on the biogeography of vertebrate body size: ecological

On the biogeography of vertebrate body size: ecological and evolutionary insights from

assemblage-level patterns

Miguel Ángel Olalla Tárraga

Memoria presentada para optar al grado de Doctor por la Universidad de Alcalá

Alcalá de Henares, Diciembre de 2008

Departamento de Ecología

Miguel Ángel Rodríguez Fernández, Profesor Titular de Ecología de la Universidad de Alcalá de Henares y Director del Departamento de Ecología

HACE CONSTAR: Que el trabajo descrito en la presente memoria, titulado: “On the biogeography of vertebrate body size: ecological and evolutionary insights from assemblage-level patterns”, ha sido realizado bajo su dirección por D. Miguel Ángel Olalla Tárraga en el Departamento de Ecología de la Universidad de Alcalá y reúne todos los requisitos necesarios para su defensa y aprobación como Tesis Doctoral.

Alcalá de Henares, a 3 de diciembre de 2008

Fdo.: Dr. Miguel Ángel Rodríguez Fernández

Departamento de Ecología

Miguel Ángel Rodríguez Fernández, Profesor Titular de Ecología de la Universidad de Alcalá de Henares y Director del Departamento de Ecología

HACE CONSTAR: Que el trabajo descrito en la presente memoria, titulado: “On the biogeography of vertebrate body size: ecological and evolutionary insights from assemblage-level patterns”, ha sido realizado por D. Miguel Ángel Olalla Tárraga dentro del Programa de Doctorado Cambio Global y Desarrollo Sostenible adscrito al Departamento de Ecología de la Universidad de Alcalá, y reúne todos los requisitos necesarios para su defensa y aprobación como Tesis Doctoral.

Alcalá de Henares, a 3 de diciembre de 2008

Fdo.: Dr. Miguel Ángel Rodríguez Fernández

AGRADECIMIENTOS

Cuenta una leyenda budista milenaria que en algún lugar inaccesible, mas allá del Tíbet, se encuentra la mítica ciudad de Shamballah (o Shangri-la). Se trataría de un paraíso perdido entre los valles y las altas montañas del Himalaya en el que habitan hombres inmortales que han sido capaces de vivir en armonía con la naturaleza. Un remanso de sabiduría universal y paz desde el que se guiarían los pasos de la humanidad. Las nieves perennes, las escarpadas pendientes y las dificultades y penurias del camino habrían contribuido a mantener oculto para la mayoría de los mortales este idílico lugar. No obstante, dichas limitaciones no han sido impedimento suficiente para que un reducido grupo de soñadores, aventureros y exploradores se hayan lanzado a la búsqueda de Shangri-la a lo largo de los años.

Quizás de forma inconsciente, hace ya muchos años que mis padres iniciaron la búsqueda de esa ciudad de conocimientos descrita por la tradición tibetana. Mi madre quería dedicarse a la investigación. En definitiva, encontrar respuestas a tantas preguntas. Y decidió emprender la aventura. Acompañada por mi padre, comenzó su peregrinaje por los senderos de la curiosidad científica. Fueron innumerables jornadas de adversidades luchando, entre otros, contra la estulticia y la envidia con las únicas armas de la inquietud por conocer, la disciplina y el trabajo concienzudo. Todo el esfuerzo, sin embargo, mereció la pena cuando un día encontraron en medio del camino lo que buscaban. Aunque no disponían de un mapa detallado del terreno para mostrarme cómo llegar de manera rápida a ese lugar, me han sabido transmitir con su ejemplo, cariño y dedicación lo que supone viajar a Shangri-la. Papá, mamá, os estaré eternamente agradecido por todas vuestras enseñanzas, cuidados y apoyo. Muchas gracias por haber motivado mi vocación por la investigación y por enseñarme que se puede ser feliz dedicando la vida a la prosecución de un sueño.

A falta de un mapa, una vez decidido a emprender el viaje hacia Shangri-la, era preciso dar con un buen conocedor del terreno para enfrentarme con éxito a las cumbres del Himalaya. Alguien me dijo que los maestros tibetanos son excelentes guías y yo tuve suerte y encontré uno de los mejores. Miguel me ha sabido orientar acertadamente con sus sabios consejos. Para evitar los traspiés, ha dirigido mis pasos sin mostrarme siempre el camino más corto. Por el contrario, ha seleccionado la travesía apropiada para enseñarme a superar los contratiempos, reservando siempre la energía y la ilusión suficiente como para emprender nuevas expediciones en el futuro. Miguel, muchas gracias por tu dedicación, por tus enseñanzas, por tu tiempo, por tu experiencia, por tu facilidad para transmitirme ilusión por la ciencia, por tu buen hacer para que esta aventura haya llegado a buen término. Ha sido un verdadero lujo tenerte como guía en mis primeros pasos por la investigación de la geografía de la naturaleza.

Durante todo este tiempo, me ha acompañado una persona especial a lo largo del camino. A mi lado siempre, o casi siempre, de la manera más desinteresada. Irene, porque no creía en las princesas hasta que me encontré con tu sonrisa. Gracias por soportarme, escucharme, sonreírme, abrazarme, quererme y estar conmigo cuando lo he necesitado. Tu compañía, en la cercanía o en la distancia, ha sido un apoyo necesario para poder avanzar con éxito en esta difícil empresa. Por eso, por todo, este trabajo también está dedicado a ti.

En la travesía también he sido muy afortunado de encontrarme con otras expediciones internacionales con muchísima experiencia. De ellas he aprendido gran parte de lo que sé. Brad, muchas gracias por tu excelencia en el trabajo, por tu disponibilidad, por tu hospitalidad y por mostrarme cómo se diferencian las preguntas interesantes de las que no lo son. Tomar una cerveza contigo es empaparse de ciencia y de vida. Espero que podamos disfrutar muchas más en los años venideros. Alexandre y Mau, ¿qué puedo deciros? Las palabras me escasean para describir lo que me ha supuesto compartir algunos meses con vosotros. Mil gracias por vuestra calurosa acogida, por vuestra humildad y brillantez, por vuestro entusiasmo contagioso por la ciencia, por hacerme sentir como en casa a miles de kilómetros de Madrid. Cleiber y Priscila, cuánta alegría haberme encontrado unos amigos extraordinarios con los que charlar, pasear, hacer turismo, fumar un charuto o beber caipirinhas sin que importe el paso del tiempo. Espero que siga siendo así por muchos años. Carina, Teresa, Ludgero & Cía., muchas gracias por vuestra disposición y ayudas, resulta difícil encontrar tanto compañerismo y cariño en estos tiempos. Cerca ya de la cumbre, en el campamento inglés, me he tropezado con un tipo genial, un anfitrión estupendo, Joaquín. Descubrir el bosque de Silwood no hubiera sido la misma experiencia fascinante y enriquecedora sin tu ayuda y consejos. Mi más sincero agradecimiento también para la gente que ha contribuido a colorear de vitalidad los nublados días londinenses (Omar, Guida, Shai, Diego, Marcia, Rich, Ally, Gavin…).

En el campamento base, el ajetreo y el movimiento de gente ha sido incesante. Algunos se disponían a comenzar la ruta, otros regresaban con mayor o menor éxito de su recorrido. Todos ellos, los que iban o venían (Itzi, Carlota, Alberto, Irene, Pedro, Marta, Luis C., Kike, Edu, Fabio, María José, Noelia, Mariano, Oscar, Dani, Luis M., Lucía, Virginia, Jorge, Irene R., Nacho, Anabella, Bárbara, Silvia, Cristina…) o los maestros que les acompañaban (Miguel Angel Z., Josevi, Nico, Pedro, Jose María, Pilar, Salva, Lorenzo, Josa, Antonio, etc…) han posibilitado un ambiente magnífico en el que intercambiar agradablemente las impresiones del viaje. De la logística en la base se ha encargado Ana Guerrero, que es sinónimo de atención, cariño y eficiencia. Estas cualidades también se pueden adjudicar a Rosa que, además, ha aportado una labor atenta y meticulosa con el GIS. Un especial reconocimiento merece también la continua predisposición de Fabio a echar una mano con las dificultades que han surgido en nuestros delicados “equipos de apoyo informático”. En la UAM, Javier Benayas y Eugenio Rico apoyaron mis primeros pasos investigadores. Muchas gracias a todos por animar cada día la ilusión en esta aventura.

Mi familia y amigos han sido compañía imprescindible en este periplo. Tania, me enorgullece ser tu hermano. Qué suerte haber tenido en la habitación de al lado todos estos años tu jovialidad, tus ganas de conocer, tu espíritu aventurero, tu imponente capacidad para el trabajo, tu inconformismo y tu madurez inocente. En realidad, qué afortunado de que puedas acompañarme siempre en la vida. Juani y Carlos, gracias por ser la familia más cercana en todos los sentidos. Tía Virginia, te mereces de veras el nombre de una calle en Cuenca y espero que te la concedan pronto.

Borja, José, Torete, Juan, Jesulín y Chinchi me habéis ayudado a desconectar del estrés y el desánimo cuando ha venido. Gracias por escucharme cuando lo he necesitado y apoyar mi excursión investigadora desde la retaguardia. Mi gente de la UAM ha sido también indispensable: Berta, Lara, José, Nuria, Patri, Helga, Jaime, Chapi, aunque no nos veamos con la frecuencia que nos gustaría, soy muy consciente de que puedo contar con vosotros siempre que lo necesite. Sé que os puedo considerar a todos, como dice la canción, amigos para siempre.

Índice

Resumen .................................................................................................................................................... 1

Abstract ...................................................................................................................................................... 2

Capítulo 1. Introducción General .............................................................................................................. 3

Capítulo 2. Regla de Bergmann y la geografía del tamaño corporal en mamíferos del Hemisferio Oeste Introducción .................................................................................................................................... 9 Material y Métodos ........................................................................................................................ 10 Resultados ...................................................................................................................................... 13 Discusión ........................................................................................................................................ 16 Referencias ..................................................................................................................................... 17 Apéndices ....................................................................................................................................... 19

Capítulo 3. Patrones a gran escala de tamaño corporal en reptiles escamados de Europa y Norte América Introducción ................................................................................................................................... 54 Material y Métodos ........................................................................................................................ 55 Resultados ...................................................................................................................................... 58 Discusión ........................................................................................................................................ 61 Referencias ..................................................................................................................................... 64

Capítulo 4. Energía y patrones de tamaño corporal interespecíficos en faunas de anfibios de Europa y Norte América: anuros siguen la regla de Bergmann, urodelos su inversa

Introducción ................................................................................................................................... 67 Material y Métodos ........................................................................................................................ 68 Resultados ...................................................................................................................................... 70 Discusión ........................................................................................................................................ 73 Referencias ..................................................................................................................................... 76 Apéndices ....................................................................................................................................... 79

Capítulo 5. Gradientes geográficos de tamaño corporal en regiones tropicales: déficit hídrico y tamaño corporal de anuros en el Cerrado Brasileño

Introducción ................................................................................................................................... 92 Material y Métodos ........................................................................................................................ 94 Resultados ...................................................................................................................................... 97 Discusión ........................................................................................................................................ 99 Referencias .................................................................................................................................... 101 Apéndices ...................................................................................................................................... 105

Capítulo 6. Regla de Bergmann en anfibios, regresión filogenética basada en autovectores y la aproximación de ensamblaje

Introducción .................................................................................................................................. 111 Material y Métodos ....................................................................................................................... 113 Resultados ..................................................................................................................................... 113 Discusión ....................................................................................................................................... 115 Referencias .................................................................................................................................... 118

Capítulo 7. Conclusiones generales ......................................................................................................... 120

RESUMEN

Antecedentes y objetivos En 1847, Bergmann sugirió que el tamaño corporal juega un importante papel en la distribución geográfica de mamíferos y aves; especies de mayor tamaño se ven favorecidas en climas fríos debido a su mejor conservación del calor corporal (menor ratio superficie/volumen). Desde entonces, el estudio de gradientes geográficos en la variación de tamaños corporales sigue suscitando el interés de la comunidad científica. Profundizar nuestro conocimiento sobre posibles mecanismos ecológicos y evolutivos requiere el uso de aproximaciones espacialmente explícitas. Esta tesis doctoral emplea un enfoque basado en ensamblaje para documentar patrones biogeográficos de tamaño corporal en faunas regionales (Paleártico occidental, Neártico y Neotrópico) de diferentes grupos de vertebrados (mamíferos, anfibios y reptiles). Asimismo, se evalúa el grado de apoyo de las hipótesis propuestas hasta la fecha para explicar la regla de Bergmann. A fin de mejorar nuestra comprensión de posibles mecanismos energéticos y fisiológicos, se presta especial atención a dos grupos de vertebrados ectotermos: anfibios y reptiles.

Metodología Se emplea un enfoque metodológico innovador basado en la recopilación de información sobre la biología y distribución de los organismos, su procesado mediante Sistemas de Información Geográfica y un posterior análisis multivariante combinado con métodos de estadística espacial para evaluar la estructura de autocorrelación espacial en los datos. Además, se usa un procedimiento de selección de modelos basado en teoría de la información y se incorpora un método de regresión filogenética basada en autovectores (PVR) que estima la señal filogenética en los datos y permite particionar el componente ecológico y filogenético de la variación de tamaños corporales.

Conclusiones En mamíferos, el mecanismo de conservación de calor explica los patrones por debajo de un umbral de temperatura ambiental en torno a 10°C, mientras que en regiones tropicales cobra mayor importancia la disponibilidad de hábitats. La energía ambiental disponible también condiciona la distribución geográfica de tamaños en vertebrados ectotermos. Entre los termoreguladores (lagartos, serpientes y anuros) la existencia de patrones de Bergmann está determinada por el balance ganancia-pérdida de calor y depende, por tanto, del tamaño relativo de los organismos. Los urodelos, con un alto grado de termoconformismo, muestran un patrón inverso de Bergmann. En salamandras Plethodon del Neártico oriental este patrón no sólo depende de las respuestas adaptativas individuales de las especies, sino de los rápidos eventos de especiación alopátrica desde el Plioceno temprano. Bajo condiciones de elevado estrés hídrico en regiones tropicales existe una ventaja selectiva para anuros de mayor tamaño, cuya reducida ratio superficie-volumen les permite disminuir la tasa de desecación. Considerando el gradiente latitudinal en la importancia relativa de la disponibilidad de energía y agua desde regiones templadas hacia los trópicos, nuestros hallazgos sugieren que los patrones biogeográficos de tamaño en anuros reflejan los efectos de la reducción de la ratio superficie-volumen en el control del balance corporal térmico e hídrico.

Palabras clave Anfibios, macroecología, mamíferos, regla de Bergmann, reptiles, termorregulación.

1

ABSTRACT

Background and aim In 1847, Bergmann first suggested that body size plays a major role in determining the geographic distribution of mammals and birds; large-bodied species are favored in colder climates due to their better heat conservation (lower surface-to-volume ratio). Since then, there has been a longstanding scientific interest in studying geographic gradients in body size. Gaining a better understanding of possible ecological and evolutionary mechanisms involves using spatially explicit approaches. This PhD thesis applies an assemblage-based approach to document biogeographic patterns of body size in regional faunas (Western Palearctic, Neartic and Neotropics) for different vertebrate groups (mammals, amphibians and reptiles). Also, we evaluate the support for several hypotheses proposed to explain Bergmann’s rule so far. Aimed at generating insights into the energetic and physiological mechanisms, we specially focus the analyses in two ectothermic vertebrate groups: amphibians and reptiles.

Methodology We employ a novel methodological approach based on compiling biological and distributional information of species, processing the data by means of Geographic Information Systems and subsequently using multivariate analysis techniques complemented with spatial statistics methods to evaluate the spatial autocorrelation structure. Moreover, we use a model selection procedure based on information theory and incorporate a Phylogenetic eigenVector Regression (PVR) method to estimate the phylogenetic signal in the data and partition the ecological and phylogenetic components of the interspecific variation in body size.

Conclusions In mammals, the heat conservation mechanism is found to be an explanation solely for body size patterns below an environmental temperature threshold around 10°C, whereas in tropical regions habitat availability becomes important. Energy available in the environment also plays a leading role in determining the geographic distribution of body size in ectothermic vertebrates. Among thermoregulators (lizards, snakes and anurans), the occurrence of Bergmannian patterns is conditioned by a trade-off between heat gain and heat conservation and, hence, depends on relative organismal size. Urodeles, which are basically thermoconformers, show converse Bergmann’s rule gradients. Our analysis of eastern Nearctic Plethodon salamanders found that these patterns do not only depend on independent adaptive responses of species, but also on rapid allopatric speciation events since the early Pliocene. Under high water deficit conditions in tropical regions there is a selective advantage for larger anurans, whose reduced surface-to-volume ratio allows them to decrease the dessication rate. In a context of latitudinal variation in the relative importance of energy and water from temperate to tropical regions, our findings suggest that anuran body size gradients reflect effects of reduced surface-to-volume ratios in larger species to control both heat and water balance.

Keywords Amphibians, Bergmann’s rule, macroecology, mammals, reptiles, thermoregulation.

2

Capítulo 1 Introducción General

INTRODUCCIÓN

El libro Geographical Ecology del eminente ecólogo Robert H. MacArthur vio la luz en 1972, el mismo año en el que se produjo la trágica muerte por cáncer de su autor, y se ha convertido desde entonces en una de las publicaciones más influyentes en el ámbito de la investigación ecológica. Este trabajo aporta una visión extremadamente novedosa al estudio de los patrones de diversidad y distribución de especies que lo convierte de manera indiscutible en el germen de la disciplina macroecológica durante los años noventa (Brown 1999). MacArthur enfatizó especialmente la necesidad de usar un enfoque a gran escala para abordar el estudio de patrones biogeográficos y alcanzar, a su vez, una mejor comprensión de los mecanismos subyacentes. Entre las posibles causas de los patrones destacó la importancia que juegan la estructura del ambiente, la morfología y el comportamiento de las especies, las dinámicas poblacionales y la historia. Quizás la originalidad del trabajo de MacArthur consistió en saber combinar de forma magistral sus conocimientos matemáticos y de historia natural para desarrollar una teoría ecológica explicada a través de simples modelos gráficos y matemáticos. De esta manera, contribuyó a unificar y dotar de cierta robustez teórica algunas de las observaciones de campo de muchos naturalistas que históricamente habían sugerido la intrigante existencia de varios patrones en la geografía de la naturaleza.

Desde los tiempos de Alexander Von Humboldt, a comienzos del siglo XIX, ha existido un interés histórico en las ciencias naturales por describir tendencias generales en la variación de rasgos biológicos a lo largo de escalas espaciales y temporales amplias. Así, biogeógrafos, biólogos evolutivos y ecólogos han establecido durante los últimos doscientos años un reducido número de

reglas ecogeográficas y evolutivas para describir ciertas regularidades en la variación de atributos biológicos a escalas geográficas o temporales amplias. Esto incluye el gradiente latitudinal de riqueza de especies observado por el propio Von Humboldt en 1807 (Hawkins, 2001), junto con los patrones (“reglas”) descritos por Gloger (1833), Bergmann (1847), Allen (1878), Cope (1887), Jordan (1892), Foster (1964) o Rapoport (1982) (véase Lomolino et al. 2006 o Gaston et al. 2008 para más detalles). No obstante, muchas de estas reglas fueron originalmente propuestas sin la suficiente evidencia empírica y han sido reiteradamente puestas en entredicho (McNab 1971, Gould 1997, Gaston et al. 1998). Las excepciones encontradas en investigaciones recientes han sugerido que la mayor parte de estas reglas son inválidas (véanse por ejemplo referencias en Ashton 2001). Otros autores, aún reconociendo que ninguno de estos patrones merecen realmente el estatus de “leyes” invariantes de la naturaleza, han subrayado el potencial de dichos estudios para identificar factores condicionantes de la evolución y diversificación en biotas regionales (Lomolino et al. 2006). Es decir, la propia existencia de patrones taxonómicamente y geográficamente contradictorios es altamente informativa sobre las causas potenciales.

En este contexto histórico, el estudio de la regla de Bergmann se plantea como un caso paradigmático. En el año 1847, el fisiólogo alemán Karl Bergmann planteó por primera vez la posibilidad de que el tamaño corporal de mamíferos y aves (organismos endotermos capaces de mantener temperaturas corporales constantes a través de la producción de energía metabólica) fuese un factor determinante en la distribución geográfica de las especies. Bergmann sugirió que, al menos entre especies de endotermos cercanamente emparentadas, las de

3

mayor tamaño serían más frecuentes en climas fríos, mientras que aquellas de menor tamaño abundarían en climas cálidos (véase traducción en James 1970). De acuerdo con su razonamiento, el mecanismo causante de este patrón geográfico se vincularía a la mejor conservación del calor corporal en aquellas especies de mayor tamaño, debido a su menor ratio superficie/volumen. Desde entonces, numerosos estudios han explorado las predicciones de la regla de Bergmann lo que la ha convertido en el patrón ecogeográfico más conocido tras el gradiente latitudinal en riqueza de especies. Al igual que sucede con otras reglas ecológicas y evolutivas, tanto el patrón como el mecanismo propuestos por Bergmann no han sido ajenos a las críticas (Scholander 1955, McNab 1971, Geist 1987). Sin embargo, tras más de 160 años, el estudio de gradientes geográficos en la variación de tamaños corporales sigue suscitando interés entre la comunidad científica (Ashton et al. 2000, Ashton 2002, Meiri & Dayan 2003, Blackburn & Hawkins 2004, Rodríguez et al. 2006, Meiri et al. 2007). Esta característica la diferencia claramente de “reglas” como la de Gloger, Allen o Jordan de interés meramente histórico y casi anecdótico.

El tamaño corporal es un rasgo de los organismos con profundas implicaciones fisiológicas, ecológicas y evolutivas (Peters 1983). Por tanto, examinar patrones biogeográficos en la variación de los tamaños corporales contribuye a mejorar nuestra comprensión de los mecanismos ecológicos y evolutivos que estructuran espacialmente los ensamblajes de especies (Lawton 1990, Brown 1995). En general, el estudio de la variación espacial del tamaño corporal a lo largo de gradientes ambientales se ha desarrollado en torno al objetivo de evaluar el grado de generalidad de la regla de Bergmann. A pesar de que originariamente el patrón fue concebido a nivel interespecífico (Blackburn et al.

1999), el enfoque tradicional ha sido esencialmente intraespecífico como resultado de la reformulación de la hipótesis original por parte de Rensch (1938) y Mayr (1942, 1956). En consecuencia, la versión interespecífica de la regla de Bergmann ha recibido escasa atención hasta el comienzo del programa de investigación macroecológica en años recientes (Brown & Maurer 1989, Brown 1995). Más allá de la información sobre la autoecología de las especies que nos proporcionan los estudios intraespecíficos, la búsqueda de patrones interespecíficos de tamaño corporal ha permitido profundizar en nuestro conocimiento sobre la organización de las comunidades animales (Zeveloff & Boyce 1988, Cotgreave & Stockley 1994, Blackburn & Gaston 1996, Blackburn & Hawkins 2004, Rodríguez et al. 2006, Medina et al. 2007).

En la actualidad, existen tres aproximaciones diferentes al estudio de la regla de Bergmann (sensu Gaston et al. 2008): intraespecíficas, interespecíficas y basadas en ensamblajes. Mientras que el abordaje intraespecífico se realiza a nivel de población (estudiando la variación de tamaño corporal entre poblaciones de una misma especie que habitan diferentes localizaciones geográficas), tanto los enfoques interespecíficos como los basados en ensamblaje comparan la variación de tamaños a nivel de especie. En consecuencia, estas dos últimas aproximaciones han sido nombradas indistintamente como “interespecíficas”, situación que ha dado lugar a cierta confusión. No obstante, las diferencias metodológicas entre ambas no son triviales (Blackburn & Hawkins 2004, Ruggiero &Hawkins 2006). El clásico enfoque interespecífico (cross-species) trata cada especie como un dato independiente (es decir, la unidad de análisis son las especies) y usa diagramas de dispersión bivariados para examinar la covariación tamaño

4

corporal-latitud (o temperatura) entre especies. Esta metodología también se conoce como aproximación del “punto medio”, puesto que conlleva obtener para cada especie una única medida espacial para describir su rango de distribución geográfica (generalmente el punto latitudinal medio de su área de distribución). Por el contrario, los estudios basados en ensamblaje usan mallas de celdas que cubren la región de estudio (es decir, la unidad de análisis es espacial) para explorar los patrones geográficos en la variación de tamaños corporales. Combinando las presencias y ausencias en cada celda de todas las especies que componen el ensamblaje faunístico del área de estudio, junto con sus respectivos tamaños corporales, se obtienen valores de tamaño corporal medio a nivel de celda (generalmente medias geométricas transformadas logarítmicamente) y se estudia su asociación con descriptores ambientales. La principal ventaja, por tanto, de los estudios basados en ensamblajes es que permiten una evaluación directa de la estructura ambiental que subyace a los patrones biogeográficos, frente a la limitación de los análisis interespecíficos que reducen a un único punto en términos geográficos o ambientales la naturaleza multidimensional del rango de distribución de las especies (Blackburn & Hawkins 2004, Ruggiero &Hawkins 2006). Puesto que los procesos ecológicos o evolutivos que podrían generar patrones en la variación del tamaño corporal tienen lugar en un contexto geográfico, varios autores han enfatizado la necesidad de emplear enfoques espacialmente explícitos basados en ensamblaje para evitar alcanzar conclusiones equívocas o contradictorias en el estudio de la Regla de Bergmann (Blackburn & Hawkins 2004, Ruggiero &Hawkins 2006).

Junto con las posibles limitaciones metodológicas anteriormente mencionadas para detectar patrones e inferir procesos, una cuestión

de fondo ha contribuido también en gran medida a alimentar el debate alrededor de la validez de la Regla de Bergmann. A comienzos de los años sesenta, dos trabajos pioneros mostraron a nivel intra- (Ray 1960) e interespecífico (Lindsey 1966) que algunos organismos ectotermos exhiben clinas de variación de tamaño similares a las de mamíferos y aves. Teniendo en cuenta los diferentes comportamientos térmicos de endotermos y ectotermos, estas investigaciones llamaron la atención sobre la posibilidad de tener que considerar alternativas al clásico mecanismo fisiológico de Bergmann para explicar los patrones en organismos que no producen calor de forma endógena. Comenzó entonces un debate que ha estado vigente durante los últimos años (véase por ejemplo Van Voorhies 1996, Mousseau 1997, Partridge and Coyne 1997, Van Voorhies 1997) alrededor de cuáles deberían ser los procesos que promueven la aparición de gradientes de tamaño corporal en ectotermos. Lamentablemente, las conclusiones de estas investigaciones proceden en su mayoría de la evidencia experimental de trabajos de laboratorio para especies concretas, mientras existe un importante desconocimiento de cuáles son los patrones reales en la naturaleza (Belk & Houston 2002).

Esta tesis doctoral emplea un enfoque basado en ensamblaje para documentar la existencia de patrones biogeográficos de variación del tamaño corporal en las faunas regionales (Paleártico occidental, Neártico y Neotrópico) de diferentes grupos de vertebrados (mamíferos, anfibios y reptiles). Asimismo, se estudia la asociación de los gradientes de tamaño corporal con una serie de descriptores ambientales para evaluar el grado de apoyo de las hipótesis que han sido propuestas para explicar la regla de Bergmann hasta la fecha (Cushman et al. 1993, Blackburn et al. 1999). Con el objetivo de explorar la generalidad de los patrones biogeográficos observados en mamíferos

5

y aves y a fin de mejorar nuestra comprensión de posibles mecanismos energéticos y fisiológicos, se ha prestado especial atención a dos grupos de vertebrados ectotermos (anfibios y reptiles) que, por definición, muestran habilidades limitadas para producir calor interno.

A nivel metodológico se ha empleado un enfoque innovador basado en la recopilación de información sobre la biología y distribución de los organismos, su procesado mediante Sistemas de Información Geográfica (GIS) y un posterior análisis multivariante combinado con métodos avanzados de estadística espacial para evaluar la estructura de autocorrelación espacial en los datos. Además, se ha recurrido al uso de un procedimiento de selección de modelos basado en teoría de la información. En los últimos años, la utilización de técnicas de selección de modelos basadas en el criterio de información de Akaike (AIC) está ganando una gran aceptación en el área de investigación ecológica cómo método de inferencia estadística que supera alguna de las limitaciones y problemas inherentes a la clásica estrategia de simplificación de modelos mediante un proceso de regresión por pasos (stepwise) (Whittingham et al. 2006, Stephens et al. 2007). Igualmente se ha incorporado, en los dos últimos capítulos, un método de regresión filogenética basada en autovectores (PVR) (Diniz-Filho et al. 1998) que evalúa el grado de inercia filogenética en los datos y permite particionar el componente ecológico y filogenético de la variación de tamaños corporales (Diniz-Filho et al. 2007). El uso de estas metodologías analíticas ha permitido abordar el estudio de la regla de Bergmann considerando de forma simultánea la distribución de las especies en un marco geográfico y filogenético.

La tesis se estructura en un formato de artículos científicos con sus correspondientes secciones de introducción, material y métodos, resultados y

discusión. Cuatro de los cinco trabajos que la componen han sido sucesivamente aceptados para su publicación en revistas internacionales con índice de impacto (Global Ecology & Biogeography, Journal of Biogeography y Ecography), mientras que el último capítulo ha sido enviado recientemente a Evolution y se encuentra en proceso de revisión.

Referencias:

Allen, J.A. (1878) The influence of physical conditions in the genesis of species. Radical Review, 1, 108–140.

Ashton, K. G., Tracy, M. C., & de Queiroz, A. (2000) Is Bergmann's rule valid for mammals? American Naturalist 156, 390-415.

Ashton, K.G. (2001) Are ecological and evolutionary rules being dismissed prematurely? Diversity and Distributions, 7, 289-295.

Ashton, K.G. (2002) Patterns of within-species body size variation of birds: strong evidence for Bergmann’s rule. Global Ecology and Biogeography, 11, 505–523.

Belk, M.C., and Houston, D.D. (2002) Bergmann’s rule in ectotherms: a test using freshwater fishes. American Naturalist, 160, 803–808.

Bergmann, C. (1847) Ueber die Verhältnisse der Wärmeökonomie der Thiere zu ihrer Grösse. Gottinger Studien, 3, 595-708.

Blackburn, T. M. & Gaston, K. J. (1996) Spatial patterns in the body sizes of bird species in the New World. Oikos, 77, 436-446.

Blackburn, T. M., Gaston, K. J. & Loder, N. (1999) Geographic gradients in body size: a clarification

6

of Bergmann’s rule. Diversity and Distributions, 5, 165-174.

Blackburn, T.M., & Hawkins, B.A. (2004) Bergmann’s rule and the mammal fauna of northern North America. Ecography 27: 715–724.

Brown, J.H & Maurer, B.A. (1989) Macroecology: The division of food and space among species on continents. Science, 243: 1145-1150.

Brown, J. H. & Nicoletto, P.F. (1991) Spatial scaling of species compostion: body masses of North American land mammals. American Naturalist, 138, 1478-1512.

Brown, J.H. (1995) Macroecology. University of Chicago Press, Chicago, IL.

Brown J.H. (1999) The legacy of Robert MacArthur: From geographical ecology to macroecology. Journal of Mammology, 80, 333-344.

Cope, E.D. (1887) The Origin of the Fittest. Appleton Press, New York.

Cotgreave, P. & Stockley, P. (1994) Body size, insectivory and abundance in assemblages of small mammals. Oikos, 71, 89-96.

Cushman, J.H., J.H. Lawton & Manly B.F.J. (1993) Latitudinal patterns in European ant assemblages: variation in species richness and body size. Oecologia, 95, 30–37.

Diniz-Filho, J. A. F., C. E. Ramos de Sant'Ana, & Bini, L. M. (1998) An eigenvector method for estimating phylogenetic inertia. Evolution, 52, 1247–1262.

Diniz-Filho, J.A.F., L.M. Bini, M.Á. Rodríguez, T.L.V.B. Rangel, and Hawkins, B.A. (2007) Seeing the forest for the trees: partitioning ecological and phylogenetic components of

Bergmann’s rule in European Carnivora. Ecography, 30, 598–608.

Foster, J.B. (1964) Evolution of mammals on islands. Nature, 202, 234–235.

Gaston, K.J., Blackburn, T.M. & Spicer, J.I. (1998) Rapoport’s rule: time for an epitaph? Trends in Ecology and Evolution 13, 70–74.

Gaston, K.J., S. Chown, and K.L. Evans. (2008) Ecogeographical rules: elements of a synthesis. Journal of Biogeography, 35, 483–500.

Geist, V. (1987) Bergmann’s rule is invalid. Canadian Journal of Zoology, 65, 1035–1038.

Gloger, C.L. (1883) Das Abandern der Vogel durch Einfluss des Klimas. A. Schulz, Breslau, Germany.

Gould, S.J. (1997) Cope’s rule as psychological artefact. Nature 385, 199–200.

Hawkins, B.A. (2001) Ecology’s oldest pattern? Trends in Ecology & Evolution, 16, 470.

James, F. C. (1970) Geographic size variation in birds and its relationship to climate. Ecology, 51, 365-390.

Jordan, D.S. (1892) Temperature and vertebrae: a study in evolution. Wilder-Quarter Century Books, New York.

Lawton, J.H. (1990) Species richness and population dynamics of animal assemblages: patterns in body size: abundance space. Philosophical Transactions of the Royal Society of London. B, 330, 283-291.

Lindsey, C. C. (1966) Body sizes of poikilotherm vertebrates at different latitudes. Evolution, 20, 456-465.

7

Lomolino, M. V., D. F. Sax, B. R. Riddle & Brown, J.H. (2006) The island rule and a research agenda for studying ecogeographic patterns. Journal of Biogeography 33, 1503-1510.

McArthur, R.H. (1972) Geographical ecology: patterns in the distribution of species. Harper and Row, New York.

McNab, B. K. (1971) On the ecological significance of Bergmann's rule. Ecology, 52, 845-854.

Mayr, E. (1942) Systematics and the origin of species. Columbia University Press. New York.

Mayr, E. (1956) Geographical character gradients and climatic adaptation. Evolution, 10, 105–108.

Medina, A.I., D.A. Martì & Bidau, C.J. (2007) Subterranean rodents of the genus Ctenomys (Caviomorpha, Ctenomyidae) follow the converse to Bergmann’s rule. Journal of Biogeography, 34, 1439–1454.

Meiri, S. & Dayan, T. (2003) On the validity of Bergmann's rule. Journal of Biogeography, 30, 331-351.

Meiri, S., Yom-Tov, Y. & Geffen, E. (2007) What determines conformity to Bergmann’s rule? Global Ecology and Biogeography, 16, 788-794

Mousseau, T. A. (1997) Ectotherms follow the converse Bergmann's rule. Evolution, 51, 630-632.

Partridge, L. & Coyne, J. A. (1997) Bergmann's rule in ectotherms: Is it adaptive? Evolution, 51, 632-635.

Peters, H.R. (1983) The ecological implications of body size. Cambridge University Press, New York.

Rapoport, E.H. (1982) Areography: geographic strategies of species. Pergamon, Oxford, UK.

Ray, C. (1960) The application of Bergmann’s and Allen’s rules to the poikilotherms. Journal of Morphology, 106, 85-108.

Rensch, B. (1938) Some problems of geographical variation and species formation. Proceedings of the Linnean Society of London 150, 275–285.

Rodríguez, M. Á., López-Sañudo, I. L. & Hawkins, B.A. (2006) The geographic distribution of mammal body size in Europe. Global Ecology and Biogeography, 15, 185–193.

Ruggiero, A., & Hawkins, B.A. (2006) Mapping macroecology. Global Ecology and Biogeography, 15, 433–437.

Scholander, P.F. (1955) Evolution of climatic adaptation in homeotherms. Evolution, 9, 15–26.

Stephens, P.A., Buskirk, S.W. & Martínez del Rio, C. (2007) Inference in ecology and evolution. Trends in Ecology and Evolution, 22, 192–197.

Van Voorhies, W. A. (1996) Bergmann size dines: A simple explanation for their occurrence in ectotherms. Evolution, 50, 1259-1264.

Van Voorhies, W. A. (1997) On the adaptive nature of Bergmann size clines: A reply to Mousseau, Partridge and Coyne. Evolution, 51, 635-640.

Whittingham, M.J., Stephens, P.A. & Bradbury, R.B. & Freckleton, R.P. (2006) Why do we still use stepwise modeling in ecology and behaviour? Journal of Animal Ecology, 75, 1182–1189

Zeveloff, S. I. & Boyce, M. S. (1988) Body size patterns in North American mammal faunas. In: Evolution of Life Histories of Mammals (ed. M. S. Boyce), pp. 123-146. Yale University Press, New Haven.

8

Capítulo 2 Regla de Bergmann y la geografía del tamaño corporal en mamíferos del Hemisferio Oeste Este capítulo reproduce íntegramente el texto del siguiente manuscrito: Rodríguez, M. Á., M. Á. Olalla-Tárraga & B. A. Hawkins. 2008. Bergmann’s rule and the geography of mammal body size in the Western Hemisphere. Global Ecology & Biogeography 17: 274 - 283.

DOI: 10.1111/j.1466-8238.2007.00363.x © 2007 The Authors

274

Journal compilation © 2007 Blackwell Publishing Ltd www.blackwellpublishing.com/geb

Global Ecology and Biogeography, (Global Ecol. Biogeogr.)

(2008)

17

, 274–283

RESEARCHPAPER

Blackwell Publishing Ltd

Bergmann’s rule and the geography of mammal body size in the Western Hemisphere

Miguel Á. Rodríguez

1

*, Miguel Á. Olalla-Tárraga

1

and Bradford A. Hawkins

2

ABSTRACT

Aim

To describe the geographical pattern of mean body size of the non-volantmammals of the Nearctic and Neotropics and evaluate the influence of five environ-mental variables that are likely to affect body size gradients.

Location

The Western Hemisphere.

Methods

We calculated mean body size (average log mass) values in 110

×

110 kmcells covering the continental Nearctic and Neotropics. We also generated cellaverages for mean annual temperature, range in elevation, their interaction, actualevapotranspiration, and the global vegetation index and its coefficient of variation.Associations between mean body size and environmental variables were tested withsimple correlations and ordinary least squares multiple regression, complementedwith spatial autocorrelation analyses and split-line regression. We evaluated therelative support for each multiple-regression model using AIC.

Results

Mean body size increases to the north in the Nearctic and is negativelycorrelated with temperature. In contrast, across the Neotropics mammals are largestin the tropical and subtropical lowlands and smaller in the Andes, generating a positivecorrelation with temperature. Finally, body size and temperature are nonlinearlyrelated in both regions, and split-line linear regression found temperature thresholdsmarking clear shifts in these relationships (Nearctic 10.9

°

C; Neotropics 12.6

°

C).The increase in body sizes with decreasing temperature is strongest in the northernNearctic, whereas a decrease in body size in mountains dominates the body sizegradients in the warmer parts of both regions.

Main conclusions

We confirm previous work finding strong broad-scaleBergmann trends in cold macroclimates but not in warmer areas. For the latterregions (i.e. the southern Nearctic and the Neotropics), our analyses also suggest thatboth local and broad-scale patterns of mammal body size variation are influencedin part by the strong mesoscale climatic gradients existing in mountainous areas.A likely explanation is that reduced habitat sizes in mountains limit the presence oflarger-sized mammals.

Keywords

Body size, climatic gradients, habitat zonation, heat tolerance, macroecology, New

World mammals.

*Correspondence: Miguel Á. Rodríguez, Department of Ecology, University of Alcalá, 28871 Alcalá de Henares, Madrid, Spain. E-mail: [email protected]

1

Department of Ecology, University of Alcalá,

28871 Alcalá de Henares, Madrid, Spain and

2

Department of Ecology & Evolutionary Biology,

University of California, Irvine, CA 92697, USA

INTRODUCTION

An increase in body size in cold climates (Bergmann’s rule) is one

of the best known empirical generalizations of geographical

ecology. Research on body size gradients has been intense in

recent decades, likely because body size can be related to key

physiological, ecological and evolutionary characteristics of

animals (McNab, 1979; Peters, 1983; Lindstedt & Boyce, 1985;

Cushman et al., 1993) and, consequently, identifying the factors

underlying geographical variation in body sizes may increase our

understanding of the organization of animal communities

(Lawton, 1990; Brown & Nicoletto, 1991).

9

Body sizes of Western Hemisphere mammals

© 2007 The Authors

Global Ecology and Biogeography

,

17

, 274–283, Journal compilation © 2007 Blackwell Publishing Ltd

275

Studies on broad-scale patterns of body size variation can be

divided into those focused on intraspecific patterns and those

focused on interspecific patterns. Intraspecific patterns have

received the most attention (Ray, 1960; Lindsey, 1966; Ashton

et al., 2000; Meiri & Dayan, 2003; Meiri et al., 2004), although

interspecific gradients have also been studied, especially for

endotherms (mammals: Zeveloff & Boyce, 1988; Cotgreave &

Stockley, 1994; Blackburn & Hawkins, 2004; Rodríguez et al.,

2006; Diniz-Filho et al., 2007; Medina, et al., 2007; birds: Cousins,

1989; Blackburn & Gaston, 1996; Gaston & Blackburn, 2000;

Ramirez et al., 2007). In the case of mammals, two recent

interspecific studies involving the northern Nearctic (Blackburn

& Hawkins, 2004) and the Western Palaearctic (Rodríguez et al.,

2006) found clear Bergmann-like patterns and strong negative

associations between mean body size and mean annual tempera-

ture. This is consistent with the heat conservation mechanism

proposed by Bergmann, that cold climates should harbour more

large-bodied species because of the reduced surface area to

volume ratios (Bergmann, 1847) and/or thicker insulation layers

(Blackburn et al., 1999) of larger animals. However, Rodríguez

et al. (2006) also suggested that this hypothesis might need to be

reformulated to include nonlinear relationships between body

size and temperature over very broad geographical scales,

because they found that the association between mammal body

size and temperature was strong in northern Europe, whereas it

virtually disappeared in the south. Although the geographical

scope of Blackburn & Hawkins (2004) did not extend into areas

with warm climates, they also reported evidence of a nonlinear

relationship between mean body size and temperature. Thus,

both studies suggest that broad-scale gradients of mammal body

size are more influenced by temperature in colder climates than

in warmer ones; this is not unexpected under the heat conservation

mechanism.

To explore this issue further, we examine geographical

variation in mean body size of the non-volant terrestrial

mammal faunas of the Western Hemisphere. Based on the

previous analyses, we expect mean body size to exhibit a negative

relationship with temperature in the far north and south, but this

relationship should weaken or disappear in more temperate and

tropical areas. In addition to the heat conservation hypothesis,

we also evaluate three other explanations for variation in body

size within the constraints imposed by a correlative approach:

(1) the heat dissipation hypothesis, which states that the higher

rates of heat loss of small-bodied species should favour them in

warm, moist climates, where evaporative cooling is more difficult

(Brown & Lee, 1969; James, 1970); (2) the resource availability

hypothesis, which proposes that large-bodied species are

favoured in areas with seasonal shortages in resources because

they metabolize fat stores at lower weight-specific rates than

smaller species (Lindstedt & Boyce, 1985; Dunbrack & Ramsay,

1993); and (3) the habitat availability hypothesis, developed to

explain geographical range size variation in the Western Hemi-

sphere avifauna (Hawkins & Diniz-Filho, 2006), which suggests

that the greater habitat zonation associated with stronger mesoscale

climatic gradients in tropical mountainous areas would limit

the occurrence of large-sized mammal species. A fifth possible

explanation emphasizes dispersal differences between large- and

small-bodied animals and the colonization of the new habitats

created during the retreat of the late Pleistocene ice sheets (e.g.

Olalla-Tárraga et al., 2006; Olalla-Tárraga & Rodríguez, 2007).

However, we excluded this hypothesis because it has been shown

to be an unlikely explanation for the mean body size patterns of

the Holarctic mammal fauna (see Blackburn & Hawkins, 2004;

and Rodríguez et al., 2006) and because only 1.2% of the

Neotropics were covered by ice during the late Pleistocene

(Adams & Faure, 1997).

MATERIALS AND METHODS

Mammal species data

Digital range maps for the 1328 non-volant terrestrial mammal

species native to the Western Hemisphere were obtained from

Patterson et al. (2005). The maps were rasterized in ArcGIS 9.2

using separate 110 × 110 km grids for North and South America.

We excluded all islands except Tierra del Fuego. We also excluded

all coastal cells containing < 50% of the land mass of inland cells

and used the mammal zoogeographical regions in Cox (2001) to

differentiate the Nearctic and Neotropics.

Average body mass (in grams) for each of 1082 species was

taken from Smith et al. (2003). For the remaining species, we

searched the literature for body mass measurements of individuals

of each species distributed across the species’ range, and then

averaged them. We found body mass or length estimates for 47

and 6 species, respectively (see Appendix S1 in Supplementary

Material). We transformed lengths into body masses using the

equations by Silva & Downing (1995). For 111 additional species,

we determined that they have been sometimes considered

subspecies, synonyms, or conspecifics of species of known mass,

and assigned these masses to them (see Appendix S2). For 61

species cited as ‘being similar in size to’, or as ‘belonging to the

group of ’ species of known body masses, we assigned these

masses to them (see Appendix S3). We were unable to find direct

or indirect measures of size for 21 species, and we assigned the

average mass of its genus in these cases (see Appendix S4). All

body masses were log

10

-transformed for analysis.

Environmental variables

We selected five variables to evaluate four hypotheses for body

size gradients, as follows.

Heat conservation

We used mean annual temperature as our indicator of heat.

The data were obtained from http://www.grid.unep.ch/data/

summary.php?dataid=GNV15. Potential evapotranspiration

has also frequently been used as an indicator of ambient energy

inputs (reviewed by Hawkins et al., 2003). However, this variable

(the Priestley–Taylor equation) is highly correlated with mean

annual temperature in the Western Hemisphere (r = 0.907,

n = 3111), so we did not use it.

10

M. Á. Rodríguez

et al.

© 2007 The Authors

276


,

17


Heat dissipation

We used cell averages of annual actual evapotranspiration (AET)

as our water-energy measure. Evaporative cooling should be

more difficult in areas with high AET. These data are available at

http://www.grid.unep.ch/data/summary.php?dataid=GNV183.

Resource availability

Following Blackburn & Hawkins (2004), we used the mean

monthly global vegetation index (GVI) and its temporal

coefficient of variation across the year (GVIcv) to test for effects

of plant production and seasonality on mammal body mass.

These variables were estimated from monthly values from April

1985 to December 1988, available at http://www.ngdc.noaa.gov/

seg/cdroms/ged_iia/datasets/a01/mgv.htm.

Habitat availability

We followed the logic and procedures used by Hawkins & Diniz-

Filho (2006) to investigate potential indirect effects of mesoscale

climatic gradients on mean body size variation occurring through

the effects of mesoclimate on habitat variability. Thus, we

used range in elevation (i.e. the difference between maximum

and minimum elevations within each grid cell) as our indicator

of variation in mesoclimate, and we combined this variable

with its interaction with mean annual temperature in multi-

factor models to account for potential differential trends of body

size variation in mountains of cold and warm areas (see Hawkins

& Diniz-Filho, 2006). Elevation data were obtained at http://

www.ngdc.noaa.gov/seg/cdroms/ged_iia/datasets/a13/fnoc.htm.

Statistical analyses

Because of regional differences in the body size gradients that

were obvious when we mapped the data (Fig. 1), the Nearctic

and Neotropical data sets were analysed separately. First, we used

simple correlation to explore associations among response and

explanatory variables. We then used least-squares multiple

regression to identify the models best supported by the data. For

this we applied model selection techniques based on information

theory which we preferred over stepwise regression because it is

well known that the results of the latter analysis are highly

dependent on the strategy chosen for model simplification, i.e.

forward and backward strategies and their combinations

(Whittingham et al., 2006). Further, the use of model selection

techniques as an alternative to hypothesis testing is becoming

increasingly common in ecology (Johnson & Omland, 2004), as

these techniques allow the relative support for each hypothesis to

be evaluated by comparing a complete set of competing models.

Moreover, in our case, given the strong collinearity among

several of the predictors in our data set (Table 1), it was imperative

to assess simultaneously the importance of all the various pre-

dictors included in the analysis. Specifically, we used the Akaike

information criterion (AIC), which we computed for all possible

models for each region using geographically effective sample

sizes (n*), given by n* = n/[(1 + p)/(1 – p)] where p is the first-

order autoregressive parameter of the residuals, approximated by

the standardized Moran’s I in the first distance class (Cressie,

1993; Haining, 2003). We also computed the ΔAIC of each

model (i.e. ΔAIC

i

= AIC

i

– minAIC; where

Δ

AIC > 10 represent

poor fits and

Δ

AIC

≤

2 correspond to models equivalent to the

best model) (Burnham & Anderson, 2001), as well as its Akaike

weighting (w

i

), a value that can be interpreted as the probability

that model

i

is actually the best explanatory model among those

evaluated. We used the 25 models with the lowest AICs for each

analysed region (and subregion, see below) to calculate w

i

. In

addition, we used the standardized regression coefficient of the

variables included in each model to rank the importance of each

variable in determining mean body size variation.

In order to investigate whether environmental influences on

mammal mean body size differ in warm and cold macroclimates

Table 1 Correlations among response (mammal mean body size) and explanatory variables (environmental predictors) in the Nearctic and Neotropics.

