polymorphism. In some specimens, a pair of lateral folds is present, and some specimens show a supraocular tubercle. In some specimens, two unfused postrictal tubercles are visible. Palmar and thenar tubercles are equal in size in some specimens, thenar slightly smaller than palmar in others. The palmar tubercle is heart-shaped in some specimens and ovoid in others. UCR 20389 with a pair of lateral folds from the axillar level to the cloaca bordered by black pigment, the area between folds and from the snout to the cloaca a lighter, cream-yellowish color in ethanol. UCR 20428 with a pattern of dark brown on the dorsum with a lighter interorbital mark and area anterior to it paler than the dorsum. The mask is absent in some specimens. The throat has a nearly uniform cream color in UCR 20401, heavily 112 mottled in UCR 20421, and nearly uniform dark brown in UCR 20389. The venter ranges from a nearly uniform cream color without dark pigment in UCR 20401 to heavily mottled in UCR 20389. Etymology: The name zunigai is a patronym honoring the field-guide Omar Zuñiga in recognition of his important aid during the fieldwork on the Cordillera de Talamanca. Omar Zuñiga took part in the fieldwork that yielded the specimens from Tres Colinas, Las Alturas, Las Tablas, and Santa María de Pittier. Natural history: Craugastor zunigai inhabits the lower montane rainforest (Holdridge 1967; Bolaños et al. 2005), which is characterized by a very short dry season (one to three months), annual precipitation ranging from 3600 to 7500 mm and annual temperature from 12 to 17 ºC. Very little is known about the natural history of C. zunigai; however, is important to note that the species was relatively abundant during the months of fieldwork. The specimens were always found on the floor jumping during the active search. We never recorded a vocalization that could be attributed to C. zunigai; however, it is possible that the species vocalize. At Las Alturas, C. zunigai occur very near the C. gabbi localities; however, we did not find them in sympatry since they are altitudinally structured. In Tres Colinas, the pattern is similar but with C. stejnegerianus. Distribution: Craugastor zunigai is restricted to Southwestern Costa Rica, on the Pacific slopes of Cordillera de Talamanca. The specimens were collected from Santa María de Pittier, Tres Colinas, Las Alturas de Cotón, and road to Las Tablas (Fig. 1). The altitudinal range of the new species is 1500–2100 m a.s.l. All populations were found in primary forests, the populations from Santa María de Pittier and Tres Colinas are in La Amistad International Park and the population from Las Alturas de Cotón and road to Las Tablas is in Zona Protectora Las Tablas. The populations of C. zunigai are fragmented over ~43 km. This species was not found on the 113 pacific slope of Cerro Utyum and Cerro Dúrika (to the west). Additional fieldwork is necessary to clarify the distribution of C. zunigai. The locality Road to Las Tablas is very close to the Costa Rica-Panama border, so C. zunigai is likely also present in Panama. Discussion With the recognition of Craugastor blairi, C. zunigai, and C. sagui, the C. podiciferus species group is now formed by 12 species, all of which are collectively distributed in Costa Rica and western Panama (Savage 2002; AmphibiaWeb 2019). The species in this group have been difficult to delineate historically because they are morphologically variable between and within populations. However, using molecular sequence data, we found well-supported clades and large genetic distances between several of these populations (Table 1). Although the sole use of genetic distances for species delimitation is not a recommended practice, we believe that large genetic distances between phylogenetically related and geographically close species are one important measure to identify cryptic species in species groups with a conservative morphology. The genetic distances we present herein between members of the C. podiciferus species group are above the thresholds of 3 % in the 16S rRNA and 10 % in the COI mitochondrial genes suggested by Fouquet et al. (2007) and Vences et al. (2005). For amphibians, the 16S rRNA gene fragment has been suggested as a DNA barcode marker for amphibian diversity inventories (Vences et al. 2005) to complement the standard COI-5’ marker used in general for animals (Smith et al. 2008). Although the species recognized herein are not cryptic sensu stricto (Pérez- Ponce de León and Nadler 2010), they are in taxonomic practice. As has been suggested by Pérez-Ponce de León and Nadler (2010), the use of molecular data within morphologically conserved groups allow us to delineate taxa that otherwise would be considered a single taxon. The use of phylogenies based on molecular data has been essential to solve taxonomic problems in this group. 114 T ab le 3 . M ai n di ag no st ic c ha ra ct er is ti c an d ch ar ac te r st at es f or s ec on da ry s ex ua l c ha ra ct er is ti cs o f th e sp ec ie s fo rm in g th e C ra ug as to r po di ci fe ru s sp ec ie s gr ou p. Sp ec ie s D or sa l s ki n te xt ur e V en te r sk in te xt ur e T he n ar a n d p al m ar t ub er cl es in s iz e re la ti on Su ba rt ic ul ar in pr of ile Su pe r- nu m er ay tu be rc le s A cc es so ry pa lm ar tu be rc le s H ee l s ki n t ex tu re N up ti a l p ad s V oc al sl it s V oc al sa c C . a en ig m at ic us Sm oo th to w ea kl y gr an ul ar Sm oo th w it h gr an ul es o n la te ra ll y T he na r tu be rc le m uc h sm al le r th an pa lm ar tu be rc le Fl at A bs en t A bs en t Sm oo th to th re e lo w gr an ul es n ot p ro je ct in g A bs en t A bs en t A bs en t C . b ra ns fo rd ii G ra nu la r to w ar ty C oa rs el y ar eo la te T he na r tu be rc le e qu al o r sl ig ht ly s m al le r th an p al m ar tu be rc le Pr oj ec ti ng Pr es en t Pr es en t G ra nu la r Pr es en t A bs en t A bs en t C . b la ir i Sm oo th to s ha gr ee n w it h sc at te re d tu be rc le s C oa rs el y ar eo la te T he na r tu be rc le e qu al o r sl ig ht ly s m al le r th an p al m ar tu be rc le Pr oj ec ti ng A bs en t Pr es en t Sm oo th to s ev er al lo w gr an ul es n ot p ro je ct in g A bs en t Pr es en t Pr es en t C . g ab bi Sh ag re en w it h sc at te re d gr an ul es to g ra nu la r C oa rs el y ar eo la te T he na r tu be rc le m uc h sm al le r th an pa lm ar tu be rc le Pr oj ec ti ng Pr es en t Pr es en t G ra nu la r A bs en t A bs en t A bs en t C . l au ra st er G ra nu la r to w ar ty C oa rs el y ar eo la te T he na r tu be rc le m uc h sm al le r th an pa lm ar tu be rc le Pr oj ec ti ng Pr es en t Pr es en t G ra nu la r A bs en t A bs en t A bs en t C . p er si m il is G ra nu la r C oa rs el y ar eo la te T he na r tu be rc le m uc h sm al le r th an pa lm ar tu be rc le Pr oj ec ti ng Pr es en t Pr es en t G ra nu la r A bs en t A bs en t A bs en t C . p od ic if er us Sm oo th to s ca tt er ed tu be rc le s Sm oo th T he na r tu be rc le e qu al o r sl ig ht ly s m al le r th an p al m ar tu be rc le Fl at A bs en t A bs en t A p ro je ct in g tu be rc le A bs en t Pr es en t Pr es en t C . p ol yp ty ch us G ra nu la r to w ar ty C oa rs el y ar eo la te T he na r tu be rc le e qu al o r sl ig ht ly s m al le r th an p al m ar tu be rc le Pr oj ec ti ng Pr es en t Pr es en t G ra nu la r A bs en t A bs en t A bs en t C . s ag ui s p. n ov . Sm oo th to s ha gr ee n w it h sc at te re d tu be rc le s Sm oo th T he na r tu be rc le e qu al o r sl ig ht ly s m al le r th an p al m ar tu be rc le U su al ly f la tt ed , bu t p ro je ct in g in so m e sp ec im en s A bs en t A bs en t Sm oo th to s ev er al lo w gr an ul es n ot p ro je ct in g Pr es en t A bs en t A bs en t C . s te jn eg er ia nu s G ra nu la r to w ar ty C oa rs el y ar eo la te T he na r tu be rc le m uc h sm al le r th an pa lm ar tu be rc le Pr oj ec ti ng Pr es en t Pr es en t G ra nu la r A bs en t A bs en t A bs en t C . u nd er w oo di G ra nu la r to w ar ty C oa rs el y ar eo la te T he na r tu be rc le e qu al o r sl ig ht ly s m al le r th an p al m ar tu be rc le Sl ig ht ly pr oj ec tin g U su al ly ab se nt Pr es en t G ra nu la r Pr es en t A bs en t A bs en t C . z un ig ai s p. n ov . Sm oo th to s ha gr ee n w it h sc at te re d tu be rc le s C oa rs el y gr an ul ar T he na r tu be rc le e qu al o r sl ig ht ly s m al le r th an p al m ar tu be rc le Sl ig ht ly pr oj ec tin g A bs en t Pr es en t G ra nu la r to tw o lo w gr an ul es n ot p ro je ct in g A bs en t Pr es en t Pr es en t 115 Taylor (1952) synonymized C. blairi under C. podiciferus without further discussion, nor was it discussed in the comprehensive seminar work of Savage (2002). Our results support the validity of Craugastor blairi (previously shown as Craugastor sp. B), which it is not a sister to C. podiciferus (Crawford and Smith 2005; Streicher et al. 2009). However, C. blairi and C. podiciferus are morphologically similar and show a high level of phenotypic polymorphism, and thus the taxonomic decision of Taylor (1952) is not surprising given that the technical capabilities at that time were limited. The distinctiveness of C. blairi is provided by its phylogenetic position. The same rules apply to C. sagui, the phylogenetic position of which provides significant evidence for its different evolutionary trajectory as a species beyond the observation that it resembles C. podiciferus and C. blairi in morphology. The morphological similarity among nonsister species can be explained either as plesiomorphy or convergence (Castroviejo-Fisher et al. 2017). All species analyzed herein have in common that they inhabit the ground leaf litter in highland forest habitats (lower montane zonation), which has likely played a role in the maintenance of this morphology. A more complicated situation is observed the recognition of Craugastor zunigai, given its morphological resemblance to its sister species C. blairi. Although the morphological divergence between C. blairi and C. zunigai is weak, we believe that both are valid species supported by phylogenetic distinctiveness based on mitochondrial analysis. These species are geographically very close (~10 airline km); thus, it seems unlikely that the large genetic distances (2.9 % in 16S and 14.3 % in COI) are explained by isolation due to distance. Our taxonomic decision to recognize these two as separate species will be strengthened with additional evidence, especially the comparison of mating calls of C. zunigai with the call of C. blairi, which we have described herein. Future sampling efforts must also be conducted on the highlands of extreme western Panama and adjacent Costa Rica between Barú Volcano and Cerro Pando to explore the species 116 distribution limits and possible contact zones of C. aenigmaticus, C. blairi, C. zunigai, and C. podiciferus. The presence of several species within Craugastor podiciferus is uncertain. Streicher et al. (2009) suggested that the current concept of C. podiciferus could be masking six distinct species that are geographically structured but with two instances of sympatry. However, the findings of Streicher et al. (2009) and our own results agree with C. podiciferus being a monophyletic entity; thus, there is currently no need to split C. podiciferus in several species until such division is supported by detailed morphological or other evidence. We strongly suggest the performance of an extensive, integrative analysis of the C. podiciferus species complex to re-evaluate the taxonomic status of its different genetic clades. For now, the recognition of C. podiciferus sensu latu as a single species results in a highly variable morphological group, given that even those few useful characters (i.e., skin on venter, subarticular tubercles, vocal slits, and nuptial pads) are highly variable between populations. The clades shown herein should be reevaluated using other types of data that can provide evidence of lineage divergence, such as the geographical distribution, ecological niche, mating calls, or detailed morphometric data. The members of the Craugastor podiciferus species group have qualities such as high abundance, broad distribution (collectively), and high genetic diversity that make them suitable for use in a variety of different studies in ecology and evolution. However, these species have been poorly studied likely because of the difficulty in clearly identifying the species. Therefore, it is necessary to clarify the taxonomy of this species group. Upon revisiting our study objectives, we report the following: 1) Craugastor podiciferus sensu stricto is restricted to the populations of Cordillera Volcánica Central, Costa Rica and Cordillera de Talamanca on Costa Rica and western Panama, and the latter is restricted to the Caribbean slopes; 2) six well- supported clades are found within the current concept of C. podiciferus, these six clades require 117 extensive taxonomic revision; and 3) the populations from southwestern Costa Rica and western Panama are grouped in three clades, one potentially referring to an existing name, C. blairi, which is resurrected herein, and two representing new species that are herein named C. sagui from western Panama and C. zunigai from southwestern Costa Rica. Acknowledgments.—We thank Laura Márquez-Valdelamar and Andrea Jiménez-Marín for their laboratory assistance and Federico Bolaños for the use of specimens from the Museo de Zoología of the Universidad de Costa Rica. We thank Linda Acker and Gunther Köhler of the Senckenberg herpetology collection (SMF) for providing photographs of some of the specimens examined herein. Sebastian Lotzkat, Frank Hauenschild, Nadim Hamad, Omar Zúñiga, Olmer Cordero, Justo Layam Gabb, and Xavier Baltodano provided valuable assistance in the field during the expeditions. We thank Rogelio Moreno, former general chief of the Ngöbe-Buglé indigenous community, for granting us access to the Comarca Ngöbe-Buglé, as well as all the Ngöbe and Buglé who helped us logistically during the field work. AH is thankful to the owners and staff of the Lost and Found Ecohostel in Valle de las Minas, Chiriquí for their enduring support and hospitality. EA thanks the Posgrado en Ciencias Biológicas for its support of this study, the CONACyT for the student grant (CVU/Becario) 626946/330343, and the Programa de Innovación y Capital Humano para la Competitividad PINN-MICITT for the student grant (PED- 0339-15-2). The laboratory work was funded by a grant from PAPIIT-UNAM (IN203617) to GP-O. The fieldwork was partially supported by the National Geography Society (Grant number W-346-14). We acknowledge the Costa Rican Ministry of Environment and Energy (MINAE) for providing the corresponding scientific collecting permits for this expedition (SINAC-SE- GAS-PI-R 007-2013 and 59-2015). Collecting permits for Panama SE/A-30-08, SC/A-8-09, SC/A-28-09, and SC/A-21-10, as well as the corresponding exportation permits, were issued by the Dirección de Áreas Protegidas y Vida Silvestre of the Autoridad Nacional del Ambiente (ANAM), recently renamed the Ministerio de Ambiente (MiAmbiente), Panama City, Panama. 118 Literature Cited AmphibiaWeb. 2019. AmphibiaWeb: Information on Amphibian Biology and Conservation [web application]. Berkeley, California. Available: http://amphibiaweb.org [Accessed: 10 January 2019]. Arias E, Chaves G. 2014. 140 years after William M. Gabbs's climb to Cerro Pico Blanco. Mesoamerican Herpetology 1: 176–180. Arias E, Chaves G, Crawford AJ, Parra-Olea G. 2016. A new species of the Craugastor podiciferus Species Group (Anura: Craugastoridae) from the premontane forest of southwestern Costa Rica. Zootaxa 4132: 347–363. Arias E, Chaves G, Parra-Olea G. 2018. A new species of Craugastor (Anura: Craugastoridae) from the montane rainforest of the Cordillera de Talamanca, Costa Rica. Phyllomedusa 17: 211–232. Barbour T. 1926 New Amphibia. Occasional Papers of the Boston Society of Natural History 5: 191–194. Barbour T. 1928. New Central American frogs. Proceedings of the New England Zoölogical Club 10: 25–31. Bolaños R, Watson V, Tosi J. 2005. Mapa ecológico de Costa Rica (Zonas de Vida), según el sistema de clasificación de zonas de vida del mundo de L.R. Holdridge), Escala 1:750 000. Centro Científico Tropical, San José, Costa Rica. vol 214 Boulenger GA. 1896. Descriptions of new batrachians collected by Mr. C. G. Underwood in Costa Rica. Annals and Magazine of Natural History 18(Series 6): 340–342. Castroviejo-Fisher S, Köhler J, De La Riva I, Padial JM. 2017. A new morphologically cryptic species of Phyllomedusa (Anura: Phyllomedusidae) from Amazonian forests of northern Peru revealed by DNA sequences. Zootaxa 4269: 245–264. Cope ED. 1875. On the Batrachia and Reptilia of Costa Rica. Journal Academy of Natural Sciences 2: 93–157. 119 Cope ED. 1886. Thirteenth contribution to the herpetology of tropical America. Proceedings of the American Philosophical Society 23: 271–287. Cope ED. 1893. Second addition to the knowledge of the Batrachia and Reptilia of Costa Rica. Proceedings of the American Philosophical Society 31: 333–347. Cossel Jr JO, Kubicki B. 2017. A field guide to the frogs and toads of Costa Rica. BookBaby, Pennsauken, New Jersey, USA. Crawford AJ. 2003. Huge populations and old species of Costa Rican and Panamanian dirt frogs inferred from mitochondrial and nuclear gene sequences. Molecular Ecology 12: 2525– 2540. Crawford AJ, Smith EN. 2005 Cenozoic biogeography and evolution in direct-developing frogs of Central America (Leptodactylidae: Eleutherodactylus) as inferred from a phylogenetic analysis of nuclear and mitochondrial genes. Molecular Phylogenetics and Evolution 35: 536–555. Dayrat B. 2005 Toward integrative taxonomy. Biological Journal of the Linnean Society 85: 407–415. de Queiroz K. 2007. Species concepts and species delimitation. Systematic Biology 56: 879–886. Duellman WE, Lehr E. 2009 Terrestrial-breeding frogs (Strabomantidae) in Peru. Natur- und Tier-Verlag, Münster, Germany. 382 p. Dunn ER. 1940. New and noteworthy herpetological material from Panama. Proceedings of the Academy of Natural Sciences of Philadelphia 92: 105–122. Frost DR. 2019. Amphibian Species of the World: an Online Reference. Version 6.0. Available: http://research.amnh.org/herpetology/amphibia/index.html [Accessed: 10 January 2019]. Fouquet A, Gilles A, Vences M, Marty C, Blanc M, Gemmell NJ. 2007. Underestimation of species richness in Neotropical frogs revealed by mtDNA analyses. PLoS ONE 2: e1109. 120 Hedges SB, Duellman WE, Heinicke MP. 2008. New World direct-developing frogs (Anura: Terrarana): Molecular phylogeny, classification, biogeography, and conservation. Zootaxa 1737: 1–182. Holdridge LR. 1967. Life Zone Ecology. Tropical Science Center, San José, Costa Rica, 206 p. Katoh K, Rozewicki J, Yamada KD. 2017. MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization. Briefings in Bioinformatics bbx108. Köhler G. 2011. Amphibians of Central America. Herpeton Verlag, Offenbach, Germany. 374 p. Lanfear R, Calcott B, Ho SY, Guindon S. 2012. PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Molecular Biology and Evolution 29: 1695–1701. Leenders T. 2016. Amphibians of Costa Rica: a field guide. Comstock Publishing Associates, Ithaca, New York, United States. 531 p. Lleonart J, Salat J, Torres GJ. 2000. Removing allometric effects of body size in morphological analysis. Journal of Theoretical Biology 205: 85–93. Lynch JD. 1980. A new frog of the genus Eleutherodactylus from western Panama. Transactions of the Kansas Academy of Science 83: 101–105. Manly BFJ. 1994. Multivariate Statistical Methods: A Primer. Chapman & Hall, London, England. 232 p. Miller MA, Pfeiffer W, Schwartz T. 2010. Creating the CIPRES Science Gateway for inference of large phylogenetic trees. Proceedings of the Gateway Computing Environments Workshop (GCE) 14 November 2010: 1–8. Padial JM, Miralles A, De la Riva I, Vences M. 2010. The integrative future of taxonomy. Frontiers in Zoology 7: 16. Pérez-Ponce de León G, Nadler SA. 2010. What we don't recognize can hurt us: a plea for awareness about cryptic species. Journal of Parasitology 96: 453–464. 121 R Core Team. 2017. R: a language and environment for statistical computing. R Foundation for Statistical Computing , Vienna, Austria. Version 3.3.3. Available: http://www.R- project.org Rambaut A, Suchard MA, Xie W, Drummond AJ. 2014. Tracer v1.6.1. Available: http://beast.bio.ed.ac.uk/Tracer. Ronquist F, Teslenko M, Van Der Mark P, Ayres DL, Darling A, Höhna S, Larget B, Liu L, Suchard M, Huelsenbeck JP. 2012. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic Biology 61: 539–542. Savage JM. 1970. On The trail of the Golden Frog: with Warszewicz and Gabb in Central America. Proceedings of the California Academy of Sciences: Festschrift for George Sprague Myers 38: 273–288. Savage JM. 1974. Type localities for species of amphibians and reptiles described from Costa Rica. Revista de Biología Tropical 22: 71–122. Savage JM. 2002. The Amphibians and Reptiles of Costa Rica: A Herpetofauna between two Continents, between two Seas. The University of Chicago Press, Illinois, USA. 934 p. Savage JM, McCranie JR, Espinal M. 1996. A new species of Eleutherodactylus from Honduras related to Eleutherodactylus bransfordii (Anura: Leptodactylidae). Proceedings of the Biological Society of Washington 109: 366–372. Schlaepfer M, Figeroa-Sandi R. 1998. Female reciprocal calling in a Costa Rican leaf-litter frog, Eleutherodactylus podiciferus. Copeia 1998: 1076–1080. Simpson GG. 1951. The species concept. Evolution 5: 285–298. Smith MA, Poyarkov Jr NA, Hebert PDN. 2008. CO1 DNA barcoding amphibians: Take the chance, meet the challenge. Molecular Ecology Resources 8: 235–246. Streicher JW, Crawford AJ, Edwards CW. 2009. Multilocus molecular phylogenetic analysis of the montane Craugastor podiciferus species complex (Anura: Craugastoridae) in Isthmian Central America. Molecular Phylogenetics and Evolution 53: 620–630. 122 Sukumaran JM, Holder MT. 2010a. DendroPy: a Python library for phylogenetic computing. Bioinformatics 26: 1569–1571. Sukumaran J, Holder MT. 2010b. SumTrees: Phylogenetic Tree Summarization. Version 4.4.0. Available: https://github.com/jeetsukumaran/DendroPy. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. 2013. MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Molecular Biology and Evolution 30: 2725–2729. Taylor EH. 1952. The salamanders and caecilians of Costa Rica. University of Kansas Science Bulletin 34: 695–791. Vences M, Thomas M, Bonett RM, Vieites DR. 2005. Deciphering amphibian diversity through DNA barcoding: chances and challenges. Philosophical Transactions of the Royal Society of London B: Biological Sciences 360: 1859–1868. Wiley EO. 1978. The evolutionary species concept reconsidered. Systematic Zoology 27: 17–26. Zwickl DJ. 2006. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. Ph.D. Dissertation, The University of Texas, Austin. 123 A pp en di x I. I ns tit ut io na l v ou ch er n um be rs , l oc al it y in fo rm at io n, a nd G en B an k ac ce ss io n nu m be rs f or th e sp ec im en s us ed in t he m ol ec ul ar ph yl og en et ic a na ly se s. M us eu m c ol le ct io n ac ro ny m s fo llo w F ro st ( 20 19 ) w ith t he a dd it io n of E A P to r ef er to E ri ck A ri as f ie ld n um be rs a nd C R A R C to r ef er to th e C os ta R ic a A m ph ib ia n R es ea rc h C en te r pr iv at e co lle ct io n. Sp ec ie s In st it ut io na l vo uc he rs C ol le ct io n lo ca lit y E le va ti on (m ) G eo gr ap hi c co or di na te s G en B an k N um be r L at L on 16 S C O I C . a en ig m at ic us S M F : 1 04 02 0 C ha ng ui no la , B oc as d el T or o, P A 23 88 8. 91 39 -8 2. 70 88 M K 21 16 15 M K 21 15 77 C . a en ig m at ic us U C R : 2 19 51 T el ir e, T al am an ca , C R 27 00 9. 34 88 -8 3. 17 50 M K 21 16 16 M K 21 15 78 C . a en ig m at ic us U C R : 2 27 37 B ue no s A ir es , P un ta re na s, C R 26 60 9. 32 24 -8 3. 20 28 M K 21 16 17 M K 21 15 79 C . b la ir i F M N H : 2 57 68 9 G ua la ca , C hi ri qu í, P A 10 00 8. 75 00 -8 2. 21 70 E F 56 23 53 N D C . b la ir i S M F : 1 02 02 4 G ua la ca , C hi ri qu í, P A 17 30 8. 67 75 -8 2. 19 80 M K 21 16 27 M K 21 15 83 C . b la ir i S M F : 1 04 02 3 G ua la ca , C hi ri qu í, P A 12 80 8. 67 81 -8 2. 21 01 M K 21 16 28 M K 21 15 84 C . b la ir i S M F : 1 04 02 7 B ug ab a, C hi ri qu í, C R 21 34 8. 84 94 -8 2. 51 54 M K 21 16 29 M K 21 15 85 C . b la ir i S M F : 1 04 03 3 G ua la ca , C hi ri qu í, P A 14 56 8. 67 40 -8 2. 21 54 M K 27 93 67 N D C . b la ir i S M F : 1 04 03 4 Ñ ur um , N gö be B ug lé , P A 15 41 8. 55 12 -8 1. 48 33 M K 27 93 68 N D C . b la ir i S M F : 1 04 03 7 B oq ue te , C hi ri qu í, P A 19 52 8. 77 57 -8 2. 39 01 M K 27 93 69 N D C . b la ir i U S N M : 5 63 03 9 B oq ue te , C hi ri qu í, P A 16 63 8. 31 36 -8 2. 40 00 E F 56 23 56 N D C . b ra ns fo rd ii U C R : 2 22 69 A la ju el a, A la ju el a, C R 46 6 10 .3 12 1 -8 4. 17 78 K T 95 02 95 M K 21 15 71 C . b ra ns fo rd ii U C R : 2 26 43 S iq uí rr es , L im ón , C R 53 7 10 .0 59 5 -8 3. 54 52 M K 21 16 10 M K 21 15 72 C . g ab bi U C R : 2 18 63 C ot o B ru s, P un ta re na s, C R 12 00 8. 78 89 -8 2. 95 83 K T 95 02 71 M K 21 15 67 C . g ab bi U C R : 2 18 64 C ot o B ru s, P un ta re na s, C R 12 00 8. 78 89 -8 2. 95 83 K T 95 02 72 M K 21 15 68 C . l au ra st er S M F : 7 97 59 M at ag al pa , M at ag al pa , N I 13 00 12 .9 99 3 -8 5. 90 92 M K 21 16 08 M K 21 15 65 C . l au ra st er U S N M : 5 59 39 3 P ue rt o L em pi ra , G ra ci as a D io s, H N 19 0 14 .9 27 5 -8 4. 53 39 K U 32 33 64 M K 21 15 66 C . p er si m il is U C R : 2 22 11 P ar aí so , C ar ta go , C R 10 50 9. 78 41 -8 3. 75 17 K T 95 02 93 M K 21 15 70 C . p er si m il is U C R : 2 26 71 T al am an ca . L im ón , C R 12 1 9. 57 73 -8 2. 93 43 M K 21 16 09 M K 21 15 69 124 A pp en di x I. C on ti nu ed . Sp ec ie s In st it ut io na l vo uc he rs C ol le ct io n lo ca lit y E le va ti on (m ) G eo gr ap hi c co or di na te s G en B an k N um be r L at L on 16 S C O I C . p od ic if er us s en su s tr ic to C R A R C : 0 01 2 T ur ri al ba , C ar ta go , C R 22 50 10 .0 19 2 -8 3. 71 32 M K 21 16 33 M K 21 15 89 C . p od ic if er us s en su s tr ic to S M F : 1 04 00 5 C ha ng ui no la , B oc as d el T or o, P A 17 66 8. 99 08 -8 2. 67 16 M K 21 16 41 M K 21 15 97 C . p od ic if er us s en su s tr ic to U C R : 1 98 53 T el ir e, T al am an ca , C R 18 17 9. 35 80 -8 3. 22 94 M K 21 16 39 M K 21 15 95 C . p od ic if er us s en su s tr ic to U C R : 1 98 56 T el ir e, T al am an ca , C R 18 17 9. 35 80 -8 3. 22 94 M K 21 16 37 M K 21 15 93 C . p od ic if er us s en su s tr ic to U C R : 1 98 60 T el ir e, T al am an ca , C R 21 08 9. 36 45 -8 3. 21 64 M K 21 16 36 M K 21 15 92 C . p od ic if er us s en su s tr ic to U C R : 1 98 62 T el ir e, T al am an ca , C R 21 08 9. 36 45 -8 3. 21 64 M K 21 16 38 M K 21 15 94 C . p od ic if er us s en su s tr ic to U C R : 2 09 92 A lf ar o R ui z, A la ju el a, C R 21 43 10 .2 27 2 -8 4. 34 82 M K 21 16 32 M K 21 15 88 C . p od ic if er us s en su s tr ic to U C R : 2 21 46 V áz qu ez d e C or on ad o, S an J os é, C R 17 00 10 .0 26 3 -8 3. 94 48 M K 21 16 35 M K 21 15 91 C . p od ic if er us s en su s tr ic to U C R : 2 22 01 D ot a, S an J os é, C R 23 95 9. 71 26 -8 3. 94 88 M K 21 16 34 M K 21 15 90 C . p od ic if er us s en su s tr ic to U C R : 2 31 75 T al am an ca , L im ón , C R 18 60 9. 36 59 -8 3. 04 17 M K 21 16 40 M K 21 15 96 C . p od ic if er us “ C hu m ac er a” U C R : 2 21 20 B ue no s A ir es , P un ta re na s, C R 18 21 9. 32 18 -8 3. 45 46 M K 21 16 42 N D C . p od ic if er us “ C hu m ac er a” U C R : 2 26 90 P ér ez Z el ed ón , S an J os é, C R 17 93 9. 32 67 -8 3. 47 06 M K 21 16 31 M K 21 15 87 C . p od ic if er us “ Fi la C os te ña ” E A P : 0 50 9 G ol fi to , P un ta re na s, C R 15 46 8. 78 78 -8 3. 03 06 N D M K 21 16 05 C . p od ic if er us “ Fi la C os te ña ” F M N H : 2 57 65 1 C ot o B ru s, P un ta re na s, C R 13 50 8. 78 33 -8 2. 98 33 E F 56 23 67 N D C . p od ic if er us “ Fi la C os te ña ” U C R : 1 65 85 D ot a, S an J os é, C R 14 00 9. 53 53 -8 3. 85 80 M K 21 16 47 N D C . p od ic if er us “ Fi la C os te ña ” U C R : 2 20 91 P ér ez Z el ed ón , S an J os é, C R 14 88 9. 44 10 -8 3. 68 30 M K 21 16 46 M K 21 16 04 C . p od ic if er us “ M on te V er de ” F M N H : 2 57 66 9 M on te V er de , P un ta re na s, C R 15 00 10 .2 77 3 -8 4. 58 91 E F 56 23 72 M K 21 15 98 C . p od ic if er us “ M on te V er de ” F M N H : 2 57 67 3 M on te V er de , P un ta re na s, C R 15 00 10 .2 77 3 -8 4. 58 91 E F 56 23 43 M K 21 16 03 C . p od ic if er us “ M on te V er de ” U C R : 1 63 61 A lf ar o R ui z, A la ju el a, C R 19 30 10 .2 17 6 -8 4. 36 71 E F 56 23 71 N D C . p od ic if er us “ M on te V er de ” U C R : 2 26 75 P un ta re na s, P un ta re na s, C R 17 26 10 .3 20 2 -8 4. 79 87 N D M K 21 16 06 C . p od ic if er us “ Pi co B la nc o” U C R : 2 22 26 E sc az ú, S an J os é, C R 22 42 9. 86 46 -8 4. 14 29 M K 21 16 44 M K 21 16 01 C . p od ic if er us “ Pi co B la nc o” U C R : 2 22 28 E sc az ú, S an J os é, C R 22 42 9. 86 46 -8 4. 14 29 M K 21 16 43 M K 21 16 00 125 A pp en di x I. C on ti nu ed . Sp ec ie s In st it ut io na l vo uc he rs C ol le ct io n lo ca lit y E le va ti on (m ) G eo gr ap hi c co or di na te s G en B an k N um be r L at L on 16 S C O I C . p od ic if er us “ Sa n G er ar do ” C R A R C : 0 24 7 T ila rá n, G ua na ca st e, C R 14 70 10 .3 60 0 -8 4. 80 00 M K 21 16 45 N D C . p od ic if er us “ Sa n G er ar do ” F M N H : 2 57 67 1 M on te V er de , P un ta re na s, C R 15 00 10 .2 77 3 -8 4. 58 91 E F 56 23 74 M K 21 15 99 C . p od ic if er us “ Sa n G er ar do ” U C R : 1 63 53 S ar ap iq uí , H er ed ia , C R 15 00 10 .2 02 2 -8 4. 16 25 E F 56 23 49 M K 21 16 02 C . p od ic if er us “ Si ol a” U C R : 2 31 69 T al am an ca , L im ón , C R 13 00 9. 39 87 -8 3. 02 00 M K 21 16 30 M K 21 15 86 C . p ol yp ty ch us U C R : 2 00 50 T al am an ca , L im ón , C R 90 0 9. 61 78 -8 3. 26 81 M K 21 16 14 M K 21 15 76 C . p ol yp ty ch us U C R : 2 26 68 T al am an ca , L im ón , C R 19 8 9. 60 64 -8 2. 91 15 M K 21 16 13 M K 21 15 75 C . s ag ui s p. n ov . S M F : 1 04 01 4 N ol e D ui m a, N gö be B ug lé , P A 17 62 8. 55 71 -8 1. 82 45 M K 21 16 23 N D C . s ag ui s p. n ov . S M F : 1 04 01 5 N ol e D ui m a, N gö be B ug lé , P A 17 00 8. 49 97 -8 1. 77 24 M K 21 16 24 M K 21 15 80 C . s ag ui s p. n ov . S M F : 1 04 01 7 N ol e D ui m a, N gö be B ug lé , P A 18 15 8. 49 55 -8 1. 76 72 M K 27 93 70 N D C . s te jn eg er ia nu s E A P : 0 51 4 O sa , P un ta re na s, C R 45 8. 96 55 -8 3. 44 11 M K 21 16 07 M K 21 15 63 C . s te jn eg er ia nu s U C R : 2 03 52 B ue no s A ir es , P un ta re na s, C R 90 0 9. 08 63 -8 3. 11 05 K T 95 02 84 M K 21 15 64 C . u nd er w oo di U C R : 2 26 19 P ar aí so , C ar ta go , C R 14 12 9. 75 18 -8 3. 77 92 M K 21 16 11 M K 21 15 73 C . u nd er w oo di U C R : 2 26 25 V áz qu ez d e C or on ad o, S an J os é, C R 17 08 10 .0 25 4 -8 3. 94 56 M K 21 16 12 M K 21 15 74 C . z un ig ai s p. n ov . U C R : 2 03 89 B ue no s A ir es , P un ta re na s, C R 15 00 9. 11 12 -8 3. 10 06 M K 21 16 25 M K 21 15 81 C . z un ig ai s p. n ov . U C R : 2 04 28 B ue no s A ir es , P un ta re na s, C R 18 00 9. 13 81 -8 3. 07 00 M K 27 93 71 N D C . z un ig ai s p. n ov . U C R : 2 27 03 C ot o B ru s, P un ta re na s, C R 17 32 8. 97 59 -8 2. 83 44 M K 27 93 72 N D C . z un ig ai s p. n ov . U C R : 2 27 09 C ot o B ru s, P un ta re na s, C R 19 80 8. 97 51 -8 2. 82 43 M K 21 16 26 M K 21 15 82 126 Appendix II. Specimens used in the morphometric analysis. Museum collection acronyms follow Frost (2019) with the addition of AH to refer to Andreas Hertz field numbers and HAU to refer to Frank Hauenschild field numbers. Craugastor blairi: PANAMA: Chiriquí: Bajo Mono, Los Naranjos, Boquete (AH: 0289–0290; SMF: 104028); Cerro La Estrella, Jaramillo, Boquete (SMF: 104037); Cerro Guayaba, Caldera, Boquete (HAU: 023; SMF: 104035); Volcán Barú, Los Naranjos, Boquete (AH: 0240–1; SMF: 104026–7); Fortuna, Hornito, Gualaca (AH: 0079–0080, 0372, 00376; HAU: 008, 010; SMF: 102024, 104023, 104025, 104029–33). Ngöbe Buglé: Guayabito, Ñurüm (SMF: 104034). Craugastor sagui sp. nov.: PANAMA: Ngöbe Buglé: Cerro Saguí, Jädeberi, Nole Duima (SMF: 104014, 104018–9); La Nevera, Nole Duima (AH: 0168; SMF: 104015–7). Craugastor zunigai sp. nov.: COSTA RICA: Puntarenas: Las Alturas, Pittier, Coto Brus (UCR: 22703–4, 22709–10); Tres Colinas, Potrero Grande, Buenos Aires (UCR: 20257–8, 20389, 20395, 20401, 20411, 20419, 20421, 20423, 20428, 23014, 23016–8); road to Las Tablas, Sabalito, Coto Brus (UCR: 23170). 127 CAPÍTULO II.II A NEW SPECIES OF THE CRAUGASTOR PODICIFERUS SPECIES GROUP (ANURA: CRAUGASTORIDAE) FROM THE PREMONTANE FOREST OF SOUTHWESTERN COSTA RICA Autores: Erick Arias, Gerardo Chaves, Andrew J. Crawford y Gabriela Parra-Olea Fuente: Zootaxa, 4132 (3): 347–363. 2016 Fecha de publicación: 30 de junio de 2016 Publicado por: Magnolia Press URL: http://doi.org/10.11646/zootaxa.4132.3.3 128 Accepted by J. Padial: 7 Jun. 2016; published: 30 Jun. 2016 ZOOTAXA ISSN 1175-5326 (print edition) ISSN 1175-5334 (online edition)Copyright © 2016 Magnolia Press Zootaxa 4132 (3): 347–363 http://www.mapress.com/j/zt/ Article 347 http://doi.org/10.11646/zootaxa.4132.3.3 http://zoobank.org/urn:lsid:zoobank.org:pub:E1BA81A4-DF66-404B-A88E-D53E4BC5BD2D A new species of the Craugastor podiciferus species group (Anura: Craugastoridae) from the premontane forest of southwestern Costa Rica ERICK ARIAS1,2,7, GERARDO CHAVES3, ANDREW J. CRAWFORD4,5,6 & GABRIELA PARRA-OLEA1 1Departamento de Zoología, Instituto de Biología, UNAM, AP 70-153 Ciudad Universitaria, CP 04510, México, D.F., México 2Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México; Av. Ciudad Universitaria 3000, C.P. 04360, Coyoacán, Distrito Federal, México 3Escuela de Biología, Universidad de Costa Rica, San Pedro, 11501-2060 San José, Costa Rica 4Departament of Biological Sciences, Universidad de los Andes, A.A. 4976 Bogotá, Colombia 5Círculo Herpetológico de Panamá, Apartado 0824-00122, Panama City, Republic of Panama 6Smithsonian Tropical Research Institute, Apartado 0843-03092, Panama City, Republic of Panama 7Corresponding author: E-mail: eapiedra@gmail.com Abstract In this report, we describe a new species of the Craugastor podiciferus species group from the premontane forest of the Pacific versant along the Costa Rican-Panamanian border. Mitochondrial DNA and karyotype analyses previously showed a marked genetic divergence between populations of the premontane forest of the Fila Costeña and the lowlands South Pacific Costa Rica near Panama. Analyses of the mitochondrial DNA sequences and the morphological variation revealed significant differences between the populations of the premontane forest relative to the other populations of C. stejnege- rianus, including the type locality. We recognize these premontane populations as a new species and show that they differ from the typical C. stejnegerianus in the coloration of the venter, the head and the body proportions, and mtDNA diver- gence. With the addition of this new species, the C. podiciferus species group now contains nine species. Key words: Brachycephaloidea, Costa Rica, Craugastor gabbi sp. nov., Craugastor stejnegerianus, cryptic species, Pan- ama, taxonomy, Terrarana Introduction Terrarana (Brachycephaloidea, sensu Padial et al. 2014) is a clade of New World frogs that use a terrestrial breeding mode and direct development (Hedges et al. 2008). Currently, this clade is composed of more than 1000 species distributed throughout the New World tropics (Padial et al. 2014). Due to its size and complexity, this group has been the target of multiple molecular phylogenetic analyses, resulting in the extensive changes of its composition at all levels; Eleutherodactylus (Duméril & Bibron), the largest vertebrate genus at one time, was split into multiple genera and families (Hedges et al. 2008; Heinicke et al. 2009; Padial et al. 2014), and numerous non- monophyletic subgeneric taxa have been identified (Padial et al. 2014). Craugastor, one of the largest components of the Terraranas, is distributed from the southern USA to northwestern Colombia (Hedges et al. 2008) and is composed of 114 species (AmphibiaWeb 2016). This genus is considered monophyletic based on jaw morphology (Lynch 1986) and mitochondrial and nuclear markers (Crawford & Smith 2005; Frost et al. 2006; Heinicke et al. 2007; Hedges et al. 2008; Pyron & Wiens 2011; Padial et al. 2014). Several species complexes are recognized within Craugastor (Hedges et al. 2008; Padial et al. 2014); however, a lack of morphological distinction among these taxa have made taxonomic work difficult (Savage 2002), and several cryptic new species that are masked under some of the nominal species remain undescribed (Savage 2002; Crawford 2003; Crawford et al. 2007; Hedges et al. 2008; Streicher et al. 2009; McCranie 2015). Craugastor stejnegerianus (Cope 1893), a member of the C. podiciferus species group, was described by Cope (1893) as Hylodes stejnegerianus, based on a specimen from Palmar de Osa, Puntarenas, Costa Rica. Günther 129 ARIAS ET AL.348 · Zootaxa 4132 (3) © 2016 Magnolia Press (1885-1902) placed this species in the synonymy of H. polyptychus (Cope 1886); subsequently, however, Taylor (1952) resurrected and re-described this species and placed it in the genus Microbatrachylus (Taylor). Based on an analysis of the color pattern polymorphisms, Savage & Emerson (1970) concluded that C. stejnegerianus, together with the rest of the C. podiciferus species group, were synonyms of Eleutherodactylus bransfordii (Cope 1886). Using biochemical data, Miyamoto (1983) determined that the populations of E. bransfordii from the Pacific versant of Costa Rica were markedly divergent from the Caribbean versants; thus the name E. stejnegerianus was resurrected and applied to the populations of the Pacific versant of Costa Rica. This finding was subsequently supported by karyotypic differences (Chen 2001, 2005). Savage (2002) found morphological differences between E. bransfordii and E. stejnegerianus and consolidated the hypothesis of specific differentiation. Phylogenetic and population genetic studies based on mitochondrial and nuclear DNA sequence data (Crawford 2003) revealed deep genetic divergences between the populations of C. stejnegerianus and showed that this species was more closely related to C. persimilis (Barbour 1926) than to C. bransfordii. In this study, we perform morphological and molecular phylogenetic analyses of all populations referred to as C. stejnegerianus of the South Pacific Costa Rica and western Panama to determine the taxonomic status of the lowland and premontane populations. We propose the recognition of the premontane populations as a distinct species closely related to C. stejnegerianus. Materials and methods Species criterion. Our definition of species follows the general metapopulation lineage species concept (de Queiroz 2007). Because we adhere to this concept, we recognize a species when there is evidence of the separation of metapopulation lineages, preferably based on multiple lines of evidence following the consensus protocol for integrative taxonomy (Padial et al. 2010). FIGURE 1. Map showing the sampling sites. Samples indicated by the solid color were included in the molecular analyses, and the sites indicated by a number plus letters were included in the morphological analyses. Samples from the sites indicated by open circles and open triangles were included only in the morphological analyses. The numbers correspond to the locality ID in Table 1. Letter codes refer to Table 2. 