William E. Duellman Editor Museum of Natural History The University of Kansas Monof^aph No. 7 HARVARD UNIVERSITY Library of the Museum of Comparative Zoology THE SOUTH AMERICAN HERPETOFAUNA: ITS ORIGIN, EVOLUTION, AND DISPERSAL THE SOUTH AMERICAN HERPETOFAUNA: ITS ORIGIN, EVOLUTION, AND DISPERSAL WILLIAM E. DUELLMAN EDITOR Museum of Natural History and Department of Systematics and Ecology The University of Kansas Lawrence, Kansas 66045, USA MONOGRAPH OF THE MUSEUM OF NATURAL HISTORY, THE UNIVERSITY OF KANSAS NUMRER 7 1979 MONOGRAPH OF THE MUSEUM OF NATURAL HISTORY, THE UNIVERSITY OF KANSAS Number 7, pages 1^85, 172 figures in text Issued December 28, 1979 © 1979 by The Museum of Natural History, The University of Kansas, Lawrence, Kansas. All rights reserved. No part of this book may be reproduced in any form or by any means without permission in writing from the publisher. ISRN Number: 0-89338-008-3 MUS. COMP. ZOOL LIBRARY Cover design by Linda Trueb JUN 51985 HARVARD UNIVERSITY PRINTED BY UNIVERSITY OF KANSAS PRINTING SERVICE LAWRENCE, KANSAS, USA Dedicated to the memories of three herpetologists who contributed so much to our knowledge of the South American herpetofauna: Roberto Donoso-Barros (1922-1975) Bertha Lutz (1894-1976) James A. Peters (1922-1972) PREFACE This volume is the result of a symposium of the same title held on 11-13 August 1977 in conjunction with the joint annual meetings of the Herpetologists' League and the Society for the Study of Amphibians and Reptiles at Lawrence, Kansas. I originally conceived the idea for such a symposium in August 1975 while returning from a 15-month sojourn in South America. My interactions with many South American biologists during that trip had convinced me that the time was appropri- ate for a thorough discussion of ideas and presentation of our existing knowledge of the South American heq^etofauna. The initial response from colleagues was heartening, so during the following year the symposium was organized. Unfortunately, owing to various circumstances not all subjects were covered; obvious omissions in this volume are chapters on the South American-North American her- petofaunal relationships and the herpetofau- nas of the Brasilian Highlands, the Atacama Desert, and the caatinga and campos cerrados of Brasil. This volume is organized in much the same way as was the symposium, except that my introductory chapter provides an overview of the South American herpetofauna. Chapter 2 deals with the fossil record of amphibians and reptiles in South America, and Chapters 3 and 4 are concerned with the relationships of the South American herpetofauna with those of Africa and Australia. The Quaternary bio- geography of the continent is the subject of Chapters 5-7. Treatments of regional herpeto- faunas are found in Chapters 8-15, and the Participants in the Symposium on the South American Herpetofauna held in Lawrence, Kansas, 11-13 August 1977. Front row (left to right): Alberto Veloso M., Beryl B. Simpson, Jaime E. Pefaur, Ana Maria Baez, Jose M. Cei. Second row: Lars Brundin, Thomas E. Lovejoy, Donn E. Rosen, Jiirgen Haffer, Thomas H. Fritts, Ramon Formas, Raymond F. Laurent. Back row: William E. Duellman, James R. Dixon, Marinus S. Hoogmoed, John D. Lynch, W. Ronald Heyer, Michael J. Tyler, Jose M. Gallardo. final chapter is devoted to the conservation of the herpetofauna. I am grateful to the contributors to this volume for their scholarly efforts and for their patience and understanding while it was being produced. For their participation in the symposium, I thank the contributors and Lars Brundin, Thomas H. Fritts, W. Ronald Heyer, Jaime E. Pefaur, Donn E. Rosen, and Alberto Veloso M. Their enthusiastic partici- pation contributed a high level of scholarly interaction, as well as much good cheer. During the editing of this volume I called upon many colleagues to review manuscripts. The quality of the papers included herein benefited from reviews by Avelino Barrio, Lars Brundin, Richard Estes, Thomas H. Fritts, Steven Gorzula, W. Ronald Heyer, Philip S. Humphrey, Jean Lescure, Alan E. Leviton, John D. Lynch, Larry D. Martin, Braulio Orejas-Miranda, Jaime E. Pefaur, Alan H. Savitzky, Beryl B. Simpson, Linda Trueb, T. van der Hammen, Alberto Veloso M. and Richard G. Zweifel. The drawings for many of the papers were executed by Debra K. Bennett, Staff Illustrator of the Museum of Natural History at The University of Kansas. Jaime E. Pefaur translated many of the sum- maries and edited the Spanish of others. Lin- da Trueb 's competent editorial review of the manuscripts is evident in their consistency and style. Rose Etta Kurtz retyped many pages of manuscript, and Rebecca A. Pyles painstakingly worked on the index. To all of these persons I owe a debt of gratitude for their endeavors in behalf of this volume. Throughout the early phases of develop- ment and organization of the symposium, as well as during the production of this volume, Philip S. Humphrey, Director of the Museum of Natural History, has provided advice, en- couragement and support. Ronald K. Cal- gaard. Vice Chancellor for Academic Affairs, and George R. Waggoner, Associate Vice Chancellor for International Programs, The University of Kansas, gave enthusiastic sup- port for the symposium. Richard F. Treece of the Bureau of Conferences and Institutes coordinated the logistics of the meetings. Without their interest and aid the symposium and this volume would not have been pos- sible. Financial support for bringing together the participants in the symposium was gen- erously provided by the National Science Foundation (DEB 76-16767), the World Wildlife Fund (WWF-US-71) and the Office of Academic Affairs, The University of Kansas. Support for the preparation of the index was provided by a grant from the General Re- search Fund of The University of Kansas. William E. Duelhnan Lawrence, Kansas September 6, 1979 CONTENTS 1. The South American Heqjetof auna : A Panoramic View. William E. Dnellman 2. The South American Herpetofauna: An Evaluation of the Fossil Record. Ana Maria Bdez and Znlma B. de Gasparini 29 3. Herpetofaunal Relationships Between Africa and South America. Baymond F. Laurent 55 4. Herpetofaunal Relations of South America with Australia. Michael J. Tyler 73 5. Quaternary Biogeography of Tropical Lowland South America. Jiirgen Haffer 107 6. Late Cenozoic Environmental Changes in Temperate Argentina. Ana Maria Bdez and Gustavo Juan Scillato Yane 141 7. Quaternary Biogeography of the High Montane Regions of South America. Beryl B. Simpson ... 157 8. The Amphibians of the Lowland Tropical Forests. John D. Lynch 189 9. Origin and Distribution of Reptiles in Lowland Tropical Rainforests of South America. James R. Dixon 217 10. The Herpetofauna of the Guianan Region. Marinus S. Hoogmoed 241 11. Origin and Distribution of the Herpetofauna of the Dry Lowland Regions of Northern South America. Carlos Rivero-Blanco and James R. Dixon 281 12. Composition, Distribution y Origen de la Herpetofauna Chaquefia. Jose M. Gallardo _ 299 13. The Patagonian Herpetofauna. Jose M. Cei 309 14. La HeqDetofauna de los Bosques Temperados de Sudamerica. /. Ramon Formas 341 15. The HeqDetofauna of the Andes: Patterns of Distribution, Origin, Differentiation and Present Communities. William E. Duellman 371 16. Refugia, Refuges and Minimum Critical Size: Problems in the Conservation of the Neotropical Herpetofauna. Thomas E. Lovejoy 461 Subject Index 465 Taxonomic Index 470 1. The South American Herpetofauna: A Panoramic View William E. Duellman Museum of Natural History and Department of Systematics and Ecology The University of Kansas Lawrence, Kansas 66045 USA A vast array of dinosaurs still inhabited the earth, ratite birds watched curiously as furry mammals experimented with new ways of reproduction, and varieties of anurans and squamates set out on diverse evolutionary courses, while turtles continued their conser- vative approach to a changing world. They witnessed the breakup of the earth as Gond- wanaland was split by the magma, giving birth to a new ocean, and somewhat later the fracture of the land again to create a large island continent — South America — destined to drift in a northwestward arc for nearly fifty million years before establishing a narrow connection with a neighbor of long ago and far away — North America. During that long period of isolation of South America, some of the archaic groups of plants and animals became extinct; some dwindled in numbers leaving only a few scat- tered relicts, and others prospered and gave rise to new and diverse kinds in the face of changing environments, for the island was not static. Tectonic events and climatic changes shaped the landscapes in an ever-changing scene. Great areas of the land were innun- dated by epeiric seas, the southern end of the continent cooled and desiccated as world zonation of climates was established; the rise of a gigantic mountain chain interrupted the winds and modified the climates and gave birth to thousands of small streams that coa- lesced in their descents to the lowlands and formed huge rivers, and finally, cooling brought glaciation and fluctuations in climate and sea level that brought about changes in the drainages and the biota. Late in this scenario man entered South America and began preying upon the animals, clearing land, cultivating plants, and building temples. So it was when Columbus "discovered" South America on his third voyage in 1498, when Ferdinand Magellan arrived in 1520, and at the beginning of the conquest of the New World by Francisco Pizarro in 1532. For nearly two centuries the decimation of the native human populace (by the sword, the Bible, and disease) was second in importance only to the frenetic search for El Dorado — the real and fabled materialistic riches of South America. With the exception of pitifully few men, no attention was given to the natural riches of the continent. But in the 18th Cen- tury, European naturalists began exploring South America, as related so eloquently by von Hagen (1948:xiii). For it was the explorer-naturalists who opened South America. It was these knowl- edge-thirsting men who, because they were deemed harmless, were permitted entry be- hind the Green Curtain when others were not. It was the naturalists who methodically and systematically pushed aside the frontiers of South America and dug it from its oblivion. With an enthusiasm that bridged every bar- rier, they climbed the Andes, they swept down the dark mysterious rivers, they trekked across the deserts and struggled through the Laocoon entanglements of its fire-fly spangled jungles. They dispelled legends, they uncovered facts, they rediscovered rubber, studied quinine and the coca leaf. They measured the earth's sur- face, they crawled into the jungle and col- lected plants, they studied the animals, they measured the tides. ... It was the naturalists who opened South America. Early collections reaching Europe formed substantial natural history cabinets, and some of these were illustrated and described; the most ambitious undertaking was that by Al- bertus Seba, who in the 1730's published his classic "Thesaurus" in four volumes. The illus- MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 (rations in Seba's work and specimens that reached Upsala formed the basis for names of South American species by Linnaeus in 1758. In the 19th Century, some of the world's most famous naturalists worked in South America — Charles Marie de La Condamine, Alexander von Humboldt, Alfred R. Wallace, Henry W. Rates, Richard Spruce, and of course Charles Darwin. These men made extensive natural history collections, but these included few, if any, amphibians and reptiles. Plants, insects, and birds were the chief goals of most of the collectors. Five European naturalists made important contributions to the early knowledge of the South American herpetofauna in the early 1800's through their collections and their writings — Maximillian A. P. zu Wied-Neu- wied, Alcide D. D'Orbigny, Johann R. von Spix, Marcos X. Jimenez de la Espada, and Johann J. Tschudi. Some of the collections made by those men, plus many small collec- tions that reached European museums pro- vided the basis for countless papers on South American amphibians and reptiles by Albert Giinther, Wilhelm Peters, Oskar Boettger, Franz Steindachner, and the most prolific of European herpetologists — George Boulenger. During the latter part of the 19th Century, South American specimens reached the United States; most of these were reported on by Edward D. Cope. Ry the beginning of the present century several centers of biological research had been established in South America. Early pioneers in herpetological research included Julio Koslowsky in Argentina, R. A. Philippi in Chile, and Alipio de Miranda-Ribeiro and Adolfo Lutz in Rrasil. By the mid-20th Cen- tury investigations on the South American herpetofauna flourished. But as research on amphibians and reptiles broadens to include studies on the ecology, life history, and be- havior, the need still remains for descriptive morphology and systematics. Ever increasing human disturbance of natural environments, especially the rainforests, eliminates forever many components of the biota before they be- come known to science. COMPOSITION OF THE HERPETOFAUNA The complex history and diverse topog- raphy and climate of South America have produced an extraordinarily rich and diverse herpetofauna. Currently more than 2,200 species are recognized in more than 300 gen- era in 37 families (Tables 1:1-1:2). These numbers are bound to increase with future discoveries. The rate of discovery of new species in South America is astonishing. As examples, of the 313 species of hylid frogs now known from South America, 100 have been named in the last two decades (1960- present); Peters and Donoso-Barros (1970) listed 71 species of Anolis, and 14 additional species have been named. Likewise, many new species of frogs, especially Centrolenella, Colostethus, and Eleatherodactylus, and of salamanders (Bolitoglossa) are being discov- ered and named yearly. A far higher percentage of the living am- phibians of the world than of the reptiles inhabits South America. In this respect am- phibians are more like birds, whereas reptiles are more like mammals (Table 1:3). Review of the Families In this brief review, each family is dis- cussed with respect to its origin (Table 1:4), temporal and geographic distribution in South America (Table 1:5, Fig. 1:1), and differen- tiation and dispersal in South America. Ma- rine reptiles are not included. Plethodontidae. — Known from Pliocene and Pleistocene deposits in North America, the family is highly differentiated there (23 genera, about 200 species). Two genera that are most speciose in Central America ( Bolito- glossa and Oedipina) also occur in South America. There, Oedipina (2 species) occurs only in the Choco, whereas Bolitoglossa (24 species) inhabits the Choco, Amazonia, and the northern Andes. Plethodontid salaman- ders entered South America from Central America after the closure of the Panamanian Portal (Wake, 1966). 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA Table 1:1. — Taxonomic Composition of the South American Herpetofauna. ( ° = endemic to South America) Table 1:3. — Comparison of Numbers of Species of Tetrapod Vertebrates in South America with World Fauna. Genera Species 37 4 2 1 2 15 1 13 2 4 6 Family Total Endemic Amphibia Plethodontidae 2 Pipidae 1 Leptodactylidae 41 Bufonidae 7 Brachycephalidae" .. 2 Rhinodermatidae° .. 1 Dendrobatidae — 3 Pseudidae" 2 Hylidae 22 Centrolenidae 2 Ranidae 1 Microhylidae 16 Rhinatrematidae° .... 2 Typhlonectidae" .... 4 Caeciliidae 9 Reptilia Pelomedusidae 1 Chelidae 7 Kinostemidae 1 Chelydridae 1 Emydidae 2 Testudinidae 1 Gekkonidae 16 Iguanidae 27 Teiidae 38 Scincidae 1 Anguidae 2 Amphisbaenidae 6 Anomalepidae 4 Leptotyphlopidae _ 1 Typhlopidae 1 Boidae 5 Aniliidae 1 Tropidophiidae 2 Colubridae 77 Micruridae 2 Viperidae 3 Crocodylidae 4 Total Endemic 7 20 28 1 5 1 2 1 39 1 24 21 5 4 411 396 95 88 2 2 2 2 75 69 4 4 313 285 57 52 1 31 28 10 10 18 18 47 44 7 15 3 1 5 5 64 240 151 8 8 45 17 34 3 10 1 4 409 32 46 7 7 15 2 1 2 4 56 220 140 7 7 44 16 32 3 6 1 3 357 27 39 5 Class Total South American Percentage South American Amphibians Reptiles 3,307 5,954 8,656' .. 4,060' 1,095 1,115 2,780b 810d 33 19 Birds 32 Mammals 20" 1 Brodkorb (1972). '" Meyer de Schauensee ( 1964). c Anderson and Jones ( 1967). d Hershkovitz (1972). " The figure 810 includes Central America and the West Indies, so the total number and percentage for South America will be lower. Leiopelmatidae. — With living representa- tives only in North America (Ascaphus) and New Zealand ( Leiopelma ) , this family is rep- resented in South America only by the fossils of Vieraella and Notobatrachus from the Ju- rassic of Patagonia (Estes and Reig, 1973). Pipidae. — At present restricted to sub- Saharan Africa and tropical America, pipids have an extensive fossil record in Gondwana- land — Early Cretaceous of Israel, Upper Cre- taceous-Miocene of southern Africa, and Late Cretaceous-Eocene of South America (Estes, 1975). Early Cenozoic fossils from South America are representatives of the living Afri- can genus Xenopus. One South American genus, Pipa, contains five species in the tropi- cal lowlands (three genera recognized by some authors). Pipa parva enters eastern Panama. Leptodactylidae. — This is the most diverse and speciose anuran family in South America, where it is known back as far as the late Paleo- Table 1:2. — Summary of the Taxonomic Composition of the South American Herpetofauna. Ordinal group Families Genera Species Total Endemic Total Endemic Total Endemic 1 2 24 21 11 3 98 75 996 930 3 2 15 12 75 72 15 5 115 87 1,095 1,023 6 13 7 36 31 5 84 56 471 430 1 6 5 45 44 9 96 44 556 484 1 4 2 7 5 22 203 114 1,115 994 37 5 318 201 2,210 2,017 Salamanders Anurans Caecilians ... total amphibians Turtles Lizards Amphisbaenians Snakes Crocodilians total reptiles total MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 cene (Baez and Gasparini, this volume). Al- though the family is unquestionably of Gond- wanan origin, the relationships of the lepto- dactylids are not clear. Lynch (1971) recog- nized the South American and Australian frogs plus the South African Heleophryne as one family, the Leptodactylidae, but he (1973) separated the Old World genera into the Myobatrachidae. This arrangement was fol- lowed generally by Savage (1973), Duellman (1975), Heyer (1975), and Heyer and Liem ( 1976), but not by Tyler (this volume). Hey- er ( 1975 ) suggested that leptodactylids might have evolved from leiopelmatids; this idea was elaborated upon by Lynch (1978). Within South America, the primitive Tel- matobiinae are primarily distributed in tem- perate regions — the tribe Telmatobiini in Pat- agonia, austral forests, and the high Andes. More advanced telmatobiines are in temper- ate and tropical regions — Odontophrynini in the Chaco, southeastern Brasil, and nonfor- ested regions in eastern Brasil, Grypiscini on the Brasilian Shield, Eleutherodactylini most diverse in northwestern South America but also occurring on the Brasilian and Guianan shields and in Amazonia, and also speciose in Middle America and the West Indies. The diversity of eleutherodactyline genera and the differentiation of Eleiitherodactyhis in Middle America are indicative of immigration of eleu- therodactylines into Central America prior to the establishment of the isthmian link in the late Pliocene (Savage, 1973; Lynch, 1976). The Ceratophryinae are widespread in Cha- coan, Amazonian, and Guianan lowlands. The Elosiinae are restricted to the Brasilian Shield. The Leptodactylinae are widespread in tropical and subtropical lowlands, with a primitive genus (Pleurodema) also inhabiting Patagonia, austral forests, and the Andes (Duellman and Veloso, 1977). Physalaemus, Pleurodema, and Leptodactylus have entered Central America, and the latter also is in the West Indies. Bufonidae. — The earliest fossil bufonids are from the Paleocene of Brasil (Estes and Reig, 1973), followed by the Oligocene Neo- procoela, which is a member of the Eurasian Bufo calamita group, according to Tihen ( 1962) and Baez and Gasparini (this volume) but referred to the telmatobiine leptodactylids by Lynch ( 1971 ) . By the Miocene, Bufo was present in South America, North America, Europe, and Africa (Tihen, 1972). The ab- sence of bufonids from the Australo-Papuan Region (except for the introduced Bufo ma- rinus), combined with the fossil history of the group, strongly suggests a western Gondwana- land origin of the family ( Blair, 1972; Savage, Table 1:4. — Postulated Geographic Origins of Families of Amphibians and Reptiles Inhabiting South America. (NA = North America; SA = South America; f= Extinct in South America) Pangaea Leiopelmatidae 1 Boidae Laurasia Plethodontidae (NA) Kinosternidae (NA) Chelydridae (NA) Emydidae Trionychidaet Anguidae Aniliidae Viperidae Gondwanaland Uncertain Pipidae Leptodactylidae Bufonidae Brachycephalidae (SA) Rhinodermandae ( SA ) Pseudidae (SA) Hylidae Centrolenidae (SA) Ranidae Microhylidae Rhinatrematidae (SA) Typhlonectidae (SA) Caeciliidae Pelomedusidae Chelidae Meiolaniidaef Iguanidae (SA) Teiidae (SA) Anomalepidae ( SA ) Tropidophiidae (SA) Micruridae (SA) Testudinidae Gekkonidae Scincidae Amphisbaenidae Typhlopidae Leptotyphlopidae Colubridae Crocodylidae 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA Table 1:5. — Distribution of Herpetofaunal Family Groups in Major Eco-physiographic Regions in South America. Family group J3 o g •c o ■c O c < 13 c/j 3 O o 2 c 03 s O 3 c u Amphibia Plethodonndae Pipidae Ceratophryinae Telmatobiinae Elosiinae Leptodactylinae . Bufonidae B rachycephalidae Rhinodermatidae Dendrobatidae Pseudidae Phyllomedusinae Hemiphractinae Amphignathodontinae Hyhnae Centrolenidae Ranidae Microhylidae Rhinatrematidae Typhlonectidae Caeciliidae Reptilia Pelomedusidae - Chelidae Kinosternidae Chelydridae Emydidae Testudinidae Gekkoninae Sphaerodactylinae I guanines Basiliscines Anolines Tropidurines Teiidae Scincidae Anguidae Amphisbaenidae Anomalepidae .._ Leptotyphlopidae Typhlopidae Boidae Aniliidae Tropidophiidae Xenodontinae Colubrinae Micruridae Crotalinae Crocodylidae + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + — + + + + — + + — — + — — + — — + + — + + + + — + + — — + — + + — + + + + — + + + — — + — — + — — + — + + + + + — — + + — + — + + + + + — + + + + + + + + — — + MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 1973; Laurent, this volume). Bufo and six other genera occur in South America, and Bufo and seven other genera occur in Africa (principally tropical western Africa), but Bufo and five other genera inhabit southeast- ern Asia and adjacent islands. One genus (Crepidophryne) is endemic to Central America, and Bufo is widespread in the Hol- arctic Region. Differing views have been ex- pressed on the dispersal of Bufo ( Blair, 1972; Savage, 1973; Duellman, this volume; Lau- rent, this volume). Bufo occurs throughout South America, but only members of the Bufo spinulosus group are present in Patagonia, the austral forests, and the high Andes (Cei, 1968, 1972). Rhamphophryne and Atelopus are primarily northern Andean; Dendrophryniscus is in Amazonia and the Brasilian Shield, Melano- phryniscus in the Chaco and adjacent areas, and Oreophrynella in the Guiana Highlands (Trueb, 1971; McDiarmid, 1971). Brachycephalidae. — Unknown in the fossil record, the two small frogs comprising this family are restricted to humid coastal low- lands of southeastern Brasil (Izecksohn, 1971). Although superficially resembling a specialized bufonid, brachycephalids lack Bidder's Organs (McDiarmid, 1971), an uniquely derived character in the Bufonidae (Lynch, 1973). The phylogenetic position of this endemic South American family is not clear, but presumably it arose from a lepto- dactylid-primitive bufonid stock. Rhinodermatidae. — Known from two spe- cies restricted to austral forests (Formas, et al., 1975), Rhinoderma is considered to be most closely related to the bufonids by Lynch ( 1971, 1973 ) and must be considered as of temperate South American origin. Dendrobatidae. — Lacking a fossil record but composed of three Recent genera, the dendrobatids are especially speciose in the northern Andes, Choco, and western Ama- zonia, but also occur in eastern Amazonia and on the Guianan and Brasilian shields. Lynch ( 1971 ) demonstrated that the dendrobatids are derived from the elosiine leptodactylids and thus arc of South American origin. Spe- cies of all three genera occur in lower Central America, presumably having arrived there after the closure of the Panamanian Portal in the late Pliocene. Pseudidae. — An autochthonous South American family containing two genera and four species (Gallardo, 1961) and widely dis- tributed in tropical and subtropical cis-An- dean lowlands, these aquatic frogs have been considered as relatives of the leptodactylids (Savage and Carvalho, 1953) or hylids (Lynch, 1973). HyUdae. — Although Estes and Reig ( 1973 ) mentioned the existence of Paleocene hylid material from Brasil, these specimens have not yet been described. The mid-Mio- cene Australobat melius from Australia has been referred to the Hylidae by Tyler ( 1974 ) ; a presumed hylid is known from the Oligo- cene of North America (Holman, 1968), and Hyla is known from the Miocene of Europe (Noble, 1928). By far the greatest diversity of hylids is in South America ( 22 genera, 313 species), as compared with Middle America ( 15 genera, 129 species; Duellman, 1970). Six Hyla, plus two endemic genera (Osteopdus and Calyptahyla) inhabit the West Indies ( Trueb and Tyler, 1974 ) . The Holarctic hylid fauna is depauperate, but in the Australo- Papuan Region 118 species are known in the genera Litoria and Nyctimystes (Duellman, 1977; Tyler and Da vies, 1978), and nine more if Cyclorana is included in the family (Tyler, et al., 1978; Tyler, this volume). Like the leptodactylids, the Australian hylids are of questionable relationship with the South American hylids. Savage (1973) resurrected the family name Pelodryadidae for the Aus- tralo-Papuan "hylids" and considered them to be derived independently from the Neotropi- cal hylids. Tyler (this volume) emphasized the lack of evidence for such an arrangement. In Australia, Cyclorana seems to be intermedi- ate between the Australian "leptodactylids" and "hylids" and may prove to establish a phylogenetic link between the two families on that continent. No such intermediates are Fig. 1:1. Major eco-physiographic regions of South America. Temperate regions: Austral forests (AF), Patagonia (PAT). Tropical evergreen forests: Amazonia (AM), Choco (CH), Atlantic coast (AC). Tropical and subtropical nonforests: Caribbean coastal desert (CD), Llanos (LL), Savannas (black), Caatinga (CA), Cerrados (CE), Gran Chaco (GC), Pampas (PA), Monte (MO), Espinal (ES), Matorral (MA), Atacama Desert (AD). Mountains (stippled): Andes (A), Guiana Highlands (G), Brasilian Highlands (B). Regiones ecofisiograficas mayores de Sudamerica. 8 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 known in South America, and even the mon- ophyly of the Neotropical hylids has been questioned. Maxson (1976) provided im- munological evidence that the phyllomedu- sines were not closely related to the other hylids. Within South America, the phyllomedusine hylids are widespread in Amazonia, Atlantic forests, and the Guianan and Brasilian shields. PhyUoinedusa is primarily South American, with only two species (one endemic) in Cen- tral America; the Central American Agalych- nis is represented by three species in the Choco and one endemic species in western Amazonia. The hemiphractines are restricted to northwestern South America with one spe- cies entering Central America (Trueb, 1974). The amphignathodontines are most speciose in northwestern South America (two species enter Central America ) , but with three genera on the Brasilian Shield and one in the Guiana Highlands. Among the hylines, all of the South American genera are endemic to the continent, except one species of Phrynohyas and several species groups of Hyla that enter Central America. Two species of the Middle American Smilisca enter South America. Assuming that at least the Neotropical hylids arose in South America, some stocks must have entered Central America by waif dispersal prior to the closure of the Panaman- ian Portal in the late Pliocene. These stocks were the ancestors of the several genera and species groups of Hyla endemic to Middle America. After the closure of the portal sev- eral groups dispersed northward into Central America (PhyUomedusa, Hemiphractus, Gas- trotheca, Phrynohyas, Hyla albomarginata, H. boans, H. bogotensis, H. leucophyllata, and H. rubra groups) and representatives of two Middle American genera (Agalychnis and Smilisca) dispersed into South America. Centrolenidae. — No fossils are known. Two genera and 46 species inhabit cloud for- ests in the Andes, whereas a few species occur on the Guianan and Brasilian shields and in Amazonia, the Choco, and Central America (Duellman, 1977). Obviously of South Amer- ican origin with Late Cenozoic dispersal into Central America, the relationships of the cen- trolenids usually are thought to be with the hylids, but no convincing evidence is avail- able. Ranidae. — Although no fossils are known before those in the Oligocene of North Amer- ica (Holman, 1968), the center of origin and dispersal of ranids quite clearly is in Africa (Savage, 1973), and presumably occurred af- ter the rift of South America and Africa in the Cretaceous. The single South American ranid, Rana palmipes, is widespread in Cen- tral America and must have entered South America after the establishment of the isth- mian link. The species is widespread in the tropical lowlands of South America. Microhylidae. — This large and diverse family presents one of the most controversial issues in anuran phylogeny and classification. Although the family is clearly of Gondwanan origin, the present interpretations of phylog- eny and biogeography are in conflict at times. Savage ( 1973 ) based his biogeography of the microhylids on Starrett's ( 1973 ) interpreta- tion of anuran phylogenv as demonstrated by larvae. Zweifel (1972), Lynch (1973), Sokol ( 1975 ) , and Tyler ( this volume ) provided compelling arguments based on diverse mor- phological, developmental, and biogeographic evidence against the Starrett and Savage model. All South American microhylids belong to the subfamily Microhylinae, which is shared with North America, tropical southeastern Asia, and the Malayan Archipelago. It is most logical biogeographically and phylogenetically that the Neotropical microhylids evolved in isolation in South America and that a stock that subsequently gave rise to the North American Ga.strophryne and Hypopachus managed to enter Central America from the south during mid-Cenozoic times; microhy- lines are known from the Miocene of Florida (Holman, 1967). The 16 genera of South American microhylids occur throughout the tropical and subtropical lowlands with the greatest diversity in the southern part of their range, particularly on the Brasilian Shield. Three genera (3 species) entered Central America after the connection of the conti- nents. The monotypic Geobatrachus in the Sierra Nevada de Santa Marta in northern Colombia tentatively was referred to the Microhylidae by Lynch (1971), but Duellman (1975) showed that this small frog has a combination of characters that precludes its assignment to 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 9 any family as presently defined. Geobatra- chus has not been included in the numerical account of the microhylids. Rhinatrematidae. — Lacking fossils and en- demic to South America, these primitive cae- cilians are a sister group of the ichthyophiids (Nussbaum, 1977), a family restricted to In- dia, tropical southeastern Asia, Malayan Arch- ipelago, and the Philippines. Roth South American genera occur on the Guianan Shield and one (Epicrionops) has four species on the forested slopes of the northern Andes. Typhlonectidae. — The specialized aquatic caecilians are autochthonous to South Amer- ica, where they are distributed discontinuous- ly in the Caribbean and Amazonian lowlands and in the Parana Rasin. Caeciliidae. — The presence of a single fos- sil from the Paleocene of Rrasil (Estes and Wake, 1972 ) , possibly referable to this family, signifies a long history of caecilians in South America, where nine genera and 46 species now occur in the humid lowland tropics and forested slopes of the northern Andes. Four genera inhabit Middle America; the endemism there in Dermophis and Gymnopis indicates that a caecilian stock entered Central America prior to the establishment of the isthmian link, whereas the other two (Caecilia and Oscaecilia) are both widespread in South America, and evidently dispersed into Central America after the closure of the Panamanian Portal. The presence of caeciliids in tropical Africa (6 genera), India (3 genera), and the Seychelles Islands (3 genera), as well as in South America, is indicative of a widespread Gondwanan distribution prior to the Late Cretaceous. Pelomednsidae. — The classical, present Gondwanaland distribution pattern of pelo- medusid turtles is complicated by their occur- rence in Cretaceous deposits in Europe and North America and in the Eocene of Asia. However, the family has an extensive fossil record beginning in the Cretaceous in both South America and Africa (Wood, 1970). The single South American genus, Podocnemis, is widely distributed in the cis-Andean tropical lowlands. Chelidae. — Considered to be a derivative of the Pelomedusidae (Gaffney, 1975, 1977), the chelids are known are fossils only from Australia (early Tertiary to Pleistocene) and South America (Eocene to Pleistocene). The seven genera endemic to South America are widely distributed in the cis-Andean tropics, mostly in Amazonia. Kinosternidae. — Known as far back as the Oligocene in North America, the great ma- jority of kinosteraids (4 genera, 19 species) occur in North America and northern Central America. Only three species of Kinosternon occur in lower Central America; one of these also is widespread in cis-Andean South Amer- ica, and two vicariant species occur in the Choco. The kinosternids obviously are a post- portal entrant into South America from the north. Chelydridae. — The snapping turtles have an extensive fossil record throughout the Cenozoic in North America, where two genera are extant. One species of Chelydra inhabits lower Central America, and one occurs in the Choco in South America. Chelydra obviously is a recent immigrant into South America. Emydidae. — This family has an extensive fossil record in the Holarctic Region and to- day is distributed mainly in North America and the Oriental Region. The genus Chry- semys (=Pseudemys) is speciose in North America and in the West Indies, occurs in Central America (same species in northern South America), and is represented by anoth- er species in the Parana Rasin. Two species of the Central American Rhinoclemys occur in the Choco, and Rhinoclemys is known from the Pleistocene of Ecuador. Possibly an early Chrysemys stock waifed to South America, but Rhinoclemys and Chrysemys scripta cer- tainly entered the continent from the north subsequent to the establishment of the isth- mian link. Meiolaniidac. — This extinct family known from South America (Late Cretaceous to early Eocene) and Australia (Miocene to Pleistocene) seems to antedate the testudi- nids. Their fossil record suggests a Gondwa- nan history similar to that of the chelids. Testudinidae. — This family is cosmopoli- tan, except in the Australo-Papuan Region, and possibly had its initial radiation in Lau- rasia (Cracraft, 1974). The genus Geoche- lone occurs today in South America, Africa, India, southeastern Asia, and on the Gala- pagos Islands; it is known from the late Oligo- cene through the Pleistocene in South Amer- 10 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 ica, where it presently occurs throughout cis- Andean lowlands southward to northern Pata- gonia. Auffenberg ( 1971 ) suggested that the South American stock probably entered the continent from the north in the Oligocene; this implies waif dispersal. Trionychidae. — Although presently wide- spread in sub-Saharan Africa and in the Ori- ental Region, trionychids have an extensive fossil history in the Holarctic Region, where Trionyx occurs today in North America. Wood and Patterson ( 1973 ) reported a tri- onychid from the late Pliocene of Venezuela; probably this was a waif, because no other fossil or Recent trionychids are known south of northeastern Mexico. Gekkonidae. — Although the gekkonids were considered to be of uncertain geographic origin by Cracraf t ( 1974 ) , reevaluation of Kluge's (1967) phylogenetic scheme of the family suggests that the gekkonids are early Gondwanan. The primitive eublepharines oc- cur in Africa, southern Asia, and North Amer- ica. The diplodactylines are dominant in Australia and have dispersed onto islands in the southwest Pacific. Sphaerodactylines are restricted to tropical America. The gekko- nines are pantropical, being most diverse in the Indian, Oriental, and Ethiopian regions but also with many representatives in Aus- tralia and on Pacific islands. Of the 11 gen- era of gekkonines in South America, only eight are endemic to the American tropics, if the Old World species presently assigned to Phyllodactyhis are not considered to be con- generic, as suggested by Dixon and Anderson (1973). Except for the speciose Phyllodac- tyhis in dry habitats in western and northern South America (also Middle America, West Indies, and Galapagos Islands), most of the endemic genera are represented by only one or two species and all live in eastern South America, save for the monotypic Thccadacty- lus in western Amazonia, the Choco, Guianan Shield, Central America, and Lesser Antilles. The only genus with more than two species is Homonota (8 species) in temperate cis-An- dean areas; the related Garthia with two spe- cies occurs in the southern Atacaman Region. The other gekkonines in South America ( Gymnodactylus — 3 species, Hemidactylus — 5, and Lygodactylus — 2) are most speciose in Africa.1 In fact, all but one of the species of Hemidactylus in America are widespread in Africa and elsewhere. The presence of gekkonids in the Paleo- cene of Brasil (Estes, 1970) suggests that gekkonines may have been present in South America prior to the separation of Africa and South America in the Cretaceous ( Bons and Pasteur, 1977). Trans-Atlantic waifing could explain the presence in South America of Af- rican genera, such as Gymnodactylus and Hemidactylus, but some of the latter most likely were transported by man. Of the five genera of sphaerodactyline geckos in South America, two (Coleodactylus and Pseudogonatodes) are endemic to for- ested cis-Andean regions in northern South America. Gonatodes is widespread in tropical lowlands, and one species has dispersed north- ward into Central America and the West In- dies. Lepidohlcpharis occurs in northwestern South America and has two species in Pana- ma. Sphaerodactylus is most speciose in the West Indies but with a few species in Central America, two of which extend into South America. Iguanidae. — This large and diverse family is first known in the fossil record from the Upper Cretaceous of Brasil (Estes and Price, 1973) and is diverse in the late Paleocene of Brasil (Estes, 1970). The earliest North American fossils are from the Eocene. Evi- dently an early iguanid stock reached North America prior to the separation of the conti- nents or waifed between the two. Informally, the iguanids are divided into five major groups (Etheridge, 1964, 1967). The sceloporines dif- ferentiated in North America and were paral- leled by an extensive radiation of tropidurines in temperate South America. The iguanines and basiliscines are primarily Middle Amer- ican with genera of the former endemic to the Galapagos Islands (Amblyrhynchus and Conolophus) and the West Indies (Cyclura); one species of Iguana and three of Basiliscus 'Smith et al. (1977) recognized the South American Lygodactylus as the sole representatives of the genus Vanzoia, but Kluge (pers. coram.) informed me that the South American species are perfectly good exam- ples of Lygodactylus. 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 11 have entered northern South America. The anolines are widely distributed in the tropics of South America, Middle America, and the West Indies. Obviously waif dispersal be- tween South America and the emerging An- tilles and Central America during the Tertiary accounted for some of the patterns of distribu- tion, many of which have been masked by more recent dispersal after the closure of the Panamanian Portal. It certainly seems safe to assume that the origin of tropidurines and some anolines was in South America. The presence of two genera of iguanids on Madagascar has been interpreted as evidence for the former occurrence of the family in Africa with subsequent extinction there (per- haps owing to competition with agamids and chamaeleontids). The similarity of caudal structure of the Madagascaran iguanids to the tropidurines (Etheridge, 1967) and the pres- ence of iguanids in the Cretaceous of South America do not contradict that hypothesis. The iguanine Brachylophus, endemic to islands in the southwest Pacific, apparently is an example of long-distance rafting via the Trans-Pacific Current (Cogger, 1974). Teiidae. — Although presumed teiids are known from the Late Cretaceous in North America, those do not seem to be ancestral to living North American teiids, whereas late Paleocene teiids of South American resemble extant Neotropical genera (Estes, 1970). All living teiid genera occur in South America, where they are widespread throughout the continent, except for Patagonia and the aus- tral forests. Five genera have representatives in the West Indies, and 10 genera extend into Central America. With the exception of Cne- midophorns, which is widespread and spe- ciose in North America, all other teiids prob- ably arrived in Central America after the formation of the isthmian link. Scincidae. — Although skinks are known from the Paleocene of Rrasil and the Late Cretaceous of North America (Estes, 1976), only four genera presently occur in the Amer- icas; this is only a small fraction of the family containing at least SO genera and more than 1,000 species (Greer, 1970). Occurring throughout South America, except for the Andes and cool temperate regions, are eight species of Mabuija, a genus containing about 75 additional species in Africa, Madagascar, southern Asia, and the Pacific islands, plus an additional two species in Central America and the Antilles. Tihen (1964) suggested an Eur- asian origin of the North American skinks (Eumeces and Scincella), but an African ori- gin of the South American Mabmja seems to be reasonable. Anguidae. — Presently distributed primar- ily in North America, western Eurasia, and southeastern Asia, the anguids have an ex- tensive fossil history dating from the Late Cretaceous in North America (Meszoely, 1970). Two genera occur in South America. Diploglossus is speciose in Central America and the West Indies, and one of the South American species is shared with Central America. The endemic South American Ophi- odes (4 species) in the south-central part of the continent apparently evolved from an anguid stock that entered South America from the north in the Cenozoic. Amphisbaenidae. — Numerous fossil am- phisbaenians are known from the Paleocene to Miocene in North America, Eocene to Plio- cene of Europe, and Oligocene of Mongolia. With the exception of Bipes and the Pale- arctic Blanus, the amphisbaenids (sensu Ber- man, 1973) are Neotropical and African — 10 genera in Africa (one ranging into Europe) and six in South America (plus one endemic to Cuba). The generic and specific differen- tiation in Africa (10 genera, 52 species) and South America (6 genera, 45 species), and the possible presence of Amphisbaena in both Africa and South America (Gans, 1967), plus 10 species endemic to the West Indies and one South American species extending into Central America, are suggestive of an African- South American amphisbaenid interchange. However, the place of origin of the amphis- baenids still remains problematic. Anomalepidae. — Unknown as fossils, 17 of the 20 species and all four genera of anoma- lepids occur in South America. One species each of Anomalepis, Helminthophis, and Liotyphlops occurs in Central America, and another species of Liotyphlops ranges from Costa Rica into northern South America. On the basis of present distributions, the anoma- lepids seem to be a South American group that only recently invaded Central America. 12 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 In South America the family is widespread in trans-Andean and eis-Andean tropical low- lands. Leptotyphlopidae. — The single genus in this family containing about 64 species is widespread in tropical and subtropical South America, Middle America, southwestern United States, Africa, and southwestern Asia. No fossils are known. On the basis of present distribution it is reasonable to suggest that the leptotyphlopids had a western Gondwa- naland origin and subsequently spread north- ward into Central and North America and independently into Asia. Typhlopidae. — These fossorial snakes are known from two genera ( Typhlina, 33 species in Australia, New Guinea, Solomon and Fiji islands) and Typhlops (about 114 species throughout tropical and subtropical parts of the world, except Australia). Only three spe- cies occur in South America; another five are in Central America, and 16 occur in the West Indies. Thus, with respect to the total differ- entiation of the family, the Neotropics are poor in typhlopids. The only fossils (Eocene- Miocene of Europe) are of no help in inter- preting the paleobiogeography of the group. In the absence of any evidence for the occur- rence of typhlopids in North America, a west- ern Gondwanaland origin for the Neotropical stocks might be suggested. Boidae. — Represented by an extensive, world-wide fossil record from the Upper Cre- taceous through the Eocene (only Pleistocene in Australia ) , the boids seem to have been the dominant snakes throughout the world in the Early Cenozoic. Evidently they had dispersed widely before the breakup of Pangaea. The Early Cenozoic South American boid Madtso- ia also is known from the Late Cretaceous of Madagascar, and a related boid, Wonambi, is known from the Pleistocene of Australia (Ty- ler, this volume). In some respects the dis- tribution of these fossils parallels that of liv- ing boines — eight genera in the Neotropics, two in Madagascar, one in New Guinea and islands in the southwest Pacific. Presently the family is widespread in tropical South Amer- ica and especially diverse in Amazonia. Aniliidae. — These problematic fossorial snakes have a long history in northern conti- nents, dating from the Middle Cretaceous in North America and the Eocene of Europe. A Late Cretaceous snake, Dinilysia from Pata- gonia, is considered to be related to aniliids ( Rage, 1977 ) , and hue aniliids were reported from the Eocene of Brasil by Baez and Gas- parini (this volume), who support Cracraft's ( 1974 ) contention that aniliids are of Laura- sian origin. Nonetheless, entry into South America possibly was by way of Africa. The one living South American aniliid is wide- spread in cis-Andean tropical lowlands. Tropidophiidae. — Structurally, the tropi- dophiids are intermediate between the boids and colubroid snakes. Tropidophis has three widely dispersed species in South America (northern Andes and southeastern Brasil) and 12 species in the West Indies. Trachyboa and Ungaliophis occur in the Choco and Central America. The tropidophiids are considered to be of South American origin with subsequent northward dispersal. Colubridae. — The poor fossil record and taxonomic chaos of the colubrids (sensu lato) permit only the most general comments to be made about this immense and important fam- ily. My use of subfamilial designations fol- lows that of Dowling (1975) but eliminates some apparent misapplications not especially germane to the Neotropical colubrids. A brief summary of the colubrid snakes follows. 1. Xenodontinae: 93 genera, about 570 species. Sixty genera occur in South America; 22 of these are shared with Central America. Seven genera are endemic to the West Indies; 13 are restricted to North America (north of the Isthmus of Tehuantepec in southern Mex- ico), and 13 are Middle American. Included in this group of rear-fanged genera are fossor- ial, terrestrial, aquatic, and arboreal snakes. Most arboreal xenodontines are nocturnal, and few (Alsophis, Dromicus, Leimadophis, Ly- gophis) are diurnal racer-like snakes. Most xenodontines feed on frogs, lizards, or other snakes. 2. Lycodontinae: 79 genera, about 285 species. These rear-fanged snakes are distrib- uted primarily in the African and Oriental regions and peripherally in northern Australia and the Palearctic Region. 3. Colubrinae: 74 genera, about 440 spe- cies. Widespread in the Holarctic Region, some genera occur in the Ethiopian and Ori- 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 13 ental regions, and two genera reach Australia. Thirty-two genera occur in the New World; of these, 12 are in South America, but only one of those (Drymoluber) does not occur in Central America. Most of the colubrines in South America (Chironius, Dendrophidion, Dn/marchon, Drymobius, Drymoluber, Lep- tophis, Masticophis, Mastigodryas, Spilotes) are diurnal racer-like snakes that are terres- trial or arboreal. 4. Natricinae: 34 genera, about 170 spe- cies (excluding from the Natricinae the Old World snakes more appropriately referred to Acrochordidae and Homalopsinae). The na- tricines are widely distributed in the Holand- ric and Oriental regions, with a few represen- tatives in Africa and one in northern Aus- tralia. Nine genera of natricines occur in North America, with Thamnophis extending to Costa Rica. Even if these subfamilial groups are mon- ophyletic, the historical biogeography of the colubrids still remains shrouded. It is evident from the distributions of the subfamilies that centers of dispersal (and perhaps of origin) can be ascertained, but ancestors cannot. Ap- parently the xenodontines evolved in South America and the lycodontines in the African- Indian-Asian Arc and had corresponding par- allel radiations in the New World and Old World, respectively. Colubrines and natri- cines probably are Holarctic in origin. If these truly are the centers of origin and dis- persal, it is possible to make some reasonable generalizations about the South American colubrid fauna. 1. Although colubrids are known from the Eocene (Rage, 1975a,b), they be- came dominant in the upper Miocene (Holman, 1976), when they first ap- pear in South America (Baez and Gasparini, this volume). 2. Xenodontines evolved in South Amer- ica, where they are the dominant snakes today and the only colubrids in the southern part of the continent. 3. Some xenodontine snakes were present in Middle America in the Cenozoic; these differentiated, and some of them dispersed as far as northeastern North America. 4. Colubrine stocks in North America in- vaded Central America, and some of them may have waifed to South Amer- ica before the closure of the Pana- manian Portal in the late Pliocene. 5. With the closure of the Panamanian Portal in the late Pliocene there was an interchange of South American xenodontines northwards and Middle American colubrines and xenodontines southward. 6. Natricines never have extended farther south than Central America. 7. The West Indian colubrid fauna is com- posed of xenodontines derived either from Central or South America (some widespread colubrine species are recent immigrants into the Lesser Antilles ) . Micruridae. — Formerly associated with the Elapidae, micrurids have been shown to be an independently derived group from Ela- pomorphus-Apostolepis xenodontines in South America (Savitzky, 1978). An upper Miocene fossil from Nebraska (Holman, 1977), to- gether with the presence of Micruroides and numerous species of Micrurus in Central America, is suggestive of dispersal of micru- rids into Central America in the Cenozoic with additional interchange after the closure of the Panamanian Portal. Micrurus has 27 endemic species in South America and ranges throughout the lowlands and moderate eleva- tions south to northern Patagonia; the mono- typic Leptomicrurus is restricted to Amazonia. Viperidae. — Although the vipers are pri- marily an Old World group, which apparently originated in the Palearctic Region (Marx and Rabb, 1965), one lineage — the crotaline vipers — may have evolved in the Oriental Re- gion and dispersed via Beringia to North America (Burger, 1971). The earliest North American fossil crotalines are of Miocene age (Holman, 1977). The presence of the pre- sumably primitive crotaline Lachesis muta in northwestern South America and lower Cen- tral America suggests that the crotaline vipers entered South America after the establishment of the isthmian link. On the other hand, the presence of many species of Bothrops throughout South America as far south as Patagonia, as well as many different species in Middle America, is suggestive of an earlier dispersal into South America. Bothrops is 14 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 known as a fossil in South America only from the Pleistocene of Bolivia. The presence of Bothrops in the West Indies attests to their abilities of over-water dispersal. Crotalus cer- tainly is a post-Pliocene immigrant into South America, where it has spread throughout non- forested tropical areas. Crocodylidae. — Present distributions and fossil records indicate that the alligatorines and Crocodyhts entered South America from the north ( Sill, 1968; Baez and Gasparini, this volume), although Paleocene crocodylines in South America are suggestive of possible Afri- can derivation. Likewise, the presence of gavial-like crocodilians in South American de- posits (Baez and Gasparini, this volume) im- plicates at least the early crocodilians in a Gondwanan distribution. Possibly the croco- dilians presently living in South America were derived from North American stocks, whereas the ancient Gondwanan crocodilians are ex- tinct in South America. Obviously, entry of crocodilians into South America from the north prior to the formation of the isthmian link was facilitated by their abilities at tra- versing open water. Only Caiman is wide- spread throughout the tropical lowlands of South America; Crocodijlus is restricted to the northern part of the continent (one spe- cies endemic to the llanos). The other two genera (Melanosuchus and Paleosnchus) are in western Amazonia. EXTRA-CONTINENTAL RELATIONSHIPS Elsewhere in this volume detailed com- parisons of the origins of the African and Australian herpetofaunas with respect to that of South America have been made by Lau- rent and Tyler, respectively. In this section I compare the compositions and taxonomic di- versities of those three faunas. Furthermore, I provide a discussion of the herpetofaunal relationships between South America and North America and between South America and the West Indies. Herpetofaunas of Gondwanan Continents South America contains 37 living families of amphibians and reptiles, Africa 26, and Australia 17. Among the amphibians only three families are shared by the three conti- nents— Leptodactylidae (only one genus with three species in Africa), Ranidae (only one species each in South America and Australia), and Microhylidae (only two genera with seven species in Australia). Two additional families are shared by South America and Africa (Pi- pidae and Bufonidae) and one (Hylidae) by South America and Australia. Among the nonmarine reptiles, six families are common to the three continents — Crocodylidae, Gek- konidae, Scincidae (only one genus with eight species in South America), Typhlopidae, Boi- dae, Colubridae (only four genera with six species in Australia). Five additional families are shared by Africa and South America (Pe- lomedusidae, Testudinidae, Amphisbaenidae, Leptotyphlopidae, and Viperidae). Three other families are shared by Africa and Aus- tralia (Agamidae, Varanidae, Elapidae), whereas only one other (Chelidae) is shared by South America and Australia. Faunal re- semblance factors (Duellman, 1966) at the family level are highest between Africa and Australia (0.56), followed by Africa and South America (0.54) and Australia and South America (0.40). South America has five endemic families of amphibians plus two others that have dispersed only to lower Cen- tral America, but no endemic families of reptiles. Africa has one endemic family of lizards (Cordylidae), one of caecilians (Sco- lecomorphidae), and one of frogs that has dispersed to Madagascar and the Seychelles Islands ( Hyperoliidae ) . Australia has no en- demic families, but the Pygopodidae is shared only with New Guinea. Examination of the amount of taxonomic diversity in anurans and lizards on each conti- nent reveals that the South American anuran fauna is much more diverse than that on the other continents but that Africa has the most species of lizards.- Analyses of diversity in- L' Data for Australia were gathered primarily from Cogger ( 1975); for African lizards chiefly from Mer- tens (1963, 1966), Wermuth (1965, 1967, 1968), and Greer (1970, 1974); for African frogs and South American frogs and lizards from my personal compilations. Owing to the absence of modern com- prehensive works on African snakes, a complete com- pazine analysis of snakes was not attempted. South America has 9 families, 96 genera, and 556 species of snakes, compared with 4 families, 36 genera, and 104 species of nonmarine snakes in Australia. The species/area values for snakes are 31.2 for South America and 13.5 for Australia. 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 15 eluded the differentiation of genera and spe- cies on each continent and the numbers of species per unit area (Table 1:6). The taxonomic diversity among South American frogs is extremely high in two fami- lies— Leptodactylidae and Hylidae; these con- tain 64 percent of the genera and 73 percent of the species of South American frogs. These same two families are the dominant compo- nents of the Australian amphibian fauna, ac- counting for 88 percent of the genera and 95 percent of the species. The dominant African families are Bufonidae and Ranidae, together accounting for 44 percent of the genera and 73 percent of the species. In South America the dominant families of lizards are the Iguanidae and Teiidae (com- bined, 77% of genera, 83% of species). In Australia the Scincidae alone accounts for 37 percent of the genera and 54 percent of the species, whereas the Gekkonidae and Agami- dae are secondary (combined 48% of the gen- era, 32% of species). Gekkonids and scincids are the most diverse African families (60% of genera, 56% of species) followed by cordylids and lacertids ( 35% of genera, 28% of species ) . The presence or absence of families on the three continents relates primarily to historical factors, whereas the diversity within families may be dependent upon the amount of time that the family has occupied the continent or perhaps also the size of the area. Further- more, ecological factors may be extremely important in the evolutionary diversity of a family provided that the family has been established on the continent for a sufficient period of time. There is no easy or objective way to measure habitat diversity within and between the three continents. One method is to compare the sizes of the continents with respect to taxonomic diver- sity. Africa is by far the largest of the con- tinents (30,264,000 km2) followed by South America (17,793,000 km2) and Australia (7,687,000 km2). Analyses of numbers of species per unit area show that South Amer- ica has an excessive number of frogs and Australia an excessive number of lizards; all other values are negative (Table 1:6). Lizards usually are most diverse and nu- merous in xeric areas, and the vast majority of Australia is xeric. Likewise, most of Africa is arid, and this is reflected in the large number of lizards (506 species) on that continent. Table 1:6. — Taxonomic Diversity of Anurans and Lizards on Gondwanan Continents. South America Africa Australia Anurans Families 11 98 996 8.9 7 63 355 9.0 4 Genera ... 26 Species 155 Genera/Family 6.5 Species/Genus 10.2 5.6 6.0 Species/Family 90.5 50.7 38.8 Species/million km" . 56.0 11.7 20.1 Deviation from expected .. . +26.7 -17.6 -9.2 Lizards Families .. 5 84 471 16.8 7 78 506 11.1 5 Genera 54 Species 364 Genera/Family 10.8 Species/Genus 5.6 6.5 6.7 Species/Family 94.2 72.3 72.8 Species/million knr 26.5 16.7 47.3 Deviation from expected _ -3.7 -13.5 + 17.1 However, proportional to size, Africa has fewer species of lizards than either Australia or South America. The comparatively few lizards in Africa might be related to the vast areas of extreme deserts (Sahara and Kala- hari), which although inhabited by some spe- cialized lizards are not species rich. Another factor might be the presence of herds of large mammals on the plains (Janzen, 1976). Fur- thermore, Pianka (1971) and Pianka and Huey (1971) suggested that lower species di- versity of lizards in the Kalahari Desert, as compared with central Australian deserts, may be a result of the comparatively richer avian fauna in the Kalahari. Frogs are most diverse in humid tropical forests. By comparison with the South Amer- ican forests, the lowland tropical forests in Africa and Australia are much smaller; the Congo Basin has about 2,000,000 km2 of rainforest, whereas the Amazon Basin has 4,500,000 km2 (Richards, 1973). Further- more, the South American lowland tropical rainforests are in three distinctly separate units — Amazonian, trans-Andean, and Atlan- tic coastal, each harboring 203, 111, and 168 endemic species of amphibians, respectively (Lynch, this volume). Quaternary climatic- vegetational changes in west African rainfor- ests (Moreau, 1963, 1969) apparently resulted in the elimination of proportionately more of the lowland tropical forests there than in South America (Haffer, 1974). Montane rain- forests (cloud forests) are much more exten- 16 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 sive in South America than in Australia, and especially, in Africa. In South America these forests support rich, localized anuran faunas. The comparative herpetofaunal diversity in the three Gondwanan continents first is the result of historical components that were ei- ther present on the continents when they were formed or emigrated there after the continents became discrete units. However, the taxonomic diversity is primarily a factor of habitat diversity. In comparison with the other continents, South America offers a much more diverse landscape, climate, and vegeta- tion. The cool temperate rainforests, Patagon- ian steppes, and high Andean punas and paramos stand in marked contrast to the pampas, llanos, caatinga, and Atacama Desert, which in turn harbor distinctly different bi- otas than do the lowland and montane rain- forests. Herpetofaunal Comparisons of North and South America It has been nearly half a century since Dunn's (1931) then classic essay on the herpetofauna of the Americas; Dunn's ap- proach was based entirely on Matthew's (1915) hypothesis of northern origin and southward dispersal of mammalian orders. Savage (1966) provided a well-documented account of the distribution patterns of am- phibians and reptiles in Central America and emphasized the degree of differentiation and endemism in that fauna by recognizing a Mesoamerican herpetofauna distinct from the Nearctic and Neotropical faunas. Although the patterns delineated by Savage are realistic, the interpretation of the origins and times of dispersal can now be modified by new paleon- tological and geomorphological information. Of primary importance to a biogeographic- al analysis of the Central-South American re- gion is the history of the connection of the two major continental masses. According to Dietz and Holden (1970), after the initial breakup of Pangaea in the Early Jurassic ( =T80 m.y.b.p. ) there was no direct land connection between North America and South America until the Late Tertiary, although the positions of the two continents converged beginning in the mid-Cretaceous. An island arc, the proto- Antilles, existed between nuclear Central America and South America in the Cretaceous and Early Tertiary (Holden and Dietz, 1972; Malfait and Dinkelman, 1972); this arc moved eastward, relative to the westward drift of the American continents, through the Tertiary and formed the present Lesser Antilles. The region of lower Central America (Costa Rica and Panama), or the isthmian link, formed as a volcanic archipelago in the Oligocene; addi- tional land emerged, and the archipelago co- alesced with nuclear Central America 10-12 m.y.b.p. and finally with South America about 5.7 m.y.b.p. During the late Mesozoic ( 180-90 m.y.b.p.) South America had direct land con- nections with Africa (Grant, 1971; Reyment and Tait, 1972; Larson and Ladd, 1973) and with Australia via Antarctica until the Eocene or Oligocene («50 m.y.b.p.) (McGowran, 1973; Veevers and McElhinny, 1976). Throughout the Cenozoic until the late Plio- cene (5.7 m.y.b.p.) there was no land con- nection with North and Central America. Thus, for about 45 million years South Amer- ica was isolated from other continents. How- ever, the island arc ( proto- Antilles ) in the Late Cretaceous and Early Tertiary and the Central American Archipelago in the Middle to Late Tertiary provided opportunities for limited faunal exchange between the conti- nents. Fossil evidence (albeit scanty or non- existent for some groups) and present pat- terns of distribution and speciation provide a basis for analysis of the herpetofaunal inter- change between Central America and South America (see preceding review of families and Savage, 1966). Examination of the family groups of amphibians and reptiles that have entered into the exchange (Table 1:7) shows two modes of entry. The first is by means of dispersal across one of two archipelagos — the early proto-Antillean island arc or the later Central American Archipelago. The second is by direct dispersal after the establishment of the isthmian link. My analysis shows no amphibians entering South America from the north via the archi- pelago but five amphibian family groups dis- persing northward via the island route. Pre- 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 17 Table 1:7. — Postulated Herpetofaunal Exchange Be- tween South America and Central America. Across Via Family Panamanian Isthmian group Portal Link N^SS ->~N N -> S S -> N Plethodonhdae — — + Pipidae — — ~ Eleutherodactylini — + +? + Leptodactylinae — — — + Bufonidae — — + + Dendrobatidae — — — + Phyllomedusinae — + + + Hemiphractinae — — — + Amphignathodontinae — — — + Hylinae — + + + Centrolenidae — — — + Ranidae — — + — Microhylidae — + — + Caeciliidae — + + + Kinostemidae — — + — Chelydridae — — + — Testudinidae + — — — Gekkoninae — + — + Sphaerodactylinae — + + + Iguanidae (primitive) — + — — Iguanines + — + — Basihscines — — + — Anolines — + + + Teiidae — + + + Scincidae — — — + ■ Anguidae + — + ~ Amphisbaenidae — — — + Anomalepidae — — — + Leptotyphlopidae — + — — Typhlopidae — + — — Tropidophiidae — — — + Xenodontinae — + + + Colubrinae +? + +? Micruridae — +? +? + Crotalinae +? - + +? Crocodylidae + — + + sumably all of these (Eleutherodactylini, Phyllomedusinae, Hylinae, Microhylidae, Caeciliidae ) dispersed northward via the Cen- tral American Archipelago, which emerged in the Oligocene. However, the Hylinae may have dispersed earlier via the proto-Antilles, for hyline frogs are known from the Oligocene in North America, have had an extensive radi- ation in North America and nuclear Central America, and have dispersed into Eurasia (presumably via Reringia). Also, it is pos- sible that a proto-pipid frog entered North America via the proto-Antillean arc; this frog could be the ancestor of the Rhinophrynidae now restricted to Mexico and nuclear Central America but known from the Paleocene-Oli- gocene of North America. Reptilian dispersal via the islands appar- ently was much more extensive than that of the amphibians. Probable dispersers via the proto-Antillean island arc are primitive igua- nid lizards ( south to north ) and anguid lizards (north to south). Dispersal via the Middle to Late Tertiary Central American Archipelago included testudinids, iguanines, crocodylids, and perhaps some colubrines and crotalines ( all north to south ) and gekkonines, anolines, teiids, leptotyphlopids, typhlopids, xenodon- tines, and perhaps sphaerodactylines and mi- crurids (all south to north). The iguanine dispersal is postulated for the migration of an Ambhjrhynchus-Conolophus stock to the Galapagos Islands from the South American mainland, but possibly this stock waifed di- rectly from Central America ( Avery and Tan- ner, 1971). Colubrine southward dispersal probably was late in the history of the archi- pelago, if indeed these snakes did enter South America prior to the isthmian link. The evo- lution of the alpha and beta groups of Anolis north and south of the isthmus bespeaks the separation of the two groups on the two land masses (Etheridge, 1959; Savage, 1966). Pos- sibly crocodilians and gekkonines also dis- persed via the proto-Antillean island arc. Overland dispersal after the establishment of the isthmian link involved more northward than southward dispersal by amphibians, but it did permit entry into South America for the first time of plethodontid salamanders (2 gen- era) and ranid frogs (1 species), all wide- spread taxa in Central America. Other groups moving southward were some phyllomedusine and perhaps some eleutherodactyline frogs that were part of the Mesoamerican fauna evolved from South American stocks that earlier had invaded Central America. Also, a member of the Bufo valliceps group (B. coniferus) invaded South America. The north- ern infusion of South American taxa includes some groups that have speciated (S) and/or dispersed widely (D) in Central America — Eleutherodactylus (SD), Leptodactylus (SD), Physalaemus (D), Bufo marinus (D), Hyla ebraccata (D) and microcephalia (SD) groups, and Centrolcnella (SD). Most of the other South American amphibians have dis- 18 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 persed only into lower Central America and have not speciated there — Pipa (eastern Pan- ama); Pleurodema, Bufo typhonius, Elachis- tocleis, Relictivomer, Caecilia, and Oscaecilia (central Panama); Hemiphractus and Gastro- theca (western Panama); Glossostoma (Costa Rica); and Bufo haematiticus (Nicaragua). The South American Phyllomedusa buckleiji group has a species endemic to lower Central America (Duellman, 1970). All three genera of South American dendrobatids occur in low- er Central America; each has undergone some speciation (Savage, 1968). The Atelopus vari- us group has invaded Central America (to Costa Rica) and has undergone a bewilder- ing diversification (Savage, 1972). Late Tertiary and Quaternary overland dispersal amongst reptiles also was extensive. Southward dispersal brought chelydrid and kinosternid turtles, basiliscine lizards, and possibly colubrine and crotaline snakes into South America for the first time, whereas si- multaneously the first Mabuija, Amphisbaena, and anomalepid and tropidophiid snakes reached Central America. Mesoamerican groups originally derived from South Amer- ican stocks (beta anoles, Cnemidophorus, and many xendodontine snakes) dispersed into South America. Many South American taxa (Thecadactylus, alpha anoles, Ameiva, micro- teiids, and xenodontine and micrurid snakes) moved northward. The herpetofauna of eastern Panama con- tains many genera that are chiefly South American — Pipa, Rhamphophryne, Hemi- phractus, Gastrotheca, Elachistocleis, Caeci- lia, Geochelone, Lepidoblepharis, Enyalioides, Echinosaura, Amphisbaena, CoraUus, Trachy- boa, Atractus, Diaphorolepis, and