Variable Mean size Temp. R. Elev AET GVI GVIcv

(a) Nearctic (n = 1529)

Mean body size 1

Mean annual temperature –0.746 1

Range in elevation –0.061 –0.010 1

Annual evapotranspiration –0.568 0.703 –0.292 1

Global vegetation index (GVI) –0.552 0.632 –0.148 0.848 1

Coefficient of variation of GVI 0.263 –0.445 –0.307 –0.052 0.106 1

(b) Neotropics (n = 1582)

Mean body size 1

Mean annual temperature 0.508 1

Range in elevation –0.578 –0.574 1

Annual evapotranspiration 0.413 0.719 –0.406 1

Global vegetation index (GVI) 0.407 0.717 –0.528 0.822 1

Coefficient of variation of GVI –0.113 –0.118 –0.013 –0.336 –0.129 1

11


© 2007 The Authors


,

17


277

(Rodríguez

et al

., 2006), we used split-line regression (Schmid

et al.

, 1994) to look for potential breakpoints in the relationship

of temperature with mammal mean body size in the Nearctic and

the Neotropics. For this, we developed a two-step procedure.

First, we utilized the quasi-Newton routine of statistica (see

StatSoft, Inc., 2003) to obtain an initial approximation to the

breakpoint in each region. Second, we checked all potential

breakpoints within a temperature range of 5 °C below and above

this approximated value (considering increments of 0.1 °C) and

retained that which maximized the coefficient of determination

as the most likely temperature breakpoint. Finally, for each

region, we performed the model selection protocol described

above for the areas above and below the breakpoint.

We performed two additional sets of analyses to investigate the

robustness of the relationships we found. First, we generated

correlograms (not shown) of the residuals of the best multiple-

regression models to evaluate how well they described spatial

variation in body size across spatial scales (Diniz-Filho et al.,

2003). These correlograms were calculated across complete

regions (i.e. the Nearctic and Neotropics) as well as to the areas

above and below the temperature threshold found for each

region. In all instances we found that our models explained mean

body size patterns at all scales except the smallest distance classes.

This result is common in map-based data, and indicates either

that variables not included in our models may be needed to

account for the spatial pattern of mean body size at smaller

scales, or that the use of range map data has made closely spaced

cells more similar than they should be. Second, because patterns

of mean body size variation are potentially sensitive to the

uneven distribution of species richness across geographical

spaces, which in turn may affect statistical analysis, we used

weighted least squares regression techniques to recalculate the

standardized regression coefficients of the variables included in

our best models, as well as the coefficient of determination of

Figure 1 Mean log10-transformed body sizes of non-volant mammals in the Western Hemisphere. Numbers included in the legend are mass values (in g) generated after antilog transformation. Labelled sites are midpoints of 110 × 110 km cells selected to examine the body size distributions on which the means are based (see Figs 2 & 3).

12

M. Á. Rodríguez

et al.

© 2007 The Authors

278


,

17


each model (

R

2

) (Olalla-Tárraga & Rodríguez, 2007). For all

regions, the results of these analyses (not shown) were similar

to those generated with ordinary least squares regression, thus

indicating that they were not affected by spatial variation in

species richness.

All statistical analyses were performed with statistica

(StatSoft, Inc., 2003) and SAM 2.0 (Spatial Analysis in Macroecology;

Rangel et al., 2006).

RESULTS

The geography of mammal mean body size

Mean body size exhibits a clear Bergmann-like pattern of spatial

variation in the Nearctic, with the largest sizes found in the far

north (e.g. the arctic tundra of Alaska and Canada, and the taiga

of Quebec) and in the northern Rocky Mountains, whereas the

smallest sizes are found in the south, particularly in mountain

ranges such as the Appalachians, the Colorado Rockies, the

Sierra Nevada and the Sierra Madre (Fig. 1). Intermediate mean

body sizes, on the other hand, are particularly found in the

lowlands of the southern half of the region. Histograms of the

frequency of body sizes in 12 size classes (Fig. 2) allow a comparison

of the body size distributions of representative cells of these

zones. The southern cells show right-skewed distributions and

generally lack species of the two larger size classes. In contrast,

the northern cells exhibit multimodal distributions with a

noticeable representation of species with intermediate to large

sizes. Therefore, the northward increase in mean body sizes

observed in the Nearctic is mainly due to a higher proportion of

small-sized species occurring in the south, coupled with an

increase in the proportion of species with intermediate to large

sizes in the north.

The pattern of body size variation is reversed in the Neotropics.

The smallest sizes are distributed across the high Andes (Fig. 1);

whereas the largest sizes cover the tropical lowlands, particularly

in the savannas of Brazil (the cerrado) and Venezuela and the

steppes of Paraguay and northern Argentina (the chacos). There

are also small clusters of larger mean body sizes in the Atacama

Desert and in the eastern half of Nicaragua. The small mean body

sizes in the Neotropical mountains are due to a high proportion

of species having small sizes (i.e. right-skewed body size distribu-

tions), whereas the larger mean sizes of the tropical lowlands are

associated with a high proportion of intermediate-sized species,

particularly of those belonging to the size class of 3.5–4 log

10

g

(i.e. with a body size between 3.2 and 10 kg) (Fig. 3).

In sum, mountain ranges in both the southern half of the

Nearctic and throughout the Neotropics support small-bodied

species, but the largest mammals occur in areas with very different

climates in each region, that is, throughout the cold, northern

Nearctic and in the warm lowlands of the central Neotropics.

Mean body size was strongly negatively correlated with mean

annual temperature and, to a lesser extent, with AET and GVI in

the Nearctic (Table 1a), which is consistent with the observed

increase in body size towards the cold, low-productive areas of

the north. Similarly, a strong negative correlation between mean

Figure 2 Histograms of mammal body sizes of species occurring in nine 110 × 110 km Nearctic cells. Species richness values (s) are provided for each cell. See Fig. 1 for cell locations.

Figure 3 Histograms of mammal body sizes of species occurring in 11 110 × 110 km Neotropical cells. Species richness values (s) are provided for each cell. See Fig. 1 for cell locations.

13


© 2007 The Authors


,

17


279

body size and range in elevation in the Neotropics (Table 1b)

reflects the small mean body sizes in the Andes and the large sizes

in the tropical lowlands. Temperature has the second strongest

correlation with mean body size among our environmental

variables in this region, but this correlation is positive.

Multiple regression models of mean body size

The Nearctic

Our multimodel selection protocol found one good model in the

Nearctic, which described 59% of mean body size variance and

was strongly favoured as the best model according to its w value

(Table 2a). Even so, collinearity among the model’s variables (see

Table 1) means that its interpretation must be cautious, so we

focused on variables with either higher standardized regression

coefficients – that is, mean annual temperature and its interaction

with topography, both with a negative sign – or low correlation

with the rest of the predictors, i.e. range in elevation which had a

positive sign (see Table 2a). The negative relationship with

temperature is what would be expected from the heat conserva-

tion hypothesis, as would also be the positive correlation with

range in elevation, which reflects a general tendency for body

sizes to increase in mountainous areas. However, the low

standardized coefficient of range in elevation reflects that its

association with body size is found only in the northern Rockies,

whereas in more southern mountains this relationship disappears,

or reverses (see Fig. 1). This was captured by the interaction

between macroclimate (measured by temperature) and range in

elevation, thus accounting for the aforementioned shift from

larger mammals in mountains in cold climates to smaller

mammals in mountains in warmer climates.

We further explored these patterns by plotting the relationship

between mean body size and temperature, and found it to be

U-shaped (Fig. 4), which led us to investigate both sides of the

temperature gradient. First, we used split-line regression in order

to find the most likely temperature breakpoint in the relationship

of this variable with mammal mean body size. This breakpoint

was found at 10.9 °C. Then, we built environmental models for

the areas below and above this threshold temperature (covering

86% and 14% of the region, respectively). In the cold Nearctic

(Table 2b), we obtained two equally likely models that described

similar proportions of variance (61%) and were overwhelmingly

dominated by a negative relationship of body size with tem-

perature, as indicated by the higher standardized regression

coefficient of this variable. Moreover, simple regression analysis

revealed that temperature alone described much of the variance

of mean body size in this region (57%). This suggests a pre-

eminent role for the heat conservation mechanism in the cold

Nearctic. In contrast, in the warmer Nearctic (Table 2c), we

found only a single ‘best’ model which described 54% of variance

(Table 2c). In this case, temperature had a positive sign but

was of secondary importance compared to range in elevation,

which had the highest standardized coefficient and a negative

association with body size. This indicated a trend for animals to

be smaller in mountain ranges than in the plains, consistent

with the habitat availability hypothesis for the warmer part of

the Nearctic.

Table 2 Environmental multiple regression models for mammal mean body size in the Nearctic, the Neotropics, and the areas of each region on each side of the break points identified by split-line regression (see text). The standardized regression coefficients of the predictors included in the ‘best’ models (i.e. with ΔAIC ≤ 2; see methods) are provided, along with the coefficient of determination (R2), AIC and the Akaike weighting of each model (wi). These information theory indices were calculated correcting for the presence of spatial autocorrelation in the model residuals. Predictor variables are: Temp., mean annual temperature; R. Elev., range in elevation; AET, annual actual evapotranspiration; GVI, mean monthly global vegetation index; and GVIcv, temporal coefficient of variation of GVI.

Predictors in model

R2 AIC wiTemp. R. Elev Temp. × R. Elev. AET GVI GVIcv

(a) Nearctic

–0.519 0.134 –0.345 –0.203 –0.104 0.589 –5265 0.949

(b) Nearctic (Temp. ≤ 10.9 °C)

–0.657 –0.149 0.195 –0.251 0.194 0.612 –4108 0.497

–0.661 –0.146 0.185 –0.266 0.029 0.185 0.612 –4108 0.495

(c) Nearctic (Temp. > 10.9 °C)

0.394 –0.523 –0.144 –0.369 0.539 –1846 0.470

(d) Neotropics

0.688 0.871 –1.163 –0.054 0.459 –4817 0.913

(e) Neotropics (Temp. ≤ 12.6 °C)

–0.156 –0.365 0.236 –577 0.282

0.044 –0.176 –0.357 0.238 –576 0.165

–0.175 –0.306 –0.078 0.240 –575 0.131

0.001 –0.156 –0.365 0.236 –575 0.109

(f) Neotropics (Temp. > 12.6 °C)

0.367 0.570 –1.006 0.124 –0.032 0.100 0.430 –4235 0.995

14

M. Á. Rodríguez

et al.

© 2007 The Authors

280


,

17


Finally, our models also captured secondary relationships of

mean body size with actual evapotranspiration (AET) and

seasonal variability in vegetation (GVIcv) in all Nearctic areas.

However, it is unclear what these relationships mean, given the

low standardized coefficients of AET and GVIcv, as well as their

collinearity with temperature (Table 1a).

The Neotropics

The single best model for this region described 46% of the

variance in mean body size, and had a 91% probability of being

the best model (Table 2d). This model included mean annual

temperature, range in elevation, the interaction term between

these variables, and GVI, although the standardized regression

coefficients identified the interaction between temperature and

topography as having the strongest relationship with body size.

Additionally, the negative sign of this interaction term reflects

that mean body sizes tend to decrease towards mountainous

areas with warm macroclimates (e.g. in the tropical Andes),

which is again consistent with the habitat availability hypothesis.

Range in elevation ranked second and had a positive coefficient,

in contrast to both the negative simple correlation between mean

body size and range in elevation in this region (see Table 1b) and

to the concentration of small-sized mammals in the mountains

and of large-sized mammals in the tropical lowlands (see Fig. 1).

A possible explanation would be that mean body sizes are

relatively larger in the mountains when controlling for the other

environmental variables. However, strong collinearity between

range in elevation and its interaction with temperature (r =

0.930, n = 1582) cast doubts about any biological interpretation.

The standardized regression coefficient for temperature was even

lower and positive, in clear contrast to the model for the whole

Nearctic (see Table 2a). The best model also included GVI, but

with a low standardized coefficient, thus suggesting that the

independent effect of plant production is marginal in this region.

Similar to the Nearctic, the relationship between body size and

temperature was U-shaped in the Neotropics (Fig. 4), although

split-line regression analysis found that the threshold temperature

marking the shift in the relationship was slightly higher in this

case (12.6 °C). The below-threshold domain included a low

number of cells (192, representing 22% of the region) that

comprised a nearly continuous strip covering the Andes south to

the equator (171 cells), southern Patagonia, and Tierra de Fuego

(not shown). Notably, temperature was not included in any of

the four best models found for this area (Table 2e), which might

be related to the reduced extent of the temperature gradient

found there (see Fig. 4). Yet, all the models did include GVI and

AET, both with negative coefficients, but this was difficult to

interpret as all models explained body size poorly (R

2

= 24% in

all cases). This low explanatory power might be another

consequence of the narrow climatic gradient in this area. As for

the above-threshold Neotropics (1390 cells), only one model was

selected, with a 99.5% chance of being the best model (Table 2f).

According to its standardized coefficients, this model was clearly

dominated by the interaction between temperature and topography,

which had a negative sign, while range in elevation and temperature

rated second and third, both with positive signs. This matches

the model found for the whole Neotropics (see Table 2c), which

is not unexpected given that 78% of the Neotropical cells had

above-threshold temperatures (see Fig. 4). Our interpretation of

this result was therefore the same as that made above for the

entire region.

Figure 4 Mean body mass (in g) as a function of mean annual temperature in the Nearctic and Neotropics. The dashed lines are threshold temperatures marking a shift in the relationships with mean body size in each region (Nearctic 10.9 °C; Neotropics 12.6 °C) identified by split-line linear regressions (see text). The below-threshold (cooler) Nearctic area is larger and colder (1144 cells; regional average temperature ± 1 SD, –1.3 ± 7.1 °C; minimum, –19.4 °C) than its Neotropical equivalent (192 cells; regional average temperature 4.6 ± 6.9 °C; minimum –11.9 °C). In contrast, the warmer part of the Nearctic to the right of the threshold is also colder but smaller (385 cells; regional average temperature 16.5 ± 3.2 °C; maximum 35.3 °C) than the Neotropical area (1390 cells; regional average temperature 22.9 ± 3.4 °C; maximum 32.3 °C).

15


© 2007 The Authors Global Ecology and Biogeography, 17, 274–283, Journal compilation © 2007 Blackwell Publishing Ltd 281

DISCUSSION

Our clearest result at the level of a biogeographical region is the

strong difference between the Nearctic and the Neotropics

with respect to patterns of mammal mean body size. Whereas

temperature is the strongest predictor of body size in the Nearctic,

with the animals becoming larger towards colder latitudes,

Bergmann’s rule is not observed in the Neotropics, where mean

body sizes increase towards the warmer tropical areas (Fig. 1).

Blackburn and Hawkins (2004) found strong support for

Bergmann’s rule in their interspecific analysis of the mammal

fauna of northern North America, but their study did not include

warm regions. However, in a study involving a larger latitudinal

span in the Western Palaearctic, Rodríguez et al. (2006) found

nonlinear relationships between mammal mean body size and

temperature, with a clear Bergmann’s trend occurring in the

northern, colder half of this region, but not in the south. They

proposed that the Bergmann’s heat-conservation mechanism

should generate gradients of increasing mean body size with

decreasing temperature only in areas with cold climates. Bearing

in mind that the Nearctic is a much colder biogeographical

region than the Neotropics (average temperatures 3.2 °C and

20.7 °C, respectively), this is supported by our data.

This said, our results also suggest that the different body size

trends of the Nearctic and Neotropics have more to do with the

particular environmental characteristics of each region than with

their dissimilar faunas (e.g. the presence of primates in the

Neotropics and the absence of them from the Nearctic), or with

their evolutionary histories before or after the late Cenozoic, i.e.

when the Great American Biotic Interchange occurred (Marshall

et al., 1982; Croft, 2001). For example, in the Nearctic, the data

indicate that mean body sizes do not display uniform trends

across the region that can be linked to regional evolutionary

history. Instead, we found a clear temperature threshold (10.9 °C)

that divides the Nearctic into two areas with opposing relation-

ships between mean body size and the environment. Thus,

whereas in the northern part of the region mean body sizes

increase with cold, in the warmer south topography is the strongest

predictor, just as the habitat availability hypothesis predicts (see

below). Moreover, temperature is positively related to mean body

size in the models generated for the warmer area, likely reflecting

a spurious side-effect of the influence of topography on body size

gradients, as within the same macroclimate mountains are

typically cooler than lowlands. Interestingly, relationships with

topography resemble the dominant trend in the Neotropics,

where the effects of this factor were even stronger, possibly

because the Neotropics are warmer and have stronger elevational

gradients (see below).

We also found nonlinear relationships between body size and

temperature in the Neotropics, which again were characterized

by a temperature threshold (12.6 °C) below and above which the

association of body size with the environment differs. Yet it

should be noted that, compared with its equivalent in the Nearctic,

the colder part of the Neotropics is much smaller and comprises

a narrower and warmer temperature gradient (see Fig. 4) mostly

represented by cells in the Andes (89%). These characteristics,

either individually or in concert, might have obscured

Bergmann’s trends at the scale of our study, or even precluded

them entirely (see Medina et al., 2007). Since all environmental

models for the cooler part of the Neotropics performed poorly,

we cannot discriminate between these possibilities. Irrespective

of this, the existence of similar temperature thresholds, as well as

the leading role of topography in the Neotropics and in the

warmer part of the Nearctic, both suggest that body size gradients

are controlled by similar environmentally driven mechanisms

(chiefly associated with combinations of temperature and

topography) across the Western Hemisphere.

The nonlinear relationships of body size with temperature

reported by Rodríguez et al. (2006) for the mammal fauna of the

Western Palaearctic suggest that the same environmental

controls might be operating in other regions of the world

(see their Fig. 3). To explore this further, we applied split-line

regression to the Palaearctic data set (for details see Rodríguez

et al., 2006) and found a temperature threshold very similar to

those found in the Western Hemisphere (10.0 °C). Below this

threshold, the mean body size–temperature relationship was very

strong and negative (r = −0.830, n = 287) while the relationship

was very weak above the threshold (r = 0.199, n = 99). Thus,

Rodríguez et al.’s (2006) supposition that the heat conservation

mechanism proposed by Bergmann operates only in cold

climates is supported in both the Old and New Worlds.

In contrast to the patterns in lowlands, body size gradients are

reversed moving into mountains. The likely explanation is that

mountains, especially in the tropics, support large numbers of

small-ranged species (Hawkins & Diniz-Filho, 2006). The

reasons for this seem straightforward: stronger climatic gradients

generate greater habitat zonation which means more, but

smaller, habitats that accumulate large numbers of habitat

specialists. However, as the macroclimate cools moving away

from the tropics, there is less habitat zonation and montane

habitats become more similar to lowland habitats, thus allowing

more broadly distributed species to inhabit both mountains and

lowlands. Because larger-sized species tend to have larger

geographical range sizes (see, e.g., Hernández-Fernández & Vrba,

2005), it is likely that mammal mean body sizes are smaller in the

mountains of the warmer Nearctic and the Neotropics because

habitats are not large enough for many larger species. The

right-skewed body size distributions observed in the mountain

ranges of these areas (see Figs 2 and 3) are consistent with this

explanation. Moreover, if reduced habitat areas lead to the

prevalence of small body sizes in warm mountains, then the

greater importance of topography for mean body size variation

in the Neotropics might be associated with the greater habitat

zonation and, hence, smaller habitats found in the Andes

(Janzen, 1967; see also Ghalambor et al., 2006). This can also

account for the increase of mean body size with increasing

temperature across South America, as it suggests that tem-

perature does not directly influence body size in warm climates

but acts indirectly via effects on habitat size and structure.

Interestingly, Medina et al. (2007) recently found reversed

Bergmann’s trends at the inter- and intraspecific levels in a genus

of rodents (Ctenomys) in the southern Neotropics. Although

16

M. Á. Rodríguez et al.

© 2007 The Authors282 Global Ecology and Biogeography, 17, 274–283, Journal compilation © 2007 Blackwell Publishing Ltd

they did not consider range in elevation, they found strong

positive associations between body size and ambient tempera-

ture variables, which they interpreted to be effects of factors not

included in their study. It is possible that topographical variability

was such a missing factor.

In conclusion, the complex geographical distribution of the

mean body size of the Western Hemisphere mammal fauna

supports previous findings in the Palaearctic reporting Bergmann’s

trends in cold macroclimates but not in warmer macroclimates

(Rodríguez et al., 2006). Moreover, in both the Western

Hemisphere and the Western Palaearctic we found similar

temperature thresholds (between 10 and 12.5 °C) marking the

transition between Bergmann-like patterns (the heat conservation

hypothesis) and no direct relationship between temperature and

average body sizes. And even in regions where temperature and

body size are not directly linked, there may still be indirect

relationships due to the influences of climate on habitat structure

(the habitat availability hypothesis). Thus, we propose that there

are at least two environmental drivers of body size gradients

across broad latitudinal extents, with their relative influence

being dependent on macroclimate; temperature influences body

size in cold regions whereas topographically driven habitat

variation influences patterns in warm areas.

ACKNOWLEDGEMENTS

This study was supported by the Spanish Ministry of Education

and Science (grants REN2003-03989/GLO and CGL2006-03000/

BOS to M.Á.R.; FPU fellowship: AP2005-0636 to M.A.O.-T.; and

sabbatical grant: SAB2003-0213 to B.A.H.). We also thank M.

Hernández Fernández and A. G. Boyer for valuable suggestions

and comments.

REFERENCES

Adams, J.M. & Faure, H. (eds) (1997) QEN members. Review and

atlas of palaeovegetation: preliminary land ecosystem maps of the

world since the Last Glacial Maximum. Oak Ridge National

Laboratory, Oak Ridge, TN (http://www.esd.ornl.gov/

projects/qen/adams1.html).

Ashton, K.G., Tracy, M.C. & de Queiroz, A. (2000) Is Bergmann’s

rule valid for mammals? The American Naturalist, 156, 390–

415.

Bergmann, C. (1847) Ueber die Verhältnisse der Wärmeökonomie

der Thiere zu ihrer Grösse. Gottinger Studien, 3, 595–708.

Blackburn, T.M. & Gaston, K.J. (1996) Spatial patterns in the

body sizes of bird species in the New World. Oikos, 77,

436–446.

Blackburn, T.M. & Hawkins, B.A. (2004) Bergmann’s rule and

the mammal fauna of northern North America. Ecography, 27,

715–724.

Blackburn, T.M., Gaston, K.J. & Loder, N. (1999) Geographic

gradients in body size: a clarification of Bergmann’s rule.

Diversity and Distributions, 5, 165–174.

Brown, J.H. & Lee, A.K. (1969) Bergmann’s rule and climatic

adaptation in woodrats (Neotoma). Evolution, 23, 329–338.

Brown, J.H. & Nicoletto, P.F. (1991) Spatial scaling of species

composition: body masses of North American land mammals.

The American Naturalist, 138, 1478–1512.

Burnham, K.P. & Anderson, D.R. (2001) Kullback–Leibler

information as a basis for strong inference in ecological studies.

Wildlife Research, 28, 111–119.

Cotgreave, P. & Stockley, P. (1994) Body size, insectivory and abun-

dance in assemblages of small mammals. Oikos, 71, 89–96.

Cousins, S.H. (1989) Species richness and the energy theory.

Nature, 340, 350–351.

Cox, C.B. (2001) The biogeographic regions reconsidered.

Journal of Biogeography, 28, 511–523.

Cressie, N.A.C. (1993) Statistics for spatial data, revised edn.

Wiley, New York.

Croft, D.A. (2001) Cenozoic environmental change in South

America as indicated by mammalian body size distributions

(cenograms). Diversity and Distributions, 7, 271–287.

Cushman, J.H., Lawton, J.H. & Manly, B.F.J. (1993) Latitudinal

patterns in European ant assemblages: variation in species

richness and body size. Oecologia, 95, 30–37.

Diniz-Filho, J.A.F., Bini, L.M. & Hawkins, B.A. (2003) Spatial

autocorrelation and red herrings in geographical ecology.

Global Ecology and Biogeography, 12, 53–64.

Diniz-Filho, J.A.F., Bini. L.M., Rodríguez, M.Á., Rangel, T.F.L.V.B.

& Hawkins, B.A. (2007) Seeing the forest for the trees: partition-

ing ecological and phylogenetic components of Bergmann’s

rule in European Carnivora. Ecography, 30, 598–608.

Dunbrack, R.L. & Ramsay, M.A. (1993) The allometry of mam-

malian adaptations to seasonal environments: a critique of the

fasting endurance hypothesis. Oikos, 66, 336–342.

Gaston, K.J. & Blackburn, T. (2000) Pattern and process in

macroecology. Blackwell Science, Oxford.

Ghalambor, C.K., Huey, R.B., Martin, P.R., Tewksbury, J.J. &

Wang, G. (2006) Are mountain passes higher in the tropics?

Janzen’s hypothesis revisited. Integrative and Comparative

Biology, 46, 5–17.

Haining, R. (2003) Spatial data analysis. Cambridge University

Press, Cambridge.

Hawkins, B.A. & Diniz-Filho, J.A.F. (2006) Beyond Rapoport’s

rule: evaluating range size patterns of New World birds in a

two-dimensional framework. Global Ecology and Biogeography,

15, 461–469.

Hawkins, B.A., Field, R., Cornell, H.V., Currie, D.J., Guégan, J.-F.,

Kaufman, D.M., Kerr, J.T., Mittelbach, G.G., Oberdorff, T.,

O’Brien, E.M., Porter, E.E. & Turner, J.R.G. (2003) Energy,

water, and broad-scale geographic patterns of species richness.

Ecology, 84, 3105–3117.

Hernández Fernández, M. & Vrba, E.S. (2005) Body size, biomic

specialization and range size of African large mammals.


James, F.C. (1970) Geographic size variation in birds and its

relationship to climate. Ecology, 51, 365–390.

Janzen, D.H. (1967) Why mountain passes are higher in the trop-

ics. The American Naturalist, 101, 233–249.

Johnson, J.B. & Omland, K.S. (2004) Model selection in ecology

and evolution. Trends in Ecology & Evolution, 19, 101–108.

17



Lawton, J.H. (1990) Species richness and population dynamics of

animal assemblages. Patterns in body size: abundance space.

Philosophical Transactions of the Royal Society B: Biological

Sciences, 330, 283–291.

Lindsey, C.C. (1966) Body sizes of poikilotherm vertebrates at

different latitudes. Evolution, 20, 456–465.

Lindstedt, S.L. & Boyce, M.S. (1985) Seasonality, fasting

endurance, and body size in mammals. The American Naturalist,

125, 873–878.

Marshall, L.G., Webb, S.D., Sepkoski, J.J. Jr & Raup, D.M. (1982)

Mammalian evolution and the Great American Interchange.

Science, 215, 1351–1357.

McNab, B.K. (1979) On the ecological significance of Bergmann’s

rule. Ecology, 52, 845–854.

Medina, A.I., Dardo A., Martí, D.A. & Bidau, C.J. (2007)

Subterranean rodents of the genus Ctenomys (Caviomorpha,

Ctenomyidae) follow the converse to Bergmann’s rule. Journal

of Biogeography, 34, 1439–1454.

Meiri, S. & Dayan, T. (2003) On the validity of Bergmann’s rule.


Meiri, S., Dayan, T. & Simberloff, D. (2004) Carnivores, biases

and Bergmann’s rule. Biological Journal of the Linnean Society,

81, 579–588.

Olalla-Tárraga, M.Á. & Rodríguez, M.Á. (2007) Energy and

interspecific body size patterns of amphibian faunas in Europe

and North America: anurans follow Bergmann’s rule, urodeles

its converse. Global Ecology and Biogeography, 16, 606–617.

Olalla-Tárraga, M.Á., Rodríguez, M.Á. & Hawkins, B.A. (2006)

Broad-scale patterns of body size in squamate reptiles of

Europe and North America. Journal of Biogeography, 33, 781–

793.

Patterson, B.D., Ceballos, G., Sechrest, W., Tognelli, M.F.,

Brooks, T., Luna, L., Ortega, P., Salazar, I. & Young, B.E. (2005)

Digital distribution maps of the mammals of the Western

Hemisphere, version 2.0. NatureServe, Arlington, VA (http://

www.natureserve.org/getData/mammalMaps.jsp).

Peters, H.R. (1983) The ecological implications of body size.

Cambridge University Press, New York.

Ramirez, L., Diniz-Filho, J.A.F. & Hawkins, B.A. (2007)

Partitioning phylogenetic and adaptive components of the

geographical body-size pattern of New World birds. Global

Ecology and Biogeography, (in press) doi: 10.1111/j.1466-

8238.2007.00346.x

Rangel, T.F.L.V.B., Diniz-Filho, J.A.F. & Bini, L.M. (2006) SAM

v.1.1 – spatial analysis in macroecology, software and user’s

guide (http://www.ecoevol.ufg.br/sam/index.htm).

Ray, C. (1960) The application of Bergmann’s and Allen’s rules to

the poikilotherms. Journal of Morphology, 106, 85–108.

Rodríguez, M.Á., López-Sañudo, I.L. & Hawkins, B.A. (2006)

The geographic distribution of mammal body size in Europe.


Schmid, B., Polasek, W., Weiner, J., Krause, A. & Stoll, P. (1994)

Modeling of discontinuous relationships in biology with

censored regression. The American Naturalist, 143, 494–507.

Silva, M.B. & Downing, J.A. (1995) CRC handbook of mammalian

body masses. CRC Press, Boca Raton, FL.

Smith, F.A., Lyons, S.K, Ernest, S.K.M., Jones, K.E., Kauffman, D.M.,

Dayan, T., Marquet, P.A., Brown, J.H. & Haskell, J.P. (2003)

Body mass of late Quaternary mammals. Ecology, 84, 3403.

StatSoft, Inc. (2003) STATISTICA (data analysis software system),

version 6 (http://www.statsoft.com).

Whittingham, M.J., Stephens, P.A., Bradbury, R.B. & Freckleton,

R.P. (2006) Why do we still use stepwise modelling in ecology

and behaviour? Journal of Animal Ecology, 75, 1182–1189.

Zeveloff, S.I. & Boyce, M.S. (1988) Body size patterns in North

American mammal faunas. Evolution of life histories of mam-

mals (ed. by M.S. Boyce), pp. 123–146. Yale University Press,

New Haven, CT.

Editor: Tim Blackburn

SUPPLEMENTARY MATERIAL

The following supplementary material is available for this article:

Appendix S1 Species with body mass or body length data found

in the literature. Body masses are provided for all species.

Appendix S2 Species considered as synonyms, subspecies or

conspecifics of species with known body masses (see column

labelled ‘Equivalent weighted species’).

Appendix S3 Species considered ‘similar to’ or ‘belonging to the

group’ of species with known body masses (see column labelled

‘Equivalent weighted species’).

Appendix S4 Species for which we were unable to find any direct

or indirect indication of species’ size.

This material is available as part of the online article

from: http://www.blackwell-synergy.com/doi/abs/10.1111/

j.1466-8238.2007.00363.x

(This link will take you to the article abstract).

Please note: Blackwell Publishing is not responsible for the

content or functionality of any supplementary materials supplied

by the authors. Any queries (other than missing material) should

be directed to the corresponding author for the article.

BIOSKETCHES

Miguel Á. Rodríguez’s main interests include the study

of factors and processes conditioning patterns of biodiversity

at local, regional and global scales. His recent research has

involved investigating the effects of habitat destruction

and fragmentation on aggregate properties of faunas at

broad scales.

Miguel Á. Olalla-Tárraga is a PhD student interested in macroecology and conservation biogeography.

Bradford A. Hawkins is interested in biogeography and geographical ecology.

18

Appendix S1 – Species with body mass or body length data found in the literature. Body masses are provided for all species. Codes under the column labelled as “source” indicate the reference where the data were found. The list of references is provided below. Codes followed by “(L)” indicate cases in which mass data were generated from body length data (see main text for details).

Family Genus Species Mass (grams) Source Muridae Abrawayaomys ruschii 63.00 1 Abrocomidae Abrocoma cinerea 250.00 1 Abrocomidae Abrocoma boliviensis 158.00 1 Abrocomidae Abrocoma bennettii 250.50 1 Muridae Abrothrix longipilis 37.60 1 Muridae Abrothrix illutea 47.80 1 Muridae Abrothrix sanborni 24.70 1 Muridae Abrothrix lanosus 32.50 1 Octodontidae Aconaemys sagei 96.50 1 Octodontidae Aconaemys fuscus 133.50 1 Octodontidae Aconaemys porteri 92.00 18 Muridae Aepeomys lugens 37.00 1 Muridae Akodon latebricola 39.00 1 Muridae Akodon torques 39.00 1 Muridae Akodon toba 51.20 1 Muridae Akodon sylvanus 39.00 1 Muridae Akodon surdus 39.00 1 Muridae Akodon spegazzinii 28.60 1 Muridae Akodon simulator 42.50 1 Muridae Akodon siberiae 34.60 1 Muridae Akodon serrensis 28.30 1 Muridae Akodon sanctipaulensis 27.10 1 Muridae Akodon orophilus 39.00 1 Muridae Akodon neocenus 42.40 1 Muridae Akodon mollis 30.40 1 Muridae Akodon molinae 33.00 1 Muridae Akodon mimus 24.00 1 Muridae Akodon lindberghi 26.40 1 Muridae Akodon kofordi 29.50 1 Muridae Akodon juninensis 39.00 1 Muridae Akodon iniscatus 28.70 1 Muridae Akodon fumeus 22.70 1 Muridae Akodon dolores 50.50 1 Muridae Akodon dayi 32.50 1 Muridae Akodon budini 26.90 1 Muridae Akodon albiventer 21.80 1 Muridae Akodon affinis 24.90 1 Muridae Akodon aerosus 60.00 1 Muridae Akodon bogotensis 13.00 1 Muridae Akodon xanthorhinus 21.30 1 Muridae Akodon varius 40.00 1 Muridae Akodon subfuscus 30.40 1 Muridae Akodon olivaceus 27.00 1 Muridae Akodon cursor 39.90 1

19

Muridae Akodon boliviensis 27.50 1 Muridae Akodon azarae 25.00 1 Cervidae Alces alces 358996.30 1 Canidae Alopex lagopus 3450.00 1 Atelidae Alouatta pigra 7000.00 1 Atelidae Alouatta sara 6611.00 1 Atelidae Alouatta caraya 5862.50 1 Atelidae Alouatta belzebul 6400.00 1 Atelidae Alouatta seniculus 6145.50 1 Atelidae Alouatta palliata 7274.90 1 Sciuridae Ammospermophilus nelsoni 160.40 1 Sciuridae Ammospermophilus leucurus 103.70 1 Sciuridae Ammospermophilus interpres 110.30 1 Sciuridae Ammospermophilus harrisii 122.00 1 Muridae Andalgalomys pearsoni 25.40 1 Muridae Andalgalomys olrogi 32.90 1 Muridae Andinomys edax 69.70 1 Muridae Anotomys leander 66.40 1 Antilocapridae Antilocapra americana 46082.90 1 Aotidae Aotus vociferans 873.00 1 Aotidae Aotus nancymaae 788.00 1 Aotidae Aotus miconax 800.00 1 Aotidae Aotus lemurinus 873.00 1 Aotidae Aotus infulatus 800.00 1 Aotidae Aotus azarae 962.90 1 Aotidae Aotus trivirgatus 900.00 1 Aotidae Aotus nigriceps 1060.00 1 Aplodontidae Aplodontia rufa 1004.00 1 Muridae Aporodon brevirostris 12.90 1 Muridae Aporodon creper 22.80 1 Muridae Aporodon tenuirostris 20.00 1 Muridae Aporodon microdon 20.00 1 Muridae Aporodon mexicanus 19.00 1 Muridae Aporodon darienensis 13.00 1 Muridae Aporodon gracilis 12.30 1 Muridae Arborimus pomo 32.50 1 Muridae Arborimus longicaudus 21.80 1 Muridae Arborimus albipes 23.00 1 Atelidae Ateles fusciceps 9100.00 1 Atelidae Ateles marginatus 6000.00 1 Atelidae Ateles chamek 6000.00 1 Atelidae Ateles paniscus 7900.10 1 Atelidae Ateles geoffroyi 5284.90 1 Atelidae Ateles belzebuth 5000.00 1 Canidae Atelocynus microtis 7750.00 1 Muridae Auliscomys sublimis 38.00 1 Muridae Auliscomys pictus 54.70 1 Muridae Baiomys musculus 9.00 1 Muridae Baiomys taylori 8.00 1 Procyonidae Bassaricyon pauli 1200.00 1 Procyonidae Bassaricyon lasius 1200.00 1

20

Procyonidae Bassaricyon gabbii 1250.00 1 Procyonidae Bassaricyon beddardi 1235.00 1 Procyonidae Bassaricyon alleni 1235.00 1 Procyonidae Bassariscus sumichrasti 900.00 1 Procyonidae Bassariscus astutus 1129.50 1 Muridae Bibimys torresi 28.00 1 Muridae Bibimys chacoensis 28.00 1 Soricidae Blarina hylophaga 14.50 1 Soricidae Blarina carolinensis 13.50 1 Soricidae Blarina brevicauda 28.00 1 Muridae Blarinomys breviceps 36.80 1 Cervidae Blastocerus dichotomus 86666.20 1 Muridae Bolomys urichi 34.00 1 Muridae Bolomys punctulatus 37.30 1 Muridae Bolomys amoenus 29.00 1 Leporidae Brachylagus idahoensis 421.30 1 Atelidae Brachyteles arachnoides 13499.90 1 Bradypodidae Bradypus torquatus 3900.00 1 Bradypodidae Bradypus variegatus 4335.00 1 Bradypodidae Bradypus tridactylus 3850.00 1 Muridae Brucepattersonius iheringi 43.00 1 Muridae Brucepattersonius paradisus 42.00 2 Muridae Brucepattersonius misionensis 34.00 2 Muridae Brucepattersonius igniventris 35.00 2 Muridae Brucepattersonius guarani 32.00 2 Muridae Brucepattersonius griserufescens 27.00 2 Muridae Brucepattersonius albinasus 20.00 2 Dasypodidae Cabassous tatouay 5350.00 1 Dasypodidae Cabassous chacoensis 1490.00 1 Dasypodidae Cabassous unicinctus 4800.00 1 Dasypodidae Cabassous centralis 3810.00 1 Pitheciidae Cacajao melanocephalus 3800.00 1 Pitheciidae Cacajao calvus 5796.00 1 Caenolestidae Caenolestes convelatus 40.00 1 Caenolestidae Caenolestes caniventer 40.00 1 Caenolestidae Caenolestes fuliginosus 27.80 1 Caenolestidae Caenolestes condorensis 48.00 23 Callitrichidae Callibella humilis 185.00 26 Pitheciidae Callicebus personatus 1350.00 1 Pitheciidae Callicebus olallae 992.40 1 Pitheciidae Callicebus oenanthe 992.40 1 Pitheciidae Callicebus modestus 992.40 1 Pitheciidae Callicebus hoffmannsi 992.40 1 Pitheciidae Callicebus dubius 992.40 1 Pitheciidae Callicebus donacophilus 795.00 1 Pitheciidae Callicebus cupreus 915.00 1 Pitheciidae Callicebus cinerascens 992.40 1 Pitheciidae Callicebus caligatus 992.40 1 Pitheciidae Callicebus brunneus 992.40 1 Pitheciidae Callicebus torquatus 1050.00 1 Pitheciidae Callicebus moloch 854.70 1

21

Pitheciidae Callicebus stephennashi 780.00 22 Pitheciidae Callicebus discolor 915.00 22 Pitheciidae Callicebus bernhardi 1200.00 22 Callitrichidae Callimico goeldii 480.00 1 Echimyidae Callistomys pictus 519.00 1 Callitrichidae Callithrix penicillata 342.00 1 Callitrichidae Callithrix kuhlii 342.00 1 Callitrichidae Callithrix jacchus 292.00 1 Callitrichidae Callithrix geoffroyi 342.00 1 Callitrichidae Callithrix flaviceps 342.00 1 Callitrichidae Callithrix aurita 342.00 1 Muridae Calomys tener 13.80 1 Muridae Calomys laucha 14.00 1 Muridae Calomys hummelincki 27.00 1 Muridae Calomys callidus 27.00 1 Muridae Calomys boliviae 27.00 1 Muridae Calomys sorellus 20.00 1 Muridae Calomys musculinus 20.10 1 Muridae Calomys lepidus 26.60 1 Muridae Calomys callosus 45.00 1 Caluromyidae Caluromys philander 229.30 1 Caluromyidae Caluromys lanatus 325.00 1 Caluromyidae Caluromys derbianus 297.00 1 Caluromyidae Caluromysiops irrupta 250.00 1 Canidae Canis lupus 42750.00 1 Canidae Canis latrans 13406.30 1 Echimyidae Carterodon sulcidens 113.80 1 Castoridae Castor canadensis 21820.00 1 Tayassuidae Catagonus wagneri 35566.40 1 Caviidae Cavia tschudii 1000.00 1 Caviidae Cavia porcellus 728.00 1 Caviidae Cavia magna 460.00 1 Caviidae Cavia aperea 549.00 1 Caviidae Cavia fulgida 282.50 1 Callitrichidae Cebuella pygmaea 125.00 1 Cebidae Cebus apella 2500.00 1 Cebidae Cebus olivaceus 2600.00 1 Cebidae Cebus capucinus 2733.30 1 Cebidae Cebus albifrons 2629.00 1 Cebidae Cebus libidinosus 3200.00 4 Cebidae Cebus xanthosternos 3200.00 4 Cebidae Cebus robustus 3200.00 4 Cebidae Cebus nigritus 3200.00 4 Cebidae Cebus macrocephalus 3200.00 4 Cebidae Cebus kaapori 2600.00 5 Canidae Cerdocyon thous 5240.00 1 Cervidae Cervus canadensis 217750.90 1 Heteromyidae Chaetodipus pernix 17.00 1 Heteromyidae Chaetodipus lineatus 23.00 1 Heteromyidae Chaetodipus goldmani 23.00 1 Heteromyidae Chaetodipus arenarius 23.00 1

22

Heteromyidae Chaetodipus spinatus 16.40 1 Heteromyidae Chaetodipus penicillatus 15.00 1 Heteromyidae Chaetodipus nelsoni 15.70 1 Heteromyidae Chaetodipus intermedius 16.50 1 Heteromyidae Chaetodipus hispidus 32.00 1 Heteromyidae Chaetodipus formosus 19.50 1 Heteromyidae Chaetodipus fallax 18.70 1 Heteromyidae Chaetodipus californicus 22.00 1 Heteromyidae Chaetodipus baileyi 26.30 1 Erethizontidae Chaetomys subspinosus 1300.00 1 Dasypodidae Chaetophractus nationi 2150.00 1 Dasypodidae Chaetophractus villosus 4540.00 1 Dasypodidae Chaetophractus vellerosus 1030.00 1 Muridae Chelemys macronyx 73.30 1 Muridae Chelemys megalonyx 50.80 1 Muridae Chibchanomys trichotis 50.00 1 Muridae Chibchanomys orcesi 27.00 28 Muridae Chilomys instans 19.00 1 Chinchillidae Chinchilla brevicaudata 500.00 1 Chinchillidae Chinchilla lanigera 485.00 1 Muridae Chinchillula sahamae 169.80 1 Didelphidae Chironectes minimus 946.00 1 Pitheciidae Chiropotes albinasus 2900.00 1 Pitheciidae Chiropotes satanas 3000.00 1 Dasypodidae Chlamyphorus truncatus 44.30 1 Dasypodidae Chlamyphorus retusus 130.00 1 Megalonychidae Choloepus hoffmanni 6000.00 1 Megalonychidae Choloepus didactylus 5160.00 1 Muridae Chroeomys jelskii 34.50 1 Muridae Chroeomys andinus 18.00 1 Canidae Chrysocyon brachyurus 23249.80 1 Muridae Clethrionomys rutilus 29.00 1 Muridae Clethrionomys gapperi 19.00 1 Muridae Clethrionomys californicus 18.30 1 Echimyidae Clyomys laticeps 201.00 1 Echimyidae Clyomys bishopi 30.00 1 Erethizontidae Coendou mexicanus 2000.00 1 Erethizontidae Coendou villosus 1750.00 1 Erethizontidae Coendou vestitus 736.00 1 Erethizontidae Coendou spinosus 750.80 1 Erethizontidae Coendou rothschildi 2000.00 1 Erethizontidae Coendou bicolor 4050.00 1 Erethizontidae Coendou prehensilis 4400.00 1 Erethizontidae Coendou insidiosus 1000.00 1 Erethizontidae Coendou melanurus 2050.00 16 Talpidae Condylura cristata 84.00 1 Mephitidae Conepatus leuconotus 3500.00 1 Mephitidae Conepatus humboldtii 328.00 1 Mephitidae Conepatus mesoleucus 2010.00 1 Mephitidae Conepatus semistriatus 1200.00 1 Mephitidae Conepatus chinga 1917.50 1

23

Geomyidae Cratogeomys castanops 251.80 1 Soricidae Cryptotis parva 5.00 1 Soricidae Cryptotis nigrescens 7.00 1 Soricidae Cryptotis mexicana 7.00 1 Soricidae Cryptotis magna 7.00 1 Soricidae Cryptotis goodwini 7.00 1 Soricidae Cryptotis goldmani 8.00 1 Soricidae Cryptotis squamipes 11.00 1 Soricidae Cryptotis montivaga 11.30 1 Soricidae Cryptotis meridensis 12.00 1 Soricidae Cryptotis thomasi 12.60 1 Ctenomyidae Ctenomys validus 232.60 1 Ctenomyidae Ctenomys tuconax 249.00 1 Ctenomyidae Ctenomys torquatus 209.50 1 Ctenomyidae Ctenomys steinbachi 385.00 1 Ctenomyidae Ctenomys sociabilis 400.00 1 Ctenomyidae Ctenomys sericeus 400.00 1 Ctenomyidae Ctenomys saltarius 230.00 1 Ctenomyidae Ctenomys porteousi 192.40 1 Ctenomyidae Ctenomys pontifex 400.00 1 Ctenomyidae Ctenomys perrensis 400.00 1 Ctenomyidae Ctenomys pearsoni 212.00 1 Ctenomyidae Ctenomys opimus 361.50 1 Ctenomyidae Ctenomys occultus 150.00 1 Ctenomyidae Ctenomys nattereri 400.00 1 Ctenomyidae Ctenomys minutus 92.00 1 Ctenomyidae Ctenomys mendocinus 177.50 1 Ctenomyidae Ctenomys lewisi 117.20 1 Ctenomyidae Ctenomys leucodon 243.60 1 Ctenomyidae Ctenomys latro 192.00 1 Ctenomyidae Ctenomys haigi 164.00 1 Ctenomyidae Ctenomys frater 172.90 1 Ctenomyidae Ctenomys emilianus 285.30 1 Ctenomyidae Ctenomys dorsalis 165.60 1 Ctenomyidae Ctenomys colburni 400.00 1 Ctenomyidae Ctenomys brasiliensis 400.00 1 Ctenomyidae Ctenomys bonettoi 202.30 1 Ctenomyidae Ctenomys boliviensis 535.00 1 Ctenomyidae Ctenomys azarae 400.00 1 Ctenomyidae Ctenomys argentinus 221.30 1 Ctenomyidae Ctenomys tucumanus 217.00 1 Ctenomyidae Ctenomys talarum 132.30 1 Ctenomyidae Ctenomys peruanus 490.00 1 Ctenomyidae Ctenomys maulinus 215.00 1 Ctenomyidae Ctenomys magellanicus 272.00 1 Ctenomyidae Ctenomys knighti 316.00 1 Ctenomyidae Ctenomys fulvus 262.00 1 Ctenomyidae Ctenomys conoveri 860.00 1 Ctenomyidae Ctenomys australis 361.50 1 Cuniculidae Cuniculus taczanowskii 8999.90 1 Cuniculidae Cuniculus paca 8000.00 1

24

Myrmecophagidae Cyclopes didactylus 329.50 1 Sciuridae Cynomys ludovicianus 1364.00 1 Sciuridae Cynomys parvidens 900.00 1 Sciuridae Cynomys mexicanus 900.00 1 Sciuridae Cynomys leucurus 908.50 1 Sciuridae Cynomys gunnisoni 925.00 1 Echimyidae Dactylomys peruanus 382.00 1 Echimyidae Dactylomys dactylinus 650.00 1 Echimyidae Dactylomys boliviensis 728.80 1 Dasyproctidae Dasyprocta mexicana 5000.00 1 Dasyproctidae Dasyprocta prymnolopha 2900.00 1 Dasyproctidae Dasyprocta kalinowskii 2650.00 1 Dasyproctidae Dasyprocta guamara 2650.00 1 Dasyproctidae Dasyprocta fuliginosa 3500.00 1 Dasyproctidae Dasyprocta punctata 2675.00 1 Dasyproctidae Dasyprocta leporina 3020.00 1 Dasyproctidae Dasyprocta azarae 2310.00 1 Dasypodidae Dasypus septemcinctus 1526.70 1 Dasypodidae Dasypus pilosus 4445.00 1 Dasypodidae Dasypus kappleri 9500.00 1 Dasypodidae Dasypus hybridus 1500.00 1 Dasypodidae Dasypus sabanicola 1150.00 1 Dasypodidae Dasypus novemcinctus 4203.80 1 Muridae Delomys sublineatus 90.00 1 Muridae Delomys dorsalis 67.50 1 Muridae Deltamys kempi 26.40 1 Muridae Dicrostonyx hudsonius 57.00 1 Muridae Dicrostonyx groenlandicus 54.40 1 Didelphidae Didelphis aurita 1164.00 1 Didelphidae Didelphis virginiana 2195.50 1 Didelphidae Didelphis marsupialis 1091.20 1 Didelphidae Didelphis albiventris 904.00 1 Dinomyidae Dinomys branickii 12500.00 1 Echimyidae Diplomys rufodorsalis 144.80 1 Echimyidae Diplomys labilis 227.50 1 Echimyidae Diplomys caniceps 394.50 1 Heteromyidae Dipodomys merriami 42.00 1 Heteromyidae Dipodomys venustus 72.00 1 Heteromyidae Dipodomys phillipsii 41.00 1 Heteromyidae Dipodomys heermanni 72.00 1 Heteromyidae Dipodomys stephensi 69.80 1 Heteromyidae Dipodomys spectabilis 135.90 1 Heteromyidae Dipodomys panamintinus 74.70 1 Heteromyidae Dipodomys ordii 60.40 1 Heteromyidae Dipodomys nitratoides 43.90 1 Heteromyidae Dipodomys nelsoni 88.60 1 Heteromyidae Dipodomys microps 54.60 1 Heteromyidae Dipodomys ingens 133.90 1 Heteromyidae Dipodomys gravipes 84.00 1 Heteromyidae Dipodomys elator 77.50 1 Heteromyidae Dipodomys deserti 104.50 1