130 Zootaxa 4132 (3) © 2016 Magnolia Press · 349NEW SPECIES OF CRAUGASTOR FROM COSTA RICA Taxon sampling. Tissue samples were collected from 26 Craugastor podiciferus species group specimens from 16 localities of Costa Rica (Table 1, Fig. 1). The frogs were collected in the field and euthanized, and liver or muscle tissue was preserved in 95% ethanol or RNAlater. Voucher specimens were fixed in 10% formalin, stored in 70% ethanol, and deposited at the Museo de Zoología, Universidad de Costa Rica (UCR) and the Division of Amphibians and Reptiles at the Field Museum of Natural History (FMNH). Additional tissue samples of five specimens of three localities of Panama and Honduras were kindly provided (see Acknowledgments). Amplification and sequencing. We extracted and sequenced a fragment of the mitochondrial 16S rRNA gene from seven samples of the new species described here, 15 samples referred to as Craugastor stejnegerianus from 10 populations, two C. persimilis, two C. bransfordii, two C. lauraster (Savage, McCranie & Espinal 1996), two C. underwoodi (Boulenger 1896), and a specimen referred to C. podiciferus as an outgroup (Table 1). The total genomic DNA was extracted from the preserved tissues (liver or muscle) using the Animal Genomic DNA Kit (BioBasic Canada Inc.). We used the polymerase chain reaction (PCR) to amplify the mitochondrial 16S rRNA gene (16S) using the primers 16Sbr and 16Sar (Palumbi et al. 1991). The PCR amplifications were performed using a total volume of 15 µL, which contained 1 µL DNA template (c. 50 ng µL-1), 0.75 U Taq polymerase (Amplificasa®, Biotecnologias Moleculares), 1X PCR buffer with 1.5 mm MgCl 2 , 0.2 mM deoxynucleotide triphosphates (dNTPs), and 0.3 µM forward and reverse primers. The PCR conditions consisted of an initial cycle of 5 min at 94ºC, followed by 35 cycles of 45 s at 94ºC, 30 or 45 s at 50ºC or 55ºC, 45 or 120 s at 72ºC, plus a final cycle of 5 min at 72ºC. The fragments were sequenced in both directions using the original amplification primers and BigDye termination reaction chemistry (Applied Biosystems). After cycle sequencing, the products were column-purified using a Sephadex G-50 (GE Healthcare) and were run on an ABI PRISM 3100 DNA Analyzer (Applied Biosystems). Consensus sequences for each individual were constructed using SEQUENCHER 5.3 (Genes Codes Corp.). The resulting sequences were deposited in GenBank (Table 1). Phylogenetic analyses. 16S sequences were aligned using the MUSCLE 3.7 software (Edgar 2004). We used the MrModelTest 2.3 (Nylander 2004) and the Akaike Information Criterion (AIC) to select an appropriate model of the DNA sequence evolution, which was the GTR + I model. Analyses were performed using both the maximum likelihood (ML) and Bayesian analyses (BA). The ML analyses were performed using RAxML 8.1.11 (Stamatakis 2014), including 1000 bootstrap replicates to evaluate nodal support. The Bayesian phylogenetic analyses were performed using MrBayes 3.2.2 (Huelsenbeck & Ronquist 2001) and 10 heated MCMC samples of every 1000 generations for 50 million generations. We examined a time-series plot of the likelihood scores of the cold chain to check stationarity using Tracer 1.6 software (Rambaut et al. 2014). We discarded the first 25% of trees as burn-in and used the remaining trees to estimate the consensus tree along with the posterior probabilities for each node and each parameter. Estimates of pairwise evolutionary genetic divergence between and within groups were computed using MEGA6 (Tamura et al. 2013), assuming corrected distances based on the Tamura 3-parameter model (Tamura 1992), with rate variation among the sites modeled as a gamma distribution with the shape parameter = 4 as the default of the software. Morphometric analyses. We examined 35 Craugastor specimens from the premontane forest of Fila Costeña and the Cordillera de Talamanca near the Costa Rican-Panamanian border and 155 specimens representing several populations from lowland South Pacific Costa Rica, including the type locality of C. stejnegerianus (Table 2, Fig. 1; Appendix). All material was deposited at the Museo de Zoología (UCR), Universidad de Costa Rica, San José, Costa Rica. The following morphological measurements were recorded, as described by Savage (2002) and Duellman & Lehr (2009): snout-vent length (SVL), head length (HL), head width (HW), eye diameter (ED), inter orbital distance (IOD), tympanum diameter (TY), width of the upper eyelid (EW), tibia length (TL), eye-nostril distance (E-N), lengths of the toes (T1, T2, T3, T4, T5), and lengths of the fingers (F1, F2, F3, F4). Measurements were made using dial calipers and were rounded to the nearest 0.1 mm. To remove the effect of body size, each morphological measurement was regressed against the SVL, and the residuals from a linear fit with the SVL were used in further analyses. The additional proportions reported here include: EW/IOD, TY/ED, E-N/ED, ED/HL, IOD/HW, and T4/TL. The sex of the individuals was determined based on the TY/ED ratio and gonadal morphology. TY/ED > 0.6 corresponds to males. The specimens with opaque seminal vesicles were assumed to be adult males, and those with developed oviducts were assumed to be adult females. The general terminology for the morphological characteristics follows Duellman & Lehr (2009). We followed Savage (2002) in our usage of the term “supernumerary tubercles,” which we use to refer to the tubercles on the phalanges (between subarticular tubercles); this is different from the tubercles referred to here as accessory palmares or plantares tubercles. 131 ARIAS ET AL.350 · Zootaxa 4132 (3) © 2016 Magnolia Press                                                                                                                          "     # $ %                &    '  (  #         $   %   ) & "         )        %                &    "             " &  +                ,          %         "   &         +     $     # ) ,    )              $                      .      / 0              C . g a b b i    # $ % 1 2  3 * 4  "   $  5  $     6       $ %  7 3 - 3 ! 8 9 3 2 ! 3 8  2 : :  ; < ! : 2 -     C . g a b b i    # $ % 1 2  3 * =  "   $  5  $     6       $ %  7 3 - 3 ! 8 9 3 2 ! 3 8  2 : :  ; < ! : 2 - 2    C . g a b b i    # $ % 1 2  3 *  "   $  5  $     6       $ %  7 3 - 3 ! 8 9 3 2 ! 3 8  2 : :  ; < ! : 2 - 4    C . g a b b i    # $ % 1 2  3 - *  "   $  5  $     6       $ %  7 3 - 3 ! 8 9 3 2 ! 3 8  2 : :  ; < ! : 2 - =  2  C . g a b b i    ) & " 1 ! ! 3 * 4  )   $   %        $   > ( 6 ,  7 3 3 4 4 ? 9 3 2 - 3 4 ?  2 : :  ; < ! : 2 -  2  C . g a b b i    ) & " 1 ! ! 3 * =  )   $   %        $   > ( 6 ,  7 3 3 4 4 ? 9 3 2 - 3 4 ?  2 : :  ; < ! : 2 - *  2  C . g a b b i    ) & " 1 ! ! 3 *  )   $   %        $   > ( 6 ,  7 3 3 4 4 ? 9 3 2 - 3 4 ?  2 : :  ; < ! : 2 - -  4  C . st ej n eg er ia n u s # $ % 1 2 2  2 3  %    6 @ A '    B   )  C @  $ %  7 ! = : = ? 9 3 4 * * ! ? 3 -  ; < ! : 2 - 3  =  C . st ej n eg er ia n u s # $ % 1 2 2  : 4  <      6 @ A '   B   )  C @  $ %  7 ! 4 :  ? 9 3 4 - -  ? - = :  ; < ! : 2 - !   C . st ej n eg er ia n u s # $ % 1 2 2 2 3 :  6   +     D   6       $ %  7 !  ! - ? 9 3 4 - 3 - ? 4 :  ; < ! : 2 3 :  *  C . st ej n eg er ia n u s # $ % 1 2 2 : - :  E  F     ,    6      $ %  7 !  2 ? 9 3 4 = 2 * ? =   ; < ! : 2 3   *  C . st ej n eg er ia n u s # $ % 1 2 2 : -   E  F     ,    6      $ %  7 !  2 ? 9 3 4 = 2 * ? =   ; < ! : 2 3 2  -  C . st ej n eg er ia n u s # $ % 1 2 : 4 = *  6           ,    6       $ %  7 ! : 3 * ? 9 3 4    ? ! : :  ; < ! : 2 3 4  -  C . st ej n eg er ia n u s # $ % 1 2 : 4 2  6           ,    6       $ %  7 ! : 3 * ? 9 3 4    ? ! : :  ; < ! : 2 3 =  3  C . st ej n eg er ia n u s # $ % 1 2 2  4 =  6      D   6       $ %  7 3 ! * * 8 9 3 4 = =  8 =  ; < ! : 2 3  3  C . st ej n eg er ia n u s # $ % 1 2 2  4 *  6      D   6       $ %  7 3 ! * * 8 9 3 4 = =  8 =  ; < ! : 2 3 *  3  C . st ej n eg er ia n u s # $ % 1 2 2  4 -  6      D   6       $ %  7 3 ! * * 8 9 3 4 = =  8 =  ; < ! : 2 3 -  !  C . st ej n eg er ia n u s " &  + 1 2 - 3 : 4  %  B   D   6       $ %  7 3 - : 8 9 3 4  2 = 8  2  ; < ! : 2 3 3  !  C . st ej n eg er ia n u s " &  + 1 2 - 3 : =  %  B   D   6       $ %  7 3 - : 8 9 3 4  2 = 8  2  ; < ! : 2 3 !   :  C . st ej n eg er ia n u s # $ % 1  * 2 - *  ,  G  D   6       $ %  7 3 * = 3 ? 9 3 4 * 2 ? 2 : :  ; < ! : 2 ! :     C . st ej n eg er ia n u s # $ % 1 2 : = =  6       6       $ %  7 3 = : - ? 9 3 4 4 4 * ?  : :  ; < ! : 2 !    2  C . st ej n eg er ia n u s # $ % 1 2  = ! =  .  +       6       $ %  7 3 = : 8 9 3 4   ! 8  :  ; < ! : 2 ! 2   ,  C . p er si m il is  # $ % 1 2 2 2    <    6     $     $ %  7 ! - 3 = ? 9 3 4 - 2 ?  : :  ; < ! : 2 ! 4   ,  C . p er si m il is  # $ % 1 2 2 2  2  <    6     $     $ %  7 ! - 3 = ? 9 3 4 - 2 ?  : :  ; < ! : 2 ! =   ,  C . b ra n sf o rd ii  # $ % 1 2 2 2 * !  )  &    , G   , G   $ %  7  : 4  2 ? 9 3 =  - 3 ? = *  ; < ! : 2 !   ,  C . b ra n sf o rd ii  # $ % 1 2 2 2 - :  )  &    , G   , G   $ %  7  : 4  2 ? 9 3 =  - 3 ? = *  ; < ! : 2 ! *   ,  C . u n d er w o o d i # $ % 1 2 2  = -  $   G  $     )  C @  $ %  7  : : 2 * ? 9 3 4 ! = ?  - : :  ; < ! : 2 ! -   ,  C . u n d er w o o d i # $ % 1 2 : 2 : 4  "   &    H  B   H  B   $ %  7 ! 3  4 ? 9 3 4  * 3 ?  2 : :  ; < ! : 2 ! 3   ,  C . p o d ic if er u s # $ % 1 2 2 2 :   .  .        )  C @  $ %  7 ! -  4 ? 9 3 4 ! = ! ? 2 = : :  ; < ! : 2 ! !   ,  C . la u ra st er  # )  & 1 ! 4 ! 4       % (  <  I F          +   7  = ! 2 3 8 9 3 =  4 = 8  :  ; # 4 2 4 4 * =   ,  C . la u ra st er  .  ) 1  : - *   )     ,    D    +   7  = ! 4 8 9 3 *  = 4 8  4  -  ; # 4 2 4 4 *  132 Zootaxa 4132 (3) © 2016 Magnolia Press · 351NEW SPECIES OF CRAUGASTOR FROM COSTA RICA                                                            H                                      "        $                       .      / 0              / # $ % 0  "   $  5  $     6       $ %  7 3 - 3 ! 8 9 3 2 ! 3 8  2 : :  4 =  3 * = =  3 * =   2 : - J  2  2   2  =   2 2 = J  2 2 *   2 ! 4   4 2 4 -   4 2 = : J  4 2 = 2   = * -  J  = * - 4   3 4 :   3 4 2 J  3 4 =   3 4  J  2  3 * 4 J 2  3 *  2  3 * -  2  3 - *    % (     $     6      $ %  7 3 3 - 3 ? 9 3 2 3 * - ?   : :    3 * 3 =         ,     6       $ %  7 ! 4 4 * ? 9 3 4 !  ! ? 4 :  2   = 4 *   = * !         ,     6       $ %  7 !  * - ? 9 3 4 3 - 2 ?  : :  *   = 4 : *   = - : *   = - 4 =   ! 3 :   ! 3 *   *  * 4   <   6     6 @ A '   B   )  C @  $ %  7 ! 4 2 3 ? 9 3 4 3 = ?  : :  *  :  :  :     = - 4 -   * 4 4 :   * 4 4 =   * 4 4 *   %    6 @ A '   B   )  C @  $ %  7 ! = : = ? 9 3 4 * * ! ? 3 -  -   * 2 - 3 J  * 2 3 :  2 2  2 -  2 2  2 3  2 2  4   2 2  4 2    )       6 @ A '   B   )  C @  $ %  7 ! 4 - 4 ? 9 3 4 - : 4 ? - : :   : ! - J  :     <     6 @ A '   B   )  C @  $ %  7 ! 4 :  ? 9 3 4 - -  ? - = :  *   * 4 4  J  * 4 4 4  2 2  : :  2 2  :   2 2  : 4   )  C     D   6      $ %  7 ! 2  2 ? 9 3 4 - = ? = : :  3   ! 4 4   ! 2 - J  ! 2 !   ! = * J  ! = 3   ! 3 !  G #   D   6       $ %  7 !  - - ? 9 3 4 - ! ?  :   :   ! 2 - 2   ! 2 - -   ! 2 - 3   ! 2 3 2 J  ! 2 3 =   ! 4 2 = J  ! 4 2 *   ! 4 2 3    C      6    6 @ A '   B   )  C @  $ %  7 ! 2 - 3 ? 9 3 4 * 4 ! ? * :  *  = 2 4 = J = 2 4 !   & G    6 @ A '    B   )  C @  $ %  7 ! 2 2 ! ? 9 3 4 * = 2 ? 3 : :  4   ! !    ! ! 4   ! ! =    6           ,    6      $ % 7 ! : 3 * ? 9 3 4    ? ! : :   2 : 4 : - J 2 : 4 : !  2 : 4 = *  2 : 4 2    $  B   D   6       $ %  7 ! : 4 * 8 9 3 4  4 ! 8 - : :  4   = 2 4 :   = 2 4   = 2 = =   6      D   6       $ %  7 3 ! * * 8 9 3 4 = =  8 =  *  - 3 *  - 3 * *  ! : -  2 : 3 3  2 2  4 *  2 2  4 -    %  B   D   6       $ %  7 3 - : 8 9 3 4  2 = 8  2   2  - ! !  3 : :   : 3  J  : 3 =    3  = - *  3 - 4 4    4 :    4     * 2   >  ,  G  D   6       $ %  7 3 * = 3 ? 9 3 4 * 2 ? 2 : :  4   * 2 -   * 2 - *   * 4 =   ,  G  D   6       $ %  7 3 * 2 2 ? 9 3 4 * 2 3 8  : :  -   4 * 3 4   4 * 3 =   * -  J  * -   )  6    D   6       $ %  7 3 * 2 = 8 9 3 4 - 4 - 8  : :    * = : :      $ 5  D   6       $ %  7 3 - : = 8 9 3 4 3 - : 8 2 :   :  4 3 J 4 * :    4 : !   4 * - 4   4 * - =   * 4 = * J  * 4 = !   )        6       $ %  7 3 = 3  8 9 3 4  ! = 8  : :   :    4 : 2    4 :    4 : *    4 : 3    4   J   4  4   * 2 - -   * 4 = :   * 4 =    6       6      $ %  7 3 = : - ? 9 3 4 4 4 * ?  : :   2   3 4 4 !   3 4 = :   3 4 * J  3 4 !   3 4 *  J  3 4 * =   ! 2 4 :  2 : = =  I     G      6       $ %  7 3 * * 8 9 3 4  3  8 4  2 4    ! * !    ! - :   2 2 = *   2 2 - 2 J  ! 2 3 !   = ! 3   = ! !  K  .  +       6      $ %  7 3 = : 8 9 3 4   ! 8  :    2  = ! =    6        6       $ %  7 3 2 =  8 9 3 2 ! = * 8  : :  4   2 -  - J  2 -  !   133 ARIAS ET AL.352 · Zootaxa 4132 (3) © 2016 Magnolia Press Dorsal color pattern. We followed Savage & Emerson (1970) in defining the dorsal color pattern polymorphisms as follows: M = mottled, U = uniform, L = striped, Z 1 = tympanic stripe, Z 2 = tympanic stripe in combination with eye mask, N = skin dorsum smooth, O = skin dorsum irregularly granular, and Q = a pair of dorsal ridges. Statistical analyses. We calculated the mean, standard deviation, and range for each morphometric variable using the program JMP 12 (SAS Institute). We compared the populations using ANOVA (α = 0.05) when the assumptions of normality (Shapiro-Wilk, α = 0.05) and homoscedasticity (Levene, α = 0.05) were met. When the assumption of normality failed, we used the Kruskal-Wallis test (α = 0.01); when there was heteroscedasticity, we used Welch’s ANOVA (α = 0.01). The following variables were subject to principal components analyses (PCA) with subsequent ANOVA of the PC1 and PC2 scores: HL/SVL, TL/SVL, E-N/SVL, F2/SVL, and IOD/HW. The same variables were used in a discriminant analysis to calculate the probability of the correct classification of the specimens to their original population. TABLE 3. Results from the morphological analyses. Significant P-values are indicated in bold. Abbreviations: S.D. = standard deviation, P = probability. The statistical test used is indicated in the last column as follows: † = ANOVA, †† = Welch’s ANOVA, § = Kruskal-Wallis test. Variable Craugastor gabbi sp. nov. Craugastor stejnegerianus P* Mean±S.D. Range Mean±S.D Range SVL 16.72±2.18 13.60–21.55 15.58±2.63 9.50–21.40 0.018† HL/SVL 0.33±0.03 0.29–0.41 0.36±0.03 0.30–0.45 <0.001§ HW/SVL 0.34±0.02 0.31–0.38 0.35±0.02 0.32–0.40 0.047§ ED/SVL 0.12±0.02 0.08–0.15 0.12±0.01 0.08–0.16 0.472§ IOD/SVL 0.11±0.01 0.10–0.13 0.11±0.01 0.08–0.17 <0.001† TY/SVL males 0.08±0.02 0.06–0.11 0.09±0.01 0.06–0.11 0.873†† TY/SVLfemales 0.06±0.01 0.05–0.08 0.06±0.01 0.03–0.08 0.113§ EW/SVL 0.08±0.01 0.06–0.11 0.08±0.01 0.05–0.10 0.273† TL/SVL 0.55±0.02 0.50–0.61 0.52±0.03 0.46–0.58 <0.001† E-N/SVL 0.09±0.02 0.04–0.11 0.10±0.01 0.06–0.12 <0.001§ T1/SVL 0.10±0.02 0.06–0.14 0.10±0.01 0.07–0.14 0.590†† T2/SVL 0.19±0.03 0.14–0.24 0.19±0.02 0.12–0.24 0.398† T3/SVL 0.32±0.03 0.26–0.41 0.31±0.03 0.25–0.38 0.433†† T4/SVL 0.48±0.04 0.37–0.55 0.47±0.04 0.37–0.54 0.228§ T5/SVL 0.30±0.03 0.23–0.35 0.29±0.03 0.22–0.34 0.226§ F1/SVL 0.10±0.02 0.08–0.14 0.11±0.01 0.07–0.15 0.006§ F2/SVL 0.13±0.02 0.11–0.17 0.14±0.02 0.09–0.18 0.001† F3/SVL 0.21±0.02 0.16–0.25 0.21±0.02 0.16–0.26 0.148† F4/SVL 0.14±0.02 0.10–0.19 0.15±0.02 0.10–0.20 0.279† EW/IOD 0.67±0.12 0.50–1.03 0.74±0.13 0.41–1.31 0.001§ TY/ED males 0.76±0.14 0.60–1.11 0.74±0.09 0.60–0.92 0.818§ TY/ED females 0.52±0.06 0.39–0.59 0.49±0.05 0.31–0.59 0.090† E-N/ED 0.74±0.16 0.36–1.05 0.81±0.12 0.50–1.20 0.002†† ED/HL 0.35±0.05 0.25–0.49 0.34±0.03 0.24–0.41 0.121§ IOD/HW 0.33±0.02 0.28–0.40 0.31±0.04 0.22–0.48 <0.001§ T4/TL 0.86±0.06 0.69–1.01 0.90±0.06 0.74–1.02 <0.001§ 134 Zootaxa 4132 (3) © 2016 Magnolia Press · 353NEW SPECIES OF CRAUGASTOR FROM COSTA RICA FIGURE 2. Bayesian phylogenetic inference of the relationships of Craugastor gabbi sp. nov. within the C. podiciferus species group based on the 16S mitochondrial DNA gene fragment. Bayesian posterior probabilities (multiplied by 100) are shown above the branch; maximum likelihood bootstrap values from the RAxML analysis are shown below the branches. The scale bar refers to the estimated substitutions per site. The support values of any node within species are not shown. Results Molecular analyses. The resulting data matrix had a total length of 539 bp, including gaps. The phylogenies inferred using ML and BA were concordant in supporting the tree shown in Fig. 2. The phylogeny shows two well- supported clades. One is formed of all the samples from the premontane forest of the South Pacific Costa Rica and western Panama (populations “a” and “b”, Fig. 1), and the second is formed of all the samples from lowland South 135 ARIAS ET AL.354 · Zootaxa 4132 (3) © 2016 Magnolia Press Pacific Costa Rica (populations “c-y”, Fig. 1). The genetic divergence between these two clades was 4.5% for the 16S gene. These two clades are nonetheless not sister taxa as Craugastor stejnegerianus + C. lauraster form a clade separated from C. gabbi and show mean-corrected genetic divergence of 4.7%, supporting that C. stejnegerianus, is more closely related to C. lauraster than to C. gabbi. Our results highlight the independent evolutionary status of the populations from the premontane forest and lead us to recognize it as a separate evolutionary unit: Craugastor gabbi sp. nov. Morphometric analyses. Morphometric variation and comparisons among the populations are shown in Table 3. The following characteristics were significantly different between the lowland and premontane populations: SVL, HL/SVL, IOD/SVL, TL/SVL, E-N/SVL, F1/SVL, F2/SVL, EW/IOD, E-N/ED, IOD/HW, and T4/TL. The PCA efficiently differentiated the premontane and lowland frogs (Fig. 3). The first principal component (PC1) explained 40.6% of the total variance, with 68.0% of the total variance explained by the first two components. In PC2, TL/SVL and IOD/HW were positively loaded, while HL/SVL, E-N/SVL, and F2/SVL were negatively loaded. PC1 and PC2 differentiated the two populations (PC1: F=12.03, df=189, p=0.001; PC2: F=114.51, df=189, p=<0.001). The discriminant analysis correctly classified 91.55% of the specimens to the populations, demonstrating a clear separation between the lowland and premontane populations studied here (Wilks’ lambda 0.512, p=<0.001). FIGURE 3. Principal component analysis showing the morphological separation among the 35 individuals of Craugastor gabbi sp. nov. (solid squares) from the premontane forest near the Costa Rican-Panamanian border and 155 individuals of C. stejnegerianus (open triangles) from the lowlands of South Pacific Costa Rica. Description of new species Craugastor gabbi sp. nov. Gabb's Dirt Frog (Figures 4–6) Craugastor stejnegerianus (part): Scott 1976; Savage 2002; Crawford 2003; Santos-Barrera et al. 2008 136 Zootaxa 4132 (3) © 2016 Magnolia Press · 355NEW SPECIES OF CRAUGASTOR FROM COSTA RICA Holotype. UCR 21864, an adult female from the Organization for Tropical Studies’ Las Cruces Biological Station (+8.7889º, -82.9583º; 1200 m elevation), Fila Costeña, San Vito de Coto Brus, Puntarenas Province, Costa Rica; collected by Erick Arias, Gerardo Chaves, Adrián García-Rodríguez, and Federico Bolaños on 16 March 2013. Paratypes. Adult male UCR 21865 and UCR 21867; adult female UCR 21863 and UCR 21876; same data as the holotype. Assignment to group. The following inclusion characteristics were used to assign frogs to the C. podiciferus species group: narrow head (31–38% SVL), presence of venter areolate, a shorter Finger I than Finger II, absence of inner tarsal fold, and absence of nuptial pads. Diagnosis. A small species of the Craugastor (Craugastor) podiciferus species group with the following characteristics: (1) skin on the dorsum is shagreen with scattered enlarged granules; venter and flanks are coarsely areolate, usually with dorsolateral and lateral folds; discoidal fold complete laterally and posteriorly; (2) tympanum round, usually with membrane differentiated and annulus prominent, (TY/ED = 39–111%), usually without supratympanic fold; (3) snout subovoid in dorsal view, rounded in profile; loreal region concave; canthus-rostralis usually rounded; (4) upper eyelid granular (EW/IOD = 50–103%); cranial crests absent; (5) vomerine teeth transverse, in two fascicles well behind the choanae; choanae smaller than the dentigerous; (6) vocal slits absent; nuptial pads absent; (7) Finger I slightly shorter than Finger II; disks usually absent, but with terminal transverse grooves; tips usually lanceolate; pads triangular; (8) fingers lacking lateral fringes; thenar and palmar tubercles ovoid, thenar much smaller than palmar; proximal supernumerary tubercles rounded; usually two accessory palmar tubercles; subarticular tubercles round, projecting and obtuse; (9) ulnar fold absent, but tubercles sometimes are visible; (10) heel lacking tubercles; inner tarsal folds usually absent; (11) inner metatarsal tubercle elongate, outer rounded, much smaller than inner; proximal supernumerary tubercles rounded; numerous (usually 10–15) rounded plantar tubercles; subarticular tubercles ovoid, projecting and obtuse; (12) Toe III larger than Toe V; disks expanded, asymmetric; disk cover usually lanceolate; disk pad triangular; toe webbing basal, usually does not reach the proximal subarticular tubercle on Toes I-II-III-IV; however, in some specimens webbing reaches the proximal subarticular tubercle or beyond; (13) dorsum gray brown to dark brown, uniform, mottled, some specimens have a middorsal light stripe, or, rarely, paired dorsolateral light stripes; venter cream with dark pigment reaching the midline; throat usually yellowish with dark mottling; upper surfaces of thighs usually with dark bars; posterior surface of thigh uniform reddish brown; usually labial bars and supratympanic mark (Fig. 4); (14) SVL in males 14.35–21.35 mm; SVL in females 13.60–21.55 mm. Comparisons with other species. Craugastor gabbi differs from all the other craugastorids of Lower Central America, except for those in the C. podiciferus species group, which have basal webbing between the toes and a narrow head, i.e., head width 31–38% SVL. Craugastor gabbi differs from other members of the C. podiciferus species group by having the following characteristics (condition for C. gabbi in parentheses, see Table 4). Craugastor gabbi differs from C. bransfordii, C. polyptychus and C. underwoodi by having a thenar tubercle equal to or slightly smaller than the palmar tubercle (thenar tubercle definitely much smaller than palmar tubercle, Fig. 5); C. gabbi differs from C. podiciferus by having a prominent calcar tubercle on the heel (calcar tubercle absent) and by the absence of supernumerary tubercles (supernumerary tubercles well defined in fingers and toes); from C. jota (Lynch 1980) by having a prominent calcar tubercle on the heel (calcar tubercle absent); from C. lauraster by having an immaculate white venter (having a cream-colored venter with dark pigment reaching the midline, Fig. 4) and by having a much shorter Finger I than Finger II (Finger I slightly shorter than Finger II); from C. persimilis by having toes unwebbed (basal webbing between toes), by having dorsum and forelimbs areolate (dorsum and forelimbs shagreen), and by having a much shorter Finger I than Finger II (Finger I slightly shorter than Finger II); from C. stejnegerianus by having an immaculate white venter, [only the 8% of the specimens reviewed here had a venter with dark pigment reaching the midline] (venter cream-colored with dark pigment reaching the midline, Fig. 6) and by having a significantly larger TL/SVL and IOD/HW and a significantly smaller HL/SVL and E-N/SVL. Is important to note that the specimens of the northernmost populations of the Central Pacific of Costa Rica had the venter with dark pigment reaching the midline; however, these populations are under taxonomic revision because they could represent a separate species. Description of holotype. Adult female; head width 36.5% of SVL; head length 38.6% of SVL; snout subovoid in dorsal view, rounded in profile; canthus-rostralis indistinct; loreal region slightly concave; nostrils small, directed laterally; vomerine teeth transverse, in two fascicles well behind the choanae; eye larger, diameter equal to 126.67% of E–N; tympanum small, 57.69% of ED, round, with membrane undifferentiated and annulus 137 ARIAS ET AL.356 · Zootaxa 4132 (3) © 2016 Magnolia Press prominent; skin on dorsum shagreen, venter coarsely areolate, throat and head smooth, flanks shagreen like the dorsum; a pair of dorsolateral folds extend from the orbit to the sacrum, and a pair of lateral folds extend from the axillar level to the sacrum; discoidal fold complete; upper eyelid granular; postrictal tubercles fused forming a short ridge posterior-ventral to the tympanum. Forelimb slim; ulnar tubercles and fold absent; thenar and palmar tubercles ovoid, with the thenar much smaller than palmar; proximal supernumerary tubercles rounded, medial supernumerary tubercles only on Finger III; three accessory palmar tubercles much smaller than the supernumerary tubercles; subarticular tubercles larger, round, projecting and obtuse; fingers slim; disks absent; fingers with grooves; tips of fingers lanceolate in dorsal view; pads triangular; fingers not webbed. Hindlimb slightly slim; heel smooth, inner tarsal fold absent; inner metatarsal tubercle elongate, outer rounded, much smaller than inner; proximal supernumerary tubercles rounded, medial supernumerary tubercles only on Toes III and IV; numerous rounded plantar tubercles; subarticular tubercles rounded, projecting and pungent; disks expanded, asymmetric; disk cover lanceolate; disk pad triangular; webbing basal between Toes I-II-III-IV, reaching the proximal subarticular tubercle on Toe I. FIGURE 4. a) Dorsal and b) ventral photos of Craugastor gabbi sp. nov. holotype (UCR 21864). The scale bar represents 0.5 cm. Photos by Mauricio Calderón-Rivera. Coloration of the holotype in ethanol (Fig. 4). Dorsum of head and back dark brown to grayish with a pair of dark spots on each flank, a pair of dark spots on the back, an oblique dark stripe crossing the sacrum, and a middorsal light stripe; upper lip with diffuse dark bars; a dark supratympanic stripe extending from the orbit to the suprascapular shoulder; venter cream-colored and dotted with dark pigment; throat cream with dark mottling, not contrasting with the venter; groin cream, with a few small dark dots, not contrasting with the flanks; flanks light brown but transition dorsally to dark brown and ventrally to cream with dark dots, which penetrate the venter; dorsal surface of hind limbs similar to dorsal background with dark bars, which extend over the tibia and feet; posterior and anterior surfaces of hind limbs uniform dark brown. Measurements of holotype (mm). SVL 19.7; HL 7.6; HW 7.2; ED 2.6; IOD 2.3; TY 1.5; EW 1.8; TL 10.4; E–N 2.1; T1 1.7; T2 3.7; T3 5.1; T4 9.50; T5 6.1; F1 2.4; F2 3.0; F3 4.2; F4 3.20. Morphometric (mm) and morphological variation of paratypes. Morphometric variation of all specimens analyzed is summarized in Table 3. Here we provide the mean and standard deviation and, in parentheses, the range of each measurement of all the paratypes. SVL 16.9±2.0 (14.8–19.3); HL 6.7±0.81 (6.0–7.8); HW 6.2±0.8 (5.4–7.1); ED 2.3±0.3 (1.9–2.5); IOD 2.1±0.3 (1.8–2.5); TY in males 1.6, in females 1.4±0.2 (1.2–1.4); EW 1.3±0.3 (1.1–1.7); TL 9.3±0.9 (8.4–10.4); E-N 1.7±0.3 (1.5–2.1); T1 1.4±0.3 (1.1–1.7); T2 2.3±0.4 (2.7–3.6); T3 138 Zootaxa 4132 (3) © 2016 Magnolia Press · 357NEW SPECIES OF CRAUGASTOR FROM COSTA RICA 5.1±0.6 (4.5–5.9); T4 7.9±0.6 (7.2–8.4); T5 4.9±0.4 (4.6–5.4); F1 2.0±0.2 (1.8–2.3); F2 2.2±0.4 (1.8–2.6); F3 3.6±0.4 (3.0–4.1); F4 2.4±0.5 (1.9–3.0). The scant variation among the paratypes is described as follows. Tympanic membrane differentiated in UCR 21863, UCR 21865, and UCR 21876; a pair of dorsolateral folds extending from the axillar level to the sacrum in UCR 21865 and UCR 21867; a pair of lateral folds extending from the orbit to the sacrum in UCR 21865; discoidal fold indistinguishable in UCR 21863, UCR 21865, UCR 21867 and UCR 21876; two tubercles postrictal in UCR 21867, semifused in UCR 21863 and UCR 21865; inner tarsal fold incomplete in UCR 21863. Fingers without medial supernumerary tubercles in UCR 21863 and UCR 21865; only two palmar accessory tubercles in UCR 21863, UCR 21865, and UCR 21876; toes without medial supernumerary tubercles in UCR 21863 and UCR 21865; dorsum with a pair of broad lateral light stripes extending from the snout to the groin, these stripes are divided by a pair of dark stripes extending from the axillar level to the groin, forming a stripe pattern in UCR 21876, with a pair of lateral light stripes extending from the orbit to the groin in UCR 21865; labial marks absent in UCR 21863; supratympanic mark absent in UCR 21876. Variation in other specimens not observed in paratypes. Some specimens with the snout subelliptical in dorsal view; canthus-rostralis rounded in some; others with vomerine teeth obtuse; some specimens with a pair of dorsolateral folds extending from the orbit to the anus; lateral folds variable to absent, from the orbit to groin, from the orbit to anus, from the axillary level to sacrum and from the axillary level to groin; supratympanic fold distinctly curved downwards present in several specimens; some with only one postrictal tubercle; a specimen with inner tarsal fold complete; the inguinal gland was elevated and prominent in some individuals, in others this was not evident, although we did not observe a relation between this characteristic and sex. Some specimens with four or five ulnar tubercles in a row; some without grooves in fingers; at least one specimen with disk expanded in Fingers III and IV; webbing between Toes I-II can reach proximal subarticular tubercle, the web between Toes II- III-IV can reach least half of the proximal phalanx. Patterns variation. In comparison with the other species of the C. podiciferus species group, C. gabbi had low levels of dorsal color pattern variation. The holotype had the formula MZ 1 OQ (mottled dorsal pattern with stripe supratympanic, skin texture irregularly granular and a pair of dorsal ridges); this formula was present in 25% of the specimens. The most common formula was UZ 1 OQ at 34.4%, described by Savage & Emerson (1970) as Morpho V. The following formulae were also observed: LZ 1 OQ present in 9.4%; UZ 1 NQ, MOQ, and MO present in 6.3%; UZ 2 OQ, LZ1NQ, LOQ, and MZ 1 NQ present in 3.1%. FIGURE 5. Comparison of the thenar and palmar tubercles. a) Craugastor gabbi sp. nov. holotype (UCR 21864) showing smaller size of thenar tubercle compared to the palmar tubercle and b) C. bransfordii (CRARC0177) of Turrialba, Costa Rica showing thenar tubercle to be equal to or slightly smaller than the palmar tubercle. This condition is also seen in C. underwoodi and C. polyptychus. Photo A by E.A and B by Brian Kubicki. 139 ARIAS ET AL.358 · Zootaxa 4132 (3) © 2016 Magnolia Press FIGURE 6. Ventral coloration. a) Craugastor gabbi sp. nov. paratype (UCR 21876), with the dark pigment reaching the midline and b) C. stejnegerianus (UCR 20885) of Palmar Norte, Costa Rica showing immaculate venter. Photos by Mauricio Calderón-Rivera. The dorsal background color ranged from light gray to blackish brown; at least one individual presented longitudinal lines from the orbit to groin, light color on a dark background, and another two specimens presented dark lines on a light background. In eight individuals, the pattern of contrasting mottles was strongly prominent. The throat color varied from cream to dark brown, strongly contrasting with the lighter-colored venter. No association between the throat color and sex was observed. At least one individual had a well-defined mask, and the head of another was completely black. The supratympanic dark mark was present in most, but not all, individuals; labial bars were also present in most, but not all, specimens. Natural history notes. Craugastor gabbi is abundant in the Fila Costeña, Coto Brus region. Santos-Barrera et al. (2008) reported this species at 21 of the 27 study sites, stating that C. gabbi was the most abundant amphibian in the forest and the coffee plantations and was also present in pastures. Usually, this species is associated with leaf litter. This species may reach densities of up 4586 individuals/ha (Scott 1976). Craugastor gabbi is diurnal (Savage 2002) and reproductively active throughout the year (Santos-Barrera et al. 2008). The type series was found together in a patch of primary forest, active at approximately 12:00 h on the leaf litter and away from any body of water. No reproductive activity was observed, and no calls were recorded. The call is unknown, but it is likely an inconspicuous single slow squeak, as reported for C. stejnegerianus (Savage 2002). We did not find gravid females, and pairs in amplexus were not observed. Geographic distribution. Craugastor gabbi is restricted to the premontane forest near the type locality of Fila Costeña, Costa Rica and the premontane forest of Cordillera de Talamanca in the extreme southwestern region of Costa Rica and western Panama near the present international border (Fig. 1). The altitudinal range of the new species is 1100–1280 m elevation. Craugastor gabbi overlaps with the range of C. underwoodi and C. podiciferus at the Las Cruces Biological Station, Fila Costeña, Coto Brus. Etymology. This species is named in honor of paleontologist William M. Gabb in recognition of his important contribution to the herpetology of Costa Rica as an explorer and collector, mainly in the Talamanca region. 