Pseudoboa. The herpetofauna of the Chocoan lowlands of northwestern South America contains many species that are familiar to the herpetologist working in lower Central America, whereas the fauna in the Amazon Basin is greatly dif- ferent (at least at the species level); com- pare data given by Savage (1966) with those presented by Lynch ( this volume ) and Dixon (this volume). Herpetofauna of the West Indies Although the West Indies are peripheral to a discussion of the South American herpeto- fauna, it is germane to this essay to ascertain the herpetofaunal relationships of the two regions inasmuch as many genera and some species are common to the two. The history of the Caribbean Plate and the tectonic move- ments in the Antillean-Caribbean region have not been resolved, but Rosen (1975) sum- marized (and extended) the existing geologi- cal data and proposed a plausible vicariance model of Caribbean biogeography. Excluding introduced taxa, the herpeto- fauna of the West Indies ( not including Trini- dad, Tobago, Bonaire, Aruba, and Curacao) consists of 505 species in 57 genera ( Schwartz and Thomas, 1975 ) ; 476 of the species and 18 of the genera are endemic to the West Indies. Schwartz ( 1978 ) gave a brief description of the herpetogeography of the West Indies. Twenty-two genera of reptiles are primar- ily mainland taxa having one or two species extending into the West Indies. Fifteen of these are South American taxa that extend into the Lesser Antilles — Phyllodactylus (also Greater Antilles), Thecadactylus, Iguana, Ba- chia, Cnemidophorus, Gymnophthabnus, Kcn- tropyx, Mabuya (also Greater Antilles), Boa, CoraUus, Chironius, Clelia, Mastigodryas, Pseudoboa, and Bothrops. Five are Central American taxa that extend into the Greater Antilles — Gonatodes, Tretanorhinus, and Ctenosaura, Boa, and Coniophanes only reach- ing Isla San Andres and/or Isla Providencia. Three are North American taxa that reach Cuba — Natrix fasciata, Kinosternon bauri, and Crocodylus acutus; the latter also has in- vaded the Lesser Antilles from South Amer- ica, and there is an endemic species of Croco- dylus on Cuba. The geckos Tarentola and Hemidactylus may have arrived by waifing from any one of many sources, although the other species of Tarentola are circum-Medi- terranean. Two of the Leptodactylus and one each of Eleuthcrodactylus and Hyla are main- land species. The five west Indian Chrysemys probably stem from an invasion from North America. Among the endemic or taxonomically rich genera in the West Indies, the hylid genera (Calyptaliyla, Osteopilas, and Hyla) were studied by Trueb and Tyler (1974), who in- ferred five invasions of the Greater Antilles 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 19 by separate hylid stocks, probably from South America by rafting; however, it is possible that the stocks for some of these were on the proto-Antilles and drifted part of the way to their present positions. The same might be true for the monotypic Cuban eleutherodac- tyline Sminthillus and some of the West In- dian stocks of Eleutherodactylus. Lynch (1971) suggested that most of the West In- dian Eleutherodactylus, plus Sminthillus and the Mexican Syrrhophus and Tomodactylus possibly represented one eleutherodactyline lineage and that the Eleutherodactylus inop- tatus group of Hispaniola and the mainland Eleutherodactylus formed another lineage. If these suppositions are correct, minimally two eleutherodactyline invasions of the West In- dies are required. The nine Greater Antillean Bufo seem to be related (Schwartz, 1972), but their affinities with mainland taxa have yet to be determined. Amongst the lizards, the dominant genus is Anolis, represented by two groups of spe- cies (alpha and beta, fide Etheridge, 1959), plus two endemic genera (Chamaeleolis and Chamaelinorops in Cuba and Hispaniola, re- spectively). The alpha anoles inhabit the Greater and Lesser Antilles and are wide- spread in South America, whereas the beta anoles occur in Central and South America and the Greater Antilles. Williams ( 1969 and in Trueb and Tyler, 1974 ) required minimally two invasions of the Greater Antilles by Anolis and two for the endemic genera. Cyclura is related to Ctenosaura of Middle America (Avery and Tanner, 1971); the ancestral stock of Cyclura presumably arrived in the Greater Antilles from Central America. This also probably is true for the ancestral xantusiid stock that gave rise to Cricosaura endemic to Cuba (Savage, 1963) and that of Diploglossus represented by some Central American and many West Indian species. Possibly an earlier or separate invasion was responsible for the endemic Hispaniolan anguid Wetnwrena. Two of the speciose Antillean genera seem to be of South American origin — Ameiva and Leiocephalus. Etheridge ( 1966 ) showed Leiocephalus to be a tropidurine related to Liolacmus (restricted to temperate South America). Sphaerodactylus, with 56 species in the West Indies, may have evolved there from an early sphaerodactyline invasion; if so, the few mainland species (Central America and Choco) are the result of dispersal of stocks back to the mainland. All of the colubrid snakes ( save the North American Natrix fasciata) in the West Indies are xenodontines. Maglio (1970) demon- strated relationships of the seven endemic genera and Alsophis with diverse mainland xenodontines and concluded that four sepa- rate xenodontine invasions of the West Indies from either Central or South America were necessary in the evolution of the West Indian xenodontines. Presumably the tropidophiid stock that gave rise to the 12 species of Tropi- dophis in the Greater Antilles and the Ba- hamas came from South America, perhaps via Central America. Too little is known about the relationships of the Antillean Aristeltiger, amphisbaenids, Leptotyphlops, and Typhlops to speculate on their origins, except that it is unlikely that they invaded from North America. CONTINENTAL PATTERNS OF DISTRIBUTION It is becoming increasingly evident that the patterns of climate and vegetation have changed drastically in South America since the beginning of the Cretaceous. Axelrod ( 1972 ) argued convincingly that the interior of the large African-American continent was arid prior to the birth of the South Atlantic Ocean, which brought maritime and mesic climates to western Africa and eastern South America for the first time. Some elements of the arid-adapted west Gondwanan flora sur- vived in South America (and Africa) (Sol- brig, 1976; Sarmiento, 1976), while much of the continent was mesic. Subsequent to the Eocene, temperate South America gradually became cooler and drier ( Axelrod and Bailey, 1969; Wolfe, 1971; Baez and Scillato Yane, this volume). The uplift of the Andes begin- ning in the Miocene drastically modified wind patterns and resulted in great changes in cli- mate and vegetation (Simpson, this volume), and the formerly widespread Tertiary-Chaco Paleoflora (Solbrig, 1976) became fragmented on the Pacific slopes as the climatic effects of 20 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 the developing Humboldt Current resulted in desiccation of the land in the Late Tertiary (Jeannel, 1967). Pleistocene climatic fluctua- tion effected the entire continent with cool and warm periods in the south (Baez and Scillato Yane, this volume), humid and dry periods in the lowland tropics (Haffer, this volume ) and extensive glaciation in the Andes (Simpson, this volume). Although the fossil record of the herpeto- fauna in South America is still fragmentary, sufficient material exists, especially when placed with the better data from mammals and the paleofloras, to give a faint impression of past distributions, especially in the southern part of the continent (Baez and Gasparini, this volume). It is evident that there has been a northward retreat of the tropical biota, espe- cially those types requiring mesic environ- ments. Conceivably much of the present arid- adapted temperate heqietofauna has evolved rather recently in response to increasingly xeric conditions, as postulated for xeric floras by Axelrod ( 1967 ) . Thus, the archaic frogs in the austral forests are relicts, like the forests themselves ( Vuilleumier, 1968; Lynch, 1971). The fossil record is especially secretive about the presently large and diverse herpeto- fauna of the tropical forests. Presumably most of this fauna evolved at the generic level by the mid-Tertiary, or at least by the Pliocene. Endemic Andean groups apparently evolved with the uplift of the Andes and probably are not older than the Pliocene. Speciation in many lowland tropical groups (Haffer, 1974, this volume) and Andean groups (Simpson, 1975, this volume; Duellman, this volume) seems to have occurred in the Pleistocene. Thus, we are faced with contrasting pic- tures— presumed recent speciation and appar- ently rapid evolutionary rates in the lowland tropics and in the Andes, as well as in some temperate groups adapted to xeric conditions, versus the survival of many old taxa in habitat refugia in the austral forests and also in the ancient Brasilian and Guianan highlands (Hoogmoed, this volume). Various contributors to this volume have analyzed distributions within certain regions (e.g., Patagonia) or biotopes (e.g., lowland tropical rainforests); here I attempt to provide a broad synthesis of patterns in the entire continent. Data for many groups and/or re- gions are inadequate for a detailed analysis; instead I present a general picture of the di- verse distribution patterns and give examples of each. Temperate Herpetofaunas Austral Forests. — The cool, moist forests of southern Chile and adjacent Argentina repre- sent an unique biotope in South America, characterized by a highly endemic herpeto- fauna composed mostly of primitive lepto- dactylid frogs ( Alsodes, Batrachyla, Caudi- verbera, Eupsophus, Hylorina, Tehnatobufo) mostly restricted to forests south of 37°S Lat (Fig. 1:2A). The distributions of all of the species are mapped by Formas ( this volume ) . The herpetofauna of the austral forests mostly is relictual and presumably consists (at least in amphibians) of remnants of groups that were widespread in temperate South America in the Early Tertiary. With few minor excep- tions (Bufo spinidosus, Tachymenis peruvi- ana), none of the species extends beyond the present limits of the region, but some species of Liolaemus and Pleurodema have congeners in adjacent regions. The Atacama Desert to the north and the Patagonian steppe to the east are effective barriers to the dispersal of groups inhabiting the austral forests. Patagonian Steppe. — The cool, dry steppes of southern Argentina interdigitate in the north with the monte ( Cei, this volume ) . The Patagonian herpetofauna contains some an- cient relicts ( telmatobiine leptodactylid frogs and some tropidurine iguanid lizards) but also many species of Liolaemus that have dif- ferentiated in the Pleistocene (Cei, this vol- ume). Some Patagonian groups have relatives in the adjacent monte and the pampas, or in the austral forests, but the major latitudinal expansion has been northward in the Andes, best exemplified by Liolaemus (Fig. 1:2B). Tropical and Subtropical Herpetofaunas Herein distinction is made between two primary biotopes, as follow: 1) Forests — tropical evergreen forests, including rainforest and cloud forest, and 2) Nonforests — the de- ciduous, scrub or thorn forests, and savannas, grasslands and deserts. Although each of these categories, especially the latter, contains diverse vegetation formations, they seem to have reality with respect to major patterns of distribution of the herpetofauna. 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 21 Fig. 1:2. Distribution patterns of the South American herpetofauna: A. Batrachyla in austral forests (black); Centrolcnclla principally in cloud forests but also entering lowland tropical rainforest (stippled). B. Liolac- mus, an austral group entering adjacent subtropical areas and extending northward in the Andes (stippled); Pleurodema brachijops, widespread in nonforests in northern South America (black). C. Hyla parviceps group with vicariant species in lowland rainforests; three species are in the Amazon Basin and one each in the other areas (stippled); Phyllopezus with disjunct populations in nonforested areas of the caatinga, cerrados, and pampas (black) (after Vanzolini, 1974). D. Osteocephalus with species on Andean and Guianan slopes and others in lowland rainforests. E. Tropidurus with species inhabiting diverse nonforested environments through- out tropical South America and the Galapagos Islands. F. Pleurodema, a temperate South American genus with vicariant species in the Andes and in nonforested environments to Panama. Patrones de distribution de la herpetofauna sudamericana. A. Batrachyla en los bosques australes (negro); Centrolenella principahncnte en bosques neblinos pero tambien entra las tierras bajas de la selva lluviosa tropical (punteado). B. Liolaemus, un grupo austral entra los areas subtropicales adyacentes y se extende hacia el norte en los Andes (punteado); Pleurodema brachyops diseminado en ambientes no forestales en el norte de Sud- america (negro). C. Hyla parviceps con especies vicarias en las tierras bajas del bosque lluvioso, tres especies en la Amazonia y una especie en cada una de los otros areas (punteado); Phyllopezus poblaciones disjuntas en areas sin bosque de caatinga, cerrado y pampas (segun Vanzolini, 1974) (negro). D. Osteocephalus con espe- cies en las laderas andinas y guianense y otras especies en las tierras bajas del bosque pluvial. E. Tropidurus con especies habitando diversos ambientes no forestales atraves de Sudamerica tropical y las Islas Galapagos. F. Pleurodema, un genero de la region templada de Sudamerica con especies vicarias en los Andes y en ambi- entes no forestales hasta Panama. 22 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Forests. — The vast Amazonian rainforests and the smaller areas of rainforest in the Choco and along the southeastern coast of Brasil contain the richest herpetofaunas in South America (Lynch, this volume; Dixon, this volume). Especially diverse in these for- ests are dendrobatid and hylid frogs, anoline and teiid lizards, and xenodontine colubrid snakes. The herpetofaunas of the montane rainforests or cloud forests in the Andes, Gui- ana Highlands, and the Brasilian Highlands are primarily altitudinal extensions of the lowland groups (Hoogmoed, this volume; Duellman, this volume). However, in the montane forests, certain groups are either en- demic or far more diverse than in the low- lands— frogs of the families Centrolenidae (Fig. 1:2A), Dendrobatidae (Colostethus), and Leptodactylidae (Eleutherodactylus) and salamanders of the genus Bolitoglossa (Andes only). Distribution patterns are highly variable (see Duellman, 1978, Fig. 197, for examples of Amazonian distributions). A few wide- spread species, such as Boa constrictor, in- habit all of the lowland forests, but these species usually also inhabit intervening non- forest areas. More commonly, vicariant spe- cies occur in the different areas of rainforest (Fig. 1:2C). Widespread and speciose gen- era, such as Eleutherodactylus, Hyla, and An- olis, are found throughout the lowland and montane forests, but usually there are distinct combinations of species at different elevations, as shown for Eleutherodactylus by Lynch and Duellman (1979). These patterns are more readily discernible in smaller genera, such as Osteocephalus (3 Amazonian species, 1 Guianan, 2 Andean, and 1 coastal Brasilian; Fig. 1:2D) or Enyalioicles (2 Amazonian spe- cies, 2 Chocoan, and 3 Andean). The differentiation of populations in Qua- ternary forest refugia (Haffer, 1969, 1974) has been postulated for lizards (Vanzolini and Williams, 1970), frogs (Duellman, 1972; Heyer, 1973; Duellman and Crump, 1974) and snakes (Dixon, this volume). Nonforests. — The tropical and subtropical nonforested biotopes are more extensive, di- verse and fragmented than the forests. In northern South America are the coastal des- erts, savannas, and the extensive llanos; south and east of the Amazon Basin are the dry areas of northeastern Brasil (Caatinga), the interior savannas (cerrados), and the Gran Chaco. Subtropical, cis-Andean, nonforests include the pampas, espinal, and monte in Argentina; west of the Andes are the matorral and the Atacama Desert (Fig. 1:1). Distributions of many species of plants and animals indicate that various combina- tions of these nonforest environments were continuous with one another in the not-too- distant past. Pleistocene climatic fluctuations resulted in drier periods ( interglacials ) that allowed for expansion of the nonforests ( Haf- fer, 1974; Vanzolini, 1976). Gallardo (1969, 1971, this volume) emphasized the faunal re- lationships among the chaco, pampas and monte, and Vanzolini (1968, 1974, 1976) demonstrated distribution patterns in the cer- rados and caatinga. The herpetofaunal rela- tionships among the coastal deserts, llanos and savannas of northeastern South America are analyzed by Rivero-Blanco and Dixon (this volume) and Hoogmoed (this volume). The trans-Andean area consists of a nar- row coastal strip and Andean slopes to 2000- 3000 m, 1-37°S Lat. The dry upper Maranon Valley and associated valleys in the Huanca- bamba Depression are separated from the trans-Andean arid zone by passes at less than 3000 m. The coastal deserts and matorral have a small, but largely endemic, herpeto- fauna including three endemic genera of liz- ards— Garthia (2 species), Callopistes (2) and Dicrodon (3). The dominant groups are two genera of lizards, Tropidurus ( Dixon and Wright, 1975), and Phylhdactylus (Dixon and Huey, 1970), both of which have repre- sentatives in the dry valleys east of the Andes and on the Galapagos Islands. For much of the length of the coastal desert in northern Chile and southern Peni, the entire herpeto- fauna is composed of solely two species of Tropidurus and one of Phyllodactylus. The Humboldt Current sweeps the coast of Chile and Peru and swings westward past the Galapagos Islands, 600 km off the coast of Ecuador. This current has been important in rafting stocks of Atacaman reptiles — Tropi- durus, Pkyllodactylus, Alsophis (=Dromi- cus) — to the Galapagos. 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 23 Several patterns of distribution are evi- dent in the nonforested regions. Some species, such as Pleitrodcma brachyops and Phimophis guianensis, are widespread in the northern regions (Fig. 1:2B), whereas others are re- stricted to the coastal deserts or llanos. Some ceratophryine frogs, some snakes, and several lizards have series of populations, subspecies or species distributed from the caatinga south- westward in the cerrados to the chaco or even into the monte or pampas (Fig. 1:2C). Such open habitats obviously were continuous, or nearly so, in the past so as to permit the dis- persal of such nonforest taxa as Tropidurus, Cnemidophorus, and Bufo granulosus (Webb, 1978 ) . Some of the taxa in the Atacama Des- ert are not represented east of the Andes, except in the upper Mar anon Valley, but Phyllodactylus also is diverse in northern South America, Middle America, and the West Indies. Tropidurus is widely distributed in tropical nonforested environments on both sides of the Andes (Fig. 1:2E). At least one temperate group (Pleurodema) has dispersed northward in nonforests to northern South America (Duellman and Veloso, 1977) (Fig. 1:2F). Montane Herpetofaunas The three major highland regions of South America — the ancient Guianan and Brasilian highlands and the young Andes — have little in common herpetologically. With few excep- tions, there are no isolated sister groups at high elevations that do not have relatives at low elevations. Hylid frogs of the genera Cryptobatra chits (northern Andes) and Ste- fania (Guiana Highlands) and teiid lizards, Euspondylus in the same regions, are primary examples. Hylid frogs of the genera Gastro- theca and Flectonotus occur in the Andes and in the Brasilian Highlands, but some of these species occur at low to moderate elevations, even though at present none lives in the inter- vening lowlands. The herpetofaunas of the highland regions seem to have been derived independently from the adjacent lowlands. In the case of the Andes, the fauna is composed of a southern assemblage derived from Pata- gonia and a northern assemblage derived from the tropical lowlands (Duellman, this vol- ume). FUTURE OF THE HERPETOFAUNA I have attempted to interpret the past and to describe the present; now I provide a prog- nosis for the study of the South American herpetofauna. South America has the richest herpetofauna of any continent, but the fauna is still poorly known taxonomically. Our knowledge of systematic and ecological rela- tionships is even less. Human devastation of vast areas of forest that a few years ago were unexplored is eliminating forever important, and in many cases unknown, aspects of the biota. Although biologists have had some in- fluence on the control of this exploitation, there is little hope that we can preserve all that we may wish to save. Thus, we are faced with two courses of action — salvage collecting and preservation of diverse natural preserves. Random collecting of the biota, even in reasonably well known areas, commonly re- sults in new information on distributions, tax- onomy, or life histories. However, collecting efforts need to be intensified and planned to sample biota before they are destroyed. In the case of amphibians and reptiles, efforts must be made to obtain not only series of well- preserved specimens but also tissues for karyo- logical and biochemical studies, colored pho- tographs, tape recordings of calls (of frogs), life history data, and extensive notes on habi- tats and behavior. We cannot necessarily ex- pect that a visit to the same region five or ten years hence, or even next year, will permit the collection of these data. The collection of these kinds of materials and data must be encouraged and supported at every level inter- nationally. The impending biological crisis has no national boundaries; responsible and effective collectors should be encouraged to make adequate collections throughout the continent. Systematic collections under re- sponsible direction of trained biologists will be one of the most important biological re- sources of the future; materials in these col- lections made available internationally to qualified investigators will be the basis for not only systematic studies but much evolutionary synthesis. The establishment of large natural reserves in areas of high species richness and ende- mism can preserve large segments of the her- 24 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 petofauna. However, such reserves should not be established strictly for conservation pur- poses. It is in these reserves that biologists can undertake long-range studies of communi- ties, population dynamics, behavior, and life history. The resulting kinds of information complement those obtained from salvage col- lecting and contribute substantially to our total understanding of the biota. Actions of these kinds are necessary now; a few years hence will be too late for some of the regions and their herpetofaunas. Without such actions the papers assembled in this vol- ume will not be the preliminary assessments as intended, but instead they might be the last word on the South American herpeto- fauna. ACKNOWLEDGMENTS In the preparation of this paper I have drawn freely on the manuscripts submitted by other contributors to this volume. I am in- debted to Richard Etheridge and John D. Lynch for some of the data and to Juan Man- uel Renjifo for translating the summary. An earlier draft of the manuscript benefited from critical review by John D. Lynch, Gregory K. Pregill, Linda Trueb, and Margaret Davies, whose austral invectives in the pits decidedly influenced the effectiveness of my writing. RESUMEN La herpetofauna sudamericana se corn- pone de 1,095 especies de anfibios distribuidos en 115 generos y 15 familias, 1,115 especies de reptiles en 203 generos y 22 familias (exclu- yendo los taxa marinos). De los 318 generos y 2,210 especies, 201 generos y 2,017 especies son endemicas de este continente. Entre las familias de reptiles, no encontramos ninguna endemica en sudamerica; en cambio existen cinco familias endemicas de anfibios. Durante 4.5-50 millones de aiios la fauna sudamericana evoluciono aislada del resto de los continentes formaban Gondwanalandia, solo hasta la relativamente reciente conexion (5.7 millones de aiios) con Norte America por la via del Lstmo de Panama. Antes del estab- lecimiento de la conexion terrestre, un archi- pielago proveo las veces de filtro entre Norte y Sur America, el cual fue cruzado en ambas direcciones por algunos grupos. El mayor in- tercambio entre las dos faunas se llevo a cabo una vez fue establecida la conexion entre los dos continentes. Seis familias de anfibios y 15 de reptiles son compartidas por Norte y Sur America. Ademas cuatro familias de anfibios y cuatro familias de reptiles sudamericanas tambien se encuentran en Centro America. Gran parte de la herpetofauna Antillana se compone de grupos neotropicales, algunos de los cuales invadieron las islas, especialmente las islas menores de las Antillas provenientes de Sudamerica; otros grupos invadieron las islas desde Centro America. De las 37 familias sudamericanas, tres de anfibios y seis de reptiles son compartidas con Africa y Australia. Un total de cinco familias de anfibios son compartidas con Africa y cu- atro familias con Australia. Entre las familias de reptiles, 11 son comunes con Africa y siete con Australia. De este modo, las relaciones a nivel de familias entre las herpetofaunas son mayores entre Sudamerica y Africa que entre Sudamerica y Australia. La mayor semejanza existe entre Australia y Africa. Comparado con los otros continentes que formaban Gond- wanalandia, Sudamerica tiene un numero des- proporcionadamente alto de anuros y Aus- tralia de saurios. La presencia o ausencia de las familias en los tres continentes se debe principalmente a factores historicos, mientras que la diversidad dentro de las familias depende del tiempo que estas hayan estado en el continente, el tamano del area de este y su diversidad ecologica. En Sudamerica la herpetofauna evoluci- ono respondiendo a los cambios de las condi- ciones climaticas durante el Cenozoico; apar- entemente muchas de las especies que existen actualmente evolucionaron en el Quaternario. Las herpetofaunas de las regiones templadas incluyen aquellas encontradas en los bosques australes y las estepas patagonicas. Los bosques estan restringidos a zonas aisladas, mientras que las estepas se han dispersado hasta el monte subtropical adyacente y hacia el norte en los Andes. Las herpetofaunas tropicales y subtropi- cales incluyen aquellas asociadas con bosques tropicales siempre verdes y bosques montano- 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 25 sos y agregaciones asociadas con ambientes sin bosqucs grandes — pampas, monte, espinal, chaco, matorral, cerrado, caatinga, sabanas, llanos y desiertos. Cada una de estas regiones tiene su fauna caracteristica. Las herpeto- faunas de las tierras altas en los Andes, en Guiana y en Rrasil tienen poco en coniun; aparentemente cada una se derivo independi- entemente de las faunas encontradas en las tierras bajas adyacentes. Con el fin de adquirir material para estu- dios en el futuro y conservar agregaciones de la herpetofauna sudamericana antes de que la destrucion causada por el hombre elimine aun mas la biota, es necesario hacer extensas y bien documentadas colecciones de la herpeto- fauna a lo largo del continente asi como establecer grandes reservas naturales en areas ricas en especies y endemismos donde estudios biologicos de la fauna puedan ser llevados a cabo en el futuro. LITERATURE CITED Anderson, S., Jones, J. K., Jr. 1967. Recent mam- mals of the world: A synopsis of families. Ronald Press, New York, 453 p. Auffenberg, W. 1971. A new fossil tortoise, with remarks on the origin of South American testu- dines. Copeia 1971 ( 1 ) : 106-117. Avery, D. F., Tanner, W. W. 1971. Evolution of the iguanine lizards ( Sauria, Iguanidae ) as deter- mined by osteological and myological characters. Brigham Young Univ. Sci. Bull. Biol. Ser. 12(3): 1-79. Axelrod, D. I. 1967. Drought, diastrophism, and quantum evolution. Evolution 21:201-209. Axelrod, D. I. 1972. Edaphic aridity as a factor in angiosperm evolution. Amer. Nat. 106:311-320. Axelrod, D. I., Bailey, H. P. 1969. Paleotempera- ture analysis of Tertiary floras. Paleogeogr. Paleo- climatol. Paleoecol. 6:163-195. Berman, D. S. 1973. Spathorhynchus fossorium, a middle Eocene amphisbaenian ( Reptilia ) from Wyoming. Copeia 1973 (4):704-721. Blair, W. F. 1972. Summary, pp. 329-343 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Bons, J., Pasteur, G. 1977. Solution histologique a un probleme de taxonomie herpetologique interes- sant les rapports paleobiologiques de I'Amerique du Sud el de l'Afrique. Acad. Sci. Paris, Ser. D 284:2547-2550. Brodkorb, P. 1971. Origin and evolution of birds, pp. 20-55 in Farner, D. S., King, J. R. (eds.). Avian biology, I. Academic Press, New York, 586 p. Burcer, W. L. 1971. Genera of pitvipers (Serpen- tes: Crotalidae). PhD Dissert. Univ. Kansas, 186 p. Cei, J. M. 1968. Remarks on the geographical dis- tribution and phyletic trends of South American toads. Pearce-Sellards Ser. Texas Mem. Mus. (13): 1-20. Cei, J. M. 1972. Bufo of South America, pp. 82-92 in Blair, W. F. (ed. ). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Cogger, H. G. 1974. Voyage of the banded iguana. Australian Nat. Hist. 18:144-149. Cogger, H. G. 1975. Reptiles and amphibians of Australia. Reed, Sydney, 584 p. Cracraft, J. 1973. Continental drift, paleoclimatol- ogy, and evolution and biogeography of birds. J. Zool. 179:455-545. Cracraft, J. 1974. Continental drift and vertebrate distribution. Ann. Rev. Ecol. Syst. 5:215-261. Dietz, R. S., Holden, J. C. 1970. Reconstruction of Pangaea: Breakup and dispersion of continents, Permian to present. J. Geophys. Res. 75:4939- 4956. Dixon, J. R., Anderson, S. C. 1973. A new species and genus of gecko (Sauria: Gekkonidae) from Iran and Iraq. Bull. South. California Acad. Sci. 72:155-160. Dlxon, J. R.; Huey, R. B. 1970. Systematics of the lizards of the gekkonid genus Phyllodactylus of mainland South America. Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (192):l-78. Dlxon, J. R., Wright, J. W. 1975. A review of the lizards of die iguanid genus Troj)idurus in Peru. Ibid. (271): 1-39. Dowling, H. G. 1975. A provisional classification of snakes. 1974 Yearbook Herpetology. Herpetol. Infor. Search Syst., New York, pp. 167-170. Duellman, W. E. 1966. The Central American herpetofauna: An ecological perspective. Copeia 1966(4):700-719. Duellman, W. E. 1970. The hylid frogs of Middle America. Univ. Kansas Mus. Nat. Hist. Monogr. (l):l-753. Duellman, W. E. 1972. South American frogs of the Hyla rostrata group (Amphibia, Anura, Hy- lidae). Zool. Meded. Rijks Mus. Nat. Hist. Leiden 47:177-192. Duellman, W. E. 1975. On the classification of frogs. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (42): 1-14. Duellman, W. E. 1977. Liste der rezenten Amphi- bien und Reptilien: Hylidae, Centrolenidae, Pseu- didae. Das Tierreich (95) : 1-225. Duellman, W. E., Crump, M. L. 1974. Speciation in frogs of the Hyla parviceps group in the upper Amazon Basin. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (23): 1-10. Duellman, W. E., Veloso M., A. 1977. The phy- logeny of Pleurodema (Anura: Leptodactylidae ) : A biogeographic model. Ibid. (64): 1-46. Dunn, E. R. 1931. The herpetological fauna of the Americas. Copeia 1931 (3): 106-119. 26 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Estes, R. 1970. Origin of the Recent North Amer- ican lower vertebrate fauna: An inquiry into the fossil record. Forma Functio 3:139-163. Estes, R. 1975. Fossil Xenopus from the Paleocene of South America and the zoogeography of pipid frogs. Herpetologica 31:263-278. Estes, R. 1976. Middle Paleocene lower vertebrates from the Tongue Formation, southeastern Mon- tana. J. Paleontol. 50:500-520. Estes, R., Price, L. I. 1973. Iguanid lizard from the Upper Cretaceous of Rrasil. Science 180:744- 751. Estes, R. Reig, O. A. 1973. The early fossil record of frogs: A review of the evidence, pp. 11-63 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Estes, R., Wake, M. 1972. The first fossil record of caecilian amphibians. Nature 239:228-231. Etheridge, R. 1959. The relationships of the anoles (Reptilia: Sauna: Iguanidae): An interpretation based on skeletal morphology. PhD Dissert. Univ. Michigan, 236 p. Etheridge, R. 1964. The skeletal morphology and systematic relationships of sceloporine lizards. Copeia 1964(4):610-631. Etheridge, R. 1966. The systematic relationships of West Indian and South American lizards referred to the iguanid genus Leiocephalus. Ibid. 1966 (1):79-91. Etheridge, R. 1967. Lizard caudal vertebrae. Ibid. 1967(4) :699-721. Formas, R., Pugin, E., Jorquera, R. 1975. La iden- bdad del batracio chileno Heminectes rufus Phi- lippi, 1902. Physis 34:147-157. Gaffnev, E. S. 1975. A phylogeny and classifica- tion of the higher categories of turtles. Bull. Amer. Mus. Nat. Hist. 155:387-436. Gaffney, E. S. 1977. The side-necked turtle family Chelidae: A theory of relationships using shared derived characters. Amer. Mus. Novit. (2620):1- 28. Gallardo, J. M. 1961. On the species of Pseudidae. Bull. Mus. Comp. Zool. Harvard Univ. 125:108- 134. Gallardo, J. M. 1969. Las especies de saurios (Reptilia) de la Provincia de Santa Fe, Argen- tina, y consideraciones sobre su ecologia y zoo- geografia. Neotropica 15:73-81. Gallardo, J. M. 1971. Composicion faunistica de los saurios de la Provincia de La Pampa, Repvib- lica Argentina. Ibid. 17:44—48. Gans, C. 1967. A check list of Recent amphisbae- nians ( Amphisbaenia, Reptilia). Bull. Amer. Mus. Nat. Hist. 135:61-105. Grant, N. K. 1971. South Atlantic, Benue Trough, and Gulf of Guinea Cretaceous triple junction. Geol. Soc. Amer. Bull. 82:229.5-2298. Greer, A. E. 1970. A subfamilial classification of scincid lizards. Bull. Mus. Comp. Zool. Harvard Univ. 139:151-183. Greer, A. E. 1974. The generic relationships of the scincid lizard genus Leiolopisma and its relatives. Australian J. Zool. Suppl. Ser. (31): 1-67. Haffer, J. 1969. Speciation in Amazonian forest birds. Science 165:131-137. Haffer, J. 1974. Avian speciation in tropical South America. Publ. Nuttall Ornith. Club (14): 1-390. Hagen, V. W. vox. 1948. The green world of the naturalists. Greenberg, New York, 392 p. Hershkovitz, P. 1972. The Recent mammals of the Neotropical Region: A zoogeographic and eco- logical review, pp. 311-^131 in Keast, A., Erx, F. C., Glass, B. (eds. ). Evolution, mammals, and southern continents. State Univ. New York Press, Albany, 543 p. Heyer, W. R. 1973. Systematics of the marmoratus group of the frog genus Leptodactylus. Nat. Hist. Mus. Los Angeles Cty. Contr. Sci. (251): 1-50. Heyer, W. R. 1975. A preliminary analysis of the intergeneric relationships of the frog family Lep- todactylidae. Smithsonian Contrib. Zool. (199): 1-55. Heyer, W. R., Liem, D. S. 1976. Analysis of the intergeneric relationships of the Australian frog family Myobatrachidae. Ibid. (233): 1-29. Holdex, J. C, Dietz, R. S. 1972. Galapagos core, NazCoPac triple junction and Carnegie/Cocos ridges. Nature 235:266-269. Holmax, J. A. 1967. Additional Miocene anurans from Florida. Q. J. Florida Acad. Sci. 30:121- 140. Holmax, J. A. 1968. Lower Oligocene amphibians from Saskatchewan. Ibid. 31:273-289. Holman, J. A. 1976. Snakes from the Rosebud For- mation (Middle Miocene) of South Dakota. Her- petologica 32:41-48. Holmax, J. A. 1977. Upper Miocene snakes (Rep- tilia, Serpentes) from southeastern Nebraska. J. Herpetol. 11:323-335. Izecksohx, E. 1971. Novo genero e nova especie de Brachycephalidae do Estado do Rio de Janeiro, Brasil (Amphibia, Anura). Bol. Mus. Nac. Rio de Janeiro, Zool. (280): 1-12. Jaxzex, D. H. 1976. The depression of reptile bio- mass by large herbivores. Amer. Nat. 110:371- 400. Jeannel, R. 1967. Biogeographie de l'Amerique australe, pp. 401-460 in Delamere Deboutte- ville, C, Rapoport, E. (eds.). Biologie de l'Amerique australe, 3. C. N. R. S. Groupe Fran- cais Argiles C. R. Reun. Etud., Paris, 834 p. Keast, A. 1972. Continental drift and the evolution of the biota on southern continents, pp. 23-87 in Keast, A., Erk, F. C, Glass, B. (eds.). Evolu- tion, mammals, and southern continents. State Univ. New York Press, Albany, 543 p. Kluge, A. G. 1967. Higher taxonomic categories of gekkonid lizards and their evolution. Bull. Amer. Mus. Nat. Hist. (135): 1-60. Larson, R. L., Ladd, J. W. 1973. Evidence for the opening of the South Atlantic in the Early Creta- ceous. Nature 246:209-212. 1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 27 Lynch, J. D. 1971. Evolutionary relationships, oste- ology, and zoogeography of leptodactyloid frogs. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): 1-238. Lynch, J. D. 1973. The transition from archaic to advanced frogs, pp. 133-182 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contem- porary research on major problems. Univ. Mis- souri Press, Columbia, 470 p. Lynch, J. D. 1976. The species groups of the South American frogs of the genus Eleutlicrodactijlus ( Leptodactylidae). Univ. Kansas Mus. Nat. Hist. Occas. Pap. (61): 1-24. Lynch, J. D. 1978. A re-assessment of the telmato- biine frogs of Patagonia. Ibid. (72): 1-57. Lynch, J. D., Duellman, W. E. 1979. The Elcu- therodactylus of the Amazonian slopes of the Ecuadorian Andes (Anura: Leptodactylidae). Univ. Kansas Mus. Nat. Hist. Misc. Publ. (in press ) . Malfait, B. T„ Dinkelman, M. G. 1972. Circum- Caribbean tectonic and igneous activity and the evolution of the Caribbean Plate. Geol. Soc. Amer. Bull. 82:251-272. Maclio, V. J. 1970. West Indian xenodontine colu- brid snakes: Their probable origin, phylogeny, and zoogeography. Bull. Mus. Comp. Zool. Har- vard Univ. (141): 1-53. Marx, H., Rabb, G. B. 1965. Relationships and zoo- geography of the viperine snakes ( Family Viperi- dae). Fieldiana Zool. 44:161-206. Matthew, W. D. 1915. Climate and evolution. Ann. New York Acad. Sci. 24:171-318. Maxson, L. R. 1976. The phylogenetic status of phyllomedusine frogs ( Hylidae ) as evidenced from immunological studies of their serum albu- mins. Experimentia 32:1149-1150. McDiarmid, R. W. 1971. Comparative morphology and evolution of frogs of the Neotropical genera Atelopus, Dendrophryniscus, Melanophryniscus, and Oreophrynclla. Nat. Hist. Mus. Los Angeles Cty. Sci. Bull. (12): 1-66. McGowran, B. 1973. Rifting and drift of Australia and the migration of animals. Science 180:759- 761. Mertens, R. 1963. Liste der rezenten Amphibien und Reptilien: Helodermatidae, Varanidae, Lan- thanotidae. Das Tierreich (79): 1-26. Mertens, R. 1966. Liste der rezenten Amphibien und Reptilien: Chamaeleonidae. Ibid. (83): 1-37. Meszoely, C. 1970. North American fossil anguid lizards. Bull. Mus. Comp. Zool. Harvard Univ. 139:87-149. Meyer de Schauensee, R. 1964. The birds of Co- lombia. Livingston Publ. Co., Narberth, Pa., 427 p. Moreau, R. E. 1963. Vicissitudes of the African biomes in the late Pleistocene. Proc. Zool. Soc. London 141:395-421. Moreau, R. E. 1969. Climatic changes and distribu- tion of forest vertebrates in West Africa. J. Zool. 158:39-61. Noble, G. K. 1928. Two new fossil Amphibia of zoogeographic importance from the Miocene of Europe. Amer. Mus. Novit. (303): 1-13. Nussbaum, R. A. 1977. Rhinatrematidae: A new family of caecilians (Amphibia: Gymnophiona). Occas. Pap. Mus. Zool. Univ. Michigan (682): 1-30. Patterson, B., Pascual, R. 1968. The fossil mam- mal fauna of South America. Q. Rev. Biol. 43: 409-451. Peters, J. A., Donoso-Barros, R. 1970. Catalogue of the Neotropical Squamata: Part II. Lizards and amphisbaenians. U.S. Natl. Mus. Bull. (297): 1-293. Pianka, E. R. 1971. Lizard species diversity in the Kalahari Desert. Ecology 52:1024-1029. Pianka, E. R., Huey, R. B. 1971. Bird species di- versity in the Kalahari and the Australian deserts. Koedoe 14:123-130. Rage, J.-C. 1975a (1974). Les serpents des Phos- phorites du Quercy. Paleovert. 6:273-303. Rage, J.-C. 1975b. Un caenophidien primirif ( Rep- tilia: Serpentes) dans l'Eocene inferieur. C. R. Somm. Seances Soc. Geol. France 1975:46-47. Rage, J.-C. 1977. La position phyletique de Dini- lysia patagonica, serpent du Cretace superieur. C. R. Acad. Sci. Paris, Ser. D, 284:1765-1768. Raven, P. H., Axelrod, D. I. 1974. Angiosperm bio- geography and past continental movements. Ann. Missouri Bot. Card. 61:539-673. Reyment, R. A., Tait, E. A. 1972. Biostratigraphic dating of the early history of the South Atlantic Ocean. Philos. Trans. Roy. Soc. London 264:55- 95. Richards, P. W. 1973. The tropical rain forest. Sci. Amer. 229:58-67. Rosen, D. E. 1975. A vicariance model of Caribbean biogeography. Syst. Zool. 24:431—464. Sarmiento, G. 1976. Evolution of arid vegetation in tropical America, pp. 65-99 in Goodall, D. W. (ed.). Evolution of desert biota. Univ. Texas Press, Austin, 250 p. Savage, J. M. 1963. Studies on the lizard family Xantusiidae IV. The genera. Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (71): 1-38. Savage, J. M. 1966. The origins and history of the Central American herpetofauna. Copeia 1966(4): 719-766. Savage, J. M. 1968. The dendrobatid frogs of Cen- tral America. Ibid. 1968(4):745-776. Savage, J. M. 1972. The harlequin frogs, genus Atelopus of Costa Rica and western Panama. Herpetologica 28:77-94. Savage, J. M. 1973. The geographic distribution of frogs: Patterns and predictions, pp. 351-445 in Vial, J. L. ( ed. ) . Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Savage, J. M. 1974. The isthmian link and the evo- lution of Neotropical mammals. Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (260):1-51. 28 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Savage, J. M, Carvalho, A. L. de. 1953. The fam- ily position of Neotropical frogs currently referred to the genus Pseudis. Zoologica 38:193-200. Savitzky, A. H. 1978. The origin of the New World proteroglyphous snakes and its bearing on the study of venom delivery in snakes. PhD Dissert. Univ. Kansas, 387 p. Schwartz, A. 1972. The native toads (Anura: Bu- fonidae) of Hispaniola. J. Herpetol. 6:217-231. Schwartz, A. 1978. Some aspects of the herpeto- geography of the West Indies, pp. 31-51 in Gill, F. B. (ed.). Zoogeography in the Caribbean. Acad. Nat. Sci. Philadelphia Spec. Publ. (13): 1-128. Schwartz, A., Thomas, R. 1975. A check-list of West Indian amphibians and reptiles. Carnegie Mus. Nat. Hist. Spec. Publ. (1):1-216. Sill, W. D. 1968. The zoogeography of the croco- dilians. Copeia 1968( 1 ):76-88. Simpson, B. B. 1975. Pleistocene changes in the flora of the high tropical Andes. Paleobiology 1:273-294. Smith, H. M., Martin, R. L., Swain, T. A. 1977. A new genus and two new species of South Amer- ican geckos (Reptilia: Lacertilia). Pap. Avulsos Zool. (Sao Paulo) 30:195-213. Sokol, O. M. 1975. The phylogeny of anuran lar- vae: A new look. Copeia 1975( 1 ) :l-23. Solbric, O. T. 1976. The origin and floristic affini- ties of the South American desert and semi-desert regions, pp. 7—49 in Goodall, D. W. (ed. ). Evo- lution of desert biota. Univ. Texas Press, Austin, 250 p. Starrett, P. H. 1973. Evolutionary patterns in lar- val morphology, pp. 251-271 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contem- porary research on major problems. Univ. Mis- souri Press, Columbia, 470 p. Tihen, J. A. 1962. A review of New World fossil bufonids. Amer. Midi. Nat. 68:1-50. Tihen, J. A. 1964. Tertiary changes in the herpeto- faunas of temperate North America. Sencken- bergiana. Biol. 45:265-279. Tihen, J. A. 1972. The fossil record, pp. 8-13 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Trueb, L. 1971. Phylogenetic relationships of cer- tain Neotropical toads with the description of a new genus (Anura: Bufonidae). Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (216): 1^40. Trueb, L. 1974. Systematic relationships of Neo- tropical horned frogs, genus Hcmiphraetus (Anu- ra: Hylidae). Univ. Kansas Mus. Nat. Hist. Occas. Pap. (29): 1-60. Trueb, L., Tyler, M. J. 1974. Systematics and evo- lution of the Greater Antillean hylid frogs. Ibid. (24): 1-60. Tyler, M. J. 1974. First fossil frogs from Australia. Nature 248:711-712. Tyler, M. J., Davies, M. 1978. Species groups with- in the Australopapuan hylid frog genus Litoria Tschudi. Australian I. Zool Suppl. 63:1—17. Tyler, M. J., Davies, M., King, M. 1978. The Aus- tralian frog Chiroleptes dahlii Boulenger: Its sys- tematic position, morphology, chromosomes and distribution. Trans. Roy. Soc. South Australia 102:17-23. Vanzolini, P. E. 1968. Geography of the South American Gekkonidae (Sauria). Arq. Zool. (Sao Paulo) 17:85-112. Vanzolini, P. E. 1974. Ecological and geographical distribution of lizards in Pemambuco, northeastern Brasil (Sauria). Pap. Avulsos Zool. (Sao Paulo) 28:61-90. Vanzolini, P. E. 1976. On the lizards of a cerrado- caatinga contact: Evolutionary and zoogeographi- cal implications (Sauria). Ibid. 29:111-119. Vanzolini, P. E., Williams, E. E. 1970. South American anoles: The geographic differentiation and evolution of the Anolis chrysolepis species group (Sauria, Iguanidae). Arq. Zool. (Sao Paulo) 19:1-298. Veevers, ]. J., McElhinny, M. W. 1976. The sep- aration of Australia from other continents. Earth Sci. Rev. 12:139-159. Vuilleumier, F. 1968. Origin of frogs of Patagonian forests. Nature 219:87-89. Wake, D. B. 1966. Comparative osteology and evo- lution of the lungless salamanders, family Pletho- dontidae. Mem. South. California Acad. Sci. (4): 1-111. Webb, S. D. 1978. A history of savanna vertebrates in the New World. Part II: South America and the great interchange. Ann. Rev. Ecol. Syst. 9: 393-436. Wermuth, H. 1965. Liste der rezenten Amphibien und Reptilien: Gekkonidae, Pygopodidae, Xantu- siidae. Das Tierreich (80) : 1-246. Wermuth, H. 1967. Liste der rezenten Amphibien und Reptilien: Agamidae. Ibid. (86):1-127. Wermuth, H. 1968. Liste der rezenten Amphibien und Reptilien: Cordylidae. Ibid. (87): 1-30. Williams, E. E. 1969. The ecology of colonization as seen in the zoogeography of anoline lizards on small islands. Q. Rev. Biol. 44:345-389. Wolfe, J. A. 1971. Tertiary climatic fluctuations and methods of analysis of Tertiary floras. Paleo- geogr. Paleoclimatol. Paleoecol. 9:27-57. Wood, R. C. 1971. The fossil Pelomedusidae (Testu- dines, Pleurodira) of Africa. PhD Dissert. Har- vard Univ., 345 p. Wood, R. C, Patterson, B. 1973. A fossil triony- chid turtle from South America. Breviora (415): 1-10. Zweifel, R. G. 1972. Results of the Archbold Ex- peditions No. 97. A revision of the subfamily Asterophryinae, family Microhylidae. Bull. Amer. Mus. Nat. Hist. 148:411-546. 2. The South American Herpetofauna: An Evaluation of the Fossil Record Ana Maria Baez Departamento de Geologia Facultad de Ciencias Exactas y Naturales Univcrsidad de Buenos Aires Buenos Aires, Argentina Zulma B. de Gasparini Facultad de Ciencias Naturales y Museo Universidad Nacional de La Plata La Plata, Argentina The presence of fossil remains of amphib- ians and reptiles related to living taxa in South America has been documented since the last century. Nevertheless, examination of this literature reveals that in many cases the names are only mentioned and the ma- terial has not been studied; in many other cases it is evident that a revision is badly needed. Partial reviews concerning several South American countries have been carried out (Argentina: Pascual, 1970; Pascual and Odre- man Rivas, 1971; Gasparini and Baez, 1975; Brasil: Paula Couto, 1970; Colombia: Hoff- stetter, 1970a; Ecuador: Hoffstetter, 1970b; Peru: Hoffstetter, 1970c; Uruguay: Mones, 1972; 1975). Also, Baez and Gasparini (1977) critically examined the Cenozoic record of amphibians and reptiles in that continent and analyzed distributional shifts, relating them to the Cenozoic environmental changes that oc- curred as a result of different geological events. In this paper, the available paleontological data are summarized in an attempt to evalu- ate the information that the fossil record can provide about the historical development of these groups in South America. The taxo- nomic assignment and the geographic and stratigraphic references are sometimes doubt- ful; questionable referrals are not considered. Some recent finds are currently under study, and identification of genera and species is not yet available; thus, only an analysis at the family level is possible at present. The main areas that have yielded fossil amphibians and reptiles referable to modern groups are shown in figures 2:4-7. A checklist of the material recorded there and the corresponding litera- ture references appear in Appendix 2:1. Inspection of the fossil record reveals that many families comprising the present South American heqoetofauna have a considerable antiquity in that continent. Their presence, and even that of some recent genera, extends back to the late Mesozoic and early Tertiary (Figs. 2:1-3). Thus, it is essential to con- sider the past changes in the geographical position of South America and its connections with other continents, especially from the Middle Mesozoic onward, for those changes must have affected distribution patterns. According to recent paleomagnetic data (Vilas and Valencio, 1978a), South America was part of Gondwanaland, which may have existed up to the Late Jurassic. It was still joined to Africa during the Early Cretaceous ( Reyment and Tait, 1972; Douglas, Moullade and Nairn, 1973; Vilas and Valencio, 1978), while faunal relations with Antarctica-Aus- tralia seem to have been possible up to the Early Tertiary (Raven and Axelrod, 1975). There is no clear evidence that a direct land connection between North and South America existed from the Jurassic to the end of the Tertiary (Malfait and Dinkelman, 1972; McKenna, 1973). Nevertheless, a discontin- uous pathway along a chain of volcanic is- lands could have been established at different times (Haffer, 1970; Malfait and Dinkelman, 1972 ) . There is no agreement concerning the time of final reconnection between both Americas. Some evidence indicates that the Isthmus of Panama was established during 29 30 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 FAMILIAL GROUPS RECENT PLEISTOCENE TERTIARY CRETACEOUS PLIOCENE MIOCENE OLIGOCENE EOCENE PALEOCENE PIP1DAE LEPTODACTYLIDAE — — ? BUFONIDAE HYLIDAE RHINODERMATIDAE PSEUDIDAE CENTROLENIDAE DENDROBATIDAE BRACHYCEPHALIDAE MICROHYLIDAE RANIDAE CAECILIDAE — Fig. 2:1. Chronological range (Cretaceous-Recent) of Anura and Gymnophiona in South America. Distribution cronologica (Cretdcico-Reciente) de Anura y Gymnophiona en America del Sur. the late Pliocene-early Pleistocene ( Patter- son and Pascual, 1968); others favor its earlier existence (Emiliani et al., 1972; Savage, 1974). Thus, it is implied that biogeograph- ical relationships with Africa and Australia may have been close until the end of the Mesozoic, whereas those with North America became more important by the Late Tertiary. Available paleomagnetic data suggest that the latitudinal position of South America did not alter significantly since the latest Paleo- zoic, although displacements of about 5°, at most, may have existed in some areas because of the different orientation of the continent (Vilas and Valencio, 1978, 1979). Thus, these changes are disregarded here because they are insufficient to account for the different distributions of many groups in the past as evidenced by the fossil record. The dispersal history of amphibians and reptiles also has been very closely related to past climates. In South America climatic- ecologic conditions were more equable dur- ing at least the Cretaceous and early Tertiary. Various kinds of evidence indicate that a warm and humid climate prevailed then at high latitudes (Pascual and Odreman Rivas, 1971; Archangelsky and Romero, 1974; Petri- ella and Archangelsky, 1975). Throughout the Cenozoic, geologic events of different magnitude provoked physiographic, and con- sequently, climatic and floristic changes. Al- though these changes restricted the dispersal of some groups, they multiplied the available environments and thereby promoted the ap- pearance of new adaptive types. 1979 BAEZ & GASPARINI: FOSSIL RECORD 31 FAMILIAL GROUPS RECENT PLEISTOCENE TERTIARY CRETACEOUS PLIOCENE MIOCENE OLIGOCENE EOCENE PALEOCENE MEIOLANIIDAE ? PELOMEDUSIDAE - CHELIDAE - TESTUDINIDAE EMYDIDAE P TRIONYCHIDAE — CHELYDRIDAE KINOSTERNIDAE SEBECIDAE CROCODYLIDAE 9 ALLIGATORIDAE GAVIALIDAE NETTOSUCHIDAE Fig. 2:2. Chronological range (Cretaceous-Recent) of Testudines and Crocodilia in South America. Fam- ily groups without Cenozoic records have not been included. Distribution cronologica (Cretdcico-Reciente) de Testudines ij Crocodilia en America del Sur. Las familias sin registros cenozoicos no han sido incluidas. THE FOSSIL RECORD OF MODERN GROUPS OF AMPHIBIANS AND REPTILES IN SOUTH AMERICA The oldest known fossils referable to fam- ily groups that comprise the present South American herpetofauna appear in that conti- nent in Cretaceous deposits. The presence of pipid frogs, iguanid lizards and pelomedusid turtles in the Late Cretaceous is documented by the fossil record. Pelomedusid turtles were quite diversified and widely distributed at that time (Fig. 2:4). It is noteworthy that the extant pelomedusid genus Podocnemis was present then. Leptodactylid frogs perhaps also are present in Cretaceous horizons. The earliest Cenozoic records of amphib- ians and reptiles come almost exclusively from southern South America — Patagonia and southeastern Brasil (Fig. 2:5). The early and late Paleocene Patagonian localities have yielded remains of turtles ( only pelomedusids can be ascertained definitely), eusuchian crocodilians (including crocodylids) and boid snakes. Most of this material is very fragmen- tary, but these records are quite interesting from a paleoclimatic point of view, for they indicate humid subtropical conditions at lati- tudes of about 45°S. On the other hand, the rich and diversified assemblage of late Paleo- cene age of Itaborai, Brasil, provides more comprehensive information about the family groups that were inhabiting South America at that time. However, most of this material has vet to be studied. 32 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 FAMILIAL GROUPS RECENT PLEISTOCENE T E RT 1 A R Y CRETACEOUS PLIOCENE MIOCENE OLIGOCENE EOCENE PALEOCENE IGUANIOAE TEIIDAE GEKKONIDAE ANGUIDAE SCINCIDAE AMPHISBAENIDAE ANILIIDAE BOIIDAE ? ? COLUBRIDAE VIPERIDAE Fig. 2:3. Chronological range (Cretaceous-Recent) of Squamata, exclusive of scolecophidian snakes, in South America. Distribution cronologica (Cretdcica-Reciente) de Squamata, cxcluyendo serpientes scolecofidias, en America del Sur. In the known Paleocene sites, fourteen families have been recorded, all of them (ex- cept the extinct sebecid crocodilians ) are rep- resented in the present-day fauna of South America. They include the Caeciliidae, Pipi- dae, Leptodactylidae, Bufonidae, Hylidae, Pelomedusidae, Crocodylidae, Alligatoridae, Iguanidae, Teiidae, Gekkonidae, Aniliidae, and Boidae. Among these families we can recognize elements that are of diverse histori- cal backgrounds, and that joined the fauna of present-day South America at different times and developed in situ. Presently-known Eocene faunas also were found mainly in the southern part of the con- tinent (Fig. 2:5). Sebecid and alligatorid crocodilians are still recorded at latitudes of about 46°S, thereby indicating that at least warm temperate conditions persisted there. Noteworthy is the appearance of chelid tur- tles, whose extinct and living representatives have been found only in Australia and South America. The Eocene remains have been re- ferred to the extant genus Hydromedusa, which now extends as far south as 36°S. The living leptodactylid genus Caudiverbera is known from Eocene deposits of Patagonia, where it was associated with crocodilian re- mains. The earliest known Cenozoic records of amphibians and reptiles in northernmost South America come mainly from Oligocene sites, especially several localities in the upper and middle Magdalena River Valley, Colom- bia (Fig. 2:6). These assemblages are char- acterized by the abundance of mesosuchian 1979 BAEZ & GASPARINI: FOSSIL RECORD 33 (Sebecidae) and eusuchian (Alligatoridae, Crocodylidae, Gavialidae) crocodilians, which is explained by the presence there of an extensive lowland with local swamps and lakes (Van Houten and Travis, 1968). All of the families recorded from those horizons also are represented in older deposits of more southern latitudes, except gavialid crocodil- ians, whose affinities are still uncertain. Oligocene localities in central Patagonia (Fig. 2:6) have yielded remains of the extant lepodactylid genera Eupsophus and Caudi- verbera, whose presence there was made pos- sible by the prevalence of more mesic condi- tions at that time. The oldest known testu- dinid turtle in South America, a species of the genus Geochelone related to the living G. chilemis (Auffenberg, 1971), is recorded in the upper Oligocene of Patagonia. Very rich, late Miocene faunas are known from the upper Magdalena River Valley, Co- lombia, where the general environmental con- ditions of Oligocene times did not change significantly, although important tectonic events took place in the Miocene (Irving, 1971). Most of the families present there have previous records in the continent, the majority being represented by extant genera or species: Bufo marinus, Podocnemis expan- sa, Tupinambis cf. T. teguixin, Dracaena, Eu- nectes, Caiman. Also a turtle, Geochelone hesterna, closely related to the living G. car- bonaria and G. denticulata (Auffenberg, 1971) was recorded. The appearance of two families, Colubridae and Nettosuchidae, in the fossil record is noteworthy. The latter are endemic eusuchian crocodilians seemingly re- stricted to tropical regions from Miocene to Plio-Pleistocene times. The southernmost records of Cenozoic am- phibians or reptiles on the continent come from early to middle Miocene deposits of southern Patagonia (Fig. 2:6). Iguanids, teiids and other lizards, turtles and snakes have been found there, but all of this ma- terial has yet to be restudied. The post-Mio- cene history of the Patagonian herpetofauna is not documented in the fossil record. The increasing aridity that developed there as a consequence of the Andean tectonic move- ments and the continuous uplift of that area throughout the late Cenozoic made the pres- ervation of remains unlikely. Numerous crocodilian fossils, all referable to the suborder Eusuchia (Alligatoridae, Gav- ialidae, Nettosuchidae and probably Croco- dylidae) are recorded in the upper Pliocene of northern Venezuela (Fig. 2:7). The gigan- tic size exhibited by many members of these taxa and the appearance of the extant alli- gatorid, Melanosuchus, is noteworthy. The possible presence of trionychid turtles there should be noted. This group is not repre- sented in the present fauna of South America, and may have reached the continent from the north. In Pliocene times the Rio Parana consti- tuted, as it does today, an important pathway for the southward migration of elements of the subtropical biota. This explains the rec- ord of alligatorid and gavialid crocodilians in Pliocene deposits near the city of Parana, Argentina (Fig. 2:7). Among them, Caiman latirostris and possibly Rhamphostomopsis were already represented in the late Miocene faunas of Colombia. Numerous remains of amphibians and rep- tiles have been reported from horizons of late Pliocene age in Monte Hermoso, Argentina (Fig. 2:7). The presence there of the living teiid genus Callopistes is especially significant considering its present range, which is on the arid Pacific lowlands of Peru and Chile. The relatively few Pleistocene sites that have yielded amphibians or reptiles are im- portant because they document the past pres- ence of families that currently inhabit South America, and that have not been recorded in older deposits. Among these families are emy- dids, viperids, and amphisbaenids. The Pleis- tocene records also suggest that significant environmental changes occurred at that time, because the distribution of some taxa is incon- sistent with the present distribution of their habitats. For example, the Pleistocene faun- ules from Talara, northwestern Peru (Lemon and Churcher, 1961; Hoffstetter, 1970c) and from the Santa Elena Peninsula, southwestern Ecuador (see Appendix 2:1), indicate that more mesic environments prevailed in those areas in the near past. 34 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 JO» 1 2 3 4 5 6 CRETACEOUS Alemanio Laquna Umayo Mossoro Vilo-Vilo Sao Jose do Rio Preto Lago Colhue Huapi Fig. 2:4. Major areas in South America that have yielded Cretaceous amphibians and/or rep- tiles of modern groups. Principales areas en America del Stir que han brindado anfibios y/o reptiles cretdcicos de grupos modernos. 1979 BAEZ & GASPARINI: FOSSIL RECORD 35 J0° A PALEOCENE • EOCENE i Golfo de San Jorge 2 Itaboroi' 3 Cerro Pan de Azucar 4 Laguna del Hunco 5 Mina Aguilor 6 Canadon Hondo -Cana- do'n Vaca 7 Lago Colhue- Huapi 8 Negritos 9 Divisadero Largo / 20" Fic. 2:5. Major areas in South America that have yielded amphibians and/or reptiles of Paleo- cene and Eocene age. Principales areas en America del Sur que han brindado anfibios y/o reptiles de cdad paleocena y eocena. 36 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 ao° A 0LIG0CENE • MIOCENE i Compo Waldo 2 Scorritt Pocket 3 Tremembe' 4 Colhue' Huopi' 5 Chaparral 6 Gaiman 7 Santa Cruz 8 Coyaima 9 Carmen de Apicola 10 La Venta n Logo Buenos Aires 12 Barranca de los Loros 13 Ingeniero Jacobacci / Fie. 2:6. Major areas in South America that have yielded amphibians and/or reptiles of Oligo- eene and Miocene age. Principales areas en America del Sur que han brindado anfibios tj/o reptiles de edad oligocena y miocena. 19(79 BAEZ & GASPARINI: FOSSIL RECORD 37 ^0° A PLIOCENE • PLEISTOCENE 1 Urumoco 2 V. de Santa Maria 3 Parana' 4 Monte Hermoso 5 Queque'n Salado 6 Chapadmalal 7 Rio Jurud 8 Rio Aguaytia 9 Arroyo Perico Flaco to Santa Elena 11 Tarijo 12 Nuapua i 40° / 20° / Fie. 2:7. Major areas in South America that have yielded amphibians and/or reptiles of Plio- cene and Pleistocene age. Principales areas en America del Stir que han brindado anfibios ij/o reptiles de edad pliocena y pleistocena. 38 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 AMPHIBIANS The fossil record of modern amphibian groups in South America is largely restricted to anurans. The only known caecilian fossil is from the late Paleocene of Brasil; this rec- ord implies that caecilians are old components of the South American herpetofauna. No salamander fossils are known from South America. The considerable diversity of anuran groups already present by the early Tertiary in South America is noteworthy. Further- more, their phylogenetic relationships and the fact that some of them were represented by taxa similar to extant South American genera or species indicate that they have had a long history there. However, many families that comprise the present anuran fauna are un- known as fossils. The aquatic frogs of the family Pipidae now live in South America and Africa; they constitute an ancient, independent lineage. The earliest pipids are from the Early Creta- ceous of Israel (Nevo, 1968), but their origin probably extends back into the Jurassic. Un- questionable pipids occur in the Cretaceous and Tertiary of both Africa and South Amer- ica; this strongly suggests that pipids were components of the Gondwanan fauna. Fur- thermore, remains referable to the African genus Xenopus have been recorded from both continents (Ahl, 1926; Vergnaud-Grazzini, 1966; Broin et al., 1974; Estes, 1975a,b; Baez, 1976). The fossil record indicates that differ- ent phyletic lines of pipids have existed in South America. The oldest known example there is Saltenia ibanezi from the Late Cre- taceous of northwestern Argentina. Members of the extant genus Xenopus have been re- corded in the late Paleocene of Brasil and Ar- gentina. No fossil taxa directly ancestral to the living Neotropical Pipinae are known. The Leptodactylidae are another ancient component of the South American fauna. The oldest known remains unquestionably refer- able to the family come from late Paleocene deposits in Brasil; undescribed taxa close to living genera and perhaps even an extant genus are present there (Estes and Reig, 1973). This group probably is represented in the Late Cretaceous of Peru (Sige, 1968), but that record has not been confirmed. Other early presumed records outside of South America have been discarded or questioned (Hecht, 1963; Estes, 1964, 1969; Reig, 1968; Lynch, 1971; Savage, 1973). Available evi- dence indicates that the Leptodactylidae, in its restricted sense (Lynch, 1973) may have differentiated in South America from a stock associated with temperate forests (Lynch, 1971; Savage, 1973; Heyer, 1975). The pro- posed relationships to leiopelmatids (Heyer, 1975) are significant in that the latter inhab- ited Patagonia by Jurassic times (Estes and Reig, 1973). Leptodactylids must have dis- persed northward early in their history. Wide- spread occurrence of subhumid environments in middle latitudes in the Late Jurassic and Early Cretaceous might have been influential in the development of xeric-adapted types of leptodactylids (Baez and Gasparini, 1977). Telmatobiine leptodactylids, which represent an ancient radiation, are recorded in the Ter- tiary of Patagonia. The extant genus Caudi- verbera appears in early Eocene, early and late Oligocene, and late Miocene deposits. By the early Oligocene two other genera were present — the living Eupsophus and the ex- tinct Neoprocoela; the assignment of Neopro- coela to this family has been discussed by Tihen (1962, 1972) and Lynch (1971). These records furnish evidence of their former wider distribution east of the Andes, and therefore reflect the mesic climate that prevailed in those now arid regions (Gasparini and Baez, 1975). Ceratophryine leptodactylids, which some authors give familial status, are first recorded in the late Miocene of northern Patagonia, but they must have originated much earlier. The basic adaptation of this group to an arid environment ( Heyer, 1975 ) could account for the paucity of paleontologi- cal evidence for its early evolution, because few fossil-bearing horizons representing those environments are known, particularly in the Late Mesozoic and Early Tertiary. Remains referable to the extant genus Ceratophrys (specific allocation undetermined) are known from the late Pliocene and middle Pleistocene of Argentina (Baez and Gasparini, 1977). The living C. ornata probably is represented in the late Pliocene of Argentina (Reig, 1958) 1979 BAEZ & GASPARINI: FOSSIL RECORD 39 and late Pleistocene of Bolivia, and C. aurita is known from the latest Pleistocene in Brasil (Lynch, 1971). Records of other leptodac- tylids are limited to the living genus Lepto- dachjlus from Pleistocene deposits. The earliest known record of the cosmo- politan bufonid toads (absent from Australia and Madagascar) is from the late Paleocene of Brasil. Although this material has not yet been described, the presence of members of living species groups of Bufo has been recog- nized (Estes and Reig, 1973). This supports the proposal that bufonids are ancient com- ponents of the South American fauna and that they could have had a southern origin (Lynch, 1971; Reig, 1972; Savage, 1973). No unquestionable bufonids are known in North America prior to the early Miocene (Tihen, 1972). Moreover, assignment of specimens of Early Tertiary age from Europe to the Bufo- nidae is highly doubtful; remains referable to the genus Bufo are unknown there before the middle Miocene (Tihen, 1972). Different evidence suggests that South America is the most likely area of origin of Bufo (Blair, 1972). Although sparse, the fossil record in- dicates that considerable diversification with- in that genus took place in South America at least since the Early Tertiary. Members of the marinus group of Bufo have been reported as far back as the late Miocene, although an earlier differentiation of that group seems likely. Tihen (1972) and Estes and Reig (1973) considered Neoprocoela to be a mem- ber of the Bufo calamita group, which is now confined to the Old World (also see Gal- lardo, 1962). The Hylidae is poorly represented in the fossil record. The earliest remains referred to this family are from the late Paleocene of Brasil, but the material is still undescribed (Estes and Reig, 1973). This testifies to the early presence of hylid frogs in South Amer- ica, and is consistent with the high degree of differentiation that they attained there during the time of isolation, as well as their possible origin on that continent (Savage, 1973). The now widespread genus Hula has been re- corded from the early Oligocene of Canada (Holman, 196S) and the early Miocene of Florida in the United States (Tihen, 1964). The paleontological data add little informa- tion to the biogeographical history of hylids. Frogs are ancient members of South Amer- ican fauna. Pipids, leptodactylids, bufonids and hylids are old components and are the only frog families represented in the South American fossil record. Other groups such as rhinodermatids and dendrobatids, differenti- ated on that continent probably during the Tertiary. On the other hand, ranids are late immigrants from the north. REPTILES Pelomedusid turtles presently inhabit South America, Africa and Madagascar, but they were more widely distributed in Late Mesozoic and Early Tertiary times (Romer, 1966; Gaffney and Zangerl, 1968; Wood, 1970, 1976b; Jimenez Fuentes, 1971, 1975; Gaffney, 1975). The earliest-known pelomedusids are from Early Cretaceous deposits in Africa (Broin et al., 1974), and remains of Late Cretaceous age have been reported from South America, North America, Africa, and Europe. The greater proximity of continents at that time (Smith, Briden and Drewry, 1973), coupled with the presumed marine habits of some of these turtles (Wood, 1972, 1976b) could have favored such wide distri- bution. The majority of the South American fossil pelomedusids that have been described are referred to the genus Podocnemis, which is still present in South America and Mada- gascar. The presence of that genus in South America extends back to the Late Cretaceous, at which time the genus was represented by a species, P. elegans, noted as being "strikingly modern in aspect" (Wood, 1971:27s).1 Nu- merous fossil remains have been assigned to Podocnemis, but there is no general agree- ment concerning the validity of species (Wood and Gamero, 1971; Baez and Gaspari- ni, 1977). By the late Miocene the extant P. expanse was already in existence. Remains referable to the genus Podocnemis are cited outside of South America, from Africa and western Europe. Pelomedusids referable to other genera have been described from the 1 Dr. F. de Broin (pers. coram.) considers that Po- docnemis elegans may belong to the extinct pelome- dusid genus Roxochehjs, related to Podocnemis. 