25

Heteromyidae Dipodomys agilis 57.80 1 Caviidae Dolichotis salinicola 1600.00 1 Caviidae Dolichotis patagonum 8000.00 1 Microbiotheriidae Dromiciops gliroides 22.30 1 Echimyidae Echimys saturnus 626.90 1 Echimyidae Echimys rhipidurus 315.00 1 Echimyidae Echimys grandis 584.00 1 Echimyidae Echimys chrysurus 652.50 1 Echimyidae Echimys semivillosus 200.00 1 Erethizontidae Echinoprocta rufescens 831.80 1 Mustelidae Eira barbara 3910.00 1 Muridae Eligmodontia puerulus 28.50 1 Muridae Eligmodontia morgani 16.50 1 Muridae Eligmodontia moreni 18.00 1 Muridae Eligmodontia typus 17.30 1 Erethizontidae Erethizon dorsatum 7085.30 1 Muridae Euneomys petersoni 83.00 1 Muridae Euneomys mordax 82.00 1 Muridae Euneomys fossor 83.00 1 Muridae Euneomys chinchilloides 87.60 1 Dasypodidae Euphractus sexcinctus 4782.90 1 Echimyidae Euryzygomatomys spinosus 187.50 1 Caviidae Galea spixii 326.20 1 Caviidae Galea flavidens 450.00 1 Caviidae Galea musteloides 480.00 1 Muridae Galenomys garleppi 59.30 1 Mustelidae Galictis vittata 3200.00 1 Mustelidae Galictis cuja 1000.00 1 Geomyidae Geomys pinetis 85.00 1 Geomyidae Geomys tropicalis 350.00 1 Geomyidae Geomys personatus 397.00 1 Geomyidae Geomys bursarius 203.80 1 Geomyidae Geomys arenarius 206.00 1 Muridae Geoxus valdivianus 31.50 1 Sciuridae Glaucomys volans 63.90 1 Sciuridae Glaucomys sabrinus 166.00 1 Marmosidae Gracilinanus microtarsus 31.00 1 Marmosidae Gracilinanus marica 23.80 1 Marmosidae Gracilinanus emiliae 7.60 1 Marmosidae Gracilinanus dryas 18.00 1 Marmosidae Gracilinanus agilis 22.00 1 Marmosidae Gracilinanus aceramarcae 20.50 1 Muridae Graomys edithae 40.40 1 Muridae Graomys griseoflavus 67.50 1 Muridae Graomys domorum 102.00 1 Mustelidae Gulo gulo 14525.10 1 Muridae Habromys simulatus 40.00 1 Muridae Habromys lophurus 40.00 1 Muridae Habromys lepturus 85.00 1 Muridae Habromys chinanteco 40.00 1 Muridae Handleyomys fuscatus 49.50 1

26

Felidae Herpailurus yaguarondi 8999.90 1 Heteromyidae Heteromys oresterus 74.80 1 Heteromyidae Heteromys goldmani 85.00 1 Heteromyidae Heteromys australis 267.50 1 Heteromyidae Heteromys anomalus 70.00 1 Heteromyidae Heteromys nelsoni 54.20 1 Heteromyidae Heteromys gaumeri 63.60 1 Heteromyidae Heteromys desmarestianus 72.50 1 Cervidae Hippocamelus bisulcus 70000.30 1 Cervidae Hippocamelus antisensis 68599.30 1 Muridae Hodomys alleni 367.60 1 Muridae Holochilus sciureus 163.50 1 Muridae Holochilus chacarius 204.00 1 Muridae Holochilus brasiliensis 155.00 1 Echimyidae Hoplomys gymnurus 240.00 1 Hydrochaeridae Hydrochaeris hydrochaeris 62449.60 1 Marmosidae Hyladelphys kalinowskii 15.50 21 Muridae Ichthyomys tweedii 118.50 1 Muridae Ichthyomys stolzmanni 84.70 1 Muridae Ichthyomys pittieri 69.10 1 Muridae Ichthyomys hydrobates 66.40 1 Muridae Irenomys tarsalis 43.10 1 Echimyidae Isothrix pagurus 210.00 1 Echimyidae Isothrix bistriata 445.00 1 Echimyidae Isothrix sinnamariensis 260.00 14 Muridae Isthmomys pirrensis 138.00 1 Muridae Juliomys pictipes 22.90 1 Echimyidae Kannabateomys amblyonyx 600.00 1 Caviidae Kerodon rupestris 800.00 1 Muridae Kunsia tomentosus 115.60 1 Muridae Kunsia fronto 168.00 1 Chinchillidae Lagidium wolffsohni 2682.00 1 Chinchillidae Lagidium viscacia 1540.00 1 Chinchillidae Lagidium peruanum 1220.00 1 Chinchillidae Lagostomus maximus 4647.50 1 Atelidae Lagothrix lagothricha 6300.00 1 Camelidae Lama guanicoe 120000.00 1 Muridae Lemmiscus curtatus 28.30 1 Muridae Lenoxus apicalis 53.60 1 Callitrichidae Leontopithecus rosalia 535.50 1 Callitrichidae Leontopithecus chrysopygus 700.00 1 Callitrichidae Leontopithecus chrysomelas 700.00 1 Felidae Leopardus tigrinus 2250.00 1 Felidae Leopardus wiedii 3250.00 1 Felidae Leopardus pardalis 11900.10 1 Leporidae Lepus flavigularis 3000.00 1 Leporidae Lepus townsendii 1555.00 1 Leporidae Lepus othus 4806.00 1 Leporidae Lepus callotis 2500.00 1 Leporidae Lepus californicus 2422.50 1 Leporidae Lepus arcticus 4405.00 1

27

Leporidae Lepus americanus 1710.00 1 Leporidae Lepus alleni 3685.00 1 Marmosidae Lestodelphys halli 76.00 1 Caenolestidae Lestoros inca 21.10 1 Heteromyidae Liomys spectabilis 65.00 1 Heteromyidae Liomys irroratus 50.00 1 Heteromyidae Liomys adspersus 65.00 1 Heteromyidae Liomys salvini 42.00 1 Heteromyidae Liomys pictus 40.00 1 Echimyidae Lonchothrix emiliae 138.20 1 Mustelidae Lontra felina 30600.00 1 Mustelidae Lontra provocax 7500.00 1 Mustelidae Lontra longicaudis 6555.00 1 Mustelidae Lontra canadensis 8087.40 1 Muridae Loxodontomys micropus 72.70 1 Muridae Loxodontomys pikumche 43.00 24 Didelphidae Lutreolina crassicaudata 537.30 1 Felidae Lynchailurus colocolo 3935.00 1 Mustelidae Lyncodon patagonicus 225.00 1 Felidae Lynx rufus 8904.10 1 Felidae Lynx canadensis 9373.20 1 Echimyidae Makalata macrura 439.00 1 Echimyidae Makalata didelphoides 108.00 1 Marmosidae Marmosa canescens 60.00 1 Marmosidae Marmosa xerophila 46.20 1 Marmosidae Marmosa tyleriana 32.40 1 Marmosidae Marmosa rubra 63.00 1 Marmosidae Marmosa murina 26.00 1 Marmosidae Marmosa andersoni 47.40 1 Marmosidae Marmosa robinsoni 56.70 1 Marmosidae Marmosa mexicana 49.50 1 Marmosidae Marmosa lepida 14.00 1 Marmosidae Marmosops noctivagus 21.00 1 Marmosidae Marmosops invictus 29.20 1 Marmosidae Marmosops impavidus 40.50 1 Marmosidae Marmosops handleyi 30.70 1 Marmosidae Marmosops cracens 25.50 1 Marmosidae Marmosops dorothea 37.50 1 Marmosidae Marmosops parvidens 15.00 1 Marmosidae Marmosops incanus 62.30 1 Marmosidae Marmosops fuscatus 60.00 1 Sciuridae Marmota olympus 6300.00 1 Sciuridae Marmota flaviventris 3350.00 1 Sciuridae Marmota monax 3801.70 1 Sciuridae Marmota caligata 7230.00 1 Mustelidae Martes pennanti 4000.00 1 Mustelidae Martes americana 1250.00 1 Cervidae Mazama nana 16499.90 1 Cervidae Mazama gouazoubira 16300.10 1 Cervidae Mazama chunyi 16499.90 1 Cervidae Mazama bricenii 16499.90 1

28

Cervidae Mazama rufina 25999.80 1 Cervidae Mazama americana 22799.70 1 Muridae Megadontomys thomasi 111.00 1 Muridae Megadontomys cryophilus 32.00 6 Soricidae Megasorex gigas 20.00 1 Muridae Melanomys zunigae 53.50 1 Muridae Melanomys robustulus 53.50 1 Muridae Melanomys caliginosus 41.00 1 Mephitidae Mephitis mephitis 2085.00 1 Mephitidae Mephitis macroura 801.30 1 Echimyidae Mesomys stimulax 108.00 1 Echimyidae Mesomys leniceps 108.00 1 Echimyidae Mesomys hispidus 175.00 1 Marmosidae Metachirus nudicaudatus 375.00 1 Callitrichidae Mico humeralifer 350.00 1 Callitrichidae Mico argentatus 440.00 1 Callitrichidae Mico nigriceps 370.00 15 Callitrichidae Mico saterei 441.25 15 Marmosidae Micoureus regina 118.60 1 Marmosidae Micoureus demerarae 75.80 1 Marmosidae Micoureus constantiae 90.00 1 Marmosidae Micoureus alstoni 132.30 1 Caviidae Microcavia shiptoni 185.00 1 Caviidae Microcavia niata 255.20 1 Caviidae Microcavia australis 286.10 1 Heteromyidae Microdipodops pallidus 12.50 1 Heteromyidae Microdipodops megacephalus 10.50 1 Muridae Microryzomys minutus 13.50 1 Muridae Microryzomys altissimus 13.50 1 Sciuridae Microsciurus santanderensis 99.80 1 Sciuridae Microsciurus mimulus 120.00 1 Sciuridae Microsciurus flaviventer 92.00 1 Sciuridae Microsciurus alfari 87.50 1 Muridae Microtus richardsoni 85.00 1 Muridae Microtus umbrosus 42.00 1 Muridae Microtus quasiater 40.00 1 Muridae Microtus miurus 41.00 1 Muridae Microtus guatemalensis 42.00 1 Muridae Microtus xanthognathus 125.80 1 Muridae Microtus townsendii 64.80 1 Muridae Microtus pinetorum 26.30 1 Muridae Microtus pennsylvanicus 36.80 1 Muridae Microtus oregoni 20.30 1 Muridae Microtus oeconomus 32.20 1 Muridae Microtus ochrogaster 38.00 1 Muridae Microtus oaxacensis 39.20 1 Muridae Microtus montanus 36.30 1 Muridae Microtus mexicanus 35.00 1 Muridae Microtus longicaudus 46.70 1 Muridae Microtus chrotorrhinus 39.00 1 Muridae Microtus canicaudus 28.40 1

29

Muridae Microtus californicus 57.40 1 Marmosidae Monodelphis unistriata 55.30 1 Marmosidae Monodelphis theresa 112.00 1 Marmosidae Monodelphis sorex 48.00 1 Marmosidae Monodelphis scalops 741.00 1 Marmosidae Monodelphis rubida 45.50 1 Marmosidae Monodelphis osgoodi 112.00 1 Marmosidae Monodelphis kunsi 11.30 1 Marmosidae Monodelphis iheringi 112.00 1 Marmosidae Monodelphis domestica 71.40 1 Marmosidae Monodelphis adusta 35.00 1 Marmosidae Monodelphis emiliae 40.00 1 Marmosidae Monodelphis dimidiata 58.00 1 Marmosidae Monodelphis brevicaudata 69.60 1 Marmosidae Monodelphis americana 19.50 1 Mustelidae Mustela felipei 211.30 1 Mustelidae Mustela africana 537.00 1 Mustelidae Mustela vison 945.00 1 Mustelidae Mustela nivalis 46.90 1 Mustelidae Mustela frenata 147.00 1 Mustelidae Mustela erminea 70.00 1 Myocastoridae Myocastor coypus 6937.50 1 Dasyproctidae Myoprocta exilis 1280.00 1 Dasyproctidae Myoprocta acouchy 600.00 1 Dasyproctidae Myoprocta pratti 908.00 9 Myrmecophagidae Myrmecophaga tridactyla 22333.20 1 Dipodidae Napaeozapus insignis 22.30 1 Procyonidae Nasua nasua 3793.80 1 Procyonidae Nasua narica 4030.00 1 Procyonidae Nasuella olivacea 1340.00 1 Muridae Neacomys spinosus 19.00 1 Muridae Neacomys guianae 14.20 1 Muridae Neacomys tenuipes 19.00 1 Muridae Neacomys paracou 14.00 21 Muridae Neacomys dubosti 14.00 21 Muridae Necromys temchuki 47.20 1 Muridae Necromys obscurus 40.70 1 Muridae Necromys lactens 32.90 1 Muridae Necromys lasiurus 39.90 1 Muridae Nectomys parvipes 248.80 1 Muridae Nectomys squamipes 190.70 1 Muridae Nelsonia neotomodon 80.00 1 Muridae Nelsonia goldmani 28.50 6 Muridae Neofiber alleni 266.00 1 Sciuridae Neotamias rufus 53.60 1 Sciuridae Neotamias canipes 70.40 1 Sciuridae Neotamias sonomae 75.00 1 Sciuridae Neotamias siskiyou 75.00 1 Sciuridae Neotamias quadrimaculatus 85.20 1 Sciuridae Neotamias umbrinus 63.00 1 Sciuridae Neotamias townsendii 74.80 1

30

Sciuridae Neotamias striatus 111.90 1 Sciuridae Neotamias speciosus 62.00 1 Sciuridae Neotamias senex 89.30 1 Sciuridae Neotamias ruficaudus 68.00 1 Sciuridae Neotamias quadrivittatus 62.20 1 Sciuridae Neotamias panamintinus 52.30 1 Sciuridae Neotamias palmeri 69.40 1 Sciuridae Neotamias ochrogenys 91.70 1 Sciuridae Neotamias obscurus 73.00 1 Sciuridae Neotamias minimus 135.30 1 Sciuridae Neotamias merriami 74.80 1 Sciuridae Neotamias dorsalis 71.10 1 Sciuridae Neotamias cinereicollis 61.70 1 Sciuridae Neotamias amoenus 50.50 1 Sciuridae Neotamias alpinus 36.60 1 Muridae Neotoma palatina 198.00 1 Muridae Neotoma nelsoni 198.00 1 Muridae Neotoma goldmani 198.00 1 Muridae Neotoma devia 200.00 1 Muridae Neotoma cinerea 299.20 1 Muridae Neotoma angustapalata 198.00 1 Muridae Neotoma stephensi 152.50 1 Muridae Neotoma phenax 227.50 1 Muridae Neotoma micropus 237.50 1 Muridae Neotoma mexicana 203.00 1 Muridae Neotoma lepida 163.70 1 Muridae Neotoma fuscipes 229.80 1 Muridae Neotoma floridana 244.70 1 Muridae Neotoma albigula 206.00 1 Muridae Neotoma magister 280.00 10 Muridae Neotomodon alstoni 40.00 1 Muridae Neotomys ebriosus 64.50 1 Talpidae Neurotrichus gibbsii 8.90 1 Muridae Neusticomys venezuelae 47.10 1 Muridae Neusticomys peruviensis 40.00 1 Muridae Neusticomys mussoi 40.00 1 Muridae Neusticomys monticolus 39.50 1 Muridae Neusticomys oyapocki 4.70 1 Muridae Notiomys edwardsii 21.30 1 Soricidae Notiosorex crawfordi 4.40 1 Muridae Nyctomys sumichrasti 60.00 1 Ochotonidae Ochotona princeps 157.60 1 Ochotonidae Ochotona collaris 129.00 1 Muridae Ochrotomys nuttalli 22.40 1 Octodontidae Octodon lunatus 233.00 1 Octodontidae Octodon degus 210.00 1 Octodontidae Octodon bridgesi 150.00 1 Octodontidae Octodontomys gliroides 150.00 1 Octodontidae Octomys mimax 131.00 1 Cervidae Odocoileus virginianus 55508.60 1 Cervidae Odocoileus hemionus 54212.60 1

31

Muridae Oecomys trinitatis 73.40 1 Muridae Oecomys superans 73.40 1 Muridae Oecomys speciosus 73.40 1 Muridae Oecomys rutilus 73.40 1 Muridae Oecomys roberti 73.40 1 Muridae Oecomys rex 73.40 1 Muridae Oecomys phaeotis 73.40 1 Muridae Oecomys paricola 73.40 1 Muridae Oecomys mamorae 62.50 1 Muridae Oecomys flavicans 73.40 1 Muridae Oecomys cleberi 73.40 1 Muridae Oecomys concolor 61.60 1 Muridae Oecomys bicolor 34.00 1 Echimyidae Olallamys edax 206.40 1 Echimyidae Olallamys albicauda 273.50 1 Muridae Oligoryzomys magellanicus 25.20 1 Muridae Oligoryzomys longicaudatus 27.00 1 Muridae Oligoryzomys griseolus 25.20 1 Muridae Oligoryzomys eliurus 30.00 1 Muridae Oligoryzomys destructor 25.20 1 Muridae Oligoryzomys delticola 29.40 1 Muridae Oligoryzomys arenalis 25.20 1 Muridae Oligoryzomys andinus 25.20 1 Muridae Oligoryzomys nigripes 20.50 1 Muridae Oligoryzomys microtis 22.50 1 Muridae Oligoryzomys fulvescens 25.00 1 Muridae Oligoryzomys flavescens 21.30 1 Muridae Oligoryzomys chacoensis 23.00 1 Muridae Oligoryzomys stramineus 25.00 20 (L) Felidae Oncifelis guigna 2230.00 1 Felidae Oncifelis geoffroyi 3590.00 1 Muridae Ondatra zibethicus 981.50 1 Muridae Onychomys arenicola 30.00 1 Muridae Onychomys torridus 25.00 1 Muridae Onychomys leucogaster 27.90 1 Felidae Oreailurus jacobita 9170.00 1 Bovidae Oreamnos americanus 72500.30 1 Atelidae Oreonax flavicauda 6800.00 1 Geomyidae Orthogeomys underwoodi 250.00 1 Geomyidae Orthogeomys heterodus 615.00 1 Geomyidae Orthogeomys cavator 650.00 1 Geomyidae Orthogeomys lanius 500.00 1 Geomyidae Orthogeomys hispidus 500.00 1 Geomyidae Orthogeomys grandis 500.00 1 Geomyidae Orthogeomys cuniculus 500.00 1 Geomyidae Orthogeomys dariensis 438.00 1 Muridae Oryzomys melanotis 50.00 1 Muridae Oryzomys intectus 60.50 1 Muridae Oryzomys yunganus 60.50 1 Muridae Oryzomys xantheolus 79.80 1 Muridae Oryzomys talamancae 55.00 1

32

Muridae Oryzomys subflavus 50.00 1 Muridae Oryzomys polius 60.50 1 Muridae Oryzomys nitidus 55.20 1 Muridae Oryzomys macconnelli 58.00 1 Muridae Oryzomys levipes 60.50 1 Muridae Oryzomys keaysi 58.30 1 Muridae Oryzomys hammondi 60.50 1 Muridae Oryzomys gorgasi 60.50 1 Muridae Oryzomys bolivaris 60.50 1 Muridae Oryzomys balneator 60.50 1 Muridae Oryzomys auriventer 60.50 1 Muridae Oryzomys albigularis 60.50 1 Muridae Oryzomys perenensis 57.80 1 Muridae Oryzomys palustris 53.90 1 Muridae Oryzomys megacephalus 57.80 1 Muridae Oryzomys couesi 69.30 1 Muridae Oryzomys chapmani 50.00 1 Muridae Oryzomys alfaroi 33.30 1 Muridae Oryzomys seuanezi 75.73 11 Muridae Oryzomys russatus 91.00 13 Muridae Osgoodomys banderanus 50.00 1 Muridae Otonyctomys hatti 36.20 1 Muridae Ototylomys phyllotis 120.00 1 Bovidae Ovibos moschatus 368502.10 1 Bovidae Ovis dalli 55650.60 1 Bovidae Ovis canadensis 74644.90 1 Muridae Oxymycterus rufus 75.40 1 Muridae Oxymycterus nasutus 68.00 1 Muridae Oxymycterus hucucha 68.00 1 Muridae Oxymycterus hispidus 36.80 1 Muridae Oxymycterus hiska 68.00 1 Muridae Oxymycterus delator 81.50 1 Muridae Oxymycterus angularis 68.00 1 Muridae Oxymycterus akodontius 68.00 1 Muridae Oxymycterus roberti 83.40 1 Muridae Oxymycterus paramensis 42.00 1 Muridae Oxymycterus incae 35.00 1 Cervidae Ozotoceros bezoarticus 40000.00 1 Felidae Panthera onca 100000.00 1 Geomyidae Pappogeomys zinseri 150.00 1 Geomyidae Pappogeomys neglectus 150.00 1 Geomyidae Pappogeomys gymnurus 600.00 1 Geomyidae Pappogeomys fumosus 150.00 1 Geomyidae Pappogeomys bulleri 150.00 1 Geomyidae Pappogeomys alcorni 150.00 1 Geomyidae Pappogeomys tylorhinus 249.00 1 Geomyidae Pappogeomys merriami 420.00 1 Talpidae Parascalops breweri 51.00 1 Muridae Pearsonomys annectens 45.83 12 Tayassuidae Pecari tajacu 21266.70 1 Heteromyidae Perognathus alticolus 24.00 1

33

Heteromyidae Perognathus parvus 21.80 1 Heteromyidae Perognathus merriami 6.80 1 Heteromyidae Perognathus longimembris 7.60 1 Heteromyidae Perognathus inornatus 10.90 1 Heteromyidae Perognathus flavus 7.70 1 Heteromyidae Perognathus flavescens 8.80 1 Heteromyidae Perognathus fasciatus 11.50 1 Heteromyidae Perognathus amplus 11.70 1 Muridae Peromyscus stirtoni 29.20 1 Muridae Peromyscus merriami 40.00 1 Muridae Peromyscus zarhynchus 40.00 1 Muridae Peromyscus winkelmanni 40.00 1 Muridae Peromyscus spicilegus 36.00 1 Muridae Peromyscus simulus 40.00 1 Muridae Peromyscus polius 40.00 1 Muridae Peromyscus perfulvus 40.00 1 Muridae Peromyscus pectoralis 39.00 1 Muridae Peromyscus ochraventer 40.00 1 Muridae Peromyscus melanurus 40.00 1 Muridae Peromyscus melanophrys 40.00 1 Muridae Peromyscus mekisturus 60.00 1 Muridae Peromyscus hooperi 36.00 1 Muridae Peromyscus gymnotis 40.00 1 Muridae Peromyscus guatemalensis 40.00 1 Muridae Peromyscus grandis 71.00 1 Muridae Peromyscus furvus 33.00 1 Muridae Peromyscus eva 22.00 1 Muridae Peromyscus bullatus 40.00 1 Muridae Peromyscus aztecus 40.00 1 Muridae Peromyscus yucatanicus 26.30 1 Muridae Peromyscus truei 27.40 1 Muridae Peromyscus polionotus 13.00 1 Muridae Peromyscus mexicanus 32.60 1 Muridae Peromyscus melanotis 39.60 1 Muridae Peromyscus melanocarpus 59.00 1 Muridae Peromyscus megalops 66.20 1 Muridae Peromyscus maniculatus 21.30 1 Muridae Peromyscus leucopus 21.20 1 Muridae Peromyscus gratus 27.40 1 Muridae Peromyscus gossypinus 29.40 1 Muridae Peromyscus eremicus 23.60 1 Muridae Peromyscus difficilis 28.00 1 Muridae Peromyscus crinitus 18.00 1 Muridae Peromyscus californicus 43.50 1 Muridae Peromyscus boylii 21.40 1 Muridae Peromyscus attwateri 27.90 1 Muridae Peromyscus nasutus 28.00 3 Muridae Peromyscus keeni 21.50 17 Muridae Phaenomys ferrugineus 93.80 1 Muridae Phenacomys intermedius 25.20 1 Muridae Phenacomys ungava 32.50 3

34

Didelphidae Philander andersoni 325.00 1 Didelphidae Philander opossum 750.00 1 Echimyidae Phyllomys thomasi 560.20 1 Echimyidae Phyllomys nigrispinus 224.30 1 Echimyidae Phyllomys lamarum 215.20 1 Echimyidae Phyllomys dasythrix 260.00 1 Echimyidae Phyllomys brasiliensis 312.50 1 Echimyidae Phyllomys blainvilii 243.30 1 Muridae Phyllotis wolffsohni 42.50 1 Muridae Phyllotis osilae 49.00 1 Muridae Phyllotis osgoodi 45.10 1 Muridae Phyllotis magister 68.50 1 Muridae Phyllotis haggardi 42.50 1 Muridae Phyllotis gerbillus 17.40 1 Muridae Phyllotis definitus 89.00 1 Muridae Phyllotis caprinus 50.80 1 Muridae Phyllotis bonaeriensis 42.50 1 Muridae Phyllotis amicus 20.20 1 Muridae Phyllotis xanthopygus 56.30 1 Muridae Phyllotis darwini 50.80 1 Muridae Phyllotis andium 53.00 1 Octodontidae Pipanacoctomys aureus 139.35 19 (L) Pitheciidae Pithecia irrorata 2241.00 1 Pitheciidae Pithecia albicans 2800.00 1 Pitheciidae Pithecia aequatorialis 2250.00 1 Pitheciidae Pithecia pithecia 1375.50 1 Pitheciidae Pithecia monachus 1537.50 1 Muridae Podomys floridanus 30.80 1 Procyonidae Potos flavus 3000.00 1 Dasypodidae Priodontes maximus 45359.70 1 Procyonidae Procyon lotor 5525.00 1 Procyonidae Procyon cancrivorus 6949.90 1 Echimyidae Proechimys warreni 285.00 1 Echimyidae Proechimys urichi 285.00 1 Echimyidae Proechimys steerei 285.00 1 Echimyidae Proechimys simonsi 285.00 1 Echimyidae Proechimys quadruplicatus 285.00 1 Echimyidae Proechimys poliopus 285.00 1 Echimyidae Proechimys oris 285.00 1 Echimyidae Proechimys oconnelli 285.00 1 Echimyidae Proechimys myosuros 285.00 1 Echimyidae Proechimys mincae 285.00 1 Echimyidae Proechimys magdalenae 285.00 1 Echimyidae Proechimys longicaudatus 205.00 1 Echimyidae Proechimys hoplomyoides 285.00 1 Echimyidae Proechimys hendeei 285.00 1 Echimyidae Proechimys gularis 285.00 1 Echimyidae Proechimys goeldii 285.00 1 Echimyidae Proechimys decumanus 285.00 1 Echimyidae Proechimys chrysaeolus 285.00 1 Echimyidae Proechimys canicollis 285.00 1

35

Echimyidae Proechimys brevicauda 285.00 1 Echimyidae Proechimys bolivianus 285.00 1 Echimyidae Proechimys amphichoricus 285.00 1 Echimyidae Proechimys semispinosus 360.50 1 Echimyidae Proechimys guairae 400.00 1 Echimyidae Proechimys cuvieri 330.00 1 Echimyidae Proechimys cayennensis 316.00 1 Canidae Pseudalopex vetulus 5350.00 1 Canidae Pseudalopex sechurae 4000.00 1 Canidae Pseudalopex gymnocercus 4690.00 1 Canidae Pseudalopex griseus 8280.00 1 Canidae Pseudalopex culpaeus 9832.40 1 Muridae Pseudoryzomys simplex 51.20 1 Mustelidae Pteronura brasiliensis 23999.90 1 Cervidae Pudu puda 9749.90 1 Cervidae Pudu mephistophiles 9600.00 1 Felidae Puma concolor 51600.00 1 Muridae Punomys lemminus 84.80 1 Cervidae Rangifer tarandus 86034.00 1 Muridae Reithrodon auritus 70.90 1 Muridae Reithrodontomys sumichrasti 19.00 1 Muridae Reithrodontomys hirsutus 20.00 1 Muridae Reithrodontomys chrysopsis 19.00 1 Muridae Reithrodontomys burti 20.00 1 Muridae Reithrodontomys raviventris 11.00 1 Muridae Reithrodontomys montanus 10.90 1 Muridae Reithrodontomys megalotis 9.40 1 Muridae Reithrodontomys humulis 8.30 1 Muridae Reithrodontomys fulvescens 11.40 1 Muridae Rhagomys rufescens 21.20 1 Muridae Rhagomys longilingua 30.00 25 Muridae Rheomys thomasi 40.00 1 Muridae Rheomys mexicanus 40.00 1 Muridae Rheomys raptor 38.00 1 Muridae Rhipidomys wetzeli 89.00 1 Muridae Rhipidomys venustus 89.00 1 Muridae Rhipidomys venezuelae 90.00 1 Muridae Rhipidomys ochrogaster 89.00 1 Muridae Rhipidomys nitela 89.00 1 Muridae Rhipidomys macconnelli 41.60 1 Muridae Rhipidomys leucodactylus 80.00 1 Muridae Rhipidomys latimanus 57.50 1 Muridae Rhipidomys fulviventer 89.00 1 Muridae Rhipidomys couesi 89.00 1 Muridae Rhipidomys caucensis 89.00 1 Muridae Rhipidomys austrinus 89.00 1 Muridae Rhipidomys mastacalis 77.50 1 Caenolestidae Rhyncholestes raphanurus 21.00 1 Leporidae Romerolagus diazi 476.70 1 Callitrichidae Saguinus tripartitus 393.50 1 Callitrichidae Saguinus oedipus 430.00 1

36

Callitrichidae Saguinus nigricollis 462.00 1 Callitrichidae Saguinus mystax 618.00 1 Callitrichidae Saguinus midas 540.00 1 Callitrichidae Saguinus leucopus 440.00 1 Callitrichidae Saguinus labiatus 575.00 1 Callitrichidae Saguinus inustus 410.50 1 Callitrichidae Saguinus imperator 400.00 1 Callitrichidae Saguinus geoffroyi 486.50 1 Callitrichidae Saguinus fuscicollis 387.00 1 Callitrichidae Saguinus bicolor 465.00 1 Callitrichidae Saguinus niger 591.20 15 (L) Callitrichidae Saguinus martinsi 603.60 15 (L) Callitrichidae Saguinus graellsi 578.75 15 (L) Cebidae Saimiri vanzolinii 769.60 1 Cebidae Saimiri ustus 1000.00 1 Cebidae Saimiri oerstedii 278.50 1 Cebidae Saimiri sciureus 743.20 1 Cebidae Saimiri boliviensis 615.00 1 Octodontidae Salinoctomys loschalchalerosorum 139.35 19 (L) Muridae Salinomys delicatus 12.50 24 Talpidae Scalopus aquaticus 91.30 1 Talpidae Scapanus townsendii 141.70 1 Talpidae Scapanus orarius 61.20 1 Talpidae Scapanus latimanus 55.00 1 Muridae Scapteromys tumidus 146.00 1 Sciuridae Sciurillus pusillus 39.00 1 Sciuridae Sciurus richmondi 205.00 1 Sciuridae Sciurus yucatanensis 225.00 1 Sciuridae Sciurus colliaei 498.00 1 Sciuridae Sciurus stramineus 433.30 1 Sciuridae Sciurus sanborni 136.00 1 Sciuridae Sciurus pyrrhinus 482.00 1 Sciuridae Sciurus pucheranii 803.00 1 Sciuridae Sciurus igniventris 700.00 1 Sciuridae Sciurus ignitus 190.00 1 Sciuridae Sciurus gilvigularis 803.00 1 Sciuridae Sciurus flammifer 4293.00 1 Sciuridae Sciurus aestuans 185.00 1 Sciuridae Sciurus variegatoides 485.00 1 Sciuridae Sciurus spadiceus 403.30 1 Sciuridae Sciurus oculatus 650.00 1 Sciuridae Sciurus niger 761.90 1 Sciuridae Sciurus nayaritensis 697.00 1 Sciuridae Sciurus griseus 731.00 1 Sciuridae Sciurus granatensis 250.00 1 Sciuridae Sciurus deppei 190.00 1 Sciuridae Sciurus carolinensis 506.50 1 Sciuridae Sciurus aureogaster 595.00 1 Sciuridae Sciurus arizonensis 647.00 1 Sciuridae Sciurus alleni 434.50 1 Sciuridae Sciurus aberti 624.00 1

37

Muridae Scolomys ucayalensis 26.50 1 Muridae Scolomys melanops 26.50 1 Muridae Scotinomys xerampelinus 15.00 1 Muridae Scotinomys teguina 11.20 1 Muridae Sigmodon mascotensis 120.00 1 Muridae Sigmodon arizonae 198.00 1 Muridae Sigmodon inopinatus 140.50 1 Muridae Sigmodon ochrognathus 122.00 1 Muridae Sigmodon leucotis 135.50 1 Muridae Sigmodon hispidus 92.40 1 Muridae Sigmodon fulviventer 214.00 1 Muridae Sigmodon alstoni 55.70 1 Muridae Sigmodon alleni 179.00 1 Soricidae Sorex sonomae 8.30 1 Soricidae Sorex pacificus 6.70 1 Soricidae Sorex bairdi 8.30 1 Soricidae Sorex ventralis 7.00 1 Soricidae Sorex stizodon 7.00 1 Soricidae Sorex sclateri 7.00 1 Soricidae Sorex preblei 3.10 1 Soricidae Sorex oreopolus 7.00 1 Soricidae Sorex milleri 7.00 1 Soricidae Sorex macrodon 7.00 1 Soricidae Sorex lyelli 5.00 1 Soricidae Sorex emarginatus 7.00 1 Soricidae Sorex alaskanus 14.00 1 Soricidae Sorex veraepacis 7.50 1 Soricidae Sorex vagrans 4.40 1 Soricidae Sorex trowbridgii 3.80 1 Soricidae Sorex tenellus 3.70 1 Soricidae Sorex saussurei 4.60 1 Soricidae Sorex palustris 13.40 1 Soricidae Sorex ornatus 5.00 1 Soricidae Sorex nanus 2.40 1 Soricidae Sorex monticolus 5.30 1 Soricidae Sorex merriami 5.90 1 Soricidae Sorex longirostris 3.60 1 Soricidae Sorex hoyi 2.60 1 Soricidae Sorex gaspensis 2.90 1 Soricidae Sorex fumeus 7.70 1 Soricidae Sorex dispar 5.00 1 Soricidae Sorex cinereus 4.70 1 Soricidae Sorex bendirii 16.10 1 Soricidae Sorex arizonae 2.40 1 Soricidae Sorex arcticus 8.20 1 Soricidae Sorex ugyunak 3.60 3 Soricidae Sorex haydeni 4.00 3 Soricidae Sorex yukonicus 3.00 7 Soricidae Sorex tundrensis 3.00 7 Octodontidae Spalacopus cyanus 127.50 1 Canidae Speothos venaticus 6000.00 1

38

Sciuridae Spermophilus richardsonii 406.00 1 Sciuridae Spermophilus mollis 165.40 1 Sciuridae Spermophilus mohavensis 190.00 1 Sciuridae Spermophilus elegans 453.60 1 Sciuridae Spermophilus perotensis 140.00 1 Sciuridae Spermophilus madrensis 275.00 1 Sciuridae Spermophilus brunneus 300.00 1 Sciuridae Spermophilus annulatus 500.00 1 Sciuridae Spermophilus adocetus 125.00 1 Sciuridae Spermophilus washingtoni 207.50 1 Sciuridae Spermophilus variegatus 691.60 1 Sciuridae Spermophilus tridecemlineatus 131.70 1 Sciuridae Spermophilus townsendii 213.00 1 Sciuridae Spermophilus tereticaudus 156.50 1 Sciuridae Spermophilus spilosoma 89.00 1 Sciuridae Spermophilus saturatus 220.00 1 Sciuridae Spermophilus parryii 760.00 1 Sciuridae Spermophilus mexicanus 189.60 1 Sciuridae Spermophilus lateralis 191.00 1 Sciuridae Spermophilus franklinii 363.00 1 Sciuridae Spermophilus columbianus 493.00 1 Sciuridae Spermophilus beldingi 280.50 1 Sciuridae Spermophilus beecheyi 578.50 1 Sciuridae Spermophilus atricapillus 551.00 1 Sciuridae Spermophilus armatus 313.00 1 Mephitidae Spilogale pygmaea 235.00 1 Mephitidae Spilogale putorius 341.00 1 Mephitidae Spilogale gracilis 467.00 8 Leporidae Sylvilagus insonus 3000.00 1 Leporidae Sylvilagus cunicularius 3000.00 1 Leporidae Sylvilagus brasiliensis 950.00 1 Leporidae Sylvilagus transitionalis 656.70 1 Leporidae Sylvilagus palustris 1500.00 1 Leporidae Sylvilagus nuttallii 755.10 1 Leporidae Sylvilagus floridanus 1172.80 1 Leporidae Sylvilagus bachmani 643.30 1 Leporidae Sylvilagus audubonii 889.60 1 Leporidae Sylvilagus aquaticus 2135.80 1 Muridae Synaptomys cooperi 31.90 1 Muridae Synaptomys borealis 21.30 1 Myrmecophagidae Tamandua tetradactyla 5515.10 1 Myrmecophagidae Tamandua mexicana 4210.00 1 Sciuridae Tamias durangae 85.00 1 Sciuridae Tamias bulleri 100.00 1 Sciuridae Tamiasciurus douglasii 225.00 1 Sciuridae Tamiasciurus hudsonicus 201.20 1 Muridae Tapecomys primus 71.50 27 Tapiridae Tapirus bairdii 299999.10 1 Tapiridae Tapirus pinchaque 148949.80 1 Tapiridae Tapirus terrestris 207500.90 1 Mustelidae Taxidea taxus 7107.60 1

39

Tayassuidae Tayassu pecari 32233.70 1 Muridae Thalpomys lasiotis 24.00 1 Muridae Thalpomys cerradensis 24.00 1 Muridae Thaptomys nigrita 19.90 1 Muridae Thomasomys vestitus 76.50 1 Muridae Thomasomys taczanowskii 77.00 1 Muridae Thomasomys silvestris 77.00 1 Muridae Thomasomys rosalinda 77.00 1 Muridae Thomasomys rhoadsi 77.00 1 Muridae Thomasomys pyrrhonotus 77.00 1 Muridae Thomasomys paramorum 77.00 1 Muridae Thomasomys notatus 77.00 1 Muridae Thomasomys niveipes 77.00 1 Muridae Thomasomys monochromos 77.00 1 Muridae Thomasomys laniger 35.50 1 Muridae Thomasomys ladewi 77.00 1 Muridae Thomasomys kalinowskii 77.00 1 Muridae Thomasomys ischyurus 77.00 1 Muridae Thomasomys incanus 77.00 1 Muridae Thomasomys hylophilus 77.00 1 Muridae Thomasomys gracilis 77.00 1 Muridae Thomasomys eleusis 77.00 1 Muridae Thomasomys daphne 77.00 1 Muridae Thomasomys cinereus 77.00 1 Muridae Thomasomys cinereiventer 77.00 1 Muridae Thomasomys bombycinus 114.50 1 Muridae Thomasomys baeops 77.00 1 Muridae Thomasomys aureus 88.00 1 Geomyidae Thomomys townsendii 263.40 1 Geomyidae Thomomys bulbivorus 359.90 1 Geomyidae Thomomys monticola 80.00 1 Geomyidae Thomomys mazama 75.00 1 Geomyidae Thomomys umbrinus 166.00 1 Geomyidae Thomomys talpoides 130.10 1 Geomyidae Thomomys bottae 114.70 1 Geomyidae Thomomys idahoensis 67.00 3 Geomyidae Thomomys clusius 58.00 3 Echimyidae Thrichomys apereoides 275.00 1 Marmosidae Thylamys velutinus 20.70 1 Marmosidae Thylamys pallidior 14.90 1 Marmosidae Thylamys macrura 20.70 1 Marmosidae Thylamys elegans 28.90 1 Marmosidae Thylamys pusilla 41.00 1 Dasypodidae Tolypeutes tricinctus 1487.00 1 Dasypodidae Tolypeutes matacus 1200.00 1 Ursidae Tremarctos ornatus 140000.60 1 Echimyidae Trinomys setosus 285.00 1 Echimyidae Trinomys iheringi 203.50 1 Echimyidae Trinomys dimidiatus 167.60 1 Echimyidae Trinomys albispinus 285.00 1 Muridae Tylomys tumbalensis 280.00 1

40

Appendix S1 references:

1) Smith, F. A., Lyons, S. K., Morgan Ernest, S. K., Jones, K. E., Kaufman, D. M., Dayan, T.,

Marquet, P. A., Brown, J. H., & Haskell, J. P. (2003) Body mass of late Quaternary

mammals. Ecology, 84: 3403. Ecological Archives E084-094.

2) Mares, M. A. & Braun J. K. (2000). Three new species of Brucepattersonius (Rodentia:

Sigmodontinae) from Misiones Province, Argentina. Occasional Papers, Sam Noble

Oklahoma Museum of Natural History, 9, 1–13.

3) http://www.mnh2.si.edu/education/mna/image_info.cfm?species_id=264

4) http://www.icb.ufmg.br/~primatas/prim15.htm

5) http://members.tripod.com/uakari/cebus_kaapori.html

6) http://www.ine.gob.mx/dgoece/con_eco/mamiferos.html

7) http://www.uaf.edu/museum/mammal/AK_Mammals/Shrews/

8) http://www.nsrl.ttu.edu/tmot1/spilgrac.htm

Muridae Tylomys bullaris 280.00 1 Muridae Tylomys mirae 183.60 1 Muridae Tylomys nudicaudus 100.00 1 Octodontidae Tympanoctomys barrerae 70.00 1 Canidae Urocyon cinereoargenteus 3833.70 1 Ursidae Ursus arctos 139440.80 1 Ursidae Ursus americanus 99949.40 1 Camelidae Vicugna vicugna 47499.60 1 Canidae Vulpes vulpes 4131.70 1 Canidae Vulpes velox 2197.50 1 Muridae Wiedomys pyrrhorhinos 46.70 1 Muridae Wilfredomys oenax 46.80 1 Muridae Xenomys nelsoni 130.00 1 Dasypodidae Zaedyus pichiy 1380.00 1 Dipodidae Zapus trinonatus 27.50 1 Dipodidae Zapus princeps 29.00 1 Dipodidae Zapus hudsonius 17.10 1 Muridae Zygodontomys brunneus 75.60 1 Muridae Zygodontomys brevicauda 52.20 1 Geomyidae Zygogeomys trichopus 500.00 1

41

9) http://www.scz.org/animals/a/acouchi.html

10) http://www.discoverlife.org/nh/tx/Vertebrata/Mammalia/Muridae/Neotoma/magister/

11) Weksler M., Geise L. & Cerqueira R. (1999) A new species of Oryzomys (Rodentia,

Sigmondontinae) from southeast Brazil, with comments on the classification of the O. capito

species group. Zoological Journal of the Linnean Society, 125, 445–462

12) D’Elía, G., Ojeda, A., Mondaca, F. & Gallardo, M. H. (2006) New data of the long-clawed

mouse Pearsonomys annectens (Cricetidae, Sigmodontinae) and additional comments on the

distinctiveness of Pearsonomys. Mammalian Biology, 71, 39-51.

13) Vieira, E.M., Pizo M.A., Izar P. (2003) Fruit and seed exploitation by small rodents of the

Brazilian Atlantic forest. Mammalia, 67 (4): 1-7

14) Richard-Hansen, C., Vié, J.-C., Vidal, N. & Kéravec, J. (1999) Body measurements on 40

species of mammals from French Guiana. Journal of Zoology, 247, 419-428

15) http://www.damisela.com/zoo/mam/primates/callitrichidae

16) Voss, R.S. & da Silva, M.N.F. (2001) Revisionary notes on Neotropical porcupines (Rodentia:

Erethizontidae). 2. A review of the Coendou vestitus group with descriptions of two new

species from Amazonia. American Museum Novitates, 3351, 1-36.

17) http://www.washington.edu/burkemuseum/collections/mammalogy/mamwash/peke.html

18) http://www.parquenacionallanin.gov.ar/inst/ed_ambiental/Fichas%20vertebrados.doc

19) Mares, M. A., Braun, J. K., Barquez, R. M. & Díaz, M. M. (2000) Two new genera and species

of halophytic desert mammals from isolated salt flats in Argentina. Occasional Papers,

Museum of Texas Tech University, 203, 1-27.

20) Weksler, M. & Bonvicino, C.R. (2005) Taxonomy of pigmy rice rats genus Oligoryzomys

Bangs, 1900 (Rodentia, Sigmodontinae) of the Brazilian Cerrado, with the description of two

new species. Arquivos do Museu Nacional, Rio de Janeiro, 63, 113-130.

42

21) Voss, R. S., Lunde, D. P. & Simmons, N. B. (2001) The mammals of Paracou, French Guiana:

A Neotropical lowland rainforest fauna. Part 2: Nonvolant species. Bulletin of the American

Museum of Natural History, 263, 1-236.

22) van Roosmalen, M. G. M., van Roosmalen, T. & Mittermeier, R. A. (2002) A taxonomic

review of the titi monkeys, genus Callicebus Thomas, 1903, with the description of two new

species, Callicebus bernhardi and Callicebus stephennashi, from Brazilian Amazonia.

Neotropical Primates, 10 (Suppl), 1-52.

23) Lunde, D. P. & Pacheco, V. (2003) Shrew opossums (Paucituberculata: Caenolestes) from the

Huancabamba region of east Andean Peru. Mammal Study, 28, 145-148.

24) Al-kahtani, M. A., Zuleta, C., Caviedes-Vidal, E., & Garland Jr., T. (2004) Kidney mass and

relative medullary thickness of rodents in relation to habitat, body size, and phylogeny.

Physiological and Biochemical Zoology, 77, 346–365.

25) Luna, W. L. & Patterson, B. D. (2003) A remarkable new mouse (Muridae: Sigmodontinae)

from Southeastern Peru: With comments on the affinities of Rhagomys rufescens (Thomas,

1886). Fieldiana: Zoology, New Series, 101, 1-34.

26) van Roosmalen M.G.M. & van Roosmalen T. (2003) The description of a new marmoset

genus, Callibella (Callitrichinae, Primates), including its molecular phylogenetic status.

Neotropical Primates, 11, 1-12.

27) Anderson, S. & Yates, T. L. (2000) A new genus and species of phyllotine rodent from Bolivia.

Journal of Mammalogy, 81,18-36.

28) Barnett A. A. (1997) The natural history and conservation of a fishing mouse Chibchanomys

spec. nov. (Ichthyomyini: Muridae) from the Andes of southern Ecuador. Zeitschrift für

Säugetierkunde, 62, 43-52.

43

Appendix S2 – Species considered as synonyms, subspecies or conspecifics of species with known body masses (see column labelled “Equivalent weighted species”). Body masses are provided for all species. Codes under the column labelled as “source” indicate the reference where the data were found. The list of references is provided below.