140 Zootaxa 4132 (3) © 2016 Magnolia Press · 359NEW SPECIES OF CRAUGASTOR FROM COSTA RICA        &                                        K                      C ra u g a st o r p o d ic if e ru s          )     <                A      "        "       A      )            )              +        E        K           E     E      C . g a b b i    <                          "                   "      6  G     6     ,     ,   ,     ,     ,     C . st ej n eg er ia n u s <                          "      >                 "      6  G     6     ,     ,   ,     ,     ,     C . p er si m il is  <                          "                      "      6  G     6     ,     ,   ,     ,     ,     C . la u ra st er  <                          "                      "      6  G     6     ,     ,   ,     ,     ,     C . b ra n sf o rd ii  <         >                       "      >                 "      6     6     ,     ,   6     ,     ,     C . u n d er w o o d i <         >                       "      >                 "      6  G        I  ,     ,     ,   6     ,     ,     C . p o ly p ty ch u s <         >                       "      >                 "      6  G             6     ,     ,   ,     ,     ,     C . p o d ic if er u s <                          "      >                 "      "     I  ,     6     )   ,     6     6     141 ARIAS ET AL.360 · Zootaxa 4132 (3) © 2016 Magnolia Press Discussion In previous studies, molecular (Crawford 2003) and cytogenetic analyses (Chen 2001) of Craugastor stejnegerianus from the premontane forest of the Fila Costeña and the lowlands South Pacific of Costa Rica near Panama showed large genetic distances. Our taxonomic re-assessment using morphological and molecular data led us to elevate to the species level populations of the premontane forest as distinct from the lowlands populations. We found that C. gabbi differ from typical C. stejnegerianus by a combination of morphological traits as determined using discriminant function analyses, differences in ventral coloration and genetic differentiation at the 16S gen. Following the criteria of Pérez-Ponce de León & Nadler (2010), C. gabbi is not a cryptic species sensu stricto, although in the past this species was masked under C. stejnegerianus due their high degree of morphological similarity (functional definition of cryptic species). Our molecular analyses rejected the null hypothesis that C. stejnegerianus represent a single species, and our detailed morphological analyses showed morphological differences between both species, allowing a proper morphological diagnosis and species description. The 16S gene fragment used here has been suggested as a DNA barcode marker for amphibian diversity inventories (Vences et al. 2005) to complement the standard COI-5’ marker used in general for animals (Smith et al. 2008). The uncorrected genetic distance of 4.5% at the 16S gene observed between the premontane populations now assigned to C. gabbi and the lowlands South Pacific populations of C. stejnegerianus, is greater than the suggested threshold of 3% for flagging the potential cryptic species of Neotropical frogs (Fouquet et al. 2007). It is also substantially higher than the 1.8% genetic distance at 16S found by Hertz et al. (2012) in separating a recently described new species of arboreal Terrarana from Panama. Thus, our single mtDNA marker offers additional support for the specific status of the new species. Furthermore, based on genetic divergence at the mitochondrial ND2 gene, Crawford (2003) estimated that the Las Cruces Biological Station (Fila Costeña) population (now C. gabbi) diverged from the lowland populations (Palmar Norte and Rincón de Osa) of C. stejnegerianus between 6.25 and 10.3 million years ago (Ma), demonstrating the long-term evolutionary independence of these two lineages, which were formerly regarded as conspecific. Although the Fila Costeña and Palmar Norte localities are only 25 km apart (Fig. 1), these populations are separated by at least 1100 m in elevation. Based on the genetic data, their estimated divergence time corresponds perfectly with the rise of the Fila Costeña at 12.80–11.67 Ma (MacMillan et al. 2004). Our findings, therefore, agree with Bickford et al. (2007) in rejecting the common assumption that cryptic species are recently diverged and remind us that morphological homoplasy presents a challenge to amphibian taxonomy (Wake 2009). We believe that the separation of C. gabbi from C. stejnegerianus is the beginning of the resolution of this group of highly similar frogs, treated as “cryptic species” in the taxonomic practice and in agreement with the functional definition of cryptic species by Pérez-Ponce de León & Nadler (2010). Craugastor stejnegerianus is distributed along the Pacific coast of Costa Rica from the Panamanian border northward to the Cordillera de Tilarán (Crawford 2003). The Cordillera de Tilarán populations could potentially represent yet one more undescribed species distinct from C. stejnegerianus (Crawford 2003; J. Savage, pers. comm.), which could eventually leave C. stejnegerianus restricted to the Pacific lowlands, roughly south of the City of Puntarenas. Additional taxonomic work is needed to complete the understanding of these diverse and locally abundant leaf- litter frogs. Acknowledgments This study will be submitted by EA in partial fulfillment of the requirements to obtain the degree of Doctor en Ciencias of the Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México. EA thanks the Posgrado en Ciencias Biológicas for its support of this study, and the CONACyT for students grant (CVU/Becario) 626946/330343. Laboratory efforts were partially funded by grant from PAPIIT-UNAM (209914) to GP-O. We thank Laura Márquez-Valdelamar and Andrea Jiménez-Marín for their laboratory assistance; Javier Guevara from the Ministerio de Ambiente y Energía (Costa Rica) who provided collecting permits; Andreas Hertz and Randy McCranie who kindly provided tissue samples from C. gabbi and C. lauraster; Federico Bolaños for the use of specimens from the Museo de Zoología of the Universidad de Costa Rica; Brian Kubicki and Maurico Calderón- Rivera who kindly provided the photographs; the Laboratorio de Autómatas y Sistemas Inteligentes en 142 Zootaxa 4132 (3) © 2016 Magnolia Press · 361NEW SPECIES OF CRAUGASTOR FROM COSTA RICA Biodiversidad (LASIB) from the Universidad de Costa Rica for access to the equipment used to prepare the photographs of the specimens. In Costa Rica, the specimens were collected according to a permit granted by SINAC to EA (007-2013-SINAC) and GC (SINAC-SE-GASP-PI-R-059-2015). References AmphibiaWeb (2016) AmphibiaWeb: Information on Amphibian Biology and Conservation [web application]. Berkeley, California. Available from: http://amphibiaweb.org (accessed 11 January 2016) Barbour, T. (1926) New Amphibia. Occasional Papers of the Boston Society of Natural History, 5, 191–194. Bickford, D., Lohman, D.J., Sodhi, N.S., Ng, P.K.L., Meier, R., Winker, K., Ingram, K.K. & Das, I. (2007) Cryptic species as a window on diversity and conservation. Trends in Ecology & Evolution, 22, 148–155. http://dx.doi.org/10.1016/j.tree.2006.11.004 Boulenger, G.A. (1896) Descriptions of new batrachians collected by Mr. C. G. Underwood in Costa Rica. Annals and Magazine of Natural History, Series 6, 18, 340–342. http://dx.doi.org/10.1080/00222939608680464 Chen, S.H. (2001) Cytogenetic study of the lower Central American frogs of the subgenus Craugastor (Anura: Leptodactylidae: Eleutherodactylus). Ph.D. dissertation, University of Miami, Florida, 369 pp. Chen, S.H. (2005) Chromosomal variation in the rhodopis group of the southern Central American Eleutherodactyline frogs (Leptodactylidae: Eleutherodactylus). In: Donnelly, M.A., Crother, B.I., Guyer, C., Wake, M.H. & White, M.E. (Eds.), Ecology and evolution in the tropics: a herpetological perspective. The University of Chicago Press, Illinois, pp. 102–144. Cope, E.D. (1886) Thirteenth contribution to the herpetology of tropical America. Proceedings of the American Philosophical Society, 23, 271–287. Cope, E.D. (1893) Second addition to the knowledge of the Batrachia and Reptilia of Costa Rica. Proceedings of the American Philosophical Society, 31, 333–347. Crawford, A.J. (2003) Huge populations and old species of Costa Rican and Panamanian dirt frogs inferred from mitochondrial and nuclear gene sequences. Molecular Ecology, 12, 2525–2540. http://dx.doi.org/10.1046/j.1365-294X.2003.01910.x Crawford, A.J. & Smith, E.N. (2005) Cenozoic biogeography and evolution in direct-developing frogs of Central America (Leptodactylidae: Eleutherodactylus) as inferred from a phylogenetic analysis of nuclear and mitochondrial genes. Molecular Phylogenetics and Evolution, 35, 536–555. http://dx.doi.org/10.1016/j.ympev.2005.03.006 Crawford, A.J., Bermingham, E. & Polanía, C. (2007) The role of tropical dry forest as a long-term barrier to dispersal: a comparative phylogeographical analysis of dry forest tolerant and intolerant frogs. Molecular Ecology, 16, 4789–4807. http://dx.doi.org/10.1111/j.1365-294X.2007.03524.x de Queiroz, K. (2007) Species concepts and species delimitation. Systematic Biology, 56, 879–886. http://dx.doi.org/10.1080/10635150701701083 Duellman, W.E. & Lehr, E. (2009) Terrestrial-breeding frogs (Strabomantidae) in Peru. Natur- und Tier-Verlag, Münster, 382 pp. Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research, 32, 1792–1797. http://dx.doi.org/10.1093/nar/gkh340 Fouquet, A., Gilles, A., Vences, M., Marty, C., Blanc, M. & Gemmell, N.J. (2007) Underestimation of species richness in Neotropical frogs revealed by mtDNA analyses. PLoS ONE, 2, e1109. http://dx.doi.org/10.1371/journal.pone.0001109 Frost, D.R., Grant, T., Faivovich, J., Bain, R.H., Haas, A., Haddad, C.F.B., De Sá, R.O., Channing, A., Wilkinson, M., Donnellan, S.C., Raxworthy, C.J., Campbell, J.A., Blotto, B.L., Moler, P., Drewes, R.C., Nussbaum, R.A., Lynch, J.D., Green, D.M. & Wheeler, W.C. (2006) The amphibian tree of life. Bulletin of the American Museum of Natural History, 297, 1–370. http://dx.doi.org/10.1206/0003-0090(2006)297[0001:TATOL]2.0.CO;2 Günther, A.C.L.G. (1885–1902) Biologia Centrali-Americana. Reptilia and Batrachia. Porter, London, 326 pp. Hedges, S.B., Duellman, W.E. & Heinicke, M.P. (2008) New World direct-developing frogs (Anura: Terrarana): Molecular phylogeny, classification, biogeography, and conservation. Zootaxa, 1737, 1–182. Heinicke, M.P., Duellman, W.E. & Hedges, S.B. (2007) Major Caribbean and Central American frog faunas originated by ancient oceanic dispersal. Proceedings of the National Academy of Sciences, USA, 104, 10092–10097. http://dx.doi.org/10.1073/pnas.0611051104 Heinicke, M.P., Duellman, W.E., Trueb, L., Means, D.B., MacCulloch, R.D. & Hedges, S.B. (2009) A new frog family (Anura, Terrarana) from South America and an expanded direct–developing clade revealed by molecular phylogeny. Zootaxa, 2211, 1–35. Hertz, A., Hauenschild, F., Lotzkat, S. & Köhler, G. (2012) A new golden frog species of the genus Diasporus (Amphibia, 143 ARIAS ET AL.362 · Zootaxa 4132 (3) © 2016 Magnolia Press Eleutherodactylidae) from the Cordillera Central, western Panama. ZooKeys, 196, 23–46. http://dx.doi.org/10.3897/zookeys.196.2774 Huelsenbeck, J.P. & Ronquist, F. (2001) MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics, 17, 754–755. http://dx.doi.org/10.1093/bioinformatics/17.8.754 Lynch, J.D. (1980) A new frog of the genus Eleutherodactylus from western Panama. Transactions of the Kansas Academy of Science, 83, 101–105. http://dx.doi.org/10.2307/3627802 Lynch, J.D. (1986) The definition of the Middle American clade of Eleutherodactylus based on jaw musculature (Amphibia, Leptodactylidae). Herpetologica, 42, 248–258. MacMillan, I., Gans, P.B. & Alvarado, G. (2004) Middle Miocene to present plate tectonic history of the southern Central American Volcanic Arc. Tectonophysics, 392, 325–348. http://dx.doi.org/10.1016/j.tecto.2004.04.014 McCranie, J.R. (2015) A checklist of the amphibians and reptiles of Honduras, with additions, comments on taxonomy, some recent taxonomic decisions, and areas of further studies needed. Zootaxa, 3931 (3), 352–386. http://dx.doi.org/10.11646/zootaxa.3931.3.2 Miyamoto, M.M. (1983) Biochemical variation in the frog Eleutherodactylus bransfordii: geographic patterns and cryptic species. Systematic Zoology, 32, 43‒51. http://dx.doi.org/10.2307/2413219 Nylander, J.A.A. (2004) MrModeltest v2. Program distributed by the author. Evolutionary Biology Center, Uppsala University, Uppsala, Sweden. Padial, J.M., Miralles, A., De la Riva, I. & Vences, M. (2010) The integrative future of taxonomy. Frontiers in Zoology, 7, 16. http://dx.doi.org/10.1186/1742-9994-7-16 Padial, J.M., Grant, T. & Frost, D.R. (2014) Molecular systematics of terraranas (Anura: Brachycephaloidea) with an assessment of the effects of alignment and optimality criteria. Zootaxa, 3825 (1), 1–132. http://dx.doi.org/10.11646/zootaxa.3825.1.1 Palumbi, S., Martin, A., Romano, S., McMillan, W.O., Stice, L. & Grabowski, G. (1991) The simple fool’s guide to PCR, version 2.0. Special Publication, Department of Zoology and Kewalo Marine Laboratory, University of Hawaii, Honolulu, Hawaii, USA. Pérez-Ponce de León, G & Nadler, S.A. (2010) What we don't recognize can hurt us: A plea for awareness about cryptic species. Journal of Parasitology, 96, 453–464. http://dx.doi.org/10.1645/GE-2260.1 Pyron, R.A. & Wiens, J.J. (2011) A large-scale phylogeny of Amphibia including over 2800 species, and a revised classification of extant frogs, salamanders, and caecilians. Molecular Phylogenetics and Evolution, 61, 543–583. http://dx.doi.org/10.1016/j.ympev.2011.06.012 Rambaut, A., Suchard, M.A., Xie, W. & Drummond, A.J. (2014) Tracer v1.6.1. Available from: http://beast.bio.ed.ac.uk/Tracer (accessed 13 June 2016) Santos-Barrera, G., Pacheco, J., Mendoza-Quijano, F., Bolaños, F., Chaves, G., Daily, G.C., Ehrlich, P.R. & Ceballos, G. (2008) Diversity, natural history and conservation of amphibians and reptiles from the San Vito Region, southwestern Costa Rica. Revista de Biología Tropical, 56, 755–778. Savage, J.M. (2002) The Amphibians and Reptiles of Costa Rica: A Herpetofauna between two Continents, between two Seas. The University of Chicago Press, Illinois, 934 pp. Savage, J.M. & Emerson, S.B. (1970). Central American frogs allied to Eleutherodactylus bransfordii (Cope): a problem of polymorphism. Copeia, 1970, 623–644. http://dx.doi.org/10.2307/1442305 Savage, J.M., McCranie, J.R. & Espinal, M. (1996) A new species of Eleutherodactylus from Honduras related to Eleutherodactylus bransfordii (Anura: Leptodactylidae). Proceedings of the Biological Society of Washington, 109, 366–372. Scott, N.J. Jr. (1976) The abundance and diversity of the herpetofaunas of tropical forest litter. Biotropica, 8, 41–58. http://dx.doi.org/10.2307/2387818 Smith, M.A., Poyarkov, N.A. Jr. & Hebert, P.D.N. (2008) CO1 DNA barcoding amphibians: Take the chance, meet the challenge. Molecular Ecology Resources, 8, 235–246. http://dx.doi.org/10.1111/j.1471-8286.2007.01964.x Stamatakis, A. (2014) RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics, 30, 1312–1313. http://dx.doi.org/10.1093/bioinformatics/btu033 Streicher, J.W., Crawford, A.J. & Edwards, C.W. (2009) Multilocus molecular phylogenetic analysis of the montane Craugastor podiciferus species complex (Anura: Craugastoridae) in Isthmian Central America. Molecular Phylogenetics and Evolution, 53, 620–630. http://dx.doi.org/10.1016/j.ympev.2009.07.011 Tamura, K. (1992) Estimation of the number of nucleotide substitutions when there are strong transition–transversion and G + C–content biases. Molecular Biology and Evolution, 9, 678–687. 144 Zootaxa 4132 (3) © 2016 Magnolia Press · 363NEW SPECIES OF CRAUGASTOR FROM COSTA RICA Tamura, K., Stecher, G., Peterson, D., Filipski, A. & Kumar, S. (2013) MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Molecular Biology and Evolution, 30, 2725–2729. http://dx.doi.org/10.1093/molbev/mst197 Taylor, E.H. (1952) A review of the frogs and toads of Costa Rica. The University of Kansas Science Bulletin, 35, 577‒942. http://dx.doi.org/10.5962/bhl.part.4328 Vences, M., Thomas, M., Bonett, R.M. & Vieites, D.R. (2005) Deciphering amphibian diversity through DNA barcoding: Chances and challenges. Philosophical Transactions of the Royal Society B: Biological Sciences, 360, 1859–1868. http://dx.doi.org/10.1098/rstb.2005.1717 Wake, D.B. (2009) What salamanders have taught us about evolution. Annual Review of Ecology, Evolution, and Systematics, 40, 333–352. http://dx.doi.org/10.1146/annurev.ecolsys.39.110707.173552 APPENDIX. Specimens examined. All voucher numbers below are ‘UCR’ numbers, and the specimens were housed at the Museo de Zoología at the Universidad de Costa Rica. Craugastor stejnegerianus COSTA RICA: Puntarenas: Bahía Ballena, Osa (19272, 19277, 19278, 19282–19284, 19324–19326, 19328, 19353, 19527–19529, 19546–19548, 19589); Potrero Grande, Buenos Aires (20307–20309, 20346, 20352); Golfito, Golfito (11969, 11970, 12246, 12272–12289, 14958, 14959); Palmar, Osa (7865, 7866, 9057, 20885, 22136, 22137); Pavón, Golfito (12717–12719, 21494); Puerto Cortés, Osa (14230, 14235, 14244); Puerto Jiménez, Golfito (11302, 11305, 11306, 11308, 11311–11313, 16277, 16340, 16341, 18339, 18340, 18356–18359, 18361–18364, 19230, 20454); Savegre, Aguirre (14306, 14536, 14569, 14706, 14734, 15980, 15986, 16163); Sierpe, Osa (799, 800, 1081–1084, 1158, 3558–3560, 4576, 6400, 8733, 11309, 11350, 11351, 11621, 13673, 13674, 13683, 13684, 16275, 16276, 16345–16349, 16571–16575). San José: Barú, Pérez Zeledón (5010, 5011, 14737, 16330–16334, 16336, 22100, 22101, 22103); Daniel Flores, Pérez Zeledón (4234–4239); Platanares, Pérez Zeledón (15991, 15993, 15994); Rivas, Pérez Zeledón (16278–16280, 22127, 22128, 22131, 22132); San Isidro del General (5097–5101) Craugastor gabbi sp. nov. COSTA RICA: Puntarenas: Aguabuena, Coto Brus (13242, 15830, 15832, 15833); Sabalito, Coto Brus (8684, 15834); San Vito, Coto Brus (8644, 8645, 12507–12512, 12514, 12524–12526, 12935, 13237, 13240, 13241, 14671–14673, 18531– 18535, 21863–21865, 21867, 21876). 145 CAPÍTULO II.III A NEW SPECIES OF CRAUGASTOR (ANURA: CRAUGASTORIDAE) FROM THE MONTANE RAINFOREST OF THE CORDILLERA DE TALAMANCA, COSTA RICA Autores: Erick Arias, Gerardo Chaves y Gabriela Parra-Olea Fuente: Phyllomedusa, 17 (2): 211–232. 2018 Fecha de publicación: 18 de diciembre de 2018 Publicado por: Universidade de São Paulo - ESALQ URL: http://dx.doi.org/10.11606/issn.2316-9079.v17i2p211-232 146 211 Phyllomedusa - 17(2), December 2018 Received 18 June 2018 Accepted 27 November 2018 Distributed December 2018 A new species of Craugastor (Anura: Craugastoridae) from the montane rainforest of the Cordillera de Talamanca, Costa Rica Erick Arias,1,2 Gerardo Chaves,2 and Gabriela Parra-Olea1 1 Departamento de Zoología, Instituto de Biología, UNAM, AP 70-153, Ciudad Universitaria, CP 04510, Ciudad de México, Mexico. E-mail: gparra@ib.unam.mx. 2 Escuela de Biología, Universidad de Costa Rica, San Pedro, 11501-2060, San José, Costa Rica. Phyllomedusa 17(2):211–232, 2018 © 2018 Universidade de São Paulo - ESALQ ISSN 1519-1397 (print) / ISSN 2316-9079 (online) doi: http://dx.doi.org/10.11606/issn.2316-9079.v17i2p211-232 Abstract A new species of Craugastor (Anura: Craugastoridae) from the montane rainforest of the Cordillera de Talamanca, Costa Rica. A new dirt frog of the Craugastor podiciferus Species Group is described from Costa Rica; it is restricted to elevations between 2330 and 2700 m a.s.l. in the montane rainforest of the Cordillera de Talamanca. Analysis of DNA sequences of the mitochondrial 16S rRNA (16S) and cytochrome oxidase I (COI) genes reveals a distinct lineage within the C. podiciferus Species Group. Additional morphological and morphometric analyses support the distinctiveness of this lineage that is described as a new species herein. The species is distinguished from other members of the C. podiciferus Species Group by its unique coloration: a violet-brown to blackish brown venter with white pigment forming blotches, and dark brown palmar surfaces with prominent white folds between subarticular tubercles in the adults. The genetic divergence of the species from other members of the C. podiciferus Species Group is significant (higher than 9.2% in 16S and 13.3% in COI). Although not closely related, it resembles C. podiciferus morphologically, a species that also inhabits montane rainforest. The discovery of this new species highlights the importance of montane rainforest as a center of species richness and endemism. Keywords: Brachycephaloidea, Central America, Craugastor podiciferus Species Group, Panama, Terrarana. Resumen Una especie nueva de Craugastor (Anura: Craugastoridae) del bosque montano lluvioso en la Cordillera de Talamanca, Costa Rica. Se describe una especie nueva para Costa Rica de rana de hojarasca perteneciente al grupo de especies Craugastor podiciferus, restringida a elevaciones entre 2330–2700 m s.n.m. en el bosque montano lluvioso de la Cordillera de Talamanca. Análisis de las secuencias del ADN de los genes mitocondriales 16S ARNr (16S) y citocromo oxidasa 1 (COI) reveló un linaje distinto dentro del grupo de especies C. podiciferus. Los análisis complementarios 147 212 Phyllomedusa - 17(2), December 2018 Introduction The highlands of isthmian Central America are characterized by a high level of species richness and endemism. The isthmian highlands (1000–3820 m a.s.l.) are an isolated topographic unit in Central America (Campbell 1999, Gutiérrez-García and Vázquez-Domínguez 2013) formed by the Guanacaste, Tilarán, Central, and Talamanca mountain ranges. The Cordillera de Talamanca extends from the central valley in Costa Rica to western Panama and contains both the highest mountain peaks of the isthmus (Campbell 1999, Savage 2002) and the most endemic amphibians of Costa Rica (Campbell de morfología y morfometría apoyaron la diferenciación de este linaje, el cual describimos aquí como una especie nueva. Esta especie se distingue de los miembros del grupo de especies C. podiciferus por su coloración única: vientre violeta-marrón a marrón negruzco con pigmento blanco formando manchas, la superficie palmar en adultos es marrón oscuro con pliegues blancos prominentes entre los tubérculos subarticulares. Geneticamente esta nueva especie es significativamente divergente de los demás miembros del grupo de especies C. podiciferus (mayores a 9.3% en el 16S y 13.3% en COI). Aunque no están estrechamente relacionadas, la nueva especie es morfológicamente similar a C. podiciferus, especie que también habita en el bosque montano lluvioso. El descubrimiento de esta nueva especie resalta la importancia del bosque montano lluvioso como un centro de riqueza de especies y endemismos. Palabras clave: América Central, Brachycephaloidea, grupo de especies Craugastor podiciferus, Panamá, Terrarana. Resumo Uma nova espécie de Craugastor (Anura: Craugastoridae) do bosque montano chuvoso da Cordilheira de Talamanca, Costa Rica. Descrevemos aqui uma nova espécie do grupo de Craugastor podiciferus para a Costa Rica, restrita a altitudes entre 2330–2700 m acima do nível do mar no bosque montano chuvoso da Cordilheira de Talamanca. Análises das sequências de DNA dos genes mitocondriais16S ARNr (16S) e da citocromo oxidase 1 (COI) revelaram uma linhagem distinta dentro do grupo de espécies de C. podiciferus. Análises morfológicas e morfométricas complementares apoiaram a diferenciação desta linhagem, que descrevemos aqui como uma espécie nova. Essa espécie distingue-se dos membros do grupo de espécies de C. podiciferus por sua coloração única: ventre marrom-violeta a marrom enegrecido com pigmento branco formando manchas e superfície palmar nos adultos marrom escura com pregas brancas proeminentes entre os tubérculos sub-articulares. Geneticamente esta nova espécie é significativamente divergente dos demais membros do grupo de espécies de C. podiciferus (maiores em 9.3% no 16S e 13.3% no COI). Ainda que não estejam estreitamente relacionadas, a nova espécie é morfologicamente similar a C. podiciferus, que também habita o bosque montano chuvoso. A descoberta dessa nova espécie ressalta a importância do bosque montano chuvoso como um centro de riqueza de espécies e endemismos. Palavras-chave: América Central, Brachycephaloidea, grupo de espécies de Craugastor podiciferus, Panamá, Terrarana. 1999, Olson et al. 2001, Savage 2002, Boza- Oviedo et al. 2012). The summits of the Cordillera de Talamanca (ranging from 2500– 3500 m a.s.l.) are dominated by Montane Rainforest Life Zone (Holdridge 1967, Bolaños et al. 2005), which is extremely fragmented and isolated. The Talamanca montane rainforest (TMR) is relatively poorly studied and is thought to have a lower species diversity than the premontane forest of Talamanca Range (Kubicki 2008, Santos-Barrera et al. 2008, Arias and Bolaños 2014). Nevertheless, the TMR is home to several micro endemic amphibians—viz. Atelopus chirripoensis Savage and Bolaños, 2009, Bolitoglossa kamuk Boza-Oviedo, Rovito, Arias et al. 148 213 Phyllomedusa - 17(2), December 2018 Chaves, García-Rodríguez, Artavia, Bolaños, and Wake, 2012, B. pesrubra Taylor, 1952, B. pygmaea Bolaños and Wake, 2009, B. robinsoni Bolaños and Wake, 2009, B. splendida Boza- Oviedo, Rovito, Chaves, García-Rodríguez, Artavia, Bolaños, and Wake, 2012, and Diasporus ventrimaculatus Chaves, García- Rodríguez, Mora, and Leal, 2009. During recent fieldwork in the TMR we found frogs of the Craugastor podiciferus Species Group Hedges et al. (2008). The anurans were collected at the summits of Cerro Arbolado, Cerro Hakú, Cerro Utyum, and Caribbean slopes of Cerro Pando. The population of Cerro Utyum is near (~ 10 km airline distance) to the type locality of C. podiciferus (Cope, 1875) (Cope 1875, Arias and Chaves 2014) and both species of anurans are morphologically similar. Craugastor podiciferus is abundant, with a broad distribution across the highlands of all the mountain ranges of Costa Rica and western Panama (1000–2650 m a.s.l.); it has been thought to represent a species complex with several unnamed species (Savage 2002, Streicher et al. 2009). Molecular analyses of two mitochondrial genes showed that despite the morphological resemblance to C. podiciferus, the recently discovered populations from the TMR are not C. podiciferus, but instead represent a new member of the species group, which is described here based on molecular and morphological data. Materials and Methods Taxon Sampling Frogs of the new species were collected at five localities along the Cordillera de Talamanca, at the summits of Cerro Arbolado, Cerro Hakú, Cerro Utyum, and the Caribbean slopes of Cerro Pando in southwestern Costa Rica (Figure 1). In addition, we collected several specimens from highlands of Costa Rica and western Panama, including several populations referred to C. podiciferus, Craugastor sp.B of Crawford & Smith (2005), and two populations that we record herein as unnamed species (Craugastor sp.1 and Craugastor sp.2). The frogs were euthanized in the field and extracted liver or muscle tissue was preserved in 95% ethanol or RNAlater. Voucher specimens were fixed in 10% formalin, stored in 70% ethanol, and deposited at the Museo de Zoología, Universidad de Costa Rica (UCR) and Senckenberg Research Institute and Nature Museum, Frankfurt, Germany (SMF). Museum codes follow those of Frost (2018), with the addition of CRARC in reference to the Costa Rica Amphibian Research Center private collection; EAP denotes field numbers of Erick Arias. Amplification and Sequencing We extracted total genomic DNA from the preserved tissue samples using the Animal Genomic DNA Kit (BioBasic Canada Inc.), DNeasy Blood & Tissue Kit (Qiagen), or the phenol-chloroform standard protocol (Sambrook & Russell 2006). We amplified the large subunit ribosomal RNA (16S) and cytochrome oxidase subunit I (COI) mitochondrial genes. The primers 16Sar and 16Sbr (Palumbi et al. 1991) were used for 16S and dgLCO and dgHCO (Meyer 2003) for COI. The PCR amplifications were performed using a total volume of 15 µL, which contained 1 µL DNA template (c. 50 ng/ µL), 0.75 U Taq polymerase (Amplificasa®, Biotecnologias Moleculares), 1X PCR buffer with 1.5 mm MgCl 2 , 0.2 mM deoxynucleotide triphosphates (dNTPs), and 0.3–0.5 µM forward and reverse primers. The PCR conditions are as follow: 16S, an initial cycle of 5 min at 94°C, followed by 35 cycles of 45 s at 94°C, 30 s at 50 or 55°C, 45 or 120 s at 72°C, plus a final cycle of 3 min at 72°C; COI, an initial cycle of 2 min at 94°C, followed by 35 cycles of 30 s at 94°C, 30 s at 50°C, 45 s at 72°C, plus a final cycle of 3 min at 72°C. PCR products were cleaned with ExoSap-IT (USB Corporation) and sequenced in both directions using the original amplification primers and BigDye termination reaction chemistry (Applied Biosystems). The cycle- A new species of Craugastor from Costa Rica 149 214 Phyllomedusa - 17(2), December 2018 Figure 1. Map showing the known populations of Craugastor aenigmaticus sp. nov. (red star = type locality) in the Cordillera de Talamanca, and the populations of C. podiciferus used in our molecular analysis. Solid (black) lines depict the geographical limits between each mountain range. sequencing products were column-purifi ed with Sephadex G-50 (GE Healthcare) and run on an ABI 3500xL Genetic Analyzer (Applied Biosystems). Consensus sequences for each individual were constructed using SEQUENCHER 5.3 (Genes Codes Corp.). Phylogenetic Analyses The sequences obtained were compared to those available in GenBank for the Craugastor podiciferus Species Group and sequences of C. gollmeri (Peters, 1863) were used as the outgroup. See Appendix I for the list of DNA voucher and GenBank Accession numbers used in this study. Sequence alignments were performed using the MUSCLE 3.7 software (Edgar 2004) with default parameters and trimmed to the point at which a majority of taxa had sequence data. We partitioned the sequence data by gene, and further partitioned COI by codon position. We used PartitionFinder v1.1.1 (Lanfear et al. 2012) and the Bayesian Information Criterion (BIC) to select the best partition scheme and the best model of sequence evolution for each partition. We used a single set of branch-lengths across all partitions (branchlengths = linked); the search of the best partition scheme was using a heuristic search (scheme = greedy). We defi ned, a priori, four partitions: one for 16S and three for COI (one for each codon). Arias et al. 150 215 Phyllomedusa - 17(2), December 2018 We performed analyses using both the maximum likelihood (ML) and Bayesian analyses (BA). For ML we used Garli 2.01 (Zwickl 2006), with 10 search replicates with the following default setting values changed: streefname = random, attachmentspertaxon = 24, genthreshfor- topoterm = 100000, significanttopochange = 0.00001. For bootstrapping, we ran 1000 replicates with the previous settings with the following changes: genthreshfortopoterm = 10000, significanttopochange = 0.01, treere- jectionthreshold = 20, as suggested in the Garli manual to speed up bootstrapping. The bootstrap consensus tree was performed using Sumtrees (Sukumaran and Holder 2010a) from DendroPy packages Version 4.4.0 (Sukumaran and Holder 2010b). Bayesian phylogenetic analysis was performed using MrBayes 3.2.6 (Ronquist et al. 2012) with the partition scheme and the model of sequence evolution for each partition as selected previously. Two separate analyses were run; each consisted of 20 million generations, sampled every 1000 generations, and four chains with default heating parameters. We examined a time-series plot of the likelihood scores of the cold chain to check stationarity using Tracer 1.6 software (Rambaut et al. 2014). We discarded the first 25% of trees as burn-in and used the remaining trees to estimate the consensus tree along with the posterior probabilities for each node and each parameter. The ML and Bayesian analyses were run on the CIPRES portal (Miller et al. 2010). Estimates of pairwise evolutionary genetic divergence between species were computed using MEGA7 (Kumar et al. 2016), assuming uncorrected distances based on the Tamura 3-parameter model (Tamura 1992), with rate variation among the sites modeled as a gamma distribution with the shape parameter = 4 as the default of the software. Morphometric Analyses We performed a morphometric analysis to compare the new species with Craugastor podiciferus because the taxa closely resemble one another and both inhabit the TMR (allopatrically). In addition we compared the new species with a third undescribed species (Craugastor sp.1) because populations of this frog are located nearby (~ 5 km airline distance) (Figure 1). We examined 20 specimens of the new species from four localities, 86 C. podiciferus from several localities in Costa Rica, and 19 Craugastor sp.1 (Appendix II). Specimens are deposited at UCR. The following morphological measurements were recorded, as described by Savage (2002), Duellman and Lehr (2009), and Arias et al. (2016): snout–vent length (SVL), head length (HL), head width (HW), interorbital distance (IOD), width of upper eyelid (EW), intercanthal distance (IC), internarial distance (IN), upper lip–nostril distance (TN), eye–nostril distance (E–N), eye diameter (ED), tympanum diameter (TY), ulna length (UL), hand length (HaL), lengths of the Fingers I (F1) and III (F3), femur length (FL), tibia length (TL), tarsus length (TaL), foot length (FoL), and lengths of the Toes III (T3) and V (T5). Measurements were taken with dial calipers and were rounded to the nearest 0.1 mm. We transformed the morphometric data using the method of Lleonart et al. (2000) to avoid allometric effects relative to the differences in the size and shape between species and between individuals. In this method, a logarithmic transformation of the continuous variables is performed to reduce the extreme values. All transformed variables are used in the allometric transformation by means of equation Y i * = Y i (X 0 /X i )b, where Y i * is the value of each of the dependent variable corrected for size and shape; Y i is the value of each of the dependent morphometric variable; X 0 is the average of the SVL variable for all populations; X i is the SVL value for each individual; and b is the regression line intercept with the Y-axis resulting from the regression of each dependent variable with X 0 . The intercept is used as an allometric transformation factor and is unique for each A new species of Craugastor from Costa Rica 151 216 Phyllomedusa - 17(2), December 2018 variable. The additional proportions reported here include: EW/IOD, IOD/HW, TY/ED, EN/ ED, ED/HL, IC/HL, IN/EN, IN/TN, FL/TL, TL/ TaL, TaL/FoL, T3/FoL, T5/FoL, UL/HaL, F1/ HaL, and F3/HaL. The sex of individuals was determined by gonadal morphology. Specimens with opaque seminal vesicles were assumed to be adult males, and those with developed oviducts were assumed to be adult females. The general terminology for the morphological characteristics follows that of Duellman and Lehr (2009). We adopted Savage’s (2002) usage of the term “supernumerary tubercles” to refer to the tubercles on the phalanges (between subarticular tubercles); this differs from the tubercles denoted as accessory palmar or plantar tubercles. We calculated the mean, standard deviation, and range for each morphometric variable without correction. We performed a discriminant analysis to determine whether the morphometric variables were effective to predict the species, using the following variables: EW/IOD, IOD/ HW, TY/ED, EN/ED, ED/HL, IC/HL, IN/EN, IN/TN, FL/TL, TL/TaL, TaL/FoL, T3/FoL, UL/ HaL, F1/HaL, and F3/HaL. We also conducted a Principal Component Analysis (PCA) to explore the degree of structure within the sample and which variables have more loads in the segregation of groups. All analyses were performed using R v3.3.3 (R Core Team 2017). Results Molecular Data The data matrix includes 56 sequences, with a total sequence length of 1222 bp, including gaps: 565 bp for 16S and 657 bp for COI. Three partition schemes were identified with the following substitution models: GTR + I + G for 16S and COI codon position 3, K80 + I + G for COI codon position 1, and HKY+I for COI codon position 2. Genetic distances between the new species and other members of the Craugastor podiciferus Species Group are of 9.2–18.5% for 16S and 18.9–24.8% for COI. Specifically, the new species is separated by a mean-uncorrected genetic distance to C. podiciferus of 11.57– 16.15% for 16S and 19.16–24.35% for COI and to Craugastor sp.1 of 12.25–12.9% in 16S and 23.57–24.72% in COI. The phylogenies inferred by Garli and MrBayes were mostly congruent (Figure 2), with six well-supported clades. The three most basal clades represent three unnamed species (Craugastor sp.B, Craugastor sp.1, and Craugastor sp.2) that occur in the highlands of the southwestern end of Cordillera de Talamanca. The fourth clade contains the samples from Cerro Hakú, Cerro Utyum, and Cerro Pando at Cordillera de Talamanca (Figure 1). A fifth major clade includes seven species of the C. podiciferus Species Group that mainly occur from the lowlands to intermediate elevations from eastern Honduras to central Panama. The sixth major clade is formed by all the samples of C. podiciferus sensu lato that are broadly distributed in the highlands of Costa Rica and western Panama (Figure 1). The main differences between the ML and Bayesian topologies are in the relationships of Craugastor sp.B, Craugastor sp.1, and Craugastor sp.2 within the phylogeny. Their placement is unresolved in the ML tree, whereas in the Bayesian tree (not shown), Craugastor sp.1 is the most basal taxon for the entire group. Craugastor sp.1 + Craugastor sp.B form a clade that is sister to the rest of the species group. Morphometric Analysis Morphometric variation among the species is shown in Table 1. The PCA did not differentiate specimens of the new species from those of Craugastor podiciferus and Craugastor sp.1. The discriminant analysis correctly classified 87.9% of the specimens to the species, showing a clear separation between the specimens of the new taxon and the specimens referred to C. podiciferus and Craugastor sp.1 (Figure 3). Arias et al. 152 217 Phyllomedusa - 17(2), December 2018 Figure 2. Maximum likelihood phylogeny of Craugastor podiciferus Species Group based on 16S and COI mitochondrial DNA genes. Bootstraps proportions are before the slash, and posterior probabilities (multiplied by 100) from MrBayes analysis are following the slash. The scale bar refers to the estimated substitutions per site. The support values of any node within species are not shown. The asterisks represent support of > 95. A new species of Craugastor from Costa Rica 153 218 Phyllomedusa - 17(2), December 2018 Figure 3. Linear discriminant analysis showing the morphological separation among the 20 individuals of Craugastor aenigmaticus sp. nov. and the individuals of C. podiciferus and Craugastor sp.1. Craugastor aenigmaticus sp. nov. Montane Dirt Frog (Figures 4–6) Holotype.—UCR 22961 (EAP 0762), an adult female from Costa Rica: Provincia de Punta- renas: Cantón de Buenos Aires: Distrito de Buenos Aires: summit of Cerro Arbolado, Parque Internacional La Amistad, (09°19'12.0'' N, 83°12'57.6'' W; 2600 m a.s.l.), collected by Erick Arias and Omar Zúñiga on 19 October 2016. Paratopotypes.—UCR 22957 (EAP 0758), subadult male; UCR 22958 (EAP 0759) and UCR 22960 (EAP 0761), adult females; UCR 22962 (EAP 0763), subadult female; UCR 22959 (EAP 0760), juvenile; same date as the holotype. Paratypes.