40 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. Cenozoic of South America. A species of the genus Taphrosphys occurs in the Eocene of coastal Peru. That genus, which includes ma- rine forms, is also recorded from the Creta- ceous of North America and Paleocene of Africa (Wood, 1975). The gigantic Pliocene Stupendemys geographicus exhibits many un- usual characteristics, and its affinities are still not clear (Wood, 1976b). The freshwater turtles of the family Cheli- dae occur today in South America, Australia and New Guinea. The geographical location of fossils assigned to this group indicates that they could have had an essentially similar former range, because the only known un- questionable records come from the early Eo- cene of southern Argentina, middle Tertiary of Australia, and Oligocene or Miocene of Tasmania (Warren, 1969). That evidence suggests an area of dispersal that included also Antarctica and that the past relations between South American and Antarctica-Aus- tralia (McGowran, 1973; Dalziel et al., 1973; Sclater and Fisher, 1974) could account for the present range. A species of the extant South American genus Hydro medusa is the oldest member ( Eocene ) of the family re- corded so far (Wood and Moody, 1976). Two extinct species of the now monotvpie genus Chelus have been described — C. co- lombianus from the late Miocene of Colombia and C. leuisi from the Pliocene of Venezuela (Wood, 1976a). Neither of these seem to have been directly ancestral to the living species, thereby indicating that different line- ages evolved within the genus (Wood, 1976a). Testudinid turtles are represented in South America by the genus Geochelone, with a world-wide distribution in the Early Tertiary. Fossils having similarities with an extant Asi- atic species have been recorded from Eocene deposits in North America and Africa, and also from the early Oligocene of North Amer- ica, Asia and Western Europe (Auffenberg, 1974 ) . The earliest representative of the genus Geochelone in South America is G. gringorum from the late Oligocene of Patagonia (Simp- son, 1942). Auffenberg (1971) suggested that the ancestors of the South American species, comprising the distinctive subgenus Cltclo- noides, probably entered that continent from the north during the Oligocene or even earlier. It is noteworthy that even though chelonians are known since the Late Creta- ceous, no testudinids have been reported be- fore the late Oligocene, whereas their pres- ence is documented frequently since the late Miocene. Assuming an entrance from the north, this partially could result from the fact that most known early Tertiary reptile-bear- ing sites are in the southern part of the conti- nent. However, an arrival earlier than late Eocene times seems improbable. Geochelone gringorum is closely related, and may be an- cestral, to the living G. chilensis, which be- came adapted to drier conditions (Auffenberg, 1971). Increasingly xeric environments devel- oped in western mid-latitudes east of the Andes since the Pleistocene (Baez and Scil- lato Yane, this volume). The related Pliocene G. gallardoi has been recorded in areas where a marked dry season presumably existed. Geochelone hesterna from the late Miocene of Colombia is thought to be ancestral to the extant species G. carhonaria and G. denticu- lata that now live in the northern part of the continent (Auffenberg, 1971). Although emydid turtles now occur in South America, their presence there during the Tertiary is still uncertain. The existence of remains referable to the Emydidae among the material collected from late Miocene de- posits in Colombia was mentioned by Medem (1966, 196S), but the record has not been substantiated. The extant genus Geocmyda is known from the late Pleistocene of Ecuador. No fossil records of the Chelydridae and Kinosternidae are known in South America. Trionychids, old members of the North Amer- ican faunas, are not present today in South America. Their presence in northern Vene- zuela in the late Pliocene could have resulted from waif dispersal, but their colonization was not successful (Wood and Patterson, 1973). In the Late Cretaceous and Early Tertiary, a peculiar group of land turtles of uncertain affinities, the Meiolaniidae, was also part of the herpetofauna of southern South America. They also have been recorded in Australia, where they survived until Pleistocene times. In summary, pelomedusids are ancient components of the ehelonian faunas of South 1979 BAEZ & GASPARINI: FOSSIL RECORD 41 America, being continuously represented since the Late Mesozoic. Podocnemis, the only liv- ing genus in South America, is known since the Late Cretaceous; it has undergone con- siderable speciation. Chelids also are old members of the fauna, even though no occur- rence prior to the early Eocene is known. No land turtles are known from the Late Meso- zoic and Early Tertiary besides meiolaniids. Their ecological role was assumed later by the testiudinid Geochelone, which entered the continent probably in Eocene-Oligocene times. The fossil record of the lizard family Iguanidae in South America is fragmentary, a situation that contrasts with its significance in the present Neotropical herpetofauna. Al- though the former presence of this group in Asia and Europe has been discarded (Hoff- stetter, 1962; Estes, 1970), fossils referable to this family have been recorded in North America, but none antedates the early Eocene (Estes and Price, 1973). The earliest record of an iguanid is from the Upper Cretaceous of Brasil, and the family is represented by at least five species in the late Paleocene fauna from Itaborai, Brasil, thereby indicating an early radiation in South America (Estes, 1970). Presently-known data are consistent with the suggested Gondwanan origin of iguanids (Cracraft, 1973). The group could have entered North America from the south by waif dispersal, becoming quite diversified by Miocene times ( Robinson and Van Deven- der, 1973). By the Miocene, iguanids attained a wide distribution in South America, but little is known of the taxa present at that time (Baez and Gasparini, 1977). Extant repre- sentatives of the family are recorded from late Pleistocene deposits: Iguana in coastal Ecua- dor ( Hoffstetter, 1970b) and Leiosaurus bellii in central-western Argentina (Van Devender, 1977). The evolution of teiids took place primar- ily in South America, where they are most diverse and widely distributed today. In the Late Cretaceous teiids were present in North America (Estes, 1964, 1969), although these early records do not seem to be of direct sig- nificance in the establishment of the Recent teiids there (Estes, 1970). On the other hand, fossil remains referable to the Teiidae and resembling primitive living South Amer- ican taxa have been recorded from the late Paleocene of Brasil (Estes, 1970). Their ab- sence in the Cretaceous is not surprising be- cause continental Middle and Late Mesozoic faunas are still poorly known in South Amer- ica. Practically all later teiid fossils (Mio- cene-Pleistocene) have been referred to the extant South American genera Tupinambis, Dracaena, Callopistes, and Dicrodon, al- though differences in distributions are evident when compared with their present ranges (Baez and Gasparini, 1977). Very few fossil records of the other lizard families now inhabiting South America are known from that continent. However, gek- konids occur in the upper Paleocene of Brasil. Although this is the earliest known record of the family, the origin of gekkonids could ex- tend back to the Late Mesozoic, upon consid- eration of their relationships with some Juras- sic European and Asiatic taxa (Hoffstetter, 1964; Kluge, 1967). Living South American representatives might have been derived from different sources; the strong African affinities of sphaerodactylines (Kluge, 1967) could be interpreted as resulting from the past connec- tions of those continents (Estes, 1970). The presence of scincids in South America was extended back to the Paleocene ( Estes, 1976 ) . These lizards are seemingly already repre- sented in the Late Cretaceous North Amer- ican faunas (Estes, 1976). Even though the fossil record of lizards in South America is extremely poor, it is evident that iguanids and teiids, both well repre- sented there today, were characteristic com- ponents of the Cenozoic herpetofaunas of that continent, where they underwent consid- erable diversification. Those groups, and also gekkonids, seem to comprise an ancient faunal element. The presence of anguids in South America has not yet been documented by the fossil record. They are present in North America since the Late Cretaceous, and seem to have had an essentially northern dispersal (Hoffstetter, 1962; Estes, 1964; Meszoely, 1970). Snake remains have been recorded in South America from the Late Cretaceous through the Pleistocene. With a few excep- tions, no good descriptions are available, and 42 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 much of the material has been assigned only to families. At least in the Early Tertiary the ophidian fauna was comprised mostly of boids. They were represented by the large snakes of the genus Madtsoia, of the extinct subfamily Madtsoiinae, which were recorded from the late Paleocene and early Eocene of Patagonia. Species of that genus also are known from the Late Cretaceous of Mada- gascar ( Hoffstetter, 1961) and of Niger (Broin et al., 1974), a distribution that sug- gests their derivation from a Gondwanan stock. Remains assigned to the Boidae have been cited from the late Paleocene of Brasil and late Pleistocene of Bolivia. The living South American genus Eunectes occurs in late Miocene deposits of Colombia. Boids probably were also represented in the early- middle Miocene of southern Patagonia and the Pliocene of Parana, Argentina, but those records have not been substantiated (Gas- parini and Baez, 1975; Baez and Gasparini, 1977). Aniliids, now restricted to the Oriental and Neotropical regions, were represented in the Early Tertiary faunas of South America. A snake of controversial phylogenetic position, Dinilysia patagonica, was recorded from the Late Cretaceous of central-western Argentina (Smith Woodward, 1901; Huene, 1929). Its affinities to modern aniliids were pointed out by Estes, Frazzeta and Williams (1970), but it was retained in the monotypic family Dini- lysiidae. However, Dinilysia was considered to be closer to boids, although not ancestral to them, by Rage ( 1977). True aniliids are pres- ent in the late Paleocene of Brasil, but the material has not yet been described. The late Miocene Colombophis portai from Colombia more closely resembles the extant Asiatic Cy- lindrophis than Anilhis, now living in South America (Hoffstetter, 1967b; Hoffstetter and Rage, 1977). The earliest known record of aniliids is from the Middle-Late Cretaceous of Canada (Fox, 1975); their presence in North America is documented from that time through the Early Tertiary (Estes, 1976). The group also occurs in Eocene deposits of Eu- rope (Hoffstetter, 1962; Rage, 1974). Rein- terpretation of relationships of Dinilysia and the known fossil record of aniliids seems to be more consistent with the postulated Laurasian origin of that group (Cracraft, 1973). The evidence, however, is still meager. The earliest known colubrid snake in South America is from late Miocene deposits in Colombia; other fossil colubrids in South America are from the late Pleistocene. True colubrids appear in the late Eocene of Europe (Rage, 1974). In North America their pres- ence has been documented since the early Miocene (Holman, 1976b); they become dominant elements of snake faunas by the late Miocene (Holman, 1976a). Thus, an entry in South America from the north during the Miocene was suggested (Hoffstetter, 1967b; Hoffstetter and Rage, 1977). However, it is noteworthy that the relationship of some Mio- cene colubrids from the United States to living Central or South American forms could indi- cate that they were derived from a more southern source (Tihen, 1964). Fossil viperids, which were referred to the Crotalinae, are known in South America only from late Pleistocene deposits in Bolivia. They could be late immigrants from the north (Reig, 1962; Baez and Gasparini, 1977), al- though their arrival in South America should have preceded that record. In the early Mio- cene, typical viperids appear in Europe (Hoffstetter, 1962). Viperid remains, tenta- tively assigned to the Crotalinae, were re- ported from the middle Miocene of North America ( Holman, 1976c ) . All evidence indi- cates, as in the case of colubrids, that the early history of this group is still largely unknown. The fossil record of snakes in South Amer- ica is not only meager but remains practically unstudied. At least during the Early Tertiary henophidians predominated, although lin- eages different from those evolving in the northern continents seem to have been present there. Available paleontological data do not give much information concerning the origin of cenophidians in the South American con- tinent. Scolecophidians are not yet recorded. The Crocodylidae in South America now are restricted to the northern part of the con- tinent. However, among the early Paleocene taxa assigned to that group, Necrosuchus ion- en-sis Simpson, 1937, inhabited southern Ar- gentina. The affinities of this crocodilian have not been clearly established. Crocodylids ap- pear in the fossil record in Cretaceous de- 1979 BAEZ & GASPARINI: FOSSIL RECORD 43 posits of comparable age in North America, Africa and Asia (Sill, 1968). Thus, it is doubtful if the ancestors of Necrosuchus came from North America, at this time is is impos- sible to determine their area of origin. From the Oligocene onward, crocodylids are re- corded north of the Amazonian Basin, with a latitudinal distribution similar to the present one. No fossil Crocodylus is known from South America, except for a doubtful record from the Pliocene of Maranhao, coastal Brasil (Maury, 1923). Available evidence suggests that the living representatives of that genus are late immigrants from the north. The rela- tionships of the peculiar crocodylid, Charac- tosuchus fieldsi Langston, 1965, from the late Miocene of Colombia are still unknown. Alligatorid crocodilians are distributed more widely than crocodylids in South Amer- ica. The earliest known representatives in that continent occur in the late Paleocene of Brasil, and they probably also are present in deposits of that age from southern Argentina. In both cases, the material has not been studied yet. The validity of the Paleocene "Notocaiman stromeri" from Patagonia, pro- posed as an ancestor of Caiman, has been dis- carded (Gasparini and Baez, 1975). The presence of alligatorids in North America ex- tends back to the Cretaceous, and a center of origin there was suggested by Sill (1968). Nevertheless, available data are inconclusive. The Eocene Eocaiman cavernensis is the old- est member of the family described from South America. Langston ( 1965 ) pointed out its modern aspect and suggested that it could have been ancestral to Caiman, Melanosuch- us, and perhaps also to the peculiar Balanero- dus. Kalin ( 1955 ) did not rule out the pos- sibility that it could be referred to the extant genus Caiman. This indicates that taxa close- ly related to living South American represen- tatives of this family were already present on that continent by the Early Tertiary. In the late Miocene of Colombia an extant genus and probably also a species ( Caiman latiro- stris) are known (Gasparini and Baez, 1975). The recent Caiman yacare is known from the Pliocene of Argentina and Melanosuchus from the Pliocene of Venezuela. Today, gavialids are restricted to the Ganges Basin, India, but they may have in- habited South America in the Tertiary. It is still controversial if the South American Ion- girostrine crocodilians are true gavialids or constitute a different lineage ( Langston, 1965; Gasparini, 1968; Sill, 1968, 1970; Hoffstetter, 1970a; Hecht and Malone, 1972; Baez and Gasparini, 1977). The oldest gavialids are known from Eocene deposits in Egypt (Hecht and Malone, 1972). The Asiatic and South American forms could have originated from an African stock; their dispersal would have been facilitated by their capacity to swim in marine waters. In South America the first gavialids appear in deposits of late Oligocene- early Miocene age from the northern part of the continent; they became extinct by the Plio-Pleistocene. The referral of South Amer- ican forms to the genus Gavialis is doubtful, because they are different from the Asiatic members of that genus. The extinct Nettosuchidae is an endemic group from mesic tropical environments of the northern and central part of the continent from the late Miocene to the earliest Pleisto- cene. Crocodilians of the more primitive and extinct suborder Mesosuchia are known in South America during a large part of the Ter- tiary. These forms belong to the family Se- becidae ( Sebecosuchia, Gasparini, 1972) and are known from the Paleocene through the Miocene. They were peculiar, mostly terres- trial crocodiles, and inhabitants of the tropical forests (Langston, 1965; Molnar, 1969, 1977; Neill, 1971; Gasparini, 1972). Sebecids seem to have had a long history on the continent, where they probably originated from bauro- suchids. The great diversity of Cenozoic crocodil- ian faunas in South America is noteworthy. Species referable to five families (Sebecidae, Crocodylidae, Alligatoridae, Gavialidae, Net- tosuchidae) have lived there, but only two of those groups (Crocodylidae, Alligatoridae) are present now. Available evidence suggests that the differentiation of sebecids and netto- suchids occurred in South America. The most conspicuous groups were the Sebecidae and Alligatoridae, the latter being represented throughout the Cenozoic and the most im- portant crocodilian family there today. On the other hand, the fossil record of crocodylids is 44 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 comparatively poor; it seems likely that they never were important components of the South American herpetof auna. The living rep- resentatives of this family are not related to the earliest forms recorded from early Terti- ary deposits in the southern part of the continent. DISCUSSION This review of the fossil record of amphib- ians and reptiles of South America discloses how poor that record actually is and how many questions concerning the history of the groups still remain unanswered. The available paleontological data are insufficient to pro- vide a comprehensive picture of the evolu- tionary, faunal, and distributional changes that occurred in South America. It is perti- nent to emphasize here that practically no special techniques for fossil collecting have been applied; for the most part, the discov- ery of amphibians and reptiles has been hap- hazard. Comparison of local faunas is diffi- cult, for in many cases little weight can be given to the differences in composition of known assemblages; thus, evolutionary and zoogeographic interpretations are still tenta- tive. The records of certain groups, such as tur- tles and crocodilians, greatly outnumber those of other groups which could be partially the result of the fact that most known sites that have yielded amphibian or reptile remains represent lowland and aquatic habitats. This could also explain the absence of other groups associated with different ecological condi- tions. In general, little is known about the heqoetofaunas that inhabited the northern and central part of the continent, not only in the cratonic areas but also in the intercratonic basins. These regions have a considerable significance from the zoogeographic point of view in that they have been considered areas of origin of diverse groups. The oldest known remains referable to modern groups of amphibians and reptiles occurring in South America are from the Late Cretaceous deposits, except for the record of leiopelmatid frogs (Vieraella and Notobatra- chus) in the Jurassic of Patagonia. The next important sample is the late Paleocene rec- ords from Itaborai, Brasil, the earliest assem- blage which offers a broad picture of the groups already comprising the fauna of the continent. According to available data from that sample, little similarity to the North American assemblages of comparable age are evident (Estes, 1976; Baez and Gasparini, 1977). Basically, the family groups recorded from the known Paleocene and Eocene sites evolved in isolation in South America and comprise the present fauna. Some of them such as leptodactylids, bufomds, pelomedu- sids, chelids, teiids, alligatorids, were already represented by forms related to their living representatives on that continent. It is evi- dent that the early history in South America of many of those groups extends back well into the Mesozoic. Unfortunately, Mesozoic records are very scarce, and the earliest known Tertiary samples (except that from Itaborai, Brasil) are badly preserved and not diverse (Gasparini and Baez, 1975). Of the additional family groups recorded from Oli- gocene and Miocene deposits, testudinids and perhaps emydids could have arrived from the north by overwater transport at different times. The known late Cenozoic records pro- vide little information concerning the faunal interchange between the Americas when the isthmian link was established. Different lines of evidence indicate that in the Early Tertiary a humid and warm tem- perate climate prevailed at high latitudes in South America as well as in North America (Dawson, et al., 1976). In the southernmost part of South America, the known assem- blages of that age are archaic, being com- prised of extinct groups that do not now exist in the area. Tectonic events throughout the Cenozoic altered the physiographic condi- tions, and the subsequent climatic and Holistic changes affected the composition of the local faunas. The modification of the Patagonian herpetofauna from the early Eocene onward clearly illustrates this point. The subtropical elements, such as crocodilians, disappeared from that region. Furthermore, the gradual desiccation of climate related to the Andean uplift presumably resulted in the confinement of taxa adapted to humid and aquatic environ- ments to the more mesic western areas. The 1979 BAEZ & GASPARINI: FOSSIL RECORD 45 change of the biota in the region of the pres- ent upper Magdalena Valley, Colombia, in relation to the uplift of the Eastern Cordillera and increasing aridity ( Howe, 1974 ) is an- other example. Many components of the rich late Miocene La Venta fauna, living on the broad flood plains that prevailed there are absent from that now semiarid area (Fields, 1959). Many fundamental aspects in the history of the South American herpetofauna have yet to be elucidated. The amount of paleontolog- ical information that has accumulated in re- cent years enables us to expect that future studies will clarify these problems. ACKNOWLEDGMENTS For their valuable comments and provision of data, we are indebted to Drs. France de Broin, Richard Estes, J. Alan Holman, Bryan Patterson and Roger Wood. Special thanks are extended to Dr. William E. Duellman, who kindly revised and corrected the manu- script. RESUMEN En este trabajo se sintetiza la information disponible sobre el registro fosil de los grupos que integran la actual herpetofauna de Amer- ica del Sur, en un intento de valorar el aporte que el mismo brinda al conocimiento del desarrollo historico de dichos grupos en ese continente. En tal sentido, se han tornado en cuenta no solo los registros sudamericanos, sino tambien aquellos otros directamente rela- cionados y provenientes de otras partes del mundo. Para integrar los resultados en un contexto coherente se considero muy espe- cialmente la disposition y relation de Amer- ica del Sur con respecto a otras masas con- tinentales a partir del Mesozoico medio. Tam- bien se tomaron en cuenta los principales eventos geologicos acaecidos desde fines del Cretacico a la actualidad en dicho continente, valorando su incidencia en los cambios fisio- graficos los que, evidentemente, actuaron sobre la composition y distribution de la herpetofauna. La mention de anfibios y reptiles ceno- zoicos en America del Sur es relativamente frecuente, sin embargo, muchas de las asigna- ciones taxonomicas como las referencias geo- graficas y cronoestratigraficas son dudosas. Ello, conjuntamente con el hecho de que la mayoria de los hallazgos son aislados, sin formar parte de asociaciones representativas, hace que las conclusiones resulten aiin tenta- tivas. En general los datos disponibles per- miten un analisis a nivel familiar. El examen del registro senala la notable antigiiedad, en America del Sur, de muchas de la familias de anfibios y reptiles que viven en ese continente. Todas las familias regis- tradas en el Terciario temprano son inte- grantes de la herpetofauna actual sudameri- cana, excepto las tortugas meiolanidas y los cocodrilos sebecidos, ambos e.xtinguidos. Es evidente que esas familias tienen distinto abolengo, habiendose integrado o diferenci- ado in situ alocronicamente, desde fines del Cretacico al menos. De acuerdo a las eviden- cias paleontologicas la antigiiedad en Ameri- ca del Sur de los pipidos, iguanidos, pelo- medusidos, meiolanidos, muy probablemente los quelidos, y posiblemente los leptodacti- lidos se remonta al Mesozoico tardio. De estas familias, los pipidos, meiolanidos y quelidos son de origen gondwanico. La diferenciacion de los sebecidos y leptodactilidos, en sentido estricto, habria tenido lugar en America del Sur. Los datos paleontologicos son aun in- suficientes para dilucidar el origen de otras familias ya presentes en el Terciario tem- prano. En el Terciario medio y tardio se constata la presencia de familias no registra- das en sedimentos mas viejos. Algunos de esos grupos pudieron haber arribado a Ameri- ca del Sur por medios de dispersion accidental en diferentes momentos durante su aislami- ento. Tal seria el caso de los testudinidos, trioniquidos, gavialidos, crocodilidos directa- mente relacionados a las formas actuales y tal vez los emididos. Los cocodrilos nettosuqui- dos, actualmente e.xtinguidos, se diferenciaron in situ. De la confrontation con la herpeto- fauna actual se desprende que numerosas familias no estan presentes en el registro. Al- gunas de ellas tales como los ranidos, angui- dos, chelidridos y tal vez kinosternidos po- 46 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 drian ser invasores tardios, llegados a traves del Istmo de Panama. Otras por el grado de endemismo y relaciones filogeneticas serian mas antiguas integrantes de la fauna de este continente, no obstante no haberselas regis- trado fosiles hasta el momento. Tal es el caso, por ejemplo, de dendrobatidos y rinoderma- tidos. La localization geografica de los depositos portadores de anfibios y reptiles ha sido tam- bien tomada en consideration. Resulta evi- dente que, con poeas excepciones, es escasa o nula la information disponible sobre las fau- nas de anfibios y reptiles que habitaron la parte norte del continente, tanto en las areas cratonicas como en las cuencas intercraton- icas. Estas regiones revisten especial interes por cuanto se ha sefialado su importancia como areas de origen de diversos grupos. Re- cien a partir del Oligoceno se conocen regis- tros de los grupos considerados en el extremo mas septentrional, fundamentalmente en Co- lombia y Venezuela. En cambio, en la parte sur del continente, en la Patagonia extran- dina, varias han sido las localidades donde depositos del Terciario temprano brindaron restos de anfibios y reptiles. Segun diversas evidancias, en ese tiempo las condiciones am- bientales fueron mas benignas que las actuales en esa region, por lo que la herpetofauna fosil que alii se registra es muy distinta de la que habita ese area en nuestros dias. El registro fosil es aim inadecuado para brindar un panorama integral de los cambios ocurridos en la composition de las diversas faunas regionales en consonancia con las modificaciones ambientales; del mismo modo limita el conocimiento de la evolucion intra- familiar a traves del tiempo. La busqueda sistematica y la aplicacion de tecnicas adecu- adas permitiran, sin duda, un mayor aporte de la paleontologia al conocimiento de la historia de la herpetofauna sudamericana. LITERATURE CITED Ahl, E. 1926. Anura; Aglossa, pp. 141-142 in Kaiser, E., Beetz, W. (eds. ). Die Diamanten- wiiste Siidwest Afrikas 2. D. Reimer, Berlin, 535 p. Ambrosetti, J. 1890. Observaeiones sobre los rep- tiles fosiles oligocenos de los terrenos terciarios antiguos del Parana. Bol. Acad. Nac. Cien. Cor- doba 10:489-499. Ameghixo, F. 1893. Sobre la presencia de verte- brados de aspecto mesozoico en la formacion Santacrueeiia de Patagonia austral. Rev. Jardin Zool. Buenos Aires 1:75-84. Ameghino, F. 1899. Sinopsis geologico-paleontolog- ica. Suplemento. La Plata, Folio 13 p. Antunes, M. T. 1975. Iberosuchus, crocodile sebe- cosuchien nouveau, de l'Eocene iberique au Nord de la Chaine Centrales et l'origine du Canyon de Nazare. Comun. Serv. Geol. Portugal 59:285-330. Archangelsky, S., Romero, E. 1974. Los registros mas antiguos del polen de Nothofagus (Fagaceae) de Patagonia (Argentina y Chile). Bol. Soc. Bot. Mexico 33:13-30. Arid, F., Vizotto, L. 1966. Um quelonio fossil de Sao Jose do Rio Preto. Rev. Cien Cultura 18: 422^28. Auffenberg, W. 1971. A new fossil tortoise, with remarks on the origin of South American testu- dines. Copeia 1971( 1 ) : 106-H7. Auffenberg, W. 1974. Checklist of fossil land tor- toises (Testudinidae). Bull. Florida State Mus. Biol Ser. 18:121-251. Baez, A. M. 1975. Los anuros de la Formacion Las Curtiembres (Cretacico superior), provincia de Salta, Republica Argentina. Evolucion de la familia Pipidae (Amphibia, Anura) en relation a la historia paleogeografiea. Tesis Doctoral Univ. Buenos Aires, 144 p. Baez, A. M. 1976. El significado paleogeografico y paleoecologico de los pipidos (Amphibia, Anura) fosiles de America del Sur. VI Congr. Argent. Geol., Bahia Blanca, 1975, 1:333-340. Baez, A. M. 1977. Sobre Teracopluys Ameghino, nomina nuda (Anura, Leptodactylidae ) de la Formacion Colhue-Huapi (Oligoceno superior), Provincia del Chubut, Republica Argentina. Rev. Asoc. Geol. Argentina, 32:145-151. Baez, A. M., Gasparini, Z. B. de. 1977. Origenes y evolucion de los anfibios y reptiles del Cenozoico de America del Sur. Acta Geol. Lilloana 14:140- 232. Blair, W. F. 1972. Summary, pp. 329-343 in Blair, W. F. (ed. ). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Bravard, A. 1858. Monografia de los terrenos ma- rinos terciarios del Parana. Diario Oficial de Gobiemo "El Nacional Argentino." (Not seen) Broin, F. de. 1971. Une espece nouvelle de tortue pleurodire (?RoxocIielys vilavilensis n.sp.) dans le Cretace superieur de Bolivie. Bull. Soc. Geol. France (7)13:445-452. Broin, F. de, Buffetaut, E., Koeniguer, J. C, Rage, J. C, Russell, D., Taquet, P., Vergnaud- Grazzini, C, Wenz, S. 1974. La faune de verte- bres continentaux du gisement din Beceten ( Se- nonien du Niger). C. R. Acad. Sci. Paris, Ser. D 279:469-472. 1979 BAEZ & GASPARINI: FOSSIL RECORD 47 Burmeister, H. 1883. Monografia de los terrenos marinos terciarios del Parana. An. Mus. Puhl. Buenos Aires 3:45-94. Burmeister, H. 1885. Examen eritico de los niamf- feros y reptiles fosiles denominados por don Augusto Bravard. Ibid. 3:95-174. Casamiquela, R. 1958. Un anuro gigante del Mio- ceno de Patagonia. Rev. Asoc. Geol. Argentina 13:171-183. Casamiquela, R. 1960. Datos preliminares sobre un pipoideo fosil de la Patagonia. Actas y Trabajos I Congr. Sudamer. Zool., La Plata 1959, 4:17-21. Casamiquela, R. 1963. Sobre un par de anuros del Mioceno de Rio Negro (Patagonia) Wawclia ger- lioldi n. gen. et. sp. ( Ceratophrydidae ) y Gigan- tohatrachus parodii ( Leptodactylidae ) . Ameghini- ana 3:141-162. Casamiquela, R. 1967. Sobre un nuevo Bufo fosil de la Provincia de Buenos Aires ( Argentina ) . Ibid. 5:161-168. Cattoi, N., Freibebc, M. 1958. Una nueva especie de "Podocncmis" del Cretaceo argentine Physis 21 (60): 58-66. Cei, J. 1972. Bufo of South America, pp. 82-90 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Cham, J. 1976. Relaciones de un nuevo Teiidae (Lacertilia) fosil del Plioceno superior, Callopistes bicuspidatus n. sp. Inst. Miguel Lillo, Univ. Nac. Tucuman, Publ. Especial, pp. 133—153. Cracraft, J. 1973. Vertebrate evolution and bio- geography in the Old World tropics, pp. 373-393 in Tarlinc, D., Runcorn, S. (eds.). Implica- tions of continental drift to the earth sciences, Vol. 1. Academic Press, New York, 622 p. Dalziel, I., Kligfield, R., Lowrie, N., Opdyke. N. D. 1973. Palaeomagnetic data from the southernmost Andes and the Antarctandes. Ibid.: 87-101. Dawson, M., West, R., Lancston, W., Hutchison, J. 1976. Paleogene terrestrial vertebrates: north- ernmost occurrence, Ellesmere Island, Canada. Science 192:781-782. Douglas, R., Moullade, M., Nairn, A. 1973. Causes and consequences of drift in the South Atlantic, pp. 513-534 in Tarling, D., Runcorn, S. (eds.). Implications of continental drift to the earth sciences, Vol. 1. Academic Press, New York, 622 p. Emiliant, C, Gaertner, S., Lidz, B. 1972. Neo- gene sedimentation on the Blake Plateau and the emergence of the Central American Isthmus. Palaeogeogr., Palaeoclimatol., Palaeoecol., 11:1- 10. Estes, R. 1961. Miocene lizards from Colombia, South America. Breviora ( 143) : 1—11. Estes, R. 1964. Fossil vertebrates from the late Cretaceous Lance Formation eastern Wyoming. Univ. Calif. Publ. Geol. Sci. 49:1-180. Estes, R. 1969. Relationships of two Cretaceous lizards (Sauria, Teiidae). Breviora (317): 1-8. Estes, R. 1970. Origin of the Recent North Amer- ican lower vertebrate fauna: An inquiry into the fossil record. Forma Functio 3:139-163. Estes, R. 1975a. Xenopus from the Palaeocene of Brazil and its zoogeographic importance. Nature 254:48-50. Estes, R. 1975b. Fossil Xenopus from the Palaeo- cene of South America and the zoogeography of pipid frogs. Herpetologica 31:263-278. Estes, R. 1976. Middle Paleocene lower vertebrates from the Tongue Formation, southeastern Mon- tana. J. Paleontol. 50:500-520. Estes, R., Frazzetta, T., Williams, E. 1970. Stud- ies on the fossil snake Dinilysia patagonica Wood- ward. Part I. Cranial morphology-. Bull. Mus. Comp. Zool. Harvard Univ. 140:25-74. Estes, R., Price, Ll. 1973. Iguanid lizard from the upper Cretaceous of Brazil. Science 180:748-751. Estes, R., Reig, O. 1973. The early fossil record of frogs: A review of the evidence, pp. 11-63 in Vial, J. L. (ed. ). Evolutionary biology of the anurans: Contemporary research on mayor prob- lems. Univ. Missouri Press, Columbia, 470 p. Estes, R., Wake, M. 1972. The first fossil record of caecilian amphibians. Nature 239:228-231. Estes, R., Wassersug, R. 1963. A Miocene toad from Colombia, South America. Breviora (193): 1-13. Fields, R. 1959. Geology of the La Venta badlands, Colombia, South America. Univ. Calif. Publ. Geol. Sci. 23:403-444. Fox, R. 1975. Fossil snakes from the Upper Milk- River Formation (Upper Cretaceous), Alberta. Canad. J. Earth Sci. 12:1557-1563. Gaffney, E. 1975. A revision of the side-necked turtle Taphrosplujs sulcatus (Leidy) from the Cretaceous of New Jersey. Amer. Mus. Novit (2571): 1-24. Gaffney, E., Zangerl, R. 1968. A revision of the chelonian genus Bothremys ( Pleurodira: Pelome- dusidae). Fieldiana Geol. 16:193-239. Gallardo, J. M. 1962. A proposito de Bufo varie- gatus (Giinther), sapo del bosque humedo An- tartandico, y las otras especies de Bufo neotropi- cales. Physis 23 (64):93-102. Gasparini, Z. B. de. 1968. Nuevos restos de Rham- plwstomopsis neogaeus (Burm.) Rusconi 1933 ( Reptilia, Crocodilia) del "Mesopotamiense" (Pli- oceno medio-superior) de Argentina. Ameghini- ana 5:299-311. Gasparini, Z. B. de. 1972. Los Sebecosuchia (Croco- dilia) del territorio argentino. Consideraciones sobre su "status" taxonomico. Ibid. 9:23-34. Gasparini, Z. B. de. 1973. Revision de los Crocodilia (Reptilia) fosiles del territorio argentino. Su evolucion, sus relaciones filogeneticas, su clasifica- cion y sus implicancias estratigrahcas. Tesis Doc- toral, Univ. Nac. La Plata, 169 p. Gasparini, Z. B. de, Baez, A. M. 1975. Aportes al conocimiento de la herpetofauna terciaria de la Argentina. Actas I Congr. Argent Paleontol. Bioestr., Tucuman 1974, 2:377-415. 48 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Haffer, J. 1970. Geologic-climatic history and zoo- geographic significance of the Uraba Region in northwestern Colombia. Caldasia 10:603-636. Hecht, M. 1963. A reevaluation of the early history of the frogs. Part II. Syst. Zool. 12:20-35. Hecht, M., Malone, B. 1972. On the early history of the gaviaiid crocodilians. Herpetologica 28: 281-284. Heyer, W. R. 1975. A preliminary analysis of the intergeneric relationships of the frog family Lep- todactylidae. Smithsonian Contr. Zool. (199): 1-55. Hoffstetter, R. 1959. Un dentaire de Madtsoia ( Serpent geant du Paleocene de Patagonie ) . Bull. Mus. Natl. Hist. Nat. Zool., Paris, (2) 31:379-386. Hoffstetter, R. 1961. Nouveau.x restes dun ser- pent Boide ( Madtsoia madagascaricnsis nov. sp. ) dans le Cretace inferieur de Madagascar. Ibid. (2) 33:152-160. Hoffstetter, R. 1962. Revue des recents acquisi- tions concernant l'histoire et la systematique des Squamates, pp. 243-279 in Problemes actuels de Paleontologie (Evolution des Vertebres). Colloq. Int. C.N.R.S. Paris 104, 474 p. Hoffstetter, R. 1963. La faune Pleistocene de Tarija (Bolivie). Note preliminaire. Bull. Mus. Natl. Hist. Nat. Zool., Paris (2) 35:195-203. Hoffstetter, R. 1964. Les Sauria du Jurassique superieur et specialment les Gekkota de Baviere et de Mandchourie. Senckenbergiana Biol. 45: 281-324. Hoffstetter, R. 1967a. Remarques sur les dates d'implantation des differents groupes de serpents terrestres en Amerique du Sud. Convp. Rend. Somm. Soc. Geol. France 1967:93-94. Hoffstetter, R. 1967b. Observations additionelles sur les serpents du Miocene de Colombie et recti- fication concernant la date d'arrivee des colubrides en Amerique du Sud. Ibid. 1967:209-210. 32:419-426. Hoffstetter, R. 1968. Nuapua, un gisement de vertebres pleistocenes dans le Chaco Bolivien. Bull. Mus. Natl. Hist. Nat. Zool., Paris (2) 40: 823-836. Hoffstetter, R. 1970a. Vertebrados cenozoicos de Colombia. Actas IV Congr. Latinoamer. Zool., Caracas 1968, 2:931-954. Hoffstetter, R. 1970b. Vertebrados cenozoicos de Ecuador. Ibid. 2:955-970. Hoffstetter, R. 1970c. Vertebrados cenozoicos y mamiferos cretacicos del Peru. Ibid. 2:971-984. Hoffstetter, R., Rage, J. C. 1977. Le gisement de vertebres miocenes de La Venta (Colombie) et sa faune da Serpents. Ann. Paleontol. (Vertebres) 63:161-190. Hol.man, J. 1968. Lower Oligocene amphibians from Saskatchewan. Q. J. Florida. Acad. Sci. 31:273-289. Holman, J. 1976a. Snakes from the Rosebud For- mation ( Middle-Miocene ) of South Dakota. Her- petologica 32:41—48. Holman, J. 1976b. A boid and a colubrid snake from the Harrison Formation ( Lower Miocene, Arikareean) of Sioux County, Nebraska. Ibid. 32:387-389. Holman, J. 1976c. Snakes of the Split Rock Forma- tion ( Middle Miocene ) , Central Wyoming. Ibid. 32:419-426. Howe, M. 1974. Nonmarine Neiva Formation (Plio- cene?) Upper Magdalena Valley, Colombia: re- gional tectonism. Geol. Soc. Amer. Bull. 85:1031- 1042. Huene, F. von. 1929. Los sarisquios y ornitisquios del Cretaceo argentino. An. Mus. La Plata 3:1- 196. Irving, E. M. 1971. La evolucion estructural de los Andes mas septentrionales de Colombia. Inst. Nac. Invest. Geol. Min. Colombia, Bol. Geol. 19: 1-90. Jimenez Fuentes, F. 1971. Los reptiles fosiles del Valle del Duero: Podocnemis carbajosai nov. sp. del Eoceno de Salamanca. Estud. Geol. (Madrid) 27:85-93. Jimenez Fuentes, E. 1975. Duerochelys arribasi, nov. gen., nov. sp., Pelomedusidae du Ludien du Bassin du Duero (Espagne). Bull. Geol. Soc. France (7) 17:410-415. Kalin, J. 1955. Crocodilia, pp. 695-784 in Piveteau, J. (ed.). Traite de Paleontologie, Masson et Cie., Paris 5:1-1113. Kluge, A. 1967. Higher taxonomic categories of gekkomd lizards and their evolution. Bull. Amer. Mus. Nat. Hist. 135:1-59. Langston, W. 1965. Fossil crocodilians from Co- lombia and the Cenozoic history of the Crocodilia in South America. Univ. Calif. Publ. Geol. Sci. 52:1-157. Langston, W. 1966. Mourasuchus Price, Ncttosuch- us Langston, and the family Nettosuchidae ( Rept. Crocodilia). Copeia 1966(4) :882-885. Lemon, R., Chubcher, C. 1961. Pleistocene geology and paleontology of the Talara region, Northwest Peru. Amer. J. Sci., 259:410-129. Lynch, J. 1971. Evolutionary relationships, osteol- ogy and zoogeography of leptodactyloid frogs. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): 1-238. Lynch, J. 1973. The transition from archaic to ad- vanced frogs, pp. 133-182 in Vial, J. L. (ed. ). Evolutionary biology of the anurans: Contempo- rary research on major problems. Univ. Missouri Press, Columbia, 470 p. Malfait, B., Dinkelman, M. 1972. Circum-Carib- bean tectonic and igneous activity and the evolu- tion of the Caribbean Plate. Geol. Soc. Amer. Bull. 83:251-272. Maury, C. 1924. Fosseis terciarios do Brasil com descripgao de novas formas cretaceas. Serv. Geol. Min. Brasil, Monogr. 4:1-711. McGowran, G. 1973. Rifting and drift of Australia and the migration of mammals. Science, 180: 759-761. 1979 BAEZ & GASPARINI: FOSSIL RECORD 49 McKenna, M. 1973. Sweepstakes, filters, corridors, Noah's arks, and beached Viking funeral ships in palaeogeography, pp. 295-308 in Tarlinc, D., Runcorn, S. (ed.). Implication of continental drift to the earth sciences. Vol. 1. Academic Press, New York, 622 p. Medem, F. 1966. Contribuciones al conocimiento sobre la ecologia y distribution geografica de Phrynaps (Batrachemys) dahli; (Testudinata, Pleurodira, Chelyidae). Caldasia 9:467-489. Medem, F. 1968. Desarrollo de la herpetologia en Colombia. Rev. Acad. Colombiana Cien. Exactas, Fis. Nat. 13:149-200. Medina, C. 1976. Crocodilians from the Late Terti- ary of Northwestern Venezuela: Melanosuchus fisheri sp. nov. Breviora (438):1-14. Meszoely, C. 1970. North America fossil anguid lizards. Bull. Mus. Comp. Zool. Harvard Univ. 139:87-149. Molnar, R. 1969. Jaw musculature and jaw mechan- ics of the Eocene crocodilian Sebecus icaeorliinus. M.A. Thesis. Univ. Texas, Austin, 89 p. Molnar, R. 1977. Crocodile with laterally com- pressed snout: First find in Australia. Science, 197:62-64. Mones, A. 1972. Lista de los vertebrados fosiles del Uruguay, I. Chondrichthyes, Osteichthyes, Rep- ulia, Aves. Comun. Paleontol. Mus. Hist. Nat. Montevideo 1:23-36. Mones, A. 1975. Notas paleontologieas uruguayas, III. Vertebrados fosiles nuevos o poco conocidos ( Chondrichthyes, Osteichthyes, Amphibia, Mam- malia). Ameghiniana 12:343—349. Mook, C. 1921. Bracliygnalliosuchus braziliensis, a new fossil crocodilian from Brazil. Bull. Amer. Mus. Nat. Hist. 44:43-50. Mook, C. 1941. A new fossil crocodilian from Co- lombia. Proc. U.S. Natl. Mus. 91:55-58. Neill, W. 1971. The last of the ruling reptiles: Alligators, Crocodiles and their kin. Columbia University Press, New York, 486 p. Nevo, E. 1968. Pipid frogs from the Early Creta- ceous of Israel and pipid evolution. Bull. Mus. Comp. Zool. Harvard Univ. 136:255-318. Parodi Bustos, R., Figueroa Caprini, M., Kraglie- vich, J., Del Corro, G. 1960. Noticia preliminar acerca del yacimiento de anuros extinguidos de Puente Morales (Depto. de Guachipas, Prov. de Salta). Rev. Fac. Cienc. Nat. Salta Univ. Nac. Tucuman 1:5-25. Pascual, R. 1970. Evolution de comunidades, cam- bios faunisticos e integraciones biocenoticas de los vertebrados cenozoicos de Argentina. Actas IV Congr. Latinoamer. Zool., Caracas 1968, 2: 991-1089. Pascual, R., Odreman Rivas, O. 1971. Evolucion de las comunidades de los vertebrados del Ter- ciario argentino. Los aspectos paleozoogeograficos y paleoclimatieos relacionados. Ameghiniana 8: 372-412. Patterson, B., Pascual, R. 1968. The fossil mam- mal fauna of South America. Q. Rev. Biol., 43: 409-451. Paula Couto, C. de. 1970. Evolucao de comuni- dades, modificacoes faunisticas e integracoes bio- cenoticas dos vertebrados cenozoicos do Brasil. Actas IV Congr. Latinoamer. Zool., Caracas 1968, 2:907-930. Petkiella, B., Archangelsky, S. 1975. Vegetation y ambiente en el Paleoceno de Chubut. Actas I Congr. Argent. Paleontol. Bioestr., Tucuman, 1974, 2:257-270. Price, Ll. 1953. Os quelonios de Formacao Bauru, Cretaceo terrestre do Brasil meridional. Dept. Nac. Prod. Min., Div. Geol. Min., Rio de Janeiro 147:1-34. Price, Ll. 1954. Um quelonio pleurodiro no Cal- cario da Serie Apodi, Cretaceo do Estado do Rio Grande do Norte. Dept. Nac. Prod. Min., Div. Geol. Min., Notas Prelim. Estud, Rio de Janeiro 85:1-12. Price, Ll. 1964. Sobre o cranio de um grande croco- dilideo extinto do Alto rio Jurua, Estado de Acre. An. Acad. Brasil. Cien. 36:59-66. Rage, J.-C. 1974. Les serpents des Phosphorites du Quercy, Palaeovert, 6:274-303. Rage, J.-C. 1975. Un serpent du Paleocene du Niger. Etude preliminaire sur l'origine des cae- nophidiens ( Reptilia, Serpentes ) . C. R. Acad. Sci. Paris, Ser. D 281:515-518. Rage, J.-C. 1977. La position phyletique de Dinily- sia patagonica, serpent du Cretace superieur. Ibid. 284:1765-1768. Raven, P., Axelrod, D. 1975. History of the flora and fauna of Latin America. Amer. Sci. 63:420- 429. Reic, O. 1958. Notas para una actualization del conocimiento de la fauna de la Formation Chapad- malal. Acta Geol. Lilloana, Tucuman 2:255-283. Reic, O. 1959. Primeros datos descriptivos sobre los anuros del eocretacico de la provincia de Salta, Republica Argentina. Ameghiniana 1:1—8. Reig, O. 1962. Las integraciones cenogeneticas en el desarrollo de la fauna de vertebrados de Ameri- ca del Sur. Ibid. 2:131-140. Reig, O. 1968. Peuplement en vertebres tetrapodes de TAmerique du Sud, pp. 215-260 in Delaware Deboutteville, C, Rapoport, E. (eds.). Biolo- gie de l'Amerique australe. C. N. R. S. Groupe Francais Argiles, C. R. Reun. Edud. 4:1-472. Reig, O. 1972. Macrogenioglottus and the South American bufonid toads, pp. 14—36, in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 457 p. Reyment, R. Tatt, E. 1972. Biostratigraphical dat- ing of the early history of the South Atlantic Ocean. Phil. Trans. Roy. Soc. London, B. Biol. Sci. 264:55-95. Roblnson, M., Van Devendeb, T. 1973. Miocene lizards from Wyoming and Nebraska. Copeia 1973:698-704. Romer, A. 1966. Vertebrate Paleontology. 3rd. ed. Univ. Chicago Press, Chicago, 468 p. Rovereto, C. 1914. Los estratos araucanos y sus fosiles. An. Mus. Nac. Hist. Nat. Buenos Aires 25:1-247. 50 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Royo Y Gomez, J. 1945-46. Los vertebrados del Terciario continental colombiano. Rev. Acad. Co- lombiana Cien. Exactas Fis. Nat. 6:496-512. Rusconi, C. 1933. Observaciones criticas sobre rep- tiles terciarios de Parana ( Alligatoridae) . Rev. Univ. Nac. Cordoba 20:57-106. Rusconi, C. 1935. Observaciones sobre los gaviales fosiles argentinos. An. Soc. Cient. Argentina 119: 203-214. Rusconi, C. 1946. Ave y reptil oligoceno de Men- doza. Bol. Paleontol., Buenos Aires 21:1-3. Savage, J. 1973. The geographic distribution of frogs: Patterns and predictions, pp. 351—445 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Savage, J. 1974. The isthmian link and the evolution of Neotropical mammals. Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (260): 1-51. Schaeffer, B. 1949. Anurans from the Early Ter- tiary of Patagonia. Bull. Amer. Mus. Nat. Hist. 93:41-68. Sclater, J., Fisher, R. 1974. Evolution of the East Central Indian Ocean, with emphasis on the tec- tonic setting of the Ninetyeast Ridge. Geol. Soc. Amer. Bull. 85:683-702. Segura, P. 1944. Estudio de la primera especie de tortuga fosil de Costa Rica con algunas generali- dades sobre el orden Testudines. Esc. Farm. Guatemala, 6:9-29. (Not seen) Sige, B. 1968. Dents du micromammiferes et frag- ments de coquilles d'oefs de dinosauriens dans la faune de vertebres du Cretace superieur de La- guna Umayo (Andes peruviennes). C. R. Aead. Sci., Paris, Ser. D 267:1495-1498. Sill, W. 1968. The zoogeography of the crocodiles. Copeia 1968(1 ):76-88. Sill, W. 1970. Nota preliminar sobre un nuevo gavial del Plioceno de Venezuela y una discusion de los gaviales sudamerieanos. Ameghiniana 7: 151-159. Simpson, G. 1933a. A new fossil snake from the Notostylops Beds of Patagonia. Bull. Amer. Mus. Nat. Hist. 67:1-22. Simpson, G. 1933b. A new crocodilian from the Notostylops Beds of Patagonia. Amer. Mus. Novit. (623): 1-9. Simpson, G. 1935. Early and middle Tertiary geol- ogy of the Gaiman region, Chubut, Argentina. Ibid. 775:1-29. Simpson, G. 1937a. New reptiles from the Eocene of South America. Ibid. 927:1-3. Simpson, G. 1937b. An ancient eusuchian crocodile from Patagonia. Ibid. 965:1-20. Simpson, G. 1938. Crossochelys, Eocene horned turtle from Patagonia. Bull. Amer. Mus. Nat. Hist. 74:221-254. Simpson, G. 1942. A Miocene tortoise from Pata- gonia. Amer. Mus. Novit. (1209): 1-6. Simpson, G., Minopiuo, L., Patterson, B. 1962. The mammalian fauna of the Oivisadero Largo Formation, Mendoza, Argentina. Bull. Mus. Comp. Zool. Harvard Univ. 127:239-293. Smith, A., Briden, J., Drewry, G. 1973. Phanero- zoic world maps, pp. 1-42 in Hughes, N. (ed.). Organisms and continents through time. Spec. Pap. Palaeontol. ( Paleo. Assoc. London) 12:1- 334. Smith Woodward, A. 1901. On some extinct rep- tiles from Patagonia of the genera Miolania, Dini- hjsia and Genyodcctes. Proc. Zool. Soc. London, 1901:169-184. Staesche, K. von. 1937. Podocnemis brasiliensis n. sp. aus der Oberen Kreide Brasiliens. Neuens Jahrb. Mineral. Geol. Palaeontol. Abh. Abt. A, 77:291-309. Reprod. by Minist. Agricult., Div. Geol. Min., Brasil, Bol. 114 (1944). Stirton, R. 1953. Vertebrate paleontology and con- tinental stratigraphy in Colombia. Geol. Soc. Amer. Bull. 64:603-622. Suarez, J. 1969. Urn quelonio da formacao Bauru. Dept. Geograf., Fac. Filos. Cien. Letras, Pres. Prudente 2:35-54. Tihen, J. 1962. New world fossil bufonids. Amer. Midi. Nat. 68:1-50. Tihen, J. 1964. Tertiary changes in the herpeto- faunas of temperate North America. Senckenber- giana Biol. 45:265-279. Tihen, ]. 1972. The fossil record, pp. 8-13 in Blair, W. F. (ed. ). Evolution in the genus Bufo. Univ. Texas Press, Austin, 457 p. Van Denvender, T. 1977. Observations on the Ar- gentine iguanid lizard Leiosaurus bellii Dumeril and Bibron ( Reptilia, Lacertilia, Iguanidae). J. Herpetol. 11:238-241. Van Houten, F., Travis, R. 1968. Cenozoic de- posits of the Upper Magdalena Valley, Colombia. Bull. Amer. Soc. Petrol. Geol. 52:675-702. Vergnaud-Grazzint, C. 1966. Les amphibiens du Miocene de Beni-Mellal. Notes Serv. Geol. Ma- roc 27 (198): 4.3-69. Vercnauu-Grazzini, C. 1968. Amphibiens pleisto- cenes de Bolivie. Bull. Soc. Geol. France, (7) 10:688-695. Vilas, J., Valencio, D. 1978. Paleomagnetism of the South American and African rocks and the age of the South Atlantic. Rev. Brasil. Geocien- cias 18:3-10. Vilas, J., Valencio, D. 1979. Paleomagnetism of South American rocks and the Gondwana Conti- nent. Seminar on Past configuration of Gondwana and geological correlation through time. IV In- termit. Gondwana Symp., Calcutta, 1977 (in press ) . Warren, J. 1969. Chelid turtles from the mid-Ter- tiary of Tasmania. J. Paleontol. 43:179-182. Wieland, G. 1923. A new Parana Pleurodira. Amer. J. Sci. 5:1-14. Williams, E. 1950. Testudo cubensis and the evo- lution of Western Hemisphere tortoises. Bull. Amer. Mus. Nat. Hist. 95:1-36. Williams, E. 1956. Podocnemis bassleri, a new species of pelomedusid turtle from the late Terti- ary of Peru. Amer. Mus. Novit. (1782): 1-10. 1979 BAEZ & GASPARINI: FOSSIL RECORD 51 Wood, R. 1970. A review of the fossil Pelomedusi- dae ( Testudines, Pleurodira ) of Asia. Breviora (357): 1-23. Wood, R. 1971. The fossil Pelomedusidae (Testu- dines, Pleurodira) of Africa. PhD Dissert. Har- vard Univ., 345 p. Wood, R. 1972. A fossil pelomedusid turtle from Puerto Rico. Breviora (392): 1-13. Wood, R. 1975. Redescription of "Bantuchelys" con- golensis, a fossil pelomedusid turtle from the Paleocene of Africa. Rev. Zool. Bot. Africaines 89:127-144. Wood, R. 1976a. Two new species of Chelus (Tes- tudines, Pleurodira) from the Late Tertiary of northern South America. Breviora (435): 1-26. Wood, R. 1976h. Stupendemys geographicus, the world's largest turtle. Ibid. (436):1-31. Wood, R., Gamero, M. de. 1971. Podocnemis vene- zueletisis, a new fossil pelomedusid ( Testudines, Pleurodira ) from the Pliocene of Venezuela and a review of the history of Podocnemis in South America. Ibid. (376)1-23. Wood, R., Moody, R. 1976. Unique arrangement of carapace bones in the South American chelid turtle Hydromcdusa maximiliani (Mikan). J. Zool. 59:69-78. Wood, R., Pattersox, B. 1973. A fossil trionychid turtle from South America. Breviora (415):1-10. Zancerl, R. 1947. Redescription of Taphrosphys olssoni, a fossil turtle from Peru. Fieldiana Geol. 10:29-40. APPENDIX Appendix 2:1. — The fossil amphibians and reptiles recorded from the areas shown in figures 4-7 and their cor- responding bibliographic references are listed below; the numbers correspond to those on the maps (Figs. 4-7). The areas have been designated by conspicuous geographic names. Taxonomic entities not recognized in re- cent studies are excluded. CRETACEOUS 1. Alemania, Provincia de Salta, Argentina (Late Cretaceous). anura: Pipidae: Saltcnia ibanezi Reig, 1959 (Reig, 1959; Parodi Bustos et al., 1960; Baez, 1975). 2. Laguna Umayo, Departamento de Puno, Peru (Late Cretaceous). anura: Leptodactylidae ? (Sige, 1968). crocodilia (Sige, 1968). 3. Mossoro, Estado do Rio Grande do Norte, Brasil ( Late Cretaceous ) . testudines: Pelomedusidae: Apodichelys lucianoi Price, 1954. 4. Vila- Vila, Departamento de Cochabamba, Bolivia ( Late Cretaceous ) . testudines: Pelomedusidae: ? Roxochelys vilavilcmis Broin, 1971. 5. Sao Jose do Rio Preto, Estado de Sao Paulo Brasil (Late Cretaceous). testudines: Pelomedusidae: Podocnemis brasilieniis Staesche, 19372 (Price, 1953; Arid and Vizotto, 1966; Broin, 1971); Roxochelys wanderleyi Price, 1953; Podocnemis elegans Suarez, 1969.2 Pieropolis, Estado de Minas Gerais, Brasil (Late Cretaceous). squamata: Sauria: Iguanidae: Prist iguana brasilicnsis Estes and Price, 1973. 6. Northwest of Lago Colhue-Huapi, Provincia del Chubut, Argentina (Late Cretaceous ?). testudines: Meiolaniidae: Niolamia patagonica Ameghino, 1899 (Smith Woodward, 1901; Simpson 1938). PALEOCE NE-EOCE NE 1. Golfo de San Jorge, Provincia del Chubut, Argentina (early Paleocene). testudines: (Gasparini and Baez, 1975; Baez and Gasparini, 1977). crocodilia: Crocodylidae: Necrosuchus ionensis Simpson, 1937 (Gasparini and Baez, 1975). 2. Itaborai, Estado de Rio de Janeiro, Brasil (late Paleocene). anura: Pipidae: Xenopus romeri Estes, 1975 (Estes, 1975a, b); Leptodactylidae (Estes, 1970); Bufoni- dae (Estes, 1970); Hylidae (Estes, 1970). gymnophiona: Caeciliidae: Apodops pricei Estes and Wake, 1972. testudines: Pelomedusidae: Podocnemis sp. (Paula Couto, 1970). squamata: Sauria: Iguanidae (Estes, 1970); Teiidae (Estes, 1970; Paula Couto, 1970); Gekkonidae (Estes, 1970). Serpentes: Boidae (Estes, 1970; Paula Couto, 1970); Aniliidae (Estes, 1970; Hoffstetter and Rage, 1977). crocodilia: Sebecidae: Sebecus sp. (Paula Couto, 1970); Alligatoridae (Paula Couto, 1970). J F. de Broin ( pers. comm. ) considers the assignment of P. brasilicnsis and P. elegans to the genus Podocnemis to be questionable. 52 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 3. Cerro Pan de Azucar, Provincia del Chubut, Argentina (late Paleocene). testudines ( Simpson, 1935). squamata: Serpentes: Boidae: Madtsoia cf. M. bai Simpson, 1933 (Simpson, 1935; Hoffstetter, 1959). crocodilia (Simpson, 1935). 4. Laguna del Hunco, Provincia del Chubut, Argentina (late Paleocene). anura: Pipidae: Xenopus pascuali ( Casamiquela, 1960) (Estes, 1975b; Gasparini and Baez, 1975; Baez, 1976). testudines: Pleurodira (Gasparini and Baez, 1975). 5. Mina Aguilar, Provincia de Jujuy, Argentina (late Paleocene-early Eocene). testudines (Gasparini and Baez, 1975). crocodilia (Gasparini and Baez, 1975). Quebrada de Humahuaca, Provincia de Jujuy, Argentina (late Paleocene-early Eocene). testudines: Pelomedusidae: Podocnemis argentinensis Cattoi and Freiberg, 1958 (Gasparini and Baez, 1975; Baez and Gasparini, 1977). 6. Canadon Hondo, near Paso Niemann, Provincia del Chubut, Argentina (early Eocene). anura: Leptodactylidae: Caudivcrbera casamayorensis (Schaeffer, 1949) (Lynch, 1971). testudines: Meiolaniidae : Crossochelys corniger Simpson, 1937 (Simpson, 1937a, 1938); Chelidae: Hij- dromcdusa sp. (Wood and Moody, 1976). crocodilia: Sebecidae: Sebecus icaeorhinus Simpson, 1937. 7. Lago Colhue-Huapi, Provincia del Chubut, Argentina ( early Eocene ) . crocodilia: Alligatoridae: Eocaiman caverensis Simpson, 1933b. 8. Negritos, Departamento de Piura, Peru (middle Eocene). testudines: Pelomedusidae: Taphrosphtjs olssoni (Schmidt) (Zangerl, 1947; Gaffney, 1975). 9. Divisadero Largo, Provincia de Mendoza, Argentina (late Eocene). testudines (Simpson et ah, 1962). squamata: Serpentes (Simpson et al., 1962). crocodilia: Sebecidae ?: Ilchunaia parca Rusconi, 1946 (Gasparini, 1972). OLIGOCE NE-MIOCE NE 1. Campo Waldo, Departamento de Santander, Colombia (Oligocene).3 crocodilia: Sebecidae: Sebecus sp. (Langston, 1965); Crocodylidae ( Langston, 1965). testudines (Stirton, 1953). 2. Scarrit Pocket, Provincia del Chubut, Argentina (early Oligocene). anura: Leptodactylidae: Caudiverbera caudiverbcra (Linnaeus) (Schaeffer, 1949; Lynch, 1971); Eu- psophus sp. (Schaeffer, 1949); Neoprocoela edentatus (Schaeffer, 1949). 3. Tremembe, Estado de Sao Paulo, Brasil (early Oligocene). testudines: Chelidae (Wood and Patterson, 1973). 4. South of Lago Colhue-Huapi, Provincia del Chubut, Argentina (late Oligocene). anura: Leptodactylidae: Caudivcrbera sp. (Schaeffer, 1949; Baez, 1977). 5. Chaparral, Departamento de Tolima, Colombia (late Oligocene-early Miocene). crocodilia: Alligatoridae: Balancrodus logimus Langston, 1965; Gavialidae (Langston, 1965). 6. Gaiman, Provincia del Chubut, Argentina (late Oligocene). testudines: Testudinidae: Geochelonc gringorum (Simpson, 1942) (Williams, 1950; Auffenberg, 1971; de la Fuente, pers. comm.). 7. Southern Provincia de Santa Cruz, Argentina (early -middle Miocene). squamata: Sauria: Iguanidae (Ameghino, 1899; Gasparini and Baez, 1975; Baez and Gasparini, 1977); Teiidae: Diasemosaurus occidentalis Ameghino, 1893 (Gasparini and Baez, 1975); Serpentes (Ameghino, 1899). 8. Coyaima, Departamento de Tolima, Colombia (late Miocene). testudines: Chelidae: Chelus colombianus Wood, 1976a. squamata: Sauria: Teiidae: cf. Tupinambis (Estes, 1961). crocodilia: Sebecidae: Sebecus sp. (Langston, 1965); Alligatoridae (Langston, 1965); Crocodylidae (Langston, 1965); Gavialidae: ? Gavialis colombianus Langston, 1965. 9. Carmen de Apicala, Departamento de Tolima, Colombia (late Miocene). testudines: Pelomedusidae (Royo y Gomez, 1945-1946; Stirton (1953); Chelidae: Chelus colombianus Wood, 1976a. crocodilia: Alligatoridae: Eocaiman sp. (Langston, 1965); Caiman neivensis Mook, 1941 (Langston, 1965). 3 The Tertiary amphibian and reptile bearing deposits of Colombia are assigned chronologically according to Van Houten and Travis (1968) and Irving (1971); in an earlier paper the authors (1977) followed Stirton (1953). 1979 BAEZ & GASPARINI: FOSSIL RECORD 53 10. Quebrada La Venta, Villavieja, Departamento de Huila, Colombia (late Miocene). anura: Bufonidae: Bufo marinus Linnaeus (Estes and Wassersug, 1963). testudines: Pelomedusidae: Podocnemis expansa ( Schvveigger, 1912) (Medem, 1966, 1968); Chelidae: Chelus colombianus Wood, 1976a; Testudinidae: Geochelone (Chelonoides) hesterna Auffenberg, 1971; Em- ydidae (Medem, 1968). squamata: Sauria: Iguanidae (Estes, 1961); Teiidae: Tupinamhis cf. T. tequixin (Estes, 1961); Dra- caena colombiana Estes, 1961; Serpentes: Aniliidae: Colombophis portai Hoffstetter and Rage, 1977; Boi- dae: Eunectes stirtoni Hoffstetter and Rage, 1977; Colubridae ( Hoffstetter, 1967b; Hoffstetter and Rage, 1977). crocodilia: Sebecidae: Sebecus huilensis Langston, 1965; Scbecus sp. (Langston, 1965); Alligatoridae: Eocaiman sp. (Langston, 1965); Caiman neivensis Mook, 1941 (Langston, 1965); Caiman cf. C. latirostris ( Daudin, 1802) (Langston, 1965; Baez and Gasparini, 1977); Crocodylidae: Charactosuchus fieldsi Lang- ston, 1965; Nettosuchidae: Mourasuchus atopus Langston, 1965 (Langston, 1966); Gavialidae: cf. Rham- phostomopsis (Langston, 1965). 11. North of Lago Buenos Aires, Provincia de Santa Cruz, Argentina (late Miocene). anura: Leptodactylidae: Caudiverbera caudiverbera Linnaeus ( Casamiquela, 1958; Lynch, 1971). 12. Barranca de los Loros, Provincia de Rio Negro, Argentina (late Miocene). anura: Leptodactylidae: Caudiverbera caudiverbera Linnaeus (Casamiquela, 1963; Lynch, 1971). 13. Ingeniero Jacobacci, Provincia de Rio Negro, Argentina (late Miocene). anura: Leptodactylidae: Wawelia gerholdi Casamiquela, 1963. PLIOCENE-PLEISTOCENE 1. Urumaco, Estado de Falcon, Venezuela (middle Pliocene). testudines: Pelomedusidae: Stupetidemys geograpliicus Wood, 1976b; Chelidae: Chelus lewisi Wood, 1976a; Trionychidae (Wood and Patterson, 1973); Testudinidae (Wood and Patterson, 1973). crocodilia: Alligatoridae: Melanosuchus fisheri Medina, 1976; Crocodylidae ?: Gryposuchus sp. (Pat- terson, pers. comm.); Gavialidae: Ikanogavialis gameroi Sill, 1970; Nettosuchia: Mourasuchus amazon- ensis Price, 1964 (Patterson, pers. comm.). 2. Valle de Santa Maria, Provincia de Catamarca, Argentina ( middle Pliocene ) . testudines: Testudinidae: Geochelone gallardoi ( Rovereto, 1914) (Auffenberg, 1974). 3. Parana, Provincia de Entre Rios, Argentina (middle-late ? Pliocene). testudines: Testudinidae : Geochelone sp. (Gasparini and Baez, 1975); Chelidae (Wieland, 1923). squamata: Sauna: Teiidae (Ambrosetti, 1890; Ga;parini and Baez, 1975; Baez and Gasparini, 1977); Ser- pentes: Boidae (Bravard, 1858; Burmeister, 1883, 1885). crocodilia: Alligatoridae: Caiman latirostris (Daudin, 1802) (Gasparini and Baez, 1975); C. australis (Burmeister, 1885); cf. C. jacare (Daudin, 1802) ( Gasparini and Baez, 1975); C. sp. (Gasparini and Baez, 1975; Baez and Gaspirini, 1977); Gavialidae: Rhamphostomopsis neogaeus (Burmeister, 1885) (Rusconi, 1933, 1935; Gasparini, 1968). 4. Monte Hermoso, Provincia de Buenos Aires, Argentina (late Pliocene). anura: Leptodactylidae: Ceratophrys prisca Ameghino, 1899 (Rovereto, 1914); Bufonidae (Gasparini and Baez, 1975). testudines: Testudinidae: Geochelone gallardoi (Rovereto, 1914) (Auffenberg, 1974). squamata: Sauria: Teiidae: Tupinamhis sp. (Rovereto, 1914); Callopistes bicuspidatus Chani, 1976. 5. Rio Quequen Salado, Provincia de Buenos Aires, Argentina (late Pliocene). anura: Bufonidae: Bufo pisanoi Casamiquela, 1967. 6. Chapadmalal, Provincia de Buenos Aires, Argentina (late Pliocene). anura: Bufonidae: Bufo pisanoi Casamiquela, 1967; Leptodactylidae: Ceratophrys sp. ( Reig, 1958). 7. Rio Jurua, Estado do Acre, Brasil (Plio-Pleistocene). testudines: Pelomedusidae: Podocnemis sp. (Paula Couto, 1970); Chelidae: Chelus sp. (Paula Couto, 1970); Testudinidae: Geochelone sp. (Paula Couto, 1970). squamata: Serpentes (Paula Couto, 1970). crocodilia: Alligatoridae: Brachygnatosuchus brasiliensis Mook, 1921; Purussaurus sp. (Paula Couto, 1970); Crocodylidae ?: Gryposuchus (Paula Couto, 1970); Gavialidae: ? Gavialis (Paula Couto, 1970; Baez and Gasparini, 1977); Nettosuchidae: Mourasuchus sp. (Paula Couto, 1970); Mourasuchus amazon- ensis Price, 1964. 8. Rio Aguaytia, West of Rio Ucayali, Peru (Pliocene ?; Pleistocene ?). testudines: Pelomedusidae: Podocnemis bassleri Williams, 1956. 9. Arroyo Perico Flaco, branch of Rio Negro, Departamento de Soriano, Uruguay (Pleistocene). anura: Leptodactylidae: Leptodactylus sp. (Mones, 1975). 54 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Geoemyda ( = Callopsis) sp. Dicrodon (Hoffstetter, 1970b). Bufonidae: Bufo cf. B. marinus 10. Peninsula de Santa Elena, Provincia de Guayas, Ecuador (late Pleistocene). testudines: Testudinidae: Geoclielone sp. (Hoffstetter, 1970b); Emydidae: (Hoffstetter, 1970b). squamata: Sauria: Iguanidae: -Iguana ( Hoffstetter, 1970b); Teiidae: crocodilia: Alligatoridae (Hoffstetter, 1970b). 11. Tarija, Departamento de Tarija, Bolivia (late Pleistocene). anura: Leptodactylidae; Ceratophrys sp. ( Vergnaud-Grazzini, 1968); horribilis Weigmann (Vergnaud-Grazzini, 1968). squamata: Sauria: Teiidae: Tupinambis tequixin (Hoffstetter, 1963). 12. Nuapua, near Carandaiti, Bolivia — Horizon 1 (middle Pleistocene). testudines: Testudinidae: Geoclielone sp. (Hoffstetter, 1968). Nuapua, near Carandaiti, Bolivia — Horizon 2 (late Pleistocene). anura: Leptodactylidae: Leptodactylus cf. L. ocellatus (Linnaeus) (Vergnaud-Grazzini, 1968); Cera- tophnjs cf. C. oiuata (Bell) (Vergnaud-Grazzini, 1968); Bufonidae: Bufo cf. B. paracnemis (Vergnaud- Grazzini, 1968). squamata: Sauria: Teiidae: Tupinambis tequixin (Linnaeus) (Hoffstetter, 1968); Serpentes: Boidae (Hoffstetter, 1968); Colubridae (Hoffstetter, 1968); Viperidae (Crotalinae) (Hoffstetter, 1968). amphisbaenia: Aiuphisbaenidae : Leposternon ? (Hoffstetter, 1968). 3. Herpetofaunal Relationships Between Africa and South America Raymond F. Laurent0 Investigator Titular Fundacion Miguel Lillo Miguel Lillo 205 4000 Tucumdn, Argentina At the height of the Matthewsian theory of continental biogeography ( Matthew, 1915; Darlington, 1957), the very title of this paper would have been almost preposterous, at least in the influential herpetological centers of North America dominated by Noble (1931), Dunn (1923, 1931), and Schmidt (1946). The dissident voices of Jeannel (1942), Du Toit (1937) and others were stifled as inconse- quential. Africa and South America were sup- posed to have had no relationships whatever for a very long time. Indeed, the differences between the Ethiopian and Neotropical her- petofaunas are striking and seem to support the idea of independent histories. Most domi- nant groups are represented in Africa and South America by pairs of adaptive equiva- lents or vicarious groups in Simpson's (1965) sense (Table 3:1). Faunal similarities between Africa and South America have been explained by ex- tinctions in the Holarctic Region. In some cases the fossil record seems to bear out this explanation (e.g., turtles of the family Pelo- medusidae). When such paleontological evi- dence was lacking, it often was implied; Dunn ( 1931 ) emphasized that it would not be sane reasoning to postulate a trans-Atlantic land bridge just because no fossils were known from the northern continents. When Dunn made his statement the fossil record was Table 3:1. — Family Groups of Amphibians and Rep- tiles that have Trans-Atlantic Counterparts. South America Africa Leptodactylidae Hylidae Iguanidae Teiidae Boinae Crotalinae Ranidae (sensu Liem, 1970) Hyperoliidae (sensu Liem, 1970) Agamidae + Chamaeleontidae Lacertidae Pythonini Viperinae Investigator Principal del CONICET. much poorer than it is now. Presently, some credence can be given to negative evidence, namely, the lack of fossils, in North America and Europe of groups that elsewhere have extensive fossil records (e.g., turtles and crocodilians). New evidence definitely points to the union of Africa and South America into a sin- gle continent from at least the Late Carbonif- erous until the Cretaceous. Then a graben formed a narrow gulf north and south of a residual bridge between northeastern Brasil and Nigeria. During the last half of the Tu- ranian, the continents were split with the birth of the South Atlantic Ocean, which initially was quite narrow. This scenario was de- scribed by Reyment ( 1975 ) and is supported by convincing geological evidence — paleo- magnetism ( Creer, 1973 ) , fit of the continents ( Dietz and Holden, 1970 ) , sea-floor spreading (Heirtzler, 1968; Francheteau, 1973), and plate tectonics with the fitting of cratons and rocks ( Hurley, 1968; Dietz and Holden, 1970; Douglas et al., 1973). The stratigraphic con- cordances are especially striking (Reyment and Tait, 1972), as well as the structure of the coastal basins with their immense amounts of fresh water sediments (Martin, 1968), which suggest a graben phase like that of the African Rift Valleys. The salt deposits in An- gola, Gabon and Brasil are reminiscent of later phases like that of Lake Turkana or the Red Sea (Reyment, 1975). Paleontological data are still more con- vincing (Baez and Gasparini, this volume). The amphi-Atlantic distribution of the meso- saurians of the early Permian (Romer, 1966) is strong evidence for a Gondwanan land mass ( Colbert, 1973 ) . The early Triassic of Argen- tina has revealed a Cynognathus fauna almost identical to that of the upper Beaufort beds of South Africa (Bonaparte, 1967). Two 55 56 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 groups of mammals provide evidence that the Atlantic was not the wide ocean of today in the Eocene and Oligocene. Primates and cav- iomorph rodents, conspicuously lacking in the early South American mammalian fauna, nev- ertheless entered the continent long before the Late Cenozoic invasion via Panama. Classically, it was assumed that they entered South America by island-hopping from North America (Wood, 1950; Simpson, 1965; Patter- son and Pascual, 1968). However, another hypothesis assuming an African origin and a trans-Atlantic migration has been maintained by Lavocat (1974) and Hoffstetter (1972, 1977). Krommelbein (1971) showed that the fossil, fresh water ostracods from the coastal basins of Reconcavo, Sergipe, Brasil, and of Gabon are almost identical. Two different am- monite faunas (one in the Potiguar Basin in northern Brasil and another in the Sergipe- Alagoas Basin) existed until the lower Turan- ian. Then the African-Brasilian isthmus disappeared, and the two faunas mingled (Beurlen, 1961; Reyment, 1958, 1970). These data provide evidence for the birth of the Atlantic Ocean and the separation of Africa and South America in the middle Turanian, about 90-95 m.y.b.p. This establishes a firm basis on which to proceed to determine which groups existed in both continents before their separation and which emigrated from one to the other before or after their separation. The Hennigian sys- tematists and biogeographers ( Brundin, 1966; Croizat, 1964; Nelson, 1973; Croizat et al., 1974) have insisted on the necessity of vi- cariance events in determining generalized tracts in reconstructing biogeographic his- tories. Their conclusions are in general agree- ment with plate tectonics, mainly because their work has been based on these premises rather than because they emphasize vicari- ance at the expense of geocenters and dis- persal. (Appendix 3:1). Their approach is applicable to the study of relationships be- tween the African and South American faunas. There was an old African-Brasilian (Inabre- sian, fide Jeannel, 1942) fauna, and there also are vicariant groups. SUMMARY OF DISTRIBUTION PATTERNS The first step in a biogeographic analysis must be a summary of distribution patterns or "generalized tracts" (Croizat, 1964). These patterns are listed below. 1. Gondwanan or West Gondwanan groups. — Geotry petes- Apodops, Pipi- dae, Pelomedusidae, Amphisbaenidae, Typhlopidae, Lcptotyphlopidae. 2. South American groups that invaded Africa before the separation of the con- tinents.— Bufonidae, Iguanidae ( ex- tinct in Africa but surviving in Mada- gascar). 3. Presumed Indian groups that invaded South America via Africa before the separation of the continents. — Microhy- lidae. 4. African groups that invaded South America after the separation of the continents. — Gavialidae (?), Gekkoni- nae, Scincidae, Amphisbaenidae (?), Colubrinae ( ? ) . 5. Holarctic groups tliat invaded Africa and South America from the north. — Testudinidae, Crocodylidae, Colubri- nae (?). 6. African groups that recently invaded South America by a northern route. — Ranidae. 7. Neotropical groups absent from Afri- ca.— Rhinatrematidae. Dermophiinae,1 Caeciliidae, Typhlonectidae, Bolito- glossini, Leptodactylidae,-' Hylidae,3 Centrolenidae, Pseudidae, Dendrobati- dae, Sphacrodactylinae, Iguanidae, Teiidae, Anguidae, Boini, Xenodonti- nae, Micrurinae, Crotalinae. 1 The reasons for splitting the Caeciliidae and recog- nizing the Herpelinae are given by Laurent (in press ) . - The African Heleophryninae are considered to be members of the Myobatrachidae by Lynch (1973). 3 The Hylidae, as well as other groups, like Disco- glossidae, Pelobatidae, Salamandridae, and Angui- dae, are present in northern Africa. That part of Africa belongs to the Palaearctica Region and is not considered to be relevant here, although the past existence of some of its fauna in the Ethiopian Re- gion cannot be ruled out entirely. 1979 LAURENT: AFRICA AND SOUTH AMERICA 57 8. African and Old World groups absent from South America. — Scolecomorphi- dae, Herpelinae, Heleophryninae, Hy- peroliidae, Agamidae, Chamaeleonti- dae, Lacertidae, Cordylidae, Varanidae, Pythonini, Lycodontinae, Dasypeltinae, Elapidae, Viperinae. 9. Groups that apparently once lived in Africa but are now extinct there. — Igua- nidae. 