Family Genus Species Mass (g) Source Equivalent weighted species Muridae Akodon reigi 28.30 9 Akodon paranaensis Muridae Akodon nucus 28.70 28 Akodon iniscatus Muridae Akodon lutescens 20.70 22 Akodon puer Muridae Akodon paranaensis 28.30 2 Akodon serrensis Atelidae Alouatta coibensis 7274.90 2 Alouatta palliata Atelidae Alouatta guariba 8000.00 2 Alouatta fusca Muridae Amphinectomys savamis 17.40 2 Neacomys savamis Atelidae Ateles hybridus 5000.00 23 Ateles belzebuth Pitheciidae Callicebus ornatus 915.00 5 Callicebus cupreus Pitheciidae Callicebus pallescens 795.00 5 Callicebus donacophilus Pitheciidae Callicebus baptista 923.55 5 Callicebus hoffmannsi Pitheciidae Callicebus nigrifrons 1350.00 5 Callicebus personatus Pitheciidae Callicebus melanochir 1350.00 5 Callicebus personatus Pitheciidae Callicebus barbarabrownae 1350.00 5 Callicebus personatus Pitheciidae Callicebus regulus 1050.00 5 Callicebus torquatus Pitheciidae Callicebus purinus 1050.00 5 Callicebus torquatus Pitheciidae Callicebus medemi 1050.00 5 Callicebus torquatus Pitheciidae Callicebus lugens 1050.00 5 Callicebus torquatus Heteromyidae Chaetodipus eremicus 15.00 21 Chaetodipus penicillatus Muridae Chelemys delfini 50.80 2 Chelemys megalonyx Erethizontidae Coendou nycthemera 3000.00 10 Coendou koopmani Soricidae Cryptotis alticola 8.00 24 Cryptotis goldmani Soricidae Cryptotis equatoris 12.60 27 Cryptotis thomasi Ctenomyidae Ctenomys goodfellowi 535.00 2 Ctenomys boliviensis Ctenomyidae Ctenomys bergi 177.50 2 Ctenomys mendocinus Ctenomyidae Ctenomys rondoni 400.00 2 Ctenomys nattereri Ctenomyidae Ctenomys coyhaiquensis 164.00 25 Ctenomys haigi Ctenomyidae Ctenomys rionegrensis 92.00 26 Ctenomys minutus Abrocomidae Cuscomys ashaninka 158.00 29 Abrocoma boliviensis Muridae Dicrostonyx richardsoni 54.40 2 Dicrostonyx groenlandicus Muridae Dicrostonyx nunatakensis 54.40 2 Dicrostonyx groenlandicus Heteromyidae Dipodomys simulans 57.80 2 Dipodomys agilis Heteromyidae Dipodomys californicus 72.00 2 Dipodomys heermanni Heteromyidae Dipodomys compactus 60.40 2 Dipodomys ordii Echimyidae Euryzygomatomys guiara 187.50 2 Euryzygomatomys spinosus Geomyidae Geomys knoxjonesi 206.00 2 Geomys arenarius Geomyidae Geomys texensis 203.80 2 Geomys bursarius Geomyidae Geomys attwateri 203.80 2 Geomys bursarius Geomyidae Geomys breviceps 203.80 2 Geomys bursarius Marmosidae Gracilinanus longicaudus 7.60 17 Gracilinanus emiliae Muridae Isthmomys flavidus 138.00 3 Isthmomys pirrensis Muridae Juscelinomys talpinus 97.30 2 Juscelinomys candango Muridae Juscelinomys vulpinus 97.30 18 Juscelinomys talpinus Atelidae Lagothrix poeppigii 6300.00 1 Lagothrix lagothricha

44

Atelidae Lagothrix lugens 6300.00 1 Lagothrix lagothricha Atelidae Lagothrix cana 6300.00 1 Lagothrix lagothricha Muridae Lemmus trimucronatus 52.30 2 Lemmus sibiricus Callitrichidae Leontopithecus caissara 535.50 1 Leontopithecus rosalia Muridae Lundomys molitor 238.50 2 Holochilus magnus Felidae Lynchailurus pajeros 3935.00 7 Oncifelis colocolo Felidae Lynchailurus braccatus 3935.00 7 Oncifelis colocolo Echimyidae Makalata occasius 400.00 1 Makalata armata Muridae Maresomys boliviensis 65.70 2 Auliscomys boliviensis Marmosidae Marmosops neblina 40.50 1 Marmosops impavidus Marmosidae Marmosops pinheiroi 15.00 2 Marmosops parvidens Marmosidae Marmosops paulensis 62.30 2 Marmosops incanus Sciuridae Marmota broweri 3600.00 2 Marmota caligata Cervidae Mazama pandora 22799.70 1 Mazama americana Muridae Megadontomys nelsoni 111.00 19 Megadontomys thomasi Callitrichidae Mico intermedius 350.00 4 Mico argentatus Callitrichidae Mico melanurus 440.00 4 Mico argentatus Callitrichidae Mico leucippe 440.00 4 Mico argentatus Callitrichidae Mico emiliae 440.00 4 Mico humeralifer Callitrichidae Mico chrysoleucus 350.00 5 Mico humeralifer Callitrichidae Mico manicorensis 339.60 2 Callithrix manicorensis Callitrichidae Mico acariensis 339.60 2 Callithrix acariensis Sciuridae Microsciurus similis 92.00 3 Microsciurus flaviventer Muridae Microtus mogollonensis 48.00 2 Microtus mexicanus Muridae Neacomys pictus 19.00 3 Neacomys tenuipes Muridae Necromys lenguarum 39.90 2 Necromys lasiurus Muridae Nectomys magdalenae 190.70 1 Nectomys squamipes Muridae Nectomys melanius 190.70 2 Notiosorex squamipes Muridae Nectomys garleppii 190.70 2 Notiosorex squamipes Muridae Nectomys apicalis 190.70 2 Notiosorex squamipes Muridae Nectomys palmipes 190.70 3 Nectomys squamipes Muridae Neotoma chrysomelas 203.00 3 Neotoma mexicana Soricidae Notiosorex villai 4.40 2 Notiosorex crawfordi Soricidae Notiosorex evotis 4.40 2 Notiosorex crawfordi Cervidae Odocoileus lasiotus 55508.60 2 Odocoileus virginianus Muridae Oecomys auyantepui 73.40 1 Oecomys paricola Muridae Oligoryzomys vegetus 25.00 3 Oligoryzomys fulvescens Geomyidae Orthogeomys thaeleri 438.00 11 Orthogeomys dariensis Geomyidae Orthogeomys cherriei 615.00 12 Orthogeomys heterodus Muridae Oryzomys laticeps 57.80 1 Oryzomys capito Muridae Oryzomys angouya 144.00 1 Oryzomys ratticeps Muridae Oryzomys rostratus 50.00 6 Oryzomys melanotis Muridae Oryzomys rhabdops 33.30 6 Oryzomys alfaroi Muridae Oryzomys saturatior 33.30 6 Oryzomys alfaroi Muridae Oryzomys devius 60.50 20 Oryzomys albigularis Muridae Peromyscus fraterculus 23.60 2 Peromyscus eremicus Muridae Peromyscus levipes 21.40 13 Peromyscus boylii Didelphidae Philander frenata 750.00 14 Philander opossum Muridae Phyllotis limatus 56.30 2 Phyllotis xanthopygus Echimyidae Proechimys roberti 285.00 15 Proechimys roberti

45

Muridae Reithrodon typicus 70.90 3 Reithrodon auritus Muridae Reithrodontomys zacatecae 9.40 8 Reithrodontomys megalotis Muridae Rheomys underwoodi 40.00 3 Rheomys mexicanus Sciuridae Sciurus alphonsei 185.00 16 Sciurus aestuans Muridae Scolomys juruaense 26.50 2 Scolomys melanops Muridae Sigmodon peruanus 92.40 3 Sigmodon hispidus Muridae Sigmodontomys alfari 59.98 2 Oryzomys alfari Muridae Sigmodontomys aphrastus 59.98 2 Oryzomys aphrastus Soricidae Sorex neomexicanus 4.85 2 Sorex vagrans Sciuridae Spermophilus canus 213.00 3 Spermophilus townsendii Leporidae Sylvilagus dicei 950.00 3 Sylvilagus brasiliensis Sciuridae Tamiasciurus mearnsi 225.00 3 Tamiasciurus douglasii Muridae Thaptomys subterraneus 19.90 9 Thaptomys nigrita Muridae Tylomys watsoni 183.60 3 Tylomys panamensis Muridae Tylomys panamensis 183.60 3 Tylomys panamensis Muridae Tylomys fulviventer 183.60 3 Tylomis mirae Canidae Vulpes macrotis 2050.00 2 Vulpes velox

Appendix S2 references:

1) http://sea.unep-wcmc.org/species/dbases/about.cfm

2) http://www.natureserve.org

3) http://nmnhgoph.si.edu/cgi-bin/wdb/msw/names/query

4) http://www.damisela.com/zoo/mam/primates/callitrichidae

5) UNEP-WCMC. (2003) Checklist of mammals listed in the CITES appendices and in EC

Regulation 338/97. 6th edition. JNCC Report No. 342.

http://www.ukcites.gov.uk/pdf_files/Mammals6th.pdf

6) Retana, O. G., & Lorenzo, C. (2002) Lista de los mamíferos terrestres de Chiapas: endemismo y

estado de conservación. Acta Zoologica Mexicana, 85, 25-49.

http://www.ecologia.edu.mx/azm/documentos/85/85-c.pdf

7) http://animaldiversity.ummz.umich.edu/site/accounts/information/Oncifelis_colocolo.html

8) Hood, C. S., Robbins, L. W., Baker, R. J., & Shellammer, H.S. (1984) Chromosomal studies

and evolutionary relationships of an endagered species, Reithrodontomys raviventris. Journal

of Mammalogy, 65, 655-667.

46

9) Cherel, J. J., Simões-Lopes, P. C., Althoff, S. & Graipel, M. E. (2004) Lista dos mamíferos do

Estado de Santa Catarina, Sul do Brasil. Mastozoología Neotropical, 11, 151-184.

10) Voss, R. S. & da Silva, M. N. F. (2001) Revisionary notes on Neotropical Porcupines

(Rodentia: Erethizontidae). 2. A review of the Coendou vestitus Group with Descriptions of

two new species from Amazonia. American Museum Novitates, 3351, 1-36.

11) Sudman, P. D., Hafner, M. S. (1992) Phylogenetic relationships among middle American

pocket gophers (genus Orthogeomys) based on mitochondrial DNA sequences. Molecular

phylogenetics and evolution, 1, 17-25.

12) Huelsenbeck, J. P. (2000) Likelihood-Based Inference of Phylogeny. Department of Biology,

University of Rochester. Unpublished Dissertation.

13) Romero-Almaráz, M. L., García-Estrada, C. & Sánchez-Hernández, C. (2004) Peromyscus

levipes (Rodentia: Muridae) in deciduous forest in southeastern Morelos, Mexico. The

Southwestern Naturalist, 49, 125-131.

14)http://www.science.smith.edu/departments/Biology/VHAYSSEN/msi/pdf/638_Philander_oposs

um.pdf

15) Weksler, M., Bonvicino, C. R., Otazu, I. B. & Silva, Jr. J. S.(2001) Status of Proechimys roberti

and P. oris (Rodentia: Echimyidae) from Eastern Amazonia and Central Brazil. Journal of

Mammalogy, 82, 109–122.

16) Valéria, F., Uarth Christoff, A., Amaro-Ghilard, R.C., Scheibler, D. R. & Yonenaga-Yassuda Y.

(2003) Multiple interstitial ribosomal sites (NORs) in the Brazilian squirrel Sciurus aestuans

ingrami(Rodentia, Sciuridae) with 2n = 40. An overview of Sciurus cytogenetics. Genetics and

Molecular Biology, 26, 253-257.

17) http://www.zipcodezoo.com/animals/g/gracilinanus_longicaudus.asp

18) http://savci.upol.cz/hlod-4.htm

47

19) http://www.ine.gob.mx/dgoece/con_eco/mamiferos.html

20) Rodríguez-Herrera, B., Chinchilla, F. A., & May-Collado, L. J. (2002) Lista de especies,

endemismo y conservación de los de mamíferos de Costa Rica. Revista Mexicana de

Mastozoología, 6, 19-41.

21) http://www.fw.vt.edu/fishex/nmex_main/species/050428.htm

22) http://www.zipcodezoo.com/animals/a/akodon_puer.asp

23) http://www.animalinfo.org/country/venezuel.htm

24) http://www.conabio.gob.mx/remib/doctos/checklist/3-1.html

25) http://www.rednova.com/news/display/?id=131909

26)http://www.rickertweb.com/~justin/School/CS360%20Database%20Systems/PC2/Data/SYNO

M/MAM70SYN.txt

27)http://bnhm.berkeley.edu/browse/vertebrates_Mammalia_Insectivora_Soricidae_Cryptotis_all.p

hp?

28)http://216.239.59.104/search?q=cache:FSobyUONjp8J:www.sib.gov.ar/fichas/fauna/sinonimos.a

sp%3Fid%3DAkodon%2520iniscatus+%22Akodon+nucus%22&hl=es

29) http://nl.wikipedia.org/wiki/Chinchillaratten

48

App

endi

x S3

– S

peci

es c

onsi

dere

d “s

imila

r to

” or

“be

long

ing

to th

e gr

oup”

of

spec

ies

with

kno

wn

body

mas

ses

(see

col

umn

labe

lled

“Equ

ival

ent w

eigh

ted

spec

ies”

). Fo

r th

ose

case

in w

hich

sim

ilari

ty o

ccur

red

with

mor

e th

an o

ne s

peci

es, w

e as

sign

ed th

e av

erag

e m

ass

valu

e of

the

se g

roup

. Bod

y m

asse

s ar

e pr

ovid

ed f

or a

ll sp

ecie

s. C

odes

und

er t

he c

olum

n la

belle

d as

“so

urce

” in

dica

te t

he r

efer

ence

w

here

the

data

wer

e fo

und.

The

list

of r

efer

ence

s is

pro

vide

d be

low

.

Fam

ily

Gen

us

Spec

ies

Mas

s (g

) So

urce

E

quiv

alen

t wei

ghte

d sp

ecie

s M

urid

ae

Aep

eom

ys

reig

i 37

.00

21

Aep

eom

ys lu

gens

M

urid

ae

Ako

don

hers

hkov

itzi

21.3

0 3

Ako

don

xant

horh

inus

M

urid

ae

Ako

don

mon

tens

is

39.9

0 10

A

kodo

n cu

rsor

M

urid

ae

Ako

don

aliq

uant

ulus

13

.00

24

Ako

don

bogo

tens

is

Mur

idae

A

kodo

n oe

nos

42.4

0 35

A

kodo

n ne

ocen

us

Mur

idae

A

kodo

n pe

rval

ens

40.0

0 32

A

kodo

n va

rius

M

urid

ae

Ako

don

mys

tax

20.6

9 35

A

vera

ge o

f fir

st q

uatil

e of

sm

alle

st A

kodo

n A

telid

ae

Alo

uatta

ul

ulat

a 64

00.0

0 5

Alo

uatta

bel

zebu

l A

telid

ae

Alo

uatta

ni

gerr

ima

6145

.50

12

Alo

uatta

sen

icul

us

Mur

idae

A

ndal

galo

mys

ro

igi

32.9

0 9

And

alga

lom

ys o

lrog

i M

urid

ae

Apo

rodo

n pa

rado

xus

12.9

0 14

A

poro

don

brev

iros

tris

M

urid

ae

Apo

rodo

n ro

drig

uezi

20

.00

14

Apo

rodo

n te

nuir

ostr

is

Mur

idae

A

poro

don

spec

tabi

lis

19.0

0 14

A

poro

don

mex

ican

us

Mur

idae

B

ibim

ys

labi

osus

28

.00

18

Bib

imys

cha

coen

sis,

Bib

imys

torr

esi

Ate

lidae

B

rach

ytel

es

hypo

xant

hus

1500

0.00

25

B

rach

ytel

es a

rach

noid

es

Pith

eciid

ae

Cal

liceb

us

luci

fer

1050

.00

7 C

allic

ebus

torq

uatu

s P

ithec

iidae

C

allic

ebus

co

imbr

ai

1350

.00

7 C

allic

ebus

cin

eras

cens

, Cal

liceb

us b

runn

eus,

Cal

liceb

us h

offm

anns

i C

aviid

ae

Cav

ia

inte

rmed

ia

591.

38

28

Cav

ia a

pere

a,C

avia

mag

na

Het

erom

yida

e C

haet

odip

us

artu

s 20

.62

8 C

haet

odip

us g

oldm

ani

Pith

eciid

ae

Chi

ropo

tes

utah

icki

30

00.0

0 5

Chi

ropo

tes

sata

nas

Pith

eciid

ae

Chi

ropo

tes

sagu

latu

s 30

00.0

0 5

Chi

ropo

tes

sata

nas

Pith

eciid

ae

Chi

ropo

tes

chir

opot

es

3000

.00

5 C

hiro

pote

s sa

tana

s E

reth

izon

tidae

C

oend

ou

prui

nosu

s 73

6.00

1

Coe

ndou

ves

titus

E

reth

izon

tidae

C

oend

ou

ichi

llus

736.

00

1 C

oend

ou v

estit

us

Ere

thiz

ontid

ae

Coe

ndou

ro

osm

alen

orum

60

0.00

1

Coe

ndou

ves

titus

, Coe

ndou

ichi

llus,

Coe

ndou

pru

inos

us

Sori

cida

e C

rypt

otis

gr

acili

s 7.

00

6 C

rypt

otis

nig

resc

ens

Sori

cida

e C

rypt

otis

en

ders

i 7.

00

6 C

rypt

otis

nig

resc

ens

Sori

cida

e C

rypt

otis

pe

ruvi

ensi

s 12

.60

13

Cry

ptot

is th

omas

i So

rici

dae

Cry

ptot

is

med

ellin

ia

12.6

0 13

C

rypt

otis

thom

asi

Sori

cida

e C

rypt

otis

m

erri

ami

7.00

13

C

rypt

otis

nig

resc

ens

49

Sori

cida

e C

rypt

otis

m

era

7.00

13

C

rypt

otis

nig

resc

ens

Sori

cida

e C

rypt

otis

m

ayen

sis

7.00

13

C

rypt

otis

nig

resc

ens

Sori

cida

e C

rypt

otis

ph

illip

sii

7.00

13

C

rypt

otis

mex

ican

a So

rici

dae

Cry

ptot

is

pere

grin

a 7.

00

13

Cry

ptot

is m

exic

ana

Sori

cida

e C

rypt

otis

ob

scur

a 7.

00

13

Cry

ptot

is m

exic

ana

Sori

cida

e C

rypt

otis

ne

lson

i 7.

00

13

Cry

ptot

is m

exic

ana

Sori

cida

e C

rypt

otis

gr

iseo

vent

ris

7.00

13

C

rypt

otis

mex

ican

a So

rici

dae

Cry

ptot

is

colo

mbi

ana

7.00

22

C

rypt

otis

nig

resc

ens

Sori

cida

e C

rypt

otis

ho

ndur

ensi

s 7.

00

29

Cry

ptot

is n

igre

scen

s C

teno

myi

dae

Cte

nom

ys

roig

i 40

0.00

15

C

teno

mys

per

rens

is

Cte

nom

yida

e C

teno

mys

do

rbig

nyi

400.

00

15

Cte

nom

ys p

erre

nsis

C

teno

myi

dae

Cte

nom

ys

flam

mar

ioni

17

7.50

16

C

teno

mys

men

doci

nus

Cte

nom

yida

e C

teno

mys

pi

lare

nsis

30

7.50

27

C

teno

mys

per

rens

i, C

teno

mys

pea

rson

i C

teno

myi

dae

Cte

nom

ys

rose

ndop

ascu

ali

177.

50

30

Cte

nom

ys m

endo

cinu

s C

teno

myi

dae

Cte

nom

ys

osva

ldor

eigi

17

7.50

30

C

teno

mys

men

doci

nus

Mar

mos

idae

G

raci

linan

us

peri

jae

30.0

0 34

G

raci

linan

us ig

nitu

s M

urid

ae

Nea

com

ys

mus

seri

17

.40

17

Nea

com

ys g

uian

ae, m

inut

us, m

usse

ri,p

ictu

s, p

usill

us, t

enui

pes

Mur

idae

N

eaco

mys

m

inut

us

17.4

0 17

N

eaco

mys

gui

anae

, min

utus

, mus

seri

,pic

tus,

pus

illus

, ten

uipe

s M

urid

ae

Ory

zom

ys

tate

i 60

.50

20

Ory

zom

ys y

unga

nus

Mur

idae

O

ryzo

mys

di

mid

iatu

s 53

.90

21

Ory

zom

ys p

alus

tris

M

urid

ae

Oxy

myc

teru

s ju

liaca

e 68

.00

19

Oxy

myc

teru

s hi

ska

Mur

idae

O

xym

ycte

rus

amaz

onic

us

81.5

0 19

O

xym

ycte

rus

dela

tor

Mur

idae

O

xym

ycte

rus

capa

raoe

68

.00

33

Oxy

myc

teru

s na

tusu

s M

urid

ae

Per

omys

cus

pseu

docr

initu

s 18

.00

4 P

erom

yscu

s cr

initu

s M

urid

ae

Per

omys

cus

may

ensi

s 33

.00

11

Per

omys

cus

furv

us

Did

elph

idae

P

hila

nder

m

cilh

enny

i 80

0.00

26

P

hila

nder

opo

ssum

, Phi

land

er m

cilh

enny

i E

chim

yida

e P

roec

him

ys

patto

ni

300.

50

31

Pro

echi

mys

ori

s, P

roec

him

ys c

ayen

nens

is.

Ech

imyi

dae

Pro

echi

mys

ga

rdne

ri

300.

50

31

Pro

echi

mys

ori

s, P

roec

him

ys c

ayen

nens

is.

Mur

idae

R

hipi

dom

ys

gard

neri

83

.25

36

Rhi

pido

mys

mas

taca

lis, R

hipi

dom

ys n

itela

Sc

iuri

dae

Synt

heos

ciur

us

broc

hus

138.

75

23

Mic

rosc

irus

alfa

ri, S

ciur

us d

eppe

i M

urid

ae

Tho

mas

omys

on

kiro

77

.00

2 T

hom

asom

ys s

ilves

tris

50

App

endi

x S3

ref

eren

ces:

1.

Vos

s, R

. S. &

da

Silv

a, M

. N. F

. (20

01)

Rev

isio

nary

not

es o

n N

eotr

opic

al P

orcu

pine

s (R

oden

tia: E

reth

izon

tidae

). 2.

A r

evie

w o

f th

e C

oend

ou v

estit

us g

roup

with

des

crip

tions

of t

wo

new

spe

cies

from

Am

azon

ia. A

mer

ican

Mus

eum

Nov

itate

s, 33

51, 1

- 36

. 2.

L

una,

L.,

& P

ache

co, V

. (2

002)

A n

ew s

peci

es o

f Tho

mas

omys

(Mur

idae

: Sig

mod

ontin

ae) f

rom

the

And

es o

f Sou

thea

ster

n P

eru.

Jo

urna

l of M

amm

alog

y, 8

3, 8

34-8

42.

3.

Alv

arez

-Cas

tañe

da, S

. T. (

1998

) Per

omys

cus

pseu

docr

initu

s. M

amm

alia

n sp

ecie

s, 60

1, 1

-3.

4.

http

://w

ww

.nat

ures

erve

.org

5.

P

ine,

R.

H.,

Woo

dman

, N

., &

Tim

m,

R.

M.

(200

2) R

edis

cove

ry o

f E

nder

s's s

mal

l-ear

ed s

hrew

, C

rypt

otis

end

ersi

(In

sect

ivor

a:

Sori

cida

e), w

ith a

red

escr

iptio

n of

the

spec

ies.

Mam

mal

ian

Bio

logy

, 67,

372

-377

. 6.

va

n R

oosm

alen

, M

. G

. M

., va

n R

oosm

alen

, T

. &

Mitt

erm

eier

, R

. A

. (2

002)

A t

axon

omic

rev

iew

of

the

titi

mon

keys

, ge

nus

Cal

liceb

us T

hom

as,

1903

, w

ith t

he d

escr

iptio

n of

tw

o ne

w s

peci

es,

Cal

liceb

us b

ernh

ardi

and

Cal

liceb

us s

teph

enna

shi,

from

B

razi

lian

Am

azon

ia. N

eotr

opic

al P

rimat

es, 1

0 (S

uppl

), 1-

52.

7.

Bes

t, T

. L. &

Lac

key,

J. A

. (19

92).

Cha

etod

ipus

art

us. M

amm

alia

n Sp

ecie

s, 41

8, 1

-3.

8.

Bra

un J

. K

. &

Día

z M

. M

. (1

999)

Key

to

the

nativ

e m

amm

als

of C

atam

arca

pro

vinc

e, A

rgen

tina.

Occ

asio

nal

Pape

rs o

f th

e O

klah

oma

Mus

eum

of N

atur

al H

istor

y, 4

, 1-1

6 9.

G

eise

, L

, W

eksl

er,

M.

& B

onvi

cino

, C

. R

. (2

004)

Pre

senc

e or

abs

ence

of

gall

blad

der

in s

ome

Oko

dont

ini

rode

nts

(Mur

idae

, Si

gmod

ontin

ae).

Mam

mal

ian

Bio

logy

, 69,

210

-214

. 10

. http

://en

cycl

oped

ia.th

efre

edic

tiona

ry.c

om/P

erom

yscu

s 11

. http

://en

cycl

oped

ia.la

borl

awta

lk.c

om/A

loua

tta

12. h

ttp://

ww

w.a

nsw

ers.

com

/topi

c/sm

all-e

ared

-shr

ew

13. h

ttp://

nmnh

goph

.si.e

du/c

gi-b

in/w

db/m

sw/n

ames

/que

ry

14. G

imén

ez, M

. D.,

Mir

ol, P

. M.,

Bid

au, C

. J.,

& S

earl

e, J

. B. (

2002

) Mol

ecul

ar a

naly

sis

of p

opul

atio

ns o

f Cte

nom

ys (C

avio

mor

pha,

R

oden

tia) w

ith h

igh

kary

otyp

ic v

aria

bilit

y. C

ytog

enet

ic a

nd G

enom

e R

esea

rch,

96,

130

-136

. 15

. D'E

lia, G

., L

essa

, E. P

. & C

ook,

J. A

. (19

99)

Mol

ecul

ar P

hylo

geny

of

Tuc

o-T

ucos

, Gen

us C

teno

mys

(R

oden

tia: O

ctod

ontid

ae):

Eva

luat

ion

of th

e m

endo

cinu

s Sp

ecie

s G

roup

and

the

Evo

lutio

n of

Asy

mm

etri

c Sp

erm

. Jou

rnal

of M

amm

alia

n E

volu

tion,

6, 1

9-38

. 16

. Vos

s, R

. S.

, L

unde

, D

. P

. &

Sim

mon

s, N

. B

. (2

001)

. T

he m

amm

als

of P

arac

ou,

Fren

ch G

uian

a :

A N

eotr

opic

al l

owla

nd

rain

fore

st fa

una.

Par

t 2 :

Non

vola

nt s

peci

es. B

ulle

tin o

f the

Am

eric

an M

useu

m o

f Nat

ural

Hist

ory,

263

, 1-2

36.

17. D

’Elia

, G

., P

ardi

nas,

U.

F. J

. &

Mye

rs,

P.

(in p

ress

) A

n in

trod

uctio

n to

the

gen

us B

ibim

ys (

Rod

entia

: Si

gmod

ontin

ae).

Phy

loge

netic

pos

ition

and

alp

ha ta

xono

my.

in E

. A. L

acey

and

P. M

yers

, eds

. Fes

tsch

rift f

or Ja

mes

L. P

atto

n. U

. C. P

ress

. 18

. Hof

fman

n, F

. G

., L

essa

, E

. P

. &

Sm

ith,

M.

F. (

2002

) Sy

stem

atic

s of

Oxy

myc

teru

s w

ith d

escr

iptio

n of

a n

ew s

peci

es f

rom

U

rugu

ay. J

ourn

al o

f Mam

mal

ogy,

83,

408

–420

.

51

19. B

onvi

cino

, C

. R

. (2

003)

A n

ew s

peci

es o

f O

ryzo

mys

(R

oden

tia,

Sigm

odon

tinae

) of

the

sub

flavu

s gr

oup

from

the

Cer

rado

of

Cen

tral

Bra

zil.

Mam

mal

ian

Bio

logy

, 68,

78-

90.

20. V

oss,

R. S

., G

ómez

-Lav

erde

, M. &

Pac

heco

, V. (

2002

) A

new

gen

us f

or A

epeo

mys

fusc

atus

Alle

n, 1

912,

and

Ory

zom

ys in

tect

us

Tho

mas

, 192

1: E

nigm

atic

Mur

id r

oden

ts fr

om A

ndea

n cl

oud

fore

sts.

Am

eric

an M

useu

m N

ovita

tes,

3373

, 1-4

2.

21. W

oodm

an, N

., C

uart

as-C

alle

, C.A

. & D

elga

do, C

.A. (

2003

) The

hum

erus

of C

rypt

otis

colo

mbi

ana

and

its b

eari

ng o

n th

e sp

ecie

s’

phyl

ogen

etic

rel

atio

nshi

ps (S

oric

omor

pha:

Sor

icid

ae).

Jour

nal o

f Mam

mal

ogy,

84,

832

–839

. 22

. Wel

ls, N

. M. &

Gia

calo

ne, J

. (19

85) S

ynth

eosc

iuru

s bro

chus

. Mam

mal

ian

Spec

ies,

249

,1-3

. 23

. Día

z, M

. M

., B

arqu

ez, R

. M

., B

raun

J.

K.,

& M

ares

, M.

A.

(19

99)

A n

ew s

peci

es o

f A

kodo

n (M

urid

ae:

Sigm

odon

tinae

) fr

om

nort

hwes

tern

Arg

entin

a. Jo

urna

l of M

amm

alog

y, 8

0, 7

86-7

98.

24. h

ttp://

ww

w.a

nim

alin

fo.o

rg/s

peci

es/p

rim

ate/

brac

arac

.htm

25

. Cas

tro-

Are

llano

, I.,

Zar

za, H

., &

Med

ell,

R.A

. (20

00) P

hila

nder

opo

ssum

. Mam

mal

ian

Spec

ies.

638,

1-8

. 26

. Con

trer

as,

J.R

. (2

000)

Cte

nom

ys p

arag

uaye

nsis,

una

nue

va e

spec

ie d

e ro

edor

exc

avad

or p

roce

dent

e de

l P

arag

uay

Ori

enta

l (M

amm

alia

, Rod

entia

, Cte

nom

yida

e). R

evist

a de

l Mus

eo A

rgen

tino

de C

ienc

ias

Nat

ural

es, 2

, 61-

68.

27. C

here

m,

J.J.

, O

limpi

o, J

., &

Xim

enez

, A

. (1

999)

Des

criç

ão d

e um

a no

va e

spéc

ie d

o gê

nero

Cav

ia P

alla

s, 1

766

(Mam

mal

ia -

C

aviid

ae) d

as I

lhas

dos

Mol

eque

s do

Sul

, San

ta C

atar

ina,

Sul

do

Bra

sil.

Bio

tem

as, 1

2, 9

5-11

7.

28. W

oodm

ann

N.

(200

0) C

rypt

otis

mer

riam

i C

hoat

e in

Cos

ta R

ica:

Syn

topy

with

Cry

ptot

is n

igre

scen

s (A

llen)

and

Pos

sibl

e C

hara

cter

Dis

plac

emen

t (M

amm

alia

: Ins

ectiv

ora)

Car

ibbe

an Jo

urna

l of S

cien

ce, 3

6, 2

89–2

99.

29. R

osi,

M. I

., C

ona,

M. I

., &

Roi

g, V

. G. (

2002

) Est

ado

actu

al d

el c

onoc

imie

nto

del r

oedo

r fo

sori

al C

teno

mys

men

doci

nus

Phi

lippi

18

69 (R

oden

tia: C

teno

myi

dae)

. Mas

tozo

olog

ía N

eotr

opic

al,

9, 2

77-2

95.

30. W

eksl

er,

M.,

Bon

vici

no,

C.

R.,

Ota

zu,

I. B

. &

Silv

a, J

r J.

S. (

2001

) St

atus

of

Proe

chim

ys r

ober

ti an

d P.

oris

(R

oden

tia:

Ech

imyi

dae)

from

eas

tern

Am

azon

ia a

nd c

entr

al B

razi

l. Jo

urna

l of M

amm

alog

y, 8

2, 1

09–1

22.

31. D

íaz,

M. M

., B

raun

, J. K

., M

ares

, M. A

., &

Bar

quez

, R. M

. (20

00)

An

upda

te o

f th

e ta

xono

my,

sys

tem

atic

s, a

nd d

istr

ibut

ion

of

the

mam

mal

s of

Sal

ta P

rovi

nce,

Arg

entin

a. S

am N

oble

Okl

ahom

a M

useu

m o

f Nat

ural

Hist

ory,

10,

1-5

2.

32. G

onça

lves

, P

. R

. &

Oliv

eira

, J.

A.

(200

5) D

iver

sific

atio

n of

Sig

mod

ontin

e ro

dent

s in

hig

h-al

titud

e en

dem

ism

cen

ters

with

in

Bra

zilia

n A

tlant

ic f

ores

t. A

bstr

act

Boo

k of

the

85t

h A

nnua

l M

eetin

g of

the

Am

eric

an S

ocie

ty o

f M

amm

alog

ists

- So

uthw

est

Mis

sour

i Sta

te U

nive

rsity

. 33

. Dia

z, M

. M

., Fl

ores

, D

. A

., &

Bar

quez

, R

. M

. (2

002)

. A

new

spe

cies

of

Gra

cile

Mou

se O

poss

um,

genu

s G

raci

linan

us

(Did

elph

imor

phia

: Did

elph

idae

), fr

om A

rgen

tina.

Jou

rnal

of M

amm

alog

y, 8

3, 8

24–8

33.

34. P

ardi

ñas,

U.

F. J

., T

eta,

P.,

Cir

igno

li, S

. &

Pod

está

, D

. H

. (2

003)

Mic

rom

amífe

ros

(Did

elph

imor

phia

y R

oden

tia)

de

Nor

pata

goni

a ex

tra

andi

na, A

rgen

tina:

Tax

onom

ía a

lfa y

bio

geog

rafía

. M

asto

zool

ogía

Neo

trop

ical

, 10,

69-

113.

35

. Cos

ta,

L.

P.

(200

3) T

he h

isto

rica

l br

idge

bet

wee

n th

e A

maz

on a

nd t

he A

tlant

ic F

ores

t of

Bra

zil:

a st

udy

of m

olec

ular

ph

ylog

eogr

aphy

with

sm

all m

amm

als.

Jour

nal o

f Bio

geog

raph

y, 3

0, 7

1–86

52

Appendix S4 – Species for which we were unable to find any direct or indirect indication of the species’ size. The compromise body mass values correspond to the average mass of each genus.

Family Genus Species Mass (g)

Muridae Brucepattersonius soricinus 33.00 Echimyidae Trinomys yonenagae 235.28 Muridae Thomasomys macrotis 79.09 Muridae Thomasomys apeco 79.09 Muridae Thomasomys oreas 77.27 Muridae Rhipidomys scandens 81.43 Geomyidae Orthogeomys matagalpae 494.13 Muridae Oryzomys emmonsae 62.88 Marmosidae Monodelphis maraxina 109.33 Callitrichidae Mico mauesi 390.95 Callitrichidae Mico marcai 390.95 Echimyidae Mesomys occultus 124.75 Muridae Juscelinomys huanchacae 97.30 Muridae Juscelinomys guaporensis 97.30 Erethizontidae Coendou sneiderni 1985.29 Caviidae Kerodon acrobata 800.00 Muridae Punomys kofordi 84.80 Muridae Microakodontomys transitorius 37.00 Echimyidae Trinomys moojeni 235.28 Echimyidae Proechimys kulinae 291.91 Echimyidae Proechimys echinothrix 291.91

53

Capítulo 3 Patrones a gran escala de tamaño corporal en reptiles escamados de Europa y Norte América Este capítulo reproduce íntegramente el texto del siguiente manuscrito: Olalla-Tárraga, M. Á., M. Á. Rodríguez & B. A. Hawkins. 2006. Broad-scale patterns of body size in squamate reptiles of Europe and North America. Journal of Biogeography 33: 781–793.

ORIGINALARTICLE

Broad-scale patterns of body sizein squamate reptiles of Europe andNorth America

Miguel A. Olalla-Tarraga1,2*, Miguel A. Rodrıguez1 and

Bradford A. Hawkins3

INTRODUCTION

Since Bergmann (1847) suggested that among closely related

endothermic animals those living in colder environments tend

to be larger than those in warmer environments, numerous

studies have explored spatial variation in body size, resulting in

a long and lively debate with respect to this ecogeographical

rule. However, despite some doubts about its existence

(Scholander, 1955; McNab, 1971; Geist, 1987), both the

intraspecific and interspecific versions of Bergmann’s rule

have received broad support in mammals (Ashton et al., 2000;

Meiri & Dayan, 2003; Blackburn & Hawkins, 2004; Rodrıguez

et al., 2006) and birds (James, 1970; Cousins, 1989; Blackburn

& Gaston, 1996; Ashton, 2002; Meiri & Dayan, 2003). Even so,

two important basic issues remain. First, whereas intraspecific

gradients in body size have been documented hundreds of

times, tests for interspecific clines are scarcer. Second, the

generalizability of geographical gradients in body size for

ectotherms remains controversial. Lindsey (1966) demon-

strated that the latitudinal variation in body size often

documented for endotherms could also be extended to at

least some ectothermic organisms. However, subsequent tests

of body size gradients in ectotherms have found a range of

patterns.

Among the seven interspecific analyses that have been

conducted for invertebrates of which we are aware (Miller,

1Department of Ecology, University of Alcala,

Alcala de Henares, Spain, 2Department of

Ecology, Universidad Autonoma de Madrid,

Madrid, Spain and 3Department of Ecology

and Evolutionary Biology, University of

California, Irvine, CA, USA

*Correspondence: Miguel Angel Olalla-Tarraga,

Departamento de Ecologıa, Facultad de

Ciencias, Universidad de Alcala, Ctra. Madrid-

Barcelona Km, 33,600, 28871, Alcala de Henares,

Madrid, Spain.

E-mail: [email protected]

ABSTRACT

Aim To document geographical interspecific patterns of body size of European

and North American squamate reptile assemblages and explore the relationship

between body size patterns and environmental gradients.

Location North America and western Europe.

Methods We processed distribution maps for native species of squamate reptiles

to document interspecific spatial variation of body size at a grain size of

110 · 110 km. We also examined seven environmental variables linked to four

hypotheses possibly influencing body size gradients. We used simple and multiple

regression, evaluated using information theory, to identify the set of models best

supported by the data.

Results Europe is characterized by clear latitudinal trends in body size, whereas

geographical variation in body size in North America is complex. There is a

consistent association of mean body size with measures of ambient energy in both

regions, although lizards increase in size northwards whereas snakes show the

opposite pattern. Our best models accounted for almost 60% of the variation in

body size of lizards and snakes within Europe, but the proportions of variance

explained in North America were less than 20%.

Main conclusions Although body size influences the energy balance of

thermoregulating ectotherms, inconsistent biogeographical patterns and

contrasting associations with energy in lizards and snakes suggest that no single

mechanism can explain variation of reptile body size in the northern temperate

zone.

Keywords

Bergmann’s rule, body size gradients, Europe, macroecology, North America,

squamate reptiles, thermoregulation.

Journal of Biogeography (J. Biogeogr.) (2006) 33, 781–793

ª 2006 The Authors www.blackwellpublishing.com/jbi 781Journal compilation ª 2006 Blackwell Publishing Ltd doi:10.1111/j.1365-2699.2006.01435.x

54

1991; Cushman et al., 1993; Barlow, 1994; Hawkins, 1995;

Hawkins & Lawton, 1995; Diniz-Filho & Fowler, 1998;

Hausdorf, 2003), decreasing body size with decreasing

latitude was found only for European ants (Cushman et al.,

1993), whereas for ectothermic vertebrates seven of eleven

data sets were consistent with Bergmann’s rule (Lindsey,

1966; McDowall, 1994; Cruz et al., 2005). Four of the seven

studies supporting the rule were for fish, and two were for

amphibians; support for Bergmann’s rule has only been

documented once for reptiles (Cruz et al., 2005), although

there have been few studies (Lindsey, 1966; Reed, 2003).

An important challenge in macroecology is to identify

and understand the large-scale variation of ecologically

relevant characteristics of organisms, such as body size,

along environmental gradients (Brown, 1995; Gaston &

Blackburn, 2000). However, as recently noted by Reed

(2003), reptiles have rarely been studied in the macro-

ecological literature. Therefore, describing patterns of

geographical variation in body size may generate insights

into the evolutionary and ecological mechanisms structuring

reptile assemblages while extending our understanding of

macroecological patterns beyond the more intensely studied

mammals and birds.

In this paper we test for the existence of broad-scale

interspecific patterns of body size in European and North

American lizards and snakes, using a grid approach.

A comparison of patterns in two regions can help to evaluate

the generalizability of the observed geographical trends and to

find region-specific differences (Schall & Pianka, 1978;

Murphy, 1985; Hawkins & Lawton, 1995). Our primary goal

is to determine if the two most speciose reptile groups follow

the pattern described by Bergmann’s rule. Further, we explore

the relationships between the observed patterns in body size

and environmental gradients that might account for geo-

graphical patterns of mean body size for these ectotherms.

Specifically, we focus the analysis on four relevant hypotheses

likely to explain large-scale gradients in body size (see e.g.

Cushman et al., 1993; Blackburn et al., 1999; Blackburn &

Hawkins, 2004):

1. Heat conservation: This is the classic physiological explana-

tion for Bergmann’s rule for endotherms. According to this

hypothesis large-bodied species can occupy northern latitudes

due to their reduced surface-to-volume ratio. This has been

considered unlikely to explain latitudinal clines in body size for

ectotherms (Cushman et al., 1993). However, reptiles actively

thermoregulate by behavioural and physiological mechanisms,

which can give them control over metabolic processes, and this

hypothesis may apply to reptiles despite their being ectother-

mic.

2. Migration abilities: This hypothesis proposes that small

species will be underrepresented at high latitudes because they

have failed to fully colonize these regions following the retreat

of the glaciers at the close of the Pleistocene. We are not aware

of any data demonstrating that small-bodied reptiles have

more limited dispersal abilities than large-bodied species, but it

is plausible and deserves to be tested.

3. Primary productivity: Rosenzweig (1968a) argued that

primary productivity could be an important selective pressure

on body size since body mass must be maintained by a

sufficient food supply. Therefore, this predicts that measures of

plant productivity should best describe variation in body size.

4. Starvation resistance: This is also sometimes referred to as

the resource availability hypothesis. It has been suggested that

measures of seasonality best predict variation in body size

(Boyce, 1978; Lindstedt & Boyce, 1985; Wigginton & Dobson,

1999; Ashton, 2001). This hypothesis assumes that larger body

mass is advantageous where there is greater seasonality in

resource abundance, because resistance to starvation increases

with body size via allometric scaling of fat reserves (Cushman

et al., 1993; Blackburn et al., 1999). This reasoning applies to

endotherms, but in the case of ectotherms a more plausible

mechanism relates the length of the growing season to the

physiological time available for development (Mousseau,

1997). In seasonal environments resources are available for

less time, and therefore animals have less time to grow. This

will be less of a problem for small species, as they require fewer

resources than larger species. Thus, small species can persist in

more seasonal environments.


Species data

Distribution maps for native squamate reptiles were obtained

from Gasc et al. (1997) and two field guides to the North

American herpetofauna (Conant & Collins, 1998; Stebbins,

2003). All islands, except Great Britain, were excluded. Data

from Belarus, Russia and Ukraine in continental Europe were

discarded because of incomplete sampling (Gasc et al., 1997;

Meliadou & Troumbis, 1997; Araujo et al., 2001). Maps were

digitized and processed using ArcGIS 8.3 in a grid comprising

110 · 110 km cells. Coastal cells containing less than 50% of

the land mass of inland cells were excluded from the analysis.

The mapped area included 386 cells in Europe and 1430 in

North America.

After removing island endemics, exotic and extinct species,

the reptile database comprised 71 species in Europe (28 snakes

and 43 lizards) and 224 species in North America (124 snakes

and 100 lizards). Lacerta vivipara in Europe and Thamnopis

sirtalis in North America were also excluded from the analysis

due to their extraordinary abilities to freeze and supercool. It is

well known that these species are freeze-tolerant, a physiolo-

gical trait that has been adduced to explain their presence in

colder climates (Churchill & Storey, 1992; Costanzo et al.,

1995; Grenot et al., 2000). Since our study seeks to establish

the role of body size in determining reptile distributions, we

excluded these species because it is already known that their

distribution is associated with another trait. Also, these are the

only reptiles found in central Canada or northern Scandinavia,

so including them would have generated many cells with

average body sizes due solely to their presence; thus it was

deemed prudent to exclude them from the analysis.

M. A. Olalla-Tarraga, M. A. Rodrıguez and B. A. Hawkins

782 Journal of Biogeography 33, 781–793ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd

55

Body sizes were extracted from field guides (Conant &

Collins, 1998; Arnold, 2002; Stebbins, 2003). The masses of

reptiles are rarely available in the literature, and previous

studies used body length as a measure of body size (Boback &

Guyer, 2003; Reed, 2003). Indeed, it has been suggested that

both body length and body mass generate similar results in

analyses of macroecological patterns (Brown, 1995). However,

some authors have stressed the fact that mass represents the

best estimate of body size (Hedges, 1985; Gaston & Blackburn,

2000), and it is especially important to use mass when making

comparisons among organisms with different basic body

shapes, as occurs in the case of lizards and snakes. Likewise,

the occurrence of some serpentiform species among lizards

may influence the overall patterns for this group if we had used

body length as an estimate of body size. Therefore, we used the

formulae proposed by Pough (1980) to convert length to mass

for reptiles. These formulae represent the best models available

for this conversion (F. H. Pough, pers. comm.). We used mass

(g) ¼ 3.1 · 10)2 length (cm)2.98 for lizards, where length is

measured snout-to-vent, and mass (g) ¼ 3.5 · 10)4 length

(cm)3.02 for snakes, where length is snout-to-tail. The masses

of serpentiform lizards were estimated from the equation for

snakes (Pough, 1980). All members of the Anguidae (three

species) and Scincidae (six species) in Europe, as well as

Anniella pulchra (Anniellidae) and the four members within

the genus Ophisaurus (Anguidae) in North America were

considered serpentiform. We used maximum lengths for each

species. In organisms with indeterminate growth maximum

values are considered to be a reasonable estimate of the size

potential for a species (Boback & Guyer, 2003; Reed, 2003). It

should be noted that this analysis does not incorporate

intraspecific variation in body size, since we assign the same

value for body size to the whole geographical range occupied

by a species. The inclusion of intraspecific variation in body

size in our analyses could potentially reinforce or distort the

observed interspecific patterns, depending on the size trends

observed in each case. However, intraspecific data could not be

included because such data exist only for a relatively small

number species (see Electronic Appendix in Ashton &

Feldman, 2003).

Body size was log10 transformed, and we calculated average

log mass (hereafter called the mean body size) in each grid cell

separately for lizards and snakes occurring in each geographical

region. Obviously, this does not reflect phylogeny because

snakes developed from lizards and the latter are thus

paraphyletic in relation to the former. Rather, it is intended

to represent two functional groups that differ markedly in

morphology and food habits (see e.g. Pough & Groves, 1983).

Additionally, most lizard species (around 80%) weigh less than

20 g, while nearly 75% of snakes have adult body masses

greater than 20 g (Pough, 1980; Pough & Groves, 1983).

Environmental variables

We generated seven environmental variables, selected because

they can be related to four relevant hypotheses likely to explain

large-scale gradients in body size (see e.g. Cushman et al.,

1993; Blackburn et al., 1999; Blackburn & Hawkins, 2004). The

variables (with their associated hypotheses) are as follows:

1. Mean annual temperature, annual potential evapotranspira-

tion and range in elevation (heat conservation): Mean annual

temperature and annual potential evapotranspiration (PET)

are both widely used as measures of ambient energy inputs

(Schall & Pianka, 1978; Currie, 1991; Rodrıguez et al., 2005,

2006). PET is an estimate of the net atmospheric energy

balance and is highly correlated with mean annual temperature

and solar radiation (Currie, 1991). We used annual PET data

calculated using the Priestley–Taylor formula, which are

available at http://www.grid.unep.ch/data/summary.php?

dataid¼GNV183. Temperature data were taken from http://

www.grid.unep.ch/data/summary.php?dataid¼GNV15.We also

used range in elevation as an estimate of mesoscale spatial

climatic variation within the cells (Turner & Hawkins, 2004).

This was calculated as the difference between maximum and

minimum elevation within a grid cell from data available at

http://www.ngdc.noaa.gov/seg/cdroms/ged_iia/datasets/a13/

fnoc.htm.

2. Time since glacial retreat (migration ability): We tested this

in Europe using the time since the most recent glaciation when

a cell became available for colonization (hereafter called the

cell age). Age was estimated using changes in ice cover at 1 ka

intervals from Peltier (1993). Cells that were not completely

covered by ice during the most recent glacial maximum were

assigned an arbitrary age of 20 ka (Hawkins, 2004; Rodrıguez

et al., 2006), which represents the age corresponding to the

time of maximum ice coverage in Europe. This hypothesis

could not be tested in North America because most of Canada

is without reptiles (see Fig. 1c,d), except for T. sirtalis, which

we excluded because it can tolerate body freezing.

3. Annual actual evapotranspiration, global vegetation index

(primary productivity): Annual actual evapotranspiration

(AET) measures water–energy balance, which drives plant

growth. Because of the strong relationship between AET and

plant productivity at large scales, this variable has been used as

a proxy for productivity (Rosenzweig, 1968b; Lieth, 1975;

Hawkins et al., 2003; Hawkins, 2004). The global vegetation

index (GVI) is derived from radiometer data from the NOAA

Polar Orbiting Environmental Satellites (Kineman & Hastings,

1992) and measures standing crop and the greenness of the

plant canopy. Therefore, it has been widely used as a proxy for

plant productivity or standing crop (Hurlbert & Haskell, 2003;

Hawkins, 2004; Rodrıguez et al., 2005). We generated annual

data for both variables. AET (Thornthwaite formula) is

available at http://www.grid.unep.ch/data/summary.php?

dataid¼GNV183 and GVI is available at http://www.ngdc.

noaa.gov/seg/cdroms/ged_iia/datasets/a01/mgv.htm.

4. Length of growing season (starvation resistance): We tested

this hypothesis using the number of months available for plant

growth in each grid cell. We followed the reasoning underlying

Gaussen ombrothermic climatic diagrams to generate this

variable (Gaussen, 1954). First, we calculated the xerothermic

season length for each cell by noting the number of months

Interspecific body size patterns in squamate reptiles

Journal of Biogeography 33, 781–793 783ª 2006 The Authors. Journal compilation ª 2006 Blackwell Publishing Ltd

56

when two times the mean monthly temperature was higher

than mean monthly precipitation. We thus eliminated those

summer months for which low precipitation limits plant

productivity. For the remaining months, we calculated the

number of months in which plant growth is limited by low

temperatures. We plotted cell monthly values of GVI against

mean monthly temperatures and used split-line regression to

find the threshold temperature value above which plant growth

occurs. We then counted the number of months having mean

monthly temperatures greater than this value in each cell, and

added this to the number of months with sufficient precipi-

tation.


The data for lizards and snakes were analysed separately.

Initially, we used simple regressions to test for relationships

between mean body size and the seven explanatory variables.