—UCR 21951 (EAP 0303), adult female from Costa Rica: Provincia de Limón: Cantón de Talamanca: Distrito de Telire: Caribbean slopes of Cerro Utyum, Parque Internacional La Amistad, (09°20'56.4'' N, 83°10'30.0'' W; 2700 m a.s.l.), collected by Erick Arias, Gerardo Chaves, Olmer Cordero, and Omar Zúñiga on 12 July 2013. UCR 22414 (EAP 0490) and UCR 22415 (EAP 0491), adult females, same data as UCR 21951 but collected on 29 March 2013. UCR 22737 (EAP 0674) and Arias et al. 154 219 Phyllomedusa - 17(2), December 2018 Table 1. Mean ± SD and range (in mm) for morphometric variables by species. Abbreviations: SVL, snout–vent length; HL, head length; HW, head width; IOD, interorbital distance; EW, width of the upper eyelid; IC, intercanthal distance; IN, internarial distance; TN, upper lip–nostril distance; E–N, eye–nostril distance; ED, eye diameter; TY, tympanum diameter; UL, ulna length; HaL, hand length; F1, length of Finger I; F3, length of Finger III; FL, femur length; TL, tibia length; TaL, tarsus length; FoL, foot length; T3, length of Toe III; T5, length of Toe V. Variable Craugastor aenigmaticus sp. nov. Craugastor podiciferus Craugastor sp.1 Mean ± SD Min–Max Mean ± SD Min–Max Mean ± SD Min–Max SVL 26.38 ± 9.33 16.10–41.10 24.27 ± 4.76 15.10–35.10 20.79 ± 3.94 13.80–26.50 HL 10.65 ± 3.52 6.20–15.95 9.91 ± 1.76 6.45–13.60 8.44 ± 1.47 5.70–10.30 HW 10.61 ± 3.68 6.25–16.50 9.66 ± 1.88 6.20–14.15 8.22 ± 1.64 5.60–10.70 IOD 3.265 ± 1.10 1.80–5.55 3.23 ± 0.53 2.10–4.55 2.74 ± 0.51 1.80–3.70 EW 2.08 ± 0.70 1.20–3.20 1.80 ± 0.38 1.15–2.90 1.62 ± 0.34 1.15–2.10 IC 5.20 ± 1.64 3.25–8.00 5.02 ± 0.77 3.55–6.95 4.22 ± 0.73 3.15–5.30 IN 3.29 ± 0.99 2.15–4.85 3.04 ± 0.51 2.20–4.30 2.57 ± 0.39 1.95–3.25 TN 1.54 ± 0.46 1.00–2.30 1.25 ± 0.24 0.90–2.10 1.04 ± 0.18 0.70–1.25 EN 2.59 ± 0.77 1.55–3.70 2.43 ± 0.46 1.50–3.60 1.99 ± 0.42 1.30–2.60 ED 3.05 ± 0.95 2.05–4.90 2.88 ± 0.49 1.90–3.90 2.46 ± 0.32 1.90–2.90 TY 2.01 ± 0.73 0.95–3.15 1.75 ± 0.41 1.00–2.70 1.96 ± 0.32 1.55–2.95 UL 6.08 ± 2.34 3.10–9.65 5.56 ± 1.09 3.50–8.20 4.80 ± 0.99 3.35–6.30 HaL 6.93 ± 2.57 3.70–11.00 5.83 ± 1.24 3.55–9.50 5.03 ± 0.99 3.10–6.55 F1 2.62 ± 1.31 1.10–4.85 2.10 ± 0.55 1.10–3.95 1.93 ± 0.56 0.90–2.95 F3 4.16 ± 1.62 2.20–6.50 3.46 ± 0.79 1.75–6.10 2.97 ± 0.62 1.80–3.90 FL 14.73 ± 5.85 7.90–23.00 12.40 ± 2.80 7.60–20.30 10.67 ± 1.96 7.05–14.00 TL 16.75 ± 6.54 8.50–25.85 14.07 ± 3.04 8.25–22.50 12.25 ± 2.26 8.25–15.25 TaL 9.44 ± 3.38 5.45–14.25 8.28 ± 1.66 5.10–12.85 7.36 ± 1.35 5.10–9.30 FoL 14.87 ± 5.76 7.95–23.50 12.79 ± 2.75 7.65–20.15 10.94 ± 2.24 7.05–14.15 T3 5.40 ± 2.32 2.05–9.70 4.46 ± 0.99 2.60–7.25 3.79 ± 0.77 2.45–5.05 T5 4.27 ± 1.82 1.70–7.00 3.67 ± 0.95 1.90–6.70 2.99 ± 0.58 1.90–3.80 EW/IOD 0.64 ± 0.07 0.46–0.75 0.56 ± 0.09 0.38–0.89 0.59 ± 0.08 0.44–0.76 EN/ED 0.86 ± 0.11 0.67–1.02 0.84 ± 0.07 0.64–1.04 0.81 ± 0.10 0.63–0.98 ED/HL 0.29 ± 0.03 0.25–0.35 0.29 ± 0.02 0.25–0.33 0.29 ± 0.02 0.26–0.33 IOD/HW 0.31 ± 0.03 0.27–0.38 0.34 ± 0.03 0.27–0.43 0.34 ± 0.02 0.30–0.38 TY/ED 0.66 ± 0.15 0.45–1.03 0.61 ± 0.10 0.41–1.00 0.81 ± 0.17 0.62–1.34 IC/HL 0.49 ± 0.02 0.43–0.53 0.51 ± 0.03 0.45–0.61 0.50 ± 0.05 0.36–0.58 IN/EN 1.27 ± 0.12 1.03–1.55 1.26 ± 0.11 0.95–1.53 1.31 ± 0.16 1.10–1.62 IN/TN 2.14 ± 0.13 1.94–2.44 2.45 ± 0.20 1.90–2.94 2.50 ± 0.22 2.21–3.00 FL/TL 0.88 ± 0.03 0.80–0.94 0.88 ± 0.05 0.54–0.96 0.87 ± 0.03 0.79–0.92 TL/TaL 1.76 ± 0.10 1.51–1.94 1.70 ± 0.07 1.48–1.83 1.67 ± 0.07 1.58–1.91 TaL/FoL 0.64 ± 0.04 0.58–0.77 0.65 ± 0.03 0.58–0.73 0.68 ± 0.03 0.59–0.72 T3/FoL 0.36 ± 0.04 0.24–0.46 0.35 ± 0.02 0.30–0.39 0.35 ± 0.01 0.32–0.36 T5/FoL 0.28 ± 0.04 0.18–0.34 0.28 ± 0.02 0.22–0.38 0.27 ± 0.02 0.22–0.31 UL/HaL 0.88 ± 0.05 0.77–0.95 0.96 ± 0.06 0.81–1.16 0.96 ± 0.08 0.82–1.11 F1/HaL 0.36 ± 0.06 0.27–0.47 0.36 ± 0.04 0.24–0.43 0.38 ± 0.05 0.29–0.46 F3/HaL 0.60 ± 0.02 0.56–0.64 0.59 ± 0.04 0.49–0.67 0.59 ± 0.04 0.51–0.68 A new species of Craugastor from Costa Rica 155 220 Phyllomedusa - 17(2), December 2018 UCR 22747 (EAP 0684), adult males from Costa Rica: Provincia de Puntarenas: Cantón de Buenos Aires: Distrito de Buenos Aires: summit of Cerro Hakú, Parque Internacional La Amistad, (09°19’19.2” N, 83°12’10.8” W; 2660 m a.s.l.), collected by Erick Arias and Omar Zúñiga on 28 December 2015. Group assignment.—Assigned to the genus Craugastor based on possessing the following characters: differentiated tympanum; absence of cranial crest; and Toe III longer than Toe V. Assigned to the C. podiciferus Species Group based on having a narrow head (HW/SVL = 36.1–43.64%) and a rugose dorsum, but lacking inner tarsal folds, webbing between the toes, nuptial pads, and vocal slits. Diagnosis.—The combination of the following characteristics distinguish Craugastor aenigmaticus from its congeners (Figure 4 and 5): (1) skin on venter smooth, but with large granules laterally; (2) vocal slits absent; (3) nuptial pads absent; (4) toes lacking webbing; (5) heels lacking enlarged calcar tubercle, but can have one to three small tubercles or granules on heels; (6) supernumerary, accessory palmar, and plantar tubercles absent; and (7) unique coloration consisting of a dark brown, olive, olive-brown or dark violet-brown dorsal ground color, violet-brown to blackish brown venter with white to white-bluish pigment forming blotches, dark brown palmar surface in adults with prominent white folds between subarticular tubercles, and in some specimens, white bones apparent through the skin (Figure 6C). Comparisons with other species.—Craugastor aenigmaticus differs from all the other craugastorids of isthmian Central America except for those in the C. podiciferus Species Group by having a narrow head (HW 36.1– 43.64% SVL) and toes that lack webbing. Craugastor aenigmaticus differs from other members of the C. podiciferus Species Group by having the following characteristics (condition Figure 4. Craugastor aenigmaticus sp. nov. Photograph taken by EA. for C. aenigmaticus in parentheses). In C. bransfordii (Cope, 1886), C. gabbi Arias, Chaves, Crawford, and Parra-Olea, 2016, C. lauraster (Savage, McCranie, and Espinal, 1996), C. persimilis (Barbour, 1926), C. polyptychus (Cope, 1886), C. stejnegerianus (Cope, 1893), and C. underwoodi (Boulenger, 1896): (1) the venter is cream (olive-brown in life, and dark brown in ethanol); (2) the venter, as well as the midline, is completely areolate to tuberculate (smooth, at least in the midline in C. aenigmaticus); and (3) they range in altitude from 0–1600 m a.s.l. (range 2330–2700 m a.s.l.). Craugastor jota (Lynch, 1980) differs from C. aenigmaticus by having a prominent calcar tubercle on the heel (evident calcar tubercle absent, although some individuals have one to three small tubercles). Craugastor podiciferus differs from C. aenigmaticus by: (1) having a prominent calcar tubercle on the heel (Figure 7) (evident calcar tubercle absent although some individuals could have one to three small tubercles); (2) venter yellow, orange, grayish or olive in adults (adults with venter violet-brown with white blotches); and (3) by lack of white prominent folds between subarticular tubercles on hands of adults (prominent white folds between the subarticular tubercles on the hands of adults) (Figure 6D). Arias et al. 156 221 Phyllomedusa - 17(2), December 2018 A C B D Figure 5. Dorsal (right) and ventral (left) color in life (above) and in ethanol (below) of the holotype (UCR 22961) of Craugastor aenigmaticus sp. nov. Photographs taken by EA. Description of the holotype.—Adult female, SVL = 40.1 mm (Figures 4, 5). Head relatively narrow, width = 41.15% SVL; snout subovoid in dorsal view, rounded in profile; snout relatively long (HL = 15.6 mm, 38.9% SL), with nostrils directed laterally; in ventral view, tip of snout protruding markedly beyond edge of lower lip. Internarial area convex (IN 4.85 mm); canthus rostralis rounded; intercanthal area flat (IC = 8.0 mm); loreal region slightly concave; vomerine teeth transverse, in two fascicles well behind the choanae. Tongue round, lacking a distinct posterior notch; teeth absent; choanae moderately large, rounded on posterior half, but flat on anterior half, hemispherical; vocal slits absent. Eye moderate (EW = 92.75% E-N), not A new species of Craugastor from Costa Rica 157 222 Phyllomedusa - 17(2), December 2018 protruding beyond dorsal and ventral outline of head, directed laterally. Tympanum distinct, round, and covered by skin; tympanic annulus prominent, round, small (54.12% of ED). Skin on all dorsal and lateral surfaces of head Figure 6. Variation of the ventral views of the hands for comparison. (A) Craugastor aenigmaticus sp. nov. holotype in life (UCR 22961); (B) holotype (UCR 22961) in ethanol; (C) paratopotype (UCR 22958); (D) C. podiciferus. Photographs taken by EA. A C B D moderately granular. Upper eyelid granular, without superciliar or supraocular tubercles. Postrictal tubercles fused forming short ridge posteroventral to tympanum. Skin on dorsum and limbs weakly granular. Skin of chest smooth, Arias et al. 158 223 Phyllomedusa - 17(2), December 2018 Figure 7. Dorsal (A) and ventral (B) color in life views of Craugastor podiciferus (EAP 0803), for comparison. Note the enlarged calcar tubercle and coloration on dorsum and venter. Photographs taken by EA. A B venter smooth and encroached on laterally by low granules; ventral surfaces of thighs smooth to weakly granular; skin of groin and ventral surfaces of arms and lower legs nearly smooth. Flanks areolate, especially along the anteroventral flank region; skin on chin smooth. A pair of complete hourglass-shaped dorsal ridges present between posterior margin of eye and sacrum. Pair of supratympanic folds extending from posterior margin of eye, above tympanum, bifurcating at the axillary level, one fold continues laterally and other continues to the venter. Discoidal fold complete. Forelimb relatively short and robust; fingers moderately long and slim without lateral fringes. Discs absent; fingers with grooves; tips of fingers unexpanded, rounded in dorsal view; pads ovoid. Proximal subarticular tubercles indistinct. Supernumerary tubercles absent; accessory palmar tubercles absent; distal subarticular tubercles rounded in basal outline, flattened, and globular in profile; thenar tubercle ovoid, flattened, palmar tubercle rounded, flattened; thenar tubercle much smaller than palmar. Ulnar tubercles and fold absent. Fingers are not webbed. Legs relatively long and robust; heel granular but lacking enlarged tubercles. Discs absent; toes with grooves; tips of toes unexpanded, lanceolate in dorsal view; pads triangular. Supernumerary tubercles absent; plantar tubercles absent; subarticular tubercles ovoid in basal outline, flattened, and globular in profile; inner metatarsal tubercle elongate, globular; outer metatarsal tubercle rounded, barely discernible, flattened; outer metatarsal tubercle much smaller than inner; inner tarsal fold absent; toes are not webbed. Cloacal opening directed posteriorly at midlevel of thighs. Coloration of the holotype in life.—Dorsal ground color is uniform dark brown, but with two dark spots on the scapular region (Figure 4). A dark interorbital mark is present, lying just posterior to an adjacent paler area. Dorsal surfaces of the legs and arms with dark bars. The surfaces below the canthus rostralis and the supratympanic fold, from the snout to the axilla, are darker (forming a type of mask); the dark pigment of the mask continues posteriorly bordering the supratympanic fold. The upper lip has dark bars with white pigment in form of A new species of Craugastor from Costa Rica 159 224 Phyllomedusa - 17(2), December 2018 faded bars. The flanks are the same color as dorsum, but with white pigment in form of small spots. Ventral surfaces of the body and legs are violet-brown with bluish-white pigment in the form of blotches; ventral surface of tibia-tarsus grayish; ventral surface of the throat violet- brown is relatively uniform, slightly paler than the venter (Figure 5). The palmar surface of the hands with violet coloration, especially on thenar, palmar, and proximal subarticular tubercles; the fingers are whitish, with a white prominent fold between subarticular tubercles (Figure 6A). Coloration of the holotype in ethanol.—After 2 years in ethanol (70%), the overall dark brown dorsum has changed little from its color in life. The violet-brown on ventral surfaces of the body and legs has faded to dark brown and the bluish- white blotches have changed to light brown. The overall patterns of the blotches and other markings on the holotype have remained identical to those that were observed while alive. Measurements of holotype (mm).—SL 40.1; HL 15.6; HW 16.5; IOD 4.65; EW 3.2; IC 8.0; IN 4.85; TN 2.3; EN 3.45; ED 4.25; TY 2.3; UL 9.65; HaL 10.25; F1 4.85; F3 6.5; FL 22.65; TL 25.85; TaL 14.25; FoL 23.5; T3 8.1; T5 7.0. Measurements in related percentages: EW/IOD 68.82%; IOD/HW 28.18%; TY/ED 54.12%; EN/ ED 81.12%; ED/HL 27.24%; IC/HL 51.28%; IN/EN 140.58%; IN/TN 210.87%; FL/TL 87.62%; TL/TaL 181.40%; TaL/FoL 60.64%; T3/FoL 34.47%; T5/FoL 29.79%; UL/HaL 94.15%; F1/HaL 47.32%; F3/HaL 63.42%. Variation.—Morphometric variation is summarized in Table 1. Craugastor aenigmaticus has a relatively high level of intraspecific poly- morphisms (Figure 8). The skin on venter ranges from almost completely smooth to heavily areolate, but always with a smooth midline. Some frogs have supraocular tubercles; additionally, some individuals have two postrictal tubercles that are not fused. The palmar tubercle is heart shaped in some specimens, but ovoid in others; the proximal subarticular tubercles are not visible in some specimens. In the adult male, UCR 22747, an accessory palmar tubercle is visible. In some specimens, the heel is areolate; in others one to three small tubercles are visible. Some specimens have flattened subarticular tubercles, whereas others have projecting subarticular tubercles. Some individuals have globular subarticular tubercles in profile, whereas others have obtuse subarticular tubercles. UCR 21951 and UCR 22734 have a pattern of dark brown lateral stripes on a pale brown background (Figure 8E); UCR 21956 has an olive-green dorsum (Figure 8F). The mask is absent in some specimens. The throat is a uniform cream coloration in some frogs, whereas it is uniform violet-brown, heavily mottled, or uniformly dark brown in others. The venter usually is dark- brown or grayish in coloration, but in some individuals, especially in juveniles, it is cream (Figure 8D). The juveniles have light brown arms in contrast to the dark brown dorsum (Figure 8C). The palmar surfaces of the hands in adults usually are dark brown with prominent white folds between subarticular tubercles, but in some frogs, the white bones are visible beneath the skin; in juveniles, the palmar surfaces are dark brown or cream. Habitat and natural history notes.—Craugastor aenigmaticus inhabits the montane rainforest of the Cordillera de Talamanca (Holdridge 1967, Bolaños et al. 2005), which has a short dry season (1 or 2 mo), annual precipitation ranging between 2200 and 4500 mm, and annual temperatures ranging from 6–12°C. The type locality (Cerro Arbolado) and the other known localities consist of primary forest dominated by oak trees (genus Quercus L.) that are abundantly covered with bryophytes and epiphytes (Figure 9). The forest floor is covered by a thick layer of leaf litter and other types of decomposing organic material. Little is known about the natural history of C. aenigmaticus, but it is important to note that the species was abundant Arias et al. 160 225 Phyllomedusa - 17(2), December 2018 Figure 8. Variation on dorsum and venter of Craugastor aenigmaticus sp. nov. (A, B) Paratopotype UCR 22958, adult female; (C, D) paratopotype UCR 22959, juvenile; (E) paratype UCR 21951, adult female; (F) UCR 21956, juvenile. Photographs taken by EA. A C E B D F A new species of Craugastor from Costa Rica 161 226 Phyllomedusa - 17(2), December 2018 during our fieldwork in the months of April and July 2013, December 2015, and October 2016. During December 2015 and October 2016, gravid females were observed. Juveniles were observed during all the periods of fieldwork; the smallest size recorded was with a SVL of 16.1 mm. All specimens of C. aenigmaticus collected were discovered as they were jumping on the forest floor. We did not record any vocalization that we could attribute to C. aenigmaticus, though we think that the frog may vocalize. Craugastor aenigmaticus is sympatric with Diasporus ventrimaculatus in Cerro Arbolado, Cerro Hakú, on the Caribbean slopes of Cerro Utyum, and Valle del Silencio, which is the type locality of D. ventrimaculatus. Figure 9. Cloud forest at type locality of Craugastor aenigmaticus sp. nov., summit of Cerro Arbolado at 2600 m a.s.l. Photograph taken by Omar Becerra Soria. Distribution.—Craugastor aenigmaticus is restricted to the summit of Cerro Arbolado, Cerro Hakú, Caribbean slopes of Cerro Utyum, Valle del Silencio, and Caribbean slopes of Cerro Pando (Figure 1). The altitudinal range of this taxon is 2330–2700 m a.s.l. The species occurs in primary forest; and the populations on Cerro Arbolado, Cerro Hakú, Cerro Utyum, and Valle del Silencio are within the La Amistad International Park and that on Cerro Pando is within the Zona Protectora Las Tablas. The distribution of C. aenigmaticus is fragmented along a line of ~ 80 km. We did not find this species during fieldwork carried out on Cerro Kamuk and Cerro Echandi. More fieldwork is required to assess the range of this species more Arias et al. 162 227 Phyllomedusa - 17(2), December 2018 accurately. Given that this species is found at Cerro Pando, which is on the Costa Rica–Panama border, C. aenigmaticus is recognized as having a distribution within southeastern Costa Rica and at least for now, marginally into Panama. Etymology.—The specific name is derived from the Latin word for enigmatic. We propose this name in light of the taxonomic confusion surrounding this species. Craugastor aenigmaticus was first collected in 2009 but was erroneously identified as C. podiciferus because of its great morphological resemblance to the latter, and its close proximity to the type locality of C. podiciferus. However, based on molecular data, we confirmed that this taxon is highly divergent genetically from other species in the C. podiciferus Species Group and represents a new species. Discussion With the recognition of Craugastor aenigmaticus, the C. podiciferus Species Group now comprises 10 species, all of which are endemic to Costa Rica and western Panama (Savage 2002, AmphibiaWeb 2018). However, the diversity within the C. podiciferus Species Group is underestimated and several species remain unnamed (Streicher et al. 2009). The high genetic divergence between C. aenigmaticus and all other members of the C. podiciferus Species Group strongly supports the distinctiveness of C. aenigmaticus. The above- mentioned genetic distances are greater than those suggested by Fouquet et al. (2007) for recognizing new taxa of Neotropical frogs, which is 3% 16S divergence as threshold to define candidate species. However, it is important to point out that although not closely related, C. aenigmaticus and C. podiciferus are morphologically similar. Furthermore, unlike other members of the group, both of these species inhabit the montane rainforest, although C. podiciferus is not restricted to it. The montane rainforest of the Cordillera de Talamanca is naturally fragmented in two relatively large patches (Bolaños et al. 2005). One extends from Cerro Vueltas to Cerro Dúrika and the other from Cerro Arbolado to Cerro Echandi, and both are separated by a depression (2300 m a.s.l.). Craugastor aenigmaticus inhabits the patch that extends from Cerro Arbolado to Cerro Echandi. Bolitoglossa kamuk, B. pygmaea, B. robinsoni, and B. splendida are endemic to the same patch of montane rainforest. Craugastor aenigmaticus occurs in sympatry with D. ventrimaculatus in four localities, with the former only being known from one additional locality (Cerro Pando). The restricted distribution of Craugastor aenigmaticus to montane rainforest highlights the importance of this habitat as center of diversification and endemism, and suggests that the highlands of the Cordillera de Talamanca have had an important role in the process of speciation, possibly in association with climatic fluctuations (Savage 2002, Streicher et al. 2009). At least eight amphibians are restricted to montane rainforest in Cordillera de Talamanca (Savage 2002, AmphibiaWeb 2018), a habitat that was naturally reduced along the last 2 million years and that is fragmented (Foster 2001). Further studies are needed to help us resolve the phylogenetic relationships of Craugastor aenigmaticus within the C. podiciferus Species Group and to evaluate the biological significance of the high genetic distances that exist within the group. Additionally, it is important to continue inventory work in the montane rainforest to justify the need to protect the flora and fauna of this life zone, much of which may remain unknown. Acknowledgments We thank Laura Márquez-Valdelamar and Andrea Jiménez-Marín for their laboratory assistance; Federico Bolaños for the use of specimens from the Museo de Zoología of the Universidad de Costa Rica; Omar Zúñiga and A new species of Craugastor from Costa Rica 163 228 Phyllomedusa - 17(2), December 2018 Olmer Cordero provided valuable assistance in the field during the expeditions; Omar Becerra Soria provided photographs of the type locality. We are especially grateful to Brian Kubicki for reviewing an early draft of the manuscript and for his numerous comments and suggestions, which greatly improved its quality. We thank Jaime Bertoluci and Linda Trueb whose comments greatly improved this manuscript. EA thanks the Posgrado en Ciencias Biológicas for support of this study, the CONACyT for the students grant (CVU/Becario) 626946/330343, and the Programa de Innovación y Capital Humano para la Competitividad PINN-MICITT for the students grant (PED-0339-15-2). Laboratory work was funded by grant from PAPIIT-UNAM (IN203617) to GP-O. Fieldwork was partially supported by the National Geography Society (Grant number W-346-14). We acknowledge the Costa Rican Ministry of Environment and Energy (MINAE) for providing the corresponding scientific collecting permits for this expedition (SINAC-SE-GAS-PI-R 007- 2013 and 59-2015). References AmphibiaWeb. 2018. Information on Amphibian Biology and Conservation. Electronic Database accessible at http://amphibiaweb.org/. Berkeley, California, USA. Captured on 04 June 2018. Arias, E. and F. Bolaños. 2014. A checklist of the amphibians and reptiles of San Isidro de Dota, Reserva Forestal Los Santos, Costa Rica. Checklist 10: 870–877. Arias, E. and G. Chaves. 2014. 140 years after William M. Gabbs’s climb to Cerro Pico Blanco. Mesoamerican Herpetology 1: 176–180. Arias, E., G. Chaves, A. J. Crawford, and G. Parra-Olea. 2016. A new species of the Craugastor podiciferus species group (Anura: Craugastoridae) from the premontane forest of southwestern Costa Rica. Zootaxa 4132: 347–363. Bolaños, R., V. Watson, and J. Tosi. 2005. Mapa Ecológico de Costa Rica (Zonas de Vida), Según el Sistema de Clasificación de Zonas de Vida del Mundo de L. R. Holdridge, Escala 1:750 000. Vol. 214. San José. Centro Científico Tropical. Boza-Oviedo, E., S. M. Rovito, G. Chaves, A. García- Rodríguez, L. G. Artavia, F. Bolaños, and D. B. Wake. 2012. Salamanders from the eastern Cordillera de Talamanca, Costa Rica, with descriptions of five new species (Plethodontidae: Bolitoglossa, Nototriton, and Oedipina) and natural history notes from recent expeditions. Zootaxa 3309: 36–61. Campbell, J. A. 1999. Distribution patterns of amphibians in Middle America. Pp. 111–210 in W. E. Duellman (ed.), Patterns of Distribution of Amphibians: A Global Perspective. Baltimore and London. The Johns Hopkins University Press. Cope, E. D. 1875. On the Batrachia and Reptilia of Costa Rica. Journal Academy of Natural Sciences 2: 93–157. Crawford, A. J. and E. N. Smith. 2005. Cenozoic biogeography and evolution in direct-developing frogs of Central America (Leptodactylidae: Eleutherodactylus) as inferred from a phylogenetic analysis of nuclear and mitochondrial genes. Molecular Phylogenetics and Evolution 35: 536–555. Edgar, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32: 1792–1797. Duellman, W. E. and E. Lehr. 2009. Terrestrial-Breeding Frogs (Strabomantidae) in Peru. Münster. Natur- und Tier-Verlag. 382 pp. Foster, P. 2001. The potential negative impacts of global climate change on tropical montane cloud forest. Earth- Science Reviews 55: 73–106. Fouquet, A., A. Gilles, M. Vences, C. Marty, M. Blanc, and N. J. Gemmell. 2007. Underestimation of species richness in Neotropical frogs revealed by mtDNA analyses. PLoS ONE 2: e1109. Frost, D. R. 2018. Amphibian Species of the World: an Online Reference. Version 6.0. Electronic Database accessible at: http://research.amnh.org/herpetology/ amphibia/index.html. American Museum of Natural History, New York, USA. Captured on 04 June 2018. Gutiérrez-García, T. A. and E. Vázquez-Domínguez. 2013. Consensus between genes and stones in the biogeographic and evolutionary history of Central America. Quaternary Research 79: 311–324. Hedges, S. B., W. E. Duellman, and M. P. Heinicke. 2008. New World direct-developing frogs (Anura: Terra- rana): molecular phylogeny, classification, biogeography, and conservation. Zootaxa 1737: 1–182. Holdridge, L. R. 1967. Life Zone Ecology. San José. Tropical Science Center. 206 pp. Arias et al. 164 229 Phyllomedusa - 17(2), December 2018 Kubicki, B. 2008. Amphibian diversity in Guayacán, Limón province, Costa Rica. Brenesia 69: 35–42. Kumar, S., G. Stecher, and K. Tamura. 2016. MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for bigger datasets. Molecular Biology and Evolution 33: 1870–1874. Lanfear, R., B. Calcott, S. Y. Ho, and S. Guindon. 2012. PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Molecular Biology and Evolution 29: 1695– 1701. Lleonart, J., J. Salat, and G. J. Torres. 2000. Removing allometric effects of body size in morphological analysis. Journal of Theoretical Biology 205: 85–93. Meyer, C. P. 2003. Molecular systematics of cowries (Gastropoda: Cypraeidae) and diversification patterns in the tropics. Biological Journal of the Linnean Society 79: 401–459. Miller, M. A., W. Pfeiffer, and T. Schwartz. 2010. Creating the CIPRES Science Gateway for inference of large phylogenetic trees. Proceedings of the Gateway Computing Environments Workshop (GCE) 14 November 2010: 1–8. Olson, D. M., E. Dinerstein, E. D. Wikramanayake, N. D. Burgess, G. V. N. Powell, E. C. Underwood, J. A. D’amico, I. Itoua, H. E. Strand, J. C. Morrison, C. J. Loucks, T. F. Allnutt, T. H. Ricketts, Y. Kura, J. F. Lamoreux, W. W. Wettengel, P. Hedao, and K. R. Kassem. 2001. Terrestrial ecoregions of the world: a new map of life on earth. BioScience 51: 933–938. Palumbi, S., A. Martin, S. Romano, W. O. McMillan, L. Stice, and G. Grabowski. 1991. The simple fool’s guide to PCR, version 2.0. Honolulu. Special Publication, Department of Zoology and Kewalo Marine Laboratory, University of Hawaii. 43 pp. R Core Team. 2017. R: a Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Version 3.3.3. URL: http:// www.R-project.org. Rambaut, A., M. A. Suchard, W. Xie, and A. J. Drummond. 2014. Tracer. Version 1.6.1. URL: http://beast.bio.ed. ac.uk/Tracer. Ronquist, F., M. Teslenko, P. Van Der Mark, D. L. Ayres, A. Darling, S. Höhna, B. Larget, L. Liu, M. Suchard, and J. P. Huelsenbeck. 2012. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic Biology 61: 539–542. Sambrook, J. and D. W. Russell. 2006. Purification of nucleic acids by extraction with phenol: chloroform. Cold Spring Harbor Protocols 2006: pdb-prot4455. Santos-Barrera, G., J. Pacheco, F. Mendoza-Quijano, F. Bolaños, G. Chaves, and G. C. Daily. 2008. Diversity, natural history and conservation of amphibians and reptiles from the San Vito Region, southwestern Costa Rica. Revista de Biología Tropical 56: 755–778. Savage, J. M. 2002. The Amphibians and Reptiles of Costa Rica: A Herpetofauna Between Two Continents, Between Two Seas. Chicago. The University of Chicago Press. 934 pp. Sukumaran, J. and M. T. Holder. 2010a. SumTrees: Phylogenetic Tree Summarization. Version 4.4.0. URL: https://github.com/jeetsukumaran/DendroPy. Sukumaran, J. and M. T. Holder. 2010b. DendroPy: a Python library for phylogenetic computing. Bioinformatics 26: 1569–1571. Streicher, J. W., A. J. Crawford, and C. W. Edwards. 2009. Multilocus molecular phylogenetic analysis of the montane Craugastor podiciferus species complex (Anura: Craugastoridae) in Isthmian Central America. Molecular Phylogenetics and Evolution 53: 620–630. Tamura, K. 1992. Estimation of the number of nucleotide substitutions when there are strong transition– transversion and G + C–content biases. Molecular Biology and Evolution 9: 678–687. Zwickl, D. J. 2006. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. Unpublished Ph.D. Dissertation. The University of Texas, USA. Editor: Jaime Bertoluci A new species of Craugastor from Costa Rica 165 230 Phyllomedusa - 17(2), December 2018 Appendix I. Institutional voucher numbers, locality information, and GenBank accession numbers for the specimens used in the molecular phylogenetic analyses. Museum codes follow those of Frost (2018), with the addition of CRARC in reference to the Costa Rica Amphibian Research Center private collection and EAP denotes field numbers of Erick Arias. Species Institutional vouchers Collection locality Geographic coordinates Elevation (m a.s.l.) GenBank Number 16S COI C. aenigmaticus sp. nov. SMF: 104020 Jurutungo, Changuinola, Bocas del Toro, PA 8°54' N, 82°42' W 2388 MK211615 MK211577 C. aenigmaticus sp. nov. SMF: 104021 Jurutungo, Changuinola, Bocas del Toro, PA 8°55' N, 82°42' W 2330 MK211621 - C. aenigmaticus sp. nov. UCR: 21951 Cerro Utyum, Talamanca, Limón, CR 9°20' N, 83°10' W 2700 MK211616 MK211578 C. aenigmaticus sp. nov. UCR: 21952 Cerro Utyum, Talamanca, Limón, CR 9°20' N, 83°10' W 2690 MK211618 - C. aenigmaticus sp. nov. UCR: 21954 Cerro Utyum, Talamanca, Limón, CR 9°20' N, 83°10' W 2627 MK211620 - C. aenigmaticus sp. nov. UCR: 21956 Cerro Utyum, Talamanca, Limón, CR 9°20' N, 83°10' W 2690 MK211619 - C. aenigmaticus sp. nov. UCR: 22737 Cerro Hakú, Buenos Aires, Puntarenas, CR 9°19' N, 83°12' W 2660 MK211617 MK211579 C. aenigmaticus sp. nov. UCR: 22739 Cerro Hakú, Buenos Aires, Puntarenas, CR 9°19' N, 83°12' W 2660 MK211622 - C. bransfordii UCR: 22269 San Miguel, Alajuela, Alajuela, CR 10º18' N, 84º10' W 466 KT950295 MK211571 C. bransfordii UCR: 22643 Guayacán, Siquirres, Limón, CR 10º03' N, 83º32' W 537 MK211610 MK211572 C. gabbi UCR: 21863 Fila Costeña, Coto Brus, Puntarenas, CR 8º47' N, 82º57' W 1200 KT950271 MK211567 C. gabbi UCR: 21864 Fila Costeña, Coto Brus, Puntarenas, CR 8º47' N, 82º57' W 1200 KT950272 MK211568 C. lauraster SMF: 79759 Selva Negra, Matagalpa, Matagalpa, NI 12º59' N, 85º54' W 1300 MK211608 MK211565 C. lauraster USNM: 559393 Bodega del Río, Tapalwás, Gracias a Dios, HN 14º55' N, 84º32' W 150 KU323364 MK211566 C. persimilis UCR: 22211 Tausito, Paraiso, Cartago, CR 9º47' N, 83º45' W 1050 KT950293 MK211570 C. persimilis UCR: 22671 Suretka, Talamanca, Limón, CR 9º34' N, 82º56' W 121 MK211609 MK211569 C. podiciferus s.l. CRARC: 012 Volcán Turrialba, Turrialba, Cartago, CR 10º01' N, 83º42' W 2250 MK211633 MK211589 C. podiciferus s.l. CRARC: 247 Monte Verde, Tilarán, Guanacaste, CR 10º21' N, 84º48' W 1470 MK211645 - C. podiciferus s.l. EAP: 509 Fila Costeña, Golfito, Puntarenas, CR 8º47' N, 83º01' W 1546 - MK211605 C. podiciferus s.l. EAP: 810 Alto Uren, Talamanca, Limón, CR 9º21' N, 83º02' W 1860 MK211640 MK211596 C. podiciferus s.l. EAP: 817 Alto Uren, Talamanca, Limón, CR 9º23' N, 83º01' W 1500 MK211630 MK211586 Arias et al. 166 231 Phyllomedusa - 17(2), December 2018 Species Institutional vouchers Collection locality Geographic coordinates Elevation (m a.s.l.) GenBank Number 16S COI C. podiciferus s.l. FMNH: 257651 Fila Costeña, Coto Brus, Puntarenas, CR 8º47' N, 82º59' W 1350 EF562367 - C. podiciferus s.l. FMNH: 257669 Monte Verde, San Ramón, Alajuela, CR 10º18' N, 84º47' W 1500 EF562372 MK211598 C. podiciferus s.l. FMNH: 257671 Monte Verde, San Ramón, Alajuela, CR 10º18' N, 84º47' W 1500 EF562374 MK211599 C. podiciferus s.l. FMNH: 257673 Monte Verde, San Ramón, Alajuela, CR 10º18' N, 84º47' W 1500 EF562343 MK211603 C. podiciferus s.l. SMF: 104005 Changena, Changuinola, Bocas del Toro, PA 8º59' N, 82º40' W 1766 MK211641 MK211597 C. podiciferus s.l. UCR: 16353 Montaña Azul, Sarapiquí, Heredia, CR 10º12' N, 84º09' W 1500 EF562349 MK211602 C. podiciferus s.l. UCR: 16361 Tapesco, Alfaro Ruiz, Alajuela, CR 10º13' N, 84º22' W 1930 EF562371 - C. podiciferus s.l. UCR: 16585 Copey, Dota, San José, CR 9º32' N, 83º51' W 1400 MK211647 - C. podiciferus s.l. UCR: 19853 Lori, Talamanca, Limón, CR 9º21' N, 83º13' W 1817 MK211639 MK211595 C. podiciferus s.l. UCR: 19856 Lori, Talamanca, Limón, CR 9º21' N, 83º13' W 1817 MK211637 MK211593 C. podiciferus s.l. UCR: 19860 Lori, Talamanca, Limón, CR 9º21' N, 83º12' W 2108 MK211636 MK211592 C. podiciferus s.l. UCR: 19862 Lori, Talamanca, Limón, CR 9º21' N, 83º12' W 2108 MK211638 MK211594 C. podiciferus s.l. UCR: 20992 Nectandra, Alfaro Ruiz, Alajuela, CR 10º13' N, 84º20' W 2143 MK211632 MK211588 C. podiciferus s.l. UCR: 22091 Quebradas, Pérez Zeledón, San José, CR 9º26' N, 83º40' W 1488 MK211646 MK211604 C. podiciferus s.l. UCR: 22120 Altamira, Buenos Aires, Puntarenas, CR 9º19' N, 83º27' W 1821 MK211642 - C. podiciferus s.l. UCR: 22146 Cascajal, Coronado, San José, CR 10º01' N, 83º56' W 1700 MK211635 MK211591 C. podiciferus s.l. UCR: 22201 El Empalme, El Guarco, Cartago, CR 9º42' N, 83º56' W 2395 MK211634 MK211590 C. podiciferus s.l. UCR: 22226 Pico Blanco, Escazú, San José, CR 9º51' N, 84º08' W 2242 MK211644 MK211601 C. podiciferus s.l. UCR: 22228 Pico Blanco, Escazú, San José, CR 9º51' N, 84º08' W 2242 MK211643 MK211600 C. podiciferus s.l. UCR: 22675 Monte Verde, Puntarenas, Puntarenas, CR 10º19' N, 84º47' W 1726 - MK211606 C. podiciferus s.l. UCR: 22690 Chumacera, Pérez Zeledón, San José, CR 9º19' N, 83º28' W 1793 MK211631 MK211587 C. polyptychus UCR: 20050 Dabagri, Talamanca, Limón, CR 9º37' N, 83º16' W 900 MK211614 MK211576 C. polyptychus UCR: 22668 Bribri, Talamanca, Limón, CR 9º36' N, 82º54' W 198 MK211613 MK211575 A new species of Craugastor from Costa Rica Appendix I. Continued. 167 232 Phyllomedusa - 17(2), December 2018 Appendix II. Specimens used in the morphometric analysis. Museum collection acronyms follow Frost (2018), with the addition of EAP refers to Erick Arias field numbers. Craugastor aenigmaticus sp. nov. COSTA RICA: Limón: Cerro Utyum, Telire, Talamanca (UCR: 21951–21952, 22413– 22415); Valle del Silencio, Telire, Talamanca (UCR: 21194, 21931–21932). Puntarenas: Cerro Arbolado and Cerro Hakú, Buenos Aires, Buenos Aires (EAP: 758–763; UCR: 22731–22734, 22737, 22747). Craugastor podiciferus. COSTA RICA: aLajueLa: Poasito, Sabanilla, Alajuela (UCR: 21317–21319); Zarcero, Palmira, Alfaro Ruiz (UCR: 20992–20994); Los Alpes, Piedades Sur, San Ramón (UCR: 13961–13962). Santa María, Aguas Claras, Upala (UCR: 10528). Cartago: Coris, Quebradilla, Cartago (UCR: 3490–3491); Empalme, San Isidro, Guarco (UCR: 22200, 22202, 22204–22205); Tapantí, Orosi, Paraíso (UCR: 11542, 12010, 21708); Vereh, Chirripó, Turrialba (EAP: 748–750). Heredia: San José de la Montaña, Barva (UCR: 21299, 21352); Cerro Chompipe, Varablanca, Heredia (UCR: 18403–18404); La Legua, Varablanca, Heredia (UCR: 17464, 17478); Montaña Azul, Cureña, Sarapiquí (UCR: 16354). Limón: Fila Matama, Matama, Limón (UCR: 20177, 20213); Cerro Pat, Telire, Talamanca (EAP: 0792, 801–803, 807, 810, 814, 817–818); Sabanas Dúrika, Telire, Talamanca (UCR: 21662, 21682, 21825, 21836); Transtalamanca, Telire, Talamanca (UCR: 19849, 19853, 19860, 19870, 19874, 19895, 19933). Puntarenas: Cerro Quemado, Biolley, Buenos Aires (UCR: 21191, 21700); Las Cruces, San Vito, Coto Brus (UCR: 12934, 12936, 13239); Las Tablas, Sabalito, Coto Brus (UCR: 8379, 8381–8382, 10738, 10747– 10748, 12560); Monte Verde, Puntarenas (UCR: 13646–13647, 17243, 21718, 22674–22676). san josé: Pico Blanco, San Antonio, Escazú (UCR: 22228); Tinamaste, Barú, Pérez Zeledón (UCR: 14510, 14513, 14518); Quebradas, San Isidro del General, Pérez Zeledón (EAP: 738–740; UCR: 16359–16360, 22606); Chumacera, San Pedro, Pérez Zeledón (EAP: 721–723; UCR: 22119); Cascajal, Vazquez de Coronado (UCR: 16090, 16092, 21918). Craugastor sp.1. COSTA RICA: Puntarenas: Las Alturas, Pittier, Coto Brus (UCR: 22703–22704, 22709–22710); Tres Colinas, Potrero Grande, Buenos Aires (EAP: 725, 727–729; UCR: 20257–20258, 20389, 20395, 20401, 20411, 20419, 20421, 20423, 20428); road to Las Tablas, Sabalito, Coto Brus (EAP: 0823). Species Institutional vouchers Collection locality Geographic coordinates Elevation (m a.s.l.) GenBank Number 16S COI C. stejnegerianus EAP: 0514 Palmar Norte, Osa, Puntarenas, CR 8º57' N, 83º26' W 45 MK211607 MK211563 C. stejnegerianus UCR: 20352 Potrero Grande, Buenos Aires, Puntarenas, CR 9º05' N, 83º06' W 900 KT950284 MK211564 C. underwoodi UCR: 22619 Tapantí, Paraíso, Cartago, CR 9º45' N, 83º46' W 1412 MK211611 MK211573 C. underwoodi UCR: 22625 Cascajal, Coronado, San José, CR 10º01' N, 83º56' W 1708 MK211612 MK211574 Craugastor sp.1 UCR: 20389 Potrero Grande, Buenos Aires, Puntarenas, CR 9º06' N, 83º06' W 1500 MK211625 MK211581 Craugastor sp.1 UCR: 22709 Las Alturas, Coto Brus, Puntarenas, CR 8º58' N, 82º49' W 1980 MK211626 MK211582 Craugastor sp.2 SMF: 104014 Cerro Saguí, Nole Duima, Ngöbe Buglé, PA 8º33' N, 81º49' W 1762 MK211623 - Craugastor sp.2 SMF: 104015 La Nevera, Nole Duima, Ngöbe Buglé, PA 8º30' N, 81º46' W 1700 MK211624 MK211580 Craugastor sp.B SMF: 102024 Fortuna, Gualaca, Chiriquí, PA 8º40' N, 82º11' W 1730 MK211627 MK211583 Craugastor sp.B SMF: 104023 Fortuna, Gualaca, Chiriquí, PA 8º40' N, 82º12' W 1280 MK211628 MK211584 Craugastor sp.B SMF: 104027 Volcán Barú, Bugaba, Chiriquí, PA 8º50' N, 82º30' W 2134 MK211629 MK211585 Arias et al. Appendix I. Continued. 