10. Groups that presumably icere pan- Gonduanan in the Jurassic. — Leiopel- matidae. Two patterns are more prevalent than others — groups present in South America but not in Africa, and present in Africa but not in South America. A third pattern involves groups that are present in both continents. RELATIONSHIPS OF PATTERNS TO CONTINENTAL HISTORIES Amphibians Caecilians. — The distribution of caecilians indicates that they are a typical Gondwanan group. Estes and Wake (1972) described the only known fossil caecilian, Apodops, from the Paleocene of southeastern Brasil. The fossil tends to confirm a Gondwanan distribution, especially because Apodops resembles the Af- rican Geotrypetes and may be closely related to it. Possibly Geotrypetes is more archaic than either of the primitive families Rhinatremati- dae and Ichthyophiidae. It is said to have a free, dentate ectopterygoid, a bone generally lost and always edentulous in other caecilians. Furthermore, it has two pairs of openings in the skull (temporal and interpterygoidal ) . According to the Lissamphibian hypothesis, such vacuities should be primitive and their disappearance a secondary adaptation to fos- sorial habits.4 ' This seems more likely than the theory- of Carroll and Currie (1975), who related the Gymnophiona to microsaurians, because the loss of so many cranial bones is more easily visualized as the result of fene- stration than an effect of further solidification of an already continuously roofed skull. Geotrypetes has as many chromosomes as the ichthyophiids, which also include the In- dian Uraeotyphlus (Nussbaum, pers. comm. ). Also, caecilians, like other lissamphibians, demonstrate a negative correlation between chromosome number and the number of de- rived character states (Laurent, in press). Thus, Apodops might belong to an African- Brasilian group of primitive caecilians, from which the African and Neotropical caecilians descended. By this reasoning, the other cae- cilians are actually different in South America and Africa. The South American Rhinatrema- tidae is more primitive than the Ichthyophii- dae (Nussbaum, 1977). The aquatic South American Typhlonectidae and the fossorial African Scolecomorphidae are specialized groups apparently derived from old stocks. The remaining genera have been placed in the Caeciliidae. The specialized Caecilia and Oscaecilia should be separated from the bulk of the family (Laurent, in press). The other genera can be divided into an Old World subfamily Herpelinae with splenial teeth (ex- cept in the specialized Boulengerula) and a New World subfamily Dermophiinae without splenial teeth (except in Gymnophis, there- fore deemed primitive ) ( Laurent, in press ) . These subfamilies are most likely sister groups. Salamanders. — All South American sala- manders are members of the Plethodontidae and obvious immigrants from Central Amer- ica (Wake, 1966). However, I must stress a recent discovery, as interesting as it is unex- pected, of a salamander in Senonian (Upper Cretaceous) beds of Niger (de Broin et al., 1975). The genus and even the family have not been determined. The age is not fixed precisely, because the Senonian is a long period following the Turanian (85 m.y.b.p.) and lasting until the end of the Mesozoic ( Maastrichian, about 65 m.y.b.p.). Salaman- ders have been considered as exclusively Hol- arctic. Therefore, their presence in equatorial Africa, perhaps quite soon after the formation of the Atlantic Ocean, indicates the possibility of an older salamander fauna in Africa and perhaps in South America. Archaeobatrachians. — The only living archaeobatrachians in South America are the 58 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 pipids, which apparently are an early special- ized derivative from Jurassic frogs. The pipids have a typical west Gondwanan distribution and are now common to Africa and South America. They are known from both conti- nents by a number of Mesozoic and Cenozoic fossils, beginning with the Early Cretaceous of Israel (Nevo, 1968). It has been argued that the pipids might have lived in the north- ern continents and subsequently migrated in- dependently into South America and Africa. However, the family is conspicuously absent in the fossil record in the north; it is replaced there by the ecologically similar palaeobatra- chids, now extinct, but apparently common from the Jurassic to the Pliocene.5 Estes (1975a) reported a fossil of Xenopus from the Paleocene of Brasil and referred the Eocene Shelania from Patagonia to the same genus. Therefore, there is little reason to doubt the Inabresian0 origin of the family. Estes (1975b) synthesized the phylogeny of the pipids. Of the two lower Cretaceous gen- era from Israel, Cordicephalus seems to be ancestral to Xenopus, whereas Thoraciliacus seems to be an early specialized derivative of the primitive stem of the family. No other fossil is known from the Cretaceous before the separation of Africa and South America. In the Late Cretaceous (Senonian, 80-85 m.y.b.p. ) there is Saltenia in northwestern Argentina and Xenopus and other pipids simi- lar to Pipa or Hijmenochirus in western Af- rica. Eoxenopoides from southwestern Africa (about the limit of the Mesozoic and Creta- ceous, 65 m.y.b.p.) is a specialized derivative of Xenopus. All more recent fossils are re- ferred to Xenopus — two in South America [Paleocene (±60 m.y.b.p.) of southeastern Brasil (Estes, 1975a) and Eocene (±50 m.y.b.p.) of Patagonia (Casamiquela, 1961)] and two in Africa [Miocene ( ±20 m.y.b.p) of southwestern Africa (Ahl, 1926) and of Mo- = According to some workers (e.g., Estes and Reig, 1973), the Palaeobatrachidae is related to the Pipi- dae, but Vergnaud-Grazzini and Hoffstetter (1972) believed that the similarities are the result of con- vergence. However, Estes (1975b) argued con- vincingly that they are in the same superfaniily. "The term "Inabresia" was coined by Jeannel (1942) for the African-Brasilian continent. rocco (Vergnaud-Grazzini, 1966)]. Xenopus presently is speciose in Africa, where the spe- cialized Hijmenochirus and Pseudhymenochi- rus occur in forests. Xenopus is extinct in South America, but perhaps through a Sal- teniaAike stock it gave rise to the modern South American pipids, now restricted to for- ested regions of northern and eastern South America (Estes, 1975b). Other archaeobatrachians have lived in South America ( Baez and Gasparini, this vol- ume). Vieraella (Lower Jurassic) and Noto- batrachus (Upper Jurassic) were placed in the Leiopelmatidae by Estes and Reig ( 1973 ) , who suggested that the leiopelmatids radiated during the Jurassic in Gondwana- land and that Ascaphus is a remnant of a northward migration at the end of the Meso- zoic. Leiopelma survived in New Zealand, where it is the only frog. Discovery of fossil leiopelmatids in other parts of Gondwana- land, especially Africa, is expected. Neobatrachians. — Estes and Reig (1973) and Laurent ( in press ) believed that the neo- batrachians are a Gondwanan group. Possibly the neobatrachians were derived from the southern Discoglossoidea ( Leiopelmatidae ) through a grade exemplified by the Australian family Myobatrachidae, which perhaps for- merly had a pan-Gondwanan range. Among the neobatrachians, only the Bu- fonidae, Microhylidae and Ranidae are pres- ent on both Africa and South America. Of these, the ranids are represented in South America only by Rana pahnipes, a Central American species of recent entry. According to Noble (1931), the bufonids and microhy- lids originated in the Holarctic and subse- quently invaded the southern continents. Blair ( 1972b ) suggested that Bufo originated in South America. Estes and Reig (1973) re- ported Bufo from the Paleocene of Brasil, whereas no other bufomd fossils are known before the Miocene or Oligocene (Hecht, 1963). The fossil record and the rich Neo- tropical radiation of bufonids (9 genera, more than 120 species) support Blair's conclusions. From a South American center of origin, three avenues of bufonid dispersal are con- ceivable— 1) through Antarctica and Austral- ia, 2) through Africa, and 3) through North America. The first obviously is out of the 1979 LAURENT: AFRICA AND SOUTH AMERICA 59 question, because the Australian region is de- void of bufonids. The last was suggested bv Rlair (1972b). Laurent (1972, 1975) did not reject dispersal through North America but proposed that bufonids also dispersed through Africa. Laurent was influenced by Estes' ( 1970, pers. coram. ) insistence that no bufo- nid entered Nortli America before the Mio- cene. There is further evidence in favor of an African dispersal route. 1) The South Amer- ican bufonid radiation is the largest (Trueb, 1971; McDiarmid, 1971; Cei, 1972) and the African is next with seven genera and about 50 species (Tihen, 1960; Tandy and Keith, 1972), followed by Eurasia with only six gen- era and about 40 species and finally North America with one genus and some 20 species. 2) Few bufonids have retained an omoster- num, a plesiomorphic character for the fam- ily. These include the Bufo haematiticus group in northern South America, the African genus Nectophrynoides, and possibly Wemer- ia in west Africa.7 3) Relationships between the Neotropical and African bufonids is sup- ported by the high degree of genetic compat- ibility between the South American Bufo arenarum and the African B. regularis (Blair, 1972a), by the striking similarity of peculiar species like the African B. superciliaris and the Neotropical B. blombergi (Blair, 1972b), and by serological affinities (Cei, 1977). Such evidence induced Laurent (in press) to em- phasize the African route rather than the North American one. A dispersal following the described route need not exclude a Mio- cene invasion of North America from South America, but the African dispersal was much earlier (±90 m.y.b.p. versus ±25 m.y.b.p.) and therefore much more important to the evolutionary biogeography of the family. The systematic position of the Microhyli- dae is the most controversial matter in the taxonomy of frogs. Boulenger (1882) con- sidered them (as the Engystomatidae) to be related to the Ranidae, because of their firmi- 'Andersson (1903) mentioned the presence of a vestigial omosternum in his description of Steno- glossa, a synonym of Wemeria, but Amiet (1976) said that the omosternum is absent in the genus. Nonetheless, the species of Wemeria resemble toads of the Bufo haematiticus group. sternal pectoral girdle. Noble (1931) sup- ported Boulenger. Orton (1957) emphasized the apparent primitiveness of the microhylid tadpoles, which are similar in many respects to those of pipids. Orton believed that it was unlikely that such an adaptive complex of features as the larval mouth in most anurans would be lost; therefore, she thought that the microhylids were related to the pipids and represented an early radiation among frogs. This contention was resisted by several herpe- tologists beginning with Griffiths (1963). The traditionalists include Griffiths and Car- valho (1965), Tihen (1965), and Kluge and Farris (1969). The "Ortonists" include Hecht (1963), Inger (1967), and Starrett ( 1973 ) , who based her opinion on a detailed study of tadpoles. Savage (1973) enthusi- astically based a new zoogeographic scheme on the apparent strengths of Starrett's conclu- sions, and even included Australia in the original Gondwanan realm of the family ( see Tyler, this volume, for contrary zoogeographic arguments ) . Lynch ( 1973 ) showed that such an evolutionary scheme of the microhylids re- quired the independent acquisition of no less than 13 characters present in the ranids. In a detailed study of tadpole structure, Sokol ( 1975 ) demonstrated that the microhylids had lost the larval papillae, denticles, and horny beaks.- Therefore, the microhylids are related to the ranoids. Microhylids have a semirelictual distribu- tion, intermediate between the scattered pat- terns of some old families, such as the Disco- glossidae and Pelobatidae, and the compact patterns of the more modern, still radiating groups, like the Ranidae and Bufonidae. Therefore, the microhylids must be older than the ranids. This idea is supported by the pres- ence of 28 chromosomes in the microhylid Kaloula; the number is not the result of sec- ondary fusions and therefore is a plesiomor- phic feature similar to the karyotype of Dk- coglossus. The presence of a variety of micro- hylids in South America also supports the ' Blommers-Schlbsser (1975) confirmed Sokol's con- clusions by discovering that in the Scaphiophryninae, the most primitive subfamily of microhylids, the tadpoles have papillae and a slightly sinistral spiracle (median in other subfamilies). 60 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 antiquity of the family; the ranids barely enter South America. Assuming that the microhylids evolved, like the ranoids, in the eastern part of Gond- wanaland, their radiation was around the Indian Ocean ( tropical Asia, East Indies, and Madagascar). The most primitive subfamilies ( Scaphiophryninae and Dyscophinae) live in Madagascar and tropical Asia. The eastern groups in New Guinea ( Sphenophryninae and Asterophryinae) are not relevant here, but the other subfamilies are — Rrevicipinae, Hop- lophryninae9 and Phrynomerinae in Africa, Cophylinae in Madagascar, and Microhylinae in Asia and America. We can dismiss as improbable an Holarc- tic origin of the family, for the primitive sub- families live in the tropics. Thus, only Mada- gascar and the Gondwanan part of Asia are likely centers of origin of the microhylids. Both are possible, for during its northward drift India apparently was connected at times with Madagascar by garlands of islands (Mc- Kenzie and Sclater, 1973), some of which continue to exist as the Seychelles, Amirante, Mascarene, Maldive and Laccadive Islands (Appendix 3:2). Later, contacts were with Malaysia and Indochina through the Nicobar and Andaman islands, allowing the Indian fauna to invade eastern Asia, Indonesia, and even the East Indies. The African invasion likely passed through the Mozambique Chan- nel and its islands (e.g., Comores). The Af- rican groups of microhylids are now highly differentiated. The American microhylines are a problem. According to the Matthewsian theory, the microhylines originated in tropical Asia, in- vaded the eastern Palaearctic Region, and passed into North America by the Bering isth- mus and into South America by island-hop- ping well before the Pliocene. There are seri- ous objections to this hypothesis. In Asia, as well as in America, the primitive genera hav- ing a complete pectoral girdle are in the trop- ics; these are Kalophrynus, Chaperina, Me- lanobatrachus and Gastrophrynoides in Asia and Otophryiie and Dermatonotus in South 0 Mclanohatrachus is included in the Microhylinae (Savage, 1973; Laurent, in press). America. The genera living in temperate Asia (Microhijla) and North America {Gas- trophryne) have reduced pectoral girdles. Carvalho (1954), Nelson (1966), Nelson and Cuellar ( 1968 ) questioned the affinities be- tween the Asiatic and American microhylines. There is a possible trans-Gondwanan path- way from India to Madagascar to Africa to South America that is marked by a series of genera having complete pectoral girdles — Kalophrynus (tropical Asia), Mclanohatra- chus (India), Scaphiophryninae and Dysco- phus (Madagascar), Brevicipinae and Parho- plophryne (Africa) and Otophryne and Dermatonotus (South America). If such a dispersal took place, it might have occurred before or slightly after the birth of the At- lantic Ocean. Chromosome numbers support this hypoth- esis ( Morescalchi, 1973; Bogart and Nelson, 1976; Bogart et al., 1976). In Asia, Kaloula has 28 chromosomes, which equals the rela- tively primitive number of Discoglossus; other Asian genera ( Uperodon, Ramanella, Micro- hijla ) are known to have 26 chromosomes. In Africa there are 26 in Phrynomerus and 24 in Breoiceps. In America there are 26 chromo- somes in the primitive genera Otophryne and Glossostoma, 24 in Chiasmocleis and 22 in seven genera, including the widespread Gas- trophryne and Elachistocleis. Considering now only those families that are present on one side of the Atlantic, we see some striking parallelisms. South American leptodactylids, rhinodermatids and dendro- batids are paralleled by the terrestrial ranids and hyperoliids in Africa; the hylids and cen- trolenids in South America are paralleled by the arboreal ranids (Chiromantis) and hyper- oliids in Africa. Moreover, some peculiar adaptations in one continent have counter- parts in the other — the aquatic South Amer- ican pseudids (coexisting with pipids) versus African pipids; the rheophilous South Ameri- can telmatobiines versus African heleophry- nines; atelopine bufonids and brachyecpha- lids in South America versus Didynamipus in Africa. In some cases the resemblances are striking. For example, compare Physalacmus biligonigerus in South America with Tomop- terna delalandii (and congeners) in Africa; compare the South American Leptodactylus 1979 LAURENT: AFRICA AND SOUTH AMERICA 61 fuscus with African species of Ptycltadena, and the Hyla leucophyllata group in South America with species of Afrixalus. On the other hand, some adaptations are unique to one continent. Africa has nothing like the marsupial tree frogs (Amphignathodontinae); South America has no frog emulating the sex- ual dichromatism of the tribe Hyperoliini (for other examples, see Laurent, 1973). Reptiles Chelonians. — The Pelomedusidae is a clas- sical case of an amphi-Atlantic distribution, which has been explained by Matthewsians as an Holarctic origin and southward migra- tions. Others have explained the distribution by the fragmentation of a primitively Gond- wanan range. The alternatives are not clear, for there are northern fossils. Some very old turtles from Germany generally classified in other suborders are really pleurodires (de Broin, pers. comm.). These are the Triassic Proterochersis ( Proganochelydia ) and the Ju- rassic Platychelys ( Amphichelydia).10 Quite an array of other northern genera are known from Upper Cretaceous to Oligocene beds be- longing to the shores of the young Atlantic Ocean. According to de Broin (pers. comm.), the marine coastal Bothremydidae is a sister family of the Pelomedusidae. The oldest pelo- medusid fossil is Platycheloides from the Low- er Cretaceous of Africa. Thus, the family was in existence before the birth of the Atlantic Ocean. Later, the exclusively African pelo- medusines (not known before Oligocene) were separated from the Podocneminae, which flourished in South America, Africa, Europe (Neochelys, Eocene-Oligocene), and even India (Schwoeboemys, Pliocene), and survived in Africa until the Pleistocene, in South America and in Madagascar (Erymno- chelys). The Cryptodira, rather common in Laura- sia in Jurassic and Cretaceous times, are pres- ent in Africa and South America; they seem to be rather recent immigrants into South America — Oligocene for Geochelone (Simp- son, 1942), Miocene for Emydidae (Medem, 1968), a short Pliocene apparition for the Tri- onychidae in Venezuela (Wood and Patter- son, 1973). Only the Testudinidae and Tri- onychidae became established in Africa, where they are known since the Miocene (Romer, 1966). Crocodilians. — The living crocodilians do not show significant similarities between South America and Africa. The Crocodylidae is present in both continents but relatively unimportant in South America; the Alligator- idae, absent from Africa, radiated impressive- ly in South America. However, before the formation of the Atlantic Ocean in the Creta- ceous, the crocodilian fauna, composed ex- clusively of mesosuchians, was much the same in South America and Africa. Thus, the Afri- can Libycosuchidae were small, blunt-snouted crocodiles, very similar to the South American Notosuchidae (Sill, 1968; Buffetaut, 1976)." The gigantic, long-snouted pholidosaurid ge- nus Sarcosuchus was common to Brasil and west Africa (Buffetaut, pers. comm.). After the severance of the last remnants of a bridge, the faunas gradually became different. Al- though the mesosuchians were dominant over the eusuchians until well into the Cenozoic in the scattered Gondwanan continents, they became subordinate to them in Laurasia ( Buf- fetaut, pers. comm.). The Dyrosauridae (Up- per Cretaceous and Early Cenozoic mesosuch- ians), extremely long-snouted, gavial-like creatures, although essentially African, also lived at the end of the Cretaceous on the west side of the still trench-like Atlantic, but this can be ascribed to their littoral habits (Buf- fetaut, 1976). Other gavial-like crocodilians, now extinct, were part of the Neotropical eu- suchian radiation in the Tertiary. Sill (1968) considered them as Gavialidae, but possibly they are another case of parallel evolution (Baez and Gasparini, this volume). If they are true gavialids, their dispersal through a 'Gaffney (1975) specifically removed Proterochersis from the Proganochelydia, because they have a fused pelvis like the pleurodires; he explicitly in- cluded Platychelys in the Pleurodira. 11 Steel (1973) suggested that the groups might have evolved in parallel, but Buffetaut (1976) and Sill ( 1968) recognized two families while admitting that they probably have a common ancestor. Nopsca (1928) recognized them as subfamilies, but Mook (1936), von Huene (1956) and Romer (1956, 1966) did not even make that distinction. 62 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 still narrow ocean in the Eocene seems to be the only explanation of their distribution, for the family is first known in the Eocene ( Hecht and Malone, 1972) of Africa. Lizards. — Only the Gekkonidae and Igua- nidae are involved in the Afro-American sep- aration. Other families of lizards seem to be parallel radiations after the separation — Aga- midae and Chamaeleontidae similar to the Iguanidae; Lacertidae, Cordylidae and Scin- cidae emulating the Teiidae and some terres- trial iguanids; the Varanidae copied by large teiids, like Tiipinaml)is. Also there is the eco- logical similarity between the numerous "mi- croteiids" and the lygosomine skinks (Lau- rent, 1973). Although some archaic lizards were pres- ent in the Triassic, these apparently have little in common with Jurassic ones, which belong to the modern infraorders (Robinson, 1967; Hoffstetter, 1955, 1967). Some of the modern families were present in the Cretaceous. Thus, when Africa and South America drifted apart, several recent families were already in existence. The iguanid, Pristiguana (Estes and Price, 1973 ) , from the Cretaceous of Bra- sil has some characters of the Teiidae. Chrom- osome morphology is similar in iguanids and teiids (Gorman, 1970). The Gekkonidae may be the oldest family of modern lizards. It existed in Brasil in the Paleocene (Estes, 1970). The Jurassic Ardeo- sauridae presumably is ancestral to the gek- konids and so similar to them (Hoffstetter, 1964) that in a cladistic system they could be included in the gekkonids. Therefore, the presence of gekkonids in western Gondwana- land before the formation of the Atlantic graben is realistic. The relict and disjunct distribution of the Eublepharinae is best explained by a northern origin. On the other hand, the Sphaerodac- tylinae is likely to be a strictly Neotropical derivative of a Gondwanan stock. Most of the South American Gekkoninae seem to have come from Africa by waif dispersal after the formation of the Atlantic Ocean. This is fairly certain and recent for the species common to both continents, such as Ilemidactylus hrooki and //. mabouia (Kluge, 1969) and hardly less obvious for Turentohx (Kluge, 1967; Vanzolini, 1968). This dispersal was easier when the Atlantic Ocean was narrower. Other gekkonine stocks (e.g., Briba and Bogertia) may have entered South America in the Late Cretaceous or Early Cenozoic. Possibly some (e.g., Homonota) immigrated into South America before the separation of the conti- nents. Bons and Pasteur (1977) suggested an early immigration for the two Neotropical species assigned to the African genus Lygo- dactylus. Estes and Price ( 1973 ) believed that the Iguanidae originated in South America when it was still united to Africa and invaded Af- rica, where they became extinct, and Mada- gascar, where they survived. Alternatively, they could have originated in Africa, where the related agamids and chamaeleontids sup- planted them. If the Iguanidae and Teiidae have a common ancestor, the presence of teiids in North America in the Cretaceous, contrasting to the absence of iguanids there (Estes, 1970), is puzzling and cannot be ex- plained with our present data. The past existence of iguanids in Africa is hardly questionable, for they are still living in Madagascar. Is their extinction in Africa the result of competition with agamids? Not likely, because the African agamids are not diverse and therefore unlikely to out-compete the diversified iguanids. Also, it is unlikely that the chameleons out-competed the igua- nids, except for possibly some arboreal types, for the chameleons are a highly specialized group of lizards. Furthermore, the agamids are probably relatively recent immigrants into Africa from Eurasia. Rafting of iguanids and teiids from South America to Africa is not possible now, because of the direction of the ocean currents, but 50-80 m.y.b.p. such an event was more likely. In the reverse direc- tion, the feasibility of a successful crossing has been proved by Ilemidactylus (Kluge, 1969), but no cases are documented for agamids, chamaeleontids, lacertids, or cordylids. If, as indicated by Estes and Price (1973), iguanids and teiids are respectively the roots of the Iguania and Scincomorpha radiations, an eastern invasion of "eoteiids" is suggested. Estes ( pers. comm. ) sees the teiids as an essentially Neotropical radiation with a lacer- toid derivation through northern Gondwana- 1979 LAURENT: AFRICA AND SOUTH AMERICA 63 land. The Scincidae have relationships be- tween the southern Atlantic continents, but the few South American species of Mahmja probably came from Africa long after the sep- aration of the continents. Amphisbaenians.- — The worm-lizards have a typical western Gondwanaland distribution. They are an old group, and their existence in the Inabresian continents is likely. Fossils are known from North America (Eocene to Pleis- tocene) and belong to several extinct genera, as well as to the extant Rhincura and Lcpo- sternon, now surviving only in South America (Romer, 1966). The extinct Omoiotyphlops is from the Eocene-Pliocene of Europe ( Rom- er, 1966; Hoffstetter, 1962). Therefore the range of the amphisbaenids underwent a con- traction similar to that of many tropical groups that lived in Europe and North Amer- ica.12 Snakes. — Among the primitive scoleco- phidians, the Leptotyphlopidae has about the same range as the amphisbaenians and pre- sumably the same history. The Typhlopidae has a pantropical range, which can be deemed pan-Gondwanan until there is contrary evi- dence. The Boidae also is considered to be a Gondwanan group; nonetheless, the present distribution is the result of complex migra- tions. Primitive fossil genera have been found in southern continents — Laparrcntophis in the Lower Cretaceous of northern Africa, Dini- Jysia in Patagonia (Upper Cretaceous), and Madtsoia in South America, Africa, and Mad- agascar (Upper Cretaceous to Paleocene). Presently, the surviving boids in South America and Africa are not closely related. The Neotropical tribe Boini may be de- scended in situ from archaic South American boids, but the African Erycinae and Pythonini probably came from elsewhere. Hoffstetter and Rage (1972) believed that the Erycinae, which may have originated in North America from a South American boine stock, was pres- ent in North America in the Paleocene or earlier and in Europe in the Eocene. Another lineage (Rage, 1977), using the Bering path- way, entered Africa in the lower Miocene. Rage ( pcrs. comm. ) contemplates three pos- sible origins for the Pythonini — Africa or Aus- tralasia, both of which he considers to be doubtful, and Asia, a choice also favored by Underwood (1976, pers. comm.). Thus, boids supposedly migrated from North America to Asia via the Bering land bridge. A fourth possibility is the Indian raft (see Appendix 3:2), which may explain why boids flourish in Australasia and how they later came to Africa from Asia, presumably in Miocene times.13 The least understood of all groups is the vast array of higher snakes, the caenophidians or Colubroidea. Until recently, the fossil rec- ord of caenophidians was exclusively Holarc- tic and only back to the Miocene. New data show that caenophidians existed in Europe in the lower Eocene, and the Colubridae (sensu lato) is known from the middle Oligocene.14 Rage (1975) described Nigerophis from the Paleocene of Africa; this genus seems to be intermediate between the caenophidians and the Palaeophidae. Such a systematic position suggests an aquatic origin of modern snakes and makes their paleogeographic history even more difficult to interpret. Rabb and Marx ( 1973 ) suggested that the group perhaps had a tropicopolitan distribution before the Gond- wanan fragmentation, but Rage (1976) dis- agreed. The scarcity and primitiveness of the few colubroids from the early Cenozoic sup- port Rage, rather than Rabb and Marx. The problem is compounded by the taxo- nomic uncertainty that prevails within the Colubridae (sensu lato). Most attempts to clarify the systematica have resulted in 1) recognition of small groups that can be sep- arated from the bulk of the genera, or 2 ) par- tition of larger groups that are highly contro- versial (Dunn, 1928; Bogert, 1940; Bourgeois, 1968; Underwood, 1967; Dowling, 1975; Smith et al., 1977). Dowling's (1975) classification ' The Oligocene fossils, Changlosaurus and Cnjthio- saurus, from Mongolia are not amphisbaenians (Gans, pers. comm.). "Rage (pers. comm.) still prefers the Bering route, but he does not reject the Indian hypothesis. "A record from the upper Eocene (Rage, 1974) is doubtful (Rage, pers. comm.). 64 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 is beautifully simple, resulting in geograph- ically discrete groups, namely New World Xenodontinae ( following Dunn's scheme ) and Old World Lycodontinae. This concept is in agreement with the probable belated birth and radiation of the family as suggested by the fossil record. Thus, parallel radiations in- to terrestrial, aquatic and fossorial groups took place in South America and Africa. On the other hand, Underwood's ( 1967 ) revolutionary classification, which was severe- ly criticized by others (namely, Dowling, 1975; Hoffstetter, 196S; Smith et al., 1977), implied trans-Atlantic relationships. The simi- larities between the fossorial African Cala- melaps and the South American Atractus, the aquatic Limnophis (or Hydraethiops) with Helicops, or the terrestrial Lycophidion and Oxyrhopus simply may be the result of con- vergence. Of course, this is the most likely hypothesis, but rafting should not be dis- missed off-handedly. Three adaptive types of arboreal colubrids correspond to three groups distinguished by Bourgeois (1968), as follows: 1) large-headed snakes with vertical pupils and slender necks (Boiginae); 2) streamlined but rather robust green snakes with round pupils (Philotham- ninae); 3) very slender snakes with round or horizontal pupils, commonly with pointed heads and venom ( Dispholidinae ) . The simi- larities between the Neotropical Leptodeira and the African Boiga-Dipsadoboa-Crotapho- peltis group are probably only convergence, but on the basis of Bourgeois' ( 1968 ) criteria, the Neotropical Oxybelis can be grouped with the African Thelotornis, in spite of its round pupil.15 Likewise, genera such as Leptophis and Chironius in South America would be- long to the African Philothamninae. Because of the relatively recent develop- ment of the colubrids, it is unlikely that any groups had an Inabresian distribution. Trans- Atlantic rafting may have occurred, but prob- ably only in one direction (Africa to South America). Hoffstetter (1972) convincingly argued that hystricomorph rodents and mon- keys entered South America across the At- ' For osteological reasons, Bourgeois (1968) put Rhamnophis and Thraso))s in the Dispholidinae, notwithstanding their round pupils. lantic Ocean. The reverse migration has never been advocated, except in the beginning of the Atlantic era, when the sea was very nar- row. Now the ocean currents are favorable for westward rafting in the tropics. If this situation prevailed for a long time, it may explain why the Neotropical fauna is so obvi- ously richer than the Ethiopian fauna. South America might have received a sizable fau- nistic contribution from Africa without giving anything in exchange, at least for the last 50 million years. Savitzky (1978) provided evidence that the micrurines are a derivative of Neotropical rear-fanged colubrids, such as Elapomorplms, rather than relatives of the Old World cobras. This removes a zoogeographic problem. Thus, Africa contributed no venomous snakes to South America, for the Crotalinae are absent in Africa. The Viperidae appears in the fossil record in the lower Miocene in northern con- tinents. Its general range suggests an open radiation without insular or peninsular traps. A Laurasian origin is likely, as clearly deduct- ible from the study of Azemiops by Liem et al. (1971). CONCLUSIONS Present geological knowledge indicates that Africa and South America were united and formed a single continent from at least Carboniferous times, during most of the Mesozoic until the Turonian in the Cretace- ous. The final split of the continents and the birth of the Atlantic Ocean occurred 90-95 m.y.b.p. Fossil evidence shows the existence of a common fauna before the Turonian — mes- osaurians of the Permian, the Cynognathns fauna of the Triassic, and mesosuchian croco- diles of the Jurassic and Cretaceous. Both pre-Atlantic and post-Atlantic distributions are hypothesized for groups of amphibians and reptiles (Figs. 3:1). Leiopelmatid frogs existed in the Jurassic in South America, and pipid frogs and pelo- medusid turtles existed in Africa in the Early Cretaceous. The continents were united then, so it is reasonable to assume that leiopelmatids also lived in Africa, and that pipids and pelo- medusids were present in South America be- 1979 LAURENT: AFRICA AND SOUTH AMERICA 65 fore the birth of the Atlantic Ocean. In Late Cretaceous beds of South America, there are doubtful leptodactylids and iguanids. In the Paleocene beds of Brasil there are caecilians, leptodactylids, hylids and bufonids. In Afri- ca there are only an unidentified salamander in the Senonian and a primitive colubroid in the Paleocene. In some groups (bufonids and iguanids) the characteristics of the fossils and the pres- ent range of the group and its inferred phylog- eny are suggestive of a pre-Atlantic western Gondwanan range. In other cases, similar conclusions can be assumed on distributional data alone without the benefit of pertinent paleontological data. Amphi-Atlantic ranges of the gekkonids, amphisbaenians, leptoptyph- lopids and typhlopids are examples. The myo- batrachid frogs have a doubtful fossil in Cre- taceous beds of India. Their presence in western Gondwanaland in Early Cretaceous times is assumed. My hypothesis is that, ac- cording to vicariance principles, a myobatra- chid stock could have become leptodactylids in western Gondwanaland (South America), ranoids in mid-western Gondwanaland (Af- rica), microhyloids in mid-eastern Gondwa- naland (Madagascar, India), and pelodrya- dids in eastern Gondwanaland (Australia). The microhylids are supposed to have radi- ated early enough to spread to Africa and South America just before or after the separa- tion of the continents. The other families generally are quite dif- ferent in Africa and South America, suggest- ing a post-Atlantic radiation. The Neotropical leptodactylids, alligatorids, iguanids, podo- cnemine turtles (which must have crossed Af- rica to reach Madagascar), teiids, anguids, Fie. 3:1. A. Pre-Atlantic Gondwanan distribu- tions— groups present in South America and Africa before the separation. B. Last pre-Atlantic faunistic exchanges — groups that migrated from one continent to the other just before the separation or perhaps soon afterwards. C. Post-Atlantic distributions and later one-way dispersals. A. Distribuciones gondwanense preatldnticas — grupos prescntes en Sudamerica y Africa largo tiempo antes de la division. B. Ultimo intercambio faunistico preatldntico — grupos que migraron de un continente al otro, justo antes dc la division o, tal vez, pronto despues de clla. C. Distribuciones postatldnticas y despues dispersion "waif" unidireccional. MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 boids, xenodontine colubrids, micrurines and crotalines are paralleled, respectively, by Ethi- opian ranoids, crocodylids, agamids, chamae- leontids, pelomedusines, lacertids, cordylids, scincids, pythonines, lycodontine colubrids, elapids and viperines. Some measure of competitive exclusion cannot be ruled out completely, although defi- nite misgivings have been expressed about the universality of Gause's principle (Fryer and lies, 1972). The expansion of three important Neotropical families outside of South America suggests the presence of rival groups is indeed an obstacle. The bufonids have no obvious competitors and are nearly cosmopolitan. The hylids invaded the entire Holarctic Region, where no other tree frogs exist, but failed to spread into the Old World tropics, where other groups of tree frogs occur. The lepto- dactylids, which met ranids when they were barely out of their Neotropical stronghold, barely encroached upon southern North America. Such an effect must be considerably magnified in colonization by rafting, because the indigenous populations have an over- whelming advantage in numbers. Colonizers can succeed only if they enter an empty eco- logical niche or if they are superior to the indigenous species. ACKNOWLEDGMENTS I am grateful to William E. Duellman for his critical comments and to the Consejo Na- cional de Investigaciones Cientificas y Tec- nicas of Argentina and the Fundacion Miguel Lillo, which authorized this work. I am also indebted to various colleagues for valuable help and information — F. de Rroin, E. Ruf- fetaut, R. Estes, C. Gans, J. C. Rage, A. Sa- vitzky, and G. Underwood. RESUMEN El contraste evidente entre las faunas herpetologicas sudamericana y africana par- ece a primera vista apoyar las teorias zoogeo- graficas de Matthew (1915) y Darlington (1957). Los grupos dominantes son complet- amente distintos aunque en general ecologica- mente similares y a menudo relacionados: Dermophiinae al oeste del Oceano y Herpe- linae al este, y asi en seguida, Leptodactylidae y Ranoidea terrestres, Hylidae y Ranoidea arboricolas, Iguanidae y Agamidae (con Chamaeleontidae), Teiidae y Lacertidae, Podocneminae y Pelomedusinae, Alligatoridae y Crocodylidae, Boinae y Pythoninae, Xeno- dontinae y Lycodontinae. Sin embargo, la tectonica de las placas y otros progresos recientes de la geologia com- probaron sin dejar lugar a duda que Africa y Sud America estaban unidas en un solo con- tinente hasta el Turoniano, es decir hasta hace mas o menos 90 millones de anos. Rastros de la comunidad faunistica de esta epoca remota persistieron en antiguos grupos que no dominan la escena, como los Gimno- fionos, Pipidae, Pelomedusidae, Gekkonidae, Amphisbaenidae, Leptotyphlopidae y Typhlo- pidae. Pero, aim en estos ejemplos, la diver- gencia debida a su evolucion por separado durante cerca de 100 millones de anos es generalmente obvia. Hay tambien familias que aparentemente despues de haber nacido en una region occi- dental o oriental del Continente de Gondwana invadieron el resto poco antes de su frag- mentacion o tal vez poco despues, ya que travesias de mares estrechos como son oceanos recien nacidos no presentan difficultades ma- yores. Asi, aparentemente los Bufonidae, Iguanidae, quizas los Teiidae nacidos en Sud- america invadieron Africa, los primeros para seguir en conquista del mundo, los lagartos para evolucionar en otros grupos y/o estar desplazados por ellos ultimamente ( Agamidae y Chamaeleontidae, Lacertidae, Cordylidae y Scincidae). Los Microhylidae, que el autor considera como un antiguo grupo de Neobatracios de origen Indico-Malgache hicieron, al parecer, el viaje inverso, ya que la selva amazonica alberga generos bastante primitivos, como Otophryne. Aun mas tarde hay pruebas de que la travesia del Atlantico no fue imposible, ya que la invasion de America por Gekkonidae de los generos Hemidactylus (Kluge, 1969) y Tarentola se hizo a fines del Cenozoico. Por consiguiente se puede suponer que tales mi- graciones ocurrieron durante todo el Ceno- 1979 LAURENT: AFRICA AND SOUTH AMERICA 67 zoico, con frecuencia decreciente, por supues- to, a medida que los continentes se alejaban. La direction de los corrientes favorece clara- mente las travesias de Este a Oeste, de ma- nera que America del Sur ceso temprano de enriquecer la fauna africana, mientras que al contrario Africa mando probablemente emi- sarios bastante numerosos a Sudamerica, no solamente salamanquesas, sino tambien escin- cidos del genero Mabuija y, tal vez varios gru- pos de culebras arboricolas. Es interesante notar que los anfibios no tuvieron exito en este tipo de aventura, porque, en general no aguantan bien el agua salada y, tal vez la competencia de los Anfibios sudamericanos no permitia lo que se llamo el acceso ecologico en su continente. LITERATURE CITED Ahl, E. 1926. Anura: Aglossa, Xenopodidae, pp. 141-142 in Kaiser, B. (ed. ). Die Diamanten- wiiste Siidwest Afrikas 2. D. Reimer, Berlin. Amiet, J. L. 1976. Observations anatomiques et biologiques sur le genre Werneria Poche, 1903 ( Amphibiens, Anoures, Fam. Bufonidae) . Rev. Zool. Africaine 90:33-45. Andersson, L. C. 1903. Neue Batrachier aus Kam- arun von den Herren Dr. Y. Sjostedt und Dr. S. Junger gesammelt. Verh. Zool. Bot. Ges. Wien 53:141-145. Beurlen, K. 1961. Die palaogeographische Entwick- Iung des siidatlantischen Ozeans. Nova Acta Leo- poldina 24:1-36. Blair, W. F. 1972a. Evidence from hybridization, pp. 196-232 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Blair, W. F. 1972b. Summary, pp. 329-343 in Blair, W. F. (ed.). Evolution in the genus Bufo. Ibid. Blommers-Schlosser, R. M. A. 1975. Observations on the larval development of some Malagasy frogs, with notes on their ecology and biology (Anura: Dyscophinae, Scaphiophryninae and Co- phylinae). Beaufortia 24:7-25. Bogart, J. P., Nelson-, C. E. 1976. Evolutionary implications from karyotypic analysis of frogs of the families Microhylidae and Rhinophrynidae. Herpetologica 32:199-208. Bogart, J. P., Pyburn, W. F., Nelson, C. E. 1976. The karyotype of Otophrync robusta (Anura: Microhylidae). Ibid. 32:208-210. Bocert, C. M. 1940. Herpetological results of the Vernay Angola Expedition, with notes on African reptiles in other collections. Part I. Snakes, in- cluding an arrangement of African Colubridae. Bull. Amer. Mus. Nat. Hist. 77:1-107. Bonaparte, J. F. 1967. New vertebrate evidence for a southern transatlantic connection during the lower and the middle Triassic. Palaeontol. 10: 554-63. Boss, J., Pasteur, G. 1977. Solution histologique a un probleme de taxinomie herpetologique interes- sant les rapports paleobiologiques de l'Amerique de Sud et de l'Afrique. C. R. Acad. Sci. Paris, Ser. D 284:2547-50. Boulenger, G. A. 1882. Catalogue of the Batrachia Salientia S. Ecaudata in the collection of the British Museum, London, 503 p. Bourgeois, M. 1968. Contribution a la morphologie comparee du crane des ophidiens de l'Afrique Centrale. Publ. Univ. Off. Congo Lubumbashi 18:1-293. Bhoin, F. de, Buffetaut, E., Koeniguer, J. C, Race, J. C, Russell, D., Taquet, P., Vercnaud-Graz- zini, C, Wenz, S. 1975. La faune de vertebres continentaux du gisement d'ln Beceten (Senonien du Niger). C. R. Acad. Sci. Paris, Ser. D 279: 469-472. Brundin, L. 1966. Transantarctic relationships and their significance. K. Sven. Vetenskapsakad. Handl. (4)ll(l):l-472. Buffetaut, E. 1976. Sur la repartition geographi- que hors d'Afrique des Dyrosauridae, Crocodiliens Mesosuchiens du Cretace terminal et du Paleo- gene. C. R. Acad. Sci. Paris, Ser. D 283:487-90. Carroll, R. L., Currie, P. J. 1975. Microsaurs as possible apodan ancestors. J. Zool. 57:229-247. Carvalho, A. L. de. 1954. A preliminary synopsis of the genera of American microhylid frogs. Occas. Pap. Mus. Zool. Univ. Mich. (555): 1-19. Casamiquela, R. 1961. Un pipoideo fosil de Pata- gonia. Rev. Mus. La Plata, Secc. Paleontol. 4:71- 123. Cei, J. M. 1972. Bufo of South America, pp. 82-90 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Cei, J. M. 1977. Serological relationships of the Patagonian toad Bufo variegatus (Giinther). Bull. Serol. Mus. 52:2. Colbert, E. H. 1973. Continental drift and the distribution of fossil reptiles, pp. 395—412 in Tarling, D. H., Runcorn, S. K. (eds.). Implica- tions of continental drift in the earth sciences, Vol. 1. Academic Press, New York, 622 p. Creer, K. M. 1973. On the arrangement of the landmasses and the configuration of the geomag- netic field during the phanerozoic, pp. 47-76 in Tarling, D. H., Runcorn, S. K. (eds.). Implica- tions of continental drift in the earth sciences. Ibid. Croizat, L. 1964. Space, time, form: The biological synthesis. Published by author, Caracas, 881 p. Croizat, L., Nelson, G., Rosen, D. E. 1974. Cen- ters of origin and related concepts. Syst. Zool. 23:265-287. Darlincton, P. J., Jr. 1957. Zoogeography: the geographical distribution of animals. John Wiley and Sons, New York, 675 p. Dietz, R. S., Holden, J. C. 1970. The break-up of Pangaea. Sci. Amer. 223:30-41. 68 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Douglas, R. G., Moullade, M., Nairn, A. E. M. 1973. Causes and consequences of drift in South Atlantic, pp. 517-537 in Tarlinc, D. H., Run- corn, S. K. (eds. ). Implications of continental drift in the earth sciences. Vol. 1. Academic Press, New York, 622 p. Dowling, H. G. 1975. A provisional classification of snakes. 1974 Yearbook Herpetology Herpetol. Infor. Search Syst., New York, pp. 167-170. Duellman, W. E. 1978. The biology of an equa- torial herpetofauna in Amazonian Ecuador. Univ. Kansas Mus. Nat. Hist., Misc. Publ. (65): 1-352. Dunn, E. R. 1923. The geographical distribution of the Amphibians. Amer. Nat. 57:129-136. Dunn, E. R. 1928. A tentative key and arrangement of the American genera of Colubridae. Bull. An- tivenin Inst. Amer. 2:18-24. Dunn, E. R. 1931. The herpetological faunas of the Americas. Copeia 1931(3) : 106-119. Du Toit, A. 1937. Our wandering continents. Oli- ver and Boyd, Edinburgh, 366 p. Estes, R. 1970. Origin of the North American lower vertebrate fauna: An inquiry into the fossil rec- ord. Forma Functio 3:139-163. Estes, R. 1975a. African frog Xenopus (Pipidae) from the Palaeocene of Brasil and its zoogeo- graphic importance. Nature 254:48-50. Estes, R. 1975b. Fossil Xenopus from the Palaeo- cene of South America and the zoogeography of pipid frogs. Herpetologica 31:263-278. Estes, R., Price, L. I. 1973. Iguanid lizard from the upper Cretaceous of Brasil. Science 180: 748-751. Estes, R., Reig, O. 1973. The early fossil record of frogs: A review of the evidence, pp. 11-63 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Estes, R., Wake, M. 1972. The first fossil record of caecilian amphibians. Nature 239:228-231. Francheteau, J. 1973. Plate tectonic model of the opening of the Atlantic ocean south of the Azores, pp. 197-202 in Tarlinc, D. H., Runcorn, S. K. (eds.). Implications of continental drift in the earth sciences. Vol. 1. Academic Press, New York, 622 p. Fryer, O, Iles, T. D. 1972. The cichlid fishes of the great lakes of Africa. T. F. H. Publ., Neptune City, N.J., 641 p. GafFx\ey, E. S. 1975. A phytogeny and classifica- tion of the higher categories of turtles. Bull. Amer. Mus. Nat. Hist. 155:391-436. Gorman, G. C. 1970. Chromosomes and the sys- tematics of the family Teiidae (Sauria, Reptilia). Copeia 1970(2):230-245. Griffiths, I. 1963. The phylogeny of the Salientia. Biol. Rev. 38:241-292. Griffiths, I., Carvalho, A. L. de. 1965. On the validity of employing larval characters as major phyletic indices in Amphibia Salientia. Rev. Bra- sileira Biol. 25:113-121. Hecht, M. 1963. A reevaluation of the early history of the frogs. Part II. Syst. Zool. 12:20-35. Hecht, M., Malone, B. 1972. On the early history of the gavialid crocodilians. Herpetologica 28: 281-284. Heirtzler, J. R. 1968. Sea-floor spreading. Sci. Amer. 219:60-70. Hoffstetter, R. 1955. Squamates de type moderne, pp. 606-662 in Piveteau, J. (ed.). Traite de Paleontologie, Paris, 1113 p. Hoffstetter, R. 1962. Revue des recentes acquisi- tions concernant l'histoire et la systematique des squamates. Colloq. Int. C. N. R. S. Paris 104, 474 p. Hoffstetter, R. 1964. Les Sauria du Jurassique superieur et speeialment les Gekkota de Baviere et de Mandchourie. Senckenbergiana Biol. 45: 281-324. Hoffstetter, R. 1967. Coup d'oeil sur les Sauriens ( = Lacertiliens) des Couches de Purbeck (Juras- sique superieur dAngleterre). Colloq. Int. C. N. R. S. Paris 163:349-71. Hoffstetter, R. 1968. [Review of] A contribution to the classification of snakes by G. Underwood. Copeia 1968(1 ):201-213. Hoffstetter, R. 1972. Relationships, origin and history of the ceboid monkeys and caviomorph rodents: A modem reinterpretation. Evol. Biol. 6:323-348. Hoffstetter, R. 1977. Primates: Filogenia e his- toria biogeografica. Stud. Geol. Univ. Salamanca 13:211-253. Hoffstetter, R., Race, J.-C. 1972. Les Erycinae fossiles de France ( Serpentes, Boidae ) . Compre- hension et histoire de la sous-famille. Ann. Paleon- tol. Vertebr. 58:81-124. Huene, F. von. 1956. Palaontologie und Phylogenie der niederen Tetrapoden. Fischer, Jena, 761 p. Hurley, P. M. 1968. The confirmation of continen- tal drift. Sci. Amer. 218:52-64. Inger, R. F. 1967. The development of a phylogeny of frogs. Evolution 21:369-383. Jeannel, R. 1942. La genese des faunes terrestres. Presses Univ. France, Paris, 514 p. Kluge, A. G. 1967. Higher taxonomic categories of gekkonid lizards and their evolution. Bull. Amer. Mus. Nat. Hist. 135:1-59. Kluge, A. G. 1969. The evolution and geographical origin of the New World Hemidactylus nxabouia- brooki complex (Gekkonidae, Sauria). Misc. Publ. Mus. Zool. Univ. Michigan (138): 1-78. Kluge, A. G., Farris, S. J. 1969. Quantitative phy- letics and the evolution of the anurans. Syst. Zool. 18:1-32. Khommelbein, K. 1971. Non-marine Cretaceous os- tracodes and their importance for the hypothesis of Gondwanaland. Proc. 2nd. IUGS Gondwana Symp. South Africa, pp. 617-619. Laurent, R. F. 1972. La distribution des amphib- ious et les translations continentales. 17th. Inter. Cong. Zool., Monaco, 1:1-16. Laurent, R. F. 1973. A parallel survey of equatorial amphibians and reptiles in Africa and South America, pp. 259-266 in Mecgers, B. J., Ayensu, E. S., Duckworth, W. D. (eds.). Tropical forest ecosystems in Africa and South America. Smith- sonian Inst. Press, Washington, D.C., 350 p. 1979 LAURENT: AFRICA AND SOUTH AMERICA 69 Laurent, R. F. 1975. La distribution des amphib- iens et les translations continentales. Mem. Mus. Natl. Hist. Nat. Paris Ser. A Zool. 88:176-191. Laurent, R. F. In press. Systematique et distribu- tion geographique (des Amphibiens ) . in Grasse, P. P. (ed.). Traite de Zoologie 14, fasc. 2 Am- phibiens. Laughton, A. S., Sclater, J. G., McKenzie, D. P. 1973. The structure and evolution of the Indian ocean, pp. 203-212 in Tarling, D. H., Runcorn, S. K. (eds.). Implications of continental drift in the earth sciences. Vol. 1. Academic Press, New York, 622 p. Lavocat, R. 1974. What is an hystricomorph ro- dent? Symp. Zool. Soc. London 34:7-20. Liem, K. F., Marx, H., Rabb, G. B. 1971. The viperid snake Azemiops: Its comparative cephalic anatomy and phylogenetic position in relation to Viperinae and Crotalinae. Fieldiana Zool. 59: 65-126. Liem, S. S. 1970. The morphology, systematics and evolution of the Old World tree frogs (Rhaco- phoridae and Hyperoliidae ) . Fieldiana, Zool. 57: 1-145. Lynch, J. D. 1973. The transition from archaic to advanced frogs, pp. 133-182 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contempo- rary research on major problems. Univ. Missouri Press, Columbia, 470 p. Martin, H. 1968. A critical review of the evidence for a former direct connection of South America with Africa, pp. 25-53 in Fittkau, E. J., Illtes, J., Klinge, H., Schwabe, G. H., Sioli, H. (eds.). Biogeography and ecology in South America. W. Junk, The Hague, 946 p. Matthew, W. D. 1915. Climate and evolution. Ann. New York Acad. Sci. 24:171-318. McDiarmid, R. W. 1971. Comparative morphology and evolution of frogs of the Neotropical genera Atelopus, Dendrophryniscus, Melanophnjniscus and Oreophrynella. Nat. Hist. Mus. Los Angeles Cty. Sci. Bull. (12;: 1-66. McKenna, M. C. 1973. Sweepstakes, filters, corri- dors, Noah's arks and beached Viking funeral ships in palaeogeography, pp. 295-308 in Tarl- inc, D. H., Runcorn, S. K. (eds.). Implications of continental drift in the earth sciences. Vol. 1. Academic Press, New York, 622 p. McKenzie, D. P., Sclater, J. G. 1973. The evolu- tion of the Indian Ocean. Sci. Amer. 228:62-72. Medem, F. 1968. Desarrollo de la herpetologia en Colombia. Rev. Acad. Colombiana Cien. Exactas Fis. Nat. 13:149-200. Mook, C. C. 1934. The evolution and classification of the Crocodilia. J. Geol. 42:295-304. Morescalchi, A. 1973. Amphibia, pp. 233-348 in Chiarelli, A. B., Cat-anna, E. (eds.). Cyto- taxonomy and vertebrate evolution. Academic Press, New York, 783 p. Nelson, C. E. 1966. The evolution of frogs of the family xMicrohylidae. PhD Dissert. Univ. Texas Diss. Abst. 27, 3719 B. Nelson, C. E., Cuellar, H. S. 1968. Anatomical comparison of tadpoles of the genera Hypopachus and Gastrophryne ( Microhylidae). Copeia 1968 (2):423-424. Nelson, G. 1973. Comments on Leon Croizat's bio- geography. Syst. Zool. 22:312-320. Nevo, E. 1968. Pipid frogs from the Early Cretace- ous of Israel and pipid evolution. Bull. Mus. Comp. Zool. Harvard Univ. 136:255-318. Noble, G. K. 1931. The biology of the Amphibia. McGraw-Hill Book Co., New York, 577 p. Nopsca, F. 1928. The genera of Reptiles. Paleobiol. 1:163-188. Nussbaum, R. A. 1977. Rhinatremidae: A new fam- ily of Caecilians (Amphibia: Gymnophiona ) . Occas. Pap. Mus. Zool. Univ. Mich. (682): 1-30. Orton, G. L. 1957. The bearing of larval evolution on some problems in frog classification. Syst. Zool. 6:79-86. Patterson, B., Pascual, R. 1968. The fossil mam- mal fauna of South America. Q. Rev. Biol. 43: 409-451. Rabb, G. B., Marx, H. 1973. Major ecological and geographic patterns in the evolution of colubroid snakes. Evolution 27:69-83. Rage, J.-C. 1974. Les Serpents des phosphorites du Quercy. Palaeovert. 6:274-303. Rage, J.-C. 1975. Un serpent du Paleocene du Niger. Etude preliminaire sur l'origine des cae- nophidiens (Reptilia, Serpentes). C. R. Acad. Sci. Paris, Ser. D 281:515-18. Rage, J.-C. 1976. Paleontologie, phylogenie et paleo- biogeographie des serpents. PhD Dissert. Univ. Paris. Rage, J.-C. 1977. L'origine des Colubroi'des et des Acrochordoides (Reptilia, Serpentes). C. R. Acad. Sci. Paris, Ser D 286:595-597. Reyment, R. A. 1958. Factors in the distribution of fossil cephalopods. Stockholm Contrib. Geol. 1: 91-184. Reyment, R. A. 1970. Vertically inbedded cepha- lopods shells. Some factors in the distribution of fossil cephalopods. Palaeogeogr. Palaeoclimatol. Palaeoecol. 7:103-111. Reyment, R. A. 1975. Paleontologie evolutive et nouvelle tectonique. Mem. Mus. Natl. Hist. Nat. Paris Ser. A Zool. 88:8-18. Reyment, R. A., Tait, E. A. 1972. Biostratigraph- ical dating of the early history of the South Atlantic Ocean. Phil. Trans. Roy. Soc. London, B. Biol. Sci. 264:55-95. Robinson, P. 1967. The evolution of the Lacertilia. Colloq. Int. C. N. R. S., Paris 163:395-407. Romer, A. S. 1956. Osteology of the reptiles. Univ. Chicago Press, Chicago, 772 p. Romer, A. S. 1966. Vertebrate Paleontology. 3rd ed. Univ. Chicago Press, Chicago, 468 p. Savage, J. M. 1973. The geographic distribution of frogs: Patterns and predictions, pp. 351-445 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, 470 p. 70 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Savitzky, A. H. 1978. The origin of the New World proteroglyphous snakes and its bearing on the study of venom delivery in snakes. PhD Dissert. Univ. Kansas, 387 p. Schmidt, K. P. 1946. On the zoogeography of the Holarctic Region. Copeia 1946 (3): 144-152. Sill W. D. 1968. The zoogeography of the Croco- dilia. Copeia 1968(1 ):76-88. Simpson, G. G. 1942. A Miocene tortoise from Pata- gonia. Amer. Mus. Novit. (1209): 1-6. Simpson, G. G. 1965. The geography of evolution. Chilton, Philadelphia, 249 p. Smith, H. M., Smith, R. B., Sawin, H. L. 1977. A summary of snake classification ( Reptilia, Ser- pentes). J. Herpetol. 11:115-121. Sokal, R. R., Crovello, T. J. 1970. The biological species concept: A critical evaluation. Amer. Nat. 104:127-153. Sokol, O. M. 1975. The phylogeny of anuran lar- vae: A new look. Copeia 1975 ( 1 ) :l-24. Starbett, P. H. 1973. Evolutionary patterns in lar- val morphology, pp. 251-271 in Vial, J. L. (ed. ). Evolutionary biology of the anurans: Contem- porary research on major problems. Univ. Mis- souri Press, Columbia, 470 p. Steel, R. 1973. Crocodylia. Handbuch der Paliio- herpetologie. 16:1-116. Tandy, M., Keith, R. 1972. Bufo of Africa, pp. 119-170 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Tihen, J. A. 1960. Two new genera of African bufo- nids with remarks on the phylogeny of related genera. Copeia 1960 (3):225-233. Tihen, J. A. 1965. Evolutionary trends in frogs. Amer. Zool. 5:309-318. Trueb, L. 1971. Phylogenetic relationships of cer- tain Neotropical toads with the description of a new genus ( Anura:Bufonidae). Nat. Hist. Mus. Los Angeles Cty. Contr. Sci. (216):l-40. Underwood, G. 1967. A contribution to the classi- fication of snakes. Brit. Mus. (Nat. Hist.), Lon- don, 179 p. Underwood, G. 1976. A systematic analysis of boid snakes, pp. 151-175 in Bellairs, A., Cox, C. B. (eds.). Morphology and Biology of Reptiles. Linnean Soc. London Symp. Series 3, 290 p. Vanzolini, P. E. 1968. Geography of the South American Gekkonidae (Sauria). Arq. Zool. (Sao Paulo) 17:85-111. Vergnaud-Grazzini, C. 1966. Les amphibiens du Miocene de Beni-Mellal. Notes Serv. Geol. Maroc 27:43-69. Vergnaud-Grazzini, C. Hoffstetter, R. 1972. Presence de Palaeobatrachidae (Anura) dans des gisements tertiaires francais. Caracterisation, dis- tribution et affinites de la famille. Palaeovert. 5:157-177. Wake, D. 1966. Comparative osteology and evolu- tion of the lungless salamanders, family Pletho- dontidae. Mem. South. Calif. Acad. Sci. (4):1- 111. Wood, A. E. 1950. Porcupines, paleogeography and parallelism. Evolution 4:87-98. Wood, R., Patterson, B. 1973. A fossil trionychid turtle from South America. Breviora (415): 1-10. APPENDICES Appendlx 3:1. — Vicariance, dispersal and centers of origin. According to the Hennigian and cladistic theories, the concepts of dispersal and centers of origin, cher- ished by so many zoologists from Darwin to Darling- ton, Simpson and Mayr, are useless (Croizat et al., 1974), whereas the concept of vicariance is the master key to biogeographic problems. Perhaps it is true that the importance of vicariance has been over- looked in the zoogeographic conjectures of the past decades, but surely not to the extent implied by these authors. The model of allopatric speciation, one of the main tenets of the synthetic evolutionary theory, is the very basis of the vicariance concept. But this does not necessitate the rejection of other causes of biotic distributions. Dispersal occurs, for there is more than one species at almost every locality (more than 80 species of frogs at Santa Cecilia, Ecuador, as noted by Duellman, 1978). Croizat and his co- workers admit that distributions are not static but fail to allow that dispersal is as important as vicariance. Furthermore, they belittle the concept of the geno- center under the pretext that a species may have a huge array of disjunct populations over an entire con- tinent. Such enormous genetic pools are not espe- cially productive evolutionarily. They impose a great deal of inertia to the spreading of genetic changes. On the contrary, innovative processes, such as genetic drift, genetic revolution, quantum and tachytelic evo- lution, occur in small, transitory and local popula- tions. Nevertheless, it is futile to seek genocenters when adequate data are lacking. It might be said for genocenters, as Sokal and Crovello ( 1970 ) did for the biological species concept, that the concept is not operational. However, this does not preclude the existence of centers of origin, even if we are unable to discover their location, exactly as the non-opera- tionality of the biological species concept does not eliminate the fact that the cessation of gene exchange between two populations is such a momentous event in evolution that it is inconceivable to ignore it, even if we are unable to pinpoint its occurrence. Appendix 3:2. — The Indian raft. It is now generally believed that India drifted away from Antarctica, Madagascar and Africa some- time at the end of the Mesozoic and travelled north- wards through the India Ocean to collide with Laura- sia in the Miocene. Tire dating of the separation is still doubtful — about 100 m.y.b.p. from Antarctica and maybe the Paleocene (60 m.y.b.p) from Mada- gascar. Little attention has been given to the impact of the Indian fauna on the evolutionary zoogeography in the Tertiary. This is an unfortunate omission, not justified by lack of evidence. The Eocene lndo- batrachus seems to belong to the Myobatrachidae, and the primitive snake family Uropeltidae survives in southern India and Sri Lanka. The rationale for my hypothesis is as follows: 1979 LAURENT: AFRICA AND SOUTH AMERICA 71 1. India broke from Madagascar and Africa dur- ing the late Cretaceous or early Tertiary. 2. As a large island, it rafted away from Mada- gascar northwards along the Mascarene Ridge, leaving behind the Seychelles, Amirante and finally the Laccadive and Maldive islands ( Laughton et al., 1973; McKenzie and Sclater, 1973). 3. The Indian fauna evolved in isolation for about 50 million years; evolution was enhanced by the changing climates. 4. Some faunistic exchanges remained with Mada- gascar and Africa through the intervening islands, like the Seychelles, and possibly others that have since disappeared. 5. These small islands provided opportunities for genetic drift and quantum and tachytelic evo- lution favoring major adaptive shifts (e.g., Microhylidae, Savage, 1973). 6. Perhaps other exchanges took place in front of and/or on the eastern side when the Indian Noah's Ark (McKenna, 1973) drew near Lau- rasia, gliding along the Ninety-east Ridge and finally along the Nicobar and the Andaman islands. 7. The fate of much of the Indian fauna must have been extinction. 8. Some elements escaped early to Madagascar and Africa and proved successful in their ex- pansion (e.g., Microhylidae). 9. Other elements escaped later in northern, northeastern and eastern directions (e.g., other Microhylidae and perhaps Agamidae, Varani- dae, Pythonini, Elapidae). Four groups of reptiles (Agamidae, Varanidae, Pythonini, Elapidae) have patterns of distribution that can be explained by an Indian differentiation. Each has a strong Indo-Malaysian component, another strong Australasian component, and a weak African component, as if there had been a late invasion of Africa from Asia. Subsequent to writing this account, I have been informed by R. Hoffstetter that recent data show that the Indian Subcontinent collided with Laurasia not later than the Eocene. 4. Herpetofaunal Relationships of South America With Australia Michael J. Tyler Department of Zoology University of Adelaide Adelaide, South Australia 5001 Australia In reviewing the extent of South American herpetofaunal relationships with Australia, while simultaneously considering South Amer- ican-African relationships (Laurent, this vol- ume), it is helpful to recognize that vast dif- ferences have existed in the opportunity for faunal exchange. South America and Africa may be regarded as lovers who experienced and exploited a large zone of contact and had considerable opportunity for interchange and exchange across it. In contrast, the South American-Australian relationship suffered from being in the form of an arranged engage- ment of longer duration. The couple never so much as touched one another at any time. The only contact was via a related intermedi- ary named Aunt Arctica, whose presence be- tween them effectively prevented a compar- able degree of intimacy, and who is now outwardly cool and distinctly secretive about revealing what took place between them. The benefit of employing such an analogy lies in emphasizing the fact that Australia and South America have always been physically separated. This separation always has been extensive, because the intervening Antarctica is a vast continent with a surface area of 1,165,500,000 km2, comparable in size to South America north of the Tropic of Capri- corn, and considerably greater than Australia (7,700,000 km2). A North to South traverse of Antarctica involves a distance of approxi- mately 4,000 km. When Antarctica was an integral com- ponent of Gondwanaland, the herpetofaunal elements shared at any one time by South America and Australia also would have oc- curred on Antarctica. Certainly a topography, climate, and vegetation equable to the main- tenance of reptiles and amphibians had to exist on Antarctica, and at least some of the modern families could just as well have orig- inated there as on the adjacent landmasses. Thus any realistic concept of intercontinental exchange avoids reference to "journeys" along "routes," and only visualizes the expansion and retraction of populations. Cartoons in an otherwise serious paper by Rich ( 1975 ) on the origins of the Australian nonpasserine avi- fauna, illustrate the errors to which some in- vestigational philosophies may have suc- cumbed. The study of intercontinental herpeto- faunal relationships faces problems of varia- tion of systematic interpretation of taxa, and these materially influence the degree of faunal similarity. For example, if the numerically dominant Australian terrestrial and arboreal frogs are regarded as members of the Lepto- dactylidae and Hylidae, respectively, all anu- ran families found in Australia are shared with South America. Superficially at least, the anuran relationship appears likely to prove a close one. However, if the names Myobatrachidae and Pelodryadidae are em- ployed for these same groups, it is difficult to avoid a bias towards a quite different inter- pretation. In fact it would appear that, for the purposes of intercontinental comparisons, there is a mystique surrounding a family name that does not extend to other nomen- clature. Over the past few years there have been substantial contributions to the study of plate tectonics, continental drift, palaeoclimate and the past flora and fauna of Australia. Many of these papers are highly relevant to the inter- pretation of evolutionary opportunities and the nature of the diversification of the herpe- tofauna. Here I have attempted to bring to- gether the most recent literature as a general background before examining the evidence to 73 74 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 lAl»0t Fig. 4:1. Australia, New Guinea and adjacent landmasses. Australia, Nueva Guinea tj iierras adijacentes. establish the origins of the modern, non- marine Australian herpetofauna, and its af- finities to the herpetofauna of South America. A map of Australia and associated landmasses is shown in figure 4:1. PALAEOENVIRONMENTAL CONSIDERATIONS Onset of drifting. — Previously there has been considerable variation in estimates of 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 75 the onset of the northwards drift of Australia away from east Antarctica, ranging from a low of 43 m.y.b.p. (Jardine and McKenzie, 1972) to a high of 180-100 m.y.b.p. (Fooden, 1972; Savage, 1973). However, it is now placed at 55-52 m.y.b.p., with most authors favoring 53 m.y.b.p. ( McGowran, 1973; Sclat- er et al., 1974; Coleman and Packham, 1976; Veevers and McElhinny, 1976). McGowran's studies of the Antarctic-Australian suture led him to suggest that, despite the onset of drift, there was no substantial barrier to the pas- sage of land animals prior to the early Eocene (49 m.y.b.p.). Climatic, floral and faunal changes. — Be- cause Nothofagus forests now occur in some temperate areas, such as southeast Australia including Tasmania, and in New Zealand, it has been possible to deduce that Nothofagus is associated classically with temperate cli- matic conditions (Axelrod, 1975). Thus, with evidence of Nothofagus occurring in the Eo- cene at several localities in southern Australia, the inference might be drawn that, at the time of the separation of Australia from East Antarctica, the southern Australian fauna was probably cool-temperate. Certainly this as- sumption would be valid for N. fusca and N. menziesi, which now exist in Australia, New Zealand, Chile and Argentina. However, the important species is N. brassi, which now exists in New Guinea and New Caledonia and clearly is a subtropical species. Formerly, its distribution was far more extensive, being known in Australia and New Zealand from the Early Cretaceous to the mid-Pliocene, in West Antarctica from the early Palaeocene to the mid-Eocene, and from Chile and Argen- tina from the Early Cretaceous to the late Oligocene (Schlinger, 1974). Further evidence of the southern Austral- ian climate being subtropical has been estab- lished by Lange (1976) from his study of microfossil epiphyllous germlings, and by Christophel and Blackburn (1978) from their assessment of the Eocene South Australian Maslin Bay flora. The geomorphological evi- dence of widespread subtropical conditions are summarized by Bowler (1976). Central Australia is another portion of the continent whose palaeoclimatic conditions have been misinterpreted. Axelrod (1960) envisaged deterioration throughout the Ceno- zoic leading to arid to semiarid conditions by the Miocene. There is now evidence that cen- tral Australia bore large, permanent lakes in the Miocene. Lungfishes, teleosts, turtles, crocodiles, lizards, and frogs shared the site with a vast diversity of marsupials and birds. The surrounding vegetation was dense, rang- ing from rainforests to extensive areas of grassland. Gallery forests extended along the watercourses, and the presence of Nothofagus and Podocarpus are interpreted to be evi- dence of high rainfall. There was southern communication with the sea at some stage (Callen and Tedford, 1976). The records of crocodiles and turtles at former freshwater sites in central Australia are particularly nu- merous (see also Newsome and Rochow, 1964). However, these represent the most conspicuous and most readily recognized rep- tile fossils; the search for smaller material has only just begun. At some stage, the area between central Australia and the north coast also was moist. This is demonstrated by the cabbage palms, Livistona mariae, now restricted to a colony of 3,000 at Finke River in central Australia. Their nearest relatives lie 1,000 km away in the northwest of the continent ( Latz, 1975 ) . Pa- laeontological and geomorphological evidence demonstrate that central Australia provided numerous niches for mesic animals until the end of the Pleistocene ( Wopfner and Twidale, 1967; Mabbut, 1967; Twidale, 1972). Cer- tainly deserts have featured in Australia for a long period and have had an essential role in lizard speciation (Pianka, 1972). It has been suggested that in the Quaternary much of the now moist extreme southwest of the continent was arid (Glassford and Killigrew, 1976). However the extent of Australia affected by aridity appears to have been exaggerated. The minimal morphological differentiation of the central Australian hylid frog fauna is wholly consistent with aridity being a late Pleistocene feature. Thus the species Litoria caerulea and L. rubella are relicts of a much richer fauna, surviving because of tolerance of adults or larvae to high temperature or 76 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 possibly for some other reason. But for the contrary evidence of Glassford and Killigrew ( 1976 ) , it is possible that Australia has not been any more arid than it is today and, to judge from the nature and abundance of vegetation cover now stabilizing sand dunes in some areas, a trend towards climatic ame- lioration has already begun. However, the Pleistocene record is of minimal relevance to this review. THE NATURE OF THE AUSTRALIAN-ORIENTAL COLLISION Modern New Guinea is composed of three distinct and roughly longitudinally arranged portions. The southern portion and the inter- vening Arafura Sea originally represented the leading edge of the Australian continental plate. The central cordillera is predominantly a much younger feature; uplift commenced in the Miocene. Finally there is a row of iso- lated mountain ranges on the north coast, each of which is composed of older volcanic rocks. The history of the evolution of New Guin- ea, and of the area to the east and west is extremely complex. Similarly only the broadest of principles of the nature of the collision of the plates has yet been established. The fol- lowing contributions provide a brief spectrum of opinions and are a source of many other references : Thompson ( 1967 ) , Falvey and Taylor (1974), Coleman (1975), Denham (1975), Mackenzie (1975), Taylor (1975), Tilbury (1975) and Coleman and Packham (1976). Only recently attempts have been made to reconstruct the nature of the plate collision. Mackenzie (1975) suggested that mountain ranges now on the north coast of New Guinea represent an arc of islands that persisted through to the Miocene, and became accreted during the collision of the plate margins (Fig. 4:2). Coleman and Packham (1976:204) favored this concept: "For the moment, we accept the likelihood that north coastal New Guinea is a piece of crust, prob- ably an arc segment, in collision with Aus- tralia-New Guinea." New Britain, to the east of New Guinea, therefore represents an island of the same arc, but which did not come di- rectly into contact with New Guinea. ^---^ TORRICELLI ~\TORRICELLI ^^^ 1 -^FINISTERRE D \=INISTERRE -uu/j \~\<3c: . NEW BRITAIN '■•■. \ ^J^riEDGE OF ^ »-* ': \F\_) MIOCENE LANDMASS /f^~~\ yi MIOCENE PLEISTOCENE * ;/ \ v3^ • EDGE OF ^""-—^V^ yr :<£ PALAEOZOIC Via r\ :' j ( CRUST n ; ^Mi^^H Fig. 4:2. Reconstruction of the collision of eastern figure represents the early Miocene prior to collision, right figure shows two of the islands ( now known as accreted into New Guinea, thereby becoming part of 1975.) Reconstruction dc la colision del cste de Nueva Gui izquierda representa el Mioccno inferior previo a la de islas. La figura de la derecha muestra dos de las islas sterre) adheridas a Nueva Guinea, llegando a formar McKenzie, 1975.) New Guinea with islands in the mid-Miocene. The left with the landmass approaching a chain of islands. The the Torricelli Mountains and the the northern New Guinea coastline Finnisterre block ) (After McKenzie, nea eon islas en el Mioceno medio. La figura de la colision, con la mass de tierra alcanzando una cadena (conocidos como Montanas Torricelli y el Bloque Finni- parte de la costa del norte de Nueva Guinea. (Dc 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 77 HERPETOFAUNAL ORIGINS In very broad terms, the ancestral stocks of the South American and Australian faunas were derived from two distinct centers. For South America there was an initial Gondwa- nan source, reinforced following the drifting of Africa, followed by a later North American infusion. For Australia there was similarly an initial Gondwanan source followed by an Oriental one. A basic problem is to determine for each continent which taxa are of Gondwanan an- cestry. The recent literature includes several assessments (Keast, 1971, 1973; Cracraft, 1973, 1974, 1975; Savage, 1973). For Aus- tralia the distinction between the Gondwanan element and the more recent Oriental one should be distinguishable on the basis of morphological divergence and by the nature of geographic distribution. This is because the Australian-Oriental collision occurred in the mid-Miocene, so that animals in Australia of Oriental origin should have distinct affinities and comparable geographic distributions with animals in the Oriental Region. Conversely, such relationships should be lacking among the Gondwanan component and there should be minimal geographic distribution outside the Australian continent. Thus I propose to establish the constituents of the Gondwanan element of the Australian fauna primarily by a process of identifying, and so eliminating, the Oriental element. On an historic and biogeographic basis the Oriental element of the Australian herpe- tofauna will fit into one of two categories, as follows: 1) Animals that occurred within the northern chain prior to the mid-Miocene col- lision with Australia. All of these would have entered that area from the west. They could range from the Philippine Islands to Fiji. To be recognizable as pre-collision components, they should be more abundant on islands east of New Guinea than in New Guinea itself. 