However, as is always the case with the grid approach, cells are

not statistically independent, resulting in an underestimation

of error variances. To obtain unbiased estimates of the

significance of the simple regressions we used the modified t-

test of Dutilleul (1993). This method corrects significance tests

for spatial autocorrelation using correlograms to estimate the

geographically effective degrees of freedom (see e.g. Legendre

et al., 2002).

We then used multiple regression to generate environmen-

tal models including multiple predictors, using a model

selection approach based on information theory (see Burham

& Anderson, 2002) to identify the set of models best

supported by the data. The use of model selection, a statistical

approach that is rapidly gaining support in ecology as an

alternative to hypothesis testing (Johnson & Omland, 2004),

allows us to evaluate the relative support for each hypothesis

by comparing a complete set of competing models. Taking

into account the strong collinearity among several of the

predictors in our data set (Table 1), it is especially necessary to

assess simultaneously the importance of all the various

(a) (b)

(c) (d)

Figure 1 Geographical patterns of species

richness at 110 · 110 km. (a) Snakes in

Europe (s ¼ 28 species). (b) Lizards in

Europe (s ¼ 43 species). (c) Snakes in North

America (s ¼ 124 species). (d) Lizards in

North America (s ¼ 100 species). These

maps do not include Lacerta vivipara in

Europe and Thamnopis sirtalis in North

America which were excluded from the ana-

lysis (see Materials and methods).



57

predictors included in the analysis. The Akaike information

criterion (AIC) was used to rank competing models and weigh

the relative support for each one. However, the presence of

spatial autocorrelation in the residuals of all our models

required us to correct the error variances before calculating

the AICs. We accomplished this by calculating geographically

effective sample sizes (n*), given by n* ¼ n/[(1 + p)/(1 ) p)],

where p is the first-order autoregressive parameter of the

residuals, approximated by the standardized Moran’s I in the

first distance class (Cressie, 1993; Haining, 2003). Spatial

correlograms of residuals were calculated using 17 distance

classes in Europe and 19 in North America, and the first class

ranged from 0 to 345 km and 0 to 451 km, respectively.

Approximate unbiased variances were obtained dividing the

residual sum of squares by n*, which were then used to

calculate corrected AICs.

We compared the resulting AIC values of each model using

DAIC, the difference between AICs of each model and

the minimum AIC found. A value of DAIC higher than

10 represents a poor fit relative to the best model, whereas a

value less than 2 indicates that a model is equivalent to the

minimum AIC model.

These DAIC values were also used to calculate Akaike’s

weighting of each model (wi), which can be interpreted as the

probability that model i is actually the best explanatory model.

The values of wi are standardized across the candidate set of

models.

Weightings also provide a way to define the relative

importance of each predictor. This measure can be calculated

as the sum of wi values over all models that include the

predictor of interest. However, it is virtually impossible to

discern the relative influence of different predictors when all of

them appear in the best set of models, so we used the

standardized regression coefficients to rank the importance of

the predictors in the best model (J. A. F. Diniz-Filho, pers.

comm.).

Finally, we used the correlograms of the residuals of our best

multiple-regression models to evaluate how these environ-

mental models control spatial variation in body size across

spatial scales (Diniz-Filho et al., 2003). We restricted this

analysis to Europe, where multiple-regression models had high

explanatory power (see Results). To do this we generated

spatial correlograms using Moran’s I coefficients calculated for

original mean body size data and residual mean body size after

fitting each model at 17 distance classes. Thus, the lower the

level of spatial autocorrelation for the residuals of the multiple-

regression models at any distance class, the greater the capacity

of the model to explain spatial structure in mean body size at

that distance. On the contrary, remaining spatial autocorrela-

tion at a distance class in the residuals of the fitted regression

models indicates the inadequacy of the model to describe the

body size pattern at that scale and, therefore, suggests that

spatially patterned variables not included in the model are

contributing to the spatial pattern.

All statistical analyses were performed using statistica

(StatSoft, Inc., 2003) and Spatial Analyses in Macroecology

(SAM) (Rangel et al., 2005).

RESULTS

Before describing the patterns in body size and their relation-

ships with environmental variables, it should be noted that

these patterns are potentially sensitive to spatial variation in

species richness. Thus, the lower cell occupancies in species-

poor areas may influence variation in mean body size because

of the smaller range of sizes present in these cells, which should

Table 1 Correlation matrices for environmental variables. Significance levels are corrected for spatial autocorrelation using the modi-

fied t-test developed by Dutilleul (1993) (*P < 0.05; **P < 0.01): (a) North America and (b) Europe

Variable Temp. AET PET Range GVI GSL

(a) North America

Mean annual temperature 1

Actual evapotranspiration 0.8* 1

Potential evapotranspiration 0.94** 0.73 1

Range in elevation )0.04 0.30 0.02 1

Global Vegetation Index 0.50* 0.62** 0.39 )0.03 1

Length of growing season 0.69* 0.88** 0.61 0.33 0.58** 1

Variable Temp. AET PET Range GVI GSL Age

(b) Europe

Mean annual temperature 1

Actual evapotranspiration 0.44 1

Potential evapotranspiration 0.79* 0.67* 1

Range in elevation )0.08 0.29 0.30 1

Global Vegetation Index 0.60 0.74* 0.58 0.06 1

Length of growing season 0.75** 0.37 0.43 )0.16 0.69* 1

Cell age 0.76* 0.71 0.74 0.18 0.71 0.67 1



58

be taken into account when interpreting trends in body size.

Therefore, we also present maps of species richness for lizards

and snakes on both continents (Fig. 1).

Bearing this in mind, Europe is characterized by clear

latitudinal trends in body size for both snakes and lizards,

although the gradients are in opposite directions (Fig. 2a,b),

i.e. whereas mean body size for lizards increases northwards,

snakes become larger towards the south. Patterns in North

America are much more complex and reveal no consistent

gradients in latitudinal space (Fig. 2c,d), although it is possible

to pick out patterns in some parts of the continent (e.g. both

snakes and lizards become larger towards the south in the

east). There is also a marked longitudinal component within

North America, with snakes being largest in the western half of

the continent.

Within Europe, after correcting the probabilities of the

simple regression coefficients between body size and the

environmental predictors for spatial autocorrelation, only PET

and mean annual temperature were significant for both groups

(Tables 2 and 3). Because the body size gradients were in

opposite directions for the two groups, with snakes smaller and

lizards larger in the north, the relationships with these two

variables were also in opposite directions (Fig. 3). PET was the

strongest predictor of mean log body size in Europe for lizards,

with mean annual temperature ranking second (Table 2).

Similarly, these two measures of energy availability were the

best single predictors for the mean size of snakes in Europe.

However, mean annual temperature accounted for 2% more

variance than PET in this case (Table 3).

Even though mean annual temperature and PET are highly

correlated (Table 1), both of them entered in all multiple-

regression models with DAIC £ 2 for mean body size of snakes

in Europe (Table 4). Based on this criterion alone we obtained

a set of four equivalent models that accounted for the same

amount of variance (59.4%). However, the Akaike weightings

suggest that the first model, which includes mean annual

temperature, PET and AET, is a better model (wi ¼ 0.427).

The higher standardized coefficients of mean annual

(c) (d)

(a) (b)

Figure 2 Geographical patterns of squamate

reptile mean body size in Europe and

North America. (a) Snakes in Europe

(s ¼ 28 species). (b) Lizards in Europe

(s ¼ 43 species). (c) Snakes in North

America (s ¼ 124 species). (d) Lizards in

North America (s ¼ 100 species). Numbers

included in the legend of each map are

back-transformed from average log-trans-

formed mass values.



59

temperature and PET provide strong evidence that these two

variables are mainly driving broad-scale variation in body size

of snakes in Europe.

The pattern of spatial autocorrelation for mean body size of

snakes in Europe was representative of a cline, with positive

autocorrelation at shorter distances and progressively becom-

ing negative at larger distances (Fig. 4a). The model including

mean annual temperature, PET and AET accounted for most

of this pattern at all distance classes except for the shortest one.

This indicates that a factor not included in the model

influences body size patterns for snakes at more local scales

in Europe.

Of the 128 possible multiple-regressions models for lizards

in Europe, three had an DAIC £ 2 and explained similar

proportions of variance (Table 4). Again, Akaike weightings

indicate that the first model is the best model (wi ¼ 0.576). It

included PET, cell age and GVI and explained 62.5% of the

variance in mean body size. In this model, the standardized

coefficients indicate that PET is consistently the strongest

predictor, followed by age and GVI. Nonetheless, it should be

noted that age and GVI together increase the variance

explained by the model by less than 4%, which lead us to

conclude that PET is the main driver of variation in body size

of lizards in Europe.

Similar to the case of snakes, the spatial autocorrelation

analysis shows a clinal pattern of variation of lizard body size

(Fig. 4b). The regression model including PET, age and GVI

accounted for most of this pattern at all distance classes except

for the shortest one. Again, this suggests that additional factors

not included in our analysis are necessary to fully explain

variation in lizard body size at more local scales in Europe.

In North America, the simple regressions were much weaker

than in Europe, and none of the environmental predictors

remained significant after correcting for spatial autocorrelation

in simple regressions (Tables 5 and 6). Energy variables (PET

and mean annual temperature for lizards and range in

elevation in the case of snakes) also represented the best single

predictors of lizard body size in North America, and the signs

of the regression coefficients were identical on both continents.

However, none of the six variables (we did not test for the

influence of cell age in this region) explained more than 10%

of the variance in body size in either group, which might be

expected given the complex spatial patterns found on this

continent.

Of the 64 possible models for each group, the models with

the lowest AIC included all variables except length of the

growing season in the case of snakes and PET, AET and length

of the growing season for lizards. Based on Akaike weightings

Table 2 Simple regressions of predictor variables against lizards

mean body size in Europe. Predictors are ranked by their coeffi-

cient of determination. Corrected probabilities are based on the

modified t-test developed by Dutilleul (1993)

Variable r2

Probabilities

Uncorrected Corrected (d.f.)

)Potential evapotranspiration 0.586 < 0.0001 0.009 (8.0)

)Mean annual temperature 0.351 < 0.0001 0.049 (9.4)

)Age 0.310 < 0.0001 0.056 (10.2)

)Actual evapotranspiration 0.137 < 0.0001 0.070 (22.5)

)Range in elevation 0.076 < 0.0001 0.078 (39.6)

)Global Vegetation Index 0.050 < 0.0001 0.298 (21.5)

)Length of growing season 0.037 < 0.0001 0.388 (20.4)

d.f. ¼ geographically effective degrees of freedom. Total number of

analysed cells ¼ 345.

Table 3 Simple regressions of predictor variables against snakes

mean body size in Europe. Presentation as in Table 2. Total

number of analysed cells ¼ 382

Variable r2

Probabilities


+Mean annual temperature 0.522 < 0.0001 0.032 (6.7)

+Potential evapotranspiration 0.504 < 0.0001 0.043 (6.3)

+Age 0.458 < 0.0001 0.058 (6.3)

+Global Vegetation Index 0.286 < 0.0001 0.069 (10.2)

+Length of growing season 0.277 < 0.0001 0.051 (11.4)

+Actual evapotranspiration 0.275 < 0.0001 0.069 (10.7)

+Range in elevation 0.008 0.0399 0.541 (34.1)

Ave

rage

log

bod

y m

ass

10

(a)

Potential evapotranspiration (mm / year)

(b)

1.9

2

2.1

2.2

2.3

2.4

2.5

2.6

2.7

2.8

2.9

0 200 400 600 800 1000 1200

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

0 200 400 600 800 1000 1200

Potential evapotranspiration (mm / year)

Ave

rage

log

bod

y m

ass

10

Figure 3 The relationship between squamate reptile mean body

size and potential evapotranspiration (PET) in 110 · 110 km cells

for Europe. (a) Snakes (s ¼ 28 species). (b) Lizards (43 species).

Similar relationships were obtained for mean annual temperature.



60

both models were the best among equivalent models

(wi ¼ 0.731 and wi ¼ 0.576, respectively). However, the

former explained only 19.6% of variance, whereas the latter

accounted for 5% of the variance.

DISCUSSION

Our results reveal clear, albeit group-specific, patterns of

interspecific body size variation in squamate reptiles through-

out Europe. In contrast, patterns are more complex and

patchier in North America. Despite this, associations of mean

body size with measures of ambient energy are similar in sign

Table 4 Multiple-regression models for mean body size (data resolved to 110 · 110 km). Models are ranked in each case by AIC from best

to worst-fitting model, and only the models with DAIC < 2 are presented. AICs have been corrected for the presence of spatial auto-

correlation in the model residuals. For each variable entering in the model we include their standardized coefficients to evaluate the relative

importance of each one. In the case of North American lizards we only show the best of seven equivalent models, because none of them

explained more than 6% of variance. Predictor variable codes are: Temp., mean annual temperature; PET, potential evapotranspiration;

Elev., range in elevation; Age, time since glacial retreat; AET, actual evapotranspiration; GVI, Global Vegetation Index; GSL, length of

growing season

Group Region

Predictors in model

AIC DAIC r2 wiTemp. PET Elev. Age AET GVI GSL

Snakes Europe 0.482 0.205 0.188 )659 0 0.594 0.427

0.418 0.246 0.169 0.054 )658 1 0.594 0.259

0.487 0.202 0.195 0.009 )657 2 0.594 0.157

0.501 0.183 0.023 0.188 )657 2 0.594 0.157

North America 0.451 )0.305 0.474 0.173 )0.331 )474 0 0.196 0.731

0.451 )0.306 0.471 0.194 )0.328 )0.027 )472 2 0.196 0.269

Lizards Europe )0.708 )0.236 0.220 )608 0 0.625 0.576

)0.713 0.013 )0.236 0.221 )606 2 0.625 0.212

)0.721 )0.211 0.244 )0.051 )606 2 0.627 0.212

North America )0.232 0.236 )0.204 )743 0 0.050 0.234

–1.2–1

–0.8–0.6–0.4–0.2

00.20.40.60.8

0 1000 2000 3000 4000 5000

Distance (km)

Mor

an's

I

(a)

–1

–0.8

–0.6

–0.4

–0.2

0

0.2

0.4

0.6

0.8

0 1000 2000 3000 4000 5000

Distance (km)

Mor

an's

I

(b)

Figure 4 Spatial correlograms using Moran’s I for mean body

size (solid circles) and residuals of the best multiple-regression

models in Europe (open circles): (a) snakes, (b) lizards.

Table 5 Simple regressions of predictor variables against mean

body size of snakes in North America. Presentation as in Table 2.

Total number of analysed cells ¼ 788

Variable r2

Probabilities


+Range in elevation 0.102 < 0.0001 0.262 (11.9)

)Global Vegetation Index 0.047 < 0.0001 0.288 (23.0)

)Length of growing season 0.035 < 0.0001 0.516 (11.9)

)Actual evapotranspiration 0.024 < 0.0001 0.626 (9.5)

+Potential evapotranspiration 0.023 < 0.0001 0.569 (14.6)

+Mean annual temperature 0.001 < 0.0001 0.729 (12.8)

Table 6 Simple regressions of predictor variables against mean

body size of lizards in North America. Presentation as in Table 2.

Total number of analysed cells ¼ 658

Variable r2

Probabilities


)Potential evapotranspiration 0.039 < 0.0001 0.220 (37.7)

)Mean annual temperature 0.036 < 0.0001 0.229 (38.6)

)Length of growing season 0.003 0.102 0.676 (43.3)

)Global Vegetation Index 0.001 0.211 0.719 (57.1)

)Actual evapotranspiration 0.001 0.336 0.804 (44.7)

)Range in elevation 0.001 0.950 0.985 (62.2)



61

in both regions. Specifically, PET was the best predictor of

mean body size for lizards in both continents, whereas mean

annual temperature, together with range in elevation in North

America, is the strongest predictor of size variation in snakes.

Moreover, the slope of the relationship between energy

variables and mean body size was positive for snakes and

negative for lizards in both regions. The proportions of

variance described by these variables were notably larger in

Europe in both cases (Europe > 57% and North America

< 10%).

The importance of energy in explaining the observed

gradients in body size was also indicated within our mul-

tiple-regression models; in all cases, the addition of extra

explanatory variables only increased slightly the amount of

variance explained by energy alone. However, our finding of

strong trends in body size in Europe and weak patterns in

North America is particularly intriguing and limits a general

explanation. Indeed, the very low explanatory power of our

models to describe variation in body size for both reptile

groups in North America strongly suggests that factors as yet

unknown account for the existence of complex spatial patterns

in this region. As a consequence, it is impossible to derive a

single explanation for such patterns of variation in body size.

A second limitation in interpreting these results is the

paucity of studies that have examined interspecific biogeo-

graphical patterns of body size in reptiles. In a pioneering

study, Lindsey (1966) tested for Bergmann’s rule across 1222

snake species and 935 lizard species throughout the world. He

found no latitudinal trends, although he reported that some

snake families weakly followed Bergmann’s rule. In spite of the

taxonomic breadth of his analysis, the results are difficult to

compare directly with ours because he did not include

complete regional faunas (Lindsey, 1966; Currie & Fritz,

1993; McDowall, 1994), and the data were binned into three

broad latitudinal bands (i.e. cool temperate, warm temperate

and tropical) without differentiating between biogeographical

regions. Indeed, the region-specific differences we find for

Europe and North America suggest that broad-scale patterns in

body size may be confounded when biogeographical regions

are pooled. Similarly, Hawkins & Lawton (1995) found

opposing patterns of variation in body size of butterflies

between continents, and concluded that evaluations of

Bergmann’s rule based on a single region may not be

generalizable to other regions.

Lizards

Lizards follow Bergmann’s rule in Europe. Moreover, mean

body size is negatively correlated with energy variables in both

Europe and North America, consistent with the heat conser-

vation hypothesis. However, heat balance in ectotherms

depends on both heat gain and heat conservation, which

means that the explanation proposed by Bergmann (1847) for

endotherms can only partially account for the observed

gradients. So, what is the relationship between body size and

heating rates? Cowles (1945) and Bogert (1949) (cited in

Ashton & Feldman, 2003) hypothesized that smaller squamate

reptiles are favoured in colder environments because their

increased surface to volume ratio allows them to heat more

rapidly. If true, we are confronted with two opposing forces in

cold climates, one that favours larger sizes (i.e. Bergmann’s

explanation to conserve internal temperature), and the other

favouring smaller sizes to gain heat. To further complicate

matters, it is well established that heat balance in reptiles can

also be strongly affected by physiological and behavioural

adjustments (Cowles & Bogert, 1944; Bartholomew, 1982;

Huey, 1982). For example, in a study of eight lacertid lizards in

Europe, Dıaz et al. (1996) found that the four species living in

the north reduced their heating times as a result of both

changes in heating rates (a physiological trait) and

selection of basking sites matching their preferred body

temperature (a behavioural trait). Consequently, northern

species can warm at faster mass-specific rates than those living

in Mediterranean areas and are able to reduce basking time by

17%. Lacertids represent 28 of the 43 lizard species in Europe.

Therefore, if the abilities observed by Dıaz et al. (1996) in the

northern lacertids they studied are common, this could

account for the pattern of variation in body size on this

continent. In other words, heat gain may not be as strongly

dependent on lizard body size as it is on behavioural and

physiological traits.

Lacertidae do not occur in North America, which has twice

as many lizard species as Europe. Another major difference

between the lizard fauna of both continents is the number of

families (6 in Europe and 11 in North America). The broader

taxonomic breadth in North America might be responsible for

the more complex pattern of body size on this continent,

particularly bearing in mind that both Bergmann’s rule and the

converse have been reported in intraspecific studies of lizards

(Ashton & Feldman, 2003) and between congenerics

(Angilletta et al., 2004; Sears & Angilleta, 2004). Although

not strictly comparable to our analysis, the variable intraspe-

cific patterns suggest that individual lizard species respond

idiosyncratically to environmental variation. If so, greater

spatial heterogeneity of mean body sizes in the richer lizard

fauna of North America is not surprising. Additionally, North

America is larger and more environmentally complex and has

habitats that are not present in Europe, including deserts and

subtropical forests. This wide environmental variation might

also contribute to the patchwork pattern in North America.

However, in spite of these complexities, energy variables

remain the best predictors of body size patterns in this

continent, which suggests that similar mechanisms to those

described for Europe might still play a role in constraining

variation in body size of North American lizards, albeit not the

dominant one.

Recently, Cruz et al. (2005) tested for interspecific body

trends within the lizard genus Liolaemus of South America and

found evidence for Bergmann’s rule for the 34 species they

analysed. Also, they observed a strong negative relationship

between latitudinal variation in body size and thermal

variables, which led them to favour heat conservation as the



62

most plausible explanation. They suggested that both

the slower rate of cooling associated with larger sizes and the

increased rate of heat gain as a result of physiological and

anatomical mechanisms may explain why lizard species from

cold climates tend to be larger. This concurs with our results

for European lizards and the explanation for the pattern is

similar in both cases. Furthermore, Cruz et al. (2005) found

that variation in the strength of this pattern strongly depended

on the phylogenetic scale of the analysis. Thus, when they

analysed all species within the subgenera Liolaemus and

Eulaemus the relationship between body size and latitude was

weaker. Therefore, they concluded that although macroeco-

logical patterns in body size for these lizards are more likely to

be the result of ecological factors related to heat conservation,

these patterns are sensitive to phylogenetic scale. A similar

result has been found for different clades of snakes (Ashton,

2001; see below) and this may at least partially account for the

marked differences in strength of the gradients when compar-

ing Europe and North America.

Another explanation for a negative relationship between

body size and energy in lizards is the temperature–size rule

(Atkinson, 1996), which describes a trend amongst ectotherms

to grow faster and to reach smaller adult sizes at higher

temperatures. Although much of the evidence for this comes

from laboratory experiments (see Atkinson, 1994), it has also

been observed in natural conditions (Angilletta et al., 2004).

Angilletta et al. (2004) compiled data on maturation time

from 18 studies of the eastern fence lizard (Sceloporus

undulatus) at different latitudes and found that lizards from

colder environments reach larger adult sizes by delaying

maturity, a characteristic they linked to lower levels of

competition and/or predation, and, hence, juvenile mortality.

Again, extending intraspecific observations to interspecific

biogeographical patterns can only be speculative, but if further

research confirms that lizard species from high latitudes reach

larger adult sizes through delayed maturation, the mechanisms

underlying the temperature–size rule would represent an

additional explanation for the increased size in lizard faunas

living in colder areas.

Snakes

As with lizards, body size in snakes shows a clear latitudinal

gradient in Europe, with snakes being larger southwards, but

with a patchwork distribution in North America. Mean annual

temperature and PET, together with range in elevation in the

case of North America, which are linked to the heat conservation

hypothesis, were the best descriptors of snake body size in both

continents, albeit the influence of these variables is much

stronger in Europe. These findings give fuel to the hypothesis

that energy availability influences body size in squamate reptiles;

although other factors must be important in North America as

well. Indeed, it is noteworthy that there is a cluster of the largest

snake species in the western half of North America, which results

in a longitudinal size cline. Like latitude, longitude per se

provides no biological or ecological information about spatial

gradients (Hawkins & Diniz-Filho, 2004), and whatever

generates this gradient appears to be largely statistically

independent of all of our environmental variables. The

observation that range in elevation, a measure of altitudinally

driven climatic variation within an area, influences variation in

body size of snakes suggests that these organisms are responding

to local effects of cold, with the largest species occurring in

higher, colder spots located in thewestern half ofNorthAmerica.

However, this explains only 10%of the variance in body size, and

the causes for this pattern remain unclear.

Although most European and North American snakes belong

to the Viperidae and Colubridae, North America has more than

four times more species of snakes than Europe (124 and 28,

respectively). Ashton (2001) observed that sister species of

rattlesnakes (Crotalus viridis and Crotalus oreganus) had oppos-

ite latitudinal clines of body size and concluded that mixing data

from separate clades may obscure biogeographical patterns of

body size. This suggests that larger numbers of species in

interspecific studies probably result in noisier data, as more

responses to environmental variation are possible. Accordingly,

the complex patterns of body size in snakes in North America

may reflect the greater variety of biologies presumably present

within the richer snake fauna of this continent.

Although body size patterns of both lizards and snakes in

Europe are mainly explained by measures of energy availab-

ility, there are obvious differences between them (snakes

become larger southwards) and we can only speculate on the

mechanisms, since no previous interspecific studies have

explored them. In a meta-analysis of intraspecific studies,

Ashton & Feldman (2003) reported that the most common

trend of body size variation within squamate species is an

increase with increasing environmental energy, consistent with

what we find interspecifically. But the question remains of why

mean body sizes of snakes and lizards show opposite relation-

ships with energy. Based on the available evidence, this

discrepancy does not exist at an intraspecific level (Ashton &

Feldman, 2003). One possibility is the differences in body mass

of these groups. Pough (1980) and Pough & Groves (1983)

found that nearly 80% of living species of lizards weigh less

than 20 g, whereas 73% of snakes are larger than this.

Similarly, 75% of the lizards in our data sets weigh £ 50 g,

whereas the same proportion of snakes weigh ‡ 80 g. Zug et al.

(2001) suggested that the behavioural control of thermal

interactions is particularly important for small lizards, whereas

other mechanisms are more important for the heat balance of

larger species. Implicit in this is that large ectotherms might

not be able to warm their bodies rapidly enough to meet their

needs in the narrow activity windows available in cold

environments. Willemsen & Hailey (1999) argued that a

converse Bergmann’s rule pattern is more likely to occur in

large thermoregulating ectotherms since they have less time

available after thermoregulation and lower food intake in

colder latitudes (see also Avery, 1976, 1978). This suggests the

existence of a body size threshold below which body size

increases with decreasing environmental energy (as in lizards)

and above which the reverse occurs (as in snakes). Again, this



63

appears not to be an adequate explanation at an intraspecific

level, since size–latitude trends do not differ among large and

small species (Ashton, 2004).

To evaluate this hypothesis at an interspecific level we

combined lizard and snake data in North America and analysed

small and large species (defined as those below and above the

mean in the approximately log-normal body size distribution)

against PET. We found no support for a threshold in North

America (slopes were negative in sign in both cases), which

might be expected given the complex spatial variation in body

size for both groups on this continent. Thus, we cannot conclude

that interspecific trends in variation in body size are size-

dependent in squamate reptiles everywhere. However, there is a

body size threshold in Europe (see Fig. 3), which suggests that

this could still be a cause for the contrasting gradients found in

this region. Clearly, analyses in other biogeographical regions are

necessary to assess whether absolute organism size is driving

interspecific body size patterns of reptiles.

In sum, in North America and Europe lizards tend to be

larger and snakes smaller in low-energy regions, indicating that

Bergmann’s mechanism is insufficient to explain the body size

patterns of these groups. But this mechanism was proposed for

endotherms and does not explain how body size influences

heat gain. Our results for Europe suggest that squamates shift

from Bergmann’s rule patterns to the converse as body size

increases, reflecting the transition from a positive influence of

increasing size on heat conservation for small squamate groups

(lizards), to a negative influence when animals are too large for

efficient heat gain in cold environments (snakes). However,

our results for North America suggest that body size plays a

secondary role in determining thermoregulation in some areas,

and that complex body size patterns can be expected when

physiological and behavioural thermoregulatory mechanisms

prevail in a regional fauna. This in turn emphasizes that there

may be no simple answer, so we need to document the patterns

in other areas of the world if we are to understand the

importance of body size in determining the biogeographical

distribution of squamate species.

ACKNOWLEDGEMENTS

M.A.O.-T. and M.A.R. thank the University of California at

Irvine for their hospitality during the preparation of the data.We

also thank Almudena Mateos and Irene L. Lopez for assistance

with data preparation, and J. Alexandre Diniz-Filho for assist-

ance with the statistical analysis. Two anonymous referees

provided valuable comments on the manuscript. This study was

supported by the Spanish CICYT (grant REN2003-03989/GLO

to M.A.R.). B.A.H. was supported by the Spanish Secretarıa de

Estado de Universidades e Investigacion (sabbatical grant SAB-

2003-0213).

REFERENCES

Angilletta, M.J., Niewiarowski, H.P., Dunham, A.E., Leache,

A.D. & Porter, W.P. (2004) Bergmann’s clines in ectotherms:

illustrating a life-history perspective with sceloporine lizards.


Araujo, M.B., Humphries, C.J., Densham, P.J., Lampinen, R.,

Hagemeijer, W.J.M., Mitchell-Jones, A.J. & Gasc, J.P. (2001)

Would environmental diversity be a good surrogate for

species diversity? Ecography, 24, 103–110.

Arnold, E.N. (2002) A field guide to the reptiles and amphibians

of Britain and Europe. HarperCollins, London.

Ashton, K.G. (2001) Body size variation among mainland

populations of the western rattlesnake (Crotalus viridis).

Evolution, 55, 2523–2533.

Ashton, K.G. (2002) Patterns of within-species body size

variation of birds: strong evidence for Bergmann’s rule.


Ashton, K.G. (2004) Sensitivity of intraspecific latitudinal

clines of body size for tetrapods to sampling, latitude and

body size. Integrative and Comparative Biology, 44, 403–

412.

Ashton, K.G. & Feldman, C.R. (2003) Bergmann’s rule in

nonavian reptiles: turtles follow it, lizards and snakes reverse

it. Evolution, 57, 1151–1163.

Ashton, K.G., Tracy, M.C. & de Queiroz, A. (2000) Is Berg-

mann’s rule valid for mammals? The American Naturalist,

156, 390–415.

Atkinson, D. (1994) Temperature and organism size – a bio-

logical law for ectotherms? Advances in Ecological Research,

25, 1–58.

Atkinson, D. (1996) Ectotherm life history responses to

developmental temperature. Animals and temperature:

phenotypic and evolutionary adaptation (ed. by I.A. Johnston

and A.F. Bennett), pp. 183–204. Cambridge University

Press, Cambridge.

Avery, R.A. (1976) Thermoregulation, metabolism and social

behaviour in Lacertidae. Morphology and biology of reptiles

(ed. by A.d’A. Bellairs and C.B. Cox), pp. 245–259. Aca-

demic Press, London.

Avery, R.A. (1978) Activity patterns, thermoregulation and

food consumption in two sympatric lizard species from

central Italy. Journal of Animal Ecology, 47, 143–158.

Barlow, N.D. (1994) Size distributions of butterfly species and

the effect of latitude on species sizes. Oikos, 71, 326–332.

Bartholomew, G.A. (1982) Physiological control of body

temperature. Biology of the reptilia, Vol. 12 (ed. by C. Gans

and F.H. Pough), pp. 167–213. Academic Press, New York.

Bergmann, C. (1847) Uber die Verhaltnisse der Warmeok-

onomie der Thiere zu ihrer Grosse. Gottinger Studien, 3,

595–708.

Blackburn, T.M. & Gaston, K.J. (1996) Spatial patterns in the

body sizes of bird species in the New World. Oikos, 77, 436–

446.


the mammal fauna of northern North America. Ecography,

27, 715–724.






64

Boback, S.M. & Guyer, C. (2003) Empirical evidence for an

optimal body size in snakes. Evolution, 57, 345–351.

Bogert, C.M. (1949) Thermoregulation in reptiles, a factor in

evolution. Evolution, 3, 195–211.

Boyce, M.S. (1978) Climatic variability and body size variation

in the muskrats (Ondatra zibethicus) of North America.

Oecologia, 36, 1–20.

Brown, J.H. (1995) Macroecology. University of Chicago Press,

Chicago, IL.

Burham, K.P. & Anderson, D.R. (2002) Model selection and

multimodel inference. Springer, New York.

Churchill, T.A. & Storey, K.B. (1992) Freezing survival of the

garter snake Thamnophis sirtalis. Canadian Journal of Zool-

ogy, 70, 99–105.

Conant, R. & Collins, J.T. (1998) A field guide to reptiles and

amphibians of eastern and central North America, 3rd edn.

Houghton Mifflin, Boston, MA.

Costanzo, J.P., Grenot, C. & Lee, R.E., Jr (1995) Supercooling,

ice inoculation and freeze tolerance in the European com-

mon lizard, Lacerta vivipara. Journal of Comparative Phy-

siology B, 165, 238–244.

Cousins, S.H. (1989) Species richness and the energy theory.

Nature, 340, 350–351.

Cowles, R.B. (1945) Surface-mass ratio, paleoclimate and heat

sterility. The American Naturalist, 79, 561–567.

Cowles, R.B. & Bogert, C.M. (1944) A preliminary study of the

thermal requirements of desert reptiles. Bulletin of the

American Museum of Natural History, 83, 265–296.

Cressie, N.A.C. (1993) Statistics for spatial data, revised edn.

Wiley, New York.

Cruz, F.B., Fitzgerald, L.A., Espinoza, R.E. & Schulte, J.A., II

(2005) The importance of phylogenetic scale in tests of

Bergmann’s and Rapoport’s rules: lessons from a clade of

South American lizards. Journal of Evolutionary Biology, 18,

1559–1574.

Currie, D.J. (1991) Energy and large-scale patterns of animal-

and plant-species richness. The American Naturalist, 137,

27–49.

Currie, D.J. & Fritz, J.T. (1993) Global patterns of animal

abundance and species energy use. Oikos, 67, 56–68.

Cushman, J.H., Lawton, J.H. & Manly, B.F.J. (1993) Latitu-

dinal patterns in European ant assemblages: variation in

species richness and body size. Oecologia, 95, 30–37.

Dıaz, J.A., Bauwens, D. & Asensio, B. (1996) A comparative

study of the relation between heating rates and ambient

temperatures in lacertid lizards. Physiological Zoology, 69,

1359–1383.

Diniz-Filho, J.A.F. & Fowler, H.G. (1998) Honey ants

(genus Myrmecocystus) macroecology: effect of spatial

patterns on the relationship between worker body size

and geographic range size. Environmental Entomology, 27,

1094–1101.




Dutilleul, P. (1993) Modifying the t-test for assessing the

correlation between two spatial processes. Biometrics, 49,

305–314.

Gasc, J.P., Cabela, A., Crnobrnja-Isailovic, J., Dolmen, D.,

Grossenbacher, K., Haffner, P., Lescure, J., Martens, H.,

Martınez Rica, J.P., Maurin, H., Oliveira, M.E., Sofianidou,

T.S., Veith, M. & Zuiderwijk, A. (eds) (1997) Atlas of

amphibians and reptiles in Europe. Collection Patrimoines

Naturels, 29, Societas Europaea Herpetologica, Museum

National d’Histoire Naturelle & Service du Petrimone Natu-

rel, Paris.

Gaston, K.J. & Blackburn, T.M. (2000) Pattern and process in

macroecology. Blackwell Science, Oxford.

Gaussen, H. (1954) Geographie des plantes. Armand Colin,

Paris.

Geist, V. (1987) Bergman’s rule is invalid. Canadian Journal of

Zoology, 65, 1035–1038.

Grenot, C.J., Garcin, L., Dao, J., Herold, J.P., Fahys, B. &

Tsere-Pages, H. (2000) How does the European common

lizard, Lacerta vivipara, survive the cold of winter?

Comparative Biochemistry and Physiology Part A, 127,

71–80.

Haining, R. (2003) Spatial data analysis. Cambridge University

Press, Cambridge.

Hausdorf, B. (2003) Latitudinal and altitudinal body size

variation among north-west European land snail species.


Hawkins, B.A. (1995) Latitudinal body-size gradients for the

bees of the eastern United States. Ecological Entomology, 20,

195–198.

Hawkins, B.A. (2004) Summer vegetation, deglaciation and the

anomalous bird diversity gradient in eastern North America.


Hawkins, B.A. & Diniz-Filho, J.A.F. (2004) ‘Latitude’ and geo-

graphic patterns in species richness. Ecography, 27, 268–272.

Hawkins, B.A. & Lawton, J.H. (1995) Latitudinal gradients in

butterfly body sizes: is there a general pattern? Oecologia,

102, 31–36.

Hawkins, B.A., Porter, E.E. & Diniz-Filho, J.A.F. (2003) Pro-

ductivity and history as predictors of the latitudinal diversity

gradient of terrestrial birds. Ecology, 84, 1608–1623.

Hedges, S.B. (1985) The influence of size and phylogeny

on life-history variation in reptiles: a response to Stearns.


Huey, R.B. (1982) Temperature, physiology, and the ecology of

reptiles. Biology of the Reptilia, Vol. 12 (ed. by C. Gans and

F.H. Pough), pp. 25–67. Academic Press, New York.

Hurlbert, A.H. & Haskell, J.P. (2003) The effect of energy and

seasonality on avian species richness and community com-

position. The American Naturalist, 161, 83–97.

James, F.C. (1970) Geographic size variation in birds and its

relationship to climate. Ecology, 51, 365–390.

Johnson, J.B. & Omland, K.S. (2004) Model selection in

ecology and evolution. Trends in Ecology and Evolution, 19,

101–108.



65

Kineman, J. & Hastings, D. (1992) Monthly Generalized

Global Vegetation Index from NESDIS NOAA-9 Weekly

GVI Data (APR 1985–DEC 1988). Digital Raster Data on a

10-minute Cartesian Orthonormal Geodetic (lat/long) 1080

· 2160 grid (Platte Carree projection). Global Ecosystems

Database Version 2.0. Boulder, CO.

Legendre, P., Dale, M.R.T., Fortin, M.J., Gurevitch, J., Hohn,

M. & Myers, D. (2002) The consequences of spatial structure

for the design and analysis of ecological field surveys.

Ecography, 25, 601–615.

Lieth, H. (1975) Modeling the primary productivity of the

world. Primary productivity of the biosphere (ed. by H. Lieth

and R.H. Whittaker), pp. 237–263. Springer-Verlag, New

York.



Lindstedt, S.L. & Boyce, M.S. (1985) Seasonality, body size and

survival time in mammals. The American Naturalist, 125,

873–878.

McDowall, R.M. (1994) On size and growth in freshwater fish.

Ecology of Freshwater Fish, 3, 67–79.

McNab, B.K. (1971) On the ecological significance of Berg-

mann’s rule. Ecology, 52, 845–854.

Meiri, S. & Dayan, T. (2003) On the validity of Bergmann’s

rule. Journal of Biogeography, 30, 331–351.

Meliadou, A. & Troumbis, A. (1997) Aspects of heterogeneity

in the distribution of diversity of the European herpeto-

fauna. Acta Oecologica, 18, 393–412.

Miller, W.E. (1991) Body size in North American Lepidoptera

as related to geography. Journal of Lepidopterists’ Society, 45,

158–168.

Mousseau, T.A. (1997) Ectotherms follow the converse to

Bergmann’s rule. Evolution, 51, 630–632.

Murphy, E.C. (1985) Bergmann’s rule, seasonality, and geo-

graphic variation in body size of house sparrows. Evolution,

39, 1327–1334.

Peltier, W. (1993) Time dependent topography through glacial

cycle. IGBP PAGES/World Data Center-A for Paleoclima-

tology data contribution series no. 93-015. NOAA/NGDC

Paleoclimatology Program, Boulder, CO.

Pough, F.H. (1980) The advantages of ectothermy for tetra-

pods. The American Naturalist, 115, 92–112.

Pough, F.H. & Groves, J.D. (1983) Specializations of the body

form and food habits of snakes. American Zoologist, 23, 443–

454.

Rangel, T.F.L.V.B., Diniz-Filho, J.A.F. & Bini, L.M. (2005)

SAM v.1.0 beta – spatial analysis in macroecology (software

and user’s guide), http://www.ecoevol.ufg.br/sam/index.htm.

Reed, R.N. (2003) Interspecific patterns of species richness,

geographic range size, and body size among New World

venomous snakes. Ecography, 26, 107–117.

Rodrıguez, M.A., Belmontes, J.A. & Hawkins, B.A. (2005)

Energy, water and large-scale patterns of reptile and

amphibian species richness in Europe. Acta Oecologica, 28,

65–70.

Rodrıguez, M.A., Lopez-Sanudo, I.L. & Hawkins, B.A. (2006)

The geographic distribution of mammal body size in Eur-

ope. Global Ecology and Biogeography, 15, 185–193.

Rosenzweig, M.L. (1968a) The strategy of body size in mam-

malian carnivores.AmericanMidlandNaturalist, 80, 299–315.

Rosenzweig, M.L. (1968b) Net primary productivity of ter-

restrial environments: predictions from climatological data.


Schall, J.J. & Pianka, E.R. (1978) Geographical trends in

numbers of species. Science, 201, 679–686.

Scholander, P.F. (1955) Evolution of climatic adaptation in

homeotherms. Evolution, 9, 15–26.

Sears, M.W. & Angilleta, M.J. (2004) Body size clines in Scelo-

porus lizards: proximate mechanisms and demographic con-

straints. Integrative and Comparative Biology, 44, 433–442.

StatSoft, Inc. (2003) STATISTICA (data analysis software system),

Version 6. http://www.statsoft.com.

Stebbins, R.C. (2003) A field guide to western reptiles and

amphibians, 3rd edn. Houghton Mifflin, Boston, MA.

Turner, J.R.G. & Hawkins, B.A. (2004) The global diversity

gradient. Frontiers of biogeography: new directions in

the geography of Nature (ed. by M.V. Lomolino and

L.R. Heany), pp. 171–190. Sinauer, Sunderland, MA.

Wigginton, J.D. & Dobson, F.S. (1999) Environmental influ-

ences on geographic variation in body size of western bob-

cats. Canadian Journal of Zoology, 77, 802–813.

Willemsen, R.E. & Hailey, A. (1999) Variation of adult body

size of the tortoise Testudo hermanni in Greece: proximate

and ultimate causes. Journal of Zoology, 248, 379–396.

Zug, G.R., Vitt, L.J. & Caldwell, J.P. (2001) Herpetology: an

introductory biology of amphibians and reptiles. Academic

Press, New York.

BIOSKETCHES

Miguel A. Olalla-Tarraga is a PhD student at the University

of Alcala particularly interested in studying large-scale patterns

and processes in ecology. He is using a broad-scale approach

to investigate the effects of Mediterranean forest fragmentation

on herpetofaunal diversity. His main research interests also

include aspects of urban ecology.

Miguel A. Rodrıguez’s main interests include the study of

factors and processes conditioning patterns of biodiversity at

local, regional and global scales. Recent research involves

investigating the effects of habitat destruction and fragmen-

tation on aggregate properties of faunas at broad scales.

Bradford A. Hawkins is interested in biogeography and

geographical ecology, with a major emphasis on broad-scale

diversity gradients.

Editor: Lawrence Heaney



66

Capítulo 4 Energía y patrones de tamaño corporal interespecíficos en faunas de anfibios de Europa y Norte América: anuros siguen la regla de Bergmann, urodelos su inversa Este capítulo reproduce íntegramente el texto del siguiente manuscrito: Olalla-Tárraga, M. Á. & Rodríguez, M. Á. 2007. Energy and interspecific body size patterns of amphibian faunas in Europe and North America: anurans follow Bergmann's rule, urodeles its converse. Global Ecology & Biogeography 16: 606–617.

DOI: 10.1111/j.1466-8238.2007.00309.x © 2007 The Authors

606

Journal compilation © 2007 Blackwell Publishing Ltd www.blackwellpublishing.com/geb

Global Ecology and Biogeography, (Global Ecol. Biogeogr.)

(2007)

16

, 606–617

RESEARCHPAPER

Blackwell Publishing Ltd

Energy and interspecific body size patterns of amphibian faunas in Europe and North America: anurans follow Bergmann’s rule, urodeles its converse

Miguel Á. Olalla-Tárraga* and Miguel Á. Rodríguez

ABSTRACT

Aim

To describe broad-scale geographical patterns of body size for European andNorth American amphibian faunas and to explore possible processes underlyingthese patterns. Specifically, we propose a heat balance hypothesis, as both heat con-servation and heat gain determine the heat balance of ectotherms, and test it alongwith five other hypotheses that have a possible influence on body size gradients: sizedependence, migration ability, primary productivity, seasonality and water availability.

Location

Western Europe and North America north of Mexico.

Methods

We processed distribution maps for native amphibian species to estimatethe mean body size in 110

×

110 km cells and calculated eight environmental predictorsto explore the relationship between environmental gradients and the observed patterns.We used least squares regression modelling and model selection approaches basedon information theory to evaluate the relative support for each hypothesis.

Results

We found consistent body size gradients and similar relationships toenvironmental variables within each amphibian group in Europe and North America.Annual potential evapotranspiration, a measure of environmental energy, was thestrongest predictor of mean body size in both regions. However, the contrastingresponses to ambient energy in each group resulted in opposite geographicalpatterns, i.e. anurans increased in size from high- to low-energy areas in both continentsand urodeles showed the opposite pattern.

Main conclusions

Our results support the heat balance hypothesis, suggestingthat the thermoregulatory abilities of anurans would allow them to reach larger sizesin colder climates by optimizing the trade-off between heating and cooling rates,whereas a lack of such strategies among urodele faunas would explain why theseorganisms tend to be smaller in cooler areas. These findings may also have implicationsfor the role of climate warming on the global decline of amphibians.

Keywords

Amphibians, bauplans, Bergmann’s rule, body size gradients, Europe, heat balance

hypothesis, macroecology, North America.

*Correspondence: Miguel Ángel Olalla-Tárraga, Departamento de Ecología, Facultad de Biología, Universidad de Alcalá, 28871 Alcalá de Henares, Madrid, SpainE-mail: [email protected]

Department of Ecology, University of Alcalá,

28871 Alcalá de Henares, Madrid, Spain

INTRODUCTION

The reduced surface area to volume ratios, and hence greater

heat conservation potential, of large-bodied animals led

Bergmann (1847) to predict that, within groups of closely related

endotherms, larger species would abound in colder climates

and smaller species in warmer areas. Following the same logic,

Rensch (1938) elaborated an intraspecific version of this rule,

stating that, within a particular species, larger individuals should

be more frequent in cold climates. Since then, the variation of

animal body sizes at broad scales has become of central interest

for biogeographers and ecologists (Mayr, 1956; Rosenzweig,

1968; McNab, 1971; Ashton et al., 2000; Meiri & Dayan, 2003,

Blackburn & Hawkins, 2004; Rodríguez et al., 2006), although

both the generality of the patterns and possible causal mechanisms

are still debated (Blackburn et al., 1999).

Bergmann’s gradients and their converse have been observed

in ectotherms, both at the intraspecific (Mousseau, 1997; Ashton,

2002; Belk & Houston, 2002; Ashton & Feldman, 2003; Angilletta

et al., 2004; Blanckenhorn & Demont, 2004) and interspecific

67

Interspecific body size patterns in amphibians

© 2007 The Authors


,

16


607

levels (Lindsey, 1966; Cushman

et al

., 1993; Barlow, 1994;

McDowall, 1994; Hawkins & Lawton, 1995; Diniz-Filho &

Fowler, 1998; Cruz et al., 2005, Olalla-Tárraga et al., 2006). These

findings not only suggest that climate might be a key deter-

minant of body size distributions within cold-blooded animals,

but also emphasize the need for hypotheses that are specifically

tailored for animals with limited capability to thermoregulate,

and whose body temperature depends on heat conservation

as well as on heat gain. Accordingly, we propose here the heat

balance hypothesis to account for the patterns exhibited by

thermoregulators (i.e. animals with good thermoregulating

abilities) and thermoconformers (animals with body temperatures

fluctuating more closely to ambient temperature). For thermo-

regulators, this hypothesis states that larger animals would

be favoured in cold environments due to their reduced surface

to volume ratios and greater heat conservation potentials.

Conversely, among thermoconformers, smaller organisms would

be favoured in cold areas as their greater surface to volume ratios

allow them to have shorter heating times. The first part of this

hypothesis coincides with the heat conservation mechanism

proposed by Bergmann for endotherms, and can also be

extended to ectotherms that are able to control body temperature

effectively through physiological and/or behavioural adjust-

ments (see, e.g., Bartholomew, 1982; Huey, 1982; Hutchinson &

Dupré, 1992). For example, the few species of anurans (frogs and

toads) in which these issues have been investigated do exhibit

thermoregulatory abilities to enhance heat gain and heat con-

servation (e.g. basking or selecting temperatures in available

microhabitats) (Brattstrom, 1963; Hutchinson & Dupré, 1992;

Stebbins & Cohen, 1995). However, since similar thermoregulating

activities are believed to be uncommon amongst urodeles (newts

and salamanders) (Brattstrom, 1963; Hutchinson & Dupré,

1992; Stebbins & Cohen, 1995), the second part of the hypothesis

would be applicable for them. This indeed generates two pre-

dictions. First, that body size would increase towards colder areas

in the case of anurans and towards warm territories in that of

urodeles and, second, that these patterns should be inversely

related to those climatic factors more directly conditioning

amphibian thermoregulation, that is, heat and sunlight (Hutchinson

& Dupré, 1992).

Along with this hypothesis, we also investigate five other

explanations for body size gradients in ectotherms. First, the size

dependence hypothesis (Olalla-Tárraga et al., 2006) suggests that

below a certain body size threshold heat conservation leads to an

increase in size towards cold environments, whereas beyond this

threshold the trend reverses due to the longer heating times of

larger animals and the narrow activity windows available at high

latitudes that leave the animals with no time to meet other needs.

Second, the migration ability hypothesis, which assumes that

large-bodied species disperse better and have been more able to

recolonize far northern territories after the retreat of the late

Pleistocene ice sheets (Cushman et al., 1993; Blackburn et al.,

1999). Third, the primary productivity hypothesis, which

assumes that body mass must be maintained by a sufficient food

supply and predicts greater body sizes in more productive areas

(Rosenzweig, 1968). Fourth, the seasonality hypothesis (also

known as the starvation resistance hypothesis), which, in the case

of amphibians, relates growing season length and the time avail-

able for physiological development, and predicts small-bodied

species to be more frequent in more seasonal environments

where animals have less time to grow (Mousseau, 1997). Finally,

the water availability hypothesis, which assumes a greater

desiccation tolerance for larger amphibians due to their reduced

surface to volume ratio, and predicts them to be more frequent

than smaller species in areas with lower water availability

(Ashton, 2002).