168 DISCUSIÓN Y CONCLUSIONES GENERALES La importancia de las tierras altas de América Central Ístmica Este trabajo contiene la mejor representatividad geográfica para cualquier grupo de anfibios en ACI, incluyendo todas las localidades tipo. Esto permitió obtener resultados sólidos sobre la distribución de las especies, pero en especial de la importancia de las tierras altas de Costa Rica y Panamá en la diversificación del grupo de especies Craugastor podiciferus. Todos los linajes encontrados contienen poblaciones por encima de los 750 m de elevación en las tierras altas de Costa Rica y el oeste de Panamá. Lo anterior apoya resultados previos que sugieren que las tierras altas de ACI tuvieron un papel importante en la diversificación de la biota regional (García-París et al. 2000; Savage 2002; Crawford et al. 2007; Boza-Oviedo et al. 2012). Los resultados obtenidos son similares a los hallazgos previos sobre la importancia de la diversificación en tierras altas de ACI . No obstante, existe controversia en la datación cronológica de los procesos responsables de esta diversificación. Savage (2002), con base en la evidencia de oscilaciones climáticas y asumiendo un rápido levantamiento de las tierras altas de América Central en el Plioceno-Pleistoceno (~5 Ma), propuso un modelo de especiación de tierras altas (especiación montana). Este autor sugirió que las oscilaciones entre periodos glaciares e interglaciares que comenzaron en el Plioceno (~5 Ma) y continuaron durante el Pleistoceno, son las responsables del origen y distribución actual de muchas especies de tierras altas. Durante periodos glaciares los bosques montanos estuvieron conectados, y periodos subsecuentes de calentamiento llevaron al aislamiento de estos bosques en las cumbres de las cordilleras en América Central, lo cual originó muchas de las especies actuales (Savage 2002). Sin embargo, los resultados encotrados en este estudio junto con los encontrados por Castoe et 169 al. (2009) y Streicher et al. (2009) rechazan que estos eventos del Plioceno-Pleistoceno estén asociados con cladogénesis dentro de especies de altura. Castoe et al. (2009) encontraron que la diversidad de serpientes de foseta neotropicales en las tierras altas de América Central, es mejor explicada por eventos tectónicos durante el Mioceno-Plioceno. Aunque esto coincide parcialmente (Plioceno) con la propuesta de Savage (2002), lo cierto es que definitivamente la diversificación es anterior al Pleistoceno. La datación cronológica del origen del grupo de especies C. podiciferus realizada en este trabajo fue similar a aquella obtenida por Streicher et al. (2009), quienes calcularon el origen del mismo en aproximadamente 20 Ma. Asumiendo un origen temprano durante el mioceno para ACI (Montes et al. 2012a,b, 2015). Nuestros datos sugieren que la llegada a ACI del ancestro común del grupo de especies C. podiciferus corresponde con las primeras colonizaciones de ACI, previo al cierre del Istmo de Panamá. Este momento histórico en América Central no debe ser entendido como un evento aislado, sino, como pulsos discontinuos de dispersión de la fauna ya presente en América Central Nuclear hacia el sur colonizando las tierras de Costa Rica y Panamá a medida que estas se levantaban. Es probable que estos pulsos de dispersión hayan sido separados por altos niveles del mar, lo cual aisló temporalmente estas dos unidades de América Central propiciando la evolución de diferentes grupos (Castoe et al. 2009; Gutiérrez-García & Vázquez-Domínguez 2013). Posiblemente el inicio de estos pulsos de dispersión se remonta al Mioceno Temprano (~23 Ma), cuando ACI había emergido y se prolongaba como una península en su posición actual (Montes et al. 2012a). Estos primeros eventos de dispersión corresponden con la aparición del grupo de especies C. podiciferus hace ~20 Ma (Streicher et al. 2009). Otros grupos apoyan esta primera dispersión desde América Central Nuclear hacia ACI. Duellman et al. (2016) 170 encontraron que el origen y diversificación del género de ranas arborícolas Isthmohyla en ACI se remonta a ~19.5 Ma. Říčan et al. (2013) encontraron que los peces dulceacuícolas de ACI llegaron desde el norte entre 19–13 Ma. Rovito et al. (2015) sugirieron que las salamandras neotropicales de la tribu Bolitoglossine se dispersaron hacia el sur, y que hace 16.4–8 Ma arribaron a Sudamérica, por lo cual debieron llegar a ACI previamente. Aunque el presente trabajo contiene una amplia representación geográfica para el grupo de especies C. podiciferus, es necesario continuar con el trabajo de campo y la recolección de muestras, especialmente en las tierras altas de Costa Rica y el oeste de Panamá. Es posible que otras especies permanecen sin describir en sitios remotos de las cordilleras, lo anterior dado que es posible que las especies de alturas tengan áreas de distribución muy restringidas como el caso de Craugastor sp. Chumacera y Craugastor sp. Siola. En particular, es necesario realizar esfuerzos intensivos en las tierras altas de la vertiente pacífica de la cordillera de Talamanca, dado que C. podiciferus parece no distribuirse sobre el pacífico de Talamanca y que existe alta estructuración con varias especies distintas. Comentarios taxonómicos al grupo de especies Craugastor podiciferus Este es el primer estudio filogenético con datos moleculares para el grupo de especies C. podiciferus en América Central, ya que estudios anteriores analizaron algunos clados (Crawford 2003; Streicher et al. 2009). Como se había sugerido previamente, se encontró gran diversidad de especies no descritas y enmascaradas por los nombres actuales dentro del grupo. Al inicio del estudio, el grupo contenía ocho especies (Padial et al. 2014) y al final de este se detectaron hasta 23 linajes independientes dentro del grupo. Con algunas excepciones (C. aenigmaticus) la gran mayoría de estos grupos han pasado desapercibidos en colecciones científicas por muchos años, 171 resaltando la complejidad de delimitar morfológicamente las especies y reconocer aquellas distintas. Este trabajo incluyó marcadores mitocondriales y una gran cantidad de SNPs nucleares, siendo el más extensivo para el grupo de especies C. podiciferus, y para el género Craugastor, hasta la fecha. El empleo de ambos marcadores en combinación con métodos de delimitación molecular permitió dilucidar la diversidad presente en el grupo, mucha de la cual ha sido enmascrada por los nombres actuales. En el presente estudio se describieron cuatro especies nuevas (C. aenigmaticus, C. gabbi, C. sagui y C. zunigai), se sinonimizaron tres especies (C. jota, C. lauraster y C. polyptychus) y dos más fueron resucitadas (C. blairi y C. rearki). Otras especies necesitan ser nombradas y descritas; sin embargo, estas deben estar acompañadas de exhaustivos análisis morfológicos que respalden su distinción y permitan una correcta diagnosis. El principal aporte de este trabajo es la comprensión de las relaciones filogenéticas dentro del grupo de especies C. podiciferus. Se encontró que el grupo está formado por cuatro clados, tres de los cuales coinciden con los tres grupos morfológicos que se mencionaron previamente (Savage 2002). Un grupo adicional, el más basal según la filogenia nuclear, fue identificado y nombrado (C. aenigmaticus). Esta especie es la más divergente del grupo y no había sido previamente colectada, es decir, no estuvo enmascarada por alguna de las especies nombradas. C. aenigmaticus es una especie microendémica a la Cordillera de Talamanca en el extremo sureste de Costa Rica, conocida únicamente entre los 2390–2650 m de elevación. Un segundo gran clado contiene dos subclados, el primero está formado por tres especies distribuidas en el extremo sureste de Costa Rica y oeste de Panamá en la Cordillera de Talamanca y Cordillera Central de Panamá. Estas tres especies están distribuidas alopátridamente sobre el eje de las Cordilleras. Dos de estas especies fueron descritas en el 172 presente estudio, C. sagui y C. zunigai, una especie adicional corresponde con C. blari (Barbour, 1928) nombre que estaba bajo sinonimia de C. podiciferus y que fue resucitado en este trabajo. Un segundo subclado contiene a C. podiciferus y una serie de grupos estrechamente relacionados a esta. Con la obtención de material topotípico de C. podiciferus se logró restringir a C. podiciferus sensu stricto a las poblaciones de la Cordillera Volcánica Central de Costa Rica y la Cordillera de Talamanca en Costa Rica y el oeste de Panamá, en esta última restringida a la vertiente caribeña. Esta especie corresponde con los clados C y D de Streicher et al. (2009). El subclado conteniendo a C. podiciferus incluye una serie de grupos que necesitan ser evaluados morfológicamente, ya que posiblemente representen al menos seis especies no nombradas. El grupo Craugastor sp. Monte Verde corresponde con el clado A de Streicher et al. (2009), un grupo restringido a la Cordillera de Tilarán y Volcánica Central. Craugastor sp. San Gerardo corresponde con el grupo B de Streicher et al. (2009), distribuido en la Cordillera de Tilarán, Volcánica Central y Valle Central. En la zona de Monte Verde los grupos Craugastor sp. Monte Verde y Craugastor sp. San Gerardo están en simpatría potencial por la cercanía de sus localidades, sin embargo, no se conocen las dos especies en el mismo punto. Lo mismo sucede en Zarcero entre los grupos C. podiciferus y Craugastor sp. Monte Verde. Craugastor sp. Fila Costeña corresponde con los clados E y F de Streicher et al. (2009), un grupo restringido al pacífico sur de Costa Rica en la Fila Costeña y las faldas de Talamanca. Adicionalmente, se encontraron tres grupos que no fueron incluidos por Streicher et al. (2009). Craugastor sp. Chumacera conocido de una única población en el pacífico de la Cordillera de Talamanca. Craugastor sp. Pico Blanco, restringido al Cerro Pico Blanco en el Valle Central de Costa Rica. Y Craugastor sp. Siola también restringido a una única población, pero en el caribe de la Cordillera de Talamanca. 173 La especie C. jota, fue atribuida al grupo de especies C. podiciferus por Hedges et al. (2008) y basado en la morfología esta debe pertenecer al clado C. podiciferus. Esta especie fue descrita con especímenes provenientes del río Changena en el oeste de Panamá sobre los 760 m, colectados por Linda Trueb en 1966. La localidad tipo dada por Lynch (1980) se basa en el mapa mostrado por Trueb (1968), sin embargo, esta localidad es discutida, ya que se basa en un mapa de poca calidad y con nombres de ríos que en la actualidad no coinciden. Aquí se incluyen especímenes provenientes de “Río Changena”, muy cerca de la localidad tipo de C. jota (Linda Trueb comunicación personal), no obstante, estos son conespecíficos con C. podiciferus sensu stricto. Por lo tanto, se decidió sinonimizar a C. jota bajo C. podiciferus. Un tercer gran clado contiene las especies relacionadas a C. bransfordii, conformado por seis especies, aunque solo dos están nombradas. La especie C. bransfordii fue descrita con especímenes del Río San Juan en la región limítrofe entre Costa Rica y Nicaragua, aquí se incluye un espécimen cuya localidad es muy cercana a la posible localidad tipo. De esta manera se pudo restringir C. bransfordii al sur de Nicaragua y el Caribe norte y central de Costa Rica. La especie nominal C. polyptychus fue descrita con especímenes de Río San Juan, Nicaragua, la misma localidad tipo de C. bransfordii. Savage (2002) resucitó este nombre de la sinonimia de C. bransfordii para asignarlos a las poblaciones del Caribe norte y sur de Costa Rica, sin embargo, mencionó que es probable que estas poblaciones deban recibir otro nombre. Dado que en la región del Río San Juan (límite entre Costa Rica y Nicaragua) no se encontraron dos clados distintos dentro del gran clado C. bransfordii se decidió sinonimizar el nombre C. polyptychus bajo C. bransfordii. La especie Craugastor sp. Fila Carbón corresponde –en parte– al C. polyptychus de Savage (2002), esta especie está conformada por especímenes provenientes del 174 Caribe Sur de Costa Rica y extremo oeste de Panamá, con una distribución de tierras bajas a intermedias. Craugastor underwoodi fue descrita con especímenes de Bajo La Hondura, Costa Rica, aquí se incluyen especímenes de Cascajal y Cinchona dos localidades geográfica y biológicamente cercanas a la localidad tipo. Craugastor underwoodi se distribuye en la región premontana de la Cordillera Volcánica Central, Cordillera de Tilarán y extremo norte de la Cordillera de Talamanca. La especie Craugastor sp. Quebradas incluye la única localidad en el Pacífico costarricense para el clado C. bransfordii. La especie Craugastor sp. Vereh es representada por dos poblaciones en el Caribe Central de Costa Rica, su gran distancia genética con respecto a los demás grupos y su cambio en la posición filogenética entre las filogenias mitocondrial y nuclear respalda su validez como nueva especie. Craugastor sp. Panamá está conformada por poblaciones de Panamá, poblaciones atribuida por Leenders (2016) a C. bransfordii, su validez como nueva especie es también apoyada por su cambio en la posición filogenética entre las filogenias mitocondrial y nuclear. El cuarto gran clado contiene a las especies relacionadas a C. stejnegerianus, conformado por seis especies, cuatro de estas nombradas. Craugastor persimilis agrupa especímenes de varias localidades en el Caribe Central y Sur de Costa Rica. Este grupo incluye la localidad tipo de C. persimilis, Suretka, Talamanca. Un caso muy particular es C. rearki (Taylor, 1952), esta fue descrita del Caribe de Costa Rica, Savage & Emerson (1970) la asignaron bajo sinonimia de C. bransfordii. En este trabajo el nombre C. rearki se asigna a las poblaciones del caribe central y norte de Costa Rica (incluyendo una localidad de donde se asignaron paratipos para C. rearki), pero además, están incluidos especímenes de Nicaragua y Honduras cercanas a la localidad tipo 175 de C. lauraster. Bajo el principio de prioridad debe prevalecer C. rearki y sinonimizar a C. lauraster bajo el primero. Las poblaciones del clado C. stejnegerianus del pacífico costarricense son altamente conflictivas, dado los cambios significativos en las posiciones filogenéticas entre las filogenias mitocondrial y nuclear. Basado en la filogenia mitocondrial y la evidencia morfológica se respalda a C. gabbi, la cual fue recientemente descrita (Arias et al. 2016) y distribuida en el hábitat premontano del Pacífico Sur de Costa Rica y el oeste de Panamá. Sin embargo, en la filogenia nuclear C. gabbi es el clado hermano a C. stejnegerianus sensu stricto, por lo cual, si se reconoce a C. gabbi como especie válida entonces se deben reconocer a Craugastor sp. Quepos y Craugastor sp. Neilly. La primera restringida a las poblaciones del Pacífico Central y el Valle Central de Costa Rica, constituyendo el clado hermano a C. rearki. Craugastor sp. Neilly está restringida a el extremo sureste del pacífico costarricense y el clado hermano a C. gabbi + C. stejnegerianus sensu stricto, el último restringido al pacífico sur de Costa Rica. El presente trabajo constituye una primera aproximación al entendimiento de las relaciones filogenéticas y los patrones biogeográficos para el grupo de especies C. podiciferus; sin embargo, persisten grandes retos por resolver en el futuro. Los resultados de este trabajo demostraron que el grupo C. podiciferus está compuesto por cuatro grandes clados, no obstante, solo el grupo basal y un subclado del segundo fueron taxonómicamente resueltos. Es necesario realizar una revisión morfológica exhaustiva de C. podiciferus con el fin de delimitar morfológicamente las distintas especies que permanecen enmascaradas por este nombre y que fueron identificadas en este estudio. Dentro del clado C. bransfordii cuatro especies carecen de nombre y formal descripción. La experiencia previa sugiere que las especies dentro del clado C. bransfordii son morfológicamente muy conservadas, por lo cual, es necesario realizar un 176 exhaustivo análisis morfológico que permita una correcta diagnosis y separación de las especies nombradas. Finalmente, dentro del clado C. stejnegerianus, es necesario determinar el estatus de los grupos asociados a C. stejnegerianus del pacífico costarricense, ya que estos fueron altamente diferentes entre la filogenia mitocondrial y la nuclear. El uso de RAD-seq permitó obtener una gran cantidad de SNPs nucleares para el grupo de especies C. podiciferus. Esto permitió obtener filogenias robustas para el grupo, pero también, hipótesis mejor apoyadas para la delimitación de las especies dentro del grupo de estudio. Es importante resaltar que la técnica utilizada ddRAD-seq (Peterson et al. 2012; Leaché et al. 2015), fue altamente eficiente en la obtención de gran cantidad de loci homólogos para este grupo de organismos no-modelo. Estos SNPs permitieron una adecuada delimitación de las especies, dado que estos marcadores son considerados una fuente de evidencia independiente a aquella obtenida con los marcadores mitocondriales. Además, la técnica de ddRAD-seq tiene como ventaja el bajo costo relativo a la cantidad de secuencias obtenidas y el poco tiempo necesario para la obtención de las mismas. Sin embargo, los datos obtenidos con RAD-seq también presentan una serie de limitaciones, como por ejemplo dado que los SNPs son de longitud corta (~100 pb) son inadecuados para ciertos métodos de delimitación molecular como BPP, donde no se pueden tratar como loci independientes. Asimismo, la longitud de los SNPs es un impedimento para la otención de árboles de máxima credibilidad a través de BEAST. Delimitación y especies crípticas Hasta la fecha ninguna de las especies del grupo C. podiciferus puede ser considerada críptica bajo el concepto sensu stricto, es decir, morfológicamente idénticas (Pérez-Ponce de León & Nadler 2010). Sin embargo, en la práctica estas han sido tratadas como crípticas y su 177 delimitación se ha basado en análisis moleculares que permitieron delimitar grupos, que posteriormente fueron revisados morfológicamente y permitió identificar características diagnósticas para todas las especies analizadas según lo sugerido por Pérez-Ponce de León & Nadler (2010). Es necesario continuar realizando análisis morfológicos para todos los clados identificados, con el fin de determinar si existen caracteres diagnósticos para todas las especies dentro del grupo. En especial para aquellas dentro del grupo C. bransfordii, las cuales son morfológicamente muy similares (sin realizar análisis exhaustivos hasta la fecha). Pero también, está latente la posibilidad de que existan especies crípticas sensu stricto dentro del grupo C. podiciferus, lo cual sería un hallazgo novedoso de un fenómeno poco documentado. El empleo de marcadores moleculares es fundamental cuando se trata con especies crípticas, al menos en el sentido funcional, es decir, aquellas que el taxónomo no puede reconocer en su tratamiento actual (Pérez-Ponce de León & Nadler 2010), dado que una de las propiedades intrínsecas de las especies crípticas es el conservadurismo morfológico. Sin embargo, desde una perspectiva de taxonomía integradora (Dayrat 2005; Padial et al. 2010) son necesarias al menos dos fuentes de evidencia independientes para el reconocimiento de separación de linajes. Lo anterior no debe ser un problema cuando se trata de especies que no son hermanas, por ejemplo, cuando se debe describir una especie para evitar la parafilia de especies. Ya que en el caso anterior la posición filogenética de la especie es evidencia suficiente para reconocer su distintividad como linaje evolucionando independientemente (de Queiroz 2007; Padial et al. 2010). Un ejemplo fue la descripción de C. gabbi, cuyas poblaciones fueron históricamente referidas como C. stejnegerianus, sin embargo, análisis mitocondriales demostraron que C. gabbi es el grupo hermano al clado formado por C. stejnegerianus + C. 178 lauraster. En el ejemplo anterior, se logró identificar un carácter diagnóstico que permite la diferenciación de la recién nombrada especie de las demás dentro del grupo. Es importante tomar en cuenta la diferencia entre caracteres diagnósticos e historia evolutiva única (Frost & Kluge 1994). Por ejemplo, en C. gabbi la principal evidencia apoyando la historia evolutiva única de esta es su posición filogenética. Una alternativa para evitar la parafilia era reconocer como una sola especie a todo el clado que contiene a C. gabbi, C. lauraster y C. stejnegerianus. Una opción que debe ser formalmente considerada y evaluada antes de nombrar alguna nueva especie, no obstante, es necesario evitar especies altamente variables en múltiples caracteres fenotípicos. Resulta complicado tomar la decisión de nombrar una nueva especie la cual es muy similar a su especie hermana, ya que en este caso la similitud morfológica es mejor explicada como el resultado de la escasa divergencia evolutiva desde el proceso de especiación. En este caso, la posición filogenética no es suficiente para justificar la presencia de dos especies hermanas morfológicamente muy similares (Castroviejo-Fisher et al. 2017). Dado que la estructura jerárquica de variación de caracteres que típicamente se observa durante la especiación, puede ser simulado por procesos como variación geográfica dado a aislamiento por distancia o por selección/deriva génica en poblaciones pequeñas (Padial et al. 2010; Castroviejo- Fisher et al. 2017). En estos casos es necesario otros tipos de evidencia que apoyen la separación de linajes, como encontrar ambas especies en simpatría, divergencia de cantos reproductivos o la combinación de marcadores mitocondriales y nucleares (Padial et al. 2010; Castroviejo-Fisher et al. 2017). Un ejemplo de este caso fue el reconocimiento de C. blairi y C. zunigai como especies válidas a pesar de que son especies hermanas y que era también válido reconocer solamente una especie sin problemas de parafilia. En este caso se logró identificar caracteres morfológicos que 179 permiten la diferenciación de ambas especies, lo cual apoya la hipótesis de que ambas especies poseen una trayectoria evolutiva única. En el caso hipotético donde ninguna evidencia adicional, más que la filogenética, apoye la divergencia de linajes entre dos clados hermanos, lo mejor es reconocer una sola especie ya que se podría caer en inflación taxonómica (Isaac et al. 2004). Lo anterior es fundamental ante el escenario de sobreestimación de especies por los métodos de delimitación molecular (Sukumaran & Knowles 2017). Los resultados encontrados en este estudio sugieren que varios de los métodos de delimitación molecular sobreestiman la diversidad real dentro de un grupo, por lo cual, es necesario la utilización de al menos dos fuentes de evidencia para dividir una especie en dos o más. En este sentido, es válida la recomendación de ser cauteloso con los resultados generados por métodos de delimitación molecular, estos deben ser entendidos como prospección molecular, mas no como evidencia irrefutable de separación de linajes. Pueden ser utilizados como evidencia preliminar para realizar otra serie de análisis (morfológico, acústico, genético, etc) que apoyen la divergencia de especies. Por último, es necesario realizar estudios con las especies del grupo C. podiciferus que permitan un mejor entendimiento de la historia natural de estas especies. A la fecha se conoce poco sobre los miembros del grupo, a pesar de que estos son relativamente abundantes y fáciles de encontrar. Por ejemplo, aunque posiblemente todas las especies del grupo realizan cantos de aviso (Savage 2002), este solo ha sido detalladamente descrito para dos especies (Schlaepfer & Figeroa-Sandi 1994; Twining & Cossel 2017) y recientemente descrito para otras dos (Cossel et al. En prensa). Adicionalmente, se desconoce la alimentación, los depredadores, uso de hábitat, picos de actividad, épocas reproductivas, cortejo, etc. Es posible que en el pasado el grupo haya 180 sido poco estudiado dada la gran complejidad para discernir entre especies y el alto grado de polimorfismos fenotípicos intrapoblacional que sugería que más de una especie estaba presente. La presente contribución aporta información sobre las relaciones filogenéticas y la delimitación geográfica de las especies del grupo, que puede servir para estudios futuros donde se ahonde en la biología particular de cada especie. 181 Conclusiones ● El presente trabajo contiene la mayor representación geográfica y la mayor cantidad de datos moleculares para el grupo Craugastor podiciferus a la fecha, y representa uno de los trabajos sistemáticos más exhaustivos dentro de la familia Craugastoridae. ● El grupo de especies C. podiciferus contiene diversidad considerable de especies bajo los nombres actuales, los cuales están conformados por al menos 23 especies distintas. ● Aunque existe alto conservadurismo morfológico dentro del grupo de especies C. podiciferus, se logró reconocer cuatro especies nuevas y se redefinieron otras ya descritas, incluyendo una adecuada diagnosis y comparación entre especies. ● El origen biogeográfico del grupo de especies C. podiciferus corresponde con las tierras altas de Costa Rica y el oeste de Panamá, área donde se distribuyen todos los linajes actualmente conocidos. ● Las especies del grupo C. podiciferus ameritan claras estrategias de conservación, ya que a pesar de que son relativamente abundantes sus áreas de distribución son restringidas. 182 LITERATURA CITADA AmphibiaWeb (2018) AmphibiaWeb: Information on Amphibian Biology and Conservation [web application]. Berkeley, California. Disponible en: http://amphibiaweb.org (accesado 10 de enero de 2019). Anger, G.R. & Dean, R. (2010) The birds of Panama: a field guide. Cornell University Press, Nueva York. Arias, E., Chaves, G., Crawford, A.J. & Parra-Olea, G. (2016) A new species of the Craugastor podiciferus species group (Anura: Craugastoridae) from the premontane forest of southwestern Costa Rica. Zootaxa, 4132: 347–363. Bacon, C.D., Molnar, P., Antonelli, A., Crawford, A.J., Montes, C. & Vallejo-Pareja, M.C. (2016) Quaternary glaciation and the Great American Biotic Interchange. Geology, 44: 375–378. Bacon, C.D., Silvestro, D., Jaramillo, C., Smith, B.T., Chakrabarty, P. & Antonelli, A. (2015) Biological evidence supports an early and complex emergence of the Isthmus of Panama. Proceedings of the National Academy of Sciences, 112: 6110–6115. Bagley, J.C. & Johnson, J.B. (2014) Phylogeography and biogeography of the lower Central American Neotropics: diversification between two continents and between two seas. Biological Reviews, 89: 767–790. Baird, N.A., Etter, P.D., Atwood, T.S., Currey, M.C., Shiver, A.L., Lewis, Z.A., Selker, E.U., Cresko, W.A. & Johnson, E.A. (2008) Rapid SNP discovery and genetic mapping using sequenced RAD markers. PLoS ONE, 3: e3376. Barbour, T. (1926) New Amphibia. Occasional Papers of the Boston Society of Natural History, 5: 191–194. 183 Barbour, T. (1928) New Central American frogs. Proceedings of the New England Zoölogical Club, 10: 25–31. Bickford, D., Lohman, D.J., Sodhi, N.S., Ng, P.K., Meier, R., Winker, K., Ingram, K.K. & Das, I. (2007) Cryptic species as a window on diversity and conservation. Trends in Ecology & Evolution, 22: 148–155. Boulenger, G.A. (1896) Descriptions of new batrachians collected by Mr. C. G. Underwood in Costa Rica. Annals and Magazine of Natural History, Series 6(18): 340–342. Boza-Oviedo, E., Rovito, S.M., Chaves, G., García-Rodríguez, A., Artavia, L.G., Bolaños, F. & Wake, D.B. (2012) Salamanders from the eastern Cordillera de Talamanca, Costa Rica, with descriptions of five new species (Plethodontidae: Bolitoglossa, Nototriton, and Oedipina) and natural history notes from recent expeditions. Zootaxa, 3309: 36–61. Campbell, J.A. (1999) Distribution patterns of amphibians in Middle America. In: Duellman, W.E. (Ed.), Patterns of Distribution of Amphibians: A Global Perspective. Johns Hopkins University Press, Maryland, pp. 111–210. Castoe, T.A., Daza, J.M., Smith, E.N., Sasa, M.M., Kuch, U., Campbell, J.A., Chippindale, P.T. & Parkinson, C.L. (2009) Comparative phylogeography of pitvipers suggests a consensus of ancient Middle American highland biogeography. Journal of Biogeography, 36: 88–103. Castroviejo-Fisher, S., Köhler, J., De La Riva, I. & Padial, J.M. (2017) A new morphologically cryptic species of Phyllomedusa (Anura: Phyllomedusidae) from Amazonian forests of northern Peru revealed by DNA sequences. Zootaxa, 4269: 245–264. Chen, S.H. (2001) Cytogenetic study of the lower Central American frogs of the subgenus Craugastor (Anura: Leptodactylidae: Eleutherodactylus). University of Florida, Coral Gables, Fla. Ph.D. dissertation, 738 pp. 184 Chen, S.H. (2005) Chromosomal variation in the rhodopis Group of the southern Central American Eleutherodactyline frogs (Leptodactylidae: Eleutherodactylus). In. Donnelly, M.A., Crother, B.I., Guyer, C., Marvalee H.W. & White, M.E. (Eds.), Evolution and ecology in the tropics. The University of Chicago Press, Chicago and London, pp. 102– 144. Coates, A.G. & Obando, J.A. (1996) The geologic evolution of the Central American Isthmus. In: Jackson, J.B.C., Budd, A.F. & Coates, A.G. (Eds.), Evolution and environment in tropical America. The University of Chicago Press, Illinois, pp. 21–56. Coates, A.G., Collins, L.S., Aubry, M.P. & Berggren, W.A. (2004) The geology of the Darien, Panama, and the late Miocene-Pliocene collision of the Panama arc with northwestern South America. Geological Society of America Bulletin, 116: 1327–1344. Cope, E.D. (1875) On the Batrachia and Reptilia of Costa Rica. Journal of the Academy of Natural Sciences of Philadelphia, Series 2(8): 93–154. Cope, E.D. (1886) Thirteenth contribution to the herpetology of tropical America. Proceedings of the American Philosophical Society, 23: 271–287. Cope, E.D. (1893) Second addition to the knowledge of the Batrachia and Reptilia of Costa Rica. Proceedings of the American Philosophical Society, 31: 333–347. Crawford, A.J. (2003) Huge populations and old species of Costa Rican and Panamanian dirt frogs inferred from mitochondrial and nuclear gene sequences. Molecular Ecology, 12: 2525–2540. Crawford, A.J. & Smith, E.N. (2005) Cenozoic biogeography and evolution in direct-developing frogs of Central America (Leptodactylidae: Eleutherodactylus) as inferred from a 185 phylogenetic analysis of nuclear and mitochondrial genes. Molecular Phylogenetics and Evolution, 35: 536–555. Crawford, A.J., Bermingham, E. & Carolina, P.S. (2007) The role of tropical dry forest as a long‐term barrier to dispersal: a comparative phylogeographical analysis of dry forest tolerant and intolerant frogs. Molecular Ecology, 16: 4789–4807. Davis, S.D., Heywood, V.H., Herrera-MacBryde, O., Villa-Lobos, J. & Hamilton, A. (1997) Centres of Plant Diversity: a Guide and Strategy for Their Conservation, Volume 3: The Americas. IUCN Publications Unit, Cambridge. Dayrat, B. (2005) Towards integrative taxonomy. Biological Journal of the Linnean Society, 85: 407–415. de Queiroz, K. (1998) The general lineage concept of species, species criteria, and the process of speciation. A conceptual unification and terminological recommendations. In: Howard, D.J. & Berlocher, S.H. (Eds.), Endless forms: Species and speciation. Oxford University Press, Oxford, pp. 57–75 de Queiroz, K. (2007) Species concepts and species delimitation. Systematic Biology, 56: 879‒ 886. Duellman, W.E., Marion, A.B. & Hedges, S.B. (2016) Phylogenetics, classification, and biogeography of the treefrogs (Amphibia: Anura: Arboranae). Zootaxa, 4104: 1–109. Ence, D.D. & Carstens, B.C. (2011) SpedeSTEM: a rapid and accurate method for species delimitation. Molecular Ecology Resources, 11: 473–480. Flot, J.F. (2015) Species delimitation's coming of age. Systematic Biology, 64: 897–899. 186 Fouquet, A., Gilles, A., Vences, M., Marty, C., Blanc, M. & Gemmell, N.J. (2007) Underestimation of species richness in Neotropical frogs revealed by mtDNA analyses. PLoS ONE, 2: e1109. Frost, D.R. 2019. Amphibian Species of the World: an Online Reference. Version 6.0. Disponible en: http://research.amnh.org/herpetology/amphibia/index.html. American Museum of Natural History, New York, USA. (accesado 10 de enero de 2019) Frost, D.R. & Kluge, A.G. (1994) A consideration of epistemology in systematic biology, with special reference to species. Cladistics, 10: 259–294. Frost, D.R., Grant, T., Faivovich, J., Bain, R.H., Haas, A., Haadad, C.F.B., de Sá, R.O., Channing, A., Wilkinson, M., Donnellan, S.C., Raxworthy, C.J., Campbell, J.C., Blotto, B.L., Moler, P., Drewes, R.C., Nussbaum, R.A., Lynch, J.D., Green, D.M. & Wheeler W.C. (2006) The amphibian tree of life. Bulletin of the American Museum of Natural History, 297: 1–291. Fujisawa, T. & Barraclough, T.G. (2013) Delimiting species using single-locus data and the Generalized Mixed Yule Coalescent approach: a revised method and evaluation on simulated data sets. Systematic Biology, 62: 707–724. García-París, M., Good, D.A., Parra-Olea, G. & Wake, D.B. (2000) Biodiversity of Costa Rican salamanders: implications of high levels of genetic differentiation and phylogeographic structure for species formation. Proceedings of the National Academy of Sciences, 97: 1640–1647. Garrigues, R. & Dean, R. (2014) The birds of Costa Rica: A field Guide. Cornell University Press, Nueva York. 187 Gutiérrez-García, T.A. & Vázquez-Domínguez, E. (2013) Consensus between genes and stones in the biogeographic and evolutionary history of Central America. Quaternary Research, 79: 311–324. Hebert, P.D.N., Cywinska, A., Ball, S.L. & DeWaard, J.R. (2003) Biological identifications through DNA barcodes. Proceedings of the Royal Society of London B, 270: 313–321. Hedges, S.B. (1989) Evolution and biogeography of West Indian frogs of the genus Eleutherodactylus: slow-evolving loci and the major groups. In: Woods, C.A. (Ed.) Biogeography of the West Indies: Past, Present, and Future. Sandhill Crane Press, Gainesville, Florida, pp. 305–370. Hedges, S.B., Duellman, W.E. & Heinicke, M.P. (2008) New World direct-developing frogs (Anura: Terrarana): molecular phylogeny, classification, biogeography, and conservation. Zootaxa, 1737: 1–182. Heinicke, M.P., Duellman, W.E. & Hedges, S.B. (2007) Major Caribbean and Central American frog faunas originated by ancient oceanic dispersal. Proceedings of the National Academy of Sciences, 104: 10092–10097. Heinicke, M.P., Duellman, W.E., Trueb, L., Means, D.B., MacCulloch, R.D. & Hedges, S.B. (2009) A new frog family (Anura, Terrarana) from South America and an expanded direct– developing clade revealed by molecular phylogeny. Zootaxa, 2211: 1–35. Heinicke, M.P., Lemmon, A.R., Lemmon, E.M., McGrath, K. & Hedges, S.B. (2018) Phylogenomic support for evolutionary relationships of New World direct-developing frogs (Anura: Terraranae). Molecular Phylogenetics and Evolution, 118: 145–155. 188 Herrera, S. & Shank, T.M. (2016) RAD sequencing enables unprecedented phylogenetic resolution and objective species delimitation in recalcitrant divergent taxa. Molecular Phylogenetics and Evolution, 100: 70–79. Huelsenbeck, J.P., Andolfatto, P. & Huelsenbeck, E.T. (2011) Structurama: Bayesian inference of population structure. Evolutionary Bioinformatics, 7, EBO-S6761. Isaac, N.J., Mallet, J. & Mace, G.M. (2004) Taxonomic inflation: its influence on macroecology and conservation. Trends in Ecology & Evolution, 19: 464–469. Kapli, P., Lutteropp, S., Zhang, J., Kobert, K., Pavlidis, P., Stamatakis, A. & Flouri, T. (2016) Multi-rate Poisson Tree Processes for single-locus species delimitation under Maximum Likelihood and Markov Chain Monte Carlo. Bioinformatics, 063875. Kluge, J. & Kessler, M. (2006) Fern endemism and its correlates: contribution from an elevational transect in Costa Rica. Diversity and Distributions, 12: 535–545. Leaché, A.D., Banbury, B.L., Felsenstein, J., Nieto-Montes de Oca, A. & Stamatakis, A. (2015) Short tree, long tree, right tree, wrong tree: new acquisition bias corrections for inferring SNP phylogenies. Systematic Biology, 64: 1032–1047. Leenders, T. (2016) Amphibians of Costa Rica: a field guide. Cornell University Press. Lynch, J.D. (1980) A new frog of the genus Eleutherodactylus from western Panama. Transactions of the Kansas Academy of Science, 83: 101–105. Lynch, J.D. (1986) The definition of the Middle American clade of Eleutherodactylus based on jaw musculature (Amphibia, Leptodactylidae). Herpetologica, 42: 248–258. Lynch, J.D. (1993) The value of m. depresor mandibulae in phylogenetic hypothesis for Eleutherodactylus and its allies (Amphibia, Leptodactylidae). Herpetologica, 49: 32–41. 189 Lynch, J.D. (2000) The relationships of an ensemble of Guatemalan and Mexican frogs (Eleutherodactylus: Leptodactylidae: Amphibia). Revista de la Academia Colombiana de Ciencias Exactas, Físicas y Naturales, 24: 129–156. Lynch, J.D. & Duellman, W.E. (1997) Frogs of the genus Eleutherodactylus in Western Ecuador: systematics, ecology, and biogeography. The University of Kansas Natural History Museum Special Publication, 23: 1–236 Martens, U., Weber, B. & Valencia, V.A. (2010) U/Pb geochronology of Devonian and older Paleozoic beds in the southeastern Maya block, Central America: Its affinity with peri- Gondwanan terranes. Geological Society of America Bulletin, 122: 815–829. Miyamoto, M.M. (1983) Biochemical variation in the frog Eleutherodactylus bransfordii: geographic patterns and cryptic species. Systematic Zoology, 32: 43‒51. Montes, C., Bayona, G., Cardona, A., Buchs, D.M., Silva, C A., Morón, S., Hoyos, N., Ramírez, D.A, Jaramillo, C.A. & Valencia, V. (2012)a Arc‐continent collision and orocline formation: Closing of the Central American seaway. Journal of Geophysical Research: Solid Earth, 117: B04105. Montes, C., Cardona, A., McFadden, R., Morón, S.E., Silva, C.A., Restrepo-Moreno, S., Ramírez, D.A., Hoyos, N., Wilson, J., Farris, D., Bayona, G.A., Jaramillos, C.A., Valencia, V., Bryan. J. & Flores, J.A. (2012)b Evidence for middle Eocene and younger land emergence in central Panama: implications for Isthmus closure. Geological Society of America Bulletin, 124: 780–799. Montes, C., Cardona, A., Jaramillo, C., Pardo, A., Silva, J. C., Valencia, V., Ayala, C., Pérez- Angel, L.C., Rodríguez-Parra, L.A., Ramírez, V. & Niño, H. (2015) Middle miocene closure of the Central American seaway. Science, 348: 226–229. 190 Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature, 403: 853–858. Nadler, S.A. & Pérez-Ponce de León, G. (2011) Integrating molecular and morphological approaches for characterizing parasite cryptic species: implications for parasitology. Parasitology, 138: 1688–1709. Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V., Underwood, E.C. & Loucks, C.J. (2001) Terrestrial Ecoregions of the World: A New Map of Life on Earth: A new global map of terrestrial ecoregions provides an innovative tool for conserving biodiversity. BioScience, 51: 933–938. Padial, J.M., Köhler, J., Munoz, A. & De la Riva, I. (2008) Assessing the taxonomic status of tropical frogs through bioacoustics: geographical variation in the advertisement calls in the Eleutherodactylus discoidalis species group (Anura). Zoological Journal of the Linnean Society, 152: 353–365. Padial, J.M., Miralles, A., De la Riva, I. & Vences, M. (2010) The integrative future of taxonomy. Frontiers in Zoology, 7: 16. Padial, J.M., Grant, T. & Frost, D.R. (2014) Molecular systematics of terraranas (Anura: Brachycephaloidea) with an assessment of the effects of alignment and optimality criteria. Zootaxa, 3825: 1–132. Pérez-Ponce de León, G. & Nadler, S.A. (2010) What we don't recognize can hurt us: a plea for awareness about cryptic species. Journal of Parasitology, 96: 453–464. Pérez-Ponce de León, G. & Poulin, R. (2016) Taxonomic distribution of cryptic diversity among metazoans: not so homogeneous after all. Biology Letters, 12: 20160371. 