2) Animals that have dispersed from west to east following the accretion of the chain within northern New Guinea. Such animals are likely to show a progressive west to east reduction in diversity and to be poorly repre- sented on the islands east of New Guinea. It is worth contemplating that some of the deficiencies of Wallace's Line and of other attempts to delineate the Oriental and Aus- tralian faunas exist because in reality there are three components. Hence to the recog- nized Australian and post-collision Oriental colonizers, biogeographers have failed to rec- ognize the existence of the additional pre- collision Oriental unit. ORIENTAL ELEMENTS Ranidae The overall distribution of the Ranidae in the Australian Region is wholly consistent with the concept of entry from the adjacent Oriental Region to the west. What is less satisfactorily explained is the existence of two endemic species of Platymantis in Fiji far to the east, whereas none occurs in Australia. In terms of diversity and abundance of ranid species, New Guinea is equally anomalous, to the extent that this component of its fauna is depauperate when compared with those of smaller islands to the west and to the east. Thus there are 20 ranids in the Philippine Islands, 10 on New Guinea, but 24 on the Solomon Islands. Those anomalies are high- lighted by the study of the genus Platy mantis, including species previously referred to Cor- nufer (Fig. 4:3). Viewing such a distribution pattern has led to the assumption that the distribution of Platymantis in New Guinea is relictual (Zweifel, 1969). In support of a concept that Platymantis was formerly far more widely distributed in New Guinea than it is today, there is evidence of close phylogenetic relationships existing between species that are geographically iso- lated from one another. An example is P. batantae of Batanta adjacent to the Vogelkop Peninsula of Irian Jaya (West New Guinea), which Zweifel ( 1969 ) considered most closely related to P. giUiardi and P. mimicus of New Britain about 2000 km distant. Platymantis punctata of northern New Guinea and P. myersi of Bougainville, Solomon Islands, with which it has affinities, are separated by a simi- lar distance. Inger (1954:355) evidently drew comparable conclusions when he suggested that the closest relations of P. meyeri of the Philippine Islands ". . . are not with other 78 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 B O R N E O p 11 H L 1 1 S 1 P A V 1 N N s E B PALAU ISLANDS BATANTA/WAIGEU N.W. NEW GUINEA CYCLOPS MTS. r1 2 m |_ NEW GUINEA 1 ^; EW BRITAIN AUSTRALIA \ A SOLOMON V \ ISLANDS FIJI {2} Fig. 4:3. Modern distribution and numbers of species of frogs of the genus Platymantis (Ranidae). Distribution actual y numeros de especies de batracios del genero Platymantis (Ranidae). Philippine Corniifer [Platymantis] but instead seem to be with non-Philippine species." Brown and Alcala ( 1970 ) proceeded a step further and declared that the distribution of Platymantis within the Philippines is relictual, thereby accounting for the predominance of the genus in the north of that group of islands. An inteq:>retation of Platymantis as relicts is most readily made if the relevant landmasses are visualized as being static and the animal populations conveniently mobile. An alterna- tive interpretation is one in which it is pos- sible to contemplate mobile landmasses and relatively static insular populations. Clearly, there are several major centers of ranoid evolution in different parts of the world; the Philippine Islands with seven gen- era and 29 species is one of them. The adja- cent and larger land mass of Sabah (Borneo) to the southwest has fewer. Platymantis prob- ably evolved within the Philippines in the Late Tertiary and subsequently dispersed southeastwards into New Britain, the Solo- mon Islands and Fiji by rafting. Presumably this would require marine equatorial currents following patterns essentially similar to those today. The primary radiation is from the Philippines to New Britain and the Solomon 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 79 Islands (Fig. 4:4). The direction of second- ary radiations is a reflection of demonstrable pliylogenetic affinities of the extant species. New Britain evidently was a major source of colonizing species, leading to the occurrence of P. batantae on Batanta and P. punctata on Waigeo Island off the Vogelkop Peninsula, P. cheesmanac on the Cyclops Mountains, and possibly P. papuensis on the Finisterre Mountains (subsequently extending through- out northern New Guinea). When the Australian continental mass col- lided with the Oriental island chain, it there- fore acquired four species (or stocks) of Platymantis (Fig. 4:4). Three of them (ba- tantae, cheesmanae and punctata) have re- mained almost entirely within the original confines of the islands on the north coast and have not spread appreciably in New Guinea. The fourth (papuensis) has extended as far as the south coast in the extreme west of New Guinea. This species ranges to New Britain and the Solomon Islands, but it is known to inhabit the intertidal zone and is well suited to dispersal by land and by sea (Tyler, 1976a). Nevertheless, with the time scale available, its dispersal in New Guinea remains modest (Fig. 4:5). Perhaps this im- plies the existence of an Australopapuan com- petitor and hence an ecological, rather than a physical, barrier to dispersal. The source of the stock that gave rise to the two endemic species on Fiji is uncertain. The intervening and florally rich New Hebri- des lacks Platymantis or any other endemic species of frogs. The Australian hylid Litoria aurea has been introduced there recently, pos- sibly from New Caledonia, where it was in- troduced at the turn of the century (Tyler, 1976a, 1979 ) . The striking success of the New Hebrides introduction tends to eliminate any possibility of extinction as an explanation for the absence of frogs there. Rana represents a more recent ranid ar- rival. The number of species on the various landmasses north of Australia exhibits a pro- gressive reduction from west to east in accord with an Oriental origin (Fig. 4:6). Microhylidae Microhylids occur in South and North America, Africa, Madagascar, Asia including Indonesia and the Philippines, New Guinea and northern Australia. The Australopapuan unit is the most prolific, with 13 genera and 102 species ( Zweifel, 1972; Menzies and Tyler, 1977), compared with 16 genera but only 32 species in South America (Walker, 1973; Walker and Duellman, 1974; Nelson, 1975). Parker (1934) was the last contributor to treat this family in its entirety. Subsequent contributions have tended to examine single geographic components, and the overall phy- logenetic relationships of the diverse genera remain obscure. Cracraft ( 1973 ) and Bogart and Nelson ( 1976 ) outlined the principal issues, of which contention has centered on interpretation of the origin of the family, and particularly its relationship to the Ranidae. In reality, the wide variety of opinions that have been offered on the origin of this family reflects the extreme morphological complexity of the constituent members and the absence of a satisfactory, modern synthesis. This is demonstrated particularly well by the varying subfamilial classifications that have been pro- posed. Within the context of South American- Australian faunal studies, the contributions of Savage (1973) must be considered in detail. Savage's conclusions differed quite strikingly from those of Parker (1934). Whereas the latter recognized two subfamilies occurring within and confined to the Australopapuan area (Asterophryinae and Sphenophryninae), Savage recognized only one, to which the name Asterophryinae was applied. Savage (1973:355) considered that the only distinc- tion between the Asterophryinae and the Sphenophryninae was that ". . . the former usually have an amphicoelous vertebra just anterior to the sacrum and the other presacral vertebrae procoelous (diplasiocoelous), while the latter have all presacral vertebrae procoe- lous." Savage reinforced his argument of the inherently trivial nature of such a distinction, by pointing out that although Genyophryne has uniformly procoelous vertebrae, it had been referred to the amphicoelous Astero- phryinae by Parker (1934). Savage wrote without the benefit of access to a contemporary study by Zweifel (1971), who reexamined the diagnostic characteristics of both subfamilies in general and of Genyo- 80 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 p H 1 1 S 1 L 1 A P N P D 1 S N E SECONDARY ^ □ RADIATION SECONDARY RADIATIONS NEW GUINEA AUSTRALIA Fig. 4:4. Dispersal routes and the distribution of Plat center of the figure represents the original Oriental is tanta + Waigeo Island; hatched = Torricelli Moun collided with these islands, the latter occupied the in leading to the situation shown in figure 3. Rutas de dispersion y la distribution de Platymantis centro de la figura represente la sistema original de Batanta + Isla Waigeo; achurado = Montanas Torri Guinea choco con est as islas, la ultima ocupada las irregu conduciendo a la situation mostrada en la figura 3. ymantis by the early Miocene. The row of squares in the land arc system. (From left to right: stippled = Ba- tains; open = Finnisterre Range.) When New Guinea dentations shown on the north coast of New Guinea, en el Mioceno inferior. La fila de cuadrados en el islas Orientales. (De izquierda a derecha: punteado = celli; bianco = Cerros Finnisterre.) Cuando Nueva laridades mostrados en la cosia norte de Nueva Guinea, 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 81 Fig. 4:5. Distribution of Platymantis papuensis on New Guinea and adjacent islands. Populations on islands east of the mainland may represent an undescribed species (R. G. Zweifel, pers. comm. ). Distribution dc Platymantis papuensis en Nucva Guinea y islas adyacentes. Las poblaciones de las islas al este del contincnte pudieran ser una especie no descrita (R. G. Zweifel, pers. com.). phryne in particular. Zweifel ( 1971 ) con- cluded that, despite certain equivocal fea- tures, Genyophryne was properly considered a member of the Sphenophryninae, and he proceeded to redefine the Asterophryinae and Sphenophryninae on the basis of distinctions of maxillae, dentaries, vertebral column and tongue. My studies of superficial mandibular musculature (Tyler, 1974a) provide addition- al data supporting such a recognition of two subfamilial units (Table 4:1). A further action by Savage ( 1973) of con- siderable impact was that of including within the Asterophryinae (sensu lato) Calluella, an Asian genus uniquely associated by Parker (1934) with the Malagasy Dyscophinae. Table 4:1. — Diagnostic Characters of Australopapuan Microhylid Frogs. (Data from Zweifel, 1971, and Tyler, 1974a) Character Asterophryinae Sphenophryninae Maxillae Dentaries Vertebral column Tongue Interhyoideus muscle Often overlapping premaxillae, and usually in contact In contact anteriorly (except in Hylophorbus) Diplasiocoelous Subcircular, entirely adherent, often with a median furrow and posterior pouch Anteriorly underlies intermandibularis ( except in Hylophorbus ) Not overlapping premaxillae Never in contact medially Not in contact Procoelous Oval, half-free behind, lacking median furrow and posterior pouch Does not underly intermandibularis 82 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 18 p H I L I P : 13 N E S (JJ PALAU BATANTA/WAIGEU NW. NEW GUINEA NEW GUINEA 6 JEW BRITAIN \ 1 AUSTRALIA *T\ SOLOMON \A ISLANDS FIJI [O] Fig. 4:6. Distribution and numbers of species of fi ana in Australia and the adjacent Oriental Region and Pacific area. Distribution y numcros de especies de Rana en Australia y la adyaccnte Region Oriental y areas del Pacifico. Unquestionably, the former union provided a biogeographic disjunction that was difficult to interpret. Nevertheless, to associate CallueUa with the Asterophryinae introduces new anomalies of even greater magnitude. This is because the Asterophryinae and Spheno- phryninae are composed exclusively of frogs exhibiting direct development. The inclusion of CallueUa among the Australopapuan spe- cies introduces species with a free-living larval stage. The magnitude of this introduc- tion can only be appreciated when the con- siderable diversity of scansorial, terrestrial, semi-aquatic and fossorial frogs is seen to be united by sharing uniformly similar ontog- enies. To accommodate CallueUa in the As- terophryinae ( sensu lato ) also has a profound biogeographic impact, extending the range of the subfamily from the Australian Region to as far as western China. Evidence contra- dicting this step can be obtained from bio- geographic and from morphological sources, 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 83 but it involves resurrecting and modifying concepts refuted by Savage. Few authors have contemplated the pos- sibility of microhylids being widely distrib- uted in Gondwanaland in the Cretaceous. Admittedly, the concept of the entry of the family into Australia from Indonesia ante- dated acceptance of continental drift and sea- floor spreading. However such an entry has been supported by many authors, of which Laurent (1975) is the most recent. Savage (1973) proposed the interesting hypothesis that the Microhylidae was present in the trop- ical portions of each of the southern land masses prior to the fragmentation of Gond- wanaland. He further put forward an in- genious account of a turbulent climatic his- tory for Australia, so as to account for the family's present abundance in New Guinea (13 genera, 95 species) and almost total ab- sence from Australia (2 genera, 7 species). Savage's hypothesis demands considerable mobility for the Australopapuan populations. In particular there is the need for extinction of the Australian component, followed by di- versification in New Guinea, and the subse- quent recolonization of Australia by immi- grants from New Guinea. The distribution of the Microhylids in Australia, New Guinea and the adjacent por- tion of the Oriental Region is shown in figure 4:7. The frogs are predominantly montane. Interpretation of the origin of the Australo- papuan microhylids must accommodate three important facts. 1) The highly adapted montane microhylids of New Guinea are unlikely to be any older than the orogeny of the mountains that they inhabit. 2) The geographically most widely distributed gen- era (Cophixalus, Oreophryne, and Spheno- phryne) all have representatives at low alti- tudes. 3) Microhylids do not occur in any part of southern Australia [therefore exclud- ing geographic areas affected by the aridity that Savage (1973) believed to have caused their demise]. Whereas Savage visualized the Australo- papuan microhylids as a Gondwanan element and thus a group whose origins involved a direct ancestry to South American frogs, I subscribe to the more orthodox opinion of an Oriental ancestry for the Papuan stock. This view can be supported on historical, biogeo- graphic, and morphological grounds. Pro- vided with the evidence of the nature of the collision of the Australian continental plate with the pre-existing chain of Oriental islands, an Oriental origin seems highly likely for the Australopapuan stock. Thus, Cophixalus, Oreophyrne, and Sphenophryne were prob- ably established within the chain at the time of the collision. The absence of these genera in the Solomon Islands and islands farther south, and the presence in New Britain of only a single species each of Oreophryne and Sphenophryne (Tyler, 1967) indicate that microhylids passed eastwards after the colo- nization of the same areas by ranids. It fol- lows that the ancestry of the Papuan micro- hylid fauna must be far less complicated than an examination of the diverse modern genera would indicate. The important criterion for their success appears to have been the in- herent ability to colonize the New Guinean montane environments that evolved during the rapid elevation immediately after the col- lision. Cophixalus, Oreophryne, Sphenophryne, and in fact all Australopapuan microhylids exhibit direct development. In this regard they differ from all Oriental microhylids. Di- rect development has enormous selective ad- vantage in situations where there is a shortage of suitable aquatic breeding sites. The first step in its evolution is probably acquisition of macrolecithal eggs without altering the met- abolic demands of the embryo. The potential for delayed emergence from the vitelline membranes would result. In terms of the anatomical structure of the tadpole, it follows that any deferment of the onset of larval life is most likely to permit economy in the elab- oration of the vast digestive system. This re- sults from the existence of increased food reserves, and of decreased demands upon the use of the larval digestive apparatus. There are several Oriental microhylids that exhibit trends towards delayed emer- gence. Inger (1966) provided ecological notes of species in which enlarged and un- pigmented ova have been found. In the genus Kalophrynus the eggs of some species are pig- mented, whereas in others such as K. pleuro- stigma they are unpigmented, and the larvae 84 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 C5 Fig. 4:7. The Australian and adjacent Oriental and Pacific areas showing the distribution of the Microhyli- dae (broken line) and the range of the most widely distributed Australopapuan genus, Oreophryne (con- tinuous line ) . Distribution de Microhylidae (linea entrccortada) y el rango del mas extendido de todos los generos austral- opapua, Oreophryne (linea continua) en Australia y las areas Oriental y Pacifico adtjacentes. have poorly developed intestines: ". . . only two loops visible ventrally and appears to be full of yolk." (Inger, 1966:135). The origin of the Asterophryinae from sphenophrynine ancestors has been consid- ered previously. Zweifel (1972:431) observed that of the asterophryine genera Hylopharbus ". . . differs from Cophixalus of the Spheno- phryninae only in having the tongue less free, and in having a diplasiocoelous rather than procoelous vertebral column. Therefore it may be that the Asterophryinae sprang from stock much like the present-day Cophixalus." Tyler ( 1974a ) similarly concluded that the super- ficial mandibular musculature of Ihjlophorhus is comparable to the uniform condition of Cophixalus and other sphenophrynines, and that the various conditions in the Astero- phryinae can be derived from the generalized sphenophrynine muscle pattern. The superficial mandibular musculature of South and North American microhylids ex- hibits a progressive trend of elongation of a single pair of slender, supplementary elements of the intermandibularis muscle (Emerson, 1976). In many respects these structures re- semble those found in Papuan sphenophry- nines, but the interhyoideus muscle is more closely involved in the vocal sac, and the structure of the vocal sac is distinctive. It forms an involuted pouch dorsal to the inter- mandibularis in at least some of the South American taxa, but there is no such trend in sphenophrynines, and in the asterophryines the interhyoideus forms a single sheet lying ventral to the intermandibularis (Fig. 4:8). Variation in microhylid muscle architecture on each of the major continents is shown in figure 4:9; the nature of the diversity is in- dicative of complex separate radiations. 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 85 suppl mand. interhy. Fig. 4:8. Superficial mandibular musculature of the Papuan mierohylid frog Phrynomantis stictogaster; ventral view with skin removed. Interhy. = inter- hyoideus; inland. = intermandihularis; mand. = man- dible; subm. = submentals; suppl. = supplementary element of intermandibularis. Musculatura mandibular superficial del batracio microhylido Phrynomantis stictogaster; visto desde la cara ventral con la piel rcmovida. Recently, our knowledge of mierohylid karyotypes has been extended to many addi- tional species, particularly by the contri- butions of Rogart and Nelson ( 1976 ) and Rlommers-Schlosser (1976). It is generally accepted that a diploid karyotype of 26 is primitive for several families, including the microhylids. This number is retained in the three Papuan and 13 Madagascan species studied, but is of variable occurrence else- where (Table 4:2). Varanidae The varanid lizards constitute a group of small to large animals that are more diversi- fied in Australia than elsewhere. The modern distribution of the family (and sole genus, Varanns) forms a broad arc from Africa to Australia; King and King (1975) supported an Asian origin. Formerly, varanids were dis- tributed far more extensively, extending far- ther north and occupying Mongolia through to Europe and to North America (McDowell and fiogert, 1954; Hoffstetter, 1961). The 19 Australian species of Varanus are placed in two groups — 1) small species lacking lateral compression of the tail and often referred to the subgenus Odatria, and 2) the large, typi- cal monitors. Hecht ( 1975 ) reviewed the fairly exten- sive history of the Varanidae, noting that the Late Cretaceous, Paleocene, and Eocene rec- ords from North America, Europe, and Mon- golia, combined with the absence of the fam- ily in South America are indicative of a Laurasian origin. The fossil record in Aus- tralia suggests that varanids may have entered Australia on two occasions. The giant Mega- lama prisca of southeastern Australia is known only from Pleistocene deposits. Al- though McDowell and Rogert (1954) synony- mized Megalania with Varanus, Hecht ( 1975) redescribed and redefined the fossil genus and provided adequate evidence to merit its recognition. Thus, it is possible that Mega- lania and Varanus represent separate inva- sions of Australia. Megalania prisca is the largest lizard known (total length up to 8 m and an estimated weight up to 600 kg). Fos- sil Varanus have been reported from the Mio- cene by Stirton, Tedford, and Miller (1961), from the Pliocene by Archer and Wade (1976), and from the Pleistocene by Smith (1976) and Molnar (1978). Thus Megalania and Varanus were contemporaneous. The success of the carnivorous and car- rion-consuming varanids in Australia was at- tributed by Storr (1964) to the absence of eutherian carnivores. Hecht (1975) suggested that Megalania represented the carnivore of the Australian megafauna preying upon some of the large herbivorous marsupials that were so abundant in the Pleistocene. Within South America the teiid genus Tupinambis is the only lizard approaching the niche filled by Varanus in Australia; Tu- pinambis is omnivorous. Scincidae Throughout the world there are over 800 species of skinks unevenly distributed among four subfamilies. The widespread distribu- tion of the family and varying interpretations of its systematics and phylogeny are biogeo- graphically undesirable attributes. The major systematic treatments are those of Mittleman (1952) and Greer (1970); I have adopted Greer's scheme. 86 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 FORMS OF MANDIBULAR MUSCULATURE IN MICROHYLIDS VENTRAL VIEW North America South America Madagascar Asia New Guinea Australia Africa Madagascar New Guinea New Guinea Fig. 4:9. Schematic representation of the orientation In each figure the muscle slip is shown on a single man plest form of this muscle is a slip at the apex of the the mandible as shown in A and B, or partially migrates and D). Representation esquenuitica cle la orientation de los crohylidos. En eada figura la portion muscular sc mucs tral. ha forma mas simple de este musculo cs como una 8. Esto migra posteriormente a lo largo de la mandi y se divide en dos porciones de varias formas (C y D). Greer considered the Scincinae to be the most primitive group within a distribution that is predominantly Laurasian but also oc- cupies the entire African continent and Mada- New Guinea of supplementary muscle elements in microhylid frogs. dible and is viewed from the ventral surface. The sim- mandibles as in figure 8. This migrates posteriorly along and then divides into two slips of various forms ( C elementos del musculo suplementario en batrachios mi- tra en una sola mandibula, y sc lo vc desde la cara ven- porcion en la punta de las mandibulas como en la figura bula como se muesira en A y B, o migra parcialmcnte gascar as well. Eumeces has a remarkably disjunct distribution, with isolates ranging from North Africa to India, China to Vietnam, and Middle to North America. Disjunctions 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 87 Table 4:2. — Microhylid Karyotypes Continent Genus Species 2n Authority North America Gastrophryne Hypopachm 2 2 22 22 Morescalchi 1968b, Bogart & Nelson 1976 Leon 1970, Bogart & Nelson 1976 South America Arcovomer Dermatoiwtus Elachistocleis Glossostoma Hamptophnjne Otophryne Stercocyclops 1 1 1 1 1 1 1 22 22 22 26 22 26 22 Bogart & Nelson 1976 Rabello 1970, Becak et al. 1970 Bogart & Nelson 1976 Bogart & Nelson 1976 Bogart & Nelson 1976 Bogart, Pyburn & Nelson 1976 Bogart & Nelson 1976 Asia Kaloula Microhyla Ramanclla Uperodon 2 1 1 1 24,28 26 26 26 Morescalchi 1968a, 1968b, Bogart & Nelson Bai 1956 Bai 1956 Natarjan 1953 1976, Sato 1936 Africa Breviceps Phrynomerus 1 1 24 26 Bogart & Nelson 1976, Morescalchi 1968a, . Morescalchi 1968a, 1968b, Bogart & Nelson 1968b 1976 Madagascar Anodontohyla Dyscophus Mantipus Platyhyla Platypelis Paracophyla Plethodontoh yla 2 4 1 1 2 1 o 26 26 26 26 26 26 26 Blommers 1971, Blommers-Schlosser 1976 Blommers 1971, Blommers-Schlosser 1976 Blommers 1971, Blommers-Schlosser 1976 Blommers 1971, Blommers-Schlosser 1976 Blommers 1971, Blommers-Schlosser 1976 Blommers 1971, Blommers-Schlosser 1976 Blommers 1971, Blommers-Schlosser 1976 New Guinea Cophixalus Phrynomantis" 2 1 26 26 Cole & Zweifel 1971 Cole & Zweifel 1971 ° Identified as "Asterophrys sp." by Cole and Zweifel, this individual was subsequently referred to Phrynoman- tis stictogaster by Zweifel (1972, p. 504). in at least some other components of the Scincinae are explained most readily in terms of independent origins from the Eumeces stock. The Acontinae and Feylininae are con- fined to southern, central, and western Africa and are regarded as independent derivatives of the Scincinae, whereas the Lygosominae are found throughout Australia, in all but the extreme south of South America, Middle America, southern North America, in most of Africa, and the entire Oriental Region. Greer (pers. comm.) considers that the present distribution and radiation of the Scin- cidae can be explained without reference to continental drift. Greer (1970:178) consid- ered the lygosomines ". . . are clearly derived from scincines and are morphologically the most advanced skinks." Thus, within Australia the problem of interpretation of the vast de- gree of diversification is the time available for that radiation and diversification. However, skinks evidently are adept at dispersal over sea water and the timing of their arrival in Australia remains unknown. No pre-Pleisto- cene fossils are known from the Australian Region which is scarcely surprising, consider- ing the state of the fossil record. Agamidae The absence of agamids from South Amer- ica and their present, virtually continuous range from Africa to Australia are highly in- dicative of the Oriental origin of the Aus- tralian component of the family. The rela- tively poor representation of the family in New Guinea reflects the ecological differences between New Guinea and Australia. The highly successful radiation of agamids in Aus- tralia is not unique to agamids, and probably reflects the wide variety of niches open to the early colonizers. Archer and Wade (1976) reported a small undetermined species similar to Amphibolurus from the Pliocene. MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Carettochelyidae and Trionychidae The carettochelyid turtles have an unusual distribution pattern, but, as yet, there is no evidence of their former presence in South America. Known from the Eocene of North America and the late Miocene to Present of New Guinea (Glaessner, 1942), the sole liv- ing representative Carettochelys insculpta was first reported from rivers of northern Aus- tralia by Cogger (1970), and since has been shown by Schodde, Mason, and Wolfe (1972) to be distributed quite widely in the Northern Territory. This species evidently has a high tolerance to salt water. Carettochelyids rep- resent a relict family, but it is not necessarily one of any great antiquity within the region. They probably represent the first trionychoid invasion that was followed by Pelochelys bi- broni, which did not extend beyond New Guinea. Both species probably entered the Australian Region in the Miocene. A triony- chid has been found in the middle Pliocene of Venezuela (Wood and Patterson, 1973). Elapidae Within Australia and New Guinea the elapid fauna is exceptionally diverse. Authors vary in the number of species and genera that they recognize, but Cogger (1975) recog- nized 26 genera and 61 species in Australia. Such numbers and diversity would seem to require a great evolutionary time span. This seems to conflict with the existence of endem- ism within the Solomon Islands (Salomone- laps and Loveridgelaps) and even as far as Fiji (Ogmodon) (McDowell, 1970). More- over, elapids extend in a slightly disjunct arc through to Africa; an Oriental origin seems likely for many of them. Whether this applies to all components of the Australian elapid fauna will have to await completion of the splendid work commenced by McDowell (1967, 1970). The elapid fossil record currently is con- fined to a species of Pseudonaja differing from P. nuchalis from the Pleistocene deposits at Naracoorte, South Australia, by Smith ( 1975 ) and Pliocene vertebrae from northern Queens- land tentatively referred to the Elapidae by Archer and Wade (1976). Colubridae The colubrids occur mainly in the northern and eastern portions of the Australian con- tinent. Some of the genera, such as Stegono- tus, range through Australia and New Guinea into the Oriental Region. The phylogenetic affinities of Stegonotus appear to be with the Oriental Dinodon (McDowell, 1972). An Oriental route of entry for the colubrids as a whole seems to be unquestionable. Acrochordidae and Uropeltidae Two acrochordids occur in New Guinea; in Australia they are restricted to the extreme north of the continent. These aquatic species are either both referred to Acrochordus, or one to that genus and the other to Chersydrus. Each species is distributed extensively in In- donesia and farther west, and they represent a recent Oriental invasion. The uropeltid genus Cylindrophis can be included within the fauna of the Australian Region because it reaches the Am Islands between Australia and New Guinea (McDowell, 1975). In other respects, it is an exclusively Oriental genus, and it has certainly entered the Aus- tralian Region very recently. Typhlopidae Typhlopids occur on almost all continents. Typhlops has a range almost equivalent to the entire family, extending throughout Asia to New Guinea. The Australian species now are referred to the genus Typhlina, which resem- bles Typhlops in external features, but differs substantially in the nature of the male geni- talia (Guibe, 1948; Robb, 1966). Cogger (1975) listed 22 species of Typhlina in Aus- tralia, and McDowell (1974) listed 11 from New Guinea and the Solomon Islands. Ty- phlina extends as far south as Fiji. The nature of the distribution pattern indicates an Orien- tal origin for the Australian component, but the date of its entry is uncertain. Its distri- bution in the southwest Pacific is extensive, and it may be well disposed to sea dispersal. Boidae The Boidae is described by Cracraft 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 89 (1973:384) as ". . . an excellent example of Gondvvanan dispersal." However, for Aus- tralia the boids may have had a rather check- ered history. The distribution of Python in an almost continuous arc from Africa to Aus- tralia provides further evidence of origin out- side Australia. There has been a fairly pro- nounced successful radiation within Australia, attributed by Storr (1964) to the absence of the Felidae. McDowell (1975) recognized three groups among the Australian Pythoni- nae, distinguished by the presence or absence of labial scale pits and prehensile adaptations to the structure of the tail. The first of these contains Liasis, and the second Python, More- lia and Chondropython. McDowell consid- ered the latter two genera to be only weakly defined and maintained that a good case could be made for referring them all to Python. Such is the state of herpetological exploration in Australia that a giant new species of Python was discovered in the Northern Territory in 1975 (Gow, 1977). However, the third group containing Aspi- clites was defined more satisfactorily by pos- sessing several features not exhibited by other genera. The Boinae is represented in New Guinea by two species of Candoia. Smith (1976) upset the concept of all of the Australian boids being completely attrib- utable to radiation from a northern entry of a Python ancestor. From the Naracoorte Caves in the southeast of South Australia, she described the fossil genus and species Wo- nambi naracoortensis. She considered Wo- nambi to be related most closely to Madtsoia bai ( Palaeocene-Eocene of Patagonia) and M. madagascariensis (Cretaceous of Mada- gascar). The Naracoorte Caves provide an ex- ceptionally rich fossil fauna and although of only late Pleistocene age, many forms of verte- brates recovered from them are now extinct. Wonambi naracoortensis was described from a series of vertebrae; Smith estimated that they were derived from a snake with a body length of approximately 5 m. A frag- ment of left maxilla associated with the verte- brae bore teeth approximately 7 mm in length. It differs from extant Australian boids in lack- ing accessory processes beneath the prezy- gapophyses, in possessing weak subcentral ridges and paracotylar foramina, and in ex- hibiting a slight posterior slope to the neural spine. If Wonambi is correctly associated with the Boidae, Australia may have been colo- nized by the family twice — an initial entry antedating the drift of Australia from Antarc- tica, and a second entry, presumably in the Miocene. If so, when Python first appeared in the north of the continent, Wonambi or its ancestors inhabited at least the southeastern part. The only other fossil record of Australian boids is the report by Archer and Wade ( 1976 ) of three vertebrae of a very large species in the lower Pliocene Allingham For- mation in north Queensland. These authors do not associate it with a modern species, but noted (op. cit: 385) that it is ". . . mor- phologically very similar to modern species of Morelia." GONDWANAN ELEMENTS As demonstrated here, the Gondvvanan elements are predominantly anuran. The fos- sil record is only just being assembled, and it is best dealt with here rather than within the individual families. Australian Fossil Frog Record Until recently there were no known fossils of frogs in the Australian Region. Tyler ( 1974a ) reported the discovery of an isolated left ilium amongst a rich vertebrate assem- blage taken at Lake Palankarinna, north of Lake Eyre, in the northern part of South Aus- tralia. Subsequently, this ilium became the holotype of the new genus and species Aus- tralohatrachus ilius, tentatively referred to the Hylidae (Tyler, 1976). More recently, 19 more ilia have been found in sediments from Lake Palankarinna (Tyler, unpublished). These include a number of additional speci- mens of A. ilius, several specimens of an un- described species of the leptodactylid Limno- dynastes and also a Litoria species closely re- lated to, and possibly representing L. caeru- lea. The age of the Lake Palankarinna fauna is uncertain but is considered to be most likely mid-Miocene. Thus, the age coincides with the collision of the Australopapuan and Ori- ental plates. 90 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 A rich Pleistocene fauna, including 166 frog ilia of extant species, has been taken at two cave sites at Naracoorte in the lower southeast of South Australia (Tyler, 1977). The fossils include Litoria ewingi, Limno- dijnastes cf. dumerili, L. tasmaniensis, Rani- della signifera, and Geocrinia cf. laevis, all of which occur in that area today. Unfortunate- ly, as yet there are no known frog fossils of an age predating the Australopapuan-Orien- tal collision. However, it is noteworthy that Limnodijnastes and Litoria were established in the Miocene at Lake Palankarinna. Hylidae Until about 1970 the confamilial status of the hylids of Australia and New Guinea ( "Australopapuan" ) with those of South America had not been disputed, and in fact Hyla was applied quite uniformly to Aus- tralopapuan and Neotropical species (creat- ing numerous problems of homonymy but ap- parently not of biogeography). The disjunct nature of the distribution of Hyla (sensu lato) was the principal area of interest long before the homogeneity of the genus (and later of the family) was really questioned. Within the framework of a concept of static continents, Parker (1929) suggested that the Hylidae might be of North American origin, radiating from there in several different direc- tions, of which South America and Australia were the ultimate destinations of two. The major gap between the depauperate Oriental Hyla and the relatively numerous Australian species remained a serious obstacle to ade- quate explanations of dispersal, although Dar- lington (1957) suggested that the arboreal rhacophorids have displaced and now replace the hylids in that intermediate area. The first suggestion of a direct association between South American and Australian hylid species was made by Reaufort ( 1951 ) , who visualized an Antarctic land bridge as a route for entry to Australia from South America. During the past decade fresh interpretations of the phylogeny and systematics of the hy- lids of Australia and South America have far exceeded the comparative faunal studies de- sirable to support some of the conclusions reached. Tyler (1971) studied a suite of characters associated with the superficial mandibular musculature, and the vocal sac that it con- tains, in representatives of numerous families and including almost all known hylid genera. He noted that anatomical divergence occurred principally in association with taxonomic units recognized as genera. He demonstrated that the Australopapuan Hyla constituted a single morph distinguishable from species from other parts of the world. Consequently he proposed the resurrection of Litoria Tschu- di to accommodate the Australopapuan spe- cies and further considered Litoria and Nyc- timystcs to be a unique, monophyletic group. Cracraft (1973) reported the above find- ings without variation, but in 1974 suggested only that the distinction of Asiatic and Aus- tralopapuan species had been demonstrated, and he introduced the topic of Australo- papuan-South American hylid affinities as though it were a new proposal. Savage ( 1973) acknowledged some of my data in press as evidence refuting the confamilial status of the two groups of frogs, and accordingly resur- rected the family name Pelodryadidae Giin- ther to accommodate them. Ragnara ( 1974 ) published rather conflict- ing data. From his discovery of the occur- rence of rhodomelanochrome in the skin of Neotropical phyllomedusines and some, but not all, Australian hylids, he made two radi- cal proposals. These were 1) that the Aus- tralian species of Litoria exhibiting rhodo- melanochrome are more closely related phylogenetically to South American phyllo- medusines than to sympatric congeners, and suggested therefore, 2) that Litoria was a highly heterogeneous assemblage. More re- cently, Ragnara (1976) reiterated the com- mon origin of the phyllomedusine-L/Yon'a component more forcefully. Laurent ( 1975 ) envisaged a totally differ- ent origin for the Australopapuan species, sug- gesting independent origin in situ from a for- merly world-wide leptodactylid ancestor, which he visualized as the ancestral stock of several families on different continents. Con- templating a wholly autochthonous origin he employed the name Nyctimystinae for the Australopapuan fauna. In an attempt to stabilize the classification 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 91 of these and other frogs, Duellman ( 1975 ) offered a number of solutions to controversial issues. Among his actions, he continued to include Australopapuan species in the Hy- lidae. In order to crystalize the current contro- versy and to aid the transcontinental study of the Hylidae, it is necessary to restate or clarify the following issues: 1) the phylogenetic re- lationships of the Indonesian hylids ( those on the periphery of the Australian population and geographically closest to the Oriental species); 2) whether the Australopapuan species genuinely constitute a monophyletic group; 3) the phylogenetic relationships of the Australian species; and 4) the phylo- genetic relationships of the Australian and South American hylid faunal units. The Indonesian hylid fauna. — The north- western geographic limit of Litoria occurs in the Indonesian islands of Timor and the Less- er Sunda Islands of Sumba, Savu and Alor. This latter assemblage represents the eastern end of an archipelago forming an intimate link to the Malaysian Peninsula far to the northwest. Therefore, the Litoria fauna of the Timor-Lesser Sunda group is of impor- tance to any contemplation of entry of hylids into Australia from the northwest. The only species (L. everetti) occurring in the relevant area is a member of the L. peroni group represented elsewhere in the Australopapuan area by five described species — peroni, everetti, amboinensis, rothii and darlingtoni. The total geographic range of this species group is exceptionally extensive (Tyler and Da vies, 1978a). It appears that the group evolved in Australia or New Guinea and is now radiating in several directions and extending its range. Thus, it is confirmed that the phylogenetic affinities of L. everetti are with other Australopapuan species and not with the southernmost Oriental hylid (Hyla chinensis). Monophyletic or polyphyletic origins. — In- sofar as the Australopapuan fauna is con- cerned, the issue is whether the frogs referred to the Hylidae are a monophyletic group. Ecologically, at least, they are incredibly di- verse, filling a spectrum of niches occupied on other continents by different families. A su- perficial examination of the diversity of struc- ture may render the casual observer critical of Australopapuan hylid systematics. Neverthe- less the magnitude of diversity need be no indication of polyphyly. Tyler ( 1971 ) pro- posed the concept of monophyletic origin on the basis of his studies of superficial mandibu- lar musculature and vocal sac structure. Sub- sequent karyotypic data assembled by Steph- enson and Stephenson (1970), Woodruff ( 1972 ) , Morescalchi and Ingram ( 1974 ) , and Menzies and Tippett (1976) have in no way caused this concept to change. All of the hylids karyotyped to date have 2n = 26, ex- cept L. infrafrenata (2n = 24), and in that instance a model for derivation from 2n = 26 has been proposed (Menzies and Tippett, 1976). Phylogenetic relationships of Australian species. — Tyler and Davies ( 1978a ) examined the morphology, osteology, myology, distri- bution, and biology of 92 of the 94 species of Litoria currently recognized. They found that these species can be associated in no less than 37 species groups. Insofar as all geographic areas occupied by Hyla are concerned, this total of groups is not exceptionally high. In reality the number of species per species group is remarkably similar in several geo- graphic areas (Table 4:3). Although more concerned with the initial step of establishing phonetic groupings, Tyler and Davies demon- strated that Australian hylids occupy an in- credible gamut of niches. Tyler (1970, 1972a) suggested that there exists a close phylogenetic relationship be- tween Australian hylids and leptodactylids. The core of this suggestion related to the lep- todactylid genus Cyclorana. That genus as then constituted comprised a group of squat- bodied and also some elongate species. Lynch ( 1971 ) regarded them an integral component of the Australian leptodactylid fauna. Tyler (1971, 1972a) studied superficial mandibular musculature in all Australian leptodactylids and his conclusions differed from those of Lynch only in his appraisal of Cyclorana, in which he noted distinct hylid affinities. Tyler (1970) suggested that the similarities of the Australian hylids and leptodactylids implied the existence of a single common ancestor. 92 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Table 4:3. — Relationship Between Number of Species and Number of Species Groups of Selected Hylid Genera on Different Continents. Geographic area Genus No. of species No. of groups Species/Group Australia Litoria 46 22 2.09 : 1 New Guinea Litoria 51 24 2.12 : 1 Australia & New Guinea" Litoria 94 37 2.54 : 1 Middle America" * Hyla 73 28 2.61 : 1 North America"" Hyla 12 4 3.00 : 1 " Includes a component common to Australia and New Guinea. *• Data from Duellman (1970). Robinson and Tyler ( 1973 ) examined the relative predominance of the catecholamines epinephrine and norepinephrine (dopamine was not detected) in the adrenal glands of various Australian hylids and leptodactylids, including Cijclorana. The found that epi- nephrine was the predominant transmitter in all hylids examined and that norepinephrine was predominant in all leptodactylids, except Cijclorana. In the absence of any biochemical convergence associated with ecological con- vergence, they regarded catecholamine selec- tion as an exceptionally conservative feature. The major issues to be explored are, as follow: 1) Do the Australopapuan hylids ex- hibit a close phylogenetic relationship with any of the sympatric leptodactylids, or 2) is there a closer relationship with South Amer- ican hylids or leptodactylids? Hence it is conceivable that Australopapuan hylids and South American hylids enjoy a reasonably close relationship, or that the Australopapuan hylids are independently derived from a South American leptodactylid stock, and that resemblance only reflects convergence. Before exploring the nature of the resem- blance of hylids from each of the continents, it is worthwhile discussing here the novel proposition introduced by Bagnara and Ferris (1975) on the basis of the significance of the presence of rhodomelanochrome in some but not all Litoria (see p. 90). Maxson (1976) rejected this conclusion on the basis of im- munological data, and Tyler and Davies (1978b) reexamined the phylogenetic rela- tionships of the same species studied by Bag- nara and Ferris, together with additional spe- cies. Employing features of adult myology, osteology, larval structure, and reproductive biology, as well as pigment, their findings conflict with those of Bagnara and Ferris. They found that Australian congeners are genuinely more closely related with one another than some Litoria are with phyllo- medusines. Their findings also tend to re- inforce further the recognition of the phyllo- medusinae as a subfamilial unit. Some of the biochemical features employed by Bagnara as indicators of close phylogenetic relationships tend to be particularly subject to convergence, and Guttman ( 1973 ) outlined the problems that arise when the relationships of higher taxa are based on such characters. Phylogenetic relationships of Australian and South American hylids. — Numerically at the level of species, genera, and families, South America is clearly the most major cen- ter in the world of evolution of modern taxa of anurans. There is no reason to suppose that the Hylidae evolved in Australia and migrated to South America, but the obvious reverse option raises a number of most inter- esting questions. When some South American hylid frogs are placed directly beside Aus- tralian frogs and compared one with another, the confamilial association is stretched to the limit. South America simply does not possess frogs that resemble the Litoria aurea, L. cae- rulea and L. freycineti groups. Similarly, there are numerous dominant components of the South American fauna that do not have coun- terparts in Australia. Irrespective of the magnitude of diver- gence between South American and Australian species, the fact remains that all possess inter- calary structures, and so are referred to the family Hylidae. On a continental basis the species are indeed different. Savage (1973), using Tyler's ( 1971 ) data, considered the dis- tinction adequate to merit family status, and so termed the Australopapuan unit "Pelodrya- didae." I too recognize the morphological distinction and the monophyletic origin of the Australopapuan unit, but I find only adequate 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 93 grounds for an intrinsic division within the current concept of the Hylidae. Therefore, I favor recognition of a subfamily to accommo- date Litoria, Nyctimystes and (as indicated in later discussion) Cyclorana. Because all family group names are of equal status for the purposes of nomenclatural priority, Pelo- dryadinae (derived from Pelodryadidae Giinther, 1858) takes priority over Nyctimy- stinae Laurent, 1975. It is worth noting that when Maxson and Wilson (1975) implemented Savage's concept of the Pelodryadidae, because of results of estimated mean albumin-immunological dis- tances between continental populations, they overestimated the continental divergence time. Their immunological distance of 100 units equates with 60 million years, so that the existence of 100 immunological units be- tween any two taxa involves acceptance of 60 million years isolation between the popula- tions. Their calculation of an immunological distance of approximately 129 units between the relevant Australian and South American populations can be interpreted in two ways, but may well be excessive. The physical separation of Australia from Antarctica is now established at 52-55 m.y.b.p. This total com- pares with 77 m.y.b.p. calculated by immu- nological techniques. If the latter is the period of isolation of the stocks, ecological or physical barriers on the Antarctic land mass are called for to explain the separation of populations prior to rifting. Wallace, Maxson, and Wilson (1971) found greater immunological distances exist- ing between South American and the adjacent North American species, than between North American and the geographically distant sin- gle Australian species examined. Maxson (1978) demonstrated a high degree of compatability between North American and European Hyla. Maxson and Wilson (1971) noted that where discrepancies exist between organismal resemblance and albumin resem- blance, it is to be attributed to differential rates of organismal evolution. As an example they cited Acris, which exhibits albumin and haemoglobin affinities to North American Hyla, and yet is strikingly different from such species in anatomy, gross structure, biology, and ecology. Duellman (1970:647) accepted such evidence with considerably less toler- ance: "Despite the divergent nature of Acris with respect to other hylids, and the super- ficial similarity of Acris to ranids, the inescap- able fact remains that Acris has procoelous vertebrae, an arciferal pectoral girdle, inter- calary cartilages and claw-shaped terminal phalanges — a combination of characters that seemingly inextricably ally the genus with the hylids." It is equally reasonable to suggest that organismal evolution is unlikely to be con- strained along any linear path of morpho- logical divergence as assessed by human ob- servers. Hence, systematists have a quandry that is of their own making, and while the Hylidae remains defined as it is now, the Pelodryadinae remains an integral component of it. Cyclorana is a problematic genus. By virtue of the fossorial habit of most of its species and the absence of intercalary struc- tures it formerly has been accommodated in the Leptodactylidae (Parker, 1940; Lynch, 1971). More recent evidence has demon- strated similarities between Cyclorana and pelodryadine hylids in myology (Tyler, 1972a), adrenal catecholamines (Robinson and Tyler, 1973), in larval structure and biol- ogy (Watson and Martin, 1973), and in cranial osteology (Fig. 4:10). A closer exam- ination of the moqjhology of Cyclorana spe- cies resulted in the discovery of intercalary structures in C. inermis, C. alboguttatus, and C. dahlii, and led to these species being re- ferred to the hylid genus Litoria by Straughan (1969), Tyler (1974b) and Tyler, Davies and King (1978) respectively.1 In consequence of these actions and of the resurrection of one species and the descrip- tion of five new species (Tyler and Martin, 1975, 1977), Cyclorana now is composed ex- clusively of robust fossorial frogs lacking in- tercalary structures but retaining a closer af- finity to hylid than to leptodactylid frogs. Awareness of this presumably led Heyer and Liem (1976) to omit Cyclorana from their 1 The customary term "intercalary cartilages" is not used because these structures are bony in 46 of 71 Australopapuan hylid species studied (Tyler and Davies, 1978a). Ossification bears no correlation with finger length, habits or geographic distribution. However, all large or moderately large arboreal species retain a cartilaginous state. 94 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Cyclorana brevipes Cyclorana australis 5mm 5mm Litoria raniformis Litoria alboguttata Fig. 4:10. Skulls of certain species of Cyclorana and Litoria. Crdneos de ciertas especies de Cyclorana ij Litoria. phylogenetic analysis of the Australopapuan leptodactylidae ( Myobatrachidae ) . Because all hylids exhibit axillary amplex- us and Australian leptodactylids (except Mix- ophyes) inguinal, it follows that the embrace of Cyclorana should be of relevance in de- termining its phylogenetic relationships. I have observed amplexus in four species of Cyclorana. Initially the grasp is high in a circumcervical position, as though the male intends to strangle his mate. However the grasp slides posteriorly to an inguinal position. Within the Hylidae, Cyclorana appears to be related most closely to the Litoria aurea species group, which now includes two spe- cies previously referred to Cyclorana. This group may prove to be the sister group of Cyclorana and merit elevation to distinct ge- neric identity. Leptodactylidae (Myobatrachidae) It is valid to describe the current state of nomenclature and phylogcny of Australian leptodactylid frogs as distinctly unstable. Even to refer to them here as Leptodactyli- dae rather than Myobatrachidae is in total opposition to the sincere efforts of many workers. I do so now because I seek a re- evaluation of the steps that led to the nomen- clatural change, and because I suspect that the examination of South American-Australian 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 95 relationships is best served by nomenclatural conservatism. Parker's (1940) review of Australasian species brought a considerable degree of stability to the nomenclature of many Aus- tralian genera. Parker recognized two sub- families— Cycloraninae (which he erected) and the Myobatrachinae. Lynch (1971) pro- vided a splendid historical account of the classification of the Leptodactylidae, and in so doing, placed Parker's contribution in an historical perspective. Following the publica- tion of Parker's monograph, A. R. Main and his colleagues undertook the first detailed biological, herpetofaunal studies. The result of their work (and of that of their students) was the description of numerous new species. Nevertheless, the genera recognized and sus- tained by Parker were in no way challenged. Certainly Neobatrachus was resurrected from the synonymy of Heleioporus by Main, Lee, and Littlejohn ( 1958 ) , but in over 25 years only one new genus was erected — Tauclac- tyJus by Straughan and Lee ( 1966 ) . At that time it was tempting to assume that Austral- ia's leptodactylid fauna was already reason- ably well established. However, as late as 1960, only 59 of the currently recognized total of 79 species had been discovered. ( Cyclo- rana has been excluded from these totals.) Close examination of some of the numerically large genera then recognized (e.g., Crinia) indicated an unsuspected heterogeneity. In fact, because most Crinia were small species, the genus had in reality become a repository for a great variety of frogs only sharing small stature. It followed that closer examination led to the erection of some new genera and the resurrection of others (Tvler, 1972b; Blake, 1973). Collection of Rheobatrachus silus by Liem (1973) represented one of the most extra- ordinary herpetological discoveries of this century. In its gross morphology as an aquatic frog with profuse dermal, mucous glands, fully webbed toes, long pointed fingers and incredible aquatic maneuverability, the re- semblance to pipids such as Xenopus, and particularly to the South American lepto- dactylid Telmatobius, is extremely striking. Rheobatrachus silus was found to be even more noteworthy when Corben, Ingram, and Tyler ( 1974 ) reported that the female broods the larvae within her stomach. Subsequently, a robust-bodied, fossorial frog was found living in coastal sandhills in a remote and arid part of Western Australia. Named Arenophryne rotunda by Tyler ( 1976c ) , this genus has affinities with both Myobatrachus and Pseudophryne. More re- cently, another new and as yet undescribed genus (Tyler et al., 1979) was discovered in the northern portion of Western Australia and the Northern Territory. This form produces a foam nest, has tadpoles with suctorial mouths and elongate tails, and the adult exhibits enor- mous tympana. The subdivision of existing genera ini- tiated by Tyler (1972b) and Blake (1973) took a further, and more radical, step with actions of Heyer and Liem ( 1976 ) who de- scribed three more new genera to accommo- date known species — Paracrinia for Crinia haswelli; Australocrinia to accommodate two southeastern species referred to Ranidella by Blake (1973); and KankanopJiryne for Pseu- dophryne occidentalis. They also resurrected Platyplectron without defining it or naming the constituent species. Unfortunately the data on which this study- by Heyer and Liem is based are, as yet, un- published, being available only in a paper by Liem cited as "in press." Thus, it is simply not possible to comprehend or assess several of the decisions reached by these authors and I am unable to recognize the genera in the following discussion. In attempting to provide a brief resume of the case for the familial and subfamilial status of the Australian frogs, I must at the outset put forth the evidence that has been provided for family distinction. Lynch ( 1973 ) and Savage (1973) well may be considered the prime initiators of the concept of the Myobatrachidae as a family unit distinct from the Leptodactylidae. In their respective views of the classification of the Anura they, and other authors, differed in many respects, and Duellman ( 1975 ) attempted to synthesize the various expressed opinions and produce a classification that constituted a compromise. Duellman recognized the distinctness of the Myobatrachidae, but his brief diagnoses did not include a single nongeographic feature by 96 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 which the majority of species of one family could be distinguished from those of the other. I draw attention to this fact not in an attempt to score a point, but solely to high- light the fact that the case for considering the Myobatrachidae a separate family still needs to be substantiated. Duellman (1975) was able to accommo- date Rheobatrachas within the Myobatrachi- dae, arguing that neither the suite of primi- tive morphological character states, nor the bizarre reproductive mode should exclude it. Lynch ( 1971 ) employed a graphic technique enabling him to compare the relative primi- tiveness of a large number of nonarchaic frogs. If Rheobatrachus is added to Lynch's frog groups and the 13 nonreproductive charac- ters are scored, Rheobatrachus has a total score of 0. It is the bizarre reproductive state that produces a positive sum. I wholly sup- port Heyer and Liem's ( 1976 ) action of plac- ing Rheobatrachus in a separate subfamily, the Rheobatrachinae. Therefore, within the Australian Region there is an enormous diver- sity of animals in terms of the nature of char- acter states, and a crucial question is whether the disjunction between the Australian and South American subfamilies is best reflected by regarding them as members of different families. Intrasubfamilial variation is more extensive among the Australian subfamilies than in any other comparable units on other continents, and it is this variation that renders their definition so difficult (Lvnch, 1973:170- 171). The South African Heleophryne forms the Heleophryninae, which Lynch placed in the Myobatrachidae. It would be extremely in- teresting to test this assignment by means of the comparison of serum albumins employing the techniques for frogs of Wallace, Maxson, and Wilson (1971). Because the African con- tinent separated from Gondwanaland be- tween the mid-Jurassic and mid-Cretaceous (100-155 m.y.b.p.), but South America from Gondwanaland in the Cenozoic (25-45 m.y.b.p.), the absence of myobatrachids from South America is curious indeed. Morescalchi ( 1973 ) and Morescalchi and Ingram ( 1974 ) noted that Cyclorana alboguttatus (now Li- toria albo guttata) is karyologically less dif- ferentiated than many Australian leptodac- tylid genera and approaches some primitive leptodactylids from other geographical areas ( Heleophryne, the Ceratophryinae, many Tel- matobiinae, all with 2n = 26). On gross morphological grounds, the re- semblance between some Australian and South American genera is striking. If the South American Ratrachyla should be found tomorrow in the cool, temperate forests of the southern section of the Australian Great Di- viding Range, it would be compared with Kyarranus and Philoria, found to be highly similar, and attract little comment. Lynch ( 1973 ) pointed out that there were several systematic options in any cladistic study of the leptodactyloid frogs, and he maintained that separating the Limnodynas- tinae ( "Cycloraninae" of Lynch without Cyclorana), Heleophryninae, and Myobatra- chinae from the Leptodactylidae, reduced the gradation of characters within the latter fam- ily. The knowledge of each of these units remains incomplete, and other more radical options are still open. For example, on bio- geographic grounds, frogs associated with he- leophrynines or myobatrachines should occur among the cool temperate Austral fauna of South America. Alternatively, other data might reinforce the existing myobatrachine- telmatobiine links, or the degree of distinc- tion between the Myobatrachinae and the Limnodynastinae, thereby meriting independ- ent family status for each of the latter. For the present, there seems to be a good case for including the Australian species in the Lep- todactylidae while the other avenues are being explored. Gekkonidae In recent years the status of the higher taxa of gekkonid lizards has attracted consid- erable attention. Underwood (1954) recog- nized three families — Eublepharidae, Sphae- rodactylidae, and Gekkonidae, with two subfamilies (Gekkoninae and Diplodactyli- nae ) . Kluge ( 1967 ) recognized only one family (the Gekkonidae) containing each of the other four units as subfamilies. Some of Kluge's concepts of the origins, dispersal, and evolutionary relationship of these subfamilies have been variously criticized by Maderson 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 97 (1972), Moffat (1973), and Russell (1976). Kluge's ( 1967 ) interpretation was made with the assumption that continental drift was neither tenable as an hypothesis, nor germane to the resolution of the study. This led to conclusions that must be reexamined within the context of continental drift as an accept- able hypothesis. For example, there is his assumption that the Australian Diplodactyli- nae evolved from a primitive southeast Asian gekkonid stock during the Late Mesozoic. However, for that entire era, Australia lay far to the south and was still united to western Antarctica; the rift and long northwards drift towards southeast Asia only commenced in the early Cenozoic. Therefore, as Cracraft ( 1975) pointed out, the prospect of a success- ful overwater dispersal of geckos from Asia to Australia in the Mesozoic is remote indeed. In addition to hosting the pantropical gek- koninae, the endemic subfamilies of South America and Australia probably exhibit simi- lar historical patterns of gekkonid evolution. Certainly a significant feature of the disjunct- ly distributed eublepharines is their absence from South America, Madagascar, and Aus- tralia. Such a distribution is explained most readily by adopting Kluge's concept of an African or Asian site of origin. The general consensus of opinion is that the sphaerodac- tylines of South America and the diplodacty- lines of Australia evolved within their present geographic ranges. The diplodactylines ex- tended to New Caledonia, the Loyalty Islands, and New Zealand evolving in New Zealand to form an ovoviviparous group. Arrival of the diplodactylines in New Zealand has been sug- gested to be a Miocene event, but if the subfamily existed in Australia in the Cenozoic it could have entered New Zealand via the Lord Howe Rise. Many of the gekkonines are remarkably well suited to transoceanic dispersal. Phyllo- dachjlus has an incredible geographic range, occurring in the Americas, Africa, Madagas- car, Australia, New Caledonia, New Zealand, and the Galapagos Islands. However, Dixon and Anderson (1973) indicated that Phyllo- dactylus may be heterogeneous; species in the Eastern Hemisphere should be separated ge- nerically from PhyUodactyhis. A number of gekkonines are dispersed by man. Recent karyotypic studies in Australia have demon- strated within well-established gekkonine and diplodactyline species the existence of numer- ous biological species, none of which has yet been accorded formal taxonomic status ( King, 1973, 1975, 1977; King and Rofe, 1976). For example, King (1977) recognized five dis- cretely distributed and chromosomally dis- tinct populations within what is now termed Diplodactylus vittatus. When all of these new taxa are named, it is clear that the Australian gekkonid fauna will be vast numerically. Chelidae Presently, the pleurodire chelonians are restricted to South America, Africa, Australia and New Guinea. Of the constituent families now extant, the Chelidae occurs in the Neo- tropical and Australian Regions, whereas the Pelomedusidae occurs in Africa, Madagascar and South America. In terms of diversity and radiation, the South American genera and spe- cies have been more labile, and study of the extant Australian members suggests an ex- ceptionally conservative history. In a phylo- genetic study, Gaffney (1977) envisaged the Australian genus Pseudemydura as possibly the sister group of all other Australian and South American