To describe amphibian body size patterns and to test these

hypotheses, we use a ‘community approach’ (sensu Blackburn

& Hawkins, 2004) by examining the mean body size of species

occurring in equal-area grid cells of 110 × 110 km each (see also

Olalla-Tárraga et al., 2006, Rodríguez et al., 2006). We analyse

separately the two major orders of living amphibians, i.e. anurans

(order Anura) and urodeles (order Caudata).


Species data

Distribution maps for native amphibian species of western

Europe and North America were obtained from the same sources

used in a previous investigation of reptile body size gradients in

these territories (Olalla-Tárraga et al., 2006), that is, Gasc et al.

(1997), Conant and Collins (1998) and Stebbins (2003). The

methodology used for map processing was described in detail

in Olalla-Tárraga et al. (2006), and rendered 386 cells in Europe

and 1430 in North America, all of them having an equal area of

110 × 110 km. Additionally, this grain size overrode all slight

discrepancies existing between different sources (Conant &

Collins 1998; Stebbins 2003) in the mapped range distributions

of species present in eastern and western North America (1

salamander and 15 anurans).

The amphibian data base comprised 49 species in Europe

(31 anurans and 18 urodeles) and 216 species in North America

(81 anurans and 135 urodeles). This data base does not include

five small North American frogs living in the far north (Rana

sylvatica, Pseudacris crucifer, Pseudacris triseriata, Hyla versicolor

and Hyla chrysoscelis). These species are known to be tolerant of

freezing, a physiological trait believed to explain their presence in

colder territories (Storey & Storey, 1996). Since our study seeks

to establish the role of body size in determining amphibian

distribution, we excluded these species because it is already

known that their distribution is associated with other traits (see

Appendix S1 in Supplementary Material).

Body size data were extracted from field guides (Behler &

King, 1994; Conant & Collins, 1998; Arnold, 2002; Stebbins,

2003, see Appendix S2 in Supplementary Material). We used

maximum snout-to-vent length (SVL) for anurans, and maxi-

mum total length (TL) for urodeles. The latter were the only size

estimates available for all species and, in the case of North Amer-

ica, the only ones that were comparable between the various field

guides that provided these data (Behler & King, 1994; Conant &

Collins, 1998; Stebbins, 2003). For those species in which length

68

M. Á. Olalla-Tárraga and M. Á. Rodríguez

© 2007 The Authors

608


,

16


measurements were provided for males and females or for differ-

ent subspecies, we averaged the values to obtain a single estimate

of body size. On the other hand, in order to assess the indicative

value of TL, we also extracted the SVL of the urodele species for

which these data were available (42 species from western North

America) and correlated both size estimates. This correlation was

very high (r = 0.92), thus indicating that both measurements are

likely to generate similar results. In addition, maximum length

values are believed to be reasonable indicators of the size poten-

tial of organisms with indeterminate growth such as amphibians,

and have been used in previous studies of macroecological

patterns in ectothermic vertebrates (Boback & Guyer, 2003;

Reed, 2003).

With regard to our response variables, we took into account

that the frequency distribution of animal body sizes is often right

skewed on large geographical scales (Brown, 1995). This indeed

makes it inadvisable to use arithmetic means as those might

be strongly influenced by the presence of large-sized species.

Blackburn and Hawkins (2004) have already tackled this

problem in a similar interspecific study of the mean body sizes

of northern North American mammals, and their solution con-

sisted of log

10

-transforming the body size values of the species

before averaging them (see also Olalla-Tárraga

et al

., 2006;

Rodríguez et al., 2006). Following the same protocol, we calculated

the average log

10

body length (hereafter called mean body size)

for anurans and urodeles for each grid cell in each biogeographical

region.

Environmental predictors

We included eight potential explanatory variables, selected

because they can be related to the six following hypotheses (see

Olalla-Tárraga et al., 2006, for further details on data definition

and sources). (1) Heat balance: to account for thigmothermy and

heliothermy; i.e. the two mechanisms involved in heat gain in

amphibians (Hutchinson & Dupré, 1992), mean annual temperature

and the Priestley–Taylor annual potential evapotranspiration

(PET; see Lu et al., 2005) were used as indicators of heat, and

heat and light inputs in the environment, respectively (Rodríguez

et al., 2005). Range in elevation within each cell was also used as

an estimate of mesoscale spatial climatic variation (Turner &

Hawkins, 2004). (2) Size dependence: the analyses were based on

the same variables, complemented with visual inspections of

their relationships with amphibian body size to look for body

size thresholds. (3) Migration ability: time since glacial retreat

(cell age) was estimated by mapped changes in ice coverage at

specific intervals since the last glacial maximum (e.g. Rodríguez

et al. 2006) in Europe (Peltier, 1993) and North America (Dyke et al.,

2003). (4) Primary productivity: the global vegetation index (GVI)

was utilized as a proxy for plant productivity. (5) Seasonality: the

number of months available for plant growth in each grid cell

was estimated to measure the length of the growing season.

(6) Water availability: annual precipitation and Thornthwaite’s

annual actual evapotranspiration (AET) were used as indicators

of water and water–energy balance, respectively. Whereas annual

precipitation reflects the limits of water stress to amphibian body

size, AET indicates the need for animals to have access to water as

well as to tolerable temperatures (Rodriguez et al., 2005).


Statistical analyses were run separately for anurans and urodeles

in each biogeographical region. We first used Pearson correla-

tions to evaluate the relationships between mean body size and

the eight explanatory variables. To control for spatial autocorre-

lation in the body size and environmental data, we used a modi-

fied t-test (Dutilleul, 1993) to obtain unbiased estimates of the

significance of the correlation coefficients.

We also used ordinary least squares multiple regression to

generate environmental models including multiple predictors.

For this we took into account the strong collinearity among

several of the variables in our data set (Table 1), which makes

the use of common model simplification procedures, such as

step-wise regression methods, inadvisable. Instead, we used the

Akaike information criterion (AIC) to evaluate the relative

support for each hypothesis by comparing the models resulting

from combinations of the environmental variables. Due to the

presence of spatial autocorrelation in the data, we calculated

AICs using corrected error variances according to methods

described in Hawkins and Diniz-Filho (2006) (see also Olalla-

Tárraga et al., 2006 for additional details). We compared the

resulting AIC values of each model using ΔAIC, the difference

between the AIC of each model and the minimum AIC found.

We also used ΔAIC values to calculate the Akaike weighting

of each model (w

i

), which is standardized across the candidate set

of models and can be interpreted as the probability that model

i is actually the best explanatory model.

Additionally, we used the standardized regression coefficients

of the models as indicators of the relative importance of each

variable for mean body size (Neter et al., 1996). Because the value

of these coefficients is sensitive to collinearity (Neter et al., 1996),

we did not include highly correlated variables in the analyses.

Thus, due to the high correlation between PET and mean annual

temperature in both continents (Table 1), we performed a sepa-

rate modelling exercise with either variable. The results were

equivalent (which was expected due to the similar biological

meaning of these variables) and we only included here those

obtained with PET, while the models generated using tempera-

ture are provided as supplementary material (see Appendix S3 in

Supplementary Material).

Because only the migration ability hypothesis is associated

with past climatic conditions, whereas the rest are linked to cur-

rent climatic variables, and present and past climates are spatially

correlated, we used the following procedure to disentangle

current and historical effects on variation in amphibian body

size. We first evaluated the relative support for each of the

hypotheses related to present climates by including only current

variables in the model selection process. Then, once we had iden-

tified the best environmental model based on present climatic

conditions, we conducted partial regression analyses to partition the

variance explained by present climate vs. cell age. This allowed us

to test the extent that the historical effect is independent of the

69


© 2007 The Authors


,

16


609

explanation provided by measures of current climatic conditions

(Hawkins & Porter, 2003). We used simple and multiple regres-

sions to calculate separately the coefficients of determination for

the best current climatic model, cell age and climate and cell age

combined. Then we used these three coefficients of determination

to partition the independent effects of present climates, cell age

and the overlap between them.

Finally, we compared spatial autocorrelation patterns in mean

body size with those of the residuals of our multiple regression

models to evaluate how these models explain spatial variation in

body size (Diniz-Filho et al., 2003, see also Olalla-Tárraga et al.,

2006, for further details).

All statistical analyses were performed with statistica

(StatSoft, Inc. 2003) and sam 1.1 (Spatial Analysis in Macroecology;

Rangel et al., 2006).

RESULTS

Anuran body size and current climate

Mean body size of the European anuran fauna increases towards

the north, exhibiting a marked Bergmann’s rule pattern (Fig. 1a),

with the smallest animals occurring in the Iberian and Italian

peninsulas, and the largest in Great Britain and eastern Scandi-

navia. Mean body size was negatively correlated with all of the

environmental variables, but only significantly with PET and

AET (Table 1a). Of 64 multiple regressions models, two had a

ΔAIC ≤ 2 and explained similar proportions of variance (Table

2). However, Akaike weightings indicated that the model includ-

ing PET, elevation range, AET, GVI and growing season length

was the best model. This model explained 6.8% more variance of

anuran body size than PET alone, (66.2% vs. 59.4%, respectively)

and its standardized coefficients identified PET as the strongest

predictor (Table 2). Moreover, these relationships remained

robust after accounting for the effects of variation in species

richness across cells with weighted least squares regression

(the results are included in Appendix S4 in Supplementary

Material).The relationship of anuran mean body size with PET

in Europe is shown in Fig. 2(a).

Spatial variation of anuran mean body size also follows

Bergmann’s rule in North America, although the pattern is

patchier than in Europe (Fig. 1b). Indeed, although the smallest

sizes appear in the south (especially in the south-east) and the

largest ones occur in the north in general, northern areas are not

uniformly occupied by large animals. Instead, they are character-

ized by a large area occupied by intermediate sizes, in which three

major patches of large mean body sizes occur (in western and

central Canada, and from the Great Lakes to the west coast).

Anuran mean body size was significantly negatively correlated

with four variables in North America (Table 1b), and the best

Table 1 Pearson correlation coefficients of environmental variables and mean body size. Significance levels are corrected for spatial autocorre-lation using the modified t-test developed by Dutilleul (1993): (a) Europe; (b) North America(a)

(b)

Variable

Size,

anurans

Size,

urodeles Temp. AET PET Range Prec. GVI GSL

Temp. −0.67 0.56 1

AET −0.66* 0.46* 0.44 1

PET −0.77* 0.73* 0.79* 0.67* 1

Range −0.11 0.21 −0.08 0.29 0.30 1

Prec. −0.17 0.07 0.06 0.47 0.17 0.41 1

GVI −0.61 0.27 0.60 0.74* 0.58 0.06 0.46 1

GSL −0.42 0.21 0.75** 0.37 0.43 −0.16 0.23 0.69* 1

Cell age −0.75 0.56 0.76* 0.71 0.74 0.18 0.33 0.71 0.67

Variable

Size,

anurans

Size,

urodeles Temp. AET PET Range Prec. GVI GSL

Temp. −0.66** 0.63 1

AET −0.61** 0.62 0.8* 1

PET −0.62** 0.65* 0.94** 0.73 1

Range 0.21 −0.36 −0.04 0.30 0.02 1

Prec. −0.24 0.43 0.44 0.65** 0.33 −0.03 1

GVI −0.27 0.15 0.50* 0.62** 0.39 −0.03 0.43** 1

GSL −0.54** 0.54 0.69* 0.88** 0.61 0.33 0.62** 0.58** 1

Cell age −0.44 0.62 0.65 0.46 0.64 0.29 0.12 0.37 0.38

Temp., mean annual temperature; AET, actual evapotranspiration; PET, potential evapotranspiration; Range, range in elevation; Prec., annualprecipitation; GVI, global vegetation index; GSL, growing season length.*P < 0.05; **P < 0.01.

70


© 2007 The Authors

610


,

16


Figure 1 Geographical patterns of amphibian mean body size in Europe and North America: (a) anurans in Europe [number of species (S) = 31]; (b) anurans in North America (S = 81); (c) urodeles in Europe (S = 18); (d) urodeles in North America (S = 135). The maps were built using cell averages of log10-transformed length values. Numbers included in the legend of each map are length values in centimetres obtained after antilog transformation. The black line in (d) represents the eastern border of the Great Plains used to conduct separate analyses for eastern North America (see text).

71


© 2007 The Authors


,

16


611

multiple regression model included PET and growing season

length and explained 46.9% of variance (Table 2). On the

other hand, the standardized coefficient for PET indicated

that this variable is the main driver of variation in anuran

body size in North America, even after accounting for species

richness patterns with weighted least squares regression (see

Appendix S4).

It is worth noting that although we excluded five anuran

species known to be freeze-tolerant from the analyses, the

patchier patterns detected in northern North America may still

Figure 2 The relationships between amphibian mean body size and potential evapotranspiration (PET) in 110 × 110 km cells for Europe and North America: (a) anurans in Europe; (b) anurans in North America; (c) urodeles in Europe; (d) urodeles in eastern North America. Similar relationships were obtained for mean annual temperature. Note that although urodeles appear to be larger than anurans, the difference arises because the tail is included for urodeles. Overall, anurans cannot be considered as small-sized amphibians when compared with urodeles (Lindsey, 1966; Pough, 1980).

Table 2 Multiple regression models for mean body size (data resolved to 110 × 110 km). Models are ranked in each case by AIC from best- to worst-fitting model, and only the models with ΔAIC < 2 are presented. AICs have been corrected for the presence of spatial autocorrelation in the model residuals. For each variable entering in the model we include their standardized coefficients to evaluate the relative importance of each one. Predictor variable codes are: PET, annual potential evapotranspiration; Range, range in elevation; AET, annual actual evapotranspiration; Prec., annual precipitation; GVI, global vegetation index; GSL, growing season length. See Appendix S2 in Supplementary Material for models including mean annual temperature instead of PET

Group Region

Predictors in model

AIC ΔAIC wi R2PET Range AET Prec. GVI GSL

Anurans Europe −0.628 0.155 −0.171 −0.168 0.063 −980 0 0.509 0.662

−0.612 0.143 −0.201 −0.117 −979 1 0.291 0.660

North America −0.475 −0.321 −2483 0 0.347 0.469

−0.480 −0.044 −0.296 −2482 1 0.235 0.471

−0.473 0.011 −0.329 −2482 1 0.221 0.461

−0.478 0.017 −0.046 −0.306 −2481 2 0.157 0.471

Urodeles Europe 0.757 −0.092 0.205 0.057 −0.320 0.103 −919 0 0.428 0.567

0.783 −0.118 0.164 0.068 −0.243 −918 1 0.343 0.562

Eastern North America 0.692 −0.192 −0.106 −828 0 0.510 0.475

72



be influenced by the occurrence of species with similar but

yet unknown capabilities to tolerate cold via physiological mech-

anisms independent of body size. As might be expected, the

results obtained after including these five species support this

idea, considering that we found similar but noisier patterns

of body size variation (map included in Appendix S1) and our

best model (i.e. with PET and length of the growing season) only

explained 21.1% of the variance. Moreover, a full model involv-

ing all six predictors only raised this amount to 37.4% (also

included in Appendix S1); that is, 10% less than the proportion

of variance described by the best model when excluding such

species (see above).

Urodele body size and current climate

In contrast to anurans, urodele mean body size clearly tends

to increase southwards in Europe (Fig. 1c), whereas in North

America the most conspicuous feature is a broad patch of larger

sizes that extends from south-central Canada throughout the

Great Plains and eastern Rocky Mountains to the northern

Chihuahuan Desert (Fig. 1d). Most of this patch was made up by

a single species, the big tiger salamander (Ambystoma tigrinum),

a widely distributed species that also appears, albeit more spot-

tily, in the eastern United States (from the Ohio Valley to the

Atlantic seaboard), the south-eastern states, and in northern

Mexico (see Conant & Collins, 1998; Stebbins, 2003). Besides

this feature, there is also a body size gradient similar to that

observed in Europe (i.e. mean body size tends to increase south-

wards) occurring between the eastern border of the Great Plains

(we followed the ecoregions division of the Commission for

Environmental Cooperation, 1997) and the east coast (Fig. 1d).

This part of North America, and particularly the Appalachian

Mountains, has the highest urodele richness in the Nearctic and

represents a global hotspot of salamander biodiversity (Duellman

& Trueb, 1986) (see richness map included in Appendix S4). In

fact, as Darlington (1957) has already noted, we found two well-

differentiated urodele faunas in North America, one occupying

the western half of the continent with no obvious geographical

trends in body size, and the other occurring in the east. Such

dissimilarities in faunal composition suggest not only that

the western mountains and deserts may have impeded faunal

exchanges between western and eastern subfaunas, but also that

they might impose different constraints on body size variation.

Taking this into account, as well as the existence of only one species

(Ambystoma tigrinum) in most cells to the west of the Great Plains,

we decided to concentrate our analysis in the eastern region.

Urodele mean body size was significantly positively correlated

with PET and AET in Europe, and with PET in eastern North

America (Table 1). The best multiple-regression models for both

regions described 56.7% and 47.5% of the variation in mean

body size, respectively (Table 2). The standardized coefficients

also identified PET as the strongest predictor of body size in both

cases (see Fig. 2c,d). Again, weighted least squares regression

analyses indicated that these results were not affected by the

geographical variation of species richness in either continent

(the results are included in Appendix S4).

Spatial analysis

In all cases the pattern of spatial autocorrelation for mean

body size was representative of a cline, with positive autocorre-

lation at shorter distances progressively becoming negative

at larger distances (Fig. 3a,c,e,g). The best multiple regression

models reduced the spatial autocorrelation in all distance

classes, although less strongly in the shortest one, which

suggests that factors not included in our analysis are needed

to account for the spatial variation in body size at more local

scales (see below).

We also mapped the residuals of these models to explore where

the unexplained variation occurs. For European anurans, we

found three clusters of positive values, in Great Britain, eastern

Scandinavia and the Balkan Peninsula, indicating that body sizes

are larger than predicted there (Fig. 3b). Similarly, the map for

North American anurans revealed four major clusters of large

residuals, three located in areas covered by the Laurentide ice

sheet in the Pleistocene (two of negative residuals and one of

positive values), and the other one in the south-east (negative

residuals) (Fig. 3d). In the case of urodeles, larger residuals were

mainly found in European peninsulas, especially in Italy (nega-

tive residuals) and Scandinavia (positive and negative residuals),

and in the region bordering the Great Plains in eastern North

America (positive residuals). On the whole, these maps rein-

forced the suggestion that factors not included in our analyses

(e.g. peninsular effects in Europe, and ‘contagion’ effects of the

western urodele fauna in eastern North America) are needed to

account for the variation in mean body size in these areas.

Body size and post-Pleistocene climatic change

Strong correlations between cell age and most current climate

variables (see Table 1) explain why large amounts of the spatial

variation in mean body size of anurans and urodeles are

described by the overlap between past and present climates

(Fig. 4). Despite this, the proportion of variance that can be

independently attributed to cell age is much lower than that

corresponding to present climatic factors in all cases. In fact, the

addition of cell age to the models only improved their explana-

tory power by between 1.5 and 4.8%, whereas current variables

independently described between 13.2 and 28.3% of the vari-

ance. This indicates that historical effects, as measured by the

pattern of glacial retreat, on amphibian body size gradients

are weak. Further, when we added cell age to the best climatic

models, the resulting AIC values were higher than those of the

original models, thus reinforcing the idea that cell age has little or

no effect on body size variation. Consequently, we conclude that

the broad-scale body size patterns of the amphibian faunas of

both continents are mainly driven by present climate, with PET

being the strongest descriptor.

DISCUSSION

The two major orders of living amphibians — anurans and urodeles

— show clear biogeographical patterns of mean body size variation

73



in North America and, especially, in Europe. Following a clear

Bergmann’s rule gradient, anurans increase their body size

northwards in both regions. In contrast, urodeles exhibit the

reverse trend, becoming larger towards southern latitudes in

Europe and eastern North America, whereas in the western part

of this continent the patterns are patchier. On the other hand,

our data also indicate a leading role of current climate in driving

these patterns. Specifically, we found a strong association

between geographical gradients in body size and the spatial varia-

tion of PET, a direct measure of ambient energy (heat and light)

which is highly correlated with temperature. However, in accord-

ance with the contrasting patterns we detect for both amphibian

groups, the relationships of PET with anuran and urodele body

sizes are opposite in sign, so that anurans tend to be larger and

urodeles smaller towards low-energy areas.

Of the six explanations for amphibian body size patterns we

have examined, the best support is for the heat balance hypo-

thesis. This takes into account that ectotherm body temperature

is limited by heat gain as well as heat conservation, and proposes

that geographical body size gradients in cold-blooded animals

are influenced by the interplay between heating and cooling

rates. An essential constraint for larger-sized ectotherms in

colder environments has to do with the fact that their low surface

to volume ratio confers upon them a reduced capability to gain

heat. This is likely to be critical for animals with low thermo-

regulatory potential (thermoconformers), which are virtually

unable to accelerate the heating process, mainly because the

difficulties these animals have in reaching active body temper-

atures might leave them with insufficient time for foraging or

reproduction in the short activity windows available in cold

Figure 3 Spatial correlograms using Moran’s I for mean body size (solid circles) and residuals of the best environmental model (open circles) for: (a) anurans in Europe; (c) anurans in North America; (e) urodeles in Europe; (g) urodeles in eastern North America. Maps below the correlograms show the geographical distribution of the higher positive (> 0.04) (white) and lower negative (< −0.04) (black) residuals of the corresponding model (b, d, f, h).

74



regions (cf. Olalla-Tárraga et al., 2006). In contrast, smaller sizes

would favour a rapid heating, thus allowing the animals to

take advantage of these short time segments in which benign

environmental conditions allow them to be active. Urodeles are

believed to have low thermoregulatory abilities (Brattstrom,

1963; Hutchinson & Dupré, 1992; Stebbins & Cohen, 1995; but

see Spotila, 1972), which agrees with the northwards decrease in

mean body size we observe in this group.

In contrast, the ability of some ectotherms to make physiological

and/or behavioural adjustments to obtain heat (thermoregulators)

could overcome the constraints imposed by large sizes, while

these sizes would become advantageous due to the positive influ-

ences that their corresponding low surface to volume ratios confer

on maintaining body temperature excesses (e.g. Digby, 1955).

Accordingly, the heat balance hypothesis predicts Bergmann’s

patterns when ectothermic organisms are able to achieve some

control over heat exchange, a characteristic that has been shown

by the anuran species for which thermoregulatory capabilities

have been documented (Brattstrom, 1963; Hutchinson & Dupré,

1992; Stebbins & Cohen, 1995). This indeed suggests that heat

balance also lies behind the body size trends we observe in this

group.

The different bauplans of both amphibian groups might also

have played a role in determining their mean body size patterns.

Anurans are tailless and squat while, in general, urodeles are

more elongated. This affects heat gain since, for the same body

mass, the anuran bauplan renders a lower surface to volume ratio

and, hence, a lower capability of obtaining heat compared with

urodeles. However, as stated before, anurans appear to be able to

actively generate body temperature excesses by performing physio-

logical and/or behavioural adjustments that enhance heat gain.

Such capabilities not only help to minimize their design limita-

tions for obtaining heat, but might also convert their bauplan

into an advantage in low-energy areas because a round body

shape is better at conserving animal temperatures. Urodeles,

on the other hand, benefit from having an elongated body

that allows a better heat acquisition. However, it appears

unlikely that, by itself, this design will be able to generate

body temperature excesses comparable to those that might

be attained through basking, and accordingly we do not observe

an increase in mean body size with cold for this group, but the

converse pattern.

On the other hand, although body size trends found at the

interspecific level have not necessarily been reproduced

intraspecifically by all species, the reasoning underlying the heat

balance hypothesis could also be extended to explain intraspecific

gradients. However, available intraspecific studies reveal

Bergmann’s trends to be the most frequent ones not only for

anuran species, but also for urodeles (reviewed by Ashton, 2002;

Morrison & Hero, 2003). This indeed concords with the inter-

specific patterns we found for anurans, but contrasts with the

counter Bergmann’s ones shown by urodeles. Bearing this in

mind, as well as the reduced number of urodele species that have

been analysed intraspecifically to date [e.g. only 18 species were

included in Ashton’s (2002) review], any conclusions regarding

the extension of the heat balance hypothesis to this level will have

to wait until more investigations have been performed.

Besides the heat balance hypothesis, environmental energy

also constitutes the basis of other explanations for body size gra-

dients. Amongst them, the size dependence hypothesis, which

was specifically thought to account for interspecific patterns of

thermoregulating ectotherms (Olalla-Tárraga et al., 2006), and

similar explanations specifically tailored to the intraspecific level

(Blanckenhorn & Demont, 2004; Walters & Hassall, 2006) both

predict Bergmann’s clines below a certain body size threshold

and the converse clines beyond it. However, we can discard the

size dependence hypothesis as an explanation for the disparate

interspecific patterns between the two amphibian groups

because there are no differences in absolute body size between

anurans and urodeles (Fig. 2; Lindsey, 1966; Pough, 1980). Like-

wise, the hypothesis is also refuted at the intragroup level, since

we found no size thresholds in the relationships of body size with

energy variables within either anurans or urodeles. For anurans,

it might be the case that the species present in the studied territo-

ries were all below the size threshold marking the transition

between Bergmann’s clines and the converse ones. This would

indeed explain why we only observed Bergmann’s trends. For

urodeles, this lack of support might be related to their reduced

Figure 4 Results of partial regression analyses using the best multiple regression model and cell age as predictors of mean body size: for (a) anurans in Europe; (b) anurans in North America; (c) urodeles in Europe; (d) urodeles in North America. In each case (a) represents the independent proportion of variation in mean body size explained by the best environmental model only; (b) indicates the overlap between the best environmental model and cell age; (c) the variation added by cell age; (d) is the unexplained variance.

75



thermoregulatory abilities, while the size dependence hypothesis

was explicitly devised for ectotherms exhibiting a good control

over heat exchanges (Olalla-Tárraga et al., 2006).

An additional intraspecific explanation relating variations in

body size with the energy available in the environment is the

temperature–size rule. According to this, cold temperatures

result in increased adult body sizes by delaying maturity through

decreasing rates of growth and development (Atkinson, 1996).

This has been experimentally observed among the majority of

ectotherm species studied to date (reviewed by Atkinson, 1994),

including amphibian species (see Ashton, 2002, and references

therein). Consequently, constraints on the rates of growth and

development have frequently been proposed to explain intra-

specific Bergmann’s clines in cold-blooded animals (see e.g. Ray,

1960; Van Voorhies, 1996). However, extrapolation from laboratory

responses to geographical gradients entails the risk of elucidating

possible mechanisms before actually knowing what the patterns

look like in nature (Belk & Houston, 2002; see also Mousseau,

1997). For instance, whereas all amphibian species reared at

lower temperatures reached larger adult sizes in controlled lab-

oratory settings (see references in Ashton, 2002), some species

studied in natural conditions did not exhibit these patterns (see

references in Morrison & Hero, 2003). In the latter review, these

exceptions were said to be explained by the effects of factors such

as habitat quality or predation pressure. This, along with the

reduced number of intraspecific investigations on variation in

amphibian body size (Ashton, 2002; Morrison & Hero, 2003),

limits the generality of the explanation involving constraints

on rates of growth and development. On the other hand, even

though this mechanism may still play a role in determining

the interspecific patterns displayed by anurans, it cannot

account for the converse Bergmann’s rule gradients we found

for urodeles.

Besides, it is worthwhile noting that the patterns observed

are potentially sensitive to phylogenetic autocorrelation effects

(i.e. closely related species tend to have similar body sizes and are

not statistically independent). Usually, phylogenetic comparative

methods have been used in cross-species analyses to control for

phylogenetic structure in the data. However, our study does not

fit into this category since we conducted a spatial grid-based

analysis in which we calculated average body sizes within each

cell instead of analysing species values. This prevents the use of

traditional phylogenetic comparative methods, and emphasizes

the need for developing techniques that allow the incorporation

of phylogenetic components in community approaches (see also

Ruggiero & Hawkins, 2006). Future studies should address this

question.

Our data are also potentially relevant for an understanding of

the mechanisms behind the global decline of amphibian popula-

tions (Alford & Richards, 1999; Houlahan et al., 2000). Increased

heat and sunlight (particularly in the ultraviolet spectrum) have

been hypothesized to drive amphibian declines (Carey & Alexander,

2003), and we found that the main predictor of body size gradients

for both anurans and urodeles is PET, a measure combining the

availability of these two variables in the environment. This

suggests that the geography of amphibian body size might be

particularly sensitive to climate change, a process that involves

dramatic changes of temperature and radiation across the planet

(Houghton et al., 2001).

In sum, our results best support the heat balance hypothesis as

an adaptive explanation for the observed body size patterns in

anurans and urodeles. We propose that greater thermoregulatory

abilities among anurans allow them to reach larger sizes in low-

energy areas by actively optimizing the trade-off between heating

and cooling rates. In contrast, the lack of such strategies among

urodele faunas can explain why their mean body sizes are smaller

in the same territories. Although this hypothesis requires further

testing, it does not alter the fact that the body size patterns and

their associations with PET are highly consistent in two conti-

nents, suggesting that the patterns are not regional idiosyncrasies

but general trends.

ACKNOWLEDGEMENTS

We thank the University of California at Irvine for their hospitality

during the preparation of the data and Bradford Hawkins for

valuable discussions on the results. We are also grateful to J. Alex-

andre Diniz-Filho, Richard J. Walters, Matthew Symonds and an

anonymous referee for their suggestions and comments that

greatly improved previous versions of the manuscript. This study

was supported by the Spanish Ministry of Education and Science

(grant REN2003- 03989/GLO to M.Á.R., and FPU fellowship

AP2005-0636 to M.Á.O.-T.).

REFERENCES

Alford, R.A. & Richards, S.J. (1999) Global amphibian declines:

a problem in applied ecology. Annual Review of Ecology and

Systematics, 30, 133–165.

Angilletta, M.J., Niewiarowski, H.P., Dunham, A.E., Leaché, A.D.

& Porter, W.P. (2004) Bergmann’s clines in ectotherms:

Illustrating a life-history perspective with sceloporine lizards.


Arnold, E.N. (2002) A field guide to the reptiles and amphibians of

Britain and Europe. HarperCollins, London.

Ashton, K.G. (2002) Do amphibians follow Bergmann’s rule?

Canadian Journal of Zoology, 80, 708–716.

Ashton, K.G. & Feldman, C.R. (2003) Bergmann’s rule in

nonavian reptiles: turtles follow it, lizards and snakes reverse it.

Evolution, 57, 1151–1163.

Ashton, K.G., Tracy, M.C. & de Queiroz, A. (2000) Is Bergmann’s

rule valid for mammals? The American Naturalist, 156, 390–

415.

Atkinson, D. (1994) Temperature and organism size — a biological

law for ectotherms? Advances in Ecological Research, 25, 1–

58.

Atkinson, D. (1996) Ectotherm life history responses to develop-

mental temperature. Animals and temperature: phenotypic and

evolutionary adaptation (ed. by I.A. Johnston and A.F. Bennett),

pp. 183–204. Cambridge University Press, Cambridge.

Barlow, N.D. (1994) Size distributions of butterfly species and

the effect of latitude on species sizes. Oikos, 71, 326–332.

76



Bartholomew, G.A. (1982) Physiological control of body

temperature. Biology of the Reptilia, Volume 12 (ed. by C. Gans

and F.H. Pough), pp. 167–213. Academic Press, New York.

Behler, J.L. & King, F.W. (1994) National Audubon Society field

guide to North American reptiles and amphibians, 12th edn.

Alfred A. Knopf, Inc., New York.

Belk, M.C. & Houston, D.D. (2002) Bergmann’s rule in ecto-

therms: a test using freshwater fishes. The American Naturalist,

160, 803–808.

Bergmann, C. (1847) Über die Verhältnisse der Wärmeökonomie

der Thiere zu ihrer Grösse. Gottinger Studien, 3, 595–708.


the mammal fauna of northern North America. Ecography, 27,

715–724.




Blanckenhorn, W.U. & Demont, M. (2004) Bergmann and con-

verse Bergmann latitudinal clines in arthropods: two ends of a

continuum? Integrative and Comparative Biology, 44, 413–

424.

Boback, S.M. & Guyer, C. (2003) Empirical evidence for an optimal

body size in snakes. Evolution, 57, 345–351.

Brattstrom, B.H. (1963) Preliminary review of the thermal

requirements of amphibians. Ecology, 44, 238–255.

Brown, J.H. (1995) Macroecology. University of Chicago Press,

Chicago, IL.

Carey, C. & Alexander, M.A. (2003) Climate change and

amphibian declines: is there a link? Diversity and Distributions,

9, 111–121.

Commission for Environmental Cooperation (1997) Ecological

regions of North America: toward a common perspective.

Department of Communications and Public Outreach of the

CEC Secretariat, Montreal, Canada (http://www.epa.gov/wed/

pages/ecoregions/na_eco.htm).

Conant, R. & Collins, J.T. (1998) A field guide to reptiles and

amphibians of eastern and central North America, 3rd edn.

Houghton Mifflin, Boston, MA.

Cruz, F.B., Fitzgerald, L.A., Espinoza, R.E. & Schulte, J.A., II

(2005) The importance of phylogenetic scale in tests of Berg-

mann’s and Rapoport’s rules: lessons from a clade of South

American lizards. Journal of Evolutionary Biology, 18, 1559–

1574.

Cushman, J.H., Lawton, J.H. & Manly, B.F.J. (1993) Latitudinal

patterns in European ant assemblages: variation in species

richness and body size. Oecologia, 95, 30–37.

Darlington, P.J. (1957) Zoogeography: the geographical distri-

bution of animals. John Wiley & Sons, New York.

Digby, P.S.B. (1955) Factors affecting the temperature excess of

insects in sunshine. Journal of Experimental Biology, 32, 279–

298.

Diniz-Filho, J.A.F. & Fowler, H.G. (1998) Honey ants (genus

Myrmecocystus) macroecology: effect of spatial patterns on the

relationship between worker body size and geographic range

size. Environmental Entomology, 27, 1094–1101.




Duellman, W.E. & Trueb, L. (1986) Biology of amphibians.

McGraw-Hill, New York.

Dutilleul, P. (1993) Modifying the t-test for assessing the correla-

tion between two spatial processes. Biometrics, 49, 305–314.

Dyke, A.S., Moore, A. & Robertson, L. (2003) Deglaciation of

North America, Geological Survey of Canada (open file 1574).

Natural Resources Canada, Ottawa.

Gasc, J.P., Cabela, A., Crnobrnja-Isailovic, J., Dolmen, D.,

Grossenbacher, K., Haffner, P., Lescure, J., Martens, H.,

Martínez Rica, J.P., Maurin, H., Oliveira, M.E., Sofianidou,

T.S., Veith, M. & Zuiderwijk, A. (eds) (1997). Atlas of

amphibians and reptiles in Europe, Collection Patrimoines

Naturels, 29. Societas Europaea Herpetologica, Muséum

National d’Histoire Naturelle & Service du Petrimone

Naturel, Paris.

Hawkins, B.A. & Diniz-Filho, J.A.F. (2006) Beyond Rapoport’s

rule: evaluating range size patterns of New World birds in a

two-dimensional framework. Global Ecology and Biogeography,

15, 461–469.

Hawkins, B.A. & Lawton, J.H. (1995) Latitudinal gradients in

butterfly body sizes: is there a general pattern? Oecologia, 102,

31–36.

Hawkins, B.A. & Porter, E.E. (2003) Relative influences of

current and historical factors on mammal and bird diversity

patterns in deglaciated North America. Global Ecology and

Biogeography, 12, 475–481.

Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., van der

Linden, P.J., Dai, X., Maskell, K. & Johnson, C.A. (2001)

Climate change 2001: the scientific basis. Contribution of Working

Group I to the Third Assessment Report of the Intergovernmental

Panel on Climate Change. Cambridge University Press,

Cambridge.

Houlahan, J.E., Findley, C.S., Schmidy, B.R., Meyer, A.H. &

Kuzmin, S.L. (2000) Quantitative evidence for global amphibian

declines. Nature, 404, 752–755.

Huey, R.B. (1982) Temperature, physiology, and the ecology of

reptiles. Biology of the Reptilia, Volume 12 (ed. by C. Gans and

F.H. Pough), pp. 25–67. Academic Press, New York.

Hutchinson, V.H. & Dupré, R.K. (1992) Thermoregulation.

Environmental physiology of the amphibians (ed. by M.E. Feder

and W.W. Burggren), pp. 206–249. University of Chicago

Press, Chicago.



Lu, J., Sun, G., McNulty, S.G. & Amatya, D.M. (2005) Com-

parison of six potential evapotranspiration methods for

regional use in the southeastern United States. Journal of the

American Water Resources Association, 41, 621–633.

Mayr, E. (1956) Geographical character gradients and climatic

adaptation. Evolution, 10, 105–108.

McDowall, R.M. (1994) On size and growth in freshwater fish.

Ecology of Freshwater Fish, 3, 67–79.

McNab, B.K. (1971) On the ecological significance of Bergmann’s

rule. Ecology, 52, 845–854.

77



Meiri, S. & Dayan, T. (2003) On the validity of Bergmann’s rule.


Morrison, C. & Hero, J.M. (2003) Geographic variation in

life-history characteristics of amphibians: a review. Journal of

Animal Ecology, 72, 270–279.

Mousseau, T.A. (1997) Ectotherms follow the converse to

Bergmann’s rule. Evolution, 51, 630–632.

Neter, J., Kutner, M.H., Nachtsheim, C.J. & Wasserman, W.

(1996) Applied linear statistical models. Irwin, Chicago.

Olalla-Tárraga, M.Á., Rodríguez, M.Á. & Hawkins, B.A. (2006)

Broad-scale patterns of body size in squamate reptiles of Europe

and North America. Journal of Biogeography, 33, 781–793.

Peltier, W. (1993) Time dependent topography through the glacial

cycle, IGBP PAGES/World Data Center-A for Paleoclimatology

Data Contribution Series no. 93-015. NOAA/NGDC Paleo-

climatology Program, Boulder, CO.

Pough, F.H. (1980) The advantages of ectothermy for tetrapods.


Rangel, T.F.L.V.B., Diniz-Filho, J.A.F. & Bini, L.M. (2006)

Towards an integrated computational tool for spatial ana-

lysis in macroecology and biogeography. Global Ecology and

Biogeography, 15, 321–327.

Ray, C. (1960) The application of Bergmann’s and Allen’s rule to

the poikilotherms. Journal of Morphology, 106, 85–108.

Reed, R.N. (2003) Interspecific patterns of species richness,

geographic range size, and body size among New World

venomous snakes. Ecography, 26, 107–117.

Rensch, B. (1938) Some problems of geographical variation

and species-formation. Proceedings of the Linnean Society of

London, 150, 275–285.

Rodríguez, M.Á., Belmontes, J.A. & Hawkins, B.A. (2005)

Energy, water and large-scale patterns of reptile and amphibian

species richness in Europe. Acta Oecologica, 28, 65–70.

Rodríguez, M.Á., López-Sañudo, I.L. & Hawkins, B.A. (2006)

The geographic distribution of mammal body size in Europe.


Rosenzweig, M.L. (1968) The strategy of body size in mam-

malian carnivores. American Midland Naturalist, 80, 299–

315.

Ruggiero, A. & Hawkins, B.A. (2006) Mapping macroecology.


Spotila, J.R. (1972) Role of temperature and water in the

ecology of lungless salamanders. Ecological Monographs, 42,

95–125.

StatSoft Inc. (2003) STATISTICA (data analysis software system),

version 6 (http://www.statsoft.com).

Stebbins, R.C. (2003) A field guide to western reptiles and amphibians,

3rd edn. Houghton Mifflin, Boston, MA.

Stebbins, R.C. & Cohen, N.W. (1995) A natural history of

amphibians. Princeton University Press, Princeton, NJ.

Storey, K.B. & Storey, J.M. (1996) Natural freezing survival in

animals. Annual Review of Ecology and Systematics, 27, 365–

386.

Turner, J.R.G. & Hawkins, B.A. (2004) The global diversity gradient.

Frontiers of biogeography: new directions in the geography

of nature (ed. M.V. Lomolino and L.R. Heaney), pp. 171–190.

Sinauer, Sunderland, MA.

Van Voorhies, W.A. (1996) Bergmann size clines: a simple

explanation for their occurrence in ectotherms. Evolution, 50,

1259–1264.

Walters, R.J. & Hassall, M. (2006) The temperature-size rule in

ectotherms: may a general explanation exist after all? The

American Naturalist, 167, 510–523.

Editor: José Alexandre F. Diniz-Filho

SUPPLEMENTARY MATERIAL

The following supplementary material is available for this article:

Appendix S1 Geographical patterns of anuran mean body size

in North America at 110 × 110 km including the five species with

freezing abilities (see main text) and corresponding multiple

regression models

Appendix S2 Species list and body size data

Appendix S3 Multiple regression models for mean body

size using mean annual temperature as a predictor instead of

PET

Appendix S4 Geographical patterns of species richness at 110 ×110 km for anurans and urodeles in Europe and North America

(Figure S4) and multiple regression results using species richness

values (S) and [1 – log10(1/S2)] as weighting factors (Table S4).

This material is available as part of the online article from:

http://www.blackwell-synergy.com/doi/abs/

10.1111/j.1466-8238.2007.00309.x

(This link will take you to the article abstract).

Please note: Blackwell Publishing is not responsible for the

content or functionality of any supplementary materials supplied

by the authors. Any queries (other than missing material) should

be directed to the corresponding author for the article.

BIOSKETCHES

Miguel Á. Olalla-Tárraga is a PhD student at the

University of Alcalá with a particular interest in

macroecology and conservation biogeography.

Miguel Á. Rodríguez’s main interests include the

study of factors and processes conditioning patterns of

biodiversity at local, regional and global scales. Recent

research has involved investigating the effects of habitat

destruction and fragmentation on aggregate properties

of faunas at broad scales.

78

Appendix S1. Geographical patterns of anuran mean body size in North America (data

resolved at 110 x 110 km) including all species (S=86); i.e. without eliminating the five

frogs with freezing abilities (see main text). The map was built using cell averages of log10-

transformed length values. Numbers in the legend are lengths in centimetres obtained

after antilog transformation. When applied to these data, our best model (see Table 2)

explained 21.1% of variance (Mean Body Size = 0.144*PET - 0.528*GSL; N=1314),

while a full model explained 37.4% (Mean Body Size = 0.233*PET - 0.496*AET +

0233*Precipitation + 0.035*GVI - 0.178*GSL + 0.286*Elevation Range; N=1314).

79

Appendix S2. Body size estimates of the amphibian species present in Europe and

North America. Body sizes are provided as maximum snout-to-vent (SVL) lengths

for anurans, and as maximum total lengths (TL) for urodeles. To allow

comparisons between both length measurements, available SVL lengths for 42

urodele species from western North America are also provided.

(a) Anurans in Europe

Species SVL max (mm) Alytes cisternasii 39 Alytes obstetricans 55 Bombina bombina 53 Bombina variegata 56 Discoglossus galganoi 75 Discoglossus sardus 70 Discoglossus pictus 80 Pelobates cultripes 95 Pelobates fuscus 73 Pelobates syriacus 85 Pelodytes punctatus 40 Bufo bufo 113 Bufo calamita 75 Bufo viridis 95 Hyla arborea 45 Hyla meridionalis 58 Rana arvalis 75 Rana balcanica 10 Rana dalmatina 85 Rana epeirotica 79 Rana kl. Esculenta 105 Rana graeca 75 Rana iberica 48 Rana italica 65 Rana latastei 65 Rana lessonae 75 Rana macrocnemis 75 Rana perezi 90 Rana ridibunda 160 Rana shqiperica 72 Rana temporaria 109

80

(b) Urodeles in Europe

Species TL max (mm) Chioglossa lusitanica 160 Euproctus asper 160 Hydromantes italicus 125 Hydromates ambrosii 125 Pleurodeles waltl 300 Proteus anguinus 250 Salamandra atra 150 Salamandra lanzai 170 Salamandra salamandra 200 Salamandra terdigitata 102 Triturus alpestris 120 Triturus boscai 100 Triturus helveticus 95 Triturus italicus 80 Triturus marmoratus 160 Triturus montandoni 100 Triturus supersp cristatus 150 Triturus vulgaris 110

(c) Anurans in North America

Species SVL max (mm) Acris crepitans 37.3 Acris gryllus 28.5 Ascaphus truei 51 Bufo alvarius 190 Bufo americanus 111 Bufo boreas 127 Bufo californicus 86 Bufo canorus 70 Bufo cognatus 114 Bufo debilis 51.8 Bufo exsul 76 Bufo hemiophrys 79.5 Bufo houstonensis 79 Bufo marinus 238 Bufo microscaphus 77 Bufo nebulifer 130 Bufo nelsoni 112 Bufo punctatus 76

81

Bufo quercicus 33 Bufo retiformis 71 Bufo speciosus 90.5 Bufo terrestris 113 Bufo woodhousii 119 Eleutherodactylus augusti 95 Eleutherodactylus cystignathoides 25 Eleutherodactylus guttilatus 32 Eleutherodactylus marnockii 38 Gastrophryne carolinensis 38 Gastrophryne olivacea 41 Hyla andersonii 51 Hyla arenicolor 57 Hyla avivoca 52 Hyla chrysoscelis 60 Hyla cinerea 64 Hyla eximia 51 Hyla femoralis 44 Hyla gratiosa 70 Hyla squirella 41 Hyla versicolor 60 Hypopachus variolosus 44 Leptodactylus labialis 51 Pseudacris brachyphona 38 Pseudacris brimleyi 32 Pseudacris cadaverina 51 Pseudacris clarkii 32 Pseudacris crucifer 37 Pseudacris nigrita 32 Pseudacris ocularis 17 Pseudacris ornata 37 Pseudacris regilla 51 Pseudacris streckeri 48 Pseudacris triseriata 39.5 Pternohyla fodiens 63 Rana areolata 92 Rana aurora 133 Rana berlandieri 114 Rana blairi 111 Rana boylii 81 Rana capito 103 Rana cascadae 76 Rana catesbeiana 203 Rana chiricahuensis 137

82

Rana clamitans 101.8 Rana grylio 162 Rana heckscheri 155 Rana luteiventris 101 Rana muscosa 89 Rana okaloosae 49 Rana onca 89 Rana palustris 87 Rana pipiens 111 Rana pretiosa 101 Rana septentrionalis 76 Rana sphenocephala 127 Rana sylvatica 82.5 Rana tarahumarae 114 Rana virgatipes 67 Rana yavapaiensis 86 Rhinophrynus dorsalis 89 Scaphiopus couchii 90 Scaphiopus holbrookii 78 Smilisca baudinii 90 Spea bombifrons 63.5 Spea hammondii 63 Spea intermontana 63 Spea multiplicata 63.5

(d) Urodeles in North America

Species SVL max (mm) TL max (mm) Ambystoma annulatum 238

Ambystoma barbouri 178

Ambystoma californiense 127 216

Ambystoma cingulatum 129

Ambystoma gracile 132 260

Ambystoma jeffersonianum 210

Ambystoma laterale 160

Ambystoma mabeei 100

Ambystoma macrodactylum 89 170

Ambystoma maculatum 248

Ambystoma opacum 127

Ambystoma talpoideum 122

83

Ambystoma texanum 178

Ambystoma tigrinum 165 330

Amphiuma means 1162

Amphiuma pholeter 330

Amphiuma tridactylum 1060

Aneides aeneus 140

Aneides ferreus 133

Aneides flavipunctatus 95 167

Aneides hardii 63 114

Aneides lugubris 101 184

Aneides vagrans 76 127

Batrachoseps attenuatus 47 140

Batrachoseps campi 61 107

Batrachoseps diabolicus 47 96

Batrachoseps gabrieli 51 110

Batrachoseps gavilanensis 66 165

Batrachoseps gregarius 46 92

Batrachoseps incognitus 48 96

Batrachoseps kawia 47 94

Batrachoseps luciae 46 126

Batrachoseps major 59 162

Batrachoseps minor 58 116

Batrachoseps nigriventris 47 141

Batrachoseps regius 35 70

Batrachoseps relictus 47 120

Batrachoseps robustus 57 100

Batrachoseps simatus 56 125

Batrachoseps stebbinsi 61 122

Batrachoseps wrighti 61 108

Cryptobranchus alleganiensis 740

Desmognathus aeneus 57

Desmognathus apalachicolae 120

Desmognathus auriculatus 163

Desmognathus brimleyorum 178

Desmognathus fuscus 141

84

Desmognathus imitator 110

Desmognathus marmoratus 146

Desmognathus monticola 149

Desmognathus ochrophaeus 111

Desmognathus quadramaculatus 210

Desmognathus santeetlah 130

Desmognathus welteri 170

Desmognathus wrighti 51

Dicamptodon aterrimus 330.2

Dicamptodon copei 120 191

Dicamptodon ensatus 173 300

Dicamptodon tenebrosus 355.6

Ensatina eschscholtzii 81 149

Eurycea aquatica 184

Eurycea bislineata 121

Eurycea cirrigera 110

Eurycea junaluska 90

Eurycea longicauda 200

Eurycea lucifuga 181

Eurycea multiplicata 102

Eurycea nana 51

Eurycea neotenes 100

Eurycea quadridigitata 90

Eurycea rathbuni 137

Eurycea robusta 101

Eurycea tridentifera 85

Eurycea tynerensis 81

Eurycea wilderae 121

Gyrinophilus palleucus 227

Gyrinophilus porphyriticus 232

Gyrinophilus subterraneus 184

Haideotriton wallacei 76

Hemidactylium scutatum 102

Hydromantes brunus 76 111

Hydromantes platycephalus 70 114

85

Hydromantes shastae 63 108

Necturus alabamensis 216

Necturus beyeri complex 222

Necturus lewisi 276

Necturus maculosus 486

Necturus punctatus 189

Notophthalmus meridionalis 110

Notophthalmus perstriatus 79

Notophthalmus viridescens 140

Phaeognathus hubrichti 256

Plethodon aureolus 151

Plethodon caddoensis 111

Plethodon cinereus 127

Plethodon dorsalis 111

Plethodon dunni 86 150

Plethodon elongatus 76 152

Plethodon fourchensis 178

Plethodon glutinosus 206

Plethodon hoffmani 137

Plethodon hubrichti 131

Plethodon idahoensis 66 113

Plethodon jordani 184

Plethodon kentucki 168

Plethodon larselli 54 103

Plethodon neomexicanus 67 143

Plethodon nettingi 111

Plethodon ouachitae 159

Plethodon petraeus 183

Plethodon punctatus 171

Plethodon richmondi 143

Plethodon serratus 125

Plethodon shenandoah 102

Plethodon teyahalee 207

Plethodon vandykei 63 123

Plethodon vehiculum 63 115

86

Plethodon websteri 110

Plethodon wehrlei 168

Plethodon welleri 92

Plethodon yonahlossee 221

Pseudobranchus striatus 251

Pseudotriton montanus 207

Pseudotriton ruber 181

Rhyacotriton cascadae 100

Rhyacotriton kezeri 100

Rhyacotriton olympicus 117

Rhyacotriton variegatus 100

Siren intermedia 686

Siren lacertina 978

Stereochilus marginatus 114

Taricha granulosa 89 216

Taricha rivularis 89 194

Taricha torosa 89 197

Typhlotriton spelaeus 135

87

A

ppen

dix

S3. M

ultip

le r

egre

ssio

n m

odel

s fo

r m

ean

body

siz

e (d

ata

reso

lved

to 1

10 x

110

km

) usi

ng m

ean

annu

al te

mpe

ratu

re a

s th

e pr

edic

tor

inst

ead

of P

ET

. Pre

sent

atio

n as

in T

able

2 (s

ee m

ain

text

).

Gro

up

Reg

ion

Pre

dict

ors

in m

odel

Anu

rans

Tem

p R

ange

A

ET

P

rec

GV

I G

SL

A

IC

ΔAIC

W

i R

2 E

urop

e

-0.6

09

-0.0

71

-0.3

51

0.17

5 -0

.216

0.

268

-9

89

0 0.

585

0.67

3 -0

.616

-0.3

81

0.14

3 -0

.203

0.

295

-9

88

1 0.

414

0.67

0

Nor

th A

mer

ica

-0.5

09

0.04

3

-0.2

77

-2

476

0 0.

463

0.47

3 -0

.514

-0

.250

-247

5 1

0.25

4 0.

472

-0.5

09

0.04

3 0.

001

-0.2

78

-2

474

2 0.

168

0.47

3

Uro

dele

s E

urop

e

0.68

8 0.

119

0.47

3

-0.3

10

-0.1

47

-8

86

0 0.

580

0.51

0 0.

688

0.11

9 0.

473

0.00

1 -0

.310

-0

.147

-884

2

0.20

4 0.

510

N

orth

Am

eric

a 0.

818

-0.1

92

-0

.191

-0

.132

-8

25

0 0.

465

0.47

8

0.

906

-0.1

86

-0

.187

-0

.135

-0

.098

-824

1

0.40

6 0.

471

88

Figure S4. Geographical patterns of species richness at 110 x 110 km. (a) Anurans in

Europe (S=31); (b) anurans in North America (S =81); (c) urodeles in Europe(S =18); (d)

urodeles in North America (S=135).

89

Tab

le S

4.1.

Bec

ause

anu

ran

and

urod

ele

spec

ies

are

not u

nifo

rmly

dis

trib

uted

in E

urop

e an

d N

orth

Am

eric

a (F

igur

e S4

abo

ve),

we

chec

ked

for

the

effe

ct t

hat

vary

ing

spec

ies

rich

ness

es a

cros

s ce

lls h

ave

had

on r

egre

ssio

n re

sults

. Sp

ecifi

cally

, w

e us

ed w

eigh

ted-

LS

regr

essi

on

tech

niqu

es (

see

Stat

Soft,

Inc

. 20

03 f

or d

etai

ls)

to r

ecal

cula

te t

he s

tand

ardi

zed

regr

essi

on c

oeffi

cien

ts o

f th

e va

riab

les

incl

uded

in

our

best

m

odel

s, a

s w

ell

as t

he c

oeffi

cien

t of

det

erm

inat

ion

of e

ach

mod

el (

R2 ).

We

used

anu

ran

and

urod

ele

cell

spec

ies

rich

ness

es a

s w

eigh

ting

fact

ors

in tw

o di

ffere

nt w

ays,

one

that

sm

ooth

es t

he v

aria

tion

of th

is v

aria

ble

acro

ss c

ells

(1-

Log

10[1

/S2 ])

, and

oth

er, m

ore

seve

re, t

hat d

irec

tly

uses

spe

cies

ric

hnes

s va

lues

(S)

. T

he r

esul

ts a

re p

rese

nted

in

the

tabl

e be

low

. In

ord

er t

o fa

cilit

ate

com

pari

sons

, th

e un

wei

ghte

d re

sults

di

scus

sed

in th

e m

ain

text

are

als

o in

clud

ed (i

n ita

lics)

. As

can

be s

een,

all

anal

yses

gen

erat

e ve

ry a

like

resu

lts, t

hus

indi

catin

g th

at th

e ob

serv

ed

rela

tions

hips

bet

wee

n an

uran

and

uro

dele

mea

n bo

dy s

izes

and

env

iron

men

tal

pred

icto

rs a

re r

obus

t, an

d no

t a

cons

eque

nce

of t

he

geog

raph

ic v

aria

tion

in s

peci

es r

ichn

ess.

Pre

dict

or v

aria

ble

code

s ar

e: P

ET

= a

nnua

l pot

entia

l eva

potr

ansp

irat

ion,

Ran

ge =

ran

ge in

ele

vatio

n,

AE

T =

ann

ual a

ctua

l eva

potr

ansp

irat

ion,

Pre

c =

annu

al p

reci

pita

tion,

GV

I =

Glo

bal V

eget

atio

n In

dex,

GSL

= g

row

ing

seas

on le

ngth

.

Gro

up

Reg

ion

Wei

ghtin

g va

riab

le

Pre

dict

ors

in m

odel

R

2

P

ET

R

ange

A

ET

P

rec

GV

I G

SL

A

nura

ns

Eur

ope

1-L

og10(1

/S2 )

-0.6

10

0.1

34

-0.1

61

-0

.163

-0

.023

0.6

41

S -0

.593

0

.095

-0

.105

-0.1

73

-0.1

11

0

.584

none

-0

.628

0

.155

-0

.171

-0.1

68

0.0

63

0

.662

N

orth

Am

eric

a 1-

Log

10(1

/S2 )

-0.4

88

-0.3

58

0

.520

S -0

.504

-0

.410

0.5

98

no

ne

-0.4

75

-0.3

21

0

.469

Uro

dele

s E

urop

e 1-

Log

10(1

/S2 )

0.7

72

-0.1

49

0.1

83

0.0

61

-0.2

76

0.0

23

0

.541

S

0.7

69

-0.1

93

0.1

47

0.0

80

-0.2

55

-0.0

25

0

.513

none

0

.757

-0

.092

0

.205

0

.057

-0

.320

0

.103

0.5

67

Nor

th A

mer

ica

1-L

og10(1

/S2 )

0.6

26

-0.2

59

-0

.110

0.4

52

S

0.4

78

-0.4

03

-0

.082

0.4

52

no

ne

0.6

92

-0.1

92

-0

.106

0.4

75

90

Capítulo 5 Gradientes geográficos de tamaño corporal en regiones tropicales: déficit hídrico y tamaño corporal de anuros en el Cerrado Brasileño Este capítulo reproduce íntegramente el texto del siguiente manuscrito: Olalla-Tárraga, M. Á., J. A. F. Diniz-Filho, R. P. Bastos & M. Á. Rodríguez. 2009. Geographic body size gradients in tropical regions: Water deficit and anuran body size in the Brazilian Cerrado. Ecography 32: DOI. 10.1111/j.1600-0587.2008.05632.x

Geographic body size gradients in tropical regions: Water deficit and

anuran body size in the Brazilian Cerrado

Miguel Á. Olalla-Tárraga1,* José Alexandre F. Diniz-Filho2, Rogério P. Bastos2, Miguel Á.

Rodríguez1

1Department of Ecology, University of Alcalá, Alcalá de Henares, 28871, SPAIN

2Departamento de Biologia Geral, ICB, Universidade Federal de Goiás, CP 131, 74.001-

970, Goiânia, GO, Brazil

ABSTRACT A recent interspecific study found Bergmann’s size clines for Holarctic anurans and proposed an explanation based on heat balance to account for the pattern. However, this analysis was limited to cold temperate regions, and exploring the patterns in warmer tropical climates may reveal other factors that also influence anuran body size variation. We address this using a Cerrado anuran database. We examine the relationship between mean body size in a grid of 1° cells and environmental predictors and test the relative support for four hypotheses using an AIC-based model selection approach. Also, we considered three different amphibian phylogenies to partition the phylogenetic and specific components of the interspecific variation in body size using a method analogous to Phylogenetic eigenVector Regression (PVR). To consider the potential effects of spatial autocorrelation we use eigenvector-based spatial filters. We found the largest species inhabiting high water deficit areas in the northeast and the smallest in the wet southwest. Our results are consistent with the water availability hypothesis which, coupled with previous findings, suggests that the major determinant of interspecific body size variation in anurans switches from energy to water towards the equator. We propose that anuran body size gradients reflect effects of reduced surface to volume ratios in larger species to control both heat and water balance. Keywords: Bergmann’s rule; heat balance hypothesis; macroecology; Phylogenetic

eigenVector Regression; South America; spatial filtering; water availability hypothesis.

91

Anuran body size variation in the Brazilian Cerrado

INTRODUCTION Body size gradients and thermoregulation have been conceptually linked since Bergmann (1847) proposed that cold macroclimates should harbor more large-bodied species than warmer areas, because the reduced surface area to volume ratios of larger animals would allow them to maintain body temperatures. Several interspecific studies in temperate to cold climates have supported Bergmann’s rule and its associated heat conservation mechanism in endothermic organisms (mammals: Blackburn & Hawkins 2004, Rodríguez et al. 2006, Diniz-Filho et al. 2007; birds: Ramírez et al., 2008). However, the interspecific formulation of Bergmann’s rule does not hold for endotherms in warm climates (Rodríguez et al. 2006, Medina et al. 2007, Greve et al. 2008, Rodríguez et al. 2008). On the other hand, although amphibians and reptiles remain largely underrepresented in the macroecological literature in comparison to mammals and birds, research on broad-scale body size gradients has not been exclusively limited to endotherms (Ashton 2002, Adams & Church 2008, Olalla-Tárraga et al. 2006, Olalla-Tárraga & Rodríguez 2007). Indeed, studies focused on vertebrates with limited capabilities for internal heat production are not only helpful to evaluate the generality of the patterns documented for mammals and birds but can generate insights into the energetic and physiological mechanisms that are likely determining the spatial variation in body size across large scales.

In ectotherms, a relationship between thermoregulation and Bergmann’s rule has also been supported to some extent in amphibians and reptiles since Lindsey (1966) first noted that some ectothermic vertebrate taxa show interspecific body size gradients in latitudinal space. For instance, anurans and lizards are believed to be thermoregulators (i.e. show behavioral control over their body temperatures through basking or selecting appropriate microhabitats) (Brattstrom 1963, Huey 1982, Hutchinson & Dupré 1992, but see Navas 2003) and Holarctic species in these groups increase in size towards colder macroclimates (Olalla-Tárraga et al. 2006, Olalla-Tárraga &

Rodríguez 2007). In contrast, urodeles, which are believed to be thermoconformers (Brattstrom 1963, Hutchinson & Dupré 1992), are smaller in colder areas of the Holarctic region (Olalla-Tárraga & Rodríguez 2007), so thermoregulating abilities may also determine interspecific body size gradients in ectotherms. Nonetheless, ectotherms, including the best thermoregulators, may be less able to control body temperatures than endotherms. Therefore, it is not surprising that anurans and lizards have shallower size gradients and weaker relationships with ambient energy variables than endotherms (see Olalla-Tárraga et al. 2006, Olalla-Tárraga & Rodríguez 2007), nor perhaps that snakes, which also thermoregulate, have shown the converse to Bergmann’s rule in the Holarctic (Olalla-Tárraga et al. 2006). In this latter case, the exception of snakes to the Bergmannian pattern displayed by anurans and lizards suggests that not only thermoregulation but the relative size of an organism determine the concordance with Bergmann’s rule in thermoregulating ectotherms. That is, considering that snakes are generally larger than lizards and anurans, Bergmann’s size clines might only be expected within groups of small sized species because longer heating times of larger animals limit the time available for other activities in colder high-latitude environments (see Olalla-Tárraga et al. 2006 for further discussion).

On the other hand, studies on amphibians and reptiles have been restricted to the Holarctic, which limits an extension of their conclusions to other regions. In this paper our primary goal is to document the body size gradients of anurans in a Neotropical region to compare with the patterns previously found for the Holarctic. Even though the Neotropical Realm has greater species richness of anurans than any other biogeographical region in the world (Duellmann 1988), our knowledge of the distribution and ecology of the anuran faunas in the Neotropics is still rather limited (see e.g. Ron 2000, Eterovick et al. 2005). In fact, this group is not an exception to the well-known Wallacean and Linnean shortfalls (i.e. a high number of species remain formally undescribed, and there is a paucity of distributional data for most species) (Bini et al.

92


2006). Furthermore, the lack of biological data is also apparent if we consider the difficulties in compiling body size data, a fundamental organismal trait, for many tropical anuran species (Cooper et al. 2008). These limitations prevent an analysis for the whole Neotropical region and lead us to focus on a single biome, the Cerrado, where enough biological information (extents of occurrence and body size) is ready available for all anuran species described so far (Diniz-Filho et al. 2004, Diniz-Filho et al. 2005). After the Amazon rainforest, this is the second largest biome in Brazil (covers approx. 204 million hectares) and comprises enough environmental heterogeneity (Nimer 1989, Silva et al. 2006) that the potential consequences for anuran body size distributions should be apparent.

Here, we use an assemblage approach by examining the variation in average body size of all anuran species occurring in a 1° x 1° grid of 181 cells in the Cerrado. Additionally, we use an information theoretic approach to investigate the relative support for a set of competing hypothesis that have been proposed to explain interspecific broad-scale body size gradients (Cushman 1993, Blackburn et al. 1999, Olalla-Tárraga & Rodríguez 2007). Specifically, we evaluate the following four potential explanations:

1. Heat balance: The heat conservation mechanism originally proposed to explain Bergmann’s rule in endotherms was considered unlikely for ectotherms as their body temperatures rely both on heat loss and heat gain (Cushman et al. 1993). The heat balance hypothesis (Olalla-Tárraga & Rodríguez 2007) incorporates both processes and proposes interspecific Bergmann’s rule patterns among those ectothermic taxa exhibiting thermoregulatory abilities when the larger species are below a certain body size threshold. Anurans are below this size threshold and many species thermoregulate, so this hypothesis predicts a negative relationship between mean body size and measures of energy availability in the environment (Olalla-Tárraga & Rodríguez 2007).

2. Water availability: Amphibians are strongly dependent on water, which may determine their size patterns in the tropics. The rationale for this hypothesis is that the reduced surface-to-volume ratio of larger species makes them less prone to water loss, conferring a stronger dessication tolerance in dryer areas (Nevo 1973, Ashton 2002). This hypothesis may be of special relevance in tropical regions with a pronounced dry season.

3. Primary productivity: In tropical climates, primary productivity is known to vary strongly depending on seasonality. Consequently, the productivity hypothesis predicts that large-bodied species will concentrate in highly productive environments because in less productive areas there may be insufficient food supply to maintain species with large body size (Rosenzweig 1968). This hypothesis was originally developed for mammals, which have much higher energetic requirements than anurans (Pough 1980). Hence, although a possible explanation that deserves testing, its relevance to anurans is unclear.

4. Habitat availability (topography x macroclimate interaction): This hypothesis has been recently proposed to explain body size gradients of mammals in the Neotropical region. Mammals conform to Bergmann’s rule in the northern parts of the Palearctic (Rodríguez et al. 2006) and the Nearctic (Blackburn & Hawkins 2004, Rodríguez et al. 2008), but not in the southern Nearctic and the Neotropics (Rodríguez et al. 2008). In the latter region, the geographic distributions of large-sized species are limited by the strong habitat zonation and reduced habitat areas in tropical/subtropical mountains and accumulate in the extensive habitat areas of the lowlands (Rodríguez et al. 2008). That is, mammal body size variation may be conditioned by topography and the effects that it exerts on habitat sizes. This hypothesis

93


predicts larger bodied assemblages in the lowlands than in the mountains, although it should be noted that all anurans are small compared to most mammals.

METHODS Study area

The Cerrado is the most extensive woodland-savanna in South America (Silva & Bates 2002). Although it ranges in elevation from 300 to 2000 m. a.s.l., the topography is generally flat and only a few locations reach altitudes beyond 900 m (Clark 1992). The most important climatic feature of the Cerrado is its strong seasonality, with a dry period extending for up to six months, which has resulted in faunas and floras adapted to persist under prolonged dry conditions (see e.g. Eiten 1972, Colli et al. 2002). In fact, because of the large number of endemic plant and vertebrate species (especially amphibians and reptiles, see Silva & Bates 2002), the Cerrado has been considered a global biodiversity hotspot, the only one dominated by tropical savanna and dry forests (Myers et al. 2000).

Anuran data

We used a distributional database for Cerrado anurans (Diniz-Filho et al. 2004, Diniz-Filho et al. 2005). The database comprises 131 species belonging to eleven families (Aromobatidae, Brachycephalidae, Bufonidae, Cycloramphidae, Dendrobatidae, Hylidae, Hylodidae, Leiuperidae, Leptodactylidae, Microhylidae and Ranidae) (Frost et al. 2006). We mapped the geographic distribution (extent of occurrence) of each species in a 1° x 1° grid of 181 cells. For each species, we also compiled mean body size data (snout-to-vent length) (see Colli et al. 2002, Diniz-Filho et al. 2004 and Diniz-Filho et al. 2005) and assigned a single body size measure (log10-transformed) to the distribution range of the species (see Appendix S1). We then calculated the average log10 body length (hereafter mean body size) across all species occurring in each grid cell (Blackburn & Hawkins 2004, Olalla-Tárraga et al. 2006, Olalla-Tárraga & Rodríguez 2007). It should be noted that our analyses do not incorporate

intraspecific variation in body size since we assume that intraspecific spatial variation in body size is small relative to interspecific variation (see e.g. Blackburn et al. 1999). Because we have insufficient data to validate fully this assumption, we did the following basic calculations to support our approach. We calculated the maximum to minimum body size ratio at the interspecific and intraspecific level for those species for which data exist. Using the complete Cerrado anurans database (131 species) we found a ratio of 13.63 for the interspecific level. We then estimated the intraspecific ratio using a subset of 16 species for which we have within-species variation in body size and we found the largest ratio for Leptodactylus podicipinus (2.25), although most ranged from 1 to 1.5. Thus, interspecific variation in Cerrado anurans appears to be much larger than intraspecific variation. Therefore, even if any intraspecific variation is geographically structured we not expect a strong influence in our analyses. This is a methodological assumption in most studies aimed to document body size patterns at an interspecific level and hence, albeit similar mechanisms may be operating at both taxonomic levels, our results must be interpreted in an explicitly interspecific context (Gaston 2008). Hypotheses and environmental variables

We selected eight environmental predictors

to evaluate the relative support for each hypothesis. They were obtained from GIS information layers in raster format and processed in ArcGIS 8.2 (ESRI, 2002) to calculate average values for each grid cell. The hypotheses and their associated variables are:

(1) Heat balance hypothesis: we used minimum and mean annual temperatures as heat indicators and annual potential evapotranspiration (PET) as a combined measure of heat plus solar radiation inputs in the environment. Temperatures were obtained from a global 0.5° resolution database interpolated from weather station data for the period from1961 to 1990 (New et al. 1999). The original PET dataset is based on a 50 year

94


time series from weather stations (1950-1999) interpolated at a spatial resolution of 0.5° (Willmott & Kenji 2001). This variable was generated using a modified version of the Thornthwaite’s formula (Willmott & Kenji 2001). (2) Water availability: Annual precipitation and relative humidity were included as direct measures of water and moisture conditions in cells, and water deficit (WD) was used to measure the joint availability of biologically usable energy and water (Stephenson 1998, Rodríguez et al. 2005). Both the precipitation and relative humidity datasets were taken from New et al. (1999) and were originally generated from spatial interpolations of weather station data at 0.5° resolution for the period 1961-1990.

We calculated WD as PET minus annual actual evapotranspiration (AET) (Stephenson 1998; Francis & Currie 2003). AET data were obtained from Willmott & Kenji (2001). (3) Primary productivity: we used the Enhanced Vegetation Index (EVI), as a surrogate for net primary productivity. This and other vegetation indices have been widely used as indicators of plant productivity or standing crop in macroecological studies (Hurlbert & Haskell 2003, Hawkins 2004, Olalla-Tárraga & Rodríguez 2007, Duro et al. 2007). The Enhanced Vegetation Index is generated from data collected by the Moderate Resolution Imaging Spectroradiometer (MODIS), and, similar to other vegetation indices, measures the greenness and density of the canopy, total standing biomass, green leaf area, and percent vegetation cover (Ferreira et al. 2003, Ratana et al. 2005). However, unlike other indices, EVI does not saturate in high biomass regions (Huete et al. 2002). EVI was calculated as monthly averaged composites for 2004-2005. (4) Habitat availability: Elevation range is often used as an indirect measure of the effects of mesoscale climatic gradients on habitat variability (Hawkins & Diniz-Filho 2006, Rodríguez et al. 2008). Based on Janzen’s hypothesis (1967), these authors argued that, in comparison to temperate montane areas, the

greater climatic zonation associated with elevational changes in tropical regions generates sharp mesoscale habitat gradients. To incorporate the interaction between macroclimate and topographic relief and capture varying effects depending on the local inputs of environmental energy, together with elevation range we included its interaction with PET in multifactor models (see Hawkins & Diniz-Filho 2006 and Rodríguez et al. 2008). Elevation data were obtained from Clark (1992) and have an original resolution of 10 minutes. Statistical analyses

Initially, we used Pearson correlations to explore the relationships between anuran mean body size and the environmental predictors. We then used ordinary least-squares multiple regression (OLS) to generate alternative models with different combinations of predictors and performed a model selection approach based on information theory to identify the best model using AIC (Burnham & Anderson 2002, Johnson & Omland 2004, Stephens et al. 2007). We also computed the ∆AIC of each model (i.e. ∆AICi= AICi– minAIC) to identify those models that are as good as the best model (i.e. with ∆AIC ≤ 2) (Burnham & Anderson, 2002). Also, to evaluate the contribution of each variable to the models and assess the relative support for each hypothesis, we used the standardized regression coefficients of the variables in the best models (Neter et al. 1996, Olalla-Tárraga et al. 2006). To minimize multicollinearity, we did not include highly correlated variables in multiple regression analyses. As a consequence, minimum and mean annual temperatures, highly correlated with PET, as well as precipitation, highly collinear with relative humidity (Table 1), were excluded. Additionally, in order to detect the presence of collinearity in our models, we calculated for each of them its condition number (CN) and the variance inflation factors (VIFs) of the predictors it includes (Belsey 1991, Neter et al. 1996, Lazaridis 2007). Variance inflation factors measure the degree to which collinearity inflates the estimated regression coefficients as compared to orthogonal predictors. The condition number is a commonly used index of the global

95


instability of the regression coefficients calculated as the square root of the largest eigenvalue divided by the smallest eigenvalue of the correlation matrix. Typically, maximum

VIF values lower than 10 and a CN lower than 5 indicate that collinearity is not a major problem (Belsey 1991, Neter et al. 1996, Lazaridis 2007).

Table 1 Pearson correlation coefficients of environmental variables and mean body size (snout-to-vent length)

Variable Anuran

SVL Prec Rel.

humid WD Mean temp

Min. temp PET Range

Annual precipitation <0.001 1 Relative humidity -0.061 0.703 1 Water deficit 0.555 -0.329 -0.406 1 Mean annual temperature 0.170 -0.001 0.100 0.596 1 Minimum temperature 0.263 0.071 0.109 0.667 0.960 1 Potential evapotranspiration 0.300 -0.093 -0.030 0.745 0.821 0.815 1 Elevation range -0.260 -0.216 -0.189 -0.181 -0.417 -0.381 -0.361 1 Enhanced Vegetation Index 0.063 0.333 0.375 -0.025 0.195 0.200 0.076 -0.124

Because cells in our spatial grid-based

approach may not be fully independent, statistical problems associated with spatial autocorrelation are possible (Legendre 1993). To take this into account, we included eigenvector-based spatial filters truncated at the 1,000 km distance as predictors in the regression models (the PCNM model – see Borcard & Legendre 2002). These spatial filters have proven to be an effective tool to capture the spatial structure of the data at different spatial scales (Diniz-Filho & Bini 2005, Griffith & Peres-Neto 2006, Dormann et al. 2007). We used the maximization of the coefficients of determination in the regression between mean body size and the filters and the minimization in residual spatial autocorrelation as criteria to select non-redundant filters for the models and avoid “overcorrecting” for spatial autocorrelation (Diniz-Filho & Bini 2005, Griffith 2003). Because of strong collinearity between some filters and the environmental variables in the fitted model, we also excluded filters that were non-significant when added to the original environmental model. Additionally, we used spatial correlograms to evaluate if this approach removed the spatial autocorrelation of the residuals of our best multiple-regression models at all distance classes, which would indicate that the fitted models adequately describe the spatial variation in body size across all spatial scales (Diniz-Filho et al. 2003).

Some workers believe that it is necessary to adjust mean body sizes in grids by the number of species (Meiri & Thomas 2007). To accommodate this point of view we used species richness values (S) and [1-log10(1/S2)] as weighting factors in a weighted least squares regression to recalculate the standardized coefficients of the variables included in our best models, as well as the coefficient of determination of each model (R2).

Finally, we also analyzed body size variation using a numerical procedure based on the Phylogenetic eigenVector Regression (PVR) approach of Diniz-Filho et al. (2007), which allows partitioning total body size variation among species into a phylogenetic (P) and a specific component (S) (Diniz-Filho et al. 1998; see also Ramirez et al. 2008). The specific component S expresses the independent evolution in the trait after each species diverged from the others, so it can indicate adaptive processes related to climatic variation not confounded by deep history of lineages. However, there is no detailed phylogeny for the Cerrado species analyzed here, and using a phylogeny at higher hierarchical levels (i.e., family), as done by Ramirez et al. (2008), would be not too informative because there is actually a small number of families and most species would be concentrated in very few families (Hylidae, and Leiuperidae + Leptodactylidae = traditional Leptodactylidae) (see Appendix S1).

96


Table 2 Multiple regression models for anuran mean body size in the Cerrado. Only the best models (i.e. ΔAIC≤2) with their corresponding coefficients of determination (R2) and the standardized regression coefficients of the predictors included in these models are shown. Env + Spa models include non-redundant spatial filters as predictors. Predictor variables are: Humidity=annual relative humidity; WD=water deficit; PET= annual potential evapotranspiration; Range = elevation range. The Enhanced Vegetation Index (EVI) is not included among the predictors because it did not enter in any of the best models.

Thus, instead of partitioning the total

variation into P and S components as in PVR, we coded the taxonomic levels (species within genera and families) as dummy variables. We also used dummy coding for expressing higher-level relationships among families, based on phylogenies by Roelants et al. (2007), Wiens (2007) and Frost et al. (2006). For the first two phylogenies, families were clustered in the same way, using two dummy variables creating the Nobleobatrachia and (Natatanura + Microhylidae) clades. For Frost’s et al. (2006) phylogeny a different arrangement between families must be employed, mainly because these authors do not consider many of the “traditional” families as monophyletic. In this case, the genera were used to define the cluster of new families (see Frost et al. 2006 pg. 40). In general, using the better-known families as a reference only, Bufonidae and Dendrobatidae are closer to some Leptodactylidae, nested within a large clade that includes Ranidae, which is finally nested within a larger clade including Hylidae. In both cases, these taxonomic coding reflects phylogenetic clusters, and thus a nested ANOVA, using body size as response variable, can be used in comparative analyses in a way analogous to PVR (Harvey & Pagel, 1991; see Diniz-Filho et al. 2005 for an application for anurans in Cerrado). In this case, the residuals of nested ANOVA estimate the S-component, and the R2 of the model can be used to indicate the amount of phylogenetic effect in body size, as in PVR. Following Diniz-Filho et al. (2007), the S-component of each

species can then be averaged for each cell and regressed against the environmental predictors.

A recent species-level phylogenetic hypothesis for hylid frogs (Wiens et al. 2006) allowed us to conduct a complementary analysis within the Hylidae family. Considering the clusters of Hylidae genera in Wiens et al. (2006) phylogeny (i.e., Phyllomedusinae, Cophomantini, the clades Dendropsophus, Scinax and Lophiohylini) we used the abovementioned nested ANOVA analysis adding additional dummy variables. However, this increased slightly the proportion of variance from 61.9 to 68% (see results) and S-components were in all cases highly correlated (r > 0.89), which suggest that including into the analyses such phylogenetic clusters within hylids does not affect qualitatively our main conclusions. Moreover, since there are no detailed species-level and generic-level relationships available for all families, we kept the analyses based on dummies for families, genus and higher order affinities as proposed by Roelants et al. (2007), Wiens (2007) and Frost et al. (2006).

All statistical analyses were performed with STATISTICA (StatSoft, Inc. 2003) and SAM 3.0 (Spatial Analysis in Macroecology; Rangel et al. 2006).

RESULTS We found a clear geographical gradient of

anuran mean body size variation in the Cerrado, with the smallest mean sizes in the southwestern areas bordering the Pantanal, a

Model Weighting variable Predictors in model Env Env + Spa Humidity WD PET Range PET x

Range AIC R2 AIC R2

1 None 0.274 0.658 -0.356 -0.098 -1206 0.455 -1343 0.765 2 None 0.269 0.650 -0.317 -0.087 -1206 0.454 -1342 0.764 1 1-Log10(1/S2) 0.458 0.747 -0.312 -0.154 0.446 0.758 2 1-Log10(1/S2) 0.449 0.735 -0.249 -0.133 0.446 0.758 1 S 0.205 0.631 -0.379 -0.105 0.424 0.748 2 S 0.200 0.624 -0.338 -0.096 0.425 0.748

97


wetland biome, and the largest sizes in the dryer areas of the northeast, near the limits of the

Caatinga semi-arid region (Fig. 1).

Table 3 Condition Numbers (CNs) and variance inflation factors (VIFs) of each predictor in our best multiple regression models. Values for environmental models (Env.) and environmental models including non-redundant spatial filters as predictors (Env+spa) are provided.

Table 4 Multiple regression models for the specific component of body size (S) obtained from the comparative phylogenetic analysis (see Methods). Results are presented for the best models in Table 2 and for two different amphibian phylogenies (Frost et al. 2006 and Roelants et al. 2007). Analyses using Wiens (2007) phylogeny provided the same results than Roelants et al. (2007) and are not reported to avoid redundancy. Note that, contrarily to Table 2, AIC values are slightly lower for model 2 (�AIC≤1). Presentation as in Table 2.

Consequently, mean body size was positively correlated with water deficit (Table 1), which described more variance than any other variable in the simple correlations (R2=0.31). WD was also included in the two “best” multifactor models and had the highest standardized regression coefficients (Table 2). Both models had identical AIC values and described 45% of variance in mean body size. Potential Evapotranspiration (PET) and relative humidity ranked second and third according to their regression coefficients (Table 2). The models also included topography, either directly or in interaction with PET, but the regression coefficients were low. Condition numbers and variance inflation factors were lower than 5 in all cases (Table 3), which indicates that collinearity among variables in these models is not a concern and we can use standardized regression coefficient values as estimates of the relative importance of each variable. Including non-redundant spatial filters in the models increased the percentage of described variance to 77% and removed

residual autocorrelation in all cases –as indicated by spatial correlograms (not shown) – but this did not alter the variables or relationships described by the best models (see e.g. Hawkins et al 2007). Likewise, the best models remained qualitatively the same after accounting for anuran species richness with weighted least squares regression (Table 2), indicating that they were not affected by the uneven distribution of species richness throughout the Cerrado. Finally, the amount of phylogenetic signal in body size slightly differed when using the dummy variables derived from the three amphibian phylogenies, being equal to 58.8% for nested ANOVAS based on Roelants et al. (2007) and Wiens (2007) phylogeny, and equal to 61.9% for nested ANOVA based on Frost et al. (2006). As expected, there was a high correlation (r = 0.99) between average S-components calculated for each ANOVA, since most body size variation can be “captured” by phylogenetic structures at genera and family levels and not by clusters of families which vary between the phylogenies

Model Predictors in model Condition Number Humidity WD PET Range PET x Range

1 Env 1.536 3.390 3.029 1.206 3.495 2 Env 1.559 3.390 2.838 1.055 3.409 1 Env + Spa 2.658 4.623 3.626 1.273 4.334 2 Env + Spa 2.663 4.644 3.753 1.437 4.400

Model Phylogeny Predictors in model Env Humidity WD PET Range PET x Range AIC R2 1 Frost et al. (2006) 0.065 0.873 -0.451 -0.033 -1250 0.339 2 Frost et al. (2006) 0.061 0.861 -0.437 -0.043 -1251 0.331 1 Roelants et al. (2007) 0.054 0.867 -0.442 -0.046 -1242 0.343 2 Roelants et al. (2007) 0.049 0.862 -0.422 -0.056 -1243 0.345

98


used. Similar to the results for total body size, water deficit also had the highest regression coefficients in our multiple regression models for the average-S component of body size, which accounted for 34% of variance irrespective of the phylogenetic hypothesis used (Frost et al. 2006, Wiens 2007 or Roelants et al. 2007) (Table 4). We found a strong positive association between the geographical patterns in the specific component (S) of body size and water deficit (see also maps in Appendix S2), which suggests adaptive responses of species to climatic conditions independently of ancestry (Diniz-Filho et al. 2007).

Figure 1 Mean anuran body size patterns in the Cerrado (131 species). Numbers in the legend are average log10-transformed snout-to-vent lengths (original units: millimeters).

DISCUSSION Our data identify a trend of increasing anuran

body size along with decreasing available environmental water, as predicted by the water availability hypothesis. That is, the average body sizes were largest in the dryer northeast and smallest in the wetter southwest of the Cerrado biome. Water deficit, a measure of dryness levels in the environment or of climate drought (Stephenson 1998), was the best predictor of this gradient. The remaining hypotheses examined here did not receive substantial support, and only the secondary role of potential evapotranspiration in our regression models (associated to the heat balance hypothesis) merits attention (see below). Neither the habitat availability nor the primary productivity hypotheses played a significant role in determining anuran body size variation in the Cerrado region and, hence, we will not speculate on possible explanations for the lack of significance. However, it is possible that both hypotheses are important in tropical areas with a pronounced altitudinal gradient such as Andean slopes, a question that remains to be tested in future analyses.

Our assemblage approach, when combined with phylogenetic information, also found that the spatial pattern for the specific component (S) of body size is strongly related to water availability in the environment. Even in the presence of a strong phylogenetic inertia in body size variation (i.e. a high proportion of variance was explained by the phylogenetic structure in the data), we detected a clear signal of unique responses of each species to water deficit in the specific component (S). That is, species larger than expected by their phylogenetic relatedness are more frequent in drier environments, whereas species smaller than expected are mostly distributed in low water deficit areas. This finding suggests that there is a selective advantage for larger anurans under high water deficit conditions in tropical regions.

Most anurans need wet conditions to maintain their water balance, because they lose water from their skin at a rate similar to evaporation from a free water surface (Adolph 1933). Thus, they have little resistance to

99


cutaneous evaporative water loss and are particularly sensitive to long droughts. The seasonal climate of the Cerrado is characterized by a six-month period of dry and hot conditions, which very likely imposes serious hydric constraints on its anuran fauna. Given that evaporative water loss is positively correlated with surface-to-volume ratios in anurans (e.g. Farrell & MacMahon 1969), a simple mechanism to reduce dessication is decreasing the surface-to-volume ratio by increasing body size. The Cerrado has a southwest-to-northeast gradient of increasing dryness that may explain larger anuran sizes in the drier, northeastern areas.

Although the water availability hypothesis is most consistent with our data, only a third of the variance in mean body size was described by water deficit alone, and our best multifactorial models explained only half of the variance (Table 2). This means that half of the spatial pattern of mean size is not associated with climatic gradients, suggesting that anuran tolerance to drought in the Cerrado is also influenced by size-independent traits. It has long been known that anurans have a wide array of adaptive mechanisms to withstand dessication (see e.g. Bentley 1966) and, observations in the neighboring, even dryer (Silva 2004), Caatinga biome indicate that such mechanisms are common within the anuran communities in the dry Neotropics. According to Navas et al. (2004), many of the ~40 anuran species in the Caatinga have developed physiological, behavioral and/or morphological adaptations to survive dry periods, including aestivation, phragmotic behavior and changes in skin morphology to improve water balance and thermal tolerance. Although the Cerrado is not as dry, it is likely that such size-independent mechanisms are common among their anuran species, which may account for the mean size variation not explained by climate.

Analyses in the Holarctic have also found relationships between climate and anuran mean body size gradients. Nonetheless, the Holarctic patterns depend more on the amounts of heat and light available in the environment (as indicated by a stronger negative relationship between body size and potential evapotranspiration) than on water availability

(see Olalla-Tárraga & Rodríguez 2007). Coupled with our present results (we found weaker relationships with PET), this suggests that anuran mean body size-climate relationships vary from temperate to tropical zones. In low energy regions (e.g. Holarctic), environmental energy availability is probably the primary driver of the body size-climate relationship, whereas water availability is the limiting factor in high energy macroclimates (e.g. Neotropics). This is not unexpected under the logic of the heat balance and water availability hypotheses. Both rely on a negative relationship between body mass and the surface-to-volume ratio, but differ in the proposed advantages of being large. That is, the heat balance hypothesis states that being large reduces heat loss, whereas the water availability hypothesis depends on a reduction in the amount of water that is evaporated from the skin. Therefore, a single explanation that incorporates both hypotheses can provide a general framework to understand body size gradients exhibited by this group across broad climatic gradients. This is similar to the water-energy conjecture of Hawkins et al. (2003) for broad-scale patterns of species richness, which proposes that water availability is more critical in warm regions, whereas environmental energy is more critical in cold areas.

Because amphibians use environmental energy to warm their bodies, a very large body size may always improve water conservation, but it may simultaneously lead to insufficient heat gain under very low energy conditions, making it difficult to maintain heat balance (see Olalla-Tárraga et al. 2006). Despite this, the fact that anurans as a group are larger towards the northern Holarctic suggests that, in general, the sizes attained by anuran species living at high latitudes are not constrained by heat gain (Olalla-Tárraga & Rodríguez 2007). Moreover, behavioural and/or physiological adjustments may help the larger, high latitude anuran species to reduce heating times, thus optimizing heat gain and converting their large bodies into an advantage for heat conservation. All in all, these arguments and results reinforce the idea that the heat balance and water availability mechanisms can be integrated into one hypothesis (the energy-water conservation

100


hypothesis): a large body provides greater heat conservation in cold macroclimates and greater water conservation in dry tropical areas.

Exceptions to the above are likely to be found in extremely cold and dry regions, particularly because size-independent behavioral and physiological adaptations appear to be necessary for anurans inhabiting these areas to endure harsh environments (Storey & Storey 1996; Navas et al. 2004). In other words, a relatively large body size might not be enough to maintain optimum heat and water balances in extremely cold and dry habitats. Thus, anurans may be similar to other vertebrates, including mammals, which have size-independent adaptations to cope with cold (e.g. the burrowing capabilities of many rodents). These special adaptations have also been proposed to explain a lack of conformity with Bergmann’s rule in some endothermic groups (e.g. Rodríguez et al. 2006; Medina et al. 2007).

In conclusion, the geographical trend in anuran body size to increase towards high water deficit areas in the Cerrado suggests that the reduced surface to volume ratios associated with larger sizes reduces evaporative water loss. This generates greater resistance to dessication, which may be critical for surviving in dry environments. In a parallel argument for colder regions, where energy availability -instead of water- is the main limiting factor, larger anuran species seem also to take advantage of a low surface to volume ratio to optimize their heat balance (Olalla-Tárraga & Rodríguez 2007). Therefore, in a context of latitudinal variation in the relative importance of energy and water from temperate to tropical regions, both hypotheses could be merged into a single explanation for anuran broad-scale body size gradients. If this proves to be general, it implies that qualitatively and quantitatively different environmental models may describe body size gradients in anurans depending on the extent and location of the study area across a latitudinal gradient. ACKNOWLEDGEMENTS M.Á.O.-T. thanks people at the Theoretical Ecology and Synthesis Lab (Universidade Federal de Goiás, Brazil) for their hospitality

during the preparation of this manuscript. We also thank P. Aragón, L.M. Bini, B.A. Hawkins, J.J. Wiens and three anonymous reviewers for critical reading and suggestions on previous drafts. This study was supported by the Spanish Ministry of Science and Innovation (grant REN2003- 03989/GLO to M.Á.R., and FPU fellowship AP2005-0636 to M.Á.O.-T.). Financial support for compiling the data came from a PRONEX program of CNPq/SECTEC-GO (Proc. No. 23234156) for establishing conservation priorities in the Cerrado area. J. A. F. Diniz-Filho is supported by CNPq productivity fellowships (301259/2005-4, 470918/2006-3). REFERENCES Adams, D.C. and Church, J.O. 2008.

Amphibians do not follow Bergmann's rule. – Evolution 62: 413–420.

Adolph, E.F. 1933. Exchanges of water in the frog. – Biol. Rev. 8: 224–240.

Ashton, K.G. 2002. Do amphibians follow Bergmann’s rule?. – Can. J. Zool. 80: 708–716.

Bentley, P.J. 1966. Adaptations of Amphibia to arid environments. – Science 152: 619 –623.

Bergmann, C. 1847. Ueber die Verhältnisse der Wärmeökonomie der Thiere zu ihrer Grösse. – Gottinger Studien 3: 595–708.

Belsey, D.A. 1991. Conditioning diagnostics, collinearity and weak data in regression. John Wiley & Sons, Inc., New York.

Bini, L.M. et al. 2006. Challenging Wallacean and Linnean shortfalls: knowledge gradients and conservation planning in a biodiversity hotspot. – Divers. Dist. 12: 475–482.

Blackburn, T.M. et al. 1999. Geographic gradients in body size: a clarification of Bergmann’s rule. – Divers. Dist. 5: 165–174.

Blackburn, T.M. and Hawkins, B.A. 2004. Bergmann’s rule and the mammal fauna of northern North America. – Ecography 27: 715–724.

Borcard, D. and Legendre, P. 2002. All-scale spatial analysis of ecological data by means of principal coordinates of neighbor matrices. – Ecol. Model. 153: 51–68.

Brattstrom, B.H. 1963. Preliminary review of the thermal requirements of amphibians. –Ecology 44: 238–255.

101


Burnham, K.P. and Anderson, D.R. 2002. Model selection and multimodel inference. – Springer.

Clark, K. 1992. FNOC Global Elevation, Terrain, and Surface Characteristics. Digital Raster Data on a 10-minute Cartesian Orthonormal Geodetic (lat/long) 1080x2160 grid. In: Global Ecosystems Database Version 2.0. Boulder, CO: NOAA National Geophysical Data Center. Available at http://www.ngdc.noaa.gov/seg/cdroms/ged_iia/datasets/a13/fnoc.htm

Colli, G.R. et al. 2002. The character and dynamics of the Cerrado herpetofauna. – In: Oliveira P.S., Marques, R.J. (eds) The Cerrado of Brazil: ecology and natural history of a neotropical savanna. Columbia Univ. Press, pp 223–241.

Cooper, N. et al. 2008. Macroecology and extinction risk correlates of frogs. – Global Ecol. Biogeogr. 17: 211–221.

Cushman, J.H. et al. 1993. Latitudinal patterns. in European ant assemblages: variation in species richness and body size. –Oecologia 95: 30–37.

Diniz-Filho J. A. F. et AL. 1998. An eigenvector method for estimating phylogenetic inertia. – Evolution 52: 1247–1262.

Diniz-Filho, J.A.F. et al. 2003. Spatial autocorrelation and red herrings in geographical ecology. – Global Ecol. Biogeogr.12: 53–64.

Diniz-Filho, J.A.F. et al. 2004. Anurans from a local assemblage in Central Brazil: linking local processes with macroecological patterns. – Braz. J. Biol. 64: 41–52.

Diniz-Filho, J.A.F. and Bini, L.M. 2005. Modelling geographical patterns in species richness using eigenvector-based spatial filters. – Global Ecol. Biogeogr.14: 177– 185.

Diniz-Filho, J.A.F. et al. 2005. Macroecological correlates and spatial patterns of anurans description dates in Brazilian Cerrado. – Global Ecol. Biogeogr. 14: 469–477.

Diniz-Filho, J.A.F. et al. 2007. Seeing the forest for the trees: partitioning ecological and phylogenetic components of Bergmann’s rule in European Carnivora. –Ecography 30: 598–608.

Dormann, C.F. et al. 2007. Methods to account for spatial autocorrelation in the analysis of species distributional data: a review. –Ecography 30: 609–628.

Duellman, W.E. 1988. Patterns of species diversity in anuran amphibians in the American tropics. –Ann. MO. Bot. Gard. 75: 79–104.

Duro, et al. 2007. Development of a large area biodiversity monitoring system driven by remote sensing. –Prog. Phys. Geog. 31: 235–260.

Eiten, G. 1972. The cerrado vegetation of Brazil. – Bot. Rev. 38: 201–341.

ESRI. 2002. ArcGIS, version 8.2. Environmental Systems Research Institute, Redlands, California.

Eterovick, P.C. et al. 2005. Amphibian declines in Brazil: An overview. – Biotropica 37, 166–179.

Farrell, M.P. and MacMahon, J.A. 1969. An eco-physiological study of water economy in eight species of tree frogs (Hylidae). – Herpetologica 25: 279–2294.

Ferreira, L.G. et al. 2003. Seasonal landscape and spectral vegetation index dynamics in the Brazilian Cerrado: An analysis within the Large-Scale Biosphere-Atmosphere Experiment in Amazonia (LBA). – Remote Sens. Environ. 87: 534–550.

Francis, A.P. and Currie, D.J. 2003. A globally-consistent richness-climate relationship for angiosperms. – Am. Nat. 161: 523–536.

Frost, D.R., et al. 2006. The amphibian tree of life. – Bull. Am. Mus. Nat. Hist. 297, 1–291.

Gaston, K.J. et al. 2008. Ecogeographical rules: elements of a synthesis. – J. Biogeogr. 35: 483–500.

Greve, M. et al. 2008. Environmental factors, regional body size distributions and spatial variation in body size of local avian assemblages. – Global Ecol. Biogeogr. 17: 514–523.

Griffith, D.A. 2003. Spatial autocorrelation and spatial filtering. Gaining understanding through theory and visualization, Springer, Berlin.

Griffith, D.A. and Peres-Neto P.R. 2006. Spatial modeling in ecology: the flexibility of eigenfunction spatial analyses. – Ecology 87: 2603–2613.

102


Harvey, P. & Pagel, M. 1991. The comparative method in evolutionary biology. Oxford University Press, Oxford.

Hawkins, B.A. 2004. Summer vegetation, deglaciation and the anomalous bird diversity gradient in eastern North America. –Global Ecol. Biogeogr. 13: 321–325.

Hawkins, B.A. et al. 2003. Energy, water, and broad-scale geographic patterns of species richness. – Ecology 84: 3105–3117.

Hawkins, B.A. and Diniz-Filho, J.A.F. 2006. Beyond Rapoport’s rule: evaluating range size patterns of New World birds in a two-dimensional framework. – Global Ecol. Biogeogr. 15: 461–469.

Hawkins, B. A. et al. 2007. Red herrings revisited: spatial autocorrelation and parameter estimation in geographical ecology. – Ecography 30: 375–384.

Huete, A. et al. 2002. Overview of the radiometric and biophysical performance of the MODIS vegetation indices. – Remote Sens. Environ. 83:195–213.

Huey, R.B. 1982. Temperature, physiology, and the ecology of reptiles. – In: Gans, C. and Pough, F.H. (eds.), Biology of the Reptilia, Vol. 12. Academic Press, pp. 25–67.

Hurlbert, A.H. & Haskell, J.P. 2003. The effect of energy and seasonality on avian species richness and community composition. – Am. Nat. 161, 83–97.

Hutchinson, V.H. & Dupré, R.K. 1992. Thermoregulation. – In: Feder, M.E. and Burggren, W.W. (eds), Environmental physiology of the amphibians. Univ. Chicago Press, pp. 206–249.

Janzen, D. H. 1967. Why mountain passes are higher in the tropics. –Am. Nat. 101:233–249.

Johnson, J.B. and Omland, K.S. 2004. Model selection in ecology and evolution. –Trends Ecol. Evol. 19: 101–108.

Lazaridis, A. 2007. A Note Regarding the Condition Number: The Case of Spurious and Latent Multicollinearity. – Quality & Quantity 41:123–135.

Legendre, P. 1993. Spatial autocorrelation: trouble or new paradigm. – Ecology 74:1659–1673.

Lindsey, C.C. 1966. Body sizes of poikilotherm vertebrates at different latitudes. –Evolution 20: 456–465.

Medina, A.I. et al. 2007. Subterranean rodents of the genus Ctenomys (Caviomorpha, Ctenomyidae) follow the converse to Bergmann’s rule. – J. Biogeogr. 34: 1439–1454.

Meiri, S. and Thomas, G.H. 2007. The geography of body size - challenges of the interspecific approach. – Global Ecol. Biogeogr. 16: 689–693.

Myers, N. et al. 2000. Biodiversity hotspots for conservation priorities. – Nature 403: 853–858.

Navas, C.A. 2003. Herpetological diversity along Andean elevational gradients: links with physiological ecology and evolutionary physiology. – Comp. Biochem. Physiol. 133: 469–485.

Navas, C.A. et al. 2004. A preliminary assessment of anuran physiological and morphological adaptation to the Caatinga, a Brazilian semi-arid environment. – Int. Congr. Ser. 1275: 298–305.

Neter, J.et al. 1996. Applied linear statistical models. Irwin.

Nevo, E. 1973. Adaptive Variation in Size of Cricket Frogs. – Ecology 54: 1271–1281.

New, M.G. et al. 1999. Representing Twentieth Century space-time climate variability. Part I: Development of a 1961-1990 mean monthly terrestrial climatology. –J. Clim.12: 829–856. Available at http://www.sage.wisc.edu/atlas/maps.php?catnum=3&type=Ecosystems

Nimer, E. 1989. Climatologia do Brasil, edn. IBGE, Departamento de Recursos Naturais e Estudos Ambientais, Rio de Janeiro.

Olalla-Tárraga, M.Á. et al. 2006. Broad-scale patterns of body size in squamate reptiles of Europe and North America. – J. Biogeogr. 33: 781–793.

Olalla-Tárraga, M.Á. and Rodríguez, M.Á. 2007. Energy and interspecific body size patterns of amphibian faunas in Europe and North America: anurans follow Bergmann's rule, urodeles its converse. – Global Ecol. Biogeogr. 16: 606–617.

103


Pough, F.H. 1980. The advantages of ectothermy for tetrapods. –Am. Nat. 115: 92–112.

Rangel, T.F.L.V.B. et al. 2006. Towards an integrated computational tool for spatial analysis in macroecology and biogeography. – Global Ecol. Biogeogr.15: 321–327.

Ramírez, L. et al. 2008. Partitioning phylogenetic and adaptive components of the geographical body-size pattern of New World birds. – Global Ecol. Biogeogr.17: 100–110.

Ratana, P. et al. 2005. Analysis of Cerrado physiognomies and conversion in the MODIS seasonal-temporal domain. – Earth Interactions 9:1–22.

Rodríguez, M.Á. et al. 2005. Energy, water and large-scale patterns of reptile and amphibian species richness in Europe. – Acta Oecol. 28: 65–70.

Rodríguez, M.Á., et al. 2006. The geographic distribution of mammal body size in Europe. – Global Ecol. Biogeogr.15: 185–193.

Rodríguez, M.Á. et al. 2008. Bergmann’s rule and the geography of mammal body size in the Western Hemisphere. – Global Ecol. Biogeogr. 17: 274-283.

Roelants K., et al. 2007. Global patterns of diversification in the history of modern amphibians. – Proc. Nat. Acad. Sci. USA 104, 887-892.

Ron, S.R. 2000. Biogeographic area relationships of lowland neotropical rainforest based on raw distributions of vertebrate groups. – Biol. J. Linn. Soc. 71: 379–402.

Rosenzweig, M.L. 1968. The strategy of body size in mammalian carnivores. – Am. Midl. Nat. 80: 299–315.

Silva,V.P.R. 2004. On climate variability in Northeast of Brazil. J. Arid Environ. 58: 575–596.

Silva, J.M.C. and Bates, J.M. 2002. Biogeographic patterns and conservation in

the South American Cerrado: a tropical savanna hotspot. – BioScience 52: 225–233.

Silva, J.F. et al. 2006. Spatial heterogeneity, land use and conservation in the Cerrado region of Brazil. – J. Biogeogr. 33: 536–548.

StatSoft Inc. (2003) STATISTICA (data analysis software system), version 6 (http://www.statsoft.com).

Stephens, P.A., et al. 2007. Inference in ecology and evolution. – Trends Ecol. Evol. 22: 192–197.

Stephenson, N.L. 1998. Actual evapotranspiration and deficit: biologically meaningful correlates of vegetation distribution across spatial scales. – J. Biogeogr.25: 855–870.

Storey, K.B. and Storey, J.M. 1996. Natural freezing survival in animals. – Annu. Rev. Ecol. Evol. Syst. 27: 365–386.

Wiens, J. J. et al. 2006. Evolutionary and ecological causes of the latitudinal diversity gradient in hylid frogs: treefrog trees unearth the roots of high tropical diversity. – Am. Nat. 168:579–596.

Wiens, J. J. 2007. Global patterns of species richness and diversification in amphibians. – Am. Nat. 170: S86–S106.

Willmott, C. J., and Kenji, M. 2001. Terrestrial water budget data archive: monthly time series (1950–1999). Available at http://www.sage.wisc.edu/atlas/maps.php?catnum=3&type=Ecosystems

SUPPLEMENTARY MATERIAL The following supplementary material is available for this article: Appendix S1. Species list and body size data. Appendix S2. Geographic patterns of the average specific component (S) of body size for the 131 anuran species in the Cerrado using two different phylogenetic hypotheses: (A) Frost et al. (2006); (B) Roelants et al. (2007).

104

Appendix S1 – Species list and body size data used for analysis (snout-to-vent length in milimeters) SVL are average values for each species, including males and females, and were compiled from a large list of primary references (species’ original descriptions or taxonomic reviews) and, for a few species, values were obtained by broadly measuring adult individuals in the collection of the Departamento de Biologia Geral, ICB, UFG, or in the herpethological collection of Museu Nacional de Rio de Janeiro (MNRJ/UFRJ). A detailed list of primary references used is available from the authors upon request. SVL= snout-to-vent length.

Species Family Description

Date SVL

Allobates goianus Aromobatidae 1975 17 Barycholos ternetzi Brachycephalidae 1937 24 Eleutherodactylus heterodactylus Brachycephalidae 1937 12 Ischnocnema juipoca Brachycephalidae 1978 26 Pristimantis crepitans Brachycephalidae 1965 16 Pristimantis dundeei Brachycephalidae 1999 20 Pristimantis fenestratus Brachycephalidae 1864 27 Ollotis nebulifer Bufonidae 1824 50 Rhaebo guttatus Bufonidae 1799 99 Rhinella crucifer Bufonidae 1821 27 Rhinella margaritifera Bufonidae 1768 50 Rhinella ocellata Bufonidae 1858 43 Rhinella ornata Bufonidae 1821 72 Rhinella rubescens Bufonidae 1925 108 Rhinella schneideri Bufonidae 1894 150 Odontophrynus americanus Cycloramphidae 1841 49 Odontophrynus cultripes Cycloramphidae 1862 43 Odontophrynus moratoi Cycloramphidae 1980 39 Odontophrynus salvatori Cycloramphidae 1996 32 Proceratophrys cururu Cycloramphidae 1998 16 Proceratophrys goyana Cycloramphidae 1937 42 Thoropa megatympanum Cycloramphidae 1984 40 Ameerega braccata Dendrobatidae 1864 20 Ameerega flavopicta Dendrobatidae 1925 26 Ameerega picta Dendrobatidae 1838 23 Aplastodiscus perviridis Hylidae 1950 68 Bokermannohyla alvarengai Hylidae 1956 42 Bokermannohyla circumdata Hylidae 1871 57 Bokermannohyla ibitiguara Hylidae 1973 42 Bokermannohyla nanuzae Hylidae 1973 42 Bokermannohyla pseudopseudis Hylidae 1937 20 Bokermannohyla ravida Hylidae 2001 66 Bokermannohyla saxicola Hylidae 1964 18 Bokermannohyla sazimai Hylidae 1982 32 Corythomanthis greeningi Hylidae 1896 45

105

Dendropsophus anataliasiasi Hylidae 1972 23 Dendropsophus araguaya Hylidae 1998 20 Dendropsophus cerradensis Hylidae 1998 19 Dendropsophus cruzi Hylidae 1998 11 Dendropsophus elianeae Hylidae 2000 23 Dendropsophus jimi Hylidae 1999 22 Dendropsophus melanargyreus Hylidae 1887 20 Dendropsophus minutus Hylidae 1872 20 Dendropsophus nanus Hylidae 1889 40 Dendropsophus rhea Hylidae 1999 48 Dendropsophus rubicundulus Hylidae 1862 55 Dendropsophus soaresi Hylidae 1983 36 Dendropsophus tritaeniatus Hylidae 1965 31 Hypsiboas albopunctaus Hylidae 1824 55 Hypsiboas buriti Hylidae 1999 40 Hypsiboas cipoensis Hylidae 1968 31 Hypsiboas crepitans Hylidae 1824 63 Hypsiboas ericae Hylidae 2000 27 Hypsiboas faber Hylidae 1821 93 Hypsiboas goianus Hylidae 1968 33 Hypsiboas lundii Hylidae 1973 52 Hypsiboas multifasciatus Hylidae 1859 23 Hypsiboas pardalis Hylidae 1824 73 Hypsiboas phaeopleura Hylidae 2000 42 Hypsiboas pulchellus Hylidae 1841 52 Hypsiboas punctatus Hylidae 1799 41 Hypsiboas raniceps Hylidae 1862 32 Hypsiboas stenocephaus Hylidae 1999 30 Phasmahyla jandaia Hylidae 1978 28 Phyllomedusa azurea Hylidae 1800 42 Phyllomedusa burmeisteri Hylidae 1882 61 Phyllomedusa centralis Hylidae 1965 110 Phyllomedusa oreades Hylidae 2002 32 Pseudis bolbodactyla Hylidae 1925 36 Pseudis caraya Hylidae 1964 63 Pseudis limellum Hylidae 1862 25 Pseudis paradoxa Hylidae 1758 55 Pseudis tocantins Hylidae 1998 35 Scinax acuminatus Hylidae 1862 27 Scinax canastrensis Hylidae 1982 37 Scinax centralis Hylidae 1996 32 Scinax duartei Hylidae 1951 43 Scinax fuscomarginatus Hylidae 1925 14 Scinax fuscovarius Hylidae 1925 81

106

Scinax luizotavioi Hylidae 1989 86 Scinax machadoi Hylidae 1973 24 Scinax maracaya Hylidae 1980 23 Scinax nebulosus Hylidae 1824 33 Scinax pinima Hylidae 1973 27 Scinax squalirostris Hylidae 1925 28 Trachycephalus nigromaculatus Hylidae 1838 50 Trachycephalus venulosus Hylidae 1768 19 Crossodactylus bokermanni Hylodidae 1985 26 Crossodactylus trachystomus Hylodidae 1862 92 Hylodes otavioi Hylodidae 1983 18 Eupemphix nattereri Leiuperidae 1863 45 Physalaemus centralis Leiuperidae 1962 26 Physalaemus cuvieri Leiuperidae 1826 37 Physalaemus deimaticus Leiuperidae 1988 28 Physalaemus evangelistai Leiuperidae 1967 24 Pseudopaludicola boliviana Leiuperidae 1927 17 Pseudopaludicola falcipes Leiuperidae 1867 30 Pseudopaludicola mineira Leiuperidae 1994 12 Pseudopaludicola mystacalis Leiuperidae 1887 42 Pseudopaludicola saltica Leiuperidae 1887 33 Leptodactylus bokermanni Leptodactylidae 1973 25 Leptodactylus camaquara Leptodactylidae 1978 34 Leptodactylus chaquensis Leptodactylidae 1950 39 Leptodactylus cunicularius Leptodactylidae 1978 47 Leptodactylus furnarius Leptodactylidae 1978 38 Leptodactylus fuscus Leptodactylidae 1799 39 Leptodactylus hylaedactylus Leptodactylidae 1868 24 Leptodactylus jolyi Leptodactylidae 1978 40 Leptodactylus labyrinthicus Leptodactylidae 1824 130 Leptodactylus martinezi Leptodactylidae 1956 24 Leptodactylus mystaceus Leptodactylidae 1824 46 Leptodactylus mystacinus Leptodactylidae 1861 48 Leptodactylus ocellatus Leptodactylidae 1758 103 Leptodactylus pentadactylus Leptodactylidae 1768 106 Leptodactylus petersii Leptodactylidae 1864 37 Leptodactylus podicipinus Leptodactylidae 1862 38 Leptodactylus pustulatus Leptodactylidae 1870 36 Leptodactylus syphax Leptodactylidae 1969 78 Leptodactylus tapiti Leptodactylidae 1978 33 Leptodactylus troglodytes Leptodactylidae 1926 50 Leptodactylus wagneri Leptodactylidae 1862 32 Physalaemus albonotatus Leptodactylidae 1864 30 Pleurodrema fuscomaculatum Leptodactylidae 1864 21

107

Chiasmocleis albopunctata Microhylidae 1884 32 Chiasmocleis centralis Microhylidae 1952 24 Chiasmocleis mehelyi Microhylidae 1997 19 Dermatonotus muelleri Microhylidae 1885 66 Elachistocleis bicolor Microhylidae 1838 30 Elachistocleis ovalis Microhylidae 1799 25 Lithobates palmipes Ranidae 1824 18

108

Appendix S2. Geographic patterns of the average specific component (S) of body size for the 131 anuran species in the Brazilian Cerrado using two different phylogenetic hypotheses: (A) Frost et al. (2006); (B) Roelants et al. (2007). The data were calculated using a comparative method analogous to Phylogenetic eigenVector Regression (PVR) (Diniz-Filho et al. 1998) (see Methods for details). Analyses using Wiens (2007) phylogeny provided the same pattern than Roelants et al. (2007) and are not reported to avoid redundancy.

109

Capítulo 6 Regla de Bergmann en anfibios, regresión filogenética basada en autovectores y la aproximación de ensamblaje Este capítulo reproduce íntegramente el texto del siguiente manuscrito: Olalla-Tárraga, M. Á., L. M. Bini, J. A. F. Diniz-Filho & M. Á. Rodríguez. Bergmann’s rule in amphibians, phylogenetic eigenvector regression and the assemblage approach: a comment on Adams and Church (2008). Evolution (In review).

Bergmann’s rule in amphibians, phylogenetic eigenvector regression and

the assemblage approach: a comment on Adams and Church (2008)

Miguel Á. Olalla-Tárraga1,*, Luis M. Bini2, José Alexandre F. Diniz-Filho2, Miguel Á. Rodríguez1

1Department of Ecology, University of Alcalá, Alcalá de Henares, 28871, SPAIN

2Departamento de Biologia Geral, ICB, Universidade Federal de Goiás, CP 131, 74.001-

970, Goiânia, GO, Brazil

ABSTRACT

There is a lively debate on the applicability of Bergmann’s rule to ectotherms. Both at intra and interspecific levels, a number of recent studies have found taxonomically and geographically conflicting patterns, which has raised questions about the generality of ecological and evolutionary explanations. In a recent issue of Evolution, Adams & Church (2008) evaluated Bergmann’s rule in Plethodon salamanders of eastern North America at different taxonomic scales. They also explored the intraspecific version of the rule for amphibians in general. They concluded that amphibians do not follow this ecogeographical “rule” and discarded the classical heat conservation hypothesis as a valid explanation in this taxon. Here, we argue it is premature to dismiss the existence of geographic body size gradients in amphibians, much less the search for causal factors. Using an assemblage-based interspecific approach complemented with phylogenetic eigenvector regression, we found a clear reversed Bergmann’s cline for Plethodon strongly associated with thermal availability in the environment. Our results are consistent with the ‘heat balance hypothesis’ as a plausible explanation for the observed pattern.

Keywords: Body size gradients, interspecific, intraspecific, Plethodon, Urodeles.

110

Plethodon salamanders and Bergmann’s rule

MAIN TEXT

Bergmann’s rule, a trend of increasing body size towards colder regions, has long attracted the attention of ecologists, biogeographers and evolutionary biologists (Bergmann 1847, Meiri and Dayan 2003, Ashton and Feldman 2003). Over the past few years, there has been a resurgence of interest in the investigation of spatial patterns in body size over large-scales to explore the validity of this ecogeographical “rule” at different taxonomic levels (Gaston et al. 2008). However, even though there is now abundant evidence for the pattern in endotherms at both the intra- (Ashton et al. 2000, Ashton 2002, Meiri and Dayan 2003) and interspecific levels (Blackburn and Hawkins 2004, Rodríguez et al. 2008, Ramírez et al, 2008), the classical heat conservation mechanism is still debated as a general explanation (Blackburn et al., 1999). Most of the controversy stems from the seminal papers of Ray (1960) and Lindsey (1966) suggesting that some ectothermic organisms also display intra and interspecific body size variation as a response to environmental gradients, which apparently requires alternative explanations to the ones offered for endotherms (Cushman et al., 1993). Since then, workers have tried to identify ecological or evolutionary mechanisms accounting for body size clines in ectotherms, but we are far from a consensus on a unifying mechanism (Atkinson and Sibly 1997, Angilletta et al. 2003, Walters and Hassall 2006, Olalla-Tárraga et al. 2006, Olalla-Tárraga and Rodriguez 2007).

A critical step before searching for underlying mechanisms is examining what the intra- and interspecific patterns look like in nature (Belk and Houston 2002). This is what Adams and Church (2008) have done for Plethodon salamander species in eastern North America. They used a large body size database (including measurements of almost 97,000 specimens) to conduct an intraspecific meta-analysis of Bergmann’s rule. They also compiled published data on the intraspecific relationship between body size and temperature for 21 additional amphibian species (including

several salamanders and anurans and a single caecilian species) and draw on the same meta-analytical approach to examine whether amphibians in general follow Bergmann’s rule. Finally, they explored the interspecific version of the rule by using a cross-species analysis complemented with phylogenetically independent contrasts (Felsenstein 1985). They concluded that amphibians do not generally follow Bergmann’s rule and discarded the classical heat conservation hypothesis as a valid explanation in this taxon (although they also emphasized that this mechanism may still be important for endotherms).

While we agree that the heat conservation hypothesis is a possible mechanistic explanation for Bergmannian patterns in mammals and birds, we disagree with Adams and Church’s conclusion that Bergmann’s clines are not generally present in amphibians. Here, we provide new complementary analyses and interpretations of the data to claim that it is premature to dismiss the applicability of either Bergmann’s rule or its converse to amphibians. We also contend that sufficient grounds exist for arguing that a thermoregulatory explanation that takes into account both heat conservation and heat gain should still be considered a plausible mechanism for the existence of both Bergmann’s rule and its converse in different amphibian clades (Olalla-Tárraga and Rodríguez 2007).

Cross-species vs. assemblage-based approaches

Bergmann’s rule has been explored using three approaches: intraspecific, interspecific and assemblage-based (sensu Gaston et al. 2008). Although the distinction between intra- and interspecific approaches seems to be clear, there is some confusion on the methodological differences between interspecific (hereafter “cross-species”) and assemblage-based studies as both have been referred to as “interspecific” analyses (see Gaston et al. 2008 and references therein). However, the differences in methodology

111


between both kinds of “interspecific” methods are not trivial. Cross-species approaches treat each species as an independent datum (i.e., the units of analysis are the species) and use bivariate scatter-plots to examine covariation of body size and latitude (or temperature) across species. This methodology has also been termed as the “midpoint approach”: it involves obtaining a single spatial measure for each species (usually the latitudinal midpoint of its geographic range) and then plotting these mid points against the species' body sizes (see Blackburn and Hawkins, 2004 and references therein). In contrast, assemblage-based studies explore geographical patterns within grids covering the study region (i.e., the units of analysis are grid-cells), and combine the species' presences/absences in the cells with their body sizes to obtain cell-mean body size values (usually log-transformed geometric means) (see Ruggiero and Hawkins 2006 and Gaston et al. 2008). Blackburn and Hawkins (2004) termed this the “community approach”, as such investigations examine the spatial distribution of summary statistics for body size across faunal assemblages (grid cells) of a particular biogeographic region.

The pros and cons of cross-species and assemblage-based approaches have been extensively discussed elsewhere (Blackburn and Hawkins, 2004; Ruggiero and Hawkins, 2006; Meiri and Thomas 2007) so we will only briefly comment on this. Notably, both methods need to circumvent potential problems associated to different sources of pseudoreplication. In assemblage-based studies, the varying proximity among units of analysis (i.e., grid cells) makes them to have different levels of spatial interdependence (i.e., the data are spatially autocorrelated and, hence, pseudoreplicated to some extent), which causes usual estimations of degrees of freedom to be inflated (i.e., each cell contributes less than one degree of freedom) and, thus, may lead to increased type I errors. Spatial statistical techniques allow to tackle this (e.g. see Diniz-Filho et al. 2003). In contrast, in cross-species analyses, there is in principle no need to evaluate sources of spatial autocorrelation, as they adopt an ‘individual species focus’ and each species contributes only

once to the overall pattern. Instead, inflated type I errors associated to pseudoreplication may arise from the phylogenetic non-independence of the data, which justifies the use of phylogenetic comparative methods (such as Felsenstein’s 1985 independent contrasts used by Adams and Church 2008).

According to several authors (Blackburn and Hawkins 2004, Ruggiero and Hawkins 2006), the main advantage of assemblage approaches is that they allow a direct evaluation of the environmental structure underlying most broad-scale geographical patterns, a feature that is severely limited in the case of cross-species analysis as environmental gradients are reduced to a single point in geographical or environmental space. That is, cross-species methods ignore the geographical structure that exists in the data by reducing the multidimensional nature of geographic ranges to single values. This can have serious implications for the interpretation of ecological and evolutionary patterns and processes such as those associated to Bergmann’s rule, even leading to incorrect or equivocal conclusions (Blackburn and Hawkins 2004, Ruggiero and Hawkins 2006). Because ecological and evolutionary processes usually take place in a geographical context, spatially explicit approaches are necessary to gain a multidimensional perception of species’ trait gradients (such as body size gradients) and to make explicit their links to environmental variation (Ruggiero and Hawkins 2006).

Adams and Church (2008) focused their investigation on the intra- and interspecific levels using meta-analytical techniques and a cross-species analysis respectively, but they did not consider geographic patterns in assemblage structure. Here, we exemplify how an assemblage-based approach complemented with phylogenetic eigenvector regression (PVR) (Diniz-Filho et al. 1998, Diniz Filho et al. 2007) can be useful to generate further insights into potential evolutionary and ecological mechanisms determining variation in body size of Plethodon salamanders in eastern North America.

112


Phylogenetic components of body size and the assemblage approach in Plethodon

We followed the methods described in Diniz-Filho et al (2007). Initially, we created a 110 x 110 km equal area grid of 320 cells and used geographic range maps (IUCN, Conservation International and NatureServe 2006) for each of the 44 Plethodon species described in eastern North America together with the body size data kindly provided by Adams and Church (2008) to calculate average body length (hereafter mean body size) in each grid cell (see Olalla-Tárraga et al. 2006, Olalla-Tárraga and Rodríguez 2007 for details). Analyses based on another measures of central tendency (median or mode) will generate similar estimates to the mean considering the log-normal distribution of body sizes. Because species of Plethodon have been extensively used in behavioral, ecological and evolutionary studies, their ranges (extents of occurrence) are well known (see Highton 1995 and references therein). We then used phylogenetic eigenvector regression (PVR) (Diniz-Filho et al. 1998) to partition the specific and phylogenetic components of the interspecific variation in body size for Plethodon assemblages across eastern North America (Diniz Filho et al. 2007).

PVR estimates the degree of phylogenetic pattern in body size data by regressing this trait against a set of orthogonal eigenvectors extracted from a pairwise phylogenetic distance matrix that describes phylogenetic relatedness among species (Diniz-Filho et al. 1998; see also Griffith and Peres-Neto 2006 for an analogous application in spatial analyses). Hence, we can partition the total body size of each species (T) into phylogenetic components (P), which represent predicted body size values according to the clade’s phylogenetic structure, and specific components (S), which are model residuals that can be interpreted as independent responses of each species (see Cheverud et al. 1985; Gittleman and Kot 1990).

We based our PVR analysis on a species-level phylogenetic tree for Plethodon

(Wiens et al. 2006) and applied the criteria described in Diniz-Filho et al. (2007) to select five eigenvectors from the initial set. These eigenvectors described 95.3% of the variation in phylogenetic structure and were subsequently used as predictors of body size. A multiple regression model of body size against the selected eigenvectors explained 63.2% of the variance in size, indicating that there is a strong phylogenetic pattern in the data. Using the predicted values (P components) and residuals (S components) of this regression, we generated mean-P and mean-S values for each cell (Fig 1a and 1b respectively). Finally, we built regression models to evaluate the relationships between the spatial patterns observed for mean-P and mean-S components and environmental variation and to evaluate the extent to which the geographical patterns in average body size are a result of phylogenetic effects at the assemblage level or are average adaptive responses of species to environmental variation.

We processed three environmental variables in ArcGIS 9.2 (ESRI 2006) to incorporate them as predictors into the analyses: mean annual temperature, annual precipitation, and the Global Vegetation Index (see Olalla-Tárraga et al. 2006 and Olalla-Tárraga and Rodríguez 2007, for further details on data description and sources). These variables were selected on the basis that they are related to three hypothesis proposed for explaining body size gradients in amphibians, heat balance, water availability and primary productivity, respectively (see Olalla-Tárraga and Rodríguez 2007). Because most Plethodon species occur in forested areas of the Appalachian and Ouachita mountains (Highton 1995), we also added to our models range in elevation within each cell as a measure of mesoscale climatic variation (Turner and Hawkins, 2004). Statistical analyses were performed with SAM (Spatial Analysis in Macroecology; Rangel et al. 2006), STATISTICA (StatSoft, Inc. 2003) and PDAP (Phenotypic Diversity Analysis Programs; Garland et al. 1993).

We found that our three-variable models including mean annual temperature,

113


annual precipitation and the Global Vegetation Index explained 57% and 61.3% of the variation in mean-P and mean-S, respectively (Table 1). Mean annual temperature was the primary predictor of the spatial pattern of mean-P and S components, being positively correlated with both. Because some workers believe that the uneven distribution of species richness across geographical space makes it necessary to adjust mean body sizes in grids by

the number of species (Meiri & Thomas 2007) we repeated these analyses using weighted-Least Square regression. Specifically, we used species richness values (S) and (1-Log10[1/S2]) as weighting factors to recalculate standardized regression coefficients (e.g., Olalla-Tárraga and Rodríguez 2007). Our results remained robust and mean annual temperature consistently had the highest standardized coefficient in the models for mean-P and mean-S (Table 1).

Table1 Multiple regression models of environmental variables against average cell values of phylogenetic (P) and specific (S) components obtained from PVR analysis for 44 Plethodon species in eastern North America. For each model we show their corresponding coefficients of determination (R2) and standardized partial regression coefficients of the predictors. Additionally, we provide results from weighted-LS regression analyses using species richness in each cell (S) and (1-Log10[1/S2]) as weighting factors.Env + Spa models include non-redundant spatial filters as predictors. Predictor variables are: Temp=mean annual temperature; Precip=Annual Precipitation; GVI=Global Vegetation Index. An asterisk (*) means that the standardized regression coefficients were significant at a probability level p < 0.001.

On the other hand, to address the statistical problems associated with spatial autocorrelation in a grid-based approach (i.e., remaining autocorrelation in model residuals, see Diniz-Filho et al 2003), we included eigenvector-based spatial filters as predictors in the regression models (the PCNM model – see Borcard & Legendre 2002). These spatial filters have shown to be an effective tool to capture the spatial structure of the data at different spatial scales (Diniz-Filho & Bini 2005, Griffith & Peres-Neto 2006). We used the maximization of the coefficients of determination in the regression between mean body size and the filters and the minimization

in residual spatial autocorrelation as criteria to select non-redundant filters for the models and avoid “overcorrecting” for spatial autocorrelation (Diniz-Filho & Bini 2005, Griffith 2003). Additionally, we built spatial correlograms based on Moran’s I coefficients (not shown) to evaluate if this approach removed the spatial autocorrelation of the residuals of our multiple-regression models at all distance classes, which would indicate that the fitted models adequately describe the spatial variation in body size across all spatial scales (Diniz-Filho et al. 2003). The inclusion of spatial filters to remove spatial autocorrelation in the residuals of our regression models

Model Weighting variable

Standardized coefficients

(P component)

Standardized coefficients

(S Component)

Temp Precip GVI R2 Temp Precip GVI R2

Env None 0.875* -0.236* -0.006 0.570 0.733* -0.178* 0.234* 0.613

Env + Spa None 0.847* -0.064 -0.136* 0.731 0.530* 0.004 0.085 0.712

Env 1-Log10(1/S2) 0.791* -0.370* 0.067 0.486 0.655* -0.230* 0.291* 0.521

Env S 0.700* -0.410* 0.139* 0.391 0.581* -0.240* 0.340* 0.452

114


(Borcard and Legendre 2002, Diniz-Filho and Bini 2005) did not change the relative importance of the environmental predictors as indicated by the standardized partial regression coefficients (Table 1). Range in elevation

contributed little to the increase of the explanatory power (adjusted R2) of the models (e.g., between 0.7 and 1.5%), hence we did not included this variable in our models.

Thus, in contrast to Adams and Church’s (2008) cross-species analysis, our assemblage approach identified a strong gradient of body size variation in Plethodon assemblages in geographical space, with decreasing size northwards. Thus, Plethodon salamanders in eastern North America follow the converse to Bergmann’s rule. Although this supports Adams and Church’s (2008) contention that this group does not follow Bergmann’s rule, it contradicts their conclusion that there is no spatial pattern in body size at any level of biological organization. Moreover, our results reveal a strong positive association between the geographical patterns in the phylogenetic (P) and ecological (S) components of body size for species assemblages and mean annual temperature. This also runs counter to Adam and Church’s (2008) analysis that was unable to detect a signal of mean annual temperature in body size variation, but it is similar to an assemblage-based study for the complete urodele fauna of Europe and eastern North America (Olalla-Tárraga and Rodríguez 2007).

Figure 1. Geographical patterns of the average phylogenetic component of body size (A) and the specific component (B) resulting from PVR for all the 44 Plethodon salamander species in eastern North America. Numbers included in the legend of each map are snout-to-vent length values in millimeters.

The heat balance hypothesis

Olalla-Tárraga and Rodríguez (2007) proposed the heat balance hypothesis to account for body size gradients exhibited by both endotherms and ectotherms at an assemblage level. For ectothermic groups with thermoregulating abilities, whose body size is not large enough to limit heat gain in low energy environments (e.g. snakes), this hypothesis parallels the traditional heat conservation mechanism originally conceived by Bergmann for endotherms. In contrast, for thermoconformers, which have limited abilities to control heat exchange and depend more closely on the thermal conditions of the environment (Ruben 1995), the hypothesis predicts a reversed Bergmann pattern as animals would minimize body size in colder regions to reduce heating times and maximize the time available for foraging and reproduction (Olalla-Tárraga and Rodríguez 2007). Consistent with these predictions, Olalla-Tárraga and Rodríguez (2007) found that anurans (thermoregulators) follow Bergmann’s rule in Europe and North America, whereas urodeles (thermoconformers) do the opposite. In the latter case, the converse Bergmann’s

115


gradients were even clearer than expected, because the elongate body plans of urodeles generate a larger surface to volume ratio, which also contributes to increased heating rates in low energy environments.

Our results for Plethodon salamanders are also consistent with the heat balance hypothesis. The pattern of the specific component (S) is strongly related to thermal availability in the environment, which suggests independent adaptive responses of species to current climatic conditions (Diniz-Filho et al. 2007). That is, species smaller than expected by their phylogenetic relatedness are more frequent in cold environments, whereas species larger than expected are mostly distributed in warmer areas.

However, the individual responses of species to temperature appear not to be the only cause for the overall gradient in body size, as we also found a climatic signal in the phylogenetic component (P). This is likely to be related to the particularities of recent diversification in this woodland-adapted salamander group since the early Pliocene. Several authors have suggested that most extant species of Plethodon salamanders in eastern North America originated as a result of rapid speciation events during the last five Mya following a history of shifts in forest cover along the Appalachian mountains associated to climatic changes (Highton 1995, Wiens et al. 2006). According to this, continuous altitudinal shifts in forest cover through the Pliocene and Pleistocene may have favoured allopatric diversification events that generated many morphologically-similar cryptic species that, at the same time, tended to conserve their ancestral climatic niches (Highton 1995, Wiens et al. 2006, Kozak and Wiens 2006). Because rapid evolutionary radiation usually involves marked changes in the diversity of morphological and ecological characters (Schluter 2000), Plethodon appears to be a particular case of rapid speciation (Wiens et al. 2006).

Under this hypothesis, the observed relationship between the spatial pattern in the

phylogenetic component of body size and mean annual temperature may simply reflect the interplay between rapid diversification and a structured trend across the phylogeny to maintain ancestral temperature-body size relationships. Even if true, this pattern in the phylogenetic structure of the data is not strong enough to obscure the signal of unique responses of each species to current climate variation found in the specific component.

Bearing in mind the above, why do we find a pattern that Adams and Church (2008) did not detect? Statistically, Adams and Church’s (2008) cross-species analyses are properly done. However, we favor the view that cross-species analyses may be conceptually inappropriate to examine Bergmann’s rule, which should be analyzed in a spatially explicit context (Blackburn and Hawkins 2004, Ruggiero and Hawkins 2006, see above). Of course, this is not to say that such approach is useless to detect correlated evolution between traits across species. For instance, cross-species analyses conducted into a phylogenetic framework have proven to be useful to detect lineage-level patterns in the relationships between a number of life history attributes and body size (see e.g. Freckleton et al. 2002; Dobson 2007 and references therein). The problem arises when we use as a trait a variable that does not actually represent the whole geographic range of a species (either the latitudinal midpoint or an overall measure of temperature as Adams and Church did). In doing so, we are restricting a multidimensional trait (i.e., a species distribution) to a single point. Therefore, a more in depth approach to study Bergmann’s rule requires employing alternative protocols that simultaneously consider the distribution of species into a geographic and phylogenetic framework, such as the assemblage approach we have described here.

Intraspecific patterns

Adams and Church (2008) found support for neither the intraspecific version of Bergmann’s rule nor its converse in this group, despite the statistical power of meta-analysis

116


(Arnqvist and Wooster, 1995). First, we have concerns about the fixed-effect model of meta-analysis used by Adams and Church (2008). Using the data provided in their Table 1, we tested the hypothesis of a single underlying correlation, as assumed by fixed effect analyses. Using the test of homogeneity described in Hedges and Olkin (1985), we found that this assumption was not met (Q = 137.6, d.f. =37, P <<0.01). Therefore, it can be misleading to pool correlation estimates to draw general conclusions about Bergmann’s rule for Plethodon in eastern North America.

Second, Adams and Church (2008) performed a meta-analysis combining their dataset for Plethodon species with data for 21 additional amphibian species to evaluate the extent to which amphibians in general follow Bergmann’s rule intraspecifically. They found no support for the rule or its converse either. However, Adam and Church’s (2008) conclusions cannot be generalized to all amphibians, as their analysis was based on a small proportion of the ~6000 amphibian species described to date (IUCN, Conservation International and NatureServe 2006). Apart from the extremely low number of species in which intraspecific trends of body size variation have been studied, Gaston et al. (2008) also noted that the available data are often geographically and taxonomically biased. Consequently, when we lack data for the vast majority of species in a group, we should be very cautious about reaching conclusions on the absence of patterns.

Last but not the least, the dissimilarities between the observed patterns at different levels of the biological hierarchy for Plethodon (i.e. absence of intraspecific pattern and existence of an interspecific gradient at the assemblage level) is an interesting question that deserves consideration. Perhaps such discordance may be seen as a limitation to support a common mechanism triggering patterns at all levels of biological organization. Gaston et al. (2008) suggested that intraspecific patterns may be more difficult to detect for species showing narrow distribution ranges or limited dispersal abilities. Nonetheless, providing any

explanation for the conflicting patterns found at the intra and interspecific levels for Plethodon is, at the very least, highly speculative and beyond the scope of our analyses. Further studies need to address this question.

Concluding remarks

In sum, we have shown how an assemblage approach complemented with phylogenetic eigenvector regression can be useful to explore both the ecological and evolutionary mechanisms associated to Bergmann’s rule in a spatially explicit context. Contrary to a cross-species approach, our assemblage-based analysis detected strong geographical trends in interspecific variation of body size of Plethodon salamanders in eastern North America. These arguments lead us to call for caution in making generalizations on the validity of Bergmann’s rule in particular taxa, unless we take into account the limitations of our database, the clade we are using as model system and the statistical techniques we employ to analyze the data. This is especially true if we are also interested in examining potential explanations. Although our analyses favor the heat balance hypothesis as a mechanism, considering the idiosyncratic patterns detected for different taxa, the disparities between phylogenetic and non-phylogenetic studies and the apparent discrepancies at different taxonomic scales, a general explanation for Bergmann’s rule still remains elusive.

ACKNOWLEDGEMENTS

We are grateful to D.C. Adams for kindly providing us his body size database. Comments by D.C. Adams, B. A. Hawkins, J. Hortal and S. Meiri greatly improved a previous draft. M.Á.O.-T. thanks people at the NERC Centre for Population Biology (Imperial College London) for their hospitality during writing of this manuscript. J. C. Nabout helped us to create the phylogenetic distance matrix. Our research is funded by the Spanish Ministry of Education and Science (grant CGL2006-03000/BOS to M.Á.R., and FPU fellowship AP2005-0636 to M.Á.O.-T.) and CNPq productivity grants to L.M.B. and J.A.F.D.-F.

117


REFERENCES

Adams, D.C., and J.O. Church. 2008. Amphibians do not follow Bergmann's rule. Evolution 62: 413–420.

Angilletta, M. J., and A. E. Dunham. 2003. The temperature-size rule in ectotherms: simple evolutionary explanations may not be general. Am. Nat. 162: 332–342.

Arnqvist, G., and D. Wooster. 1995. Meta-analysis: synthesizing research findings in ecology and evolution. Trends Ecol. Evol. 10: 236–240.

Ashton, K.G. 2002. Patterns of within-species body size variation of birds: strong evidence for Bergmann’s rule. Global Ecol. Biogeogr. 11: 505–523.

Ashton, K.G., and C.R. Feldman. 2003. Bergmann’s rule in nonavian reptiles: turtles follow it, lizards and snakes reverse it. Evolution 57: 1151–1163.

Ashton, K.G., M.C. Tracy, and A. de Queiroz. 2000. Is Bergmann’s rule valid for mammals? Am. Nat. 156: 390–415.

Atkinson, D., and R. M. Sibly. 1997. Why are organisms usually bigger in colder environments? Making sense of a life history puzzle. Trends Ecol. Evol. 12: 235–239.

Belk, M.C., and Houston, D.D. 2002. Bergmann’s rule in ectotherms: a test using freshwater fishes. Am. Nat. 160: 803–808.

Bergmann, C. 1847. Ueber die Verhältnisse der Wärmeökonomie der Thiere zu ihrer Grösse. –Gottinger Studien 3: 595–708.

Blackburn, T.M., Gaston, K.J., and N. Loder. 1999. Geographic gradients in body size: a clarification of Bergmann’s rule. Divers. Dist. 5: 165–174.

Blackburn, T.M., and B.A. Hawkins. 2004. Bergmann’s rule and the mammal fauna of northern North America. Ecography 27: 715–724.

Borcard, D., and P. Legendre. 2002. All-scale spatial analysis of ecological data by means of principal coordinates of neighbor matrices. Ecol. Model. 153: 51–68.

Cheverud, J.M., M. M. Dow and W. Leutenegger. 1985. The quantitative

assessment of phylogenetic constraints in comparative analyses: sexual dimorphism in body weight among primates. Evolution 39: 1335-1351.

Cushman, J.H., J.H. Lawton, and B.F.J. Manly. 1993. Latitudinal patterns in European ant assemblages: variation in species richness and body size. Oecologia 95: 30–37.

Diniz-Filho, J. A. F., C. E. Ramos de Sant'Ana, and L. M. Bini. 1998. An eigenvector method for estimating phylogenetic inertia. Evolution 52:1247–1262.

Diniz-Filho, J.A.F., L.M. Bini, and B.A. Hawkins. 2003. Spatial autocorrelation and red herrings in geographical ecology. Global Ecol. Biogeogr.12: 53–64.

Diniz-Filho, J.A.F., and L.M. Bini. 2005. Modelling geographical patterns in species richness using. eigenvector-based spatial filters. Global Ecol. Biogeogr.14: 177– 185.

Diniz-Filho, J.A.F., L.M. Bini, M.Á. Rodríguez, T.L.V.B. Rangel, and B.A. Hawkins. 2007. Seeing the forest for the trees: partitioning ecological and phylogenetic components of Bergmann’s rule in European Carnivora. Ecography 30: 598–608.

Dobson, F.S. 2007. A lifestyle view of life-history evolution. Proc Nat Acad Sci USA 104, 17565–17566.

ESRI (2006) ArcGIS, version 9.2. Environmental Systems Research Institute, Redlands, California.

Felsenstein, J. 1985. Phylogenies and the comparative method. Am Nat 125: 1–15.

Freckleton, R.P., P. H. Harvey and M. Pagel. 2002. Phylogenetic analysis and comparative data: a test and review of evidence. Am. Nat.160: 712-726.

Garland, T., Jr., A. W. Dickerman, C. M. Janis, and J. A. Jones. 1993. Phylogenetic analysis of covariance by computer simulation. Syst. Biol. 42: 265-292.

Gaston, K.J., S. Chown, and K.L. Evans. 2008. Ecogeographical rules: elements of a synthesis. J. Biogeogr. 35: 483–500.

118


Gittleman, J.L. and M. Kot. 1990. Adaptation: statistics and a null model for estimating phylogenetic effects. Syst. Zool. 39: 227-241.

Griffith, D.A. 2003. Spatial autocorrelation and spatial filtering. Gaining understanding through theory and visualization, Springer, Berlin.

Griffith, D.A. and Peres-Neto P.R. 2006. Spatial modeling in ecology: the flexibility of eigenfunction spatial analyses. Ecology 87: 2603–2613.

Hedges, L. V., and I. Olkin. 1985. Statistical methods for meta-analysis. Academic Press, Orlando, FL

Highton, R. 1995. Speciation in eastern North American salamanders of the genus Plethodon. Annu. Rev. Ecol. Syst. 26: 579–600.

IUCN, Conservation International, and NatureServe. 2006. Global Amphibian Assessment. Available at www.globalamphibians.org, version 1.1. Downloaded on 15 October 2006.

Kozak, K. H., and J. J. Wiens. 2006. Does niche conservatism drive speciation? A case study in North American salamanders. Evolution 60: 2604–2621.

Lindsey, C.C. 1966. Body sizes of poikilotherm vertebrates at different latitudes. Evolution 20: 456–465.

Meiri, S., and T. Dayan. 2003. On the validity of Bergmann’s rule. J. Biogeogr. 30: 331–351.

Meiri, S. and Thomas, G.H. 2007. The geography of body size - challenges of the interspecific approach. Global Ecol. Biogeogr. 16: 689–693.

Neter, J., M.H. Kutner, C.J. Nachtsheim, and W. Wasserman 1996. Applied linear statistical models. Irwin, Chicago.

Olalla-Tárraga, M.Á., M.Á. Rodríguez and B.A. Hawkins. 2006. Broad-scale patterns of body size in squamate reptiles of Europe and North America. J. Biogeogr. 33: 781–793.

Olalla-Tárraga, M.Á., and Rodríguez, M.Á. 2007. Energy and interspecific body size patterns of amphibian faunas in Europe and North America: anurans follow

Bergmann's rule, urodeles its converse. Global Ecol. Biogeogr. 16: 606–617.

Ramírez, L., Diniz-Filho, J.A.F., and B.A. Hawkins. 2008. Partitioning phylogenetic and adaptive components of the geographical body-size pattern of New World birds. Global Ecol. Biogeogr.17: 100–110.

Rangel, T.F.L.V.B., J.A.F. Diniz-Filho, L.M. Bini. 2006. Towards an integrated computational tool for spatial analysis in macroecology and biogeography. Global Ecol. Biogeogr.15: 321–327.

Rodríguez, M.Á., M.Á. Olalla-Tárraga, and B.A. Hawkins. 2008. Bergmann’s rule and the geography of mammal body size in the Western Hemisphere. Global Ecol. Biogeogr. 17: 274-283.

Ruben, J. 1995. The evolution of endothermy in mammals and birds: from physiology to fossils.Annu. Rev. Physiol. 57: 69–95.

Ruggiero, A., and B.A. Hawkins. 2006. Mapping macroecology. Global Ecol. Biogeogr. 15: 433–437.

Turner, J.R.G., and B.A. Hawkins. 2004. The global diversity gradient. Frontiers of biogeography: new directions in the geography of Nature (ed. by M.V. Lomolino and L.R. Heany), pp. 171–190. Sinauer, Sunderland, MA.

Schluter, D., 2000. The ecology of adaptive radiation. Oxford University Press, Oxford, UK.

StatSoft Inc. 2003. STATISTICA (data analysis software system), version 6

Walters, R.J., and M. Hassall. 2006. The temperature-size rule in ectotherms: may a general explanation exist after all? Am. Nat. 167: 510–523.

Wiens, J. J. 2004. What is speciation and how should we study it? Am. Nat. 163:914–923.

Wiens, J. J., T. N. Engstrom, and P. T. Chippindale. 2006. Rapid diversification, incomplete isolation, and the "speciation clock" in North American salamanders (genus Plethodon): testing the hybrid swarm hypothesis of rapid radiation. Evolution 60: 2585–2603.

119

Capítulo 7 Conclusiones Generales

Conclusiones generales

1. En la fauna de mamíferos de la región Neártica septentrional existe una relación negativa entre la temperatura ambiental y la variación en el tamaño corporal medio. Esta asociación desaparece según descendemos latitudinalmente hacia el Neártico meridional y el Neotrópico, donde las especies son menores en las montañas andinas. Nuestros análisis confirmaron la existencia de un umbral de temperatura en torno a 10°C, por debajo del cual encontramos patrones consistentes con la Regla de Bergmann que no son detectables en macroclimas más cálidos.

2. El mecanismo de conservación de calor propuesto por Bergmann es una explicación plausible para la variación de tamaño en mamíferos que habitan macroclimas fríos, mientras que en regiones tropicales los gradientes de tamaño parecen estar condicionados por la disponibilidad de hábitats. En el Neártico meridional y el Neotrópico, la interacción entre topografía y macroclima origina una fuerte zonación climática a nivel de mesoescala, que genera tamaños de hábitat más reducidos en áreas montañosas y limita la presencia de especies de mayor tamaño en montañas tropicales.

3. Los ensamblajes de especies de reptiles escamados del Neártico y, especialmente, del Paleártico occidental muestran gradientes geográficos en la variación del tamaño corporal asociados a la disponibilidad de energía ambiental. No obstante, los lagartos exhiben una variación de tamaños de acuerdo con la regla de Bergmann, mientras que las serpientes muestran el patrón inverso. En consecuencia, el clásico mecanismo de Bergmann es insuficiente para explicar la distribución geográfica de tamaños corporales en organismos cuyo balance térmico no sólo depende de la conservación sino también de la ganancia de calor.

4. La lenta tasa de calentamiento vinculada a un mayor tamaño corporal sería un factor limitante para la presencia de serpientes de mayor porte en climas fríos, puesto que ven restringido el tiempo disponible para actividades como la alimentación o la reproducción, al precisar periodos de termorregulación más largos. Nuestros hallazgos sugieren, por tanto, que la existencia de patrones de Bergmann en ectotermos termoreguladores depende del tamaño relativo de los organismos.

5. Nuestros resultados para las faunas de anfibios del Neártico y el Paleártico occidental apoyan, igualmente, la hipótesis del balance térmico. En ambas regiones encontramos una fuerte asociación entre la disminución de energía disponible en el ambiente y el incremento del tamaño corporal medio de anuros (regla de Bergmann). Esto sugiere que las habilidades termorreguladoras de este grupo permiten optimizar con éxito el balance entre tasas de calentamiento y enfriamiento en climas fríos. Los urodelos poseen, por el contrario, un alto grado de termoconformismo y siguen el patrón opuesto, al aumentar en tamaño hacia ambientes con mayor disponibilidad energética.

6. En el Cerrado, sabana tropical de Brasil, detectamos un aumento en el tamaño medio de las especies de anuros hacia áreas con un mayor déficit hídrico. Este patrón fue robusto a la presencia de una fuerte señal filogenética en la variación de tamaños de las especies analizadas. Es decir, las especies de mayor tamaño aparecen con mayor frecuencia de lo esperado en base a su parentesco filogenético en ambientes más secos, mientras que las especies más pequeñas de lo esperado seleccionan ambientes más húmedos. Este hallazgo sugiere que existe una ventaja selectiva para los anuros de mayor tamaño corporal en condiciones de alto estrés hídrico en regiones tropicales.

120

Conclusiones generales

7. En zonas tropicales, donde la escasez temporal de agua en vez de la disponibilidad energética

juega un papel limitante, un incremento del tamaño corporal permitiría reducir las tasas de desecación, puesto que la pérdida de agua evaporativa se correlaciona de forma positiva con la ratio superficie-volumen en anuros. Considerando la variación latitudinal en la importancia relativa de la disponibilidad de energía y agua desde regiones templadas hacia los trópicos, proponemos que los gradientes de tamaño corporal en anuros reflejan los efectos de la reducción de la ratio superficie-volumen en el control del balance corporal térmico e hídrico.

8. Las especies de salamandras pertenecientes al género Plethodon que habitan en el Neártico oriental siguen un patrón inverso a la regla de Bergmann que está fuertemente vinculado a la variación térmica ambiental. El gradiente observado no sólo resultaría de las respuestas adaptativas individuales de las especies a la temperatura, sino de los rápidos eventos de especiación alopátrica en este grupo desde el Plioceno temprano.

121

on the biogeography of vertebrate body size: ecological

Documents