191 Peterson, B.K., Weber, J.N., Kay, E.H., Fisher, H.S. & Hoekstra, H.E. (2012) Double digest RADseq: an inexpensive method for de novo SNP discovery and genotyping in model and non-model species. PLoS ONE, 7: e37135. Pie, M.R., Ströher, P.R., Bornschein, M.R., Ribeiro, L.F., Faircloth, B.C. & McCormack, J.E. (2017) The mitochondrial genome of Brachycephalus brunneus (Anura: Brachycephalidae), with comments on the phylogenetic position of Brachycephalidae. Biochemical Systematics and Ecology, 71: 26–31. Pinto-Sánchez, N.R., Ibáñez, R., Madriñán, S., Sanjur, O.I., Bermingham, E. & Crawford, A.J. (2012) The great American biotic interchange in frogs: multiple and early colonization of Central America by the South American genus Pristimantis (Anura: Craugastoridae). Molecular Phylogenetics and Evolution, 62: 954–972. Pons, J., Barraclough, T.G., Gomez-Zurita, J., Cardoso, A., Duran, D.P., Hazell, S., Kamoun, S., Sumlin, W.D. & Vogler, A.P. (2006) Sequence-based species delimitation for the DNA taxonomy of undescribed insects. Systematic Biology, 55: 595–609. Puillandre, N., Lambert, A., Brouillet, S. & Achaz, G. (2012) ABGD, Automatic Barcode Gap Discovery for primary species delimitation. Molecular Ecology, 21: 1864–1877. Pyron, R.A. & Wiens, J.J. (2011) A large-scale phylogeny of Amphibia including over 2800 species, and a revised classification of extant frogs, salamanders, and caecilians. Molecular Phylogenetics and Evolution, 61: 543–583. Reid, W.V. & Miller, K.R. (1989) Keeping options alive: the scientific basis for conserving biodiversity. World Resources Institute, Washington. 192 Říčan, O., Piálek, L., Zardoya, R., Doadrio, I. & Zrzavý, J. (2013) Biogeography of the Mesoamerican Cichlidae (Teleostei: Heroini): colonization through the GAARlandia land bridge and early diversification. Journal of Biogeography, 40: 579–593. Rovito, S.M., Parra‐Olea, G., Recuero, E. & Wake, D.B. (2015) Diversification and biogeographical history of Neotropical plethodontid salamanders. Zoological Journal of the Linnean Society, 175: 167–188. Savage, J.M. (1966) The origins and history of the Central American herpetofauna. Copeia, 1966: 719–766. Savage, J.M. (1982) The enigma of the Central American herpetofauna: dispersals or vicariance?. Annals of the Missouri Botanical Garden, 69: 464–547. Savage, J.M. (1987) Systematics and distribution of the Mexican and Central American rainfrogs of the Eleutherodactylus gollmeri group (Amphibia, Leptodactylidae). Fieldiana Zoology, 33: 1–57. Savage, J.M. (2002) The Amphibians and Reptiles of Costa Rica: a Herpetofauna Between Two Continents, Between Two Seas. The University of Chicago Press, Illinois. Savage, J.M. & Emerson, S.B. (1970) Central American frogs allied to Eleutherodactylus bransfordii (Cope): a problem of polymorphism. Copeia, 1970: 623–644. Savage, J.M., McCranie, J.R. & Espinal, M. (1996) A new species of Eleutherodactylus from Honduras related to Eleutherodactylus bransfordii (Anura: Leptodactylidae). Proceedings of the Biological Society of Washington, 109: 366–372. Schlaepfer, M.A. & Figeroa-Sandí, R. (1998) Female reciprocal calling in a Costa Rican leaf- litter frog, Eleutherodactylus podiciferus. Copeia, 1998: 1076–1080. 193 Sites, J.W. & Marshall, J.C. (2003) Delimiting species: a Renaissance issue in systematic biology. Trends in Ecology & Evolution, 18: 462–470. Smith, M.A., Poyarkov Jr., N.A. & Hebert, P.D.N. (2008) CO1 DNA barcoding amphibians: Take the chance, meet the challenge. Molecular Ecology Resources, 8: 235–246. Sukumaran, J. & Knowles, L.L. (2017) Multispecies coalescent delimits structure, not species. Proceedings of the National Academy of Sciences, 114: 1607–1612. Streicher, J.W., Crawford, A.J. & Edwards, C.W. (2009) Multilocus molecular phylogenetic analysis of the montane Craugastor podiciferus species complex (Anura: Craugastoridae) in Isthmian Central America. Molecular Phylogenetics and Evolution, 53: 620–630. Streicher, J.W., Devitt, T.J., Goldberg, C.S., Malone, J.H., Blackmon, H. & Fujita, M.K. (2014) Diversification and asymmetrical gene flow across time and space: lineage sorting and hybridization in polytypic barking frogs. Molecular Ecology, 23: 3273–3291. Taylor, E.H. (1948) New Costa Rican salamanders. Proceedings of Biological Society of Washington, 61: 177–180. Taylor, E.H. (1952) A review of the frogs and toads of Costa Rica. The University of Kansas Science Bulletin, 35: 577‒942. Trueb, L. (1968) Variation in the tree frog Hyla lancasteri. Copeia, 1968: 285–299. Twining, J.E. & Cossel Jr, J.O. (2017) Temporal and spectral analysis of the advertisement call of Craugastor stejnegerianus (Anura: Craugastoridae) in Costa Rica. Mesoamerican Herpetology, 4: 129–136. Vences, M., Thomas, M., Bonett, R.M. & Vieites, D.R. (2005) Deciphering amphibian diversity through DNA barcoding: chances and challenges. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 360: 1859–1868. 194 Webb, S.D. (1991) Ecogeography and the Great American Interchange. Paleobiology, 17: 266– 280. Webb, S.D. (2006) The Great American Biotic Interchange: patterns and processes. Annals of the Missouri Botanical Garden, 93: 245–257. Weber, B., Iriondo, A., Premo, W.R., Hecht, L. & Schaaf, P. (2007) New insights into the history and origin of the southern Maya block, SE México: U–Pb–SHRIMP zircon geochronology from metamorphic rocks of the Chiapas massif. International Journal of Earth Sciences, 96: 253–269. Will, K.W., Mishler, B.D. & Wheeler, Q.D. 2005 The perils of DNA barcoding and the need for integrative taxonomy. Systematic Biology, 54: 844‒851. Yang, Z. (2015) The BPP program for species tree estimation and species delimitation. Current Zoology, 61: 854–865. Yang, Z. & Rannala, B. (2010) Bayesian species delimitation using multilocus sequence data. Proceedings of National Academy of Sciences USA, 107: 9264–9269. Yang, Z. & Rannala, B. (2014) Unguided species delimitation using DNA sequence data from multiple loci. Molecular Biology and Evolution, 31: 3125–3135. Zhang, J., Kapli, P., Pavlidis, P. & Stamatakis, A. (2013) A general species delimitation method with applications to phylogenetic placements. Bioinformatics, 29: 2869–2876. 195 APÉNDICES 196 APÉNDICE I 140 YEARS AFTER WILLIAM M. GABB`S CLIMB TO CERRO PICO BLANCO Autores: Erick Arias y Gerardo Chaves Fuente: Mesoamerican Herpetology, 1 (1): 176–180. 2014 Fecha de publicación: septiembre de 2014 197 176 Mesoamerican Herpetology September 2014 | Volume 1 | Number 1 Other Contributions Nature Notes MISCELLANEOUS NOTES 140 years after William M. Gabb’s climb to Cerro Pico Blanco On 4 February 1873, the American geologist William M. Gabb arrived in Puntarenas, Costa Rica, commissioned by the Keith brothers and the Costa Rican president, Tomás Guardia. He remained in Costa Rica until August of 1874, and returned to the United States where he died on 30 May 1878 at the age of 39. We will not provide a more extensive biography on W. M. Gabb, since other authors have written extensively about his life (Dall, 1909) and contributions to herpetology (Savage, 1970), geology and cartography (Denyer and Soto, 1999; Denyer and Lücke, 2007), and ethnic and socio-political studies in Costa Rica (Gabb, 1978). Instead, the purpose of this note is to an- alyze the route William M. Gabb traveled on his expedition to Cerro Pico Blanco, because of its importance in the study of the Costa Rican herpetofauna. During his 19 months in Costa Rica, Gabb primarily was engaged in studying the Talamanca region (Gabb, 1894). As part of his research he was required to climb Cerro Pico Blanco, currently known as Cerro Kamuk. Whether he actually went to Kamuk or another peak has been debated (see below). During this trip, Gabb collected specimens of 20 species of amphibians and reptiles (Cope, 1875), of which 15 were described as new species (Cope, 1875; Savage, 1970). To this day, 12 specimens from Gabb’s collection are maintained as holotypes (Frost, 2014; Table 1). Due the great scientific value of Gabb’s collections, it is important to clarify the actual collecting localities from where the specimens came. Identifying the type locality of a species is an essential consideration when attempting to visit the site to collect new specimens to replace lost material, or to obtain tissues for molecular studies. Thus, the goals of this contribution are as follows: (1) to clarify which peak Gabb actually climbed during his trip to Cerro Pico Blanco in Costa Rica, (2) to suggest the possible ascent route he followed, and (3) to suggest the possible type localities for the majority of the specimens he collected. Table 1. Species described with specimens Gabb collected. The elevations are according to the sites suggested on Fig. 3. Species Locality Amphibia Craugastor gulosus Near an unnamed hill close to the Río Lari canyon, elev. 1,830 m Craugastor megacephalus Near an unnamed hill close to the Río Lari canyon, elev. 1,830 m Craugastor melanostictus Near another unnamed hill at the headwaters of the Río Lari, elev. 2,135 m Craugastor podiciferus Between Cerro Pat and the headwaters of the Río Lari, elev. 1,520–2,135 m Diasporus hylaeformis Near another unnamed hill at the headwaters of the Río Lari, elev. 2,135 m Incilius fastidiosus Around the small village of Ourut, elev. 760 m Incilius epioticus Near Cerro Pat, elev. 1,520 m Isthmohyla pictipes Between Cerro Pat and the headwaters of the Río Lari, elev. 1,520–2,135 m Pristimantis cerasinus Slope of Cerro Kamuk, elev. possibly 760–1,520 m Reptilia Anolis pachypus Slope of Cerro Kamuk, elev. possibly 760–1,520 m Ninia psephota Between Cerro Pat and the headwaters of the Río Lari, elev. 1,520–2,135 m Mesaspis monticola Summit of Cerro Kamuk, elev. ca. 3,500 m 198 177 Mesoamerican Herpetology September 2014 | Volume 1 | Number 1 Other Contributions Nature Notes Among the documents Gabb generated during his stay in Costa Rica was a report on the geology of the country (Gabb, 1874). This report, however, remained unpublished until it was transcribed in 2007 (Gabb, 2007). In this report, Gabb wrote that he reached the summit of Cerro Pico Blanco or Kamuk on 13 June 1874, after 12 days of the hardest work in his life. He reported an elevation of 11,877.8 feet (3,620 m), close to the 3,549 m in elevation calculated for Kamuk by the Instituto Geográfico Nacional of Costa Rica (IGN). Nonetheless, in a map Gabb produced from his studies on Costa Rica (Petermann 1877, Fig. 1) an elevation of 9,652 feet (2,942 m) is assigned to Cerro Pico Blanco. This confusion grew when Enrique Pittier translated Gabb’s reports (Gabb, 1894), since Pittier attributed an elevation of 9,652 feet to Cerro Pico Blanco. This discrepancy in the reported elevations led some researchers to conclude that William Gabb ascended Cerro Utyum and not Cerro Kamuk (Carballo, 1960; Gutiérrez, 1960; Savage, 1970), a conclusion largely motivated by the similarity of elevations in the 1894 transla- tion of Gabb’s report and that of the Cerro Utyum (3,060 m). We believe, however, that this interpretation from the 1960s and 70s is the result of the difficulty in accessing Gabb’s unpublished 1874 manuscript. The inaccessibility of the original copy at the U.S. Geological Survey obligated researchers to consult only Pittier’s translation (Gabb, 1894) in referring to Gabb’s work. Denyer and Lücke (2007) suggested that Gabb actually climbed Cerro Kamuk, based on the similarity of the hydrography and position of the mountains between Gabb’s map (Petermann, 1877, Fig. 1) and recent ones. These authors, however, did not discuss the issues raised by other researchers (Carballo, 1969; Gutiérrez, 1960; Savage, 1970). Hence, Gabb’s zoological contributions from Cerro Pico Blanco still are attributed to Cerro Utyum (Frost, 2014). We performed a comprehensive analysis of Gabb’s original manuscript (1874), his map (Petermann, 1877), Pittier’s translation (Gabb, 1894), Savage’s interpretation (1974), Denyer and Lücke’s analysis (2007), and our per- sonal surveys at the summits of Kamuk and Utyum. Based on our re-evaluation of these sources, we conclude that Gabb indeed climbed Cerro Kamuk, as he reported. This conclusion is based on several lines of reasoning. First, when Gabb sent the specimens he collected to the Smithsonian Institution, he attributed an elevation of at least 11,800 feet to Cerro Pico Blanco (Cope 1875), which is very close to the actual elevation of Cerro Kamuk. Also, Fig. 1. A portion of Gabb’s map showing the region between Sipurio and Cerro Kamuk. Compare sites underlined in red with those in Fig. 3. Reproduced from Petermann (1877). 199 178 Mesoamerican Herpetology September 2014 | Volume 1 | Number 1 Other Contributions Nature Notes in some of Gabb’s other documents (Gabb, 1894; Petermann, 1877), he consistently miscalculated the elevation of several peaks, other than Cerro Kamuk. For example, the well-known Volcán Barva was reported at an elevation of 2,827 m in Gabb (1894), but it was reduced to 2,652 m in his map (Petermann, 1877); indeed, it would be diffi- cult to believe that Gabb misidentified Volcán Barva for another peak. Thirdly, Gabb (1874) reported that the Lari and Uren rivers originate on Cerro Pico Blanco; both are known to originate at Cerro Kamuk, and it is difficult to imagine that Gabb could have reported this accurately if he actually had reached the summit of another peak. For example, if he in fact had climbed Cerro Utyum his descriptions of the Lari and Coen rivers would have been the most detailed, but the Río Coen is poorly described in his map (Petermann, 1877). Finally, a detailed description of the summit of Cerro Pico Blanco in the original 1874 report is not included in the translation of Pittier (Gabb, 1894). Gabb described the difficult access to the summit, with large cliffs and vegetation similar to that in American deserts. This description differs greatly from the summit of Cerro Utyum, where we currently conduct studies. An extensive peat bog is present on Cerro Utyum, and the summit is accessed easily from the treeline (Fig. 2a). The description, however, closely matches that of Cerro Kamuk (G. Chaves, pers. observ.), in which the slopes are very steep and it is difficult to reach the summit, the vegetation is limited, and cliffs dominate the landscape (Fig. 2b). Fig. 2 Summits of (A) Cerro Utyum and (B) Cerro Kamuk in southwestern Costa Rica. ' © (A) E. Arias, (B) Luis G. Artavia A B 200 179 Mesoamerican Herpetology September 2014 | Volume 1 | Number 1 Other Contributions Nature Notes We analyzed Gabb’s map (Petermann, 1877, Fig. 1) to determine the possible route he followed while ascend- ing Cerro Kamuk. We inferred that the more accurate Gabb’s description of a landmark, the better he was able to inspect it, and therefore, the closer his route took him to it. Gabb (1874) indicated that he ascended between the Lari and Uren rivers. According to Denyer and Lücke (2007), Gabb’s map (Petermann, 1877) adheres closely to current knowledge of the hydrography of the Talamanca region; they especially noted the excellent fit of the Río Lari with the IGN’s current maps (Fig. 1). A noteworthy consideration, however, is that the section of the Río Uren between Sipurio (9.5360°N, -82.9513°W, WGS84, elev. 60 m) and Dipuk (see below) is different than in the current maps (Figs. 1 and 3). Given this detail, we suggest that Gabb did not use the river line of the Uren to reach Dipuk, but rather walked along the Lari. We propose (see Fig. 3) that he started in the town of Sipurio, walked to the small village of Sarke or Alto Lari (9.4345°N, -83.0466°W, elev. 400 m) following the Río Lari, then walked to Cerro Pat (9.3957°N, -83.0231°W, elev. 1,500 m). At an elevation of about 2,000 m, he climbed to Cerro Kamuk following a ridge that divides the Uren and Lari rivers. When Gabb’s map (Petermann, 1877) is superimposed on IGN’s current map, the locations Gabb called Isikoilset and Dipuk correspond to Cerro Pat and the small village of Guachalaba (9.3730°N, -82.9987°W, elev. 900 m), respectively (Figs. 1 and 3). Gabb likely visited Dipuk on a separate trip taken before climbing Cerro Kamuk, since he indicated that he crossed the Río Uren and continued southeastward until reaching Panama. We infer that this visit to Dipuk was part of an earlier trip, because the journey to Panama would have been very difficult to complete following the strenuous trip to Kamuk. Notably, the trail between Sipurio and Guachalaba Fig. 3 Map of the Costa Rica–Panama border region showing the route we suggest Gabb followed. 1–500 500–1,000 1,000–1,500 1,500–2,000 2,000–2,500 2,500–3,000 3,000–3,500 201 Other Contributions Nature Notes 180 Mesoamerican Herpetology September 2014 | Volume 1 | Number 1 is an old indigenous route recognized on IGN maps from 1968. Finally, Gabb (1874) noted that he climbed to Pico Blanco by the narrow ridge dividing the Lari and Uren, a crest that runs from Cerro Pat to Cerro Kamuk. Based on the proposed route, we suggest some localities as potential collection sites (Table 1, Fig. 3). The specimens collected by Gabb on his trip to Cerro Kamuk are restricted to four sites, at 760, 1,520, 1,830 and 2,135 m in elevation (Cope, 1875). We suggest the following four areas are an approximation of Gabb’s original sites (Table 1, Fig. 3): the site at 760 m is around the small village of Ourut (9.4149°N, -83.0514°W, elev. 700 m); the 1,520 m site is near the Cerro Pat; the 1,830 m site is near an unnamed hill close to the Río Lari canyon (9.3657°N, -83.0415°W, elev. 1,800 m); and the 2,135 m site is near another unnamed hill at the headwaters of the Río Lari (9.3488°N, -83.0434°W, elev. 2,100 m). Clarifying the route followed by Gabb on his trip to Cerro Kamuk allows us to understand the fieldwork of a pioneer collector of Costa Rica’s herpetofauna. On 13 June 2014 we celebrated the 140th anniversary of William Gabb’s climb to Cerro Pico Blanco. This feat fills us with admiration, but also imposes the task of replicating his expedition, particularly since the comforts and technologies of the present facilitate fieldwork. LITERATURE CITED CARBALLO, E. 1960. Viaje a los Vértices Dudu y Kamuk, en la Cordillera de Talamanca. Instituto Geográfico de Costa Rica. San José, Costa Rica. COPE, E. D. 1875. On the Batrachia and Reptilia of Costa Rica. Journal of the Academy of Natural Sciences of Philadelphia 8: 93–157. DALL, W. H. 1909. Biographical memoir of William More Gabb. Natural Academy of Sciences, Biographical Memoirs 6: 347– 361. FROST, D. R. 2014. Amphibian Species of the World: An Online Reference. Version 6.0. American Museum of Natural History, New York, New York, United States. (www.research.amnh. org/herpetology/amphibia/index.html; accessed 27 May 2014). GABB, W. M. 1874. On the geology of the Republic of Costa Rica. Unpublished manuscript at U.S. Geological Survey. GABB, W. M. 1894: Informe sobre la exploración de Talamanca verificada durante los años 1873–1874 (Introduction by E. Pittier). Tipografía Nacional, San José, Costa Rica. GABB, W. M. 1978. Talamanca, el espacio y los hombres (Presented by L. Ferrero). Ministerio de Cultura de Juventud y Deportes, San José, Costa Rica. GABB, W. M. 2007. On the Geology of the Republic of Costa Rica. (Transcription of the original manuscript by O. H. Lücke, V. Gutiérrez, and G. Soto). Revista Geológica de América Central 37: 103–118. GUTIÉRREZ, F. 1960. Ascensión de Gabb al Kamuk o Pico Blanco. Instituto Geográfico de Costa Rica, San José, Costa Rica. PERCY, D., AND G. SOTO. 1999. Contribución pionera de William M. Gabb a la geología y cartografía de Costa Rica. Anuario de Estudios Centroamericanos 25: 103–138. PERCY, D. AND O. H. LÜCKE. 2007. Comentario sobre William M. Gabb: legado y contribuciones inéditas y olvidadas. Revista Geológica de América Central 37: 91–102. PETERMANN, A. 1877. W. M. Gabb, Collins & Martínez’s aufnahme von Talamanca un der kartographische standpunkt von Costa Rica in 1877. Gotha 23: 385–387 + map. SAVAGE, J. M. 1970. On the trail of the Golden Frog: with W. Arszewicz and Gabb in Central America. Proceedings of the California Academy of Sciences: Festschrift for George Sprague Myers 38: 273–288. SAVAGE, J. M. 1974. Type locality for species of amphibians and reptiles described from Costa Rica. Revista de Biología Tropical 22: 71–122. ERICK ARIAS 1,2 AND GERARDO CHAVES 2 1Departamento de Zoología, Instituto de Biología, Universidad Nacional Autónoma de México, Distrito Federal 04510, Mexico. E-mail: eapiedra@gmail.com (Cooresponding Author) 2Escuela de Biología, Universidad de Costa Rica, San Pedro, 11501–2060 San José, Costa Rica. E-mail: cachi13@gmail. com 202 APÉNDICE II BIOGEOGRAFÍA HISTÓRICA DE LOS ANFIBIOS DE AMÉRICA CENTRAL: UNA REVISIÓN DE HIPÓTESIS ACTUALES Y PERSPECTIVAS FUTURAS Autores: Erick Arias y Gabriela Parra-Olea Ensayo evaluado durante el examen de candidatura. 203 Biogeografía histórica de los anfibios de América Central: una revisión de hipótesis actuales y perspectivas futuras. 1,2,3Erick Arias & 1Gabriela Parra-Olea 1 Departamento de Zoología, Instituto de Biología, UNAM, AP 70-153 Ciudad Universitaria, CP 04510, Ciudad de México, México. E-mail: eapiedra@gmail.com 2 Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México, Av. Ciudad Universitaria 3000, C.P. 04360, Coyoacán, Ciudad de México, México. 3 Escuela de Biología, Universidad de Costa Rica, San Pedro, 11501-2060 San José, Costa Rica. 204 CONTENIDO Introducción .......................................................................................................................... 3 América Central .................................................................................................................... 5 Definición ....................................................................................................................... 5 Unidades ecogeográficas ................................................................................................ 5 Nivel del mar y clima en el pasado ................................................................................. 7 Historia geológica ........................................................................................................... 8 Diversidad de anfibios en América Central...................................................................... 10 Diversidad de anfibios y sus patrones de distribución en América Central ..................... 10 Áreas de endemismo para los anfibios de América Central ............................................. 15 Biogeografía histórica de los anfibios de América Central ............................................. 17 Principales estudios durante el siglo XX .......................................................................... 17 Hipótesis actuales a luz de la evidencia molecular ........................................................... 19 La primera dispersión y diversificación en América Central........................................ 19 La colonización de América Central Ístmica previo al cierre del Istmo de Panamá .... 23 El cierre del Istmo de Panamá y el Gran Intercambio de la Biota Americana ............. 24 Perspectivas actuales: una mirada desde las ranas de desarrollo directo ..................... 26 Especies crípticas .............................................................................................................. 26 Filogeografía como herramienta para descubrir especies crípticas .................................. 29 Retos futuros........................................................................................................................ 31 Conclusiones ........................................................................................................................ 32 Agradecimientos .................................................................................................................. 33 Literatura citada ................................................................................................................. 33 205 3 INTRODUCCIÓN La biogeografía histórica estudia la distribución de los seres vivos, trata de establecer las causas y factores responsables de esta distribución en una escala de tiempo geológico (Crisci et al. 2003). Su origen multidisciplinario hace de la biogeografía histórica una disciplina sensible a los avances en otros campos de conocimiento como la geología, geografía y biología. Según Wiley (1988) el desarrollo de la teoría de la tectónica de placas y los avances conceptuales- metodológicos en la sistemática filogenética explican el surgimiento de la biogeografía cladística a principios de 1980, misma que dio origen a la gran mayoría de los enfoques teóricos y prácticos de la actualidad (Morrone 2005). Además, el entendimiento de la historia biogeográfica de una región se beneficia con nuevas y más robustas reconstrucciones filogenéticas y geológicas para dicha región. Tres características hacen de América Central una región de gran interés para la biogeografía. La primera de estas es el cierre del Istmo de Panamá en el extremo sureste de América Central que permitió el Gran Intercambio de la Biota Americana, el intercambio entre dos grandes masas continentales (Norte y Sudamérica) hasta entonces aisladas. El interés por este gran evento biogeográfico ha generado mucha investigación en la región, desde sus orígenes como modelo en mamíferos hasta recientemente con otros organismos como anfibios (Simpson 1940; Marshall et al. 1982, 1988; Webb 1991, 2006; Pinto-Sánchez et al. 2012; Bacon et al. 2016). La segunda de estas características es su compleja historia geológica con fuertes implicaciones para la biogeografía regional. Las discusiones generadas desde la geología sobre la formación de América Central y la posibilidad de una conexión temporal entre Norteamérica y Sudamérica previo al cierre del Istmo de Panamá (Pindell & Barrett 1990; Iturralde-Vinent & MacPhee 1999) han sido de gran interés para la biogeografía (Savage 1966, 1982; Hedges 2006; Alí 2012). La tercera característica de América Central es su extremadamente rica biodiversidad actual. Por ejemplo, alberga el 7% de los anfibios del mundo, muchos de los cuales son endémicos a la región (Duellman 2001; Savage 2002). Esta alta diversidad se ha atribuido a su compleja historia geológica y su amplia heterogeneidad geográfica y climática (Savage 1966, 1982; Campbell 1999). La presencia de una biodiversidad rica y única ha propiciado diversos estudios biogeográficos con diferentes grupos de organismos (Savage 1996, 1982; Bussing 1976, 1985; Bagley & Johnson 2014). 206 4 Los anfibios son un grupo con alta diversidad en América Central (Duellman 2001), con un alto nivel de endemismo, lo que motivó a Savage (1966) a sugerir que esta fauna debe ser considerada una unidad independiente al mismo nivel de la fauna neotropical o neártica. Uno de los grupos diversos de anfibios en la región son las ranas de desarrollo directo (Brachycephaloidea), que en su totalidad representan el 23% de las especies de anfibios de América Central (Frost 2019). Siguiendo a Padial et al. (2014) las ranas de desarrollo directo constituyen un clado de 25 géneros en tres familias neotropicales (Craugastoridae, Eleutherodactylidae y Brachycephalidae). Cinco de estos géneros (Craugastor, Diasporus, Eleutherodactylus, Pristimantis y Strabomantis) habitan América Central, pero, Craugastor es el más diverso con 96 especies en la región. Estos géneros tienen orígenes distintos y su presencia en América Central responde a diferentes momentos espacio-temporales (Heinicke et al. 2007; Hedges et al. 2008). Los distintos orígenes e historias biogeográficas hacen de las ranas de desarrollo directo un grupo ideal para una revisión de la biogeografía histórica de los anfibios en América Central. El presente trabajo revisa la biogeografía histórica de los anfibios en América Central, mas no se pretende repetir lo que de forma extensa y detallada se ha documentado en las revisiones sobre biogeografía histórica de anfibios para la región (Savage 1966, 1982, 2002; Campbell 1999; Flores-Villela & Goyenechea 2001; Duellman 2001). Este trabajo se enfoca en la revisión de las hipótesis actuales y como estas responden a avances en reconstrucciones filogenéticas y geológicas para la región. Para alcanzar este objetivo el presente trabajo se divide en cinco secciones. La primera sección da un marco histórico para América Central, su definición, unidades ecogeográficas, su historia geológica, el cambio en el clima y el nivel del mar en el pasado. En la segunda sección se revisa brevemente la diversidad de anfibios en América Central, sus patrones de diversidad en la región y las principales hipótesis biogeográficas que surgieron previo al advenimiento de las herramientas moleculares. En la tercera sección se revisan las hipótesis actuales de la biogeografía histórica de América Central a la luz de las nuevas evidencias moleculares y geológicas, haciendo especial énfasis a estudios realizados con ranas de desarrollo directo. En la cuarta sección se analiza la biogeografía de la región desde la perspectiva actual con nuevas herramientas metodológicas y conceptuales utilizando ejemplos puntuales con ranas de desarrollo directo. En la quinta y última sección se 207 5 revisan los principales retos que debe atender el estudio de la biogeografía histórica de los anfibios en América Central. AMÉRICA CENTRAL Definición. Siguiendo la definición de Savage (1982) y Gutiérrez-García & Vázquez- Domínguez (2013) América Central es el área comprendida entre el Istmo de Tehuantepec, México y la Cordillera de los Andes en Colombia (Fig. 1). Es importante remarcar que en la literatura biogeográfica existen nombres ambiguos para la delimitación geográfica de la región. El término en inglés “Middle America” que se puede traducir como Medio de América o Centro de América, se ha utilizado para referirse al área comprendida entre los Estados Unidos de América y Colombia (Villa et al. 1988; Duellman 2001), a veces incluyendo las Islas Caribeñas (Winker 2011). El término Mesoamérica (“Mesoamerica” en inglés) ha sido utilizado como un sustituto para “Middle America” (Johnson et al. 2001; Wilson et al. 2010), sin embargo, Winker (2011) recomienda el uso de “Middle America” sobre “Mesoamerica”, principalmente porque este último término se acuñó en antropología para referirse a una entidad social. A pesar de la confusión que puede causar el uso de América Central, especialmente con respecto a “Middle America”, la delimitación geográfica aquí propuesta corresponde con la región nombrada “Central America” por Savage (1982) y Gutiérrez-García & Vázquez-Domínguez (2013); Savage (19829 además distinguió entre “Central America” y “Middle America”. El término centroamericano, se utilizará como relativo a América Central. América Central Nuclear hace referencia a las tierras entre el Istmo de Tehuantepec y el sur de Nicaragua y América Central Ístmica se refiere a las tierras de Costa Rica y Panamá en su conjunto. Unidades ecogeográficas. América Central, a pesar de ser una región relativamente pequeña, poco más de 770 077 km2, alberga una compleja y heterogénea fisiografía, que a su vez, genera una amplia gama de climas y hábitats. La región se ubica entre los 7º y 21º N, corriendo de Noroeste a Sureste, separando el Océano Pacífico al Noroeste y el Golfo de México y el Mar Caribe al Noreste. Su rango de elevación va desde el nivel del mar hasta los 4 220 m.s.n.m. (Volcán Tajumulco, Guatemala). En el extremo sureste, la franja terrestre se reduce a 65 km formando el Istmo de Panamá. 208 6 Figura 1. Mapa de América Central, se muestran los cuatro bloques tectónicos (delimitados en azul) y las principales unidades ecogeográficas de la región (en blanco). Las abreviaciones corresponden a: IT = Istmo de Tehuantepec, FMP = Falla Motagua-Polochic, EH = escarpada de Hess, FNP = fractura norte de Panamá, CA = Cordillera de los Andes, MC = macizo de Chiapas, AVMC = arco volcánico moderno de Chiapas, TAB = tierras altas de Belice, PY = península de Yucatán, AVCA = arco volcánico de Centro América, DP= depresión de Nicaragua, TAI = tierras altas del Istmo, IP = Istmo de Panamá y TAD = tierras altas del Darién. 209 7 Las unidades ecogeográficas de la región han recibido diversos nombres. Siguiendo a Campbell (1999) y Duellman (2001) con algunas modificaciones a partir de Gutiérrez-García & Vázquez-Domínguez (2013) (Fig. 1), las tierras altas de la región se pueden agrupar en seis unidades geomorfológicas: 1) el Macizo de Chiapas, 2) las Tierras Altas de Belice, 3) el Arco Volcánico Moderno de Chiapas [Campbell (1999) y Duellman (2001) trataron estas últimas dos unidades como una sola “Tierras Altas del Este de América Central Nuclear”], 4) el Arco Volcánico de Centro América [las “Tierras Altas del Oeste de América Central Nuclear” en Campbell (1999) y Duellman (2001)], 5) las Tierras Altas del Istmo y 6) las Tierras Altas del Darién. Asimismo, las tierras bajas de la región se agrupan en ocho unidades: 1) la Península de Yucatán, 2) las Tierras Húmedas del Caribe Norte, 3) las Tierras Húmedas del Caribe Sur, 4) las Tierras Húmedas del Pacífico, 5) las Tierras Secas de América Central, 6) las Tierras Húmedas del Golfo Dulce, 7) las Tierras Secas de Panamá y 8) las Tierras Húmedas del Chocó. Nivel del mar y clima en el pasado. El nivel del mar desde el Paleoceno hasta el Mioceno ha sido controversial. Miller et al. (2005) sugirieron que el nivel del mar desde el Paleoceno (~60 Ma) hasta el presente se mantuvo casi de forma constante bajo el nivel actual con un descenso acelerado durante los últimos 2 Ma y con un aumento hasta su nivel actual tan solo en los últimos 18 ka –miles de años antes–. Posteriormente, Kominz et al. (2008) revisaron y corrigieron la curva de Miller et al. (2005) y sugirieron que el nivel del mar desde el Cretácico Superior (~100 Ma) hasta el Oligoceno (~30 Ma) se mantuvo siempre por encima del nivel actual. A partir del Oligoceno y hasta el presente, el nivel del mar se mantuvo oscilando cerca de los niveles actuales, con al menos dos momentos donde el nivel del mar fue significativamente inferior al actual, uno al inicio del Mioceno (~23 Ma) y otro en el Mioceno Medio (~10 Ma). Los resultados de Kominz et al. (2008) son concordantes con trabajos previos (Kominz 1984; Haq et al. 1987) y apoya la hipótesis de que la Depresión de Nicaragua permaneció inundada desde el Cenozoico Tardío hasta el comienzo del Pleistoceno (Duellman 2001; Gutiérrez-García & Vázquez-Domínguez 2013). Savage (1966) sugiere que durante el Cenozoico el clima en la región fue mésico y argumenta a favor de una barrera semi-árida a árida entre América Central y Norteamérica en el Oligoceno Medio (~28 Ma). El cambio climático mejor documentando en América Central es la transición de la región hacia condiciones más xéricas, especialmente en las tierras bajas de la costa pacífica, a partir del Plioceno (5 Ma) (Savage 1966; Graham & Dilcher 1995; Duellman 210 8 2001). Durante el Pleistoceno Tardío (39.4-28.1 ka) se asume que las temperaturas fueron elevadas y el nivel del mar superior al actual, posteriormente (28-14.5 ka) y coincidente con la última máxima glaciación las temperaturas fueron inferiores y el nivel del mar disminuyó (González et al. 2006), en los últimos 18 ka se dio un aumento acelerado del nivel del mar hasta su nivel actual (Miller et al. 2005; González et al. 2006). Durante la última máxima glaciación las tierras altas de América Central Nuclear e Ístmica estuvieron cubiertas por glaciares que terminaron su desglaciación hace ~10 ka (Lachniet & Vazquez-Selem 2005; Kapelle & Horn 2005). Historia geológica. La historia geológica de América Central es compleja e incompleta. Existen diversos trabajos extensos revisando este tema (Coates & Obando 1996; Coates et al. 2004, 2005; Bundschuh & Alvarado 2007; Rogers et al. 2007; Silva-Romo 2009; Montes et al. 2012a,b, 2015) y otros que han realizado recopilaciones de la historia geológica de la región como marco para análisis biogeográficos (Savage 1982; Campbell 1999; Duellman 2001; Gutiérrez-García & Vázquez-Domínguez 2013; Bagley & Johnson 2014). Sin embargo, es necesario dar una breve reseña de la geología regional. América Central se puede subdividir en 4 bloques tectónicos (Fig. 1): Maya, Chortís, Chorotega y Chocó. En conjunto los bloques Maya y de Chortís conforman lo que se denomina América Central Nuclear, representando las tierras más antiguas de la región. Estas tierras por su mayor tiempo emergidas, han jugado un papel importante en la biogeografía de la región, recibiendo biodiversidad de las grandes masas norte y sudamericanas. Por su parte, los bloques Chorotega y Chocó conforman América Central Ístmica, tierras relativamente más jóvenes y que han tenido un papel más reciente en la conformación actual de la biodiversidad regional, en especial permitiendo el Gran Intercambio de la Biota Americana. El bloque Maya se extiende desde el Istmo de Tehuantepec hasta la Falla Motagua- Polochic (Fig. 1), son las tierras más viejas de América Central. Se ha sugerido que el Macizo de Chiapas y el domo de los Cuchumatanes existieron durante el Pérmico-Triásico (~250 Ma) (Weber et al. 2007; Martens et al. 2010). El bloque de Chortís se extiende desde la Falla Motagua-Polochic hasta el Escarpa de Hess, su evolución es más compleja y discutida. Rogers et al. (2007) revisaron las diferentes hipótesis del origen del bloque de Chortís, ellos sugirieron que el bloque colisionó con el extremo sur de la placa Norteamérica (sur de México) en el Cretácico Tardío (~72 Ma) y se mantuvo en contacto con el continente hasta alcanzar su posición actual. 211 9 Keppie & Morán-Zenteno (2005) sugirieron una hipótesis alternativa, en la que el bloque de Chortis llegó a su posición actual por deriva desde el Océano Pacífico Este entre el Cretácico Tardío y el Eoceno Temprano (65-55 Ma). Otra hipótesis sugiere que el bloque de Chortís estuvo en contacto con el bloque Maya, en su posición actual, al menos desde el Cretácico Tardío (~66 Ma) (James 2005). Más allá de la discusión sobre el origen del Bloque de Chortís, es probable que este haya estado en su posición actual o al menos adherido al sur del actual México desde el Eoceno (~56 Ma). Los bloques Chorotega y Chocó se extienden desde el Escarpa de Hess hasta la Fractura Norte de Panamá y desde esta última hasta la zona de subducción entre las placas Nazca y Sudamericana en Colombia (Cordillera de los Andes), respectivamente (Fig. 1). Aunque ambos bloques tienen características únicas, se revisan en conjunto dada su estrecha relación histórica. El conocimiento sobre la historia geológica de la región está cambiando aceleradamente. La hipótesis tradicional, denominada “modelo de isla” propone un origen para América Central Ístmica reciente, del Mioceno Medio (~15 Ma) a ~3 Ma (Coates & Obando 1996, Coates et al. 2004), bajo este modelo en el Mioceno Medio (~15 Ma ) América Central Ístmica estaba conformada por un archipiélago de islas. Este modelo plantea que el levantamiento regional generó un continuo de tierras emergidas pero desconectadas de Sudamérica hace ~6.5 Ma y data el cierre del Istmo de Panamá hace ~3 Ma. Una hipótesis más reciente, denominada “modelo de península” propone un levantamiento y un cierre mucho más antiguo (Montes et al. 2012a,b, 2015). Este modelo expone que América Central Ístmica existe como tierra emergida pero desconectada de Sudamérica desde el Oligoceno Tardío al Mioceno Temprano (~25 Ma) y data el cierre del Istmo de Panamá entre 15-13 Ma. Aún persisten grandes vacíos de conocimiento sobre la formación geológica de América Central. Es de especial importancia para las hipótesis biogeográficas de la región, un modelo que explique si hubo una conexión terrestre entre América Central Nuclear y Sudamérica previo al cierre del Istmo de Panamá, y si la hubo cómo fue. La evidencia es concordante en apoyar que diversos grupos biológicos llegaron a América Central Nuclear desde Sudamérica previo al cierre del Istmo de Panamá (Duellman 2001; Savage 2002). Dos hipótesis se han sugerido como modelos para explicar esta primera conexión. La primera, la hipótesis de las “proto-Antillas” asume que existió una conexión entre América Central Nuclear y Sudamérica al menos hace 80- 70 Ma, este puente terrestre estuvo ubicado en la posición actual de América Central Ístmica 212 10 (Pindell & Barrett 1990; Hedges 2006). Una segunda hipótesis denominada “Greater Antilles + Aves Ridge GAARlandia” menciona que las Antillas mayores estuvieron conectadas a Sudamérica al menos hace 35-33 Ma (Iturralde-Vinent & MacPhee 1999). Esta última hipótesis no implica una conexión directa entre América Central Nuclear y Sudamérica, pero sí una ruta intermedia que acortó la distancia y aumentó el éxito de dispersión desde las Antillas Mayores hacia América Central Nuclear. La conexión temporal entre América Central Nuclear y Sudamérica sí existió, sigue siendo controversial y la hipótesis de las proto-antillas es altamente improbable, asumiendo un origen temprano de América Central Ístmica bajo el modelo de península de Montes et al. (2012a,b). DIVERSIDAD DE ANFIBIOS EN AMÉRICA CENTRAL Diversidad de anfibios y sus patrones de distribución en América Central Los anfibios (Amphibia Linnaeus) conforman un grupo monofilético, distribuido de forma cosmopolita excepto en la Antártida, Groenlandia, el desierto del Sahara y la Meseta Tibetana (Frost et al. 2006; Pyron & Wiens 2011). Los anfibios actuales se remontan al carbonífero (~320 Ma) (San Mauro 2010) y el registro fósil presenta una gran diversidad de grupos anfibios (Pyron 2011). La diversidad de anfibios vivientes es de 7 959 especies nombradas (Frost 2019) distribuidas en tres órdenes: 1) Anura (Fischer von Waldheim) con 7021 especies, 2) Caudata (Fischer von Waldheim) con 726 especies y 3) Gymnophiona (Müller) con tan solo 212 especies. La diversidad de anfibios en América Central es de 529 especies, representando el 7% de los anfibios en el mundo en tan solo el ~0.7% de la superficie mundial habitada por anfibios. De esta diversidad el 65.5% (346 especies) corresponde a Anura, el 31.5% (167 especies) corresponde a Caudata y tan solo el 3% (16 especies) corresponde a Gymnophiona. 213 11 Figura 2. Patrones de distribución de riqueza de anfibios en América Central. A) riqueza total de especies, B) riqueza total de géneros y subgéneros, C) riqueza autóctona de especies y D) riqueza autóctona de géneros y subgéneros. Las gráficas de barras corresponden a los conteos por bloque tectónico. Rangos de distribución obtenidos de la IUCN (http://www.iucnredlist.org/). 214 12 Los anfibios de América Central se pueden agrupar con base en tres orígenes distintos (ver abajo): 1) el grupo centroamericano (CA) correspondiendo a aquellos que se originaron en la región, 2) el grupo Norteamericano (NA) que representan especies que ampliaron su rango de distribución hacia el sur y 3) el grupo Sudamericano (SA) que representan especies que ampliaron su rango de distribución hacia el norte. En la Tabla 1, se resume la conformación de los tres grupos de anfibios de América Central siguiendo a Duellman (2001), Savage (2002) y Rovito et al. (2015a). De las 529 especies de anfibios habitando en América Central, 401 (75.8%) se atribuyen a géneros que se originaron en la región. Las restantes 128 especies corresponden a especies pertenecientes a géneros de origen sudamericano (114) o norteamericano (14). Aunque muchas de estas especies atribuidas a Sudamérica son endémicas a la región, no se consideran parte del grupo originario de América Central. Esta alta diversidad de anfibios en América Central retrata una historia evolutiva regional compleja que incluye sitios de alto endemismo, principalmente en las tierras altas (Savage 1982, 2002; García-París et al. 2000; Rovito et al. 2015a). La diversidad de anfibios en América Central no está distribuida de forma proporcional, la riqueza de especies de Costa Rica y Panamá (América Central Ístmica) es por mucho superior al resto de la región (Fig. 2). La mayor cantidad de especies en Costa Rica y Panamá se ha explicado históricamente como el producto de múltiples factores, el primero, es la colonización por especies que se originaron en América Central Nuclear, segundo, una diversificación in situ y tercero, la colonización por especies precedentes de Sudamérica que alcanzan sus límites más norteños en Costa Rica o Panamá (Savage 1982, 2002; Campbell 1999). Sin embargo, el factor que posiblemente más ha contribuido a la conformación actual de la fauna anfibia en estos dos países, es la diversificación in situ promovida por una alta heterogeneidad ambiental en las tierras altas del Istmo (Costa Rica y oeste de Panamá) (García-París et al. 2000; Savage 2002; Boza-Oviedo et al. 2012). Con el fin de ilustrar el patrón desproporcional de la riqueza de anfibios en América Central, se presenta en la Tabla 2 el número de especies de anfibios por país en la región. Se realizó un análisis de riqueza de especies para 459 anfibios distribuidos en América Central. Las áreas de distribución son aquellas disponibles en IUCN y el análisis fue realizado con lestR (Vilela & Villalobos 2015). Se obtuvieron acumulados de especies para cuadrículas de 1ºx1º, y el análisis fue replicado excluyendo las especies de Sudamérica. Al analizar los patrones 215 13 de distribución de los anfibios en América Central (Fig. 2) se puede observar lo siguiente. Cuando se analiza toda la diversidad actual es evidente una mayor diversidad de especies y linajes mayores (géneros y subgéneros) en Costa Rica y Panamá. Al excluir las especies de origen sudamericano se observa como estas tienen una gran influencia en la diversidad de Panamá pero contribuye poco al resto de América Central (Fig. 2A vs Fig. 2C) y en especial, se sigue observando como Costa Rica sigue siendo la región con más riqueza de especies. Otro resultado llamativo es que al analizar solamente la fauna autóctona, se puede observar que existe mayor cantidad de linajes mayores habitando los bloques geológicos que conforman América Central Nuclear (Fig. 2B vs Fig. 2D). Otro patrón llamativo es la mayor predominancia de especies en zonas de elevación intermedia (Fig. 3), patrón que se pronuncia cuando se incluyen únicamente las especies autóctonas. Figura 3. Cantidad de especies según rango altitudinal (m.s.n.m) en América Central, en la izquierda se observa el conteo para el total de especies en la región y en la derecha se observa solamente las especies autóctonas. 216 14 Tabla 1. Componente de géneros en América Central según sus tres distintos orígenes. El número entre paréntesis corresponden con el número de especies. Norte América América Central Sur América Anura Lithobates (11) Agalychnis (6) Allobates (1) Rana (1) Anotheca (1) Ameerega (1) Rhinophrynus (1) Bromeliohyla (1) Andinobates (4) Charadrahyla (1) Anomaglossus (2) Craugastor (96) Atelopus (8) Cruziohyla (1) Cochranella (2) Ctenophryne (1) Colostethus (3) Diasporus (8) Dendrobates (1) Dryophytes (2) Dendrosophus (5) Duellmanohyla (7) Elachistocleis (3) Ecnomiohyla (7) Engystomops (1) Eleutherodactylus (3) Espadarana (1) Exerodonta (5) Gastroteca (2) Gastrophryne (2) Hemiphractus (1) Hypopachus (4) Hyalinobatrachium (8) Incilius (29) Hyloscirtus (2) Isthmohyla (15) Hyloxalus (1) Plectrohyla (18) Hypsiboas (5) Ptychohyla (9) Leptodactylus (7) Smilisca (6) Oophaga (5) Tlalocohyla (2) Osteopilus (1) Triprion (1) Phyllobates (2) Phyllomedusa (1) Pipa (1) Pleuroderma (1) Pristimantis (14) Rhaebo (1) Rhinella (6) Sachatamia (2) Scinax (6) Silverstoneia (2) Strabomantis (2) Teratohyla (2) Trachycephalus (1) 217 15 Tabla 1. Continua. Caudata Bolitoglossa (93) Bradytriton (1) Cryptotriton (6) Dendrotriton (8) Ixalotriton (2) Nototriton (17) Nyctanolis (1) Oedipina (35) Pseudoeurycea (4) Gymnophiona Dermophis (6) Caecilia (4) Gymnopis (2) Oscaecilia (4) Tabla 2 Riqueza de especies anfibias en América Central por país. País Área (Km2) Familias Géneros Especies Especies / 1000 Km2 México1 246 226 12 37 119 0.48 Guatemala 108 889 12 36 167 1.53 Belice 22 966 12 24 41 1.78 Honduras 112 492 12 41 144 1.28 El Salvador 21 041 11 21 40 1.90 Nicaragua 129 494 14 37 76 0.59 Costa Rica 51 100 16 47 203 3.97 Panamá 78 569 16 55 217 2.76 1 Incluye los Estados al este del Istmo de Tehuantepec (Campeche, Chiapas, Quintana Roo, Tabasco y Yucatán). Áreas de endemismo para los anfibios de América Central Un área de endemismo es definida como el área de congruencia distribucional no azarosa entre al menos dos taxones (Morrone 1994), esta congruencia implica simpatría. Las áreas de endemismo son unidades básicas para los estudios evolutivos y biogeográficos, para identificar estas áreas de congruencia es sugerido el uso del Análisis de Parsimonía de Endemismos (PAE, Morrone 1994). Se realizó un PAE para identificar las áreas de endemismo de los anfibios de América Central. América Central fue dividido en 110 celdas de 1ºx1º y utilizamos las mismas 459 áreas 218 16 de distribución de los anfibios de América Central. La matriz de presencia-ausencia fue analizada con NONA y WINCLADA, los taxones no-informativos fueron descartados y el análisis de parsimonía fue usado con 50 réplicas y la búsqueda heurística TBR+TBR. Los clados con al menos dos sinapomorfías (especies) fueron considerados áreas de endemismos, la simpatría de estas áreas de endemismo fueron evaluadas en el espacio geográfico. Nosotros identificamos siete áreas de endemismo (Fig. 4), cuatro de estas en América Central Nuclear y tres en América Central Ístmica. Estas áreas de endemismo no fueron correlacionadas con la riqueza de especies, excepto en Costa Rica donde hay alta riqueza de especies y dos áreas de endemismo. Todas las áreas de endemismo corresponden con áreas de elevación intermedia resaltando la importancia de la complejidad topográfica en la diversificación de los anfibios de América Central. La simpatría de las especies en estas áreas de endemismo fue alta, resaltando el rol de estas áreas en la diversificación de diferentes taxones, por ejemplo, la simpatría de cinco especies apoya el área de endemismo del caribe de Costa Rica (Fig. 5) Figura 4. Áreas de endemismo de los anfibios de América Central. 219 17 Figura 5. Mapa de Costa Rica mostrando un área de endemismo apoyada por cinco especies de anfibios. BIOGEOGRAFÍA HISTÓRICA DE LOS ANFIBIOS DE AMÉRICA CENTRAL Principales estudios durante el siglo XX El primer gran trabajo abordando la historia biogeográfica de los anfibios de América Central fue presentado por Savage (1966). El autor concluyó que la fauna anfibia de la región es producto de la confluencia de tres grandes grupos: 1) un grupo centroamericano con su origen en la región, 2) un grupo sudamericano el cual el autor considera que contribuye poco a la diversidad y se restringe principalmente en el extremo este de Panamá y 3) un grupo norteamericano que tiene sus más cercanos ancestros fuera de América Tropical (ej. salamandras). Savage (1966) concluye que los anfibios de América Central derivan en su gran mayoría de un ancestro en común con clados presentes en Sudamérica, él propone que esto es producto de una conexión terrestre entre América Central Nuclear y Sudamérica en el Paleoceno (66-56 Ma). Savage (1966) sugiere que posterior a la conexión entre América Central Nuclear y 220 18 Sudamérica estas estuvieron separadas durante el Eoceno al Plioceno (56-5 Ma), permitiendo la evolución de forma aislada. Por último, el autor sugiere que durante el Plioceno (~5 Ma) se dio el inicio de la formación del Istmo, permitiendo la colonización de América Central por especies sudamericanas. Savage (1966) también basado en su análisis de distribución y endemismo de los géneros de anfibios y reptiles en América Central, plantea que la fauna anfibia de esta región debe catalogarse como una unidad independiente al mismo nivel que los ensamblajes neárticos o neotropicales. El trabajo de Savage (1966) sirvió como base para la revisión de la historia biogeográfica de los anfibios y reptiles de América Central por el mismo autor (Savage 1982). En esta extensa revisión el autor discute las diferentes posiciones metodológicas y teóricas de la biogeografía en aquel momento (vicarianza vs. dispersión) utilizando como modelo las especies de anfibios y reptiles de la región. Savage (1982) llega a las mismas conclusiones que él alcanzó en Savage (1966), él refinó que los eventos que en conjunto originaron la fauna anfibia de América Central son los siguientes: 1) Un evento de dispersión desde Sudamérica, aunque evita sugerir la forma en que esta dispersión ocurrió, 2) un evento de vicarianza entre América Central y Sudamérica, 3) un segundo evento de dispersión desde Norteamérica, 4) un segundo evento de vicarianza entre América Central y Norteamérica y 5) un tercer evento de dispersión desde Sudamérica al completarse el cierre del Istmo de Panamá. Savage (1982) también reanaliza los datos de angiospermas de Raven & Axelrod (1974) y peces de agua dulce de Bussing (1976) y demuestra que estos dos grupos son concordantes con los patrones que él encontró. El modelo de conformación propuesto por Savage (1966, 1982) se ha mantenido como el modelo más aceptado. Otros trabajos biogeográficos para la región Bussing (1985), Duellman (1979, 2001) y Savage (2002) han respaldado en lo general esta propuesta. Sin embargo, la principal discrepancia se ha centrado en la forma en que ocurrió la primera dispersión desde Sudamérica hacia América Central. Duellman (1979) argumentó a favor de una primera dispersión sobre el mar o a través de islas. Savage (1982) evita argumentar a favor de alguna hipótesis y enumera las posibles explicaciones incluyendo una conexión por las proto-Antillas (~80 Ma) o el modelo propuesto por Duellman (1979). Savage (1982) sugiere que una hipótesis alternativa sería una primera dispersión post-Mioceno (~23 Ma) a través de la actual América Central (asumida como un complejo de islas) que conllevó a una rápida diversificación en la 221 19 región. Más recientemente Duellman (2001) y Savage (2002) han argumentado a favor de una primera dispersión a través de las proto-Antillas durante el Cretácico Tardío (~84 Ma). Hipótesis actuales a la luz de la evidencia molecular El siglo XXI trajo un uso cada vez más generalizado del ADN en la construcción de filogenias. Filogenias moleculares existen hoy para los principales grupos anfibios de América Central como las ranas arborícolas (Faivovich et al. 2005, 2010; Duellman et al. 2016), salamandras neotropicales (Parra-Olea et al. 2004; Wiens et al. 2007; Rovito et al. 2015a) y los sapos mesoamericanos (Mendelson et al. 2011). Posiblemente, el entendimiento de las relaciones filogenéticas en las ranas de desarrollo directo ha sido el más beneficiado con el uso de las herramientas moleculares. El alto grado de homoplasia y polimorfismos fenotípicos entre dichas ranas impidió resolver sus relaciones filogenéticas hasta el uso de ADN en el siglo presente. Casi 1000 especies conformaban hasta entonces el gran género Eleutherodactylus (Frost et al. 2004). Las relaciones filogenéticas entre las ranas de desarrollo directo continúa cambiando (Hedges et al. 2008; Pyron & Wiens 2011; Padial et al. 2014), pero existe amplia concordancia a favor de muchos de sus géneros, familias y subfamilias. Actualmente, dichas ranas se agrupan en tres familias y 25 géneros (Padial et al. 2014). Las hipótesis biogeográficas para América Central también se ha beneficiado del uso de las herramientas moleculares. En la biogeografía cladística un requisito fundamental es conocer la monofilia de los grupos en cuestión. Con relaciones filogenéticas pobremente entendidas entre las ranas de desarrollo directo estas carecían de utilidad el estudio biogeográfico de la región, pero, a medida que sus relaciones filogenéticas son mejor entendidas su importancia para evaluar hipótesis biogeográficas se hace notable. Además, las ranas de desarrollo directo como grupos de estudio tienen las siguientes ventajas: su alta diversidad en la región, su alta abundancia relativa y su ubicuidad. En esta sección se revisa cómo el estudio de las ranas de desarrollo directo en la era de la sistemática molecular han tenido un papel importante en apoyar las hipótesis actuales de la biogeografía de América Central. Además, otros grupos biológicos son utilizados para respaldar dichos hallazgos. La primera dispersión y diversificación en América Central. Como propuso Savage (1966), la explicación de la diversidad actual en América Central requirió una primera dispersión hacia la 222 20 región desde Sudamérica y Norteamérica (salamandras). El único grupo de origen norteamericano con diversificación regional son las salamandras de la tribu Bolitoglossini (Savage 1966, 1982, 2002; Rovito et al. 2015a). El arribo de las salamandras a América Central desde Norteamérica se entiende como un evento de dispersión sobre las tierras continentales actuales, esto debido a la conexión terrestre entre dichas unidades es anterior al arribo de las salamandras a América Central. Rovito et al. (2015a) reconstruyen la historia filogenética y biogeográfica para las salamandras neotropicales (Bolitoglossini), encontrando que todas las salamandras de América Central constituyen un grupo monofilético y rechazando la parafilia encontrada por Wiens et al. (2007). Además, encontraron que el origen de las salamandras neotropicales se remonta al Eoceno Medio (47-37 Ma) y resaltaron que América Central Nuclear tuvo un papel importante en la diversificación temprana de las salamandras neotropicales, en la diversificación de géneros y subgéneros. Basado en los hallazgos de Rovito et al. (2015a) y en los argumentos utilizados por Savage (1966, 1982) para clasificar los orígenes de la fauna anfibia, las salamandras neotropicales deben ser consideradas como un clado centroamericano y no un clado norteamericano como lo consideró Savage (1966, 1982, 2002). La primera dispersión desde Sudamérica hacia América Central es más conflictiva; Savage (1966) concluyó que una primera dispersión tuvo que ocurrir previo al cierre del Istmo de Panamá, seguido de un aislamiento que llevó al origen de la mayor parte de la fauna anfibia autóctona de la región. El origen sudamericano de todo el clado de las ranas de desarrollo directo (Hedges et al. 2008) y su presencia actual en América Central lo convierte en un grupo ideal para evaluar esta primera dispersión. Crawford & Smith (2005) evaluaron la evolución y biogeografía de ranas de desarrollo directo en América Central. Estos autores asumen a priori que la primera dispersión hacia América Central desde Sudamérica ocurrió a través de las proto-Antillas y siguiendo a Savage (1966, 1982) restringen el origen de Craugastor al intervalo de 80-60 Ma. Con esta calibración ellos encontraron que el origen de Craugastor se remonta al Cretácico- Paleoceno (75-66 Ma) y argumentan a favor de una primera dispersión sobre las proto-Antillas. Crawford & Smith (2005) argumentan que el movimiento de las proto-Antillas hacia el este, aisló los ancestros de Craugastor en América Central Nuclear lo que llevó a su origen. Esta conclusión la basan en el hecho de que los clados más basales de Craugastor (C. milesi y C. augusti + C. alfredi ) tienen como extremo sur de su distribución a Honduras. Crawford & Smith (2005) también proponen que los ancestros de Eleutherodactylus arribaron a las Antillas 223 21 mediante GAARlandia hace unos 46-25 Ma y posteriormente los Eleutherodactylus continentales llegaron a América Central Nuclear desde las Antillas en el Oligoceno (~35-26 Ma). El estudio de Crawford & Smith (2005) tenía dos limitantes, primero, la calibración restringida por un modelo biogeográfico y segundo, la falta de fósiles o grupos filogenéticamente cercanos a Craugastor en las Antillas como lo predice un modelo de colonización mediante las proto-Antillas. Heinicke et al. (2007) aumentaron el muestreo de especies y caracteres moleculares y revisaron el origen de las ranas de desarrollo directo en América Central y el Caribe, pero esta vez calibrando con el registro fósil. Los autores encontraron que los orígenes de Craugastor en América Central y de Eleutherodactylus en el Caribe son producto de dispersiones oceánicas. Heinicke et al. (2007) sugieren que el origen de Craugastor y Eleutherodactylus se ubica en el Eoceno, ~42-31 Ma y ~47-29 Ma respectivamente, mientras que el origen del clado continental de Eleutherodactylus se data en ~19 Ma producto de otra dispersión oceánica desde Cuba a Yucatán. Los autores rechazan las hipótesis de dispersión sobre las conexiones terrestres de proto-Antillas y GAARlandia. Sin embargo, es importante resaltar que los tiempos propuestos para la dispersión de Sudamérica hacia América Central y las Antillas (~47-29 Ma) son coincidentes con la propuesta de GAARlandia que se ubica en ~35-33 Ma. Recientemente Duellman et al. (2016) revisaron las relaciones filogenéticas de las ranas arborícolas (Hylidae), encontrando que el origen de Hylinae (subfamilia con origen en América Central) se remonta al Oligoceno Temprano (35.6-30.2 Ma). Aunque esta dispersión es congruente también con la hipótesis de GAARlandia los autores no la sugieren como posible medio, quizás, porque tampoco existen fósiles ni grupos filogenéticamente cercanos a Hylinae en las Antillas. Říčan et al. (2013) encontraron que la dispersión de peces dulceacuícolas hacia América Central y las Antillas desde Sudamérica se ajusta a la hipótesis de GAARlandia. Říčan et al. (2013) proponen que una primera dispersión desde Sudamérica hacia las Antillas corresponde con la presencia de GAARlandia (~35-29 Ma), posteriormente, América Central Nuclear fue colonizado por al menos dos eventos de dispersión desde las Antillas. El trabajo de Říčan et al. (2013) rechaza los hallazgos de Chakrabarty (2006) quien había sugerido que el origen de estos peces se remontaba al Cretácico-Paleoceno (~66 Ma) y quien propuso que la vía de colonización era coincidente con las proto-Antillas. Alonso et al. (2012) al revisar la filogenia 224 22 y biogeografía del género de sapos Peltophryne en Cuba, encontraron que la dispersión de estos anfibios a las Antillas desde Sudamérica es concordante con la conexión de GAARlandia. La primera dispersión desde Sudamérica hacia América Central sigue bajo discusión, los tres ejemplos anteriores (Heinicke et al. 2007; Říčan et al. 2013; Duellman et al. 2016) rechazan la hipótesis de las proto-Antillas. Más discutida sigue siendo la hipótesis de GAARlandia (Iturralde-Vinent & MacPhee 1999), ejemplificado anteriormente (Říčan et al. 2013) y como lo notó Alí (2012) dicha hipótesis ha sido bien recibida como explicación espacio-temporal de dispersión desde Sudamérica hacia las Antillas y América Central. Alí (2012) argumenta en contra de GAARlandia como una conexión continua entre Sudamérica y las Antillas y como medio generalizado de dispersión entre estas dos masas terrestres. Alí (2012) sugiere que si bien pudo existir la conexión de GAARlandia, es difícil que haya sido ampliamente utilizada como medio de dispersión y expone que muchos otros eventos de dispersión debieron ocurrir sobre el agua. La evidencia plantea que el origen de Craugastor y la subfamilia Hylinae, se explica mejor con eventos aislados de dispersión oceánica. También, es concordante la evidencia de que los Eleutherodactylus continentales tienen su origen en las Antillas, pero el origen de estos en las Antillas es incierto, Heinicke et al. (2007) argumentan a favor de dispersión oceánica, pero también es cierto que pudieron arribar por GAARlandia. Más investigación será necesaria para esclarecer el origen de este clado. Una hipótesis alternativa que Savage (1982) sugirió pero sin detallar, es que los ancestros de Craugastor e Hylinae hayan dispersado por una América Central Ístmica primigenia durante el Oligoceno Temprano (~34-28 Ma). Está hipótesis alternativa tiene la ventaja de que ubica una conexión directa entre Sudamérica y América Central Nuclear y evita recurrir a la dispersión de largas distancias sobre el agua, eventos que son altamente improbables para fisiologías de organismos como anfibios o peces de agua dulce. La desventaja de esta hipótesis es que los centros de diversificación basal para Craugastor e Hylinae se encuentran en América Central Nuclear, lo cual obliga a recurrir a una extinción de los ancestros en América Central Ístmica posterior a la colonización de América Central Nuclear. La evidencia revela que durante el Eoceno (~56-33 Ma), América Central Ístmica existía como un estrecho cinturón de tierras emergidas de baja elevación que variaba de forma peninsular en el Norte (Costa Rica) a insular en el sureste (Panamá) y separado de Sudamérica por el Canal Marítimo Centroamericano (Montes et al. 2012a,b, 2015). El levantamiento continuó y para el Mioceno Temprano (~23 Ma) 225 23 América Central Ístmica ya existía como una península que se extendía desde el extremo sur de América Central Nuclear hasta el Darién (extremo este de Panamá) separado de Sudamérica por ~200 Km de océano (Kirby & MacFadden 2005; Montes et al. 2012b). Se ha propuesto que durante el Oligoceno Temprano (~30 Ma) el nivel del mar fue inferior al actual (Haq et al. 1987), descenso que fue argumentado a favor de la hipótesis de GAARlandia (Alí 2012). A pesar de que los estudios sobre la reconstrucciones del nivel del mar han sido contrastantes (ver Clima y nivel del mar en el pasado), es posible que este descenso –si existió– pudo haber permitido la primera dispersión desde Sudamérica hacia América Central Nuclear. El posterior aumento del nivel del mar pudo haber inundado las tierras apenas superficiales de América Central Ístmica y extinguido su fauna incipiente. La colonización de América Central Ístmica previo al cierre del Istmo de Panamá. Este momento histórico en América Central no debe ser entendido como un evento aislado, sino, como pulsos discontinuos de dispersión de la fauna ya presente en América Central Nuclear hacia el sur colonizando las tierras de Costa Rica y Panamá a medida que estas se levantaban. Es probable que estos pulsos de dispersión hayan sido separados por altos niveles del mar lo cual aislaba temporalmente estas dos unidades de América Central, propiciando la evolución de diferentes grupos (Castoe et al. 2009; Gutiérrez-García & Vázquez-Domínguez 2013). Posiblemente el inicio de estos pulsos de dispersión se remonta al Mioceno Temprano (~23 Ma), cuando América Central Ístmica había emergido y se prolongaba como una península en su posición actual (Montes et al. 2012a). Estos primeros eventos de dispersión corresponden con la aparición del grupo de especies Craugastor podiciferus hace ~20 Ma (Streicher et al. 2009). El grupo de especies C. podiciferus contiene diez especies de ranas de desarrollo directo que tiene su diversificación en Costa Rica y el oeste de Panamá, el grupo hermano del clado C. podiferus es el clado C. rhodopis que habita desde el sur de México hasta el norte de Honduras (Hedges et al. 2008; Padial et al. 2014). Otros grupos apoyan esta primera dispersión desde América Central Nuclear hacia América Central Ístmica, Duellman et al. (2016) encontraron que el origen y diversificación del género Isthmohyla en América Central Ístmica data de ~19.5 Ma. Říčan et al. (2013) encontraron que los peces dulceacuícolas de América Central Ístmica llegaron desde el norte entre 19-13 Ma. Rovito et al. (2015a) sugirieron que las salamandras se dispersaron hacia 226 24 el sur y que entre 16.4-8 Ma arribaron a Sudamérica, por lo cual debieron llegar a América Central Ístmica previamente. El cierre del Istmo de Panamá y el Gran Intercambio de la Biota Americana. El cierre del Istmo de Panamá se asumió históricamente en ~3 Ma (Coates & Obando 1996; Coates et al. 2004), sin embargo, evidencia reciente ubica el cierre entre 15-13 Ma (Montes et al. 2015). Previo a los descubrimientos del cierre del Istmo de Panamá en el Mioceno (Montes et al. 2012a,b, 2015) algunos estudios indicaban que varias dispersiones desde Sudamérica hacia América Central habían ocurrido previo a ~3 Ma (excluyendo las primeras dispersiones de Craugastor e Hylinae). Wang et al. (2008) estudiaron la filogeografía de la rana de desarrollo directo Pristimantis ridens, género de origen sudamericano, y encontraron que el origen del clado conteniendo las poblaciones de Costa Rica y Honduras ocurrió ~12 Ma, por lo que esta especie dispersó desde Sudamérica mucho antes del cierre del Istmo de Panamá. Pinto-Sánchez et al. (2012) realizaron un estudio intensivo con ranas Pristimantis tanto al norte como al sur del Istmo de Panamá, en busca de dispersiones desde Sudamérica hacia América Central previas al cierre del Istmo de Panamá. Ellos descubrieron que el género Pristimantis efectivamente tiene su origen en Sudamérica, donde habita la mayoría de su diversidad (489 especies), y encontraron que los Pristimantis de América Central son producto de múltiples invasiones desde Sudamérica entre 12-6 Ma. Estos autores refirieron ocho eventos individuales de dispersión de Pristimantis que se distribuyen tanto en América Central como en Sudamérica, pero también, encontraron tres eventos de dispersión que han generado linajes endémicos de América Central, entre ellos el grupo de especies P. pardalis que diversificó en al menos tres especies. Los resultados anteriores son concordantes con otros encontrados en peces dulceacuícolas (Bermigham & Martin 1998) y pseudoescorpiones (Zeh et al. 2003). Estas primeras dispersiones desde Sudamérica hacia América Central originaron algunos géneros de anfibios y reptiles autóctonos para la región como Agalychnis, Bothriechis, Atropoides y Cerrophidion. Duellman et al. (2016) en un estudio con las ranas arborícolas, encontraron que posterior a la primera dispersión que dio origen a Hylinae, al menos otros seis eventos de dispersión ocurrieron desde Sudamérica hacia América Central, todos durante el Mioceno (23-5 Ma). Una de estas dispersiones dio origen al género Agalychnis que divergió de sus ancestros sudamericanos entre 12.3-7.9 Ma. Castoe et al. (2009) encontraron dos eventos de dispersión en serpientes de foseta 227 25 neotropicales entre 15-10 Ma, que dieron origen a los géneros Bothriechis, Atropoides y Cerrophidion. La única dispersión bien documentada de un anfibio desde América Central hacia Sudamérica es la del género de salamandras Bolitoglossa, que ha sido datada entre 16.4-8 Ma (Rovito et al. 2015a), concordante con los resultados anteriores en haber ocurrido previo a ~3 Ma. El Gran Intercambio de la Biota Americana se refiere al gran intercambio biótico que siguió al cierre del Istmo de Panamá (Webb 1991, 2006). Bacon et al. (2015) basados en la evidencia reciente que ubica el cierre del Istmo de Panamá entre 15-13 Ma (Montes et al. 2015), evaluaron si este cierre temprano es concordante con un Gran Intercambio de la Biota Americana previo a ~3 Ma. Bacon et al. (2015) encontraron que el Gran Intercambio de la Biota Americana ocurrió de forma generalizada antes de los 3 Ma, en forma de pulsos entre 20-6 Ma, resultados concordantes para anfibios, reptiles, aves, artrópodos y plantas. Sin embargo, múltiples estudios realizados con mamíferos apoyan el Gran Intercambio de la Biota Americana entre 3.5-2.5 Ma (Webb 1991, 2006). Recientemente, Bacon et al. (2016) analizaron por qué el intercambio de mamíferos entre Norte- y Sudamérica no concuerda con el cierre del Istmo de Panamá hace 15- 13 Ma, habiendo encontrado que el cruce tardío de los mamíferos se debió a factores climáticos y medioambientales, más que la configuración geológica de la región. Todos los resultados anteriores son contundentes en apoyar una serie de pulsos discontinuos de dispersión desde Sudamérica hacia América Central y viceversa, posterior al cierre del Istmo de Panamá hace 15-13 Ma. Esto demuestra que para los anfibios, el Gran Intercambio de la Biota Americana es concordante con el cierre del Istmo de Panamá en el Mioceno. Sin embargo, existe un grupo grande de especies de origen sudamericano que carecen de estudios que daten su presencia en América Central. Estas especies pertenecen a géneros (ej. Atelopus, Elachistocleis y Gastroteca) o inclusive familias (Dendrobatidae, Leptodactylidae y Centrolenidae) con sus centros de origen y mayor diversidad en Sudamérica. Muchas de estas especies poseen distribución compartida entre América Central y Sudamérica, lo cual se ha interpretado como el producto de dispersiones recientes (<~3 Ma) desde Sudamérica hacia América Central. Es necesario datar la presencia de dichas especies en América Central, porque podría ser que efectivamente un grupo grande de estas especies hayan ingresado tan recientemente como ~3 Ma, quizás respondiendo a los mismos factores climáticos que impulsaron la migración de mamíferos (Bacon et al. 2016). Pero quizás, a pesar de que 228 26 efectivamente estas especies pertenecen a géneros de origen sudamericano, su presencia en América Central corresponda con el cierre del Istmo de Panamá hace 15-13 Ma, mucho antes de ~3 Ma, tal como lo encontró Pinto-Sánchez et al. (2012) en Pristimantis. PERSPECTIVAS ACTUALES: UNA MIRADA DESDE LAS RANAS DE DESARROLLO DIRECTO Especies crípticas Con el desarrollo de nuevas tecnologías que facilita el obtener secuencias moleculares en poco tiempo y a bajo costo, surgió un cambio de paradigma taxonómico: transformar la taxonomía tradicional en una taxonomía integradora (Dayrat, 2005; Will et al. 2005). El objetivo es claro, robustecer las conclusiones taxonómicas con el apoyo de evidencia filogenética, morfológica, filogeográfica y demás en cuanto sea posible (Dayrat, 2005; Will et al. 2005). A pesar de que la implementación de la taxonomía integradora traería grandes beneficios para la taxonomía y el entendimiento de la biodiversidad (Padial et al. 2010) aún su implementación no se generaliza en la práctica taxonómica (Pante et al. 2015). La implementación de herramientas taxonómicas integrativas en grupos con alta diversidad críptica puede ser especialmente beneficiosa (Padial & De la Riva 2009; Padial et al. 2010). Según Bickford et al. (2007) las especies crípticas son aquellas que han sido tratadas como un único taxón dado que son al menos superficialmente indistinguibles. Esta característica intrínseca de la diversidad críptica, justifica el atender este fenómeno desde un enfoque integrativo con adecuados protocolos de delimitación, que incluya el uso de herramientas moleculares, morfológicas, ecológicas u otras fuentes de evidencia particulares a cada grupo biológico. El fenómeno de especies crípticas tiene profundas implicaciones en la sistemática, la conservación, la biología evolutiva y en la biogeografía (Nadler & Pérez-Ponce de León 2010). La presencia de especies crípticas no delimitadas correctamente implica el enmascaramiento de biodiversidad, puede conllevar a conclusiones biogeográficas erróneas y a la pérdida de evidencia para apoyar o refutar patrones biogeográficos. Recientemente se han propuestos enfoques integradores para enfrentar el fenómeno de las especies crípticas. Pérez-Ponce de León & Nadler (2010) proponen atender el problema desde un enfoque molecular de prueba de hipótesis, sugiriendo que si la hipótesis nula de un solo taxón se rechaza, entonces una re-evaluación morfológica de los clados moleculares puede revelar 229 27 diferencias hasta entonces ocultas. También, se han propuesto protocolos detallados para delimitar diversidad críptica en anfibios (Padial et al. 2010; Ortega-Andrade et al. 2015). Y de forma general, grandes avances se han dado en el campo de la delimitación de especies, cuyas propuestas pueden contribuir a delimitar la diversidad críptica (Sites & Marshall 2003; Wiens 2007; Flot 2015). El consenso sugiere que la delimitación de la diversidad críptica en anfibios requiere en un primer paso, generar una filogenia molecular que permita agrupar la diversidad en grupos monofiléticos. Posteriormente, con base en estos clados se selecciona una muestra de la cual se recupera información morfológica, morfométrica, acústica o ecológica y finalmente la delimitación es basada en la concordancia de la evidencia. Se ha sugerido que la distribución de especies crípticas no es proporcional entre distintos grupos biológicos (Pérez-Ponce de León & Poulin 2016) y que los anfibios, en especial los Anuros, pueden albergar mayor diversidad críptica entre los vertebrados (Bickford et al. 2007; Bagley & Johnson 2014). Las ranas de desarrollo directo son uno de los grupos sugeridos a albergar gran diversidad de especies crípticas (Padial & De la Riva 2009). En América Central el género Craugastor es el más diverso de las ranas de desarrollo directo y se presume que contiene muchas especies no descritas, ocultas por las especies actualmente reconocidas (Savage 2002; Crawford et al. 2007; Streicher et al. 2009). En este apartado del ensayo se revisa como la taxonomía integrativa, la delimitación de especies y la filogeografía, contribuyen a resolver problemas de especies crípticas y con esto a mejorar el entendimiento de la sistemática y biogeografía en Costa Rica y Panamá. Como se comentó anteriormente (ver Diversidad de anfibios en América Central y sus patrones de distribución) Costa Rica y Panamá representan la región más diversa en anfibios de América Central, en especial Costa Rica y el oeste de Panamá son los sitios de mayor diversificación y endemismo regional. Un estudio actualmente en desarrollo, donde se revisa la sistemática y taxonomía de las ranas del grupo de especies Craugastor podiciferus en América Central, sugiere la presencia de varias especies no descritas dentro del grupo (Datos sin publicar). Una de estas especies, C. gabbi Arias, Chaves, Crawford & Parra-Olea, 2016 estuvo históricamente bajo el nombre C. stejnegerianus (Cope, 1893) producto de una morfología extremadamente conservada. Sin embargo, análisis moleculares apoyan su monofilia y ubican al clado C. stejnegerianus + C. rearki (Taylor, 1952) como su grupo hermano (Fig. 6). Con base en la inferencia molecular se realizaron análisis morfométricos y de morfología externa y se encontró una característica 230 28 diagnóstica que permitió su correcta delimitación y descripción. La delimitación de esta especie tiene implicaciones biogeográficas, pues, esta es endémica de la región premontana (1200-1500 m.s.n.m) del pacífico sur de Costa Rica, respaldando descubrimientos previos de anfibios endémicos para la región (Brame & Duellman 1970; García-París & Wake 2000; Wake et al. 2007). Crawford (2003) calculó que el origen de la nueva especie data entre 10.3-6.3 Ma, respaldando la importancia biogeográfica del levantamiento de la Fila Costeña hace 12.8-11.7 Ma (MacMillan et al. 2004). El uso de herramientas moleculares en el grupo de especies C. podiciferus, también ha permitido esclarecer especies mal clasificadas dado su alto grado de similitud morfológica. Gracias a las inferencia moleculares se pudo determinar que la especie C. rearki, hasta entonces sinónimo de C. bransfordii, es una especie válida (Fig. 6). Además, se demostró que C. rearki se distribuye hasta el este de Honduras incluyendo las poblaciones anteriormente asignadas a C. lauraster, la cual se colocó bajo sinonimia de la primera. Figura 6. Mapa de América Central mostrando la distribución de un clado dentro del grupo de especies Craugastor podiciferus. Las flechas indican poblaciones que fueron en el pasado clasificadas como C. stejnegerianus, incluyendo a la especie C. gabbi. La línea blanca señala el río Savegre. 231 29 Filogeografía como herramienta para descubrir especies crípticas La filogeografía estudia la distribución espacial de linajes genéticos dentro de una especie o especies estrechamente relacionadas (Avise 2009), esto hace de la filogeografía una herramienta especialmente útil para el descubrimiento de especies crípticas. Bagley & Johnson (2014) revisaron los estudios filogeográficos realizados en América Central Ístmica y encontraron que en promedio los estudios registraron 2.1 linajes genéticos independientes por cada taxón. Los grupos con mayor diversidad críptica registrada son los peces dulceacuícolas y los anuros. Las ranas de desarrollo directo se han convertido en un excelente modelo para estudios filogeográficos en la región (Wang et al. 2008; Crawford et al. 2007; Streicher et al. 2009; Crawford et al. 2013; Paz et al. 2015). En un estudio filogeográfico con la rana Craugastor podiciferus en Costa Rica y el oeste de Panamá, se encontró la presencia de una especie críptica y se dató la divergencia entre 11.8–6.6 Ma (Streicher et al. 2009). En otro estudio filogeográfico con Craugastor en Costa Rica y Panamá, se encontró que al menos tres especies crípticas permanecen sin describir (Crawford et al. 2007). Estas tres especies no descritas de Craugastor se distribuyen por encima de los 1000 m de elevación, resaltando la importancia de las tierras intermedias en la diversificación regional (ver Fig. 3). Los ejemplos anteriores junto con la distribución altitudinal de las especies de América central (Fig. 3) son concordantes en apoyar la importancia de la diversificación en tierras intermedias. Sin embargo, existe controversia en la datación cronológica de los procesos responsables de esta diversificación. Savage (2002), basado en la evidencia de oscilaciones climáticas y el rápido levantamiento de las tierra altas de América Central en el Plioceno- Pleistoceno (~5 Ma), propuso un modelo de especiación de tierras altas (especiación montana). Savage (2002) sugiere que las oscilaciones entre periodos glaciares e interglaciares que comenzaron en el Plioceno (~5 Ma) y se continuaron durante el Pleistoceno, son las responsables del origen y distribución de muchas especies de tierras altas. Durante periodos glaciares los bosques montanos estuvieron conectados, periodos subsecuentes de calentamiento llevaron al aislamiento de estos bosques en las cumbres de las cordilleras centroamericanas lo cual conllevó al origen de muchas de las especies actuales (Savage 2002). Sin embargo, análisis filogeográficos con la rana Craugastor podiciferus (Streicher et al. 2009) y con serpientes de foseta neotropicales (Castoe et al. 2009) han rechazado que estos eventos del Plioceno- 232 30 Pleistoceno estén asociados con cladogénesis dentro de estas especies de altura. Más estudios son necesarios para entender la diversificación en las tierras altas de la región. En un estudio actualmente en desarrollo, la evidencia molecular sugiere que los eventos durante el Pleistoceno provocaron la formación de al menos 11 especies de salamandras de musgo (Nototriton) en la tierras altas de Costa Rica (Datos sin publicar). Estos resultados respaldan la propuesta de Savage (2002) de que los eventos climáticos del Plioceno-Pleistoceno generaron diversificación en los anfibios de América Central. La discrepancia entre los resultados encontrados en Nototriton y aquellos encontrados por Streicher et al. (2009) y Castoe et al. (2009) podrían reflejar las diferentes capacidades de dispersión entre los grupos estudiados. Las salamandras del género Nototriton son altamente miniaturizadas (longitud estándar = ~30mm), semifosoriales y de baja movilidad (Savage 2002). Estos resultados con las salamandras Nototriton de Costa Rica, también respaldan la evidencia que sugiere que existe alta diversidad críptica dentro de las salamandras miniaturizadas del neotrópico en los géneros Chiropterotriton, Nototriton y Thorius (Darda 1994; Savage 2002; Townsend et al. 2011; Rovito et al. 2013). Nuestros resultados sugieren la presencia de cuatro especies no descritas, todas con alta similitud morfológica entre sí y con las demás especies descritas para el país (Datos sin publicar). Pero la evidencia molecular y la historia de vida de estos organismo como: la restricción altitudinal (1000-2500 m.s.n.m), la baja movilidad, y la ausencia de conectividad biogeográfica (Savage 2002) respalda el reconocimiento de las especies y la descripción de las mismas. La filogeografía comparada busca patrones filogeográficos compartidos dentro de especies codistribuidas (Gutiérrez-García & Vázquez-Domínguez 2011). El objetivo principal, es la búsqueda de patrones generales que permitan construir explicaciones para una gama amplia de organismos (Bagley & Johnson 2014). El filtro de Bocas del Toro en el oeste de Panamá se ha propuesto como una barrera geográfica, asociada a eventos de altos niveles del mar que afectó las poblaciones de tierras bajas (Coates et al. 2005). Esta barrera se ha correlacionado con fragmentaciones filogeográficas en peces (Perdices et al. 2002) y anfibios (Crawford et al. 2007), pero en los anfibios el nivel de fragmentación varió entre tres especies de Craugastor con diferentes requerimientos ecológicos (Crawford et al. 2007). Sin embargo, según Bagley & Johnson (2014) en América Central Ístmica hay una carencia generalizada de historia 233 31 biogeográfica compartida y sugieren que la respuesta a barreras varía entre especies y en temporalidad. A pesar de la carencia de patrones generales, es importante continuar en la búsqueda de historias compartidas que apoyen ciertas fragmentaciones, al menos de forma parcial. El descubrimiento de nuevas zonas de fragmentación donde no existen claras barreras actuales, obliga a profundizar en el estudio geológico y paleoclimático de la región. Profundas fragmentaciones entre poblaciones de la lagartija Anolis aquaticus en la costa pacífica de Costa Rica, llevó a la delimitación de una nueva especie cuya única barrera visible es la cuenca del río Savegre (Chaves et al. En revisión). Este mismo río es la única barrera visible para explicar la fragmentación encontrada en la rana Craugastor stejnegerianus en el pacífico costarricense (Fig. 6), cuya distancia genética entre poblaciones de ambos márgenes del río asciende al 1.8% en el gen mitocondrial 16S y 5.5% en el gen COI (Datos sin publicar). No existe explicación para soportar estas fragmentaciones concordantes, más allá del río, mismo que no varía de los otros ríos que se encuentran dentro del rango geográfico de estas especies. Quizás alguna explicación climática del Plioceno-Pleistoceno pueda ser hallada en el futuro. RETOS FUTUROS Se han realizado grandes avances en el entendimiento biogeográfico de América Central y en especial para anfibios. Sin embargo, persisten grandes vacíos de conocimiento por llenar. Los novedosos y controversiales hallazgos sobre el cierre temprano en el Mioceno Medio (15-13 Ma) del Istmo de Panamá (Montes et al. 2015) tiene implicaciones cruciales para el entendimiento biogeográfico de la región. Es necesario más evidencia que soporte o refute estos hallazgos y que consolide las reconstrucciones paleogeográficas del Istmo de Panamá. Asimismo, para entender las primeras dispersiones desde Sudamérica hacia América Central Nuclear es necesario generar un cuerpo de evidencia que apoye o refute la posibilidad de una conexión terrestre entre Sudamérica y las Antillas (GAARlandia), esto cuanto la evidencia actual es controversial y dudosa (Iturralde-Vinent & MacPhee 1999; Alí 2012). Como se detalló anteriormente, muchas de las hipótesis biogeográficas para América Central tienen supuestos sobre el nivel del mar en distintos momentos históricos, no obstante, las reconstrucciones sobre el nivel del mar carecen de concordancia (Miller et al. 2005; Kominz et al. 2008). La 234 32 biogeografía histórica y las hipótesis sobre América Central se verán beneficiadas de futuros hallazgos. Desde el campo de la sistemática filogenética son necesarios grandes avances. Más y mejores reconstrucciones filogenéticas son necesarias. Aún persisten vacíos de conocimiento en las relaciones filogenéticas de diversos grupos de anfibios en la región. Diversos grupos de ranas de desarrollo directo (Craugastor, Pristimantis y Diasporus) y salamandras (Bolitoglossa) carecen de filogenias sólidas (Padial et al. 2014; Boza-Oviedo et al. 2012). Si bien, la carencia de fósiles anfibios centroamericanos es generalizada (Rovito et al. 2015a; Duellman et al. 2016) las dataciones moleculares pueden mejorar significativamente. Recientemente Sheng et al. (2016) encontraron que el origen de la familia de salamandras neotropicales Plethodontidae es al menos 24 Ma más recientes que los cálculos anteriores. Dataciones moleculares a partir de grandes matrices de ADN nuclear, generó dataciones más recientes que aquellas alcanzadas con la utilización de ADN mitocondrial, los autores sugieren que las dataciones a partir de ADN mitocondrial deben ser analizadas con precaución y construir nuevas dataciones con el uso de ADN nuclear (Shen et al. 2016). Avances recientes han permitido el desarrollo de modelos y software que reconstruyen la evolución de rangos geográficos a partir de árboles filogenéticos (Ree et al. 2005; Ree & Smith 2008; Ree & Sanmartín 2009). Estos avances aún no se han implementado extensivamente en estudios con anfibios de América Central, a excepción de algunos trabajos con salamandras neotropicales (Rovito et al. 2012, 2015a, 2015b). Estos avances en biogeografía histórica tienen la ventaja de que permiten probar hipótesis sobre centros de dispersión y los posibles rangos ancestrales de un clado particular (Ree & Sanmartín 2009). Además, la implementación de modelos de dispersión-cladogénesis-extinción en el software Lagrange (Ree & Smith 2008) tiene la ventaja de que se basa en máxima verosimilitud para reconstruir el rango ancestral para una especie utilizando un árbol filogenético. En el futuro las reconstrucciones biogeográficas pueden tomar ventaja de estos avances en biogeografía paramétrica. CONCLUSIONES El análisis de los patrones de distribución de anfibios en América Central apoyó una distribución desproporcional, con una mayor concentración de especies en Costa Rica y Panamá. Costa Rica y el oeste de Panamá constituyen el sitio con la mayor riqueza de anfibios en la 235 33 región, tanto en el análisis con el total de las especies, como también al analizar únicamente las especies autóctonas. Esto resalta la importancia biogeográfica de Costa Rica y el oeste de Panamá, visto como un reservorio de especies que alcanzan en esta región su límite de la distribución y también como sitio de alta diversificación. De igual forma, en el ensayo se comprobó como el uso de herramientas moleculares han ayudado a un mejor entendimiento biogeográfico para América Central. Se pudo observar como los estudios moleculares recientes han rechazado la posibilidad de las proto-Antillas como medio de dispersión desde Sudamérica hacia América Central y como la evidencia sobre GAARlandia es todavía controversial. Se demostró como el uso de secuencias de ADN junto con los descubrimientos geológicos están cambiando aceleradamente el entendimiento sobre el Gran Intercambio de la Biota Americana, ahora entendida como pulsos de dispersión posterior al cierre del Istmo de Panamá hace 15-13 Ma. Las herramientas moleculares también han permitido descubrir gran diversidad de especies crípticas dentro de las ranas de desarrollo directo en América Central, resultando en mejores hipótesis biogeográficas para la región. AGRADECIMIENTOS EA agradece al Posgrado en Ciencias Biológicas por su apoyo para este estudio, la CONACyT por la beca de estudio (CVU/Becario) 626946/330343 y al Programa de Innovación y Capital Humano para la Competitividad PINN-MICITT por la beca de estudio (PED-0339-15- 2). LITERATURA CITADA Ali, J.R. (2012) Colonizing the Caribbean: is the GAARlandia land‐bridge hypothesis gaining a foothold? Journal of Biogeography, 39: 431–433. Alonso, R., Crawford, A.J. & Bermingham, E. (2012) Molecular phylogeny of an endemic radiation of Cuban toads (Bufonidae: Peltophryne) based on mitochondrial and nuclear genes. Journal of Biogeography, 39: 434–451. Arias, E., Chaves, G., Crawford, A.J. & Parra-Olea, G. (2016) A new species of the Craugastor podiciferus species group (Anura: Craugastoridae) from the premontane forest of southwestern Costa Rica. Zootaxa, 4132: 347–363. Avise, J.C. (2009) Phylogeography: retrospect and prospect. Journal of Biogeography, 36: 3–15. 236 34 Bacon, C.D., Silvestro, D., Jaramillo, C., Smith, B.T., Chakrabarty, P. & Antonelli, A. (2015) Biological evidence supports an early and complex emergence of the Isthmus of Panama. Proceedings of the National Academy of Sciences, 112: 6110–6115. Bacon, C.D., Molnar, P., Antonelli, A., Crawford, A.J., Montes, C. & Vallejo-Pareja, M.C. (2016) Quaternary glaciation and the Great American Biotic Interchange. Geology, 44: 375–378. Bagley, J.C. & Johnson, J.B. (2014) Phylogeography and biogeography of the lower Central American Neotropics: diversification between two continents and between two seas. Biological Reviews, 89: 767–790. Bermingham, E. & Martin, A.P. (1998) Comparative mtDNA phylogeography of Neotropical freshwater fishes: testing shared history to infer the evolutionary landscape of lower Central America. Molecular Ecology, 7: 499–517. Bickford, D., Lohman, D.J., Sodhi, N.S., Ng, P.K., Meier, R., Winker, K., Ingram, K.K. & Das, I. (2007) Cryptic species as a window on diversity and conservation. Trends in Ecology and Evolution, 22: 148–155. Boza-Oviedo, E., Rovito, S.M., Chaves, G., García-Rodríguez, A., Artavia, L.G., Bolaños, F. & Wake, D.B. (2012) Salamanders from the eastern Cordillera de Talamanca, Costa Rica, with descriptions of five new species (Plethodontidae: Bolitoglossa, Nototriton, and Oedipina) and natural history notes from recent expeditions. Zootaxa, 3309: 36–61. Brame, A.H. & Duellman, W.E. (1970) A new salamander (genus Oedipina) of the uniformis group from western Panama. Contributions in Science. Natural History Museum of Los Angeles County, 201: 1–8. Bundschuh, J. & Alvarado, G.E. (2007) Central America: Geology, Resources and Hazards. CRC Press, London. Bussing, W.A. (1976) Geographic distribution of the San Juan ichthyofauna of Central America with remarks on its origin and ecology. In: Thorson, E.B. (Ed.), Investigations of the Ichthyo-fauna of Nicaraguan Lakes. Universty of Nebraska-Lincoln, pp 157–175. Bussing, W.A. (1985) Patterns of distribution of the Central American ichthyofauna. In: Stehli, F.G. & Webb, S.D. (Eds.), The Great American Biotic Interchange, Springer, New York, pp. 453–473. 237 35 Campbell, J.A. (1999) Distribution patterns of amphibians in Middle America. In: Duellman, W.E. (Ed.), Patterns of Distribution of Amphibians: A Global Perspective. Johns Hopkins University Press, Maryland, pp. 111–210. Castoe, T.A., Daza, J.M., Smith, E.N., Sasa, M.M., Kuch, U., Campbell, J.A., Chippindale, P.T. & Parkinson, C.L. (2009) Comparative phylogeography of pitvipers suggests a consensus of ancient Middle American highland biogeography. Journal of Biogeography, 36: 88–103. Chakrabarty, P. (2006) Systematics and historical biogeography of Greater Antillean Cichlidae. Molecular Phylogenetics and Evolution, 39: 619–627. Coates, A.G. & Obando, J.A. (1996) The geologic evolution of the Central American Isthmus. In: Jackson, J.B.C., Budd, A.F. & Coates, A.G. (Eds.), Evolution and environment in tropical America. The University of Chicago Press, Illinois, pp. 21–56. Coates, A.G., Collins, L.S., Aubry, M.P. & Berggren, W.A. (2004) The geology of the Darien, Panama, and the late Miocene-Pliocene collision of the Panama arc with northwestern South America. Geological Society of America Bulletin, 116: 1327–1344. Coates, A.G., McNeill, D.F., Aubry, M.P., Berggren, W.A. & Collins, L.S. (2005) An introduction to the geology of the Bocas del Toro Archipelago, Panama. Caribbean Journal of Science, 41: 374–391. Crawford, A.J. (2003) Huge populations and old species of Costa Rican and Panamanian dirt frogs inferred from mitochondrial and nuclear gene sequences. Molecular Ecology, 12: 2525–2540. Crawford, A.J. & Smith, E.N. (2005) Cenozoic biogeography and evolution in direct-developing frogs of Central America (Leptodactylidae: Eleutherodactylus) as inferred from a phylogenetic analysis of nuclear and mitochondrial genes. Molecular Phylogenetics and Evolution, 35: 536–555. Crawford, A.J., Bermingham, E. & Carolina, P.S. (2007) The role of tropical dry forest as a long‐term barrier to dispersal: a comparative phylogeographical analysis of dry forest tolerant and intolerant frogs. Molecular Ecology, 16: 4789–4807. Crawford, A.J., Cruz, C., Griffith, E., Ross, H., Ibáñez, R., Lips, K R., Driskell, A.C. Bermingham, E. & Crump, P. (2013) DNA barcoding applied to ex situ tropical amphibian conservation programme reveals cryptic diversity in captive populations. Molecular Ecology Resources, 13: 1005–1018. 238 36 Crisci, J., Katinas, L., Posadas, P. & Crisci, J.V. (2003) Historical biogeography: an introduction. Harvard University Press, Massachusetts. Darda, D.M. (1994) Allozyme variation and morphological evolution among Mexican salamanders of the genus Chiropterotriton (Caudata: Plethodontidae). Herpetologica, 50: 164–187. Dayrat, B. (2005) Towards integrative taxonomy. Biological Journal of the Linnean Society, 85: 407–415. Duellman, W.E. (1979) The South American herpetofauna: Its origen, evolution and dispersal. The University of Kansas, Kansas. Duellman, W.E. (2001) The hylid frogs of Middle America. Society for the Study of Amphibians and Reptiles, New York. Duellman, W.E., Marion, A.B. & Hedges, S.B. (2016) Phylogenetics, classification, and biogeography of the treefrogs (Amphibia: Anura: Arboranae). Zootaxa, 4104: 1–109. Faivovich, J., Haddad, C.F., Garcia, P.C., Frost, D.R., Campbell, J.A. & Wheeler, W.C. (2005) Systematic review of the frog family Hylidae, with special reference to Hylinae: phylogenetic analysis and taxonomic revision. Bulletin of the American Museum of Natural History, 294: 1–240. Faivovich, J., Haddad, C.F., Baêta, D., Jungfer, K.H., Álvares, G.F.R., Brandao, R.A., Sheil, C., Barrientos, L.S., Barrio-Amorós, C.L., Cruz, C.A.G. & Wheeler, W.C. (2010). The phylogenetic relationships of the charismatic poster frogs, Phyllomedusinae (Anura, Hylidae). Cladistics, 26: 227–261. Flores-Villela, O. & Goyenechea, I. (2001) A comparision of hypotheses of historical area relationships for México and Central America, or in search for the lost pattern. In: Johnson, J.D., Webb, R.G. & Flores-Villela, O. (Eds.), Mesoamerican Herpetology: Systematics, Natural History and Conservation. University of Texas at El Paso, Texas, pp. 171–181. Flot, J.F. (2015) Species Delimitation's Coming of Age. Systematic Biology, 64: 897–899. Frost, D.R. (2019) Amphibian Species of the World: an Online Reference. Disponible en: http://research.amnh.org/herpetology/amphibia/index.html. (accesado 10 de enero de 2019). Frost, D.R., Grant, T., Faivovich, J., Bain, R.H., Haas, A., Haadad, C.F.B., de Sá, R.O., Channing, A., Wilkinson, M., Donnellan, S.C., Raxworthy, C.J., Campbell, J.C., Blotto, 239 37 B.L., Moler, P., Drewes, R.C., Nussbaum, R.A., Lynch, J.D., Green, D.M. & Wheeler W.C. (2006) The amphibian tree of life. Bulletin of the American Museum of Natural History, 297: 1–291. García-París, M., Good, D.A., Parra-Olea, G. & Wake, D.B. (2000) Biodiversity of Costa Rican salamanders: implications of high levels of genetic differentiation and phylogeographic structure for species formation. Proceedings of the National Academy of Sciences, 97: 1640–1647. García-París, M. & Wake, D.B. (2000) Molecular phylogenetic analysis of relationships of the tropical salamander genera Oedipina and Nototriton, with descriptions of a new genus and three new species. Copeia, 2000: 42–70. González, C., Urrego, L.E. & Martínez, J.I. (2006) Late quaternary vegetation and climate change in the Panama basin: palynological evidence from marine cores ODP 677 and TR 163-38. Palaeogeography, Palaeoclimatology, Palaeoecology, 234: 62–80. Gutiérrez-García, T.A. & Vázquez-Domínguez, E. (2011) Comparative phylogeography: designing studies while surviving the process. BioScience, 61: 857–868. Gutiérrez-García, T.A. & Vázquez-Domínguez, E. (2013) Consensus between genes and stones in the biogeographic and evolutionary history of Central America. Quaternary Research, 79: 311–324. Graham, A. & Dilcher, D. (1995) The Cenozoic record of tropical dry forest in northern Latin America and the southern United States. In: Bullock, S.H., Mooney, H.A. & Medina, E. (Eds.), Seasonally dry tropical forests, Cambridge University Press, Cambridge, pp. 124- 145. Haq, B.U., Hardenbol, J. & Vail, P.R. (1987) Chronology of fluctuating sea levels since the Triassic. Science, 235: 1156–1167. Hedges, S.B. (2006) Paleogeography of the Antilles and origin of West Indian terrestrial vertebrates. Annals of the Missouri Botanical Garden, 93: 231–244. Hedges, S.B., Duellman, W.E. & Heinicke, M.P. (2008) New World direct-developing frogs (Anura: Terrarana): Molecular phylogeny, classification, biogeography, and conservation. Zootaxa, 1737: 1–182. 240 38 Heinicke, M.P., Duellman, W.E. & Hedges, S.B. (2007) Major Caribbean and Central American frog faunas originated by ancient oceanic dispersal. Proceedings of the National Academy of Sciences, 104: 10092–10097. Iturralde-Vinent, M.A. & MacPhee, R.D. (1999) Paleogeography of the Caribbean region: implications for Cenozoic biogeography. Bulletin of the American Museum of Natural History, 238: 1–95. James, K.H. (2005) A simple synthesis of Caribbean geology. Caribbean Journal of Earth Science, 39: 69–82. Johnson, J.D., Webb, R.G. & Flores-Villela, O. (2001) Mesoamerican Herpetology: Systematics, Natural History and Conservation. University of Texas at El Paso, Texas. Kapelle, M. & Horn, S.P. (2005) Páramos de Costa Rica. Editorial INBio, Costa Rica. Keppie, J.D. & Morán-Zenteno, D.J.( 2005) Tectonic implications of alternative Cenozoic reconstructions for southern Mexico and the Chortis block. International Geology Review, 47: 473–491. Kirby, M.X. & MacFadden, B. (2005) Was southern Central America an archipelago or a peninsula in the middle Miocene? A test using land-mammal body size. Palaeogeography, Palaeoclimatology, Palaeoecology, 228: 193–202. Kominz, M.A. (1984) Oceanic ridge volumes and sea level change – an error analysis. In: Schlee, J.S. (Ed.), Interregional Unconformities and Hydrocarbon Accumulation, American Association of Petroleum Geologists Memoir 36, pp. 109–127. Kominz, M.A., Browning, J.V., Miller, K.G., Sugarman, P.J., Mizintseva, S. & Scotese, C.R. (2008) Late Cretaceous to Miocene sea-level estimates from the New Jersey and Delaware coastal plain coreholes: an error analysis. Basin Research, 20: 211–226. Lachniet, M.S. & Vazquez-Selem, L. (2005). Last glacial maximum equilibrium line altitudes in the circum-Caribbean (Mexico, Guatemala, Costa Rica, Colombia, and Venezuela). Quaternary International, 138: 129–144. MacMillan, I., Gans, P.B. & Alvarado, G. (2004) Middle Miocene to present plate tectonic history of the southern Central American Volcanic Arc. Tectonophysics, 392: 325–348. Marshall, L.G., Webb, S.D., Sepkoski, J.J. & Raup, D.M. (1982) Mammalian evolution and the great American interchange. Science, 215: 1351–1357. 241 39 Marshall, L.G. (1988) Land mammals and the great American interchange. American Scientist, 76: 380–388. Martens, U., Weber, B. & Valencia, V.A. (2010) U/Pb geochronology of Devonian and older Paleozoic beds in the southeastern Maya block, Central America: Its affinity with peri- Gondwanan terranes. Geological Society of America Bulletin, 122: 815–829. Mendelson III, J.R.M., Mulcahy, D.G., Williams, T.S. & Sites Jr, J.W. (2011) A phylogeny and evolutionary natural history of mesoamerican toads (Anura: Bufonidae: Incilius) based on morphology, life history, and molecular data. Zootaxa, 3138: 1–34. Miller, K.G., Kominz, M.A., Browning, J.V., Wright, J.D., Mountain, G.S., Katz, M.E., Sugarman, P.J., Cramer, B.S., Christie-Blick, N. & Pekar, S.F. (2005) The phanerozoic record of global sea-level change. Science, 310: 1293–1298. Montes, C., Bayona, G., Cardona, A., Buchs, D.M., Silva, C A., Morón, S., Hoyos, N., Ramírez, D.A, Jaramillo, C.A. & Valencia, V. (2012)a Arc‐continent collision and orocline formation: Closing of the Central American seaway. Journal of Geophysical Research: Solid Earth, 117: B04105. Montes, C., Cardona, A., McFadden, R., Morón, S.E., Silva, C.A., Restrepo-Moreno, S., Ramírez, D.A., Hoyos, N., Wilson, J., Farris, D., Bayona, G.A., Jaramillos, C.A., Valencia, V., Bryan. J. & Flores, J.A. (2012)b Evidence for middle Eocene and younger land emergence in central Panama: implications for Isthmus closure. Geological Society of America Bulletin, 124: 780–799. Montes, C., Cardona, A., Jaramillo, C., Pardo, A., Silva, J. C., Valencia, V., Ayala, C., Pérez- Angel, L.C., Rodríguez-Parra, L.A., Ramírez, V. & Niño, H. (2015). Middle miocene closure of the Central American seaway. Science, 348: 226–229. Morrone, J.J. (1994) On the identification of areas of endemism. Systematic Biology, 43: 438– 441. Morrone, J.J. (2005) Cladistic biogeography: identity and place. Journal of Biogeography, 32: 1281–1284. Nadler, S.A. & Pérez-Ponce de León, G. (2011) Integrating molecular and morphological approaches for characterizing parasite cryptic species: implications for parasitology. Parasitology, 138: 1688–1709. 242 40 Ortega-Andrade, H.M., Rojas-Soto, O.R., Valencia, J.H., de los Monteros, A.E., Morrone, J.J., Ron, S.R. & Cannatella, D.C. (2015) Insights from integrative systematics reveal cryptic diversity in Pristimantis frogs (Anura: Craugastoridae) from the upper Amazon basin. PloS One, 10: e0143392. Padial, J.M. & De la Riva, I. (2009) Integrative taxonomy reveals cryptic Amazonian species of Pristimantis (Anura: Strabomantidae). Zoological Journal of the Linnean Society, 155: 97– 122. Padial, J.M., Miralles, A., De la Riva, I. & Vences, M. (2010) The integrative future of taxonomy. Frontiers in zoology, 7: 16. Padial, J.M., Grant, T. & Frost, D.R. (2014) Molecular systematics of terraranas (Anura: Brachycephaloidea) with an assessment of the effects of alignment and optimality criteria. Zootaxa, 3825: 1–132. Pante, E., Schoelinck, C. & Puillandre, N. (2014) From integrative taxonomy to species description: one step beyond. Systematic Biology, 64: 152–160. Parra-Olea, G., García-París, M. & Wake, D.B. (2004) Molecular diversification of salamanders of the tropical American genus Bolitoglossa (Caudata: Plethodontidae) and its evolutionary and biogeographical implications. Biological Journal of the Linnean Society, 81: 325–346. Paz, A., Ibáñez, R., Lips, K.R. & Crawford, A.J. (2015) Testing the role of ecology and life history in structuring genetic variation across a landscape: a trait‐based phylogeographic approach. Molecular Ecology, 24: 3723–3737. Perdices, A., Bermingham, E., Montilla, A. & Doadrio, I. (2002) Evolutionary history of the genus Rhamdia (Teleostei: Pimelodidae) in Central America. Molecular Phylogenetics and Evolution, 25: 172–189. Pindell, J.L. & Barrett, S.F. (1990) Geological evolution of the Caribbean region; A plate- tectonic perspective. In: Dengo, G. & Case, J.E. (Eds.), The Caribbean Region. Geological Society of America, pp. 405-432. Pinto-Sánchez, N.R., Ibáñez, R., Madriñán, S., Sanjur, O.I., Bermingham, E. & Crawford, A.J. (2012) The great American biotic interchange in frogs: multiple and early colonization of Central America by the South American genus Pristimantis (Anura: Craugastoridae). Molecular Phylogenetics and Evolution, 62: 954–972. 243 41 Pérez-Ponce de León, G. & Nadler, S.A. (2010) What we don't recognize can hurt us: a plea for awareness about cryptic species. Journal of Parasitology, 96: 453–464. Pérez-Ponce de León, G. & Poulin, R. (2016) Taxonomic distribution of cryptic diversity among metazoans: not so homogeneous after all. Biology Letters, 12: 20160371. Pyron, R.A. (2011) Divergence time estimation using fossils as terminal taxa and the origins of Lissamphibia. Systematic Biology, 60: 466–481. Pyron, R.A. & Wiens, J.J. (2011) A large-scale phylogeny of Amphibia including over 2800 species, and a revised classification of extant frogs, salamanders, and caecilians. Molecular Phylogenetics and Evolution, 61: 543–583. Raven, P.H. & Axelrod, D.I. (1974) Angiosperm biogeography and past continental movements. Annals of the Missouri Botanical Garden, 61: 539–673. Ree, R.H., Moore, B.R., Webb, C.O. & Donoghue, M.J. (2005) A likelihood framework for inferring the evolution of geographic range on phylogenetic trees. Evolution, 59: 2299– 2311. Ree, R.H. & Smith, S.A. (2008) Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Systematic Biology, 57: 4–14. Ree, R.H. & Sanmartín, I. (2009) Prospects and challenges for parametric models in historical biogeographical inference. Journal of Biogeography, 36: 1211–1220. Říčan, O., Piálek, L., Zardoya, R., Doadrio, I. & Zrzavý, J. (2013) Biogeography of the Mesoamerican Cichlidae (Teleostei: Heroini): colonization through the GAARlandia land bridge and early diversification. Journal of Biogeography, 40: 579–593. Rogers, R.D., Mann, P. & Emmet, P.A. (2007) Tectonic terranes of the Chortis block based on integration of regional aeromagnetic and geologic data. The Geological Society of America Special Paper, 428: 65–88. Rovito, S.M., Wake, D.B., Papenfuss, T.J., Parra‐Olea, G., Muñoz‐Alonso, A. & Vásquez‐ Almazán, C.R. (2012). Species formation and geographical range evolution in a genus of Central American cloud forest salamanders (Dendrotriton). Journal of Biogeography, 39: 1251–1265. Rovito, S.M., Parra‐Olea, G., Hanken, J., Bonett, R.M. & Wake, D.B. (2013) Adaptive radiation in miniature: the minute salamanders of the Mexican highlands (Amphibia: Plethodontidae: Thorius). Biological Journal of the Linnean Society, 109: 622–643. 244 42 Rovito, S.M., Parra‐Olea, G., Recuero, E. & Wake, D.B. (2015)a Diversification and biogeographical history of Neotropical plethodontid salamanders. Zoological Journal of the Linnean Society, 175: 167–188. Rovito, S.M., Vásquez‐Almazán, C.R., Papenfuss, T.J., Parra‐Olea, G. & Wake, D.B. (2015)b Biogeography and evolution of Central American cloud forest salamanders (Caudata: Plethodontidae: Cryptotriton), with the description of a new species. Zoological Journal of the Linnean Society, 175: 150–166. San Mauro, D. (2010) A multilocus timescale for the origin of extant amphibians. Molecular Phylogenetics and Evolution, 56: 554–561. Savage, J.M. (1966) The origins and history of the Central American herpetofauna. Copeia, 1966: 719–766. Savage, J.M. (1982) The enigma of the Central American herpetofauna: dispersals or vicariance?. Annals of the Missouri Botanical Garden, 69: 464–547. Savage, J.M. (2002) The Amphibians and Reptiles of Costa Rica: a Herpetofauna Between Two Continents, Between Two Seas. The University of Chicago Press, Illinois. Shen, X.X., Liang, D., Chen, M.Y., Mao, R.L., Wake, D.B. & Zhang, P. (2016) Enlarged Multilocus Data set Provides Surprisingly Younger Time of Origin for the Plethodontidae, the Largest Family of Salamanders. Systematic Biology, 65: 66–81. Silva-Romo, G. (2009) Guayape-Papalutla fault system: a continuous Cretaceous structure from southern Mexico to the Chortís block? Tectonic implications. Geology, 36: 75–78. Simpson, G.G. (1940) Mammals and land bridges. Journal of the Washington Academy of Sciences, 30: 137–163. Sites, J.W. & Marshall, J.C. (2003). Delimiting species: a Renaissance issue in systematic biology. Trends in Ecology & Evolution, 18: 462–470. Streicher, J.W., Crawford, A.J. & Edwards, C.W. (2009) Multilocus molecular phylogenetic analysis of the montane Craugastor podiciferus species complex (Anura: Craugastoridae) in Isthmian Central America. Molecular Phylogenetics and Evolution, 53: 620–630. Townsend, J.H., Medina-Flores, M., Murillo, J.L. & Austin, J.D. (2011) Cryptic diversity in Chortís Highland moss salamanders (Caudata: Plethodontidae: Nototriton) revealed using mtDNA barcodes and phylogenetics, with a new species from eastern Honduras. Systematics and Biodiversity, 9: 275–287. 245 43 Vilela, B. & Villalobos, F. (2015) letsR: a new R package for data handling and analysis in macroecology. Methods in Ecology and Evolution, 6: 1229–1234. Villa, J., Wilson, L.D. & Johnson, J.D. (1988) Middle American Herpetology. A bibliographic checklist. University of Missouri Press, Columbia. Wang, I.J., Crawford, A.J. & Bermingham, E. (2008) Phylogeography of the pygmy rain frog (Pristimantis ridens) across the lowland wet forest of isthmian Central America. Molecular Phylogenetics and Evolution, 47: 992–1004. Wake, D.B., Savage, J.M. & Hanken, J. (2007) Montane salamanders from the Costa Rica- Panama border region, with descriptions of two new species of Bolitoglossa. Copeia, 2007: 556–565. Webb, S.D. (1991) Ecogeography and the Great American Interchange. Paleobiology, 17: 266– 280. Webb, S.D. (2006) The Great American Biotic Interchange: patterns and processes. Annals of the Missouri Botanical Garden, 93: 245–257. Weber, B., Iriondo, A., Premo, W.R., Hecht, L. & Schaaf, P. (2007) New insights into the history and origin of the southern Maya block, SE México: U–Pb–SHRIMP zircon geochronology from metamorphic rocks of the Chiapas massif. International Journal of Earth Sciences, 96: 253–269. Wiens, J.J. (2007) Species delimitation: new approaches for discovering diversity. Systematic Biology, 56: 875–878. Wiens, J.J., Parra-Olea, G., García-París, M. & Wake, D.B. (2007) Phylogenetic history underlies elevational biodiversity patterns in tropical salamanders. Proceedings of the Royal Society Biological Sciences Series B, 274: 919–928. Wiley, E.O. (1988) Vicariance biogeography. Annual Review of Ecology and Systematics, 19: 513–542. Will, K.W., Mishler, B.D. & Wheeler, Q.D. (2005) The perils of DNA barcoding and the need for integrative taxonomy. Systematic Biology, 54: 844–851. Wilson, L.D., Townsend, J.H. & Johnson, J.D. (2010) Conservation of Mesoamerican Amphibians and Reptiles. Eagle Mountain Publishing, Utah. Winker, K. (2011) Middle America, not Mesoamerica, is the accurate term for biogreography. The Condor, 113: 5–6. 246 44 Zeh, J.A., Zeh, W. & Bonilla, M.M. (2003) Phylogeography of the harlequin beetle-riding pseudoscorpion and the rise of the Isthmus of Panama. Molecular Ecology, 12: 2759–2769. 247