genera. At present, there are four recognizable genera of extant Australian chelids — Chelo- dina, Elseya, Emydura and Pseudemydura. Fossil records are becoming quite numerous, but the majority of them are based upon totally disarticulated shell fragments lacking sufficient detail to provide adequate diagnos- tic characters for generic determinations. The first report of fossil chelids from Aus- tralia is that of Lydekker (1889) who re- ported fragments of CheJodina and Emydura from various localities. De Vis (1894) pro- posed Trionyx australiemis for a substantial quantity of fragments (probably chelid) taken at Darling Downs, and then (1897) described one new genus and four new spe- cies from the same or adjacent localities. These specimens are in urgent need of re- view. Records of Pleistocene chelids from Queensland reported by Longman ( 1929 ) are fragmentary and lack diagnostic characters (Warren, 1969). Evidence of the conservative nature of the 98 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 chelid fauna was provided by Warren ( 1969 ) who reported Emydura sp. aff. macquari from siltstones in Tasmania. Reported to be of Oligocene-Miocene age they are now con- sidered somewhat older and of Early Tertiary age (J. W. Warren, pers. coram.). Chelids are no longer extant in Tasmania, and E. mac- quari is now confined to the Murray-Darling drainage system of the southeastern Australian mainland. With the discovery of freshwater turtle remains from Tasmania and also from various Miocene to Pleistocene deposits on the mainland (Callen and Tedford, 1976; Archer and Wade, 1976), it is likely that previously chelids have occupied almost all of the Aus- tralian continent and New Guinea as well. Contraction of their ranges is probably a Pleistocene phenomenon. The intercontinental relationships of che- lids is, superficially at least, extremely close (e.g., Elseya and Platemys). However, the Lower Cretaceous Chelycarapookus arcuatus Warren (1975) (Chelycarapookidae) (previ- ously identified erroneously as Emydura mac- quari by Chapman, 1919) needs to be exam- ined. Unfortunately, with the posterior por- tion of the plastron of that fossil missing, whether the pelvic girdle was attached or not remains unknown, so that even the infra-order position of the family is uncertain. Warren ( 1975 ) noted that in Chelycarapookus neurals probably were present between all costals, and suggested tentatively that with loss of the neurals, Australian chelids could have been derived from a chelycarapookid ancestor. However Rhodin and Mittermeier (1977), ap- parently without sighting the description of Chelycarapookus, reported that neurals occur regularly in one species of Australian chelid and irregularly in several other species. Gaff- ney's ( 1975 ) study of the phytogeny and classification of turtles almost exclusively re- lied upon cranial characters, and, in common with most Australian chelid remains, the head and neck of Chelycarapookus remain un- known. Crocodiles The crocodiles include two quite distinct components differing in their ancestry and arc probably of separate Gondwanan and Orien- tal origins. The extant species of Crocodylus are clearly of Oriental origin or derived from an Oriental stock, and are confined to the north of the Region — C. porosus of south- eastern Asia, New Guinea and Australia, C. novaeguineae of New Guinea and C. johnsoni of Australia. The fossil fauna is substantial both in quantity and diversity, and a number of highly significant finds has been reported re- cently. Molnar ( 1977 ) described from Chil- lagoe in North Queensland an incomplete skull, with a high and laterally compressed snout and probably xiphodont dentition. The subsequent discoveiy of Pleistocene Palor- chestes cf. P. azael at the same site was in- terpreted as evidence of a Pleistocene age for the xiphodont (Molnar, 1978), and indi- cated that xiphodonts had survived in Aus- tralia long after their extinction elsewhere in the world. Hecht and Archer ( 1977 ) reported two forms of xiphodonts from the Pleistocene of South Australia and southeast Queensland, re- spectively. The former is reported to compare favorably with the type of the sebecosuchian Sebecus icaeorhinus from the Eocene of Pata- gonia. The authors suggested that the as- sumed sebechosuchian Planocrania datangen- sis from the Early Tertiary of China is in reality probably an eusuchian, so that origin of the Sebecosuchia is clearly from Gond- wanaland. AUTOCHTHONOUS ELEMENT Pygopodidae The legless, fossorial lizard family Pygo- podidae is the only reptilian family restricted to the Australian Region. Kluge (1974) rec- ognized eight genera and 30 species of pygo- podids. Kluge's (1976) analysis of phylo- genetic relationships within the family led to the recognition of only six genera. Following the work of Underwood (1957), there has been general acceptance that the pygopodid- gekkonid relationship is extremely close, and it follows that the most likely origin for the pygopodids is within Australia directly from a gekkonid stock. Many pygopodid species are restricted 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 99 either to the southwest or to the southeast of the continent, a distribution pattern common to numerous vertebrates and attributed to speciation in the Pleistocene, associated with the glacial-interglacial climatic oscillations. Thus, the implication of the existence of such distribution patterns is that speciation of these same populations is not of any great antiquity. Somewhat in conflict is the total absence of pygopodids from the southern island of Tas- mania and the presence of only a single spe- cies on Kangaroo Island. If these indicate that pygopodids occupied the adjacent main- land only after isolation of these islands from the mainland (8,000-10,000 y.b.p.), the southern speciation pattern evidently is more complex. Western Pacific Island Faunas and Oceanic Dispersal Although South America now is separated from Australia by the vast expanses of the Pacific Ocean, the nature of the major oceanic surface currents provides a mechanism for westward dispersal of animals by rafting from South America in the direction of Australia. The west coast of South America is swept by the north flowing Humboldt Current which meets the transpacific southern equatorial cur- rent at mid-latitudes. The latter current travels westward and eventually disperses around all of the islands of the Pacific south of the Equator. Thirty years ago Thor Heyerdahl's raft Kon-Tiki demonstrated the transport potential of the southern equatorial current by travel- ling from Peru to the Tuamotu Archipelago in French Polynesia south of the Marquesas Islands. Had his vessel been driven two or three degrees northwards, he woidd have passed between the Marquesas and the Tua- motu Archipelago, and continued much far- ther west on a longer journey terminating in Western Samoa, Fiji, or Tonga. This longer journey is probably the route of the ancestral stocks of the iguanids of Fiji and Tonga — Brcichylophus fasciatus and B. brevicephalus, respectively. South American derivates are not evident elsewhere, and the remaining ele- ments of the herpetofauna of the islands in the west and southwest Pacific area are repre- sentative of two other sources — 1) overwater dispersal principally from the north and north- east, and 2) land communication with Aus- tralia via the exposed Lord Howe Rise. Fiji and New Zealand are the most south- erly of the large landmasses in the Pacific and merit specific mention. Fiji. — In addition to the iguanid Brachy- lophus (discussed by Cogger, 1974), Gorham ( 1965) stated that an additional 14 lizards oc- cur on Fiji. These are gekkonids and scincids, but only one possibly is endemic. The pres- ence of two boids, one elapid (endemic), and one typhlopid so far southeast tends to sup- port the hypothesis that these are all of Ori- ental source and not Gondwanan elements. As noted previously, the two endemic ranids on Fiji (Platy mantis vitiensis and P. vitianus) were derived from the same source route. In the case of the ranids, the absence of frogs from the intermediate potential stepping stones of New Hebrides and New Caledonia remains an apparently inexplicable anomaly. Additional references dealing with the Fijian herpetofauna are Barbour ( 1923 ) , Brown and Myers (1949), and Gorham (1968). Neic Zealand. — The most well known com- ponent of the New Zealand herpetofauna is the rhynchocephalian reptile, the Tuatara Sphenodon punctatus. This relic was prob- ably quite widely distributed elsewhere in the Mesozoic, as evidenced by the extensive fos- sil record of rhynchocephalians at that time. The remaining terrestrial reptiles are 35 spe- cies of lizards (almost all are endemic) of the families Gekkonidae and Scincidae. The geckos include three endemic genera that are unique among gekkonids in being ovovivipar- ous, whereas the endemic skinks are members of genera widely distributed outside New Zealand. In their recent survey of New Zealand vertebrates, Bull and Whittaker (1975) ac- cepted Kluge's (1967) interpretations of the origin of the gekkonids, resulting in their statement (op. cit: 239): "The geckos form part of the Malayo-Pacific element of the New Zealand fauna and probably entered New Zealand in the Miocene when the climate was warmer and the land more extensive than now." They further visualized fairly extensive oceanic dispersal by rafting from New Cale- 100 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 donia or directly from Australia. A totally dif- ferent interpretation would result by consider- ing the connection between New Zealand and the Lord Howe Rise and the exceptionally close proximity of the Lord Howe Rise to Australia in Early Cenozoic (Griffiths and Varne, 1972). If geckos really arrived in New Zealand no earlier than the Miocene, they have been evolving rapidly ever since along a unique path. Therefore, acceptance of an Early Ceno- zoic entry avoids any concept of an explosive radiation, and provides an adequate time span for speciation in situ along novel lines. Kluge's (1967) hypothesis of gekkonid evolution and dispersal did not accommodate continental drift. Accordingly, it is not surprising that his interpretations of the source of the faunal ancestors is at variance with one incorporat- ing this phenomenon. The endemic frog fauna of New Zealand is even more bizarre than the reptile fauna and is represented by three species of Leio- pelma. Whether Leiopelma and Ascaphus of North America should be placed in a single family ( Ascaphidae), or whether Leiopelma should constitute the Leiopelmatidae remains a matter of debate. However, there is no argument to the concept that these genera represent relics of a fauna that was widely distributed. Estes and Reig ( 1973 ) referred Vieraella and Notobatrachus of the Early and Late Jurassic of Patagonia to the Ascaphidae. Until recently, Leiopelma hamiltoni was known only from a small heap of stones oc- cupying one quarter of a hectare on the South Island. Now it is known from an additional fifteen hectares on the small Maud Island off the coast (Rull and Whitaker, 1975). Leio- pelma archeyi and L. hochstetteri are dis- tributed somewhat more widely on the North Island, but Bull and Whitaker (op. cit: 235) stated that in the recent past (". . . probably within the last 1,000 years. . . .") Leiopelma was far more widespread, being known from five sites where it no longer occurs today. They described subfossil material as being about twice the size of the extant species, but otherwise similar in skeletal features. Thus, the animals involved would have been as much as 100 mm in length. Either the mor- phological change to existing species was ef- fected in a millenium or the subfossils repre- sent extinct species; in either case it seems to be unprofitable to speculate about the char- acteristics of their immediate postdrift an- cestors. Within the context of discussion of south- west Pacific biological origins, the concepts of Nur and Ben-Avraham (1977) on a lost Pa- cific continent (formerly lying close to the east coast of Australia) must be considered. Nevertheless the present study of the anurans has not required such a landmass to explain their origins. CONCLUSIONS When the amphibian and reptile families now found in Australia are examined, one by one, to determine whether their affinities lie with South American or with Oriental stocks, it rapidly becomes apparent that Oriental sources predominate, and that Australian- South American links are few indeed. At the commencement of drifting in the Eocene, the Australian herpetofauna included the follow- ing families shared with South America. Gekkonidae (Diplodactylinae). — The na- ture of the extensive radiation within Aus- tralia may well support the concept that the Gekkonidae was the only lizard family pres- ent in Australia. Boidae (P Madtsoinae). — The presence of this family hinges upon the Pleistocene Wo- nambi whose phylogenetic affinities are with snakes outside the Oriental Region. Extant genera certainly arrived in the Miocene. Chelidae. — A Gondwanan component probably of considerable antiquity. Within Australia, fossils extending to the Early Ter- tiary represent modern species. The origin of the family is uncertain, but the Lower Cre- taceous Chehjcarapookus arcuatus, for which Warren (1975) erected the Chelycarapooki- dae, exhibits postcranial features that render it a potential chelid ancestor. Crocodijlidac. — The first fossils of a sebe- cosuchian crocodile fauna have just been dis- covered. Hylidae. — Previous concepts of South American and Australian tree frogs represent- ing a single family do not as yet appear to have been refuted. 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 101 Leptodactylidae. — Irrespective of the final assessment of the familial disposition of the Australian genera and species, the South American affinities of the Australian members are indisputable. These six families represent the elements of the Australian herpetofauna destined to persist through to the Holocene. It follows that by modern standards the total herpe- tofauna was incredibly depauperate, and lacked many significant elements. Pygopodids evolved in Australia probably during the period of isolation. The serious deficiencies were remedied from the Miocene onwards as Australia and the eastern outliers of the Oriental Region approached one another on their collision course. The commencement of colonization by families such as the Scincidae probably antedated the mid-Miocene. Others such as the Typhlopidae, Carettochelyidae, Microhy- lidae, Varanidae, and Crocodylidae probably were acquired at the time of the collision, while the Ranidae entered then and again in a subsequent wave. In broad terms, the Ori- ental influence upon Australia probably was more significant than the North American in- fluence upon South America, because South America had a more diverse and better estab- lished herpetofauna before the time of con- tact. ACKNOWLEDGMENTS For the invitation to participate in the symposium "The South American Herpeto- fauna," and for the funding making this pos- sible, I am greatly indebted to The University of Kansas, and particularly to the convenor, William E. Duellman. Many of the ideas and assessment pro- posed in this paper arose from or were stimu- lated by discussions or correspondence with many of my colleagues. For this great help I gratefully acknowledge the contributions of Bob Lange, John M. Legler, Brian McGow- ran, Allen E. Greer, Rowley Twidale and George R. Zug. Margaret Davies prepared figures 4:1 and 4:10, whereas figures 4:3-4:7 and 4:9 are the work of Debra Bennett. The manuscript was typed by Mrs. J. Russell-Price. I am also indebted to Richard G. Zweifel for permis- sion to reproduce the map of New Guinea upon which figure 4:5 was prepared. RESUMEN Las oportunidades para un intercambio herpetofaunistico gondwanico entre Sud America y Australia fueron muchos menores que aquellos entre Sud America y Africa, debido a la expansion de la Antartica. Al comienzo de la separation de Australia de Sud America (53 m.a.a.p.) el continental su- reno Australiano era subtropical. Condiciones humedas con una diversidad de anfibios y reptiles occupaban el centra de Australia hasta el Pleistoceno superior cuando un incremento de la aridez elimino muchos sitios acuaticos. De todos los posibles modos de establecer cuales segmentos de la herpetofauna australi- ana eran compartidas con Sud America como resultado de un intercambio gondwanico, se ha seleccionado determinar que familias que estan hoy presentes en Australia y Nueva Guinea, fueron adquiridas cuando Australia coludio con la Region Oriental en el Mioceno medio. Un analisis de tal evento demuestra que un grupo de islas adheridas a lo que es la actual costa norte de Nueva Guinea. Por ende, si estas islas estaban poblados con an- fibios y reptiles algunos, al menos, debieron haberse integrado a la heipetofauna australo- papua a traves de esta via. Se propone un modelo que involucra la dispersion de algunos generos a traves de las Islas Orientales desde las Filipinas a Fiji en el Mioceno (el genera de batracios ranidos Platy mantis). La colision de tierras arego este rana a la fauna de Nueva Guinea y los actuales extremos de diversidad al Oeste y al Este de Nueva Guinea reflejan la edad de especiacion. En contraste otros generos arri- baron desde el Oeste despues de la colision, resultando en una progresiva reduction del numero de especies desde el Oeste al Este. Entonces, los patrones de distribution geo- grafica de los elementos Orientales en Aus- tralia varian de acuerdo con la fecha de en- trada relativa a la colision de Australia. Sin embargo, el topico mas importante es las rela- 102 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 cion de los anfibios y reptiles de Australia que aquellos de la Region Oriental adyacente. Bajo esta criterio, las Ranidae, Microhylidae, Varanidae, ? Scincidae, Agamidae, Caretto- chelyidae, Trionychidae, Elapidae, Colubri- dae, Acrochordidae, Uropeltidae, Typhlopi- dae, y Boidae entraron a la region geografica australiana despues del Mioceno medio. Los Boidae son linicos, por tener un genuino com- ponente gondwanico (si Wonambi del Pleis- toceno se relaciona con Madtsoia), y otro Oriental. Correspondientemente, la hcrpetofauna gondwanica de Australia no poseia la gran mayoria de los elementos de esa fauna. Un buen ejemplo esta dado por la presencia de los Hylidae y Leptodactylidae ( y por la man- tencion del uso de estos nombres familiares para los animales modernos de Australia). Aparte de estos, existian solo los Cheliidae y los geckos diplodactylinos, de los cuals los Pygopodidae probablemente evolucionaron como los linicos representantes australianos. Una deriva pasiva permitio a los largartos iguanidos viajar largas distancias desde la costa oeste de Sud America a Fiji y a Tonga. A pasar de que la direccion de las corrientes transpacificas (Este a Oeste) favorece Is dis- persion desde Sud America a Australia mas que la direccion opuesta, los inmigrantes sud- americanos han alcanzado las islas del Pa- cifico Sur, pero aim no han filtrado has al- canzar las costas del continente australiano. LITERATURE CITED Archer, M., Wade, M. 1976. Results of the Ray E. Lemley Expeditions, Part 1. The Allingham For- mation and a new Pliocene vertebrate fauna from northern Queensland. Mem. Queensland Mus. 17(3): 379-397. Axelrod, D. I. 1960. The evolution of flowering plants, pp. 227-305 in Tax, S. (ed.). Evolution after Darwin, 1, Univ. Chicago Press, Chicago, 629 p. Axelrod, D. I. 1975. Plate tectonics and problems of angiospeim history. Mem. Mus. Natl. Hist. Nat., Paris Sec. C 88:72-86. Bagnara, J. F., Ferris, W. 1975. The presence of phyllomedusine melanosomes and pigments in Australian hylids. Copeia 1975(3) :592-595. Bagnara, J. T. 1976. Color change, pp. 1-52 in Lofts, B. (ed.). Physiology of the Amphibia. Vol. Ill, Academic Press, New York, 597 p. Bai, A. R. K. 1956. Analysis of chromosomes in two genera of Microhylidae (Amphibia: Anura). Proc. Natl. Inst. Sci. India, Part B 22:1-5. Barbour, T. 1923. The frogs of the Fiji Islands. Proc. Acad. Nat. Sci. Philadelphia 75:111-115. Beaufort, L. F. de. 1951. Zoogeography of the land and inland waters. Sidgwick and Jackson, London, 208 p. Becak, M. L., Demaro, L., Becak, W. 1970. Poly- ploidy and mechanisms of karyotypic diversifica- tion in Amphibia. Cytogenetics 9:225-238. Blake, A. J. D. 1973. Taxonomy and relationships of myobatrachine frogs ( Leptodactylidae ) : A numerical approach. Australian ]. Zool. 21:119- 149. Blommers, R. 1971. Karyotype de Anodontohyla montana Angel (Anura, Microhylidae) de massif de TAndringitra (Madagascar). Terre Mai. 10: 261-265. Blommers-Schlosser, R. M. A. 1976. Chromosomal analysis of twelve species of Microhylidae (Anu- ra) from Madagascar. Genetica 46:199-210. Bocart, J. P., Nelson, C. E. 1976. Evolutionary implications from karyotypic analysis of the fami- lies Microhylidae and Rhinophrynidae. Herpeto- logica 32:199-208. Bogart, J. P., Pyburn, W., Nelson, C. E. 1976. The karyotype of Otophnjne robusta (Anura: Micro- hylidae). Ibid. 32:208-210. Bowler, J. M. 1976. Aridity in Australia: Age, ori- gins and expression in aeolian landforms and sedi- ments: Earth-Sci. Rev. 12:279-310. Brown, W. C, Alcala, A. C. 1970. The zoogeog- raphy of the herpetofauna of the Philippine Islands, a fringing archipelago. Proc. California Acad. Sci. 38(6) : 10.5-130. Brown, W. C, Myers, G. S. 1949. A new frog of the genus Cornufer from the Solomon Islands, with notes on the endemic nature of the Fijian frog fauna. Amer. Mus. Novit. (1418): 1-10. Bull, P. C, Whitaker, A. H. 1975. The amphib- ians, reptiles, birds and mammals, pp. 231-276 in Kuschel, G. (ed.). Biogeography and ecology in New Zealand. W. Junk, The Hague, 689 p. Callen, R. A., Tedford, R. H. 1976. New late Cainozoic rock units and depositional environ- ments, Lake Frome area, South Australia. Trans. Roy. Soc. South Australia 100:125-168. Chapman, F. 1919. On a fossil tortoise in ironstone from Carapook, near Casterton. Proc. Roy. Soc. Victoria 32 (newser.) (1):11-13. Christophel, D. C, Blackburn, D. T. 1978. Terti- ary megafossil flora of Maslin Bay, South Australia: A preliminary report. Alcheringa 2:311-319. Cogger, H. G. 1970. First record of the pitted- shelled turtle Carcitocliehjs insculpta from Aus- tralia. Search 1:41. Cocger, H. G. 1974. Voyage of the banded iguana. Australian Nat. Hist. 18:144-149. Cogger, H. G. 1975. Reptiles and amphibians of Australia. Reed, Sydney, 584 p. Cole, C. J., Zweifel, R. G. 1971. Chromosomes of a New Guinean microhylid frog, Cophixalus ri- IHirius Zweifel. Herpetol. Rev. 3:15-16. 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 103 Coleman, P. J. 1975. The Solomons as a non-arc. Bull. Australian Soc. Explor. Geophys., 6(2-3): 60-61. Coleman, P. J., Packham, G. H. 1976. The Melane- sian borderlands and India-Pacific plates' bound- ary. Earth-Sci. Rev. 12:197-233. Corben, C. J., Ingram, G. J., Tyler, M. J. 1974. Gastric brooding: Unique form of parental care in an Australian frog. Science 186:946-947. Cracraft, J. 1973. Vertebrate evolution and bio- geography in the Old World tropics: Implications of continental drift and palaeoclimatology, pp. 373-393 in Tarling, D. H., Runcorn, S. K. (eds. ). Implications of continental drift to the earth sciences. Vol. 1. Academic Press, London, 622 p. Cracraft, J. 1974. Continental drift and vertebrate distribution. Ann. Rev. Ecol. Syst. 5:215-261. Cracraft, J. 1975. Mesozoic dispersal of terrestrial faunas around the southern end of the world. Mem. Mus. Natl. Hist. Nat. Paris, Ser A 88:29- 54. Darlington, P. J. 1957. Zoogeography: The geo- graphical distribution of animals. Wiley, New York, 675 p. Denham, D. 1975. Distribution of underthrust litho- spheric slabs and focal mechanisms — Papua New Guinea and Solomon Islands Region. Bull. Aus- tralian Soc. Explor. Geophys. 6(2-3) : 78-79. De Vis, C. W. 1894. The lesser chelonians of the Nototherian drifts. Proc. Roy. Soc. Queensland 10:123-127. De Vis, C. W. 1897. The extinct fresh-water turtles of Queensland. Ann. Queensland Mus. 3:3-7. Dixon, J. R., Anderson, S. C. 1973. A new species and genus of gecko (Sauria: Gekkonidae) from Iran and Iraq. Bull. South. California Acad. Sci. 72:155-160. Duellman, W. E. 1970. The hylid frogs of Middle America. Univ. Kansas Mus. Nat. Hist. Monogr. ( 1 ) : 1-753. Duellman, W. E. 1975. On the classification of frogs. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (42): 1-14. Emerson, S. B. 1976. A preliminary report on the superficial throat musculature of the Microhylidae and its possible role in tongue action. Copeia 1976 (3):546-551. Estes, R., Reig, O. A. 1973. The early fossil record of frogs: A review of the evidence, pp. 11-63 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Falvey, D. A., Taylor, L. W. H. 1974. Queens- land Plateau and Coral Sea Basin: Structure and time — stratigraphic patterns. Bull. Australian Soc. Explor. Geophys. 5(4) : 123-126. Fooden, J. 1972. Breakup of Pangaea and isolation of relict mammals in Australia, South America, and Madagascar. Science 175:894-898. Gaffney, E. S. 1975. A phylogeny and classification of the higher categories of turtles. Bull. Amer. Mus. Nat. Hist. 155:387-436. Gaffney, E. S. 1977. The side-necked turtle family Chelidae: A theory of relationships using shared derived characters. Amer. Mus. Novit. ( 2620 ) : 1-28. Claessner, M. F. 1942. The occurrence of the New Guinea turtle (Carettochchjs) in the Mio- cene of Papua. Rec. Australian Mus. 21:106-109. Glassford, D. K., Killigrew, L. P. 1976. Evidence for Quaternary westward extension of the Aus- tralian desert into south-western Australia. Search 7:394-396. Gorham, S. W. 1965. Fiji frogs (with synopses of the genera Cornufcr and Plat y mantis). Zool. Beitr. 11:381-435. Gorham, S. W. 1968. Fiji frogs. Life history data from field work. Ibid. 11:427-446. Gow, G. F. 1977. A new species of Pytlion from Arnhem Land. Aust. Zool. 19:133-139. Greer, A. E. 1970. A subfamilial classification of scincid lizards. Bull. Mus. Comp. Zool. Harvard Univ. 139:151-184. Griffiths, J. R., Barne, R. 1972. Evolution of the Tasman Sea, Macquarie Ridge and Alpine Fault. Nature (Phys. Sci.) 235:83-86. Guibe, J. 1948. Contribution a l'etude de l'appareil genital des tvphlopides (Ophidiens). Bull. Soc. Zool. France 73:224-228. Guttman, S. I. 1973. Biochemical techniques and problems in anuran evolution, pp. 183-203 in Vial, J. L. (ed. ). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Hecht, M. K. 1975. The morphology and relation- ships of the largest known terrestrial lizard, Mega- lania prisca Owen, from the Pleistocene of Aus- tralia. Proc. Roy. Soc. Victoria (new ser.) 87(2): 239-250. Hecht, M. K, Archer, M. 1977. Presence of xipho- dont crocodilians in the Tertiary and Pleistocene of Australia. Alcheringa 10:383-385. Heyer, W. R., Liem, D. S. 1976. Analysis of the intergeneric relationships of the Australian frog family Mvobatrachidae. Smithsonian Contri. Zool. 233:1-29. Hoffstetter, R. 1961. Nouveaux restes d'un ser- pent boide (Madtsoia madagascahcnsis nov. sp.) dans la Cretace superieur de Madagascar. Bull. Mus. Natl. Hist. Nat. Zool, Paris (2) 33:152-160. Inger, R. F. 1954. Systematics and zoogeography of Philippine amphibia. Fieldiana Zool. 33(4): 181- 531. Incer, R. F. 1966. The systematics and zoogeog- raphy of the amphibia of Borneo. Ibid. 52:1-402. Jardine, N., McKenzie, D. 1972. Continental drift and the dispersal and evolution of organisms. Nature 234(5332) : 20-24. Keast, A. 1971. Continental drift and the evolu- tion of the biota on southern continents. Q. Rev. Biol. 46:335-378. Keast, A. 1973. Contemporary biotas and the sep- aration sequence of the southern continents, pp. 309-343 in Tabling, D. H., Runcorn, S. K. ( eds. ) . Implications of continental drift to the earth sciences. Vol. 1. Academic Press, London, 622 p. 104 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 King, M. 1973. Chromosomes of two Australian lizards of the families Scincidae and Gekkonidae. Cytologia 38:20.5-210. Kinc, M. 1975. Chromosomal studies on Australian lizards. Ph.D. Dissert. Univ. Adelaide, 150 p. King, M. 1977. Chromosomal and morphometric variation in the Gekko Diplodactylus vittatus Gray. Australian J. Zool. 25:43-57. King, M., King, D. 1975. Chromosomal variation in the lizard genus Varanus (Reptilia). Australian J. Biol. Sci. 28:89-108. King, M., Rofe, R. 1976. Karyotypic variation in the Australian Cecko Phyllodactylus marmoratus (Gray) (Gekkonidae: Reptilia). Chromosoma 54:75-87. Kluge, A. G. 1967. Higher taxonomic categories of gekkonid lizards and their evolution. Bull. Amer. Mus. Nat. Hist. 135:1-60. Kluge, A. G. 1974. A taxonomic revision of the lizard family Pvgopodidae. Misc. Publ. Mus. Zool. Univ. Michigan (147): 1-221. Kluge, A. G. 1976. Phylogenetic relationships in the lizard family Pygopodidae: An evaluation of theory, methods and data. Ibid. (152): 1-72. Lance, R. T. 1976. Fossil epiphyllous "Germlings," their living equivalents and their palaeohabitat indicator value. Neues Jahrb. Geol. Palaeontol. Abh. 151:142-165. Latz, P. K. 1975. Notes on the relict palm Livistona mariae F. Muell. in central Australia. Trans. Roy. Soc. South Australia 99:189-195. Laurent, R. F. 1975. La distribution des amphib- iens et les translations continentales. Mem. Mus. Natl. Hist. Nat., Paris 88:176-191. Leon, P. E. 1970. Report of the chromosome num- bers of some Costa Rican anurans. Rev. Biol. Trop. 17:119-124. Liem, D. S. 1973. A new genus of frog of the family Leptodactylidae from SE. Queensland, Australia. Mem. Queensland Mus. 16:459-470. Longman, N. A. 1929. Palaeontological notes. Specimens from a well at Brigalow. Ibid. 9:247- 252. Lydekker, R. L. 1889. Catalogue of fossil amphib- ians and reptiles in the British Museum ( Natural History) Part 3. British Museum, London, 254 p. Lynch, J. D. 1971. Evolutionary relationships, oste- ology and zoogeography of leptodactyloid frogs. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): 1-238. Lynch, J. D. 1973. The transition from archaic to advanced frogs, pp. 133-182 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contem- porary research on major problems. Univ. Mis- souri Press, Columbia, 470 p. Mabbutt, J. A. 1967. Denudation chronology in Central Australia. Structure, climate and land- form inheritance in the Alice Springs area, pp. 144-181 in Jennings, J. N., Mabbutt, J. A. (eds.). Landform studies from Australia and New Guinea. Australian Natl. Univ. Press, Can- berra, 434 p. MacKenzie, D. E. 1975. Volcanic and plate tectonic evolution of central Papua and New Cuinea. Bull. Australian Soc. Explor. Geophys. 6(2-3) :65-67. Maderson, P. F. A. 1972. The structure and evolu- tion of holocrine glands in sphaerodaetyline and eublepharine gekkonid lizards. Copeia 1973 (3): 559-571. Main, A. R., Lee, A. K, Littlejohn, M. J. 1958. Evolution in three genera of Australian frogs. Evolution 12:224-233. Maxson, L. R. 1976. The phylogenetic status of phyllomedusine frogs ( Hylidae ) as evidenced from immunological studies of their serum albu- mins. Experientia 32:1149-1150. Maxson, L. R., Wilson, A. C. 1975. Albumin evo- lution and organismal evolution in tree frogs (Hylidae). Syst. Zool. 24:1-15. McDowell, S. B. 1967. Aspidomorphus, a genus of New Guinea snakes of the family Elapidae, with notes on related genera. J. Zool. 151:497-543. McDowell, S. B. 1970. On the status and relation- ships of the Solomon Island elapid snakes. Ibid. 161:145-190. McDowell, S. B. 1972. The species of Stegoiwtus (Serpentes, Colubridae) in Papua New Guinea. Zool. Meded. Rijks Mus. Nat. Hist. Leiden 47: 6-26. McDowell, S. B. 1974. A catalogue of the snakes of New Guinea and the Solomons, with special reference to those in the Bernice P. Bishop Mu- seum. Part 1, Scolecophidia. J. Herpetol. 8:1-57. McDowell, S. B. 1975. A catalogue of the snakes of New Guinea and the Solomons, with special reference to those in the Bernice P. Bishop Mu- seum. Part II, Anilioidea and Pvthoninae. Ibid. 9:1-79. McDowell, S. B., Bocert, C. M. 1954. The sys- tematic position of Lanthanotus and the affinities of the anguimorphan lizards. Bull. Amer. Mus. Nat. Hist. 105:1-142. McGowran, B. 1973. Rifting and drift of Australia and the migration of mammals. Science 180: 759-761. Menzies, J. I., Tippett, J. 1976. Chromosome num- bers of Papuan hylid frogs and the karyotype of Litoria infrafrenata (Amphibia, Anura, Hvlidae). J. Herpetol. 10:167-173. Menzies, J. I., Tyler, M. J. 1977. The systematica and adaptations of some Papuan microhylid frogs which live underground. J. Zool. 183:431-464. Mittleman, M. B. 1952. A generic synopsis of the lizards of the subfamily lygosominae. Smithson- ian Misc. Collect. 117:1-35. Moffatt, L. 1973. The concept of primitiveness and its bearing on the phylogenetic classification of the Gekkota. Proc. Linn. Soc. New South Wales 97:275-301. Molnar, R. E. 1977. Crocodile with laterally com- pressed snout: First find in Australia. Science 197:62-64. Molnar, R. E. 1978. Age of the Chillagoe crocodile. Search 9(4) : 156-158. 1979 TYLER: SOUTH AMERICA AND AUSTRALIA 105 Morescalchi, A. 1968a. Initial cytotaxonomic data on certain families of amphibious Anura ( Diplas- iocoela, after Noble, T. C. ). Experientia 24:280- 283. Morescalchi, A. 1968b. Hypothesis on the phytog- eny of the Salientia, based on karyological data. Ibid. 24:964-966. Morescalchi, A. 1973. Amphibia, pp. 233-248 in Chiarelli, A. B., Capanna, E. (eds.). Cytotax- onomy and vertebrate evolution. Academic Press, London, 783 p. Morescalchi, A., Ingram, G. J. 1974. New chro- mosome numbers in Australian Leptodactylidae (Amphibia, Salientia). Experientia 30:1134— 1135. Natarjan, R. 1953. A note on the chromosomes of Cacopus systoma. Proc. 40th Indian Sci. Congress, Lucknow pp. 180-181. Nelson', C. E. 1975. Another new miniature 4-toed South American microhylid frog (Genus: Syn- cope). J. Herpetol. 9:81-84. Newsome, A. E., Rochow, K. A. 1964. Vertebrate fossils from Tertiary sediments in central Aus- tralia. Australian J. Sci. 26:252. Nur, A., Ben-Avraham, Z. 1977. Lost Pacifica con- tinent. Nature 270:41-45. Parker, H. W. 1929. Two fossil frogs from the Lower Miocene of Europe. Ann. Mag. Nat. Hist. (10) 4 (21):270-281. Parker, H. W. 1934. A monograph of the frogs of the family Microhylidae. British Museum (Na- tural History), London, 208 p. Parker, H. W. 1940. The Australasian frogs of the family Leptodactylidae. Novit. Zool. 42:1-106. Pianka, E. R. 1972. Zoogeography and speciation of Australian desert lizards: An ecological perspec- tive. Copeia 1972 (1):127-144. Rabello, M. N. 1970. Chromosomal studies in Bra- zilian anurans. Caryologia 23:45-59. Rhodin, A. G. J., Mittermeier, R. A. 1977. Neural bones in chelid turtles from Australia and New Guinea. Copeia 1977 (2):370-372. Rich, P. V. 1975. Antarctic dispersal routes, wan- dering continents, and the origin of Australia's non-passeriform avifauna. Mem. Natl. Mus. Vic- toria (36):63-125. Robb, J. 1966. The structure and possible function of the cloacal pouches of male Australian typhlo- pids. Australian J. Zool. 14:27-30. Robinson, R. L., Tyler, M. J. 1972. The catecho- lamine content of the adrenal glands of frogs as an index of phylogenetic relationships. Comp. Gen. Pharmacol. 3:167-170. Russell, A. P. 1976. Some comments concerning interrelationships amongst gekkonine geckos, pp. 217-244 in Bellaxrs, A. d'A., Cox, B. (eds.). Morphology and biology of reptiles. Linn. Soc. Symp. (Ser. 3), 290 p. Savage, J. M. 1973. The geographic distribution of frogs: Patterns and predictions, pp. 351-445 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Schllnger, E. I. 1974. Continental drift, Notho- fagus, and some ecologically associated insects. Ann. Rev. Entomol. 19:323-343. Schodde, R., Mason, I., Wolfe, T. O. 1972. Fur- ther records of the pitted-shelled turtle (Caretto- chclys insculpta) from Australia. Trans. Roy. Soc. South Australia 96:115-117. Sclater, J. G., Von Der Borch, C, Veevers, J. J., Hekinian, R., Thompson, R. W., Pimm, A. C., McGowran, B., Gartner, S., Johnson, D. A. 1974. Regional synthesis of the deep sea drilling results from leg 22 in the eastern Indian Ocean, pp. 815-831 in Von Der Borch, C, Sclater, J. G., Veevers, J. J. Initial reports of the Deep Sea Drilling Project 22, 890 p. Smith, M. J. 1975. The vertebrae of four Australian elapid snakes. Trans. Roy. Soc. South Australia 99:71-84. Smith, M. J. 1976. Small fossil vertebrates from Victoria Cave, Naracoorte, South Australia. IV Reptiles. Ibid. 100:39-51. Stephenson, E. M., Stephenson, N. G. 1970. Kar- yotypes of two Australian hylid frogs. Chromo- soma 30:38-50. Stirton, R. A., Tedford, R. H., Miller, A. H. 1961. Cenozoic stratigraphy and vertebrate paleontology of the Tirari Desert, South Australia. Geogr. Rev. 62:40-70. Storr, G. M. 1964. Some aspects of the geography of Australian reptiles. Senckenbergiana Biol. 45: 577-589. Strauchan, I. R. 1969. Hyla inermis (Peters), a species hitherto erroneously referred to the lepto- dactylid genus Cyclorana (Anura, Hylidae/Lep- todactylidae). Zool. Meded. Rijks Mus. Nat. Hist. Leiden 43:207-212. Straughan, I. R., Lee, A. K. 1966. A new genus and species of leptodactylid frog from Queens- land. Proc. Roy. Soc. Queensland 77:6.3-66. Taylor, L. W. H. 1975. Deposition and tectonic- patterns in the western Coral Sea. Bull. Australian Soc. Explor. Geophys. 6:33-35. Tilbury, L. A. 1975. Lineations in the Bismarck Sea. Ibid. 6:72. Twidale, C. R. 1972. Evolution of sand dunes in the Simpson Desert, Central Australia. Trans. Instit. British Geogr. (56):77-109. Tyler, M. J. 1967. Microhylid frogs of New Britain. Trans. Roy. Soc. South Australia 91:187-190. Tyler, M. J. 1970. Patterns of distribution and the origins of the Papuan hylid frog fauna. Search 1:246-247. Tyler, M. J. 1971. The phylogenetic significance of vocal sac structure in hylid frogs. Univ. Kansas Mus. Nat. Hist. Occas. Pap. ( 19) :319-360. Tyler, M. ]. 1972a. Superficial mandibular muscu- lature, vocal sacs and the phylogeny of Australo- Papuan leptodactylid frogs. Rec. South Australian Mus. 16:1-20. Tyler, M. J. 1972b. A new genus for the Australian leptodactylid frog Crinia darlingtoni. Zool. Me- ded. Rijks. Mus. Nat. Hist. Leiden 47:193-201. 106 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Tyler, M. J. 1974a. Superficial mandibular mus- culature and vocal sac structure in the Anura. M.Sc. Thesis, Univ. Adelaide, 134 p. Tyler, M. J. 1974b. First frog fossils from Aus- tralia. Nature 248 (5450) :711-712. Tyler, M. J. 1974c. The systematic position and geographic distribution of the Australian frog Chirolcptes alboguttatus Gunther. Proc. Roy. Soc. Queensland 85:27-32. Tyler, M. J. 1976a. Frogs. William Collins, Syd- ney, 236 p. Tyler, M. J. 1976b. Comparative osteology of the pelvic girdle of Australian frogs and description of a new fossil genus. Trans. Roy. Soc. South Australia 101:3-14. Tyler, M. J. 1976c. A new genus and two new spe- cies of leptodactylid frogs from Western Aus- tralia. Rec. Western Australian Mus. 4:45-52. Tyler, M. J. 1977. Pleistocene frogs from caves at Naracoorte, South Australia. Trans. Roy. Soc. South Australia 101:85-89. Tyler, M. J. 1979. The introduction and current distribution in the New Hebrides of the Aus- tralian hylid frog Litoria aurea. Copeia. (In press ) Tyler, M. J., Davies, M. 1978a. Species groups within the Australopapuan hylid frog genus Li- toria Tschudi. Australian J. Zool. Suppl. 63:1—47. Tyler, M. J., Davies, M. 1978b. Phylogenetic rela- tionships of Australian hyline and Neotropical phyllomedusine frogs of the family Hylidae. Her- petologica 34:219-224. Tyler, M. J., Davies, M., King, M. 1978. The Aus- tralian frog Chiroleptes dahlii Boulenger: Its sys- tematic position, morphology, chromosomes and distribution. Trans. Roy. Soc. South Australia 102:17-23. Tyler, M. J., Martin, A. A. 1975. Australian lepto- dactylid frogs of the Cyclorana australis complex. Ibid. 99:93-99. Tyler, M. J., Martin, A. A. 1977. Taxonomic studies of some Australian leptodactylid frogs of the genus Cyclorana Steindacfmer. Rec. South Australian Mus. 17:261-276. Tyler, M. J., Rorerts, J. D. 1973. Noteworthy range extensions for some South Australian frogs. South Australian Nat. 99:143-147. Tyler, M. J., Martin, A. A., Davies, M. 1979. Biol- ogy and systematics of a new limnodynastine genus (Anura: Leptodactylidae) from northwestern Australia. Australian J. Zool. 27 ( 1 ) : 135-150. Underwood, G. L. 1954. On the classification and evolution of geckos. Proc. Zool. Soc. London 124: 469-492. Underwood, G. L. 1957. On lizards of the family Pygopodidae. A contribution to the morphology and phvlogeny of the Squamata. J. Morphol. 100:207-268. Veevers, J. J., McElhinny, M. W. 1976. The sep- aration of Australia from other continents. Earth- Sci. Rev. 12:139-159. Walker, C. F. 1973. A new genus and species of microhylid frog from Ecuador. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (20): 1-7. Walker, C. F., Duellman, W. E. 1974. Descrip- tion of a new species of microhylid frog, Cliiasmo- cZeis, from Ecuador. Ibid. (26): 1-6. Wallace, D. C, Maxson, L. R., Wilson, A. C. 1971. Albumin evolution in frogs: A test of the evolutionary clock hypothesis. Proc. Natl. Acad. Sci. USA 68:3127-3129. Warren, J. W. 1969. Chelid turtles from the mid- tertiary of Tasmania. J. Palentol. 43:179-182. Warren, J. W. 1975. A fossil chelonian of probably Lower Cretaceous age from Victoria, Australia. Mem. Natl. Mus. Victoria 29:23-28. Watson, G. F., Martin, A. A. 1973. Life history, larval morphology and relationships of Australian leptodactylid frogs. Trans. Roy. Soc. South Aus- tralia 97:33-45. Wood, R., Patterson, B. 1973. A fossil trionychid turtle from South America. Breviora 415:1-10. Woodruff, D. S. 1972. Australian anuran chromo- some numbers. Herpetol. Rev. 4:208. Wopfner, H., Twidale, C. R. 1967. Geomorpho- logical history of the Lake Eyre Basin, pp. 119- 143 in Jennings, J. N., Mabrutt, J. A. (eds.). Landform studies from Australia and New Guinea. Australian National Univ. Press, Canberra, 434 p. Zweifel, R. G. 1969. Frogs of the genus Platyman- tis ( Ranidae ) in New Guinea, with the descrip- tion of a new species. Amer. Mus. Novit. (2374): 1-19. Zweifel, R. G. 1971. Results of the Archbold Ex- peditions. No. 96. Relationships and distribution of Genyophryne tliomsoni, a microhylid frog of New Guinea. Ibid. (2469): 1-13. Zweifel, R. G. 1972. Results of the Archbold Ex- peditions. No. 97. A revision of the frogs of the subfamily Asterophryinae Familv Microhvlidae. Bull. Amer. Mus. Nat. Hist. 148:411-546. 5. Quaternary Biogeography of Tropical Lowland South America Jiirgen Haffer Tommesweg # 60 4300 Essen-1 West Germany Research into the Quaternary biogeog- raphy of the Neotropical Region has been intensified during recent years as biologists became increasingly aware of the fact that Pleistocene climatic-vegetational fluctuations caused vast changes in the distribution of forest and nonforest biotas. Comparatively restricted populations of previously widely distributed plants and animals were isolated in remnant habitats during adverse climatic periods and differentiated at a varying rate depending upon the size of the restricted population (i.e., the size of the "refuge" area), the degree of isolation, and the varying "plasticity" of systematic groups following the model of geographic speciation (Mayr, 1942, 1963). The interpretation of Quaternary for- est and savanna fragmentation provides biol- ogists with a mechanism to explain extensive recent speciation in the South American low- lands, the occurrence of widely disjunct popu- lations of related taxa and other biogeo- graphical phenomena that could not be ac- counted for in the absence of natural barriers to interbreeding and dispersal (Meggers, 1977). The studies recently completed in the fields of Neotropical ornithology, herpetology, entomology, and botany yield comparable re- sults and, hopefully, will stimulate further investigations needed to test the model of Quaternary differentiation proposed. Prob- ably only few Neotropical plants and animals have survived from the Late Tertiary until the present time without evolutionary change — a notion popular among biologists only one or two decades ago. It is held that speciation events leading to extensive differentiation of faunas and floras during the Tertiary con- tinued into the Quaternary, possibly at a somewhat accelerated pace because of rapid environmental changes. In this review I characterize, in a brief introductory section, the climate and vegeta- tion of tropical lowland South America and their Quaternary history as far as it is known today. The main portion of this paper is a detailed discussion of recent research into the Quaternary biogeography of various groups of South American animals and plants. The re- sults of these studies concern biologists in- terested in the historical aspects of tropical biotas and should prove useful for compara- tive purposes to students of the Neotropical herpetofauna in particular. CLIMATE AND VEGETATION The climate of the lowlands of tropical South America varies from wet, humid, or moist, especially near mountain ranges and in the equatorial Amazon region, to dry or even arid in northeastern Brasil and along the Car- ibbean coast of northern Venezuela and Co- lombia. The northern tradewind belt moves southward into northern South America dur- ing the northern winter causing a pronounced dry season. Alternating wet and dry seasons occur over most of tropical South America, even to some extent in portions of the lower Amazon Valley. The dry seasons are least pronounced and the annual rainfall corre- spondingly high in the upper Amazon Valley, near the Atlantic coast of northeastern and southeastern tropical South America as well as in the vicinity of the Andes, especially in the Pacific Choco region of western Colombia, and along the Caribbean slope of the Middle American mountains. Tall evergreen forests grow in areas of high rainfall (Fig. 5:1) and grade through semi-evergreen and deciduous forests into 107 108 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 thorn forest and scrub as the annual rainfall decreases and the influence of a prolonged dry season increases. Variations are caused by local conditions of soil and topography. Characteristic plant associations of the non- forest regions of interior Brasil are the cer- rado, a typical woodland savanna, and the caatinga, an open thorn woodland rich in cacti. Extensive grass savannas are found in the north and south of the Amazon forest in areas where flooding lasts several months each year and alternates with a severe dry season (eastern Colombian and central Venezuelan llanos; the savannas of eastern Bolivia and the varzea campos of the lower Amazon Val- ley). Extensive discussions of the climate and vegetation of South America have been pub- lished by Schwerdtfeger (1976), Hueck (1966), and Hueck and Seibert (1972); see also the recent literature review by Haffer (1974). The vast Amazonian forest covers some 6,000,000 km2 of central South American low- lands from the Andes to the Atlantic coast including the upper Orinoco region of south- ern Venezuela and the Guianas to the north- east of the Solimoes-Amazon basin itself. Riv- ers bordered by characteristic vegetation zones of varying width and isolated savanna en- claves interrupt the immense and superficially uniform forests. Interspersed savannas are concentrated in a transverse zone of reduced annual precipitation extending from southern Venezuela across the lower Amazon River into northeastern Brasil; others occur between the upper Madeira and Purus rivers (Fig. 5:1). Extensive forests in Amazonia grow on infer- tile leached soils of the terra firme in areas where Tertiary strata and "basement" rocks of the Guianan and Brasilian shields form the subsoil. Only rather small areas of Amazonia are underlain by fertile soil, especially along major river valleys and in the Andean fore- land. Erosive material from the Andes is transported eastward into the Amazonian low- lands and forms soils that are considerably richer in nutrients than the soils of the adja- cent terra firme between large river courses (Fittkau, 1969, 1974). This simplified scheme is currently being modified through detailed interpretation of radar images, field controls, and extensive mapping in Amazonia (Ham- mond, 1977). CLIMATIC-VEGETATIONAL HISTORY OF THE NEOTROPICAL LOWLAND REGION DURING THE QUATERNARY Humid tropical vegetation, perhaps some- what drier in midlatitudes, covered most of the exposed land area of South America dur- ing early Tertiary time (Wolfe, 1971; Solbrig, 1976). Forests slowly retreated northward in the Patagonian region during the second half of the Tertiary when the Andes were gradu- ally uplifted and the climate became cooler and drier, possibly in response to periodic polar glaciations which began during the Miocene. The glacial phases gained momen- tum with time until, during the Quaternary, vast polar glaciers repeatedly advanced to- ward lower latitudes and extensive montane glaciers covered the higher slopes of tropical mountains. Extensive continental shelf areas were emergent during the glacial periods of lowered world sea level and were submerged during interglacial phases. Interglacial seas even encroached over low lying coastal plains, such as northern Colombia, and covered a huge portion of the Amazon Valley. Although temperatures in the tropical lowlands remained "tropical" during glacial periods («3°C lower than today), alternating humid and arid climatic phases of the Quater- nary caused vast changes in the distribution of forest and nonforest vegetation. Forests broke into isolated remnants during cool dry periods (glacial phases) and expanded and coalesced during warm, humid phases (inter- glacial periods). Conversely, nonforest vege- tation expanded during glacials and retreated during interglacial phases. Geoscience data are insufficient so far to map the changing distribution of forest and nonforest vegetation during the various climatic periods and, in particular, to locate accurately areas of rem- nant forests during arid phases which served as "refugia" for animal populations. From the location of current rainfall maxima and the topographic relief (which was already in existence during most of the Pleistocene) one would tentatively conclude that several areas along the northern slopes and foreland of the mountains in the interior Guianas, along the eastern slopes and foreland of the Andes, as well as along the northern margin of the Brasilian tableland remained humid and for- 1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 109 Fig. 5:1. Distribution of humid tropical lowland forest and location of rainfall centers in Middle and South America. Explanations: Shaded = humid forest, often semideciduous around savanna regions. Hatched vertically = areas receiving over 2500 mm of rain per year. Solid = Andean Cordilleras and Middle American mountains of more than 2000 m elevation. Heavy dashed lines delimit the dry transverse zone of lower Ama- zonia characterized by numerous isolated savanna enclaves. Letters designate areas of paleoecological research. (See text for details.) Distribution de selva humeda y position de centros de Uuvia en las tierras bajas tropicales de Centro y Sud America. Explicaciones: Matizado = selva humeda; frecuentemente selva semidecidua alrededor de regiones de savanas. Rayado vertical = areas recibiendo mas que 2500 mm de Uuvia anuales. Negro = Cordilleras andinas y montafias de Centro America con alturas mayores que 2000 m. Lineas rayadas anchas delimitan la zona seca transversal de Amazonia baja caracterizada por numerosas savanas aisladas. Las letras indican las areas de investigaciones paleoecologicas. (Ver el texto para detalles.) 110 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Copoera + Katiro (Rondonia. Brasil) I0 2O3O4O5O6O7O809O KDO% Loguxi de Agua Suoa (Llanos Onerrfoles .Gotomtoo ) Loke Moreru iRupsKn. Guyana) 21601 23401 ■ i^ Eem of tie wt tropical foresf .'. , [ Maunna I 1 Savarmo elem (pnnc Gtamineae) 50 100%^° 38301 4IK)| 0-H00 j Oftier forest elements Gramineae (savannas) □ B G/amneae , Cyperoceoe orel crtTier 5 / / • x y v 4. ^ 400 KM Fie. 5:5. Reconstruction of Quaternary refugia in tat (dashed outline). Schematic representation. Explan parapatric species and hybridizing subspecies. Contact suture zones. 2. Mapping of distribution centers for ( contours indicate numbers of sympatric species and of location of suture zones and distribution centers. 4. ing into consideration all available data on relief, dim under consideration. Reconstruction de refugios cualernarios en medio de savana; delimitada por un margen raijado). Explication especies parapdtricas ij subespecies con hibridacion. Zo ureas formando asi zonas de sutura. 2. Mupeo de cen subespecies localizadas (Hncas de contorno indican el grupos de organismos bajo consideration). 3. Compara de distribution. 4. Mapeo de refugios rclacionados a datos disponibles sobre el relieve, clima, geomorfologia, an extensive and fairly uniform forest or nonforest habi- ations: 1. Mapping of secondary contact zones of allied zones often cluster in certain regions forming faunal med by fairly localized species and subspecies clusters subspecies in the groups considered). 3. Comparison Mapping of refugia related to the dispersal centers tak- ate, geomorphology and palynology related to the areas tin extenso ambiente y mas o menos uniforme (sclva o es: 1. Mapeo de zonas de contacto secundaria entre nav de contacto frccucntcmente se encuentran en ciertas tros de distribution formadas por grupos de especies y niimero de especies y subespecies simpdtricas en los cion de la position de zonas de sutura y de los centros los centros de dispersion teniendo en cucnta todos los y palinologia de la region. 120 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 contact zones and distribution centers. — Often clusters of secondary contact zones fall be- tween core areas, thereby supporting the in- terpretation that the distribution centers func- tioned as centers of dispersal in the past. This interpretation is particularly applicable in the case of hybridizing subspecies and parapatric species that characterize neighboring centers and meet in the intervening area where other contact zones of more wide ranging forms are clustered as well. Recause the contact zones and centers often involve different species, full complementarity between them cannot be expected. Thus, there are many contact zones between forms whose present ranges comprise more than one distribution center. This is schematically indicated in the diagram (Fig. 5:5, no. 3), where several forms occupying the entire eastern portion of the habitat with a northern and a southern center meet western forms along contact zones that are not re- stricted to the area between the centers. If biogeographical analyses are based ex- clusively on the population structure of sev- eral widespread and geographically variable species, a patchwork of subspecies ranges sep- arated by more or less extensive hybrid zones exists. Superimposition of the various species maps may show that the ranges of pure sub- specific populations with uniform character expression and the location of separating hy- brid belts more or less coincide in the differ- ent species (Fig. 5:6). Contour lines illustrate varying hybrid levels of populations around the centers. Analyses also may be based on the population structure of a single species, and distribution maps of several individual characters may be prepared (e.g., Vanzolini and Williams, 1970, for the lizard Anolis chrij- solepis). The character maps may then be superimposed and contoured in a similar manner as those of subspecies. Coinciding areas of uniform characters with low variabil- ity often cluster in core areas that are sep- arated by zones of high character variability coinciding with hybrid belts between sub- species ranges. Step 4: Mapping of the refngia related to the dispersal centers. — Ideally, this can be accomplished using geoscience data exclu- sively. Having established on the basis of zoogeographical data that a given distribution 300 KM Fig. 5:6. Schematic representation of centers of subspecies endemism (stippled) in an extensive and fairly uniform habitat (dashed outline). Explana- tions: Superimposed ranges of "pure" subspecies populations and separating hybrid belts in different species often cluster in certain areas. Number of dif- ferentiated forms of various species superimposed is indicated (n = 11 or 12). Average hybrid level of combined species populations is mapped by contour lines ( 0 = "pure" ) . A barrier to gene-flow ( broad river or mountain range) is schematically shown as a black bar; the populations on either side of the barrier are mostly pure, as limited gene-flow takes place between subspecies of only a few species whose ranges have been superimposed. Diagrama esquemdtico de centros de endemismo subespecifico (puntcado) dentro de tin habitat rela- tivamente extenso y uniforme (delimitado por una linea entrecortada). Explicaciones: Rangos supcr- impuestos de poblaciones de subespecies "puras" y los cinturones de hibridizacion en varias especies muchas veces se juntan en determinadas rcgiones. El numero de jormas difcrenciadas de las especies superpuestas estd indicado (n = 11 6 12). El nivel promedio de hibridizacion en las poblaciones de especies combinadas estd mapcado por tineas de con- lomo (0 = "puro"). Una barrera para el flujo genetico (un gran rio 6 una montana) estd indicada esquematicamente por una faja ncgra. ha matjoria de las poblaciones en ambos lados de la barrera son "puras"; un flujo genetico restringuido ocurrc sola- mente entre unas pocas poblaciones separadas. center probably was a center of dispersal in the past, we proceed to postulate the approxi- mate location, size and shape of the corre- sponding refuge area of forest or nonforest vegetation within the central portion of the center, taking into consideration all available geoscience data from the region, however de- ficient these data may be. Medium elevation areas with high rainfall near the base of mountains or plateau regions are prime candi- dates for forest refugia whose postulated ex- tent during the maximum of the arid climatic 1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 121 period remains highly speculative. The model assumes that the species and subspecies char- acterizing a given center were confined to the postulated refuge prior to dispersal and prior to establishing secondary contact with other forms spreading from distant centers. Obvi- ously, the former existence and changing size of the refugia ultimately can be traced only through detailed palynological, pedological, and geomorphological studies rather than through zoogeographical analyses. After all, the refugia are geological-paleoclimatological, not biological, phenomena. However, as long as only scattered geoscience data are avail- able, zoogeographical analyses as outlined above will help our understanding of biotic differentiation in lowland tropical South America during the Quaternary. In general, the locations of distribution centers of the Amazonian forest biota correlate well with the locations of forest refugia tentatively de- rived from rainfall, relief and geoscience data alone (p. 10S), thus strengthening an histor- ical interpretation of the biogeographical core areas (Simpson and Haffer, 1978; Brown and Ab'Saber, 1979). Endler (1977) suggested that parapatric speciation occurs frequently in nature and that the hybrid zones in Amazonia might ac- tually be zones of primary intergradation caused by strong environmental gradients. However, the geoscience data reviewed above (but not discussed by Endler) favor the inter- pretation of allopatric rather than parapatric speciation for the many hybridizing and non- hybridizing populations of plants and animals in contact. The Amazonian forest refugia are geological-paleoclimatical, not biological phenomena. Biogeographical Studies on the Forest Fauna and Flora Some of the major regional conclusions re- garding the Quaternary differentiation and dispersal of Neotropical biotas are summar- ized, as follows: 1) Extensive speciation took place in many groups of Neotropical animals and plants that were repeatedly isolated in refugia and later expanded their ranges dur- ing favorable expansive phases. 2) During humid climatic periods, extensive forests probably permitted a direct connection of the Amazonian fauna and flora with those of the Atlantic forest in eastern Brasil across the central Brasilian Plateau and along the coastal region of northeastern Brasil, where scattered forests are still preserved on isolated moun- tains (Muller, 1973; Haffer, 1974; Brown, 1976, 1977b). 3) The cis- Andean and trans- Andean forest biotas probably were connected repeatedly both north of the Andes in the Caribbean lowlands of northern Colombia and through the north Peruvian Andes from the upper Maranon Valley via the low Por- culla Pass so as to reach the forested Pacific lowlands of Colombia from the south (Chap- man, 1926; Haffer, 1967, 1975; Muller, 1973). 4) The relations between the Middle and South American nonforest bird faunas are less pronounced than those of the forest faunas of these two areas (Mayr, 1964). 5) There are numerous conspecific popula- tions inhabiting the nonforest regions of northern South America and central Brasil, respectively, being separated by the entire width of the Amazon forest. Because their dispersal under present climatic-vegetational conditions is highly unlikely, a rather recent direct communication of nonforest faunas across Amazonia during one or more arid cli- matic periods seems probable. This theory also explains the close relationship of the fauna and flora of the Amazonian savanna enclaves with those of the nonforest regions to the north and south of Amazonia ( Hueck, 1966; Haffer, 1969; Muller, 1973). The authors of the studies reviewed below are aware of the tentative nature of the results obtained and of the suggestions made. More- over, the nascent theory of ecological refugia during the Quaternary, like any theory, cannot be proven but only disproven by the results of additional studies on the paleo- climatology, vegetational history, and zooge- ography of South America. I wish to empha- size with Meggers (1977) that the efforts by biologists and archeologists at present are not more than a search for correlations and pat- terns useful as guides for investigation. The strikingly similar results as to the basic pat- terns of differentiation in various unrelated groups of Neotropical animals tend to sup- port the refuge theory and justify continued 122 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 research into this field of enquiry. Biogeog- raphers working on the Neotropical fauna proposed for the lowlands of Middle and South America a total of 40 areas that are assumed to have served as refugia for the forest fauna at various times during the Qua- ternary. These areas are listed and briefly described in Appendix 5:1. Birds. — Numerous secondary contact zones of avian species and subspecies pairs are clustered in north-central Amazonia — in southern Venezuela and in the Rio Negro region of northern Brasil (Fig. 5:7), where Guianan forms from the east established con- tact with western forms that had spread from upper Amazonia (Haffer, 1969, 1974). These contact zones represent range limits either of hybridizing subspecies or of non-hybridizing competing species; the ranges were mapped in conjunction with studies of patterns of geographic variation in several groups of Amazon forest birds. South of the Rio Ama- zonas, contact zones are scattered over a more extensive area, where we may distin- guish an upper Amazonian and a south-cen- tral Amazonian suture zone. A distributional analysis of the Amazonian forest avifauna (Haffer, 1978) indicated the existence of six core areas of distribution (Fig. 5:7) or local- ized species clusters, each of them composed of 10 to 50 species. The six clusters together are characterized by a total of around 150 species or about 25 percent of the Amazon forest bird fauna. Most of the remaining forest birds (75%) are more widely distrib- uted, their ranges comprising two or more distribution centers. There are conspicuous clusters of lower and upper Amazon forest birds, as well as smaller groups of northern and southern Amazon forest birds, in addition to those species that inhabit even larger areas of Amazonia and beyond (see Haffer, 1978, for more details). A number of other species have an irregular, spotty distribution or are known from only a single locality. The mapped contact zones in Amazonia (Fig. 5:7) are mostly located between distributional core areas which probably functioned as centers of dispersal. The complementarity of contact zones and distribution centers is less well de- veloped in central Amazonia, where a num- ber of birds spreading from the species-rich Fig. 5:7. Location of secondary contact zones (above) and of distribution centers in the Amazon forest avifauna (below). Explanations: Secondary contact zones (dashed) are concentrated in north- central Amazonia (a), upper Amazonia (b), and south-central Amazonia (c). A continuous line de- limits the Amazon forest (above). Fairly localized species clusters (each composed of 10 to 50 species) form the distribution centers A to F (below). Adapted from Haffer, 1974, Fig. 9.13 and Haffer, 1978. Additional clusters of endemic species char- acterize the Atlantic forests of southeastern Brasil and the trans-Andean forests of northwestern South America. Position de zonas de contacto secundario (arriba) y de centras de distribution en la avifauna de la sclva amazonica (abajo). Explicaciones: Zonas de contacto secundario (ttneas rayadas) se encucntran en la Ama- zonia norcentral (a), en la alta Amazonia (b) y en la Amazonia surccntral (c). Una linea continua de- limita la sclva amazonica (arriba). Grupos de especies hastante localizadas forman los centros de distribu- tion (A-F; abajo) cada uno de cllos formado pot It) a 50 especies. Adaptado de Haffer, 1974, Fig. 9.13 y Haffer, 1978. Otros grupos de especies en- demicas caracterizan la selva atldntica de Brasil meri- dional y la sclva transandina de Sudamerica norocci- dental. upper and lower Amazonian centers are in contact within the area of the "weak" Imeri and Rondonia centers (characterized by com- paratively few species). 1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 123 Based on geosciencc data, Quaternary for- est refugia in Amazonia probably were lo- cated near the windward base of hilly ranges and mountains, such as the northern margin of the Brasilian Plateau, the eastern base of the Andes, and the northern base of the moun- tains in the interior Guianas. These areas correlate well with the location of the zoo- geographically defined distribution centers which, therefore, are assumed to indicate the approximate location of Quaternary forest refugia. Additional refugia for the avifauna (Fig. 5:8) based on similar geological and zoogeographical criteria have been proposed for the Atlantic forest of southeastern Brasil (Miiller, 1973; Jackson, 1978) and for the trans-Andean forest region in western Colom- bia and in Middle America (Haffer, 1967, 1969, 1974, 1975). A recent study of a com- plex Neotropical avian genus on the basis of the refuge concept is Fitzpatrick's ( 1976 ) analysis of the Todirostrum flycatcher group; also see the recent review by Dorst ( 1976 ) . Lizards and frogs. — Analysing the popula- tion structure and character variation in the Amazonian Anolis chrysolepis group, Vanzo- lini and Williams (1970) and Vanzolini (1970, 1973) arrived at conclusions regarding the history of faunal differentiation during the Quaternary that are strikingly similar to those reached by Haffer (1969, 1974). Sev- eral extensive areas of uniform character ex- pression (core areas) in Anolis chrysolepis are separated by regions where complex char- acter variation suggests hybridization or in- trogression along zones of secondary contact. Vanzolini and Williams (1970) interpreted this situation as being the result of secondary intergradation of populations that had differ- entiated in geographic isolation, and they as- sumed changes in the distribution of forest in Amazonia during the course of climatic fluctu- ations. These authors reconstructed several forest refugia around the periphery of Ama- zonia (Fig. 5:8), most of which coincide closely with those proposed by Haffer ( 1969, 1974) for birds. Additional forest refugia probably determined the differentiation of Neotropical forest reptiles and will be iden- tified when other species are studied in detail. Miiller (1972, 1973) using mostly herpetolog- ical and ornithological data described several broad centers of dispersal but did not sug- fest specific refugia. A comparison shows that the boundaries of Miiller's centers rarely co- incide with those of avifaunal core areas but fall in the peripheral gradient of decreasing species numbers or coincide with river bar- riers. Miiller (1973) did not state the criteria he used to delimit the various broad centers, which on his maps are frequently separated only by narrow corridors. In recent years, several herpetologists ac- cepted the notion of a Pleistocene origin of numerous species and subspecies of Neo- tropical frogs, lizards and snakes in forest refugia: Duellman (1972, 1978), Heyer (1973), Duellman and Crump (1974), and Silverstone (1975, 1976) for certain Amazon- ian and trans-Andean frogs, Gallardo ( 1965, 1972 ) and Lynch ( this volume ) for the South American amphibian fauna generally, Dixon (this volume) for Amazonian reptiles, Ech- ternacht (1973) for Middle American lizards (Ameiva), Hoogmoed (1973, this volume) for the herpetofauna of the Guianas, Jackson (1978) for two genera of eastern Brasilian iguanid lizards, and C. W. Myers (1973, 1974) for two genera of snakes. Dixon (this volume) and Lynch (this volume) mapped clusters of endemic reptile and amphibian species in upper and lower Amazonia. These centers of distribution more or less coincide with similar centers established for the forest avifauna. At this stage, a more quantitative treatment of available distributional data would be desirable, such as the derivation of herpetological distribution centers by means of contoured diversity maps of localized spe- cies clusters and a comparison of the location of secondary contact zones with that of dis- persal centers and postulated refugia. Insects. — The distribution and population structure of two groups of Neotropical insects are relatively well known and permit initial biogeographic analyses — 1) Certain Dro- sophila flies that have been collected and analysed extensively in the course of genetic investigations, and 2) butterflies of the genus Heliconius. Spassky et al. (1971) summarized the distribution of several semispecies and closely related species of Drosophila. These authors and Winge (1973) concluded that the various forms may have originated in for- 124 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fig. 5:8. Location of presumed Quaternary forest refugia (hatched) in tropical lowland South America. Explanations: 1. Reconstruction based on the distribution patterns of Neotropical birds ( Haffer, 1967, 1969, 1974). 2. Reconstruction based on the population structure of Amazonian lizards (Anolis chnjsolcpis group, Vanzolini and Williams, 1970, Vanzolini, 1970; wide hatches indicate core areas). 3. Reconstruction based on the distribution patterns of subspecies and species of Heliconius butterflies ( Brown et al., 1974, Brown 1977a,b). 4. Reconstruction based on the distribution patterns of four families of Amazonian trees (Prance, 1973). Position de presuntos refugios cuaternarios de selva humeda (rayado) en las Metros hajas tropicales de Sudamerica. Explicaciones: 1. Reconstruction basada en los patrones de distribution de aves neotropicales (Haffer, 1967, 1969, 1974). 2. Reconstruction basada en la estructura de poblaciones de lagaitos amazonicos (grupo de Anolis chrysolepis, Vanzolini y Williams, 1970, Vanzolini, 1970; Hneas verticales indican areas nu- clearcs). 3. Reconstruction basada en la distribution de subespicies y espeties VOROHUE FM. "PRE-ENSENAOENSE" tc UQUIAN 111 BARRANCAS DE LOS LOBOS FM. PUELCHES FM. Fig. 6:3. Pleistocene South American Provincial Ages and main referable mammal-bearing lithostratigraphical units in the regions under study. Edades Provinciates pleistocenicas sudamericanas y principalcs unidades litoestratigrdficas mamaliferas re- feribles en las regiones estudiadas. of Nearctic mammals in the South American fossil record is the basis for correlation be- tween the Land-Mammal Ages of both conti- nents. Immigrants of known South American ancestry ( megalonychid and mylodontid edentates) are first recorded in North Amer- ica in the Hemphillian Mammal Age (mid- dle Pliocene) (Hirschfeld and Webb, 1968). In South America, mammals of Nearctic ori- gin ( procyonid carnivores ) appear in deposits corresponding to the Huayuerian Age, which mainly for that reason has been tentatively referred to the middle Pliocene (Pascual et al., 1965). A greater number of northern im- migrants (cricetine rodents, a skunk and a peccary) are recorded in the Montehermosan (Reig, 1952) and many more in the Uquian. Many taxa of South American origin occur in Blancan-early Irvingtonian deposits in North America; thus, the Uquian has been grossly correlated with those faunal ages (Webb, 1976). According to this and the radiometric dates established for western North America, the Plio-Pleistocene boundary could be placed in the early Uquian (Mar- shall and Pascual, 1979). The recognition of the Holocene as a dis- tinct epoch has been criticized by some work- ers, who consider the Holocene to be only an interglacial interval within the Pleistocene. The massive extinction of the megafauna is the paleontological criterion used to place the Pleistocene-Holocene boundary. This extinc- tion is verified at the end of the Lujanian Age in South America and in late Rancho- labrean Age in North America (Martin, 1975), thereby providing some possible corre- lation of both ages (Pascual et al., 1965). 146 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Nevertheless, a few elements of the mega- fauna survived into Holocene times. For example the ground-sloth Mylodon is known from deposits younger than 5394 ± 55 y.b.p. in southern Chile (Saxon, 1976). Glaciations Very often the concept of the Quaternary has been equated with the Ice Age and the base of the Pleistocene considered synchro- nous with the onset of glaciation. But, in fact, climatic deterioration began to accelerate about 10 million years ago (early Pliocene) (Berggren and Van Couvering, 1974). Dif- ferent lines of evidence indicate the occur- rence of glaciation in polar and mountainous areas of both hemispheres at least since Plio- cene times (Curry, 1966; Hays and Opdyke, 1967; Fleck et al., 1972; Mercer and Sander, 1975), but this does not imply the synchro- nous presence of lowland ice sheets in tem- perate regions. The four classic Pleistocene glaciations of Europe (Giinz, Mindel, Riss, Wiirm) tradi- tionally have been correlated to those of North America ( Nebraskan, Kansan, Illinoian, Wisconsin), but the problem is complex and has been oversimplified. Recent evidence has led to a complete revision of that scheme, and accurate intercontinental correlation of glacial events is still under debate (Berggren and Van Couvering, 1974). Widely differing opinions have been expressed concerning the number of major glacial episodes as well as the extension of glaciated areas in southern South America. Four postulated major glaciations (Valli- manca, Colorado, Diamante, Atuel) on the Argentinian side of the Andes have been cor- related with the classic four glaciations in the Northern Hemisphere by Groeber ( 1952, 1954). He contended that during the first two assumed glacial events, which were ascribed to the "Eoquaternary," extensive ice sheets covered the Andean region and Patagonia. In northern Patagonia glaciers were supposed to have reached the Atlantic Ocean south of Rio Colorado. The other two glaciations were re- ferred to the "Neoquaternary" and were re- stricted mainly to the Andean region. The existence of local centers of glaciation in Pata- gonia and the advance of a postulated "great Somuncura glacier" extending to the sea dur- ing the youngest glaciations also have been suggested by Auer (1956, 1958, 1960, 1974). Such schemes have been mostly discarded for recent investigations have demonstrated that much of the evidence of glacial action upon which those hypothesis were based are the result of different geomorphological proc- esses (Polanski, 1963, 1965; Methol, 1967; Fi- dalgo and Riggi, 1970; Fidalgo, 1973). There is now general agreement that east of the Patagonian Andes glaciation extended for a relatively short distance, reaching the present Atlantic coast only south of about 52°S (An- tevs, 1929; Caldenius, 1932; Flint and Fidalgo, 1963, 1968). On the other hand, westward flowing glaciers reached the Pacific Ocean as far north as 43°S (Mercer, 1976). The glaci- ated areas progressively decreased north- wards, where they were confined mainly to mountainous areas even during full glacial times (Groeber, 1936; Frenguelli, 1957b). Generally, three or four glaciations have been recognized on the eastern side of the Patagonian Andes, although the interpretation of those glacial events has differed. They have been considered either as minor episodes roughly equivalent to the Wurm-Wisconsin and perhaps also to an older event (Calden- ius, 1932; Flint and Fidalgo, 1963, 1968) or as different major glaciations ( Feruglio, 1949; Auer, 1956, 1960, 1970). Moraine belts in the Chilean lake district have been attributed to three major glaciations (Laugenie, 1971; Mer- cer, 1972, 1976); the last was named the Llanquihue Glaciation by Heusser (1974). Recently, glacial deposits in southwestern Patagonia (about 50°S) have been dated be- tween 1.2 and probably 1.0 m.y.b.p.; thus, they correspond to an early Pleistocene glaci- ation ( Mercer and Sander, 1975 ) . The avail- able data permit reliable intercontinental comparison of only the most recent glacial fluctuations. Environmental Changes from Pliocene to Holocene Times It is premature to attempt a detailed ac- count of the climatic and environmental changes in temperate South America from 1979 BAEZ & SCILLATO YANE: CENOZOIC OF ARGENTINA 147 Pliocene to Holocene times and their in- fluence on biogeographic patterns. Many important factors and much of the evidence for the evaluation of past climates have been studied inadequately, and interpretations fre- quently are contradictory. Therefore, only an account of major events and general trends is possible. Few palynological, micropaleontological or isotopic dating studies have been carried out in this part of South America. Mainly, we have considered evidence provided by fossil mammals, especially distributional shifts of taxa whose ecological requirements are known, because the fossil record of other groups is very fragmentary. Closer attention to the stratigraphy of deposits in which fossils are collected and more accurate data on cor- relations and ages of assemblages are neces- sary in order to reconstruct the climatic fluc- tuations of such a short interval of the earth's history. In many parts of the world, mammalian faunas have furnished interesting clues to possible late Cenozoic climatic fluctuations, and the inferences based on them have been supported by other kinds of evidence. Thus, a basic chronological framework of the cli- matic record has been established. It is note- worthy that in southern South America, Pleis- tocene mammals have only been recorded quite far from the glaciated area. No taxa indicative of extreme conditions have been recognized. Although in Pleistocene times glaciations were mainly restricted to the mountainous areas, temperature and moisture changes during glacial and interglacial inter- vals probably caused important shifting of the biotas of the nonmountainous regions. In order to provide a better understanding of the Late Cenozoic environmental changes, each region is considered separately. Patagonia. — By the late Miocene, the de- velopment of increasingly xeric conditions had begun in this region. Nevertheless, in northern Patagonia, especially in the area of influence of the pre-Colorado and pre-Negro rivers, a warmer and more humid climate than that of today seems to have persisted during early and middle Pliocene. This is sug- gested by the record of a megalonychid eden- tate of the subfamily Orthotheriinae, which also could indicate a woody environment (Scillato Yane, Uliana and Pascual, 1976). According to sedimentological and geomor- phological data provided by Andreis (1965) and Volkheimer ( 1971 ) , by the late Pliocene the climate at those latitudes had changed to drier and colder conditions. Although the general trend in this region has been towards a drier climate as a result of the Andean uplift, some evidence suggests more humid conditions during certain phases of the Pleistocene. Few mammalian faunas of that age are known, but at least those tenta- tively referred to the upper Pleistocene ( Ame- ghino, 1902, 1906; Parodi, 1930) support the presence of open environments. The occurrence of paleargids aridisols, a group of soils of Pleistocene age, in southern Patagonia testifies to the existence of a period of more humid conditions than at the present. Paleosoils from the tablelands of central and northern Patagonia are being studied by J. A. Ferrer (pers. comm.). Prevalence of dry conditions in eastern Patagonia during Quaternary times is sug- gested by palynological analysis of sea cores from the Argentine Basin in the southwestern Atlantic Ocean (Groot and Groot, 1964, 1966). A cyclic alternation of two types of pollen zones occurs with depth. Those indica- tive of more pronounced arid environments (high percentage of Chenopodiaceae pollen) were considered to represent "glacial stages." Furthermore, certain features suggest eustatic lowering of sea level, an inference also sup- ported by the study of diatoms (Groot et al., 1967). More humidity could account for the relative increase of the arboreal pollen in the samples attributed to the "interglacial stages." This agrees with the climatic interpretation of the palynology of interglacial peat beds in northwestern Patagonia and Tierra del Fuego (Auer, 1958, 1970). Humid glacial ages and relatively warm and dry interglacial ages in the Andean region of Argentina and central and southern Chile were proposed in accordance with a pollen diagram and the fluctuations of the carbonate content of a deep-sea core from the southeast- ern Pacific Ocean (Groot and Groot, 1966). However, palynological spectra of strati- graphic sections in the southern Chilean lake 148 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 district reflect more complex climatic changes since the last interglacial, as well as drier conditions during the Llanquihue Glaciation (Heusser, 1974). Evidently, eustatic descent of sea level caused the emersion of much of the present continental shelf; thus, eastern Patagonia sup- porting a steppe vegetation became larger. The probable connection of the Malvinas Islands to the mainland could explain the presence there of terrestrial species with rela- tives on the continent, for example the canid, Dusycion australis (Auer, 1958:223-224). The sub-Antarctic forests, which now extend along the eastern slopes of the Andes as far north as about 36° (Fernandez, 1976), could have descended to lower altitudes in glacial times (Vuilleumier, 1971). Sub-Andean Region. — The rich mamma- lian faunas recorded in the Pliocene of this region have a marked subtropical character, thereby suggesting different environmental conditions from those existing at present. It is worth mentioning the occurrence of giant erethizontid and echimyid rodents, as well as anteaters (Rovereto, 1914), both myrmeco- phagids and cyclopodids (Hirschfeld, 1976), in areas where their living relatives are not found. Also dasypodids closely related to the living tropical and subtropical Euphractus were recorded by Scillato Yane ( 1975 ) . Fos- sil trunks occur in Pliocene horizons but most of them have not been identified (Boden- bender, 1924; Frenguelli, 1937; Menendez, 1962; Ramos, 1970). Those assigned to the leguminosid genus Acacioxylon exhibit growth rings that indicate wide seasonal dif- ferences of rainfall (Menendez, 1962, 1971). These data suggest that environments similar to those of the present Western Chacoan dis- trict (sensu Cabrera, 1971) could have existed in the Subandean Region. At that time, many of the mountain ranges that comprise the eastern border of this region were not so high as today and consequently did not constitute an effective barrier to the humid Atlantic winds. Thus, some Chacoan species could have ranged farther west than they do now. The prevalence of semiarid to arid climates in intermontane basins in the northern areas is indicated by a variety of geological data (Bossi, 1969; Ramos, 1970; Caminos, 1972; Gordillo and Lencinas, 1972). The regions along the eastern side of the Andes, then lower and wanner than today, constituted an important pathway for migra- tion between the subtropical regions and the more temperate southern Pampean Plain. Moreover, similar ecological conditions to the present phytogeographic Chacoan Province (sensu Cabrera, 1971), with open xerophytic woodland, apparently existed in the south- western part of Provincia de Buenos Aires. Upper Pliocene elements recorded from there, such as cariamid birds (Tonni, 1974) myrme- cophagid anteaters ( Kraglievich, 1934), and the anuran Bufo paracnemis (Gasparini and Baez, 1975) now are absent in that area. The Sub-Andean Region was affected by important tectonic events during, and subse- quent to, the Plio-Pleistocene. These had con- siderable physiographic and climatic conse- quences. The increasing rain-shadow effect accentuated the desiccation of climate. A fauna referred to the earliest Pleistocene in the northernmost part of the region has elements indicative of relatively humid conditions, such as a beaver (Kraglievich, 1934). By that time, a dry regime prevailed in the southern areas (Polanski, 1963). The few known later faunas denote a significant impoverishment in contrast to those of the Pliocene. Some mammals there at present are the caviid Microcavia, the mustelid Lyncodon, and the dasypodid Chaetophractus vcllerosus. Fauna] connections with the southwestern Pampean Plain and Patagonia were enhanced by the development of an arid regime. The dry periods that affected the former area per- mitted a more eastern distribution of the xeric-adapted biota from the western regions. Similar expansion to the east of the semi-arid vegetation of the Monte during Pleistocene dry phases was suggested by Solbrig (1976). Mesopotamia. — Pliocene vertebrates of this region are almost exclusively known from cliffs along the Rio Parana. Most of the re- mains of continental vertebrates have been found in sandstones and conglomerates that constitute the so-called "Mesopotamiense." They overlie typical marine deposits of the Atlantic transgression, the "Paranense Sea," 1979 BAEZ & SCILLATO YANE: CENOZOIC OF ARGENTINA 149 which extended over a great part of the Chaco-Pampean Plain during the late Mio- cene (Frenguelli, 1920; Rossi de Garcia, 1966). Within the "Mesopotamiense" there is an evident alternation of sediments corre- sponding to different environments but most are fluvio-deltaic sediments. Fossil remains are indicative of humid subtropical conditions (Pascual and Odre- man Rivas, 1971; Gasparini and Baez, 1975; Baez and Gasparini, 1977). Presumably, the gallery forests along the river favored the southward extension of the tropical biota ( Pascual and Odreman Rivas, 1971 ) . A south- ward extension of tropical flora (Cabrera, 1971) and fauna (Ringuelet, 1955, 1961) is evident today, but the southward penetration probably was greater in Pliocene times, when the climate was warmer. There is an abun- dance of fossil tree trunks, turtles, large croco- diles, megalonychid and pampatheriine eden- tates, and dinomyid rodents in the Pliocene deposits. There is an absence of chinchillid and octodontid rodents and hegetotheriid and mesotheriid notoungulates, which are characteristic of open environments. The composition of these assemblages is quite different from those of about the same age in the Sub-Andean and Pampean regions. These differences reveal that these areas had distinct ecological and biogeographical fea- tures. Pleistocene deposits are widely distributed in Mesopotamia, but in most cases neither their stratigraphy nor their paleontological content has been studied adequately; there- fore, their chronological assignment is still uncertain. Even so, the presence of certain taxa is quite interesting from the paleoen- vironmental point of view. The record of the caviid rodent Dolichotis in the middle Pleis- tocene of northwestern Mesopotamia (Alva- rez, 1974) indicates that at certain times the climate was drier than at present; drier con- ditions also are indicated by sedimentological data (Alvarez, 1974). At present Dolichotis is restricted to the Patagonian, Sub-Andean and southwestern Pampean regions. Chaco-Pampean Plain. — T his region started to acquire its present features after the regression of the late Miocene "Paranense Sea" (Pascual and Odreman Rivas, 1971, 1973). Pliocene and Pleistocene stratigraphic sequences and their paleontological content are best known in the Provincia de Buenos Aires; therefore, our analysis will deal prin- cipally with that portion of the Chaco-Pam- pean Plain. The available studies of successive faunas and sediments of Pliocene and Pleistocene age have not yet furnished evidence of drastic changes. In general there was an overall cooling trend from the more warm temperate conditions of the early Pliocene to the more moderate present ones with colder periods probably occurring during the Pleistocene. Concerning precipitation, at present the re- gion has a permanently humid climate, with a change to drier conditions and some pre- dominantly dry months in the western part. This regime seems to have prevailed during most of the Pliocene and Pleistocene, but an increase of xeric conditions, especially during certain intervals, is evident (see below). Since this region came into existence, the dominant vegetation type has been grass steppe. However, the climatic fluctuations just mentioned would have favored greater eastward expansion of xerophytic forest, which presently borders the Pampean Plain to the west and north. Echimyid rodents, now living in tropical and subtropical regions, are recorded from this area in the Pliocene and lower Pleisto- cene. Also in early Pliocene times (Chasi- coan Age) megalonychid and nothrotheriine tree sloths were abundant there, but they are inconspicuous in later ages (Scillato Yane, 1977, 1979). These groups of rodents and edentates are found frequently in Miocene deposits of southern Patagonia; the north- ward shift of their southern limit of distribu- tion reflects the climatic deterioration in the south. The occurrence of large dasypodids, closely related to the tropical and subtropical living Enphractus, in the Pliocene and of procyonids in the middle and upper Pliocene and lower Pleistocene is also significant. Ac- cording to these and other faunal evidence, it can be postulated that throughout the Plio- cene and early Pleistocene ( Uquian Age ) the climate was wanner than today followed by 150 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 a gradual cooling. This agrees with the more pronounced deterioration of climatic condi- tions in Patagonia after early Pleistocene time, as indicated by palynological data (Groot and Groot, 1966). The evidence points to a desiccation of climate during the Pliocene. The abundance of sloths in the Chasicoan fauna suggests a humid regime (Scillato Yane, 1977, 1979) but contrasts with their absence in the Huayquer- ian. The presence of some sub-Andean ele- ments in deposits assigned to the latter age also could be indicative of drier conditions at that time (Bondesio et al., 1979). By the late Pliocene, at least in the southwestern part of the Provincia de Buenos Aires, a regime characterized by seasonal differences in rainfall seems to have prevailed. An envi- ronment similar to that of the present Cha- coan phytogeographic province, with open xerophytic woodlands, has been inferred there from the composition of the fossil fauna (Tonni, 1974; Baez and Gasparini, 1977). The stratigraphical units of early Pleisto- cene to Recent age, especially in the coastal area and in the northeastern part of the Provincia de Buenos Aires exhibit an alterna- tion of subaqueous (fluvial, lacustrine, palu- dal ) and aeolian deposits ( Kraglievich, 1952, 1953; Frenguelli, 1957a). These alternations of continental deposits of different origins have been associated with fluctuating climatic conditions, but there is no agreement as to what climatic condition, especially tempera- ture, corresponds to each of the lithological changes. The existence of alternating humid and arid cycles during the late Pleistocene and Holocene also is indicated by geomorpho- logical data (Tricart, 1973). Intercalated marine and brackish-water levels occur discontinuously in different areas and have been correlated with the above mentioned continental subaqueous deposits which probably indicate humid conditions (Frenguelli, 1957a; Tricart, 1973). If the marine levels are interpreted as resulting from eustatic sea level rise, the correlation of the humid phases with the interglacial stages fol- lows (Tricart, 1973). Accordingly, cold and arid or semiarid climates seem to have alter- nated with warm and humid ones. Prelim- inary palynological analysis of some samples from the Argentine shelf off the mouth of the Rio de la Plata also suggests more pronounced aridity during the glacial stages (Groot et al., 1967:208). The latest Pleistocene is represented in this area by the lower section of the Loberia Formation (Fig. 6:3), where the last repre- sentatives of the megafauna still occur. Sedi- mentological characteristics (sand dunes and loess) and geomorphological evidence (Tri- cart, 1973) indicate extremely arid conditions at that time. This dry phase must have dis- turbed the equilibrium of the biota in that reduction of grazing land affected primarily the large herbivores, thus contributing to the extinction of the local megafauna. The alternation of arid and humid cli- mates can be explained by the changes in the distribution of ocean currents and wind sys- tems over the South American continent dur- ing glacial and interglacial stages (Damuth and Fairbridge, 1970). During glacial phases the Chaco-Pampean Plain was under the in- fluence of the dry southwesterly winds that acted as the depositional agent for the loess. Even though the fossil record is rich, the available paleozoological data do not provide clear support to the environmental fluctua- tions during the Pleistocene and Holocene, for much of the material is not accurately as- signed to stratigraphic levels. Furthermore, formational units have a long time range, during which climatic fluctuations took place. This probably explains the records of taxa that are indicative of opposing climatic con- ditions from a single locality and formation. Nevertheless, significant distributional shifts in the biota can be attributed to Pleistocene environmental changes. Drier conditions in the Pampean Plain are indicated by the occur- rence of characteristic elements of the present Sub-Andean and Patagonian regions, such as the mustelid Lijncodon and the caviids Do- lichotis and Microcavki, in Pleistocene de- posits. Dolichotis also occurs in sediments of the same age in southern Uruguay (Calca- terra, 1972). On the other hand, more equa- ble climatic conditions would have allowed procyonids and the dasypodid Euphractus to live on the Pampean Plain. 1979 BAEZ & SCILLATO YANfi: CENOZOIC OF ARGENTINA 151 Other regions of temperate South Amer- ica.— Available data from other areas within temperate South America do not permit us to outline a sequence of the climatic and en- vironmental changes that occurred during the late Cenozoic. On the other hand, many geo- morphological, sedimentological and phyto- geographic studies have provided evidence of Quaternary fluctuations. In the states of Sao Paulo, Parana and Rio Grande do Sul in southeastern Brasil nu- merous mammalian remains have been re- corded from Pleistocene continental units that tentatively have been referred to the Lujanian Age (late Pleistocene) of Argentina (Paula Couto, 1975). The mammalian fauna is pre- dominantly one of savanna taxa and has Pampean affinities, especially the assemblage from Rio Grande do Sul. Some taxa could be indicative of different conditions than those existing at present; wild llamas of the genus Palaeolama are recorded from low alti- tudes thus suggesting a colder climate for the region. Paula Couto (1975) related the east- ward shift of their range to a wider extension of the glaciated areas in the Andean region and adjacent plateaus where their close rela- tives are restricted today. Also, a former wider expansion towards the east of the "Andean forests" has been postulated to ac- count for the presence of some Andean plant genera in the Araucaria forests of southeast- ern Brasil (Klein, 1975). Sedimentological and geomorphological studies provide evidence for climatic changes and fluctuations during the Quaternary in southern Brasil. Distinct epochs of pedimen- tation, during which xeric conditions pre- vailed, have been established. These periods seem to be related to times of lowered sea level and thus are correlated with Pleistocene glacial phases (Bigarella, 1964). The evi- dence points to an alternation of semiarid and humid episodes during glacial and inter- glacial times, respectively, with smaller cli- matic fluctuations within both the xeric and humid phases (Bigarella, 1964; Ab'Saber, 1977). These cyclic changes are also docu- mented by phytogeographic data, although the chronology is far from clear (Klein, 1975). CONCLUSIONS The general trend of climatic change since the Pliocene has been towards a colder and drier regime, but this trend evidently has not been uniform, and many fluctuations took place. Rain-shadow effects produced by the rising mountains constituted an important factor leading to the change of environmental conditions. Even though our knowledge of past climates in temperate South America is incomplete, the main environmental changes during the Late Cenozoic can be broadly out- lined, as follows. 1. Development of an increasingly arid and cold regime began by late Miocene in Patagonia. Many forest-savanna inhabitants retreated north and west. In northern Patagonia wanner and more humid conditions persisted during early to middle Pliocene. In Quaternary times a dry and cold climate prevailed, especially in the eastern areas although there is evidence of more humid phases during the Pleistocene. 2. During Pliocene times the regions situated to the north of Patagonia had at least a warm- temperate climate. Many elements of the sub- tropical fauna reached a more southern dis- tribution than they do today. 3. Plio-Pleistocene tectonic events accentuated the differentiation of the sub-Andean region, which formerly had connections in its north- ern part with the Chacoan region. Subsequent to that time an essentially dry regime was established there. 4. During the Pleistocene, glaciations extended over the Andean region and reached the pres- ent Atlantic coast only south of 52°S. In the northwestern ranges of Argentina glaciers were restricted to the highest elevations and upper portions of fluvial valleys. 5. In the middle and late Pleistocene and Holo- cene, the climate fluctuated in the Pampean Plain, especially the humidity. Dry phases favored geographic extension of xeric-adapted species, to the east and north. Also they may have caused the isolation or migration of more mesic types. 6. Evidence of important Quaternary climatic fluctuations in southeastern Brasil has been provided mainly by sedimentological, geo- morphological, paleontological and phytogeo- graphical studies. An alternation between warm and humid and relatively cold and dry periods, with minor fluctuations within those phases, has been proposed. 152 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 ACKNOWLEDGMENTS We thank Mrs. M. Guiomar Vucetich, Drs. Rosendo Pascual, Francisco Fidalgo, Eduardo Tonni, Geol. Jose A. Ferrer (Museo de La Plata), Dr. Edgardo Romero (Facultad de Ciencias Exactas y Naturales, Buenos Aires) y Prof. Guillermo del Corro (Museo Argen- tino de Ciencias Naturales, Buenos Aires) for their valuable comments. Special thanks are extended to Dr. Victor Ramos (Servicio Na- tional Minero Geologico, Buenos Aires), who critically read the manuscript and provided useful information. RESUMEN La actual zona templada de America del Sur incluye (de acuerdo con la clasificacion climatica de Koppen): Uruguay, la mayor parte de Paraguay, partes de Bolivia, Brasil y Argentina, asi como una portion territorial a lo largo de la costa pacifica al sur de los 38°S. Para los fines del presente estudio, se han considerado tambien la region subandina y Patagonia, donde actualmente prevalecen climas aridos. Esto se debe a que su historia paleoambiental se halla intimamente ligada a la de los aludidos ambitos actualmente tem- plados. La mayor parte de las evidencias disponi- bles indicadoras de cambios ambientales, especialmente durante el Terciario tardio, se refieren al territorio argentino, en el que pue- den considerarse cuatro regiones principales: 1) Patagonia, 2) Area Subandina, 3) Meso- potamia, y 4) Llanura Chaco-Pampeana. Una adecuada comprension de los cam- bios climatico-ambientales acaecidos en Sud- america templada durante el Cenozoico tardio requiere inevitablemente, la consideration de los tiempos geologicos anteriores. Los datos disponibles indican un pronunciado deterioro de las condiciones que, aun hacia el Mioceno medio a tardio, eran mucho mas benignas y uniformes que las actuates. Cabe seiialar que en esta parte de Ameri- ca del Sur la mayoria de las unidades lito- estratigraficas y cronoestratignificas del Ceno- zoico continental no han sido formalmente establecidas. De hecho, ha sido el reconoci- miento de "Edades mastozoologicas" (tenta- tivamente referidas a las Epocas de la escala geocronologica mundial) lo que ha provisto, para toda Sudamerica el principal criterio de correlation. Estudios de diversa indole (litologicos, geomorfologicos, paleontologicos, fitogeo- graficos) han suministrado information que permite esbozar un relato de la evolution ambiental en las regiones consideradas. Hasta el Mioceno medio a tardio las con- diciones prevalecientes en Patagonia fueron mucho mas calidas y humedas que las hoy alii vigentes; asi lo indica el registro de nu- merosos vertebrados cuyos parientes actuales viven en areas intertropicales, parcialmente boscosas o selvaticas. A partir del Mioceno tardio se produjo un notable desecamiento (motivado por la circunstancia de que los Andes australes alcanzaron ya una altura suficiente como para actua como efectiva ba- rrera a los vientos humedos del Pacifico ) y una mas progresiva atemperacion (en parte coin- cidente con un partial englazamiento de Antartida occidental). Estos cambios deter- minaron la retraction hacia el norte y oeste de muchos elementos de la biota. La consideracion en particular del Ceno- zoico tardio nos remite, primeramente, al problema de los limites entre las Epocas im- plicadas (Plioceno, Pleistoceno y Holoceno). El limite plio-pleistocenico no ha sido aun claramente determinado en Sudamerica; tentativamente se lo ubica hacia el Uquiense temprano (Blanquense tardio de USA). Tal incertidumbre responde tanto a la ausencia de un criterio unanime a nivel mundial en esta materia, como a las dificultades implicitas en correlacionar las Edades locales sudameri- canas con las de otros continentes. El limite pleisto-holocenico coincide aproximadamente con el fin de la Edad Lujanense, serialado por una extincion masiva de la megafauna. El estudio de los eventos glaciales — cuya importancia fundamental en la evolution climatica del Cenozoico tardio es universal- mente reconocida — se halla aun en Sudameri- ca en una etapa preliminar. No existe acuerdo en cuanto al numero y extension de las gla- eiaciones, pero puede desde ya desecharse el esquema simplista que establecia una exacta correspondencia con las cuatro clasicamente 1979 BAEZ & SCILLATO YANE: CENOZOIC OF ARGENTINA 153 reconocidas en el Pleistoceno del Hemisferio Norte. En el SO de Patagonia fueron reciente- mente reconocidos varios cpisodios glaciales que datan del Plioceno tardio y Pleistoceno temprano. Mucho mejor documentada esta la glaciacion suprapleistocenica, groseramente correlacionada con Wiirm-Winsconsin del Hemisferio Norte; pero debe remarcarse que afecto especialmente a la region andina, y que solo al sur de los 52°S los glaciares al- canzaron la actual costa atlantica. La evolucion ambiental durante el Ceno- zoic tardio en las cuatro areas del territorio argentino mencionadas precedentemente pue- de sintetizarse asi: Patagonia. — Con exception dc la region mas septentrional, el clima fue frio y semi- arido a arido durante la mayor parte del Plioceno. Estas condiciones parecen haberse acentuado durante el Cuaternario, si bien ciertas evidencias sugieren algunos lapsos mas hiimedos. Los datos disponibles no indican un extenso englazamiento. Area Subandina. — Durante el Plioceno el registro paleomastozoologico incluye numero- sos taxa tipicamente intertropicales, que in- dican condiciones parcialmente boscosas y calidas, posiblemente algo mas humedas que las actuales; de tal modo, esta area (continua entonccs con la chaquefia hacia el norte) obro corao una "via de conexion" del SO de la llanura pampeana con ambitos francamente subtropicales. No obstante, como consecuen- cia de los Movimientos Andinos, existian ya cuencas intermontanas en la parte septentrio- nal donde pudieron prevalecer condiciones mas aridas. Los eventos tectonicos del Plio- ceno tardio-Pleistoceno temprano acentuaron notablemente el bloqueo a los vientos hii- medos del NE, con la consecuente desecacion progresiva del clima en toda esta region. Mesopotamia. — Durante el Plioceno las condiciones fueron aiin mas calidas y hume- das que las hoy vigentes. El area influenciada por el rio pre-Parana obro ya entonces como una importantisima via de conexion" entre la llanura bonaerense y ambitos biogeograficos mas septentrionales. En el transcurso del Cuaternario siguieron prevaleciendo condi- ciones calidas y humedas, pero muy probable- mente algunos lapsos mas secos hicieron sen- tir su influencia tanto en esta region como en territorio uruguayo. Llanura Chaco-Pampeana. — Casi toda la documentation disponible se refiere a la mitad austral de esta area (Pampasia). Los estudios de las sucesivas "faunas" y sedi- mentitas pliocenicas y pleistocenicas no han revelado la ocurrencia de cambios drasticos. No obstante, a partir del Plioceno temprano tuvo lugar una progresiva atemperacion. Desde el Pleistoceno medio hasta el Holoceno es factible reconocer varias fluctuaciones cli- maticas: lapsos relativamente calidos y hii- medos (con frecuente depositation fluvio- lacustre) alternaron con otros relativamente frios y secos (en los que predomino amplia- mente la sedimentation de "limos loessoides," o aun de verdaderos loess). Un analisis pre- liminar del contenido paleomastozoologico de las respectivas unidades estratigraficas avala, en general, esta interpretation. Es muy im- portante seiialar que los lapsos presumible- mente calidos y hiimedos parecen coincidir con ingresiones marinas, en tanto que los frios y secos corresponden a regresiones. Estos ascensos y descensos del nivel marino podrian ser de origen eustatico. En lo que ataiie a otras areas fuera del territorio argentino, no existe information suficiente sobre el Terciario tardio. En cam- bio, hay claras evidencias de fluctuaciones climaticas ciclicas especialmente para el Cua- ternario del sudeste de Brasil. De acuerdo a datos de diversa indole, lapsos calidos y hiimedos alternaron con otros relativamente frios y secos. LITERATURE CITED Ab'Sabeb, A. N. 1977. Espacos ocupados pelo ex- pansao dos climas secos na America do Sul, por ocasiao dos periodos glaciales quatemarios. Paleo- climas 3:1-19. Alvarez, B. B. 1974. Los mamiferos fosiles del Cuaternario de arroyo Toropi, Corrientes (Argen- tina). Ameghiniana 11(3) :295-311. Amechino, F. 1902. L'age des formations sedimen- taires de Patagonie. An. Soc. Cient. Argentina 54(6):283-342. Amechino, F. 1906. Les formations sedimentaires du Cretace superieur et du Tertiaire de Pata- gonie. An. Mus. Nac. Hist. Nat. Buenos Aires (3a)8: 1-568. 154 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Andreis, R. 1965. Petrografia y paleocorrientes de la Formacion Rio Negro ( Tramo Genera] Conesa- Boca del Rio Negro). Rev. Mus. La Plata, Secc. Geol. 5(36):245-310. Antevs, E. 1929. Maps of the Pleistocene glacia- tions. Geol. Soc. America Bull. 40:631-720. Auer, V. 1956. The Pleistocene of Fuego-Patagonia. Part I: The Ice and Interglacial Ages. Ann. Acad. Sci. Fennicae, Ser. A3 45:1-226. Auer, V. 1958. The Pleistocene of Fuego-Patagonia. Part II: The history of the Flora and Vegetation. Ibid., 50:1-239. Auer, V. 1960. The Quaternary history of Fuego- Patagonia. Proc. Roy. Soc. London, Ser. B152: 507-516. Auer, V. 1970. The Pleistocene of Fuego-Patagonia. Part V: Quaternary Problems of Southern South America. Ann. Acad. Sci. Fennicae Ser. A3 100: 1-194. Auer, V. 1974. The isorhythmicity subsequent to the Fuego-Patagonia and Fenoscandian ocean level transgressions and regressions of the latest glaciation. Ibid. 115:1-88. Baez, A. M., Gasparini, Z. B. de. 1977. Origenes y evolucion de los anfibios y reptiles del Ceno- zoico de America del Sur. Acta Geol. Lilloana 14:140-232. Berggren, W. A., Van Cowering, J. A. 1974. The Late Neogene. Palaeogeogr., Palaeoclimatol., Pa- laeoecol. 16:1-216. Bicarjella, J. J. 1964. Variacoes climaticas no Qua- ternario e suas implicacoes no revestimento floris- tico do Parana. Bol. Paranaense Geogr. Curitiba: 10-15. Bodenbender, G. 1924. El calchaqueno y los Estra- tos de la Puna de Penck. Bol. Acad. Nac. Cien. Cordoba 26:405-468. BONDESIO, P., LAZA, J. H., SCLLLATO YaNE, G., TON- ni, E., Vucetich, M. G. 1979. Estado actual del conocimiento de los vertcbrados de la Formacion Arroyo Chasico (Edad Chasiquense, Plioceno Temprano) de la Provincia de Buenos Aires, Re- publica Argentina. Act. II Congr. Argentino Paleontol. y Bioestr. I Congr. Latinoamer. Pa- leontol. (in press). Bossi, G. E. 1969. Geologia y Estratigrafia del sec- tor sur del valle de Choromoro. Parte I: Tecto. Acta Geol. Lilloana 10:17-64. Cabrera, A. L. 1971. Fitogeografia de la Republica Argentina. Bol. Soc. Argentina Bot. 14(1-2): 1-42. Calcaterba, A. 1972. Dos roedores fosiles nuevos para Uruguay y confirmation de otro. Comun. Paleontol. Mus. Hist. Nat. Montevideo 1:11-21. Caldenius, C. 1932. Las glaciaciones cuaternarias en la Patagonia y Tierra del Fuego. Publ. (Bue- nos Aires) Dir. Nac. Minas Geol. 95:1-150. Camacho, H. H. 1967. Las transgresiones del Cre- tacico Superior y Terciario de la Argentina. Rev. Asoc. Geol. Argentina 22:253-280. Caminos, R. 1972. Sierras Pampeanas de Tucuman, Catamarca, La Rioja y San Juan, pp. 41-79 in Leanza, A. L. (ed.). Geologia Regional Argen- tina. Acad. Nac. Ciena Cordoba, 869 p. Curry, R. P. 1966. Glaciation about 3,000,000 years ago in the Sierra Nevada. Science 154:770-771. Damuth, J. E., Fairbridge, R. W. 1970. Equatorial Atlantic deep-sea arkosic sands and Ice-Age arid- ity in tropical South America. Geol. Soc. Amer. Bull. 81:189-206. Eidt, R. C. 1969. The climatology of South America, pp. 54-81 in Fittkay, E. J., Illies, J., Klinge, H., Schwabe, G., Sioli, H. (eds.). Biogeography and ecology in South America. Vol. I, Junk, The Hague, 447 p. Fernandez, J. 1976. El limite boreal de los bosques andino-patagonicos. Bol. Soc. Argentina Bot. 17: 307-314. Feruglio, E. 1949. Description geologica de la Patagonia. Dir. Grab Yac. Petrol. Fiscales, Buenos Aires 2:1-349. Fidalgo, F. 1973. Consideraciones sobre los bajos situados al Norte de la provincia de Santa Cruz. Act. V Congr. Geol. Argentino 5:123-137. Fidalgo, F., Riggi, J. C. 1970. Consideraciones geomorficas y sedimentologicas sobre los Rodados Patagonicos. Rev. Asoc. Geol. Argentina 25:430- 443. Fleck, R. J., Mercer, J. H., Nairn, A. E. M., Peter- son, D. N. 1972. Chronology of Late Pliocene and Early Pleistocene glacial and magnetic events in Southern Argentina. Earth Planet. Sci. Lett., 16:1.5-22. Flint, R. F. 1965. The Plio-Pleistocene boundary, pp. 497-533 in International Studies on the Quaternary. Geol. Soc. Amer. Spec. Pap. 84. Flint, R. F., Fidalgo, F. 1963. Geologia glacial de la zona de borde entre los paralelos 39° 10' y 41°20' de latitud sur en la Cordillera de los Andes (Argentina). Bol. Dir. Nac. Geol. Min. (Buenos Aires) 93:1-35. Flint, R. F., Fidalgo, F. 1968. Drift glacial al Este de los Andes entre Bariloche y Esquel. Min. Ec. y Trab., Secret. Est. Energ. y Min., Inst. Nac. Geol. Min., Argentina, Bol. 119:1-19. Frenguelli, J. 1920. Contribution al conocimiento de la geologia de Entre Rios. Bol. Acad. Nac. Ciena Cordoba 24:55-256. Frenguelli, J. 1937. Investigaciones geologicas en la zona saltena del Valle de Santa Maria. Inst. Mus. Univ. Nac. La Plata. Obra Cincuentenario 2:215-572. Frenguelli, J. 1957a. Neozoico, pp. 1-115 in Geografia de la Republica Argentina. Soc. Arg. Est. Geogr. GAEA, 2, 218 p. Frenguelli, J. 1957b. El glaciarismo cuaternario, pp. 117-218 iit Geografia de la Republica Argen- tina. Ibid. 2, 218 p. Gasparini, Z. B. de, Baez, A. M. 1975. Aportes al conocimiento de la herpetofauna terciaria de la Argentina. Act. I Congr. Argentino Paleontol. Bioestr. 2:377-115. 1979 BAEZ & SCILLATO YANE: CENOZOIC OF ARGENTINA 155 Gordillo, C, Lencinas, A. 1972. Sierras Pam- peanas de Cordoba y San Luis, pp. 1-39 in Leanza, A. F. (cd.). Geologia Regional Argen- tina. Acad. Nac. Cienc. Cordoba, 869 p. Groeber, P. 1936. Oseilaciones de clima en la Ar- gentina desde el Plioceno. Holmbergia 1:71-84. Groeber, P. 1952. Glacial, Tardio y Postglacial en Patagonia. Rev. Mus. Municipal Cienc. Nat. Tra- ditional de Mar del Plata 1:79-103. Groeber, P. 1954. Duration de las glaciaciones cuartarias en America del Sud y discusion de las condiciones Astronomico-climaticas conducentes a englazamiento. Ibid. 1:141-159. Groot, J. J., Groot, C. R. 1964. Quaternary stratig- raphy of sediments of Argentine Basin. A paly- nological investigation. Trans. New York Acad. Sci. 26:881-886. Groot, J. J., Groot, C. R. 1966. Pollen spectra from deep-sea sediments as indicators of climatic changes in Southern South America. Mar. Geol. 4:525-537. Groot, J. J., Groot, C. R., Ewing, M., Burckle, L., Conolly, J. R. 1967. Spores, pollen diatoms and provenance of the Argentine Basin sediments, Prog. Oceanogr. 4: 179-217. Hays, J. D., Ondyke, N. D. 1967. Antarctic Radio- laria, magnetic reversals and climatic change. Science 158:1001-1011. Hays, J. D., Berggrex, W. A. 1971. Quaternary boundaries and correlations, pp. 669-691 in Fun- nel, B. M., Riedel, W. R. (eds.). Micropaleon- tology of the Oceans. Cambridge Univ. Press, Cambridge, 828 p. Heusser, C. J. 1974. Vegetation and climate of the Southern Chilean Lake district during and since the last Interglaciation. Quat. Res. ( New York ) 4:290-315. Hirschfeld, S. E. 1976. A new fossil anteater (Edentata, Mammalia) from Colombia, S. A. and evolution of the Yermilingua. J. Paleontol. 50: 419-432. Hirschfeld, S. E., Webb, S. D. 1968. Plio-pleisto- cene megalonychid sloths of North America. Bull. Florida State Mus., Biol. Sci. 12:213-296. Klein, R. M. 1975. Southern Brazilian phytogeo- graphic features and the probable influence of upper Quaternary climatic changes in the floristic distribution, pp. 67-88 in Bigarella, J. J., Beck- er, R. D. (eds.). Internat. Symp. Quaternary. Bol. Paranaense Geocien., Curitiba 33, 370 p. Kraglievich, J. L. 1952. El perfil geologico de Chapadmalal y Miramar, Provincia de Buenos Aires. Rev. Mus. Municipal Cienc. Nat. Tradi- tional de Mar del Plata 1:8-37. Kraglievich, J. L. 1953. La llanura bonaerense a traves de nn perfil geologico. Rev. Mundo Atomico (Buenos Aires) Ario 4, 14:88-95, 99. Kraglievich, L. 1934. La antiguedad pliocena de las faunas de Monte Hermoso y Chapadmalal, deducidas de su comparation con las que le pre- cedieron y sucedieron. Imprenta "El Siglo Ilus- trado," Montevideo, 1-136. Laugenie, C. 1971. Elementos de la cronologia glaciar en los Andes Chilenos meridionales. Cuad. Geogr. Sur 1:7-20. Marshall, L. G., Pascual, R. 1979. Una preliminar escala temporal radiometrica de las Edades-mam- ifero del Cenozoico medio y tardio sudamericano. Obra Centenario Museo La Plata (in press). Martin, P. S. 1975. Palaeolithic players on the American stage: Man's impact on the late Pleis- tocene megafauna, pp. 669-700 in Ives, J. D., Barry, R. G. (eds.). Arctic and Alpine Environ- ments. Methuen & Co., Ltd., 999 p. Menendez, C. A. 1962. Lefio petrificado de una leguminosa del Terciario de Tiopunco, Provincia de Tucuman. Ameghiniana, 2:121-126. Menendez, C. A. 1971. Floras terciarias de la Ar- gentina. Ibid. 8:357-368. Mercer, J. H. 1972. Chilean glacial chronology 20,000-11,000 C-14 years ago: Some global com- parisons. Science 176:1118-1120. Mercer, J. H. 1976. The last glaciation in Chile: A radiocarbon-dated chronology. Act. I Congr. Geol. Chileno. LD55-68. Mercer, J. H., Sander, W. 1975. Pliocene and early Pleistocene glaciations in southern Argentina, and the origin of the Patagonia gravel. VI Congr. Geol. Argentino, Resumenes. 20. Methol, E. J. 1967. Rasgos geomorfologicos de la Meseta de Somuncura, Rio Negro. Consideraci- ones acerca de los origenes de los "pequeiios bajos sin salida." Rev. Asoc. Geol. Argentina 22:295- 311. Parodi, L. 1930. Sobre nuevos restos de mamiferos de la fauna pampeana en Patagonia. Physis 10 (35):21-34. Pascual, R. 1970. Evolution de comunidades, cam- bios faunisticos e integraciones biocenoticas de los vertebrados cenozoicos de Argentina. Act. IV Congr. Latinoamer. Zool. 2:991-1088. Pascual, R., Fidalgo, F. 1972. The problem of the Plio-Pleistocene Boundary in Argentine (South America). Coll. Pap. Int. Colloq. on the Problem "The Boundary between Neogene and Quaternary" (Moscow) 2:205-262. Pascual, R., Ortega Hinojosa, E. I., Gondar, D., Tonnt, E. 1965. Las Edades del Cenozoico ma- malifero de la Argentina, con especial atencion a aquellas del territorio bonaerense. An. Com. Inv. Cient. Prov. Buenos Aires 6:165-193. Pascual, R., Odreman Rivas, O. 1971. Evolution de las comunidades de los vertebrados del Terci- ario argentino. Los aspectos paleozoogeograficos v paleoclimaticos relacionados. Ameghiniana 8: 372-412. Pascual, R., Odreman Rivas, O. E. 1973. Las uni- dades estratigraficas del Terciario portadoras de mamiferos. Su distribution y sus relaciones con los acontecimientos distroficos. Act. V Congr. Geol. Argentino 3:293-338. Patterson, B., Pascual, R. 1968. The fossil mam- mal fauna of South America. Q. Rev. Biol. 43: 409-451. 156 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Patterson-, B., Pascual, R. 1972. The Fossil Mam- mal Fauna of South America, pp. 247-309 in Keast, A., Erk, F., Glass, B. (eds.). Evolution, Mammals and Southern Continents. State Univ. New York Press, Albany, 543 p. Paulo Couto, C. de. 1975. Mamiferos fosseis do Quaternario do Sudeste Brasileiro, pp. 89-132 in Bigarella, J. J., Becker, R. D. (eds.). Internat. Symp. Quat. Bol. Paranaense Geocien., Curitiba 33, 370 p. Polanski, J. 1963. Estratigrafia, Neotectonica y Geomorfologia del Pleistoceno Pedemontano entre los rios Diamante y Mendoza (Provincia de Mendoza). Rev. Asoc. Geol. Argentina 17:127- 349. Polanski, J. 1965. The maximum glaciation in the Argentine Cordillera, pp. 453-472 in Wright, H. E., Jr., Frey, D. (eds.). International Studies on the Quaternary. Geol. Soc. America. Spec. Pap. 84. Ramos, V. 1970. Estratigrafia y estructura del Ter- ciario en la Sierra de los Colorados (Provincia de La Rioja, Republica Argentina). Rev. Asoc. Geol. Argentina 25:358-382. Reig, O. A. 1952. Sobre la presencia de mustelidos mefitinos en la Formacion Chapadmalal. Rev. Mus. Municipal de Cienc. Nat. Traditional Mar del Plata 1:45-51. Ringuelet, R. 1955. Panorama zoogeografico de la provincia de Buenos Aires. Notas Mus. La Plata 18 (Zool. 156):1-15. Ringuelet, R. 1961. Rasgos fundamentales de la zoogeografia de la Argentina. Physis 22:151-170. Rossi de Garcia, E. 1966. Contribution al conoci- miento de los Ostracodos de la Argentina. I. — Formacion Entre Rios de Victoria, Provincia de Entre Rios. Rev. Asoc. Geol. Argentina 21:194- 208. Rovereto, C. 1914. Los estratos araucanos y sus fosiles. An. Mus. Nac. Hist. Nat. Buenos Aires 25:1-250. Rutland, R. W. R., Guest, J. E., Crastley, R. L. 1965. Isotopic ages and Andean uplift. Nature 208 (5001): 677-678. Saxon, E. C. 1976. La prehistoria de Fuego-Pata- gonia: colonizacion de un habitat marginal. An. Inst. Patagonia 7:63-73. Scillato Yane, G. J. 1975. Nuevo genero de Dasy- podidae (Edentata, Xenarthra) del Plioceno de Catamarca (Argentina). Algunas consideraciones filogeneticas y zoogeograficas sobre los Euphrac- tini. Act. I Congr. Argentino Paleontol. y Bioestr. 2:449-461. Scillato Yane, G. J. 1977. Nuevo Megalonychidae (Edentata, Tardigrada) de Edad Chasiquense (Plioceno temprano) del sur de la Provincia de Buenos Aires (Argentina). Su importancia filo- genetica, biostratigrafica y paleobiogeografiea. Rev. Asoc. Cienc. Nat. Litoral (Santo Tome, Santa Fe, Argentina) 8:45-54. Scillato Yane, G. J. 1979. Nuevo Nothrotheriinae ( Edentata, Tartigrada ) de Edad Chasiquense ( Plioceno temprano ) del sur de la Provincia de Buenos Aires, Republica Argentina. Su impor- tancia bioestratigrafica, filogenetica y paleobio- geografica. Act. VII Congr. Geol. Argentino ( in press ) . Scillato Yane, G. J., Uliana, M. A., Pascual, R. 1976. Un Megalonychidae (Edentata, Pilosa) del Plioceno de la Provincia de Rio Negro (Argen- tina). Su importancia bioestratigrafica y paleo- biogeografica. Act. VI Congr. Geol. Argentino 1:579-591. Simpson, G. G. 1940. Review of the Mammal-bear- ing Tertiary of South America. Proc. Amer. Phil. Soc. 83:649-709. Solbric, O. T. 1976. The origin and floristic affini- ties of the South American temperate desert and semidesert regions, pp. 65-99 in Goodall, D. W. ( ed. ) . Evolution of the desert biota. Univ. Texas Press, Austin, 250 p. Tonni, E. 1974. Un nuevo cariamido (Aves, Grui- formes) del Plioceno Superior de la Provincia de Buenos Aires. Ameghiniana 11:366-372. Tricart, J. L. F. 1973. Geomorfologia de la Pampa deprimida. Base para los estudios edafologicos y agronomicos. Seer. Est. Agric. y Ganad. de la Nation, Inst. Nac. Tecnol. Agrop., Coleccion Cient., 12, Imprenta "CONI," Buenos Aires, 202 p. Volkheimer, W. 1971. Aspectos paleoclimaticos del Terciario argentino. Rev. Mus. Arg. Cienc. Nat. "B. Rivadavia," Paleontol. 1:243-262. Vuilleumier, B. S. 1971. Pleistocene changes in the fauna and flora of South America. Science 173:771-780. Webb, S. D. 1976. Mammalian faunal dynamics of the great American interchange. Paleobiology 2:220-234. 7. Quaternary Biogeography of the High Montane Regions of South America Beryl B. Simpson1 Department of Botany Smithsonian Institution Washington. D.C. 20560 USA From 11°N to 55°S, the South American Andes span 66 degrees of latitude, a distance of over S000 km (Fig. 7:1). At their widest point, they are about 500 km broad. Such an enormous land mass, extending across one hemisphere and projecting into the other, must necessarily exhibit great diversity. Else- where in South America, significantly high mountains are found only in southeastern Brasil and along the southern edge of the Guiana Shield (Fig. 7:1), but the area of these mountains above 2000 m elevation is very small. In several previous papers (Vuil- leumier, 1971; Simpson, 1973, 1975), I dis- cussed Pleistocene events in the Andes, but within the last few years, several studies have added to the knowledge of Andean Tertiary and Quaternary history. In this review of the Quaternary history of the high montane re- gions of South America, I first review the current information about the formation of the Andean Cordillera and the other montane regions. A background knowledge of the historical geology of the regions is necessary because their differing ages have an im- portant bearing on the composition and di- versity of the floras and faunas that were affected by Quaternary climatic changes. After outlining the historical geology, I briefly describe the modern climate and vegetation of these regions and then turn to a discussion of Pleistocene biogeographical events. GEOLOGICAL HISTORY OF THE MONTANE REGIONS The Andes As mentioned by numerous authors (e.g., 1 Present address: Department of Botany, University of Texas, Austin, Texas 78712, USA. Simpson, 1975), the "Andes" actually consist of a complex array of mountains that can be partitioned in several ways for purposes of discussion. It is meaningful to discriminate units of the Andes by partitioning the Cor- dillera into geomorphological units as done by Harrington (1956), Gansser (1973) and Simpson (1975) (Fig. 7:2). Although in this treatment, mountains often are referred to by the country in which they occur, the use of the country is merely for convenience as to lo- cality; the units being discussed are the geo- morphological units or their parts in the regions mentioned. A different and somewhat more detailed subdivision of the Cordillera into "tectonic segments" was proposed by Sillitoe (1974), Gansser (1973) and others. The boundaries of the major tectonic seg- ments correspond with the boundaries of the geomorphological units (Fig. 7:2). Perhaps the single most important aspect of the geological history of all the Andean units is their recency of uplift (see Simpson, 1975, Fig. 2). Although there is some dis- crepancy in the exact timing of final uplift, the last major upheaval of almost all units was at the end of the Tertiary or within the Quaternary. The final uplift of the Sierra Nevada de Santa Marta, the most northerly part of the Cordillera, has been dated toward the end of the Pleistocene (Gansser, 1955), although the use of late moraines as evi- dence can be questioned. Recent studies (van der Hammen, 1974; van der Hammen et al., 1973) have continued to document that the high elevation zones (above 2000 m) of the eastern Andes of Colombia were produced during the middle and later part of the Pliocene. Garner (1975) questioned the evi- dence for late uplift of the northern moun- tain ranges. The absence of traces of glacia- 157 158 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 tions before the last part of the Pleistocene, elusive evidence, because earlier moraines used as evidence for late uplift (Herd and can be easily eliminated. Few studies ex- Naeser, 1974), can not be considered as con- tending to the beginning of the Pleistocene 800 Kilometers 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 159 have been made, but careful examination of one important long pollen core (Fig. 7:3) clearly shows that plants characteristic of high elevations were absent in the Colombian Andes until the beginning of the Pleistocene (van der Hammen et al., 1973). Limited studies have been made in the Cordillera Occidental of Colombia, but the available evidence seems to suggest that they were ini- tially uplifted later than the ranges to the east (Shagam, 1975) and that they reached their present elevations after the end of the Tertiary (Burgl, 1961; Haffer, 1970a; Herd and Naeser, 1974). In Venezuela, geological investigations of the Merida Andes (Shagam, 1975) show several early orogenies with the last initiated in the Oligocene and continuing to the present. Significantly high elevations were achieved only relatively late in the Pleistocene (Schubert, 1974a). The portions of the eastern and western Cordilleras extending through Ecuador have received little attention in recent years, but a summary of previous studies in this region (Saner, 1971) indicated three stages of final uplift, the last of which occurred during the Pleistocene. This last uplift was postulated to have accounted for all elevations above 2000 m (Sauer, 1965, diagram 7). The youth- fulness of the high volcanic peaks in Ecuador is attested to, in part, by still active Volcan Cotopaxi. Earlier investigations of both the Eastern and Western Cordillera through Peru and Bolivia (Steinmann, 1930; Dollfus, 1959/1960; Ahlfeld, 1970) indicated Pliocene-Pleistocene final uplifts of about 1000 to 3000 m. More recently the orogenic history of the Andes has been interpreted in relation to plate tec- tonics (James, 1971, 1973; Gansser, 1973). James, in his interpretation of the uplift of the Bolivian Andes ( 1971 ) , envisioned an ini- Fig. 7:1. Distribution of high elevation areas in South America. Areas above 2000 m are shaded on the western side of the continent. For the southeastern Brasilian Highlands and the Guiana Tablelands, areas above 1000 are indicated. Inserts give climate diagrams showing mean monthly temperature and precipitation regimes at several high elevation stations. The changes in annual dispersion of rainfall and yearly patterns of temperature from north to south along the Andes as well as the aridity of the western part of the central Andes are evident. In the inset diagrams, mean monthly values of both temperature and precipitation are con- nected. Areas shaded with vertical bars indicate times of excess precipitation; stippled areas are times of mois- ture deficit. The left hand, vertical axis is calibrated in units of 10°C; the right hand vertical axis is calibrated in units of 20 mm of precipitation. The horizontal axis indicates the month of the year ( reversed in different hemispheres). Shaded areas along the base of the diagram show the months from which freezing temperatures have been recorded. For each of the stations with an inset, the following are given: name, latitude, longitude, altitude, number of years from which data were recorded, mean annual temperature and mean annual precipi- tation for this period. A. Bogota, Colombia, 4°38' N X 74 "05' W, 2556 m, 94 years, 13.°2C, 940.9 mm. B. Riobamba, Ecuador 00°22'S X 78°34'W, 3058 m, 8 years, 11.5°C, 1361 mm. C. Cajamarca, Peru, 07°08'S X 78°28'W, 2621 m, 9 years, 14°C, 716 mm. D. Cuzco, Peru, 13°33'S X 71°59'W, 3312 m, 17 years, 12.5°C, 750 mm. E. Arequipa, Peru, 16°19'S X 71°33'W, 2525 m, 37 years, 13.8°C, 104 mm. F. El Alto (airport, La Paz), Bolivia, 16°30'S X 68°12'W, 4105 m, 28 years, 7.5°C, 564 mm. G. Oruro, Bolivia, 17°58'S X 67°07'W, 3708 m, 10 years, 7.5°C, 282 mm. //. La Quiaca, Argentina, 22°06'S X 65°36'W, 3459 m, 8 years, 9.5°C, 322 mm. I. Cristo Redentor, Argentina, 32°50'S X 70°5'W, 3829 m, 27 years, -1.7°C, 354 mm. /. Itatiaia, Brasil, 22°20'S X 44°43'W, 2200 m, 20 years, 12.9°C, 2417 mm. Data for Figure insets A-I from Schwerdtfeger ( 1976) and inset J from Brade (1956). Climate data for a complete year are not available from the Guiana Highlands. Distribution de his regiones de altura de Sudamerica. En el lado oeste del contincnte las areas por encima de los 2000 m cstdn sombreadas. Para la region sudeste brasileho y los tablazos de la Cuayana, se indican las areas mayores de 1000 m. Las inclusiones representan diagramas climdticos, que muestran la temperatura media mensual y los regimenes de precipitation, para varias estaciones de altura. Los eambios en la dispersion anual de la lluvia y los patrones anuales de temperatura del norte a sur a lo largo de los Andes asi como la aridez del lado oeste de los Andes Centrales son evidentes. En los diagramas, los valorcs medios mensuales de la temperatura y precipitation cstdn conectados. Areas achuradas indican tiempo con exceso de precipitation; areas punteadas son tiempos con deficit de humedad. El eje vertical izquierdo estd calibrado en unidades de 10°C; el derecho estd calibrado en unidades de 20 mm de precipitaeion. El eje horizontal indica los rneses del aiio (iiwersos en difcrentcs hemisferios). La linea oscura en la base de los diagramas represcnta a los meses con registros de tempcraturas de congelation. A cada estation representada por una letra, le eorrcsponde su nombre, latitud, longitud, elevation, numero de ahos con registros, temperatura media anual y precipitation anual. Information para las inclusiones A-l proviene de Schwerdfeger (1976), para J de Brade (1956). Datos climdticos de tin aiio cotnpleto no cstdn disponiblc para las Alturas Guayanensis. 160 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 CARNEGIE RIDGE CHILE RIDGE SANTA MARTA EASTERN COROILLERA 1 WESTERN CORDILLERA ] CENTRAL COROILLERA CORDILLERA OCCIDENTAL CORDILLERA ORIENTAL PRINCIPAL COROILLERA PAMPEAN RANGES I COASTAL CORDILLERA 1 PATAGONIAN CORDILLERA "1 ALTIPLANO OR PUNA SURFACE 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 161 tial Cretaceous collision between the Nazca and the South American plates (Fig. 7:2) leading to the original outline of most of the Cordillera. He dated the final "crunch" and major upheaval as occurring at the end of the Tertiary. During this last period, the Altiplano was postulated to have been raised to its present elevation of 3000 and 4500 m. Although James' (1971) study was directed particularly at the Rolivian/Peruvian Andes, his conclusions probably extend to the ma- jority of the Andean Cordillera, because the contact between the Nazca and South Amer- ican plates extends from Ecuador to southern Chile. Gansser (1973) also discussed the Andean orogenies in terms of plate tectonics, but he included the areas of the Andes where the South American Plate is in contact with other plates (Fig. 7:2) and stressed the re- cent evidence that initial movements caused by contact of the plates were very early, probably Mesozoic, in some areas. In northwestern Argentina, the Pampean Ranges (Fig. 7:2) were raised (Simpson and Vervoorst, 1977) after final Quarternary up- lift of the Principal Cordillera to the west (Turner, 1972; Yrigoyen, 1972). Paralleling the Principal Cordillera from about latitude 14°S to about 44°S is the Coastal Cordillera (Fig. 7:2), which apparently was uplifted somewhat earlier than most of the Andes (Okada, 1971). Recent evidence (Cobbing et al., 1977; Dalmayrac et al., 1977) indicates that two small sections of this cordillera, one in Peru and one in Chile, are of Precambrian age. Cobbing et al. (1977) suggested that the Arequipa Massif in Peru is a rifted por- tion of the old South American shield areas. Plio-Pleistocene movements were primarily involved in altering preexisting structure (Dessanti, 1972). Dott et al. (1977) indicated that the up- lift of the Patagonian Cordillera of southern Chile was later than 77 to 81 million years ago and that most orogenic movements were completed by the Miocene. The Mountains of Southeastern Brasil In southeastern Brasil, the Serra do Mar, Serra da Mantiqueira, and associated ranges were formed by the arching and fracturing Fig. 7:2. Schematic drawing of the major geomorphological units of the Andean Cordillera (redrawn from Simpson, 1975) and the principal shields of the South American continent (from Putzer, 1968). The tectonic segments are drawn from Sillitoe (1974) and Gansser (1973). Numbers to the left of the dotted lines refer to boundaries of tectonic segments ( transverse faults that segment the descending lithosphere ) as numbered in Sillitoe (1974). Missing numbers were assigned to segments omitted here. Numbers included the following: 1. Amotape Zone, the northern limit of the Central Andes. 2. Huancabamba deflection, locality of the change in direction of the Andes from NW to NNE. 4. Pisco or Abancay deflection, a proposed divi- sion of the Central Andes; a sharp step marks the beginning of the Coastal Cordillera. 5. Northern limit of the Altiplano-Puno block. 6. A change in strike of the Andes from N to NW and the narrowing of the Eastern Cordillera (the Ichilo Fault or the Arica Elbow line). 10. The southern edge of the Puno block. 13. Boundary of the Norte Chico and Central Chilean regions, the northern limit of the Central Valley and the southern edge of the Precordillera and Sierra de Cordoba in Argentina. 16. Southern limit of the Cordillera Principal. In Southeastern Brasil letters refer to tectonic belts (from de Almeida, 1966). Dashed line is the Tropic of Capricorn. Dibujo esquemdtico de las unidades gcomorfologicas maijores de la Cordillera de los Andes (redibujado de Simpson, 1975) y de los principales escudos del continente Sudamericano (de Putzer, 1968). Los segmentos tectonicos estdn dibujados de Sillitoe (1974) y Ganscr (1973). Los numeros indican las fronteras de los seg- mentos tectonicos: 1. Zona de Amotape, el limitc nortc de los Andes Centrales. 2. Desviacion de Huancabam- ba, localidad del cambio de direccion de los Andes de NO a NNE. 4. Desviacion de Pisco o Abancay, una division de la Cordillera Central propuesta; un marcado paso sehala el cc-mienzo de la Cordillera de la Costa. 5. Limite norte del bloque Altiplano-Puno. 6. Cambio direccional de los Andes de N a NO y el cstrechamiento de la Cordillera Oriental (la falla de Ichilo o la linea de Codo de Arica). 10. Mdrgen sureno del bloque del Puno. 13. Separacion del Norte Chico y Region Central en Chile, limite norte del Valle Central y borde sur de la Precordillera y la Sierra de Cordoba en Argentina. 16. Limite sur de la Cordillera Principal. En el su- destc del Brasil, las letras indican cordones tectonicos (de de Almeida, 1966). La linea entrecortada senala al Tropico de Capricornio. 162 MONOGRAPH MUSEUM OF NATURAL HISTORY ELEMENTS NO. 7 OUERCUS ALNUS 500 000 Years Ago PLEISTOCENE PLIOCENE 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 163 of the ancient Brasilian Shield along north- east-southwest lines. Rubidium-Strontium (Rb/Sr) age determinations of rock samples from several areas in southeastern Brasil have yielded dates that range in age from 500 to 2,000+ million years (de Almeida et al., 1973). Thus, although differing some- what in age, all of these basement rocks are essentially Precambrian in age, and the Bra- silian platform must have consolidated fol- lowing an Precambrian orogenic cycle, pos- sibly about 1,800 million years ago. More restricted orogenies apparently occurred be- tween 900 and 1,300 million years ago. A final Precambrian orogeny that affected 40 per cent of the shield area (de Almeida et al., 1973) has been dated at about 450 to 700 million years ago. Because most of the Brasilian Shield above 1200 m is covered with Cretaceous sediments, uplift to appreciable elevations must have oc- curred primarily during the Cenozoic. De Freitas (1951) concluded that three periods of epeirogenic movements account for the uplift to 1200 m or higher. The first of these probably occurred at the end of the Creta- ceous and the last two in the Tertiary. The third may have extended into the Quaternary. It was during these last three phases of uplift that the Serras do Mar, Mantiqueira, Espin- haco and Borborema wer fully formed. The Guiana Highlands Like the mountains of southeastern Brasil, the basement of the Guiana Highlands, or Tepuis, appears to be very old. For many years, the age of the rock formations com- prising these abrupt table mountains emerg- ing from the Guiana Shield was disputed. Most South American geologists tended to favor an interpretation of very ancient initial uplift (e.g., Oliveira and Leonardos, 1943) but some geologists such as Gansser (1954) concluded that they were first raised more recently, perhaps in the Cretaceous. Part of the difficulty in interpreting the history of these mountains was the lack of fossils or geological stratigraphic sequences that could be correlated with those of known ages else- where on the continent. Since 1960, rock samples from a portion of the highlands in Guyana have been Rb/Sr dated as between 1,500 to 2,000 million years of age (Snelling, 1963). In southern Venezuela, Rb/Sr dates from several localities indicate rock ages of 1,700 to 2,100 million years (Hurley and Rand, 1973). The modern interpretation of the geological history of the Guiana High- lands (Lexico Estrat. Venezuela, 1970) is of a Precambrian basement about 3,000 to 3,100 million years old that was initially uplifted about 2,000 to 2,100 million years ago. Sub- sequent uplifts apparently occurred in the Fig. 7:3. Pollen core diagrams (redrawn from van der Hammen, 1974, Figs. 3 and 5) from the high plain of Bogota, Colombia. The base of the right hand core has been dated as older than 4 million years (Pliocene). The uppermost ( separated ) part of the core on the left is from Fuquene, Colombia. High elevation ( subpara- mo) elements first appear in the uppermost Pliocene. Appreciable amounts of pollen (indicating species that are abundant) of paramo plants are found only at the beginning of the Pleistocene. In the diagrams, percent- age of the total pollen recovered from a piece of a core that is represented by plants of a given vegetation type is recorded from left to right. Zero percent (absence) is on the left, and 100 percent would fill the box. The diagram also shows the times of first appearance in this area of the North American tree genera, Alnus and Quercus which migrated into South America after the closure of the Panama Portal. During the Pleistocene, interglacials are indicated by increases in the percentage of the total pollen that belongs to Andean forest genera. In glacial times, pollen of paramo plants dominate. Increases in the amounts of Polylepis indicate coo] but quite moist conditions. The dotted arrow shows the corresponding portions of two cores taken from the high plain. On the right core, letters are designations of disjunct core segments. Diagramas de cortes palinologicos (redibujados de van der Hammen, 1974, Figs. 3 y 5) de la Meseta de Bogota, Colombia. La base del corte de la derecha tiene una cdad mayor que 4 millones de aiios (Plioceno). La parte aha del corte de la izquierda es de Fequene, Colombia. Elementos de altura (subpdramo) aparecen recien en el Plioceno tardio. Considerables cantidades de polen (indicadores de las especies mas abundantes) de plantas del paramo se encuentran solo al comienzo del Plcistoccno. En los diagramas, el porcentaje del total de polen cstd registrado en la base de izquierda a derecha; cero por cicnto (ausencia) estaria a la izquierda, micntras que un 100 por cicnto llcnaria el diagrama. El diagrama tambien mucstra los tiempos de la primera aparicion de los generos de drbolcs norteamcricanos Alnus y Quercus los cuales migraron a Sudamerica luego de establccida la comunicacion en el 1st mo de Panama. Durante el Plcistoceno los inter glaciales estun indicados por incrementos en el porcentaje del total de polen perteneciente a los generos forcstales andinos. En tiempos glaciales, el polen de plantas del paramo domina. Los incrementos en la cantidad dc Polylepis indica condi- ciones frias pero humedas. La flecha entrecortada muestra las porciones correspondientcs de los cortes tornados en la sabana. En el corte de la derecha, las letras representan seementos de cortes disjuntos. 164 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Mesozoic, the Paleocene, and again later in the Tertiary. The recent history has been one of intense erosion and minor eustatic changes. Summary This cursory analysis of the geological history of the high mountains of South Amer- ica has pointed out that there are two very distinct sets of montane areas. All of the Andean Cordillera, extending the length of the western edge of the continent is very young; in fact, it was raised above sea level only after the end of the Cretaceous, and elevations above 2000 m were achieved only within the last 2 to 5 million years. The faunas and floras of these high elevations could not have migrated into high elevation habitats nor begun to differentiate until the later Pliocene or Pleistocene. Quaternary cli- matic changes did not, for the most part, modify a preexisting biota but were part of the actual story of the initiation and de- velopment of the high Andean flora and fauna. For the mountains of southeastern Brasil and the Guiana Highlands, which are composed of Precambrian/ Cambrian rocks and which were progressively uplifted throughout the Mesozoic and Cenozoic, the Pleistocene was merely the latest stage of a long developmental history. Prior to the Pleistocene, there was a well-developed and established biota on which the climatic changes of the Quaternary had the combined effects of augmentation, decimation and mod- ification. MODERN CLIMATE The Andean Cordillera, flanking the west- ern coast of South America, is strongly in- fluenced by its proximity to the Pacific Ocean. The highlands of southeastern Brasil, 10° to 20° longitude farther east, come under the influence of Atlantic Ocean pressure systems. In contrast to both, the Guiana Highlands, far from any ocean, exhibit continental cli- matic patterns. When all of the high montane areas of South America are considered, they encompass most of the climatic patterns pres- ent across the continent. Moreover, because of their elevations, they have added com- plexities due to orographic effects. A brief description of the major climatic regimes pro- vides a background against which Pleistocene climatic changes can be assessed. This out- line consists of climate diagrams (Fig. 7:1) and brief descriptions of the major tempera- ture and precipitation regimes. A much more complete description of South American cli- mates and references to local meterological data can be found in Schwerdtf eger ( 1976 ) . Trewartha ( 1961 ) still provides one of the most lucid accounts of the more complex climates. Tropical Andes Most of the Colombian Andes and the Merida Range of Venezuela lies within the Equatorial Trough (zone over which the Intertropical Convergence Zone travels dur- ing the course of a year) and receives rain- fall during two periods each year. The time of the heaviest precipitation is October to November; a second maximum occurs in April and May (Fig. 7:1A). Throughout this area, precipitation is greatest at intermediate ele- vations (500-1500 m) and decreases with elevation. Both the Caribbean and the Pa- cific slopes receive about 1500 to 2000 mm of rainfall annually. The very high precipita- tion ( up to 14,000 mm per year ) on the Pacific coastal lowlands decreases inland and with elevation. The Cordillera Central of Colom- bia has slightly higher amounts of annual rainfall (up to 2600 mm per year) than the eastern or western Cordilleras. The major inter-Andean valleys of Colombia ( Cauca and Magdelena) are dry at low elevations, but their slopes receive as much as 1000 to 3000 mm of rain (respectively) a year (Johnson, 1976). The depth of these valleys and the aridity of their floors have acted as barriers to east-west migrations of montane forest and paramo elements throughout the Pleistocene (Fig. 7:4). At high elevations ( >3500 m), the mean monthly temperatures are very constant over the course of a year. However, within one day, temperatures at high elevations within 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 165 the tropics (Fig. 7:5) can exhibit a range equal to that displayed over the course of a year in temperate lowland areas. Maximum and minimum daily temperatures can differ by more than 20°C with the lowest nocturnal values below freezing. The pattern of bimodal rainfall and con- stant mean monthly temperatures is found along the eastern slopes of the Andes from Colombia and Venezuela to southern Ecuador (Fig. 7:1A,B). Beginning in south-central Ecuador, the bimodal pattern shifts to one of unimodality. At high elevations and along the western slopes, a pronounced dry season, from April to August, develops (Fig. 7:1C- H). Along the eastern slopes, total rainfall remains above 900 mm, although it is more or less concentrated into one season. Thus, as one goes south along the Andes, there is a change in the dispersion pattern in the rainfall, a reduction in the total precipitation, and an increasing difference in the amount of moisture that falls on the eastern and western slopes. The discrepancy between the two sides of the Andes reaches its maximum across the southern Altiplano (Fig. 7:1E,F). On the western slopes of southern Peru and northern Chile, the climate is extremely arid and produces a barren desert. At the same latitude on the eastern slopes of southern Bolivia, enough precipitation falls to support upper montane cloud forest. The severe arid- ity of the western slopes is caused by a com- bination of several factors— 1) the presence of a cold upwelling along the coast, 2) the flow of a cold oceanic current parallel to the coast, 3 ) the production of a rainshadow by the Eastern Cordillera, and 4) the strong Pacific anticyclone that lies off the coast of Peru. During some years, the Pacific anti- cyclone, the force of the Humboldt current and the amount of upwelling are lessened. At such times, the Intertropical Convergence Zone extends farther south than in normal years, thereby resulting in relatively heavy rains on the coast and coastal mountains. The sporadic occurrence of this phenomenon, known as "El Nino," has led some researchers to postulate that a similar climatic pattern was prevalent during glacial periods (see section on Theories of Glacial Climatology). Temperate Andes As one goes south of the Tropic of Capri- corn, major changes in climate are evident. Along the Pacific coast, the Pacific anticyclone splits and the westerlies begin to exert an influence. Rainfall remains sparse between the latitudes of 27° and 31 °S, but 90 percent of the moisture falls in the winter. At lati- tudes south of 31°S, total precipitation in- creases as the summer months also begin to receive rainfall. At the latitude of Valdivia (40°S), the annual rainfall is as high as 3000 mm (5000 mm on the coast). The amount of summer precipitation increases to western Tierra del Fuego, although the total amount per year is somewhat less than slightly farther north. In the southernmost regions, total pre- cipitation decreases, but it falls almost con- stantly during the year, with 24 to 28 days of each month receiving some form of pre- cipitation (Miller, 1976). On the eastern side of the Andes south of the tropic (actually south of 29°S), the Andean slopes become dry. The moisture- laden winds of the tropical Atlantic reach only to about 29°S. In addition, the eastern slopes are the rain shadow flank of the Andes and are too far inland to receive any oceanic influences. As in the case of the western slopes at about the same latitude, rainfall be- comes briefly bimodal and then shifts to a pattern of predominantly winter rainfall near the Rio Colorado and Rio Negro (37°S) (Prohaska, 1976). Precipitation increases southward. South of 37°S, the Andean slopes intercept sufficient moisture to support de- ciduous beech forest. From 27 °S to the southern tip of the continent, the variation in the annual march of temperature becomes more and more pro- nounced, and yearly effects overshadow any diurnal temperature differences. In southern areas of the continent, the growing season is restricted to a few months of the year. Eastern Tropical Mountains Along the coast of southeastern Brasil, 70 to 80 percent of the rain falls between No- vember and April with July and August being the driest part of the year. Total precipitation 166 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 167 is quite high (2000 mm) and is thought to — 6°C have been recorded. Annual rainfall be attributable to the buildup of the sub- at this elevation exceeds 2400 mm (Brade, tropical anticyclone over the "hump" of Bra- 1956; Fig. 7:1J). Highlands farther inland sil (Trewartha, 1961). At elevations above from the coast are influenced primarily by 2200 m on Itatiaia, mean monthly tempera- the flow of unstable equatorial air that an- ture differences during a year are between nually moves across central Brasil. Here, 15°C and 27°C. Minimum temperatures of rainfall is lower than along the coast, and Fig. 7:4. Barriers to population expansions of high elevation taxa during either glacial or interglacial periods. 1. Lowland area of northeastern Colombia separating Sierra Nevada from the Eastern Cordillera of Colombia. 2. Low arid area separating the Eastern Cordillera of Colombia and the Merida Andes of Vene- zuela. 3. Rio Magdalena Valley separating the Cordillera Oriental from the Cordillera Central of Colombia. 4. Rio Cauca Valley separating the Cordilleras Occidental and Central of Colombia. 5. Northern Peru low area. In all of the cases of barriers 1-5, east-west or north-south dispersal of high elevation elements is now pre- vented by areas of unsuitable low elevation habitat. During glacial periods, the effects of these barriers would have been less severe. 6. Upper Maranon Valley, an arid valley that was a glacial and interglacial barrier preventing east-west exchange of elements, probably more effective in integlacial than in glacial times. 7. Eastern side of Lake Titicaca where glacial ice flowed into the lake and prevented north-south biotic ex- change. This area is not a barrier at the present time. 8. Clacial barriers for arid elements formed by lakes and bogs across the surface of the Altiplano during cold periods. 9. Zone of continuous aridity across the Cor- dillera which has served as a barrier to north-south migration during glacial and interglacial times. 10. Rio Bio-Bio. A modern and glacial barrier to north-south dispersal because of the climatic change at this lati- tude, the presence of the river itself and the glacial ice which followed the course of the river. Asterisks indi- cate habitats isolated on mountain peaks during interglacial periods (paramos in the north and alpine habitats in the south). During glacial times, these habitats expanded, facilitating exchange. The letter A refers to the area of lowland rainforest and llanos separating the Andes from the Guiana Highlands and preventing east-west colonization. In glacial times, stepping stones of subtropical habitat may have been present in this region. Let- ter B indicates thorn scrub ( monte/chaco ) separating the Andes from the highlands of southeastern Brasil. A modern barrier, presumably of narrower extent in glacial times. Letter C designates the Amazon lowlands which separate the Guiana and the southeastern Brasil highlands. An effective filter zone in both glacial and inter- glacial times but with stepping stones possible in glacial periods formed on the low tablelands of central Brasil. The triangles indicate arid barriers between relic woodlands on the western slopes of the Peruvian Andes. In humid periods, presumably during glacial times, these arid areas received sufficient moisture to allow continu- ous forest growth. Ban-eras para la expansion de las poblacioncs de biota de altura durante los periodos glaciates o intcrglacialcs. 1. Area de tierras bajas del noreste de Colombia separando la Sierra Nevada de la Cordillera Oriental de Co- lombia. 2. Area drida ij baja separando la Cordillera Oriental de Colombia y los Andes de Merida de Vene- zuela. 3. Valle del Rio Magdalena separando la Cordillera Oriental de la Cordillera Central de Colombia. 4. Valle del Rio Cauca separando las cordilteras Occidental y Central de Colombia. 5. Areas bajas del norte del Peru. En todos los casos de las barreras 1—5, la dispersion este-oeste o nortc-sur de los clcmentos de alturas estd prevenido por areas de poca clcvacion. Durante los periodos glaciates, los efectos de estas barreras hab- rian sido menos severos. 6. Valle del Alto Maranon, un valle drido en cual durante la epoca glacial c intergla- cial fue barrcra la cual evito el intercambio de etementos de este-oeste; probablemente la nuis effectiva de las dos fue la epoca interglacial. 7. Lado oriental del Lago Titicaca donde el hiclo glacial flujo y previno el inter- cambio biotico del norte a sur. Esta area no es una barrcra al tiempo presente. 8. Barreras glaciates para etementos dridos fonnados por lagos y pantanos en el Altiplano durante periodos frios. 9. Zona de aridez continua a lo largo de la cordillcra que ha servido de barrcra a la migracion norte-sur durante los tiempos glaciates e intcrglacialcs. 10. Rio Bio-Bio, una barrcra moderna y glacial para la dispersion norte-sur a causa del cambio climdtico a esta latitude, la presencia del rio mismo y del hiclo glacial que siguio el curso del rio. Los astcriscos indican ambicntes aislados sobrc los topes de montanas durante los periodos intcrglacialcs (paramos en el norte y ambientes alpuws en el sur). Durante el tiempo glacial estos ambientes se expandieron facilitando el intercambio. La letra A reficre a las areas de sclva pluvial de tierras bajas y de llanos separando los Andes de las Alturas Guayanesas y previmiendo la colonizacion este-oeste. En tiempos glaciates, rutas de migracioncs cntrccortadas de ambicntes subtropieales pucden habcr estado presente en esta region. La letra B indica un matorral espinoso (monte/chaco) separando los Andes de las alturas del sudeste del Brasil. Una barrcra mod- erna, probablemente de extension menor en tiempos glaciates. La letra C designa las tierras bajas amazonicas que separan la Guayana y las alturas del sudeste brasileno. Un filtro efjectivo en los tiempos glaciates e intcr- glacialcs pero con rutas de migracioncs entrecortadas posiblcmente en periodos glaciates formadas en las mesctas bajas del Brasil central. Los triangidos indican las barreras dridas cntre selvas relictuales sobrc las ladcras oc- cidentates de los Andes peruanos. En periodos humedos, probablemente durante los periodos glaciates, estas areas humedas recibidn suficicnte humedad como para permitir el crecimiento forestal continuo. 168 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 OXFORD, ENGLAND 5 "N, 'W (63 Ml F MAMJJA SOND ORURO, BOLIVIA 17°S, 67°W (3703 M) S O N D J F M M J Fie. 7:5. Comparison of diurnal and annual tem- perature regimes in a lowland temperate and high elevation tropical area using thermoisopleths. In these diagrams, hourly mean temperatures are recorded (vertical axis) throughout the year (horizontal axis) and similar values are connected with lines. These diagrams have been somewhat simplified to make the patterns more apparent. The total temperature range of the two localities is about equal, 16°C, but at Oxford, England, the greatest temperature differences occur between summer and winter. At Oruro, Bo- livia, the greatest differences between mean maximum and mean minimum temperatures occur within one day (during the summer). Upper figure from Troll, 1965, Fig. 1; lower figure from Troll, 1959, Fig. 4. Comparacion tie regimenes de temperaturas di- urnas y anuales en tierras hajas templadas y en tierras altas tropicalcs usando termoisoyetas. En estos dia- gramas las temperaturas medias par hora esttin regis- tradas (eje vertical) a troves del aho (eje horizontal) y lof> valores similares estdn conectados por lima.',. Los dibujos han sido simplificados para destacar mds los patrones. El rango total de temperatura es cast the majority falls in spring and summer (September-March) with the fall and winter (April-August) being the drier seasons. The tropical precipitation pattern is combined with a more temperate pattern of tempera- ture at high elevations inland. Interior Tropical Mountains On the summits of the Guiana Highlands, rainfall is higher than on the surrounding lowlands. The eastern group of these table- lands receives about 2000 to 2500 mm per year and the western portion about 3300 to 3500 mm. At elevations above 1000 m, rain falls almost continuously throughout the year with slight reductions in March and Septem- ber (Maguire, 1970; Snow, 1976). Because this area is so far removed from any oceanic influences, it is probable that the moisture has been derived, in part, from reevaporation or from evapotransporation of precipitation that previously fell as rain over the Amazon Basin (Snow, 1976). Temperatures are cool throughout the year, often dipping to 5°C at night (Maguire, 1970). MODERN VEGETATION OF THE HIGH ELEVATION MOUNTAINS OF SOUTH AMERICA In order better to understand how Pleisto- cene changes in the climate might have af- fected the vegetation of the highlands of South America, it is necessary to have some idea of the modern dominant vegetation types in the Andes and the eastern mountain ranges. In general, I confine my remarks to the vege- tation above 2000 m. Naturally, such a con- densed account must be rather general and somewhat superficial. The principal objec- tive is to present an idea of where major breaks, both altitudinal and latitudinal, occur in the vegetation. In some cases, only a few characteristic plant genera are mentioned; in others, species are given. In the latter there iqual — 16°C — en ambas loealidades pew en Oxford, Inglaterra, las mas grandes diferendas se dan entre oerano e inoierno, En Oruro, Bolivia, las mayores diferendas entre la media de las mdximas y la media de las minimus suceden en un dia (durante el oerano). La figura superior estd tornado de la figura 1 de Troll, 1965; la figura inferior de la figura 4 de Troll, 1959. 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 169 are often changes in species of characteristic genera at key biogeographical points. In each case, a description of the physiognomy of the vegetation is given so as to present a visual aspect of the habitats. More detailed treatments can be found in Hueck ( 1966 ) , Hueck and Seibert ( 1972 ) and Cabrera and Willink (1973). Recause so much detail is lost in maps of large scale, I have not at- tempted to include a vegetation map. Hueck and Seibert (1972) produced the most com- prehensive continent-wide vegetation map. The Eastern Andes from Colombia and Venezuela to Northern Argentina All of the eastern slopes of the Andes not hidden in valleys or obscured by a higher, more easterly ranges, are covered by upper montane forests. In some descriptions of the vegetation (Cabrera and Willink, 1973) this entire region is described simply as "humid montane forest" or "ceja." Yet, changes in the dominant taxa, if not in the physiognomy, do occur. In general, throughout the span (10°N-27°S), there is tall, humid, evergreen upper montane forest up to about 3500 m. Above the forest is a band of humid low scrub dominated by members of the Erica- ceae (Gaultheria), Escalloniaceae (Escal- lonia), and Melastomataceae (Miconia). This scrub, composed of shrubs about 0.5 to 2 m tall, gives way to various supraforest vegeta- tion types, described below, as one proceeds north to south. Although the upper montane forest is dominated throughout its length by members of the genera Podocarpus (Podo- carpaceae), Weimnannia (Cunoniaceae), Drimys ( Winteraceae ) and various members of the Guttiferae such as Clusia, notable breaks do occur, and smaller, cohesive units can be distinguished. The first, in Colombia and Venezuela (10°N-1°N) spans altitudes of 2400 to 3800 m and locally ascends to 4200 m. The species of Weinmannia that dominate in Venezuela are W. jahnii and W. microphylla, along with Podocarpus olei- folius, P. montanus, and P. rospiglossii. Spe- cies of Oreopanax ( Araliaceae ) , Ilex (Aqui- foliaceae), and Brunellia (Rrunelliaceae) are commonly mixed with other characteristic elements. From northern Ecuador to central Bolivia (1°N-16°S), the montane forest has an upper limit at a somewhat lower elevation, usually only to 3600 m, and can be consid- ered the true "ceja." The dominant Podo- carpus is P. nubigenus and the predominant Weimnannia, W. fagaroides. Oreopanax, Al- nus (Juglandaceae) and Clusia (Guttiferae) are also very common. South of Santa Cruz, Bolivia, and continuing into northern Argen- tina (16°S-27°S) there is a noticeable differ- ence in the vegetation. At lower elevations, Podocarpus (P. parleteoreii) , Eugenia (Myr- taceae) and Weinmannia remain common along with Ilex. However, at higher eleva- tions, the forest becomes a semideciduous woodland with Juglans (}. australis, Juglanda- ceae), Pohjlepis australis (Rosaceae) and Sambucus (S. peruviana, Caprifoliaceae), mixed with the evergreen trees and Alnus jorullensis. The abrupt southern termination of the eastern montane forest coincides with a major climatic break that drastically reduces the rainfall on the eastern slopes (Fig. 7:1). The High Elevation Habitats of Colombia, Venezuela and Parts of Ecuador and Pern Above tree line (3500-4700 m) in the northern Andes, a low, humid, herbaceous vegetation covers the high elevation areas known as paramo ( Cuatrecasas, 1968). Scat- tered through this zone are characteristic "rosette" plants, primarily of the genus Espe- letia (Compositae). Most of the vegetation is about 0.5 to 1 m tall with the rosette frailejones (Espeletia) emergent and up to 7 m tall. In areas of suitable microhabitat, open woodlands of Pohjlepis (P. sericea and P. cocuyensis) occur. The dominant groups of plants are grasses (particularly Calama- grostis and Sicallenochloa) and Compositae (Espeletia, Diplostephium, Gynoxys, Lori- caria, Aster, Baccharis, Senecio). Several common genera such as Lupinus ( Legumino- sae), Geranium (Geraniaceae), and Hyperi- cum ( Hypericaceae ) are familiar to temper- ate botanists. In wet paramos, the ground is often spongy with accumulated sphagnum and plant debris, and the landscape is usually shrouded in mist, if not drizzle. Analyses of the vegetation have shown that the dominant elements have not been derived by point by point vertical differentiation of the upper montane flora (Simpson, 1975). Rather, the 170 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 differentiation of the flora has primarily been horizontal within the paramo zone, although several of the characteristic genera were orig- inally derived from Neotropical stocks that grew at lower elevations. Many genera and species are restricted to one or two paramos (e.g., Espeletia, Dipholostephium, and Lori- caria) or have undergone immense radiations within this regions. Other elements belong to genera common in high elevations through- out the Andes (Senecio, Baccharis, Festuca, Mutisia, Polylepis) or to those which migrated from North or Central America after the clos- ing of the Panamanian portal (Lapinus). The High Elevations of Ecuador and Peru and Easternmost Bolivia There has been argument for years as to whether the area of Ecuador and Peru with low, humid, herbaceous vegetation dominated by grasses and Compositae should or should not be called paramo, but several authors (Cuatrecasas, 1968), consider the paramo zone to extend from Costa Rica to Peru. The most conspicuous difference between the areas south of northern Ecuador and those to the north is the absence of Espeletia to the south. Weberbauer ( 1945 ) tended to call humid high elevation areas in Peru "jalca." However, in Ecuador the term "paramo" often is used as the common name for supra- forest grasslands. Many of the same grass genera found in Colombia are dominant there, and the typical high elevation Ranuncula- ceae, Gentianaceae, Geraniaceae, legumes and Compositae are common. Beginning in southwestern Ecuador and increasing in breadth across Bolivia to Argen- tina (4 or 5°S-27°S) the supraforest vegeta- tion becomes (sporadically) drier and usual- ly is called "puna." As in the case of the paramo, grasses and members of the Com- positae dominate the vegetation. However, the most abundant grasses here are Poa (P. humilis), Stipa (S. ichu) and Festuca. The puna extends from elevations of about 3400 m to 4400 m or even higher. Dominant Com- positae are Werneria, Chuquiraga, Baccha- ris, Senecio and LepidophijUum. The aspect of the vegetation is more coarse than that of the paramo. Grasses tend to be clumped; shrubs have many xerophytic characteristics. Tightly compacted mat plants, such as Azor- ella (Umbelliferae), Pycnophyllum (Caryo- phyllaceae) and Werneria (Compositae), are common. Along the eastern side of Altiplano, the vegetation is often termed "wet puna" (Troll, 1959), because it is more lush than that in the west. In the western portions, cacti, often ground-hugging, and various Bromeliaceae (Puya, TiUandsia) are locally abundant. Trees (Polylepis) are found in humid microsites up to elevations of 5200 m, but individuals are widely spaced, small, and very gnarled with reduced leaves. The southwestern altiplano of Bolivia and adjacent regions of northern Argentina and Chile, are exceedingly dry (see Modern Climate ) ; even grasses are sparse. The domi- nant plants are scattered Compositae (Para- stephia, LepidophijUum and Nordophyllum) . Shrubs (1-1.5 m tall) of these genera, spaced up to 6 m apart, grow with small twisted shrubs of Adesmia horrida (Leguminosae). Cacti are sometimes found, and in southern Peru, bromeliads are locally dominant. In intermontane valleys at lower elevations (3400 m-1000 m) in southern Bolivia and Argentina, columnar cacti (Trichocereus terscheckii, T. pascana) and scrubby trees and shrubs ( Prosopis ferox, Zuccagnia, Legu- minosae; Gochnatia and Baccharis; Composi- tae) are common. This semiarid scrub vege- tation becomes dominant (with different spe- cies of Prosopis) along the eastern slopes of the Argentine Andes south of the ceja forest (27°S) until the latitude (36°S) where the deciduous Nothofagus forest begins to appear. The West Slopes of the Andes of Colombia Because of peculiar climatic events (Tre- wartha, 1961), the westernmost slopes of the Colombia Andes (extending into northern Ecuador) are covered by extremely humid montane forest, which extends narrowly in- land and to about 1500 m elevation. Domi- nant elements include Annonaceae, Legum- inosae, Moraceae, and palms. Above this zone there is upper montane forest merging into an ericaceous zone and, finally, paramo simi- lar to that found on the other northern ranges. 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 171 The Western Slopes of the Andes from Ecuador to Central Chile Along the west coast from south-central Ecuador to northern Chile (2°S-21°S), the vegetation becomes increasingly xerophytic and sparse. In southern Ecuador into Peru to about the latitude of Lima (12°S). the western slopes are covered with thorn scrub and cacti. Legumes such as Prosopis and Acacia, Bromeliaceae (Puya), Cactaceae (Trichoccreus), and Compositae such as Proustia, Franscria and Diplostephium, are dominant ( Weberbauer, 1945; Tosi, 1960). This scrub extends to, and merges with, the dry puna at elevations of about 3000 m. In southern Peru, the vegetation becomes more and more sparse until, west of Arequipa, only scattered bromeliads and a few cacti are visi- ble. In northern Chile, vegetation of any prominence is virtually absent. When the infrequent rains occur, ephemerals appear, but such periods of flowering can be decades apart. Once the latitude of about 21° to 27°S is reached, vegetation reappears along the slopes and southward becomes increasingly dense. At first, thorn scrub appears, primarily around areas where streams descend from the mountains. This scrub extends elevationally from 300 m to about 3000 m and merges with the southern extension of the puna. The height of the vegetation is about 2 to 3 m and is quite open until it merges into the true evergreen "matorral" below. The dominant genera include Adesmia, Acacia and Prosopis (Leguminosae), Cactaceae {Trichoccreus, Eulychnia) and Bromeliaceae (Puya). South of 36°, the deciduous southern beech forests begin along the Pacific slopes with the appearance of Nothofagus obliqua and N. procera (Hueck, 1966). These same two taxa are the most common Nothofagus on the Argentine side of the Andes. South- ward between 38 and 40°S, the forest be- comes more lush. At the latitude of Valdivia (40°S) the species of Nothofagus (predomi- nantly N. betuloides and IV. dombeiji) are evergreen and the association is consequently called the evergreen rainforest or the Val- divian Forest. In addition to the evergreen species of Nothofagus I>etuloides and Podo- carpus again becomes a dominant element (P. nubigenus) as well as Drimys winteri. Coni- fers such as Fitzroya cupressoides, and Arau- caria become conspicuous. Along with these stately trees are laurels, including Laurclia and Mijrceugenella (Myrtaceae). The forest becomes diminished in species and lower in stature as the Andes descend and the climate becomes harsher from 40°S to the tip of Tierra del Fuego. The beeches, Nothofagus pumilio and N. antarctica, although reduced in size, continue to Tierra del Fuego. The High Southern Andes In the southernmost Andes (26°S to about 51°S) the area above tree line and below the level of the glaciers is covered by humid meadows similar in aspect to those of the Alps. As would be expected, rosette herbs are abundant, especially members of the Compositae (Perezia, Leucheria and Nassau- via). Genera of families typical of the north temperate zone are common — Ranunculus (Ranunculaceae), Cardamine ( Crucif erae ) , Epilobium ( Onagraceae ) , Primula (Primula- ceae), Pinguicula (Lentibulariaceae). In contrast to the paramos, these plants are covered by snow for much of the year and flower and fruit during late southern spring and summer (January to March). The Highlands of Southern Brasil In the extreme southern part of Brasil, the highlands (600-1800 m) are covered with an open woodland with scattered trees of genera common in the Andes — Araucaria (A. angus- tifolia), Podocarpus (P. lambertii), Drimys (D. brasiliensis) and Ilex (I. paraguariensis) . Grasses and various perennial herbs form the understory. Farther north, along the ex- treme coastal area of southeastern Brasil, at elevations of 200 to 1800 m, is a band of ever- green humid forest dominated in the northern part by characteristic tropical legumes trees such as Caesalpinia echinata, Apuleia fcrra, Piptadenia peregrina, Parkia pcndula, Ma- chacrium and Cecropia ( Urticaceae), Teco- ma (Bignoniaceae), Geonoma, and other palms (Euterpe, Cocus). In the more south- ern sector, Eugenia (Myrtaceae), Roupala 172 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 ( Proteaceae ) , Tibouchina and Miconia (Mc- lastomataceae), various members of the Mal- pighiaceae, Cunoniaceae (Weinmannia) and numerous Compositae are dominant. On the uplands of the mountains ( > 1800 m ) , grasses dominate, especially Rambusoideae (Chusquea penifolia and Cortaderia modes- ta). Numerous shrubs, among them species of Escallonia and Compositae (e.g., Senecio) are common. There is much endemism, but some herbaceous elements, such as Perezia, Dasijphijllum and Trichocline (Compositae), are shared with the Andes. The strange mem- bers of the Vochysiaceae and Velloziaceae, often cited as characteristic, are usually found at lower elevations and belong, properly speaking, to the Cerrado vegetation ("campos rupestres" ) . The Guiana Highlands The Guiana Highlands are the least bio- logically explored of any of the South Amer- ican montane regions. Ry 1970, over 2,000 plant species had been described from the few explored tepuis. Maguire (1970) esti- mated that at least 4,000 taxa eventually would be found. In contrast to the other regions described, endemism here is extreme- ly high — perhaps as much as half of the flora. Certain families such as the Rapateaceae are almost confined to this area and have many genera with species restricted to the various mountain tops. Among the other families with especially large numbers of endemic genera and species are the Melastomaceae and the Myrtaceae. Many primitive and endemic Compositae also are found in this region. Most of these, Stenopadus, Glossarion, Neblina, Chimantea, Quelchia, Achropogon and Duidea belong to the tribe Mutisieae and are considered by some (Carlquist, 1957, 1974) to be among the most primitive members of the family. The relationships of the flora of these high- lands are obscure, because of the high per- centage of endemic genera and families. There are few taxa closely related to groups found in the high Andes and only slightly more show floristic similarities with elements in the surrounding lowlands and mountains of southeastern Rrasil (Brade, 1956; Maguire, 1970; Carlquist, 1974). QUATERNARY OF THE HIGH MONTANE AREAS OF SOUTH AMERICA Within the last few years, several studies have supplied new data that have altered or refined previous conclusions about the num- ber, extent, and duration of glacial periods in South America and the moisture regimes dur- ing cold or warm periods. The most signifi- cant data come from the Colombian Andes and from the lake region of southern South America. In these and in a few additional areas, modern dating techniques combined with careful geological studies have contrib- uted to the modification of many earlier ideas and to the resolution of some old problems. As one might expect from the discussion of the climate and vegetation of the different mountainous areas throughout the continent, their Pleistocene histories differed consider- ably depending on elevation, latitude, longi- tude and exposure. In the most general terms, we can say that within the Pleistocene, and during times of world wide sea-level lowering (glacial periods), cooler conditions existed with or without, an absolute increase in moisture. In times of high, world-wide sea levels (interglacials), there was a warming of climate sometimes accompanied by in- creased precipitation. The Northern Andes As a result of 20 years of careful palyno- logical and geological work of van der Ham- men and his colleagues, a firmly dated and fairly complete sequence of changes over the last 4 million years is available for the East- ern Andes of Colombia (van der Hammen, 1974, and references therein). The sequences shown in a representative pollen core (Fig. 7:3) indicate several things. First, high ele- vation, cool-climate-adapted taxa appear only at the very end of the Pliocene. The earliest of these elements includes Myrica sp. (Myri- caceae). As the plain around Bogota, from which these cores were made, became pro- gressively uplifted (van der Hammen, 1974; van der Hammen et al., 1973), more and more plant taxa characteristic of high elevations appeared in the fossil record [e.g., Lycopodi- urn (Lycopodiaceae); Gunnera (Haloragida- 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 173 ceae), Geranium (Geraniaceae), GeiUiana ( Gentianaceae ) and Lysipomia (Campa- nulaceae)]. Later (Fig. 7:3), woody ele- ments from North America such as Alnus (Betulaceae at > 500,000 years b.p.) and still later Quercus (Fagaceae at > 250,000 years b.p. ) appeared. The arrival of these montane forest trees indicates the southward migration of Northern Hemisphere stocks after the clos- ing of the Panamanian Portal at the end of the Pliocene ( ca. 5.7 million years ago, Raven and Axelrod, 1974). A characteristic element of the high Andes that appears at more than 600,000 years b.p. is the genus Polylepis. This genus, a wind-pollinated member of the Rosaceae, had a complex history in Colombia that has been useful in interpreting changes in the climatic regime, because the northern species of this genus are restricted to cool and moist habitats (Fig. 7:3). By the end of the Pleistocene, the present flora had been estab- lished with 1 ) a mixture of species derived from North American genera that migrated southward (Berberis, Berberidaceae; Draba, Cruciferae; Hypericum, Hypericaceae; Genti- ana, Gentianaceae; Bartsia, Scrophulariaceae; Valeriana, Valerianaceae); 2) south temper- ate taxa that migrated northward along the Andes as the increasing elevations provided progressively cool habitats (Muehlenbergia, Gramineae; Acaena, Rosaceae; Azorella, Um- belliferae); and 3) a large number of en- demic species derived by speciation within the high elevation zone from original Neo- tropical stocks (Puya, Bromeliaceae; Bhizo- cephahim, Campanulaceae; and Espeletia, Diplostephium, Loricaria, Compositae; see Simpson, 1975). As an analysis by Simpson (1975) and discussions by van der Hammen ( 1974 ) have indicated, this complex flora was produced, in large part, by Quaternary cli- matic changes. Although moraines do not provide evidence for glaciation in the region of the Sabana de Bogota before the penulti- mate glaciation about 250,000 years b.p., van der Hammen (1974) interpreted the presence of cold-adapted species in the pollen cores at earlier dates as evidence of glaciations elsewhere. The ultimate glaciation began about 130,000 years b.p. The pollen-core composition at this time indicates initial cool and wet conditions followed by cold dry cli- mates (Fig. 7:3). Woodlands of Polylepis covered the area about 30,000 years b.p., when conditions were cold, and not exces- sively dry. The driest period seems to have been about 20,000 years b.p. At that time, traces of Polylepis disappeared altogether. During the very cold dry periods, vegetation zones were lowered about 1200 to 1500 m, and the temperatures probably were about 6 to 7°C cooler than at present. Following this last major glaciation, there were minor advances and retreats; a notable advance oc- curred about 14,000 years b.p. Since that time, the climate has been cool and moist. During glacial periods, when vegetation zones were lowered (and at the same time undoubtedly compressed elevationally), mi- grations of high elevation elements from one peak or range to another were facilitated. However, as indicated in Fig. 7:4 most of such migrations were north-south rather than east-west, because of the orientation of the cordilleras and the barrier effects of the deep inter-Andean valleys produced by the Magda- lena, Cauca and Atrato rivers (Fig. 7:4). Differentiation of elements or distribution patterns that show the effects of these bar- riers can be seen in Polylepis (Simpson, in press) and numerous other plant genera (Simpson, 1975). In Venezuela and in the Sierra Nevada de Santa Marta of Colombia, evidence has established with certainty the existence of only two glacial episodes assumed to be co- incident with the last two major advances elsewhere (Shagam, 1975). Only late strati- graphic sequences have been analysed paly- nologically from Venezuela (Labouriau and Schubert, 1977), but geological evidence from moraines located in the Merida Andes at Sierra de Santo Domingo (8°48'N, 70°48'W; Schubert, 1974a), Pico Bolivar (8°30'N, 71°00'W; Schubert, 1974a), and Paramo La Culata (8°45'N, 70°60'W; Schubert, 1974b) provide documentation for only the last major glacial advance. This means that there have been at least two periods of facilitated migra- tion into these high elevation habitats and that these periods would have occurred dur- ing the final production (uplift) of these habitats. It should be noted in this context that the plant species diversity of the Merida 174 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Andes is lower than that of the Colombian Andes. It is possible that there were fewer opportunities for speeiation in situ (fewer times of effective glaciation) than in the Colombian Andes. Sauer's ( 1971 ) summary of the research on Ecuadorian Quaternary history indicated three glacial periods, the last of which re- duced snow line 1500 to 2000 m. According to his reconstructions (Sauer, 1971, Fig. 7:8), the Ecuadorian Andes were being continu- ously uplifted throughout the Pleistocene, and the first glacial advance had only minor ef- fects because of the lack of appreciable land above 2000 m. By the time of the last glacia- tion, the mountains had reached their present elevation. Despite the differing intensities felt from the three glaciations in Ecuador, Sauer ( 1971 ) correlated them with the Min- del, Riss and Wiirm glaciations of Europe. During the last glaciation at least, southward and northward (and probably to some extent across the central valley) migrations of high elevation taxa occurred in cooler and/ or more humid periods. Because the high elevation areas of Ecuador are now in the form of rela- tively isolated peaks (Fig. 7:4), many taxa exhibit differentiation from mountain to mountain. The restriction of taxa to moun- tains of certain regions can be seen in Poly- lepis (Simpson, in press), numerous other plant genera such as Niphogeton and Arraca- cia (Umbelliferae), Llerasia and Mutisia (Compositae) (see Simpson, 1975), and in some avian groups such as the Atlapetes schistaceous superspecies (Paynter, 1972). In Peru, where more stratigraphic work has been done, researchers ( Hastenrath, 1967; Clapperton, 1972; Dollfus, 1976; and Nogami, 1972), have interpreted levels of moraines as phases of the last or last two world-wide glacial advances (see below), although Stein- mann (1930), postulated that there were at least three (one very ancient) glaciations in Peru. In any event, moraines show that snow line was lowered about 1000 to 1500 m in the east and 500 to 1000 m in the west (the drier slope of the Andes) several times during the Quaternary. In a north-south direction across Peru, the snow line depression was fairly constant, despite the present decreasing annual precipitation southward (Hastenrath, 1967). Mercer and Palacios (1977) dated times of ice advance during the last glaciation in the Cordillera Vilcanota. The major ad- vance ended between 28,000 and 14,000 years b.p. but minor advances occurred at about 11,500 and between 600 and 300 years b.p. There is still much dispute about the amounts of precipitation received in Peru during glacial periods. Nevertheless, with the exception of Nogami (1972), most authors have come to the conclusion that there was an increase in total moisture received on the western slopes of the Peruvian and adjacent Chilean Andes at some time during the vari- ous glacial cycles. One or more periods of precipitation increase were postulated by Garner (1959) and Dollfus (1976), and ac- cepted by Gansser (1973). Mercer and Pala- cios ( 1977 ) were unable to come to con- clusions about ice age wetness. Possible moisture increases on the western montane flanks, as outlined by Simpson (1975), would have allowed southward expansion of upper montane forest down the Pacific facing slopes (Fig. 7:4). The present occurrence of relic patches of upper montane forest in sheltered, humid canyons along the western Andes prompted Koepcke (1961) to postulate that glacial periods had been accompanied by in- creases in precipitation in this region. An- thropological finds now indicate that rivers flowing from the Andes to the Pacific carried much more water at earlier times in the Pleistocene than at present and were able to support human settlements (Gansser, 1973). In eastern Peru, which normally receives a plentiful supply of rain, the depression of snow line and the increases in moisture would have been relatively less pronounced. It is doubtful that even the "arid" phases experi- enced by the Eastern Cordillera as envisioned by Garner ( 1959, Fig. 8 ) would have altered the habitat sufficiently so as to have created a north-south migratory corridor for arid or semi-arid elements. However, during periods of increased precipitation on the highest por- tions of the Cordillera, it is likely that there was an exchange of mesophytic taxa from east to west across the nudos of the Peruvian Andes (Koepcke, 1961, Fig. 2). These nudos are east-west extensions of the Cordilleras Oriental and Occidental that form "bridges" 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 175 across low inter-Andean regions. The most pronounced nudos occur slightly north of Lima (the Nudo de Pasco connecting the Cordillera Blanca and Negra with the Cordil- lera Oriental) and at about 15°S (Nudo de Vilcanota, connecting the Cordilleras Occiden- tal, Oriental and Auzangate). A common distribution pattern for plant taxa that indi- cates such a migratory route is the presence of populations in mesophytic habitats on the western slopes near Lima and then, disjunctly, across the Andes on the eastern slopes in or around the Urubamba Valley (Simpson, 1975). Data from the entire puna surface is pri- marily geological, although several of the lakes on the Altiplano eventually may prove to be reliable sources of continuous pollen cores. The careful work of Kessler ( 1963 ) and the earlier work of Troll (1927), provide a basis for interpretations about the changing regimes and fluctuating snow lines of the Quaternary. Their findings indicate that dur- ing "climax" phases of the last glaciation ( and perhaps earlier) lake systems covered much of the Altiplano (Simpson, 1975, Fig. 20) and that snow line was lowered 700 m below its present level on mountains flanking the plain. Vegetational changes on the nonflooded por- tions of the Altiplano during such periods may have been similar to those recorded near Lago Junin in central Peru (cited in Dollfus, 1976), namely, the production of bogs, per- haps seasonal bogs, dominated by members of the Cyperaceae and Juncaceae. Under such conditions, elements of the arid-scrub associations would have been restricted to higher, west-facing slopes of the Andes. The distributional ranges of xerophytic members of genera such as Lepidophyllum, Mutisia, Perezia, Baccharis, Senecio (all Compositae) and other genera, such as Cremolobus (Cru- ciferae), would have been severely frag- mented (see maps of present distributions of some of these taxa in Simpson, 1975; Fig. 7:4). In contrast to these taxa, elements such as species of Pohjlepis, which are now re- stricted to river valleys and/ or cloud belt areas, would have expanded their ranges. However, in the case of Pohjlepis, the exceed- ingly restricted modern distribution pattern is a result of both the present localized, patchy nature of suitable microhabitats and drastic cutting by man. In the southern part of the Altiplano near Tarija, fossil remains of Megatherium, Scelidontherium, Macrauchen- ia, and Equus (Ahlfeld and Branisva, 1960) indicate the presence of grasslands in glacial times. This region is now covered by desert scrub vegetation. In addition to humidity changes, studies of moraines on the mountains to the east of Lake Titicaca indicate at least four significant advances of ice, some of which extended into the lake itself. Unfortunately, no correlations of these advances have been made with ad- vances elsewhere and it is uncertain whether they represent stages in a late glaciation or independent major glacial advances. Northwestern Argentina and the northern part of Chile differed in their Pleistocene his- tory just as they now differ in their climate and vegetation. In this area of northern Argentina, both the vegetation of the puna and the ceja forests reach their southernmost limit. As in the case of the entire upper mon- tane forest from Colombia to this latitude, Pleistocene climatic changes probably had little disruptive effect on the continuity of the band of forest. A more significant change that occurred during the Quaternary would have been the southward spread of many North American forest taxa that entered South America after the closing of the Pana- manian Portal. Important genera such as Alnus, Jiiglans, and Quercus merged with native tropical and subtropical taxa such as Drimys and Podocarpus as they spread south. Alders and walnuts are now among the most characteristic trees of the upper forests in northern Argentina. Although the addition of these trees would not have altered the physi- ognomy of the forest, the fact that they be- came dominants and increased the amount of deciduousness would have had a pronounced effect on the insects and other animals for- merly associated with the montane forests. In the inter-Andean valleys of northern Argentina, conditions were quite different. An analysis of many of the internally drained basins, such as the Bolson de Pipanaco (27°S), indicates that at least once in the Pleistocene, presumably during glacial peri- ods, some were covered by lakes (Simpson 176 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 and Verhoorst, 1977). In addition, many of the higher peaks of the region such as the Nevados del Aconquija and the Sierra de Ambato were glaciated, with permanent snow lines reaching 500 m lower than any modern vestiges of ice. The combination of the for- mation of these lakes and the increase in permanent snow indicates that cold periods at the latitude of the tropic were accompanied by increases in precipitation. If conditions were appreciably wetter as well as colder, the monte vegetation now found in these valleys would have been eliminated or, at least, greatly restricted. The arid-scrub monte asso- ciations presumably persisted farther south on the east side of the Andes and/or in parts of the modern Chaco. The depauperate na- ture of the faunas now found in these isolated northern inter-Andean valleys has been at- tributed, in part, to the slow recolonization of the isolated pockets, following Quaternary periods of decimation (Mares, 1976). On the western side of the Andes be- tween latitudes 18° and 27°S there is evidence of slight glaciation in the Chilean Provinces of Tarapaca and Antofagasta. Most of the evidence of Pleistocene glaciations has been eroded away by modern weathering processes. However, those moraines that remain indicate that on the Payachata Volcanos (18°10'S), glaciations reached to 4500 m. Ice is now found only at elevations over 5500 m. On Volcan Sajama as well, ice lobes may have reached as low as 4500 m and on Yaricova 20°S down to 4000 m (Paskoff, 1977). Near the Salar de Huasco on the western edge of the Altiplano, Tricart (cited in Paskoff, 1977) found evidence of two, an older and a younger, glaciations, both of which left mo- raines at about 4200 m. Although most gla- cial advances in this area remain undated, Tricart correlated these with times of high shorelines (presumably equivalent to lower sea levels) and hence, glacial periods on a world-wide scale. Lakes were formed during cold periods in Tarapaca near the Peruvian border, as indicated by cold-water-adapted diatoms found fossilized in dry lake basins. The Cen- tral Valley in the northern part of Anto- fagasta also was covered by a lake (Paskoff, 1977). Farther to the south in Chile between 27° and 30°S, there is a poor record of Pleistocene conditions, although there are remnants of al- luvial terraces in Copiapo (Paskoff, 1977). Still farther to the south (30°-33°S), evidence of pronounced Quaternary events begin to appear. This area, as indicated above, now is subject to pronounced changes resulting from latitudinal movements of the nearby atmos- pheric pressure system. During the Pleisto- cene, these movements must have been al- tered in some fashion. All of the areas above 4000 m in this area were glaciated with ice extending to 2100 m elevation at 30°S, 2800 m at 33°S, 1700 m at 33°30'S (Maipo Valley), and to 1200 m at 34°S in the lower Central Valley (Paskoff, 1977). This substantial in- crease in the amount of ice formation indi- cates both a significant glacial depression in ambient temperature and an increase in total moisture received during the year. Terraces on the Pacific shore caused by increased pre- cipitation have been shown to correspond in time to periods of eustatic sea-level changes dated as glacial. The combined effect of cooler and wetter conditions in the region would have led to an environment amenable to growth of southern beech forests (Simp- son, 1973; Fig. 7:6). Remnants of the Notho- fagus forest, which was pushed to the north and capable of surviving in this area during glacial periods, can now be seen in small areas such as Fray Jorge (30°30'S), where woodlands with Drimys, Myrceugenia, Gun- nera, and Escallonia, still persist. Recently there has been discussion as to whether or not these woodlands are Pleistocene or pre- Quaternary relicts (Kummerow, et al., 1961). Geological data (Rirot, 1970) and paleon- tological evidence ( Hoffstetter and Paskoff, 1966) support an hypothesis of a Pleistocene age for these woodlands. Consequently, the last time(s) forests reached these areas would have been during cold, wet phases of the Pleistocene. In the centralmost portion of Chile at about the latitude of Santiago (33°S) and southward (to 39°S), evidence of glaciation is abundant. Alpine glaciers extended down to elevations of 2800 m at Portillo and to 1600 m at Guardia Vieja (Paskoff, 1977). In the valley of the Rio Bio-Bio, a low area of 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 177 u. X '. ' l r ■ . i *-» ul i 5 O Q ■ IDDI < < _l GC LU z LU C/) LU cc Q_ vnovoNoov < < NI1N31VA o0 VWIVIl uft vrvToia oia soia jo NOiioNnr voiaiamoNii saiaiamoNii „o VflOVONOOV OBOi Oo CD 0) I ■£•£ £ 60"d ^3 c c O 5) U ^ N a 05 a* 1-1 — _. „ o — "3 -^ > 3 o 3 (ui) gamin v o o o o o o fufl aannnv &-» to "B-T3 to Saps fc a *" 5 <-> G « § o to .2 c E o H •rl o 3 c C ■" a S O 0) o n3 o .a to u to a a 2 C co o 60 It T3 J5 e to >-3 CD ' ' Kft a Co a C cj N T3 g _2 to 'C O OS -to ■2 2, 'to "to 3 to -9 O O C ■" •2 « cs 3 2 -2 c »to C t) rt o c 0) > OJ U ** a 2 U <-Ui ^ o a .2 to ,3 £ C 55 0) sj5 ■X3 5 -a "2 -2 T ° a „ to 1- n 3 CD o & o c O T1 a's ? °5 » fc a w to e •£ .2 to ij E a c c ^-2 ' IB 0) a , — '0 to Co to §§^ '£ « -2 "a =-> 8 » " «> £(2 to^ o "= 3 -— ^ /^ "3 — - to - 60 Si to .£ a '■a s ° ■= -g -c S a CD a; 60 > to ~ ° C to co < 60 ij o 6 c > fe £ •£ .2 N | a 2 o-S- to 178 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 the Cordillera (Simpson, 1973), an ice tongue apparently reached as low as 1100 to 1200 m (Fig. 7:6). Although so postulated by Briig- gen ( 1950 ) , there has been no recent evidence in this part of Chile to support an hypothesis of an ancient glaciation. Paskoff ( 1977 ) ac- cepted the evidence for only two, relatively recent, major advances in this region. Fossil remains of cool-habitat-adapted vertebrates (Mastodon, Mylodon, Equus) have been as- sociated with the last of these advances (Casamiquela, 1969-1970, in Paskoff, 1977). Traces of these large vertebrates disappeared after 11,000 years b.p. In the southern lake region there is firm documentation for four glaciations. Between latitudes 39°-40°S, ice reached out from the Andes on several occasions westward to the base of the Coastal Cordillera. The last two of these advances have been named the Rio Negro and El Salto, respectively. The latter corresponds to the Llanquihue Glaciation of Heusser (1974). The earlier advances have neither been named nor dated. South of 40°, the Pleistocene geology is simplified. Ice from the Andes flowed west- ward and downward into the Pacific, virtu- ally eliminating all of the biota on the western slopes (Fig. 7:6). However, the flora and fauna of the southern forest and bog associa- tions persisted, because the climatic condi- tions allowed it to expand toward the Equator north of the sheet of ice (Figs. 7:6-7; see also Vuilleumier, 1971, Fig. 2, for a more detailed map of the extent of glacial ice). For biogeographers interested in Quater- nary biotic changes, one area in southern Chile has long remained controversial — the Island of Chiloe. Many authors have postu- lated that the island was free of glacial ice and served as a refugium for southern ele- ments at times of glacial maxima. Others have claimed that the island was covered by ice and could have served no such function. Recent geological and palynological investi- gations by Heusser and Flint (1977) settled the debate. The northern part of the island was, indeed, covered by glacial ice during parts of three independent advances. The first advance has not been dated, but the last two have been radiocarbon dated as having occurred at 57,000 and 43,000 years b.p. Paly- nological studies from cores showed that the coldest period was during the earliest ad- vance and that grasses and Compositae become dominant during all of the glacial stages. Several taxa now present, including some species of Nothofagus and Podocarpus, disappeared from the island. However, dur- ing the middle advance, other taxa, such as Poclocarpus andinus, now found only in the Andes at elevations of 1200 to 1800 m, were present. Other high latitude herbaceous spe- cies, such as Lycopodium fuegiana and Dra- pctes muscosa (Thymelaeaceae), were found in fossil associations with Podocarpus andi- nus. Thus, it is possible that parts of Chiloe did harbor in glacial periods plant (and ani- mal) taxa that are normally characteristic of higher elevations and latitudes. On the eastern side of the Andes, a wealth of new data has been gathered and sum- marized by Mercer (1976) about Pleistocene conditions of the lake region between 39° and 50°S. With the exception of the Sabana de Bogota in Colombia, this area now has the most complete and well-dated Quaternary sequence of any region of South America. Mercer and his colleagues have dated the old- est glacial advance with geological remains (49°28'S) at 3.5 million years. The glacial advance which left the most extensive re- mains occurred 1.2 million years ago. Sub- sequent to this, several smaller advances, cul- minating at about 13,000 years b.p., took place. The most extensive of this last series was about 56,000 years b.p. In his analysis, Mercer (1976) concluded that the Patagonian gravel, hypothesized by Auer ( 1970 ) to have been deposited by a piedmont glacier, is, in fact, a composite mixture of glacial outwash deposited over a time span dating from the mid-Pliocene. Correlation of the Argentine glacial se- quences with those of the Northern Hemi- sphere is still uncertain, but three advances during the Recent period (4,600-4,200 years b.p., 2,700-2,000 years b.p., and ca. 1250 AD) coincide with advances of glacial ice else- where in the world (Mercer, 1976). Mercer's work does not include palyno- logical investigations but some conclusions 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 179 STRONG GLACIAL Fig. 7:7. Climatology of a maximum glacial (A) and interglacial (B) cycle following Fairbridge (1972) showing the proposed ocean current, pressure systems and wind direction changes. Continental outlines and the extent of glacial ice were drawn from Vuilleumier (1971). It is assumed in this model that during glacial periods, the intensities of the Atlantic and Pacific anticyclones (high pressure systems) and the force of the Humboldt Current were increased. The Amazon Basin was drier than at present because the influx of humid air from the northeast was prevented. Dotted arrows are January wind directions, solid lines are wind direc- tions in July. CUmatologia de un pcriodo glacial maxima (A) y de un interglacial (B) siguicndo a Fairbridge (1972). Se muestran las corrientes oceanicas propuestas, los sistemas de )ircsion, y los cambios de direction de los vientos. Los limitcs continentales y la extension de los hielos glaciates fueron tornados de Vuilleumier (1971). Se asume en cste modelo que durante los periodos glaciates, las intensidades de los anticiclones (sistemas de alto prcsion) del Atlantico y del Pacifico, y la fuerza de la corriente de Humboldt se incrementaron. La cuenca amazonica fue mas scca que a lo presente debido a la auscncia de los vientos humedos dcsde el noreste. Las flcchas entrecortado.s indiean direction del viento en Enero; las flcchas continuas, la direction de los vientos en Julio. about the effects of the long series of glacial advances on the biota of southeastern South America can be drawn. First, the biota obvi- ously has been subjected to periods of intense cold interspersed with periods as warm as, or warmer than, the present, since the middle of the Pliocene. Such fluctuations, as previ- ously outlined by Simpson (1973), would have caused the high elevation biota of the region to migrate northward and/or down- ward several times. For elements in habitats above tree line, a succession of downward and outward migrations followed by retreat upward would have led to rapid differentia- tion of populations that were reduced in size and restricted in distribution durinff inter- glacial periods (Fig. 7:6). Forest elements, in contrast, would have been shifted in distri- bution, but their ranges would not have been subjected to fragmentation either during gla- cial or interglacial periods. The investigations of Mercer (1976) also confirm the glacial limit at the base of the Andes on the eastern flanks as first established by Caldenius ( 1932 ) . This limit of glacial ice provides a guide to the eastern limit of the southern beech forests during glacial periods. The tableland of Patagonia would have been neither a solid sheet of ice nor a broad ex- panse of forest. Rather, it probably was a boggy moorland periodically subjected to flooding and deposition of rubble during gla- 180 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 cial periods. For further discussion of Qua- ternary climates in Patagonia, see Baez and Scillato Yane (this volume). Eastern Highlands In Eastern South America, there is poor documentation for events during the Qua- ternary and some debate as to the effects of various climatic changes. In the Serra do Mar, Bigarella and Mousinho (1966) de- scribed a series of terraces that they ascribed to changes in humidity during the Pleisto- cene. When an arid climatic cycle occurred, valleys were filled with debris (from short- lived stream flow) and sudden but sporadic showers produced slope retreat at the base of mountains leading to the formation of pedi- ments. In succeeding humid phases these pediments were partially degraded (because of continuous washing by rainfall), and the debris in the valleys were partially swept and scoured away. A series of such climatic changes led to the superposition of eroded pediments and stream-bed patterns that these authors correlated with times of low sea level or glacial periods. However, as Beurlen (1970) has cautioned, beach levels per se are not necessarily an accurate indication of gla- cial advances. Other corroborative evidence of temporal correlations is needed. This evi- dence is provided, in part, by pollen se- quences found elsewhere in South America that indicate dry phases were coincident (in tropical areas) with glacial advances in the Andes (van der Hammen, 1974). At higher elevations, in the Serra da Man- tiqueira, there are geomorphological remains that have led some authors ( cited in Beurlen, 1970) to postulate the formation of a perma- nent glacier on Itatiaia during glacial times. Such formations include the presence of huge scattered boulders and U-shaped valleys (Ebert, 1960). However, there are no actual moraines. It is possible, in opposition to the conclusion of Ebert, that the boulders and the U-shaped valleys were products of mud flows and solifluction rather than glacial ice. Moreover, the presumed elevation of the Itatiaia glaciation was 2100 to 2500 m, an elevation lower than any Andean glaciation at similar latitudes. Odeman (cited in Beur- len, 1970) argued that the cooling on this iso- lated mountain could not have been sufficient to have caused ice formations at this low elevation. Consequently, the presence or absence of an alpine glacier on Itatiaia during Pleistocene glacial periods still remains un- certain. There is no doubt, however, that the high- est parts of Itatiaia, the rest of the Serra da Mantiqueira, and other highlands in south- eastern Brasil were subjected to periods of cold and (at least) increased aridity relative to that of the present. In terms of the biota, the periods of cold would have led to a de- pression of vegetation zones accompanied by an increase in the superficial extent of the vegetation zones of the upper elevations. This increase in superficial distribution would have promoted migration to, and exchange of ele- ments with, neighboring mountains (Fig. 7:4). Several authors (Brade, 1956; Smith, 1962; Miiller, 1968; Klein, 1975) have pointed out the relationships of components of the biota of the southeastern highlands with the biotas of the Andes and the lowland south temperate regions. Many genera, such as Araucaria, Poclocarpus, which show connec- tions with the Andes, are remnants of ancient, widely distributed temperate forests (i.e., Early Tertiary), but many of the herbaceous genera, such as Azara (Flacourtiaceae), Boopis (Calyceraceae), Escallonia (Escallon- iaceae), which are now found disjunctly in the highlands of southeastern Brasil may have spread to this region during cool periods of the Pleistocene. Other plant genera that are more widely distributed in high elevation and/or temperate regions of South America and also occur disjunctly in the highlands of southeastern Brasil (e.g., Jamesonia, Gyno- grama, Polypodiaceae; Anemone, Clematis, Ranunculus, Ranunculaceae; Berberis, Ber- beridaceae; and Lepicliina, Labiatae) also may have reached these mountains during cold periods. In the Guiana Highlands, virtually no Pleistocene geological studies have been con- ducted. All of the presumed reconstructions of Quaternary climatology have been made from inferences derived from data from other areas. Such data from the Andes to the west, the southeastern Brasilian Highlands to the 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 181 south, and the Caribbean to the north (Ro- natti and Gartner, 1973) leave little choice but to assume that the Pantepui tablelands were cooler in times of glacial advance else- where in the tropics of South America and that vegetation zones of the upper elevations were lowered similarly to those elsewhere. As in the case of the paramos and the campos of the Brasilian highlands, a lowering of the upper elevation vegetation zones on these tablelands would have slightly increased their areal distribution and produced patches of subtropical forest on low mountains that are now covered by tropical forest. These changes in vegetation patterns would have fostered exchange among the tepuis and increased exchange of some elements with the Andes (Mayr and Phelps, 1967; Haffer, 1970b). Cook ( 1974 ) , in a study of the origin of the avifauna of these highlands, pointed out that there is little correlation between the areal extent of the modern upper elevation zones, or the complexity of the vegetation of the vari- ous tepuis, and the number of breeding bird taxa found on each. This lack of correlation implies that there is not an equilibrium as would be expected following the predictions of the model of island biogeography (Mac- Arthur and Wilson, 1967). Cook interpreted the lack of an equilibrium as an indication of the recent arrival of the avifauna. Mayr and Phelps ( 1967 ) also pointed out that the ma- jority of the avian taxa on these highlands is derived from Andean, and hence young, stocks. Presumably, during glacial periods, many successful colonizers from the Andes reached the Guiana Highlands and replaced an original, ancient avifauna. Nevertheless, it is interesting that there is little floristic con- nection between the tepuis and the Andes, despite the evidence from avian relationships. In contrast to the relatively low level of en- demicity in the avifauna (10-30%), over 50 percent of the flora is estimated to be en- demic. Moreover, the relationships of the flora of the Guiana Highlands, albeit distant, are primarily with the highlands of south- eastern Brasil or even Africa rather than with the Andes. In view of the presumed recency of the avifauna and its successful migration eastward during the Pleistocene, it is puzzling that the flora has remained so distinct and remotely related to the flora of either the Andes or the surrounding lowlands. THEORIES OF GLACIAL CLIMATOLOGY Refore extensive studies in the Southern Hemisphere had been made, many authors (e.g., Budel, 1951) assumed that Pleistocene climatic changes in both hemispheres involved simply an equatorial shift and latitudinal compression of the earth's climatic belts. The net effect of such shifts would have been to push high latitude climatic zones toward the Equator and to reduce the latitudinal extent of the mid-latitude semi-arid zones. This model of Quaternary climatology was based on European data and incorporated the ap- parent synchrony of cold and wet periods. It is now clear that in many regions portions of glacial cycles (often the coldest part) were arid rather than wet (Hammond, 1976). In South America, investigations have shown that the Quaternary climatic picture was par- ticularly confusing, because the continent spans two hemispheres and is dominated by exceedingly high mountain ranges. A further impediment to an elucidation of the Pleisto- cene climates across the entire continent has been the lack of precise temporal correlations for what appear to be actual discrepancies in temperature-moisture relationships. Several hypotheses, not necessarily conflicting, have been proposed to explain the different cli- matic regimes in various parts of the con- tinent. Some authors (e.g., Nogami, 1972) have suggested that there was no major change at any time in the Pleistocene in the atmospheric circulation pattern over western South Amer- ica. On the basis of observations that the snow line in the Andes during glacial ad- vances was parallel to that now found, No- gami ( 1972 ) concluded that the "apparent" increase in moisture was only a reduction in evaporation caused by lower temperatures. However, most authors disagree, favoring in- stead changes in the force and/or location of the various high pressure systems that af- fect the South American climate. Fairbridge ( 1972 ) divided, for purposes of glacial clima- 182 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 tology, cycles of climatic change (considered to be periods of 100,000 years) into four sub- stages — the kataglacial with a cold, wet cli- mate; the peniglacial with a very cold, very dry climate; the anaglacial or initial inter- glacial phase with a cool, diy climate changing to mild and wet conditions; and the true interglacial with conditions similar to those of the present. As this complex scheme indi- cates, times of glacial advance would involve both arid and humid periods; Fairbridge (1972) attributed the older associations of glacial periods and "pluvials" to interpreta- tions made on data from anaglacial or kata- glacial phases of the cycle. Fairbridge further proposed that the aridity of the full glacial phase in the low latitudes of southern conti- nents was caused by ocean-surface cooling and the buildup of semi-stable high pressure areas near tropical continents (Fig. 7:7). In South America, semi-stable highs would have been located off the north coast of the Guianas, in the Atlantic Ocean off the tip of Brasil, and off the coast of Peru (Fig. 7:7). Although the model proposed by Fair- bridge ( 1972 ) explains the climates that have been proposed for eastern tropical South America, it does not account well for the com- plexities exhibited in the western tropical Andes nor for conditions found at high lati- tudes. Garner (1959) argued that Pleistocene glacial advances on the eastern and western slopes of the Peruvian Andes had been out of phase (Fig. 7:8) and suggested that times of expansion of glacial ice on the Eastern Cordil- lera were synchronous with those in Antarc- tica (i.e., in phase with glacial advances in most areas of the world). He hypothesized that during such periods, the southern oceans and the Humboldt Current (originating in southern waters ) would have become increas- ingly cold as the glacial cycle progressed, causing a general cooling of the atmosphere in South America. The resulting general tem- perature depression would have led to the accumulation of ice (reduced melting) in areas where moisture was plentiful, such as over the Eastern Cordillera. He further pro- posed that the ^Vestern Cordillera would have experienced increased aridity. During the in- terglacial cycles in the Antarctic region, the southern oceans and the Humboldt Current would have become comparatively warm, and glaciers would have retreated in Antarctica and on the Eastern Cordillera. Garner (1959) proposed that at this time there was increased precipitation on the Western Cordillera and the expansion of glaciers there (Fig. 7:8). Garner's hypothesis, while taking into ac- count the complexities of the Andes and the apparent nonsynchrony of glacial advances on the two sides of the Peruvian Andes, ap- pears to conflict with data from the Brasilian lowlands (Haffer, 1974) that there was arid- ity, rather than increased humidity, during full glacial phases. The Eastern Cordillera is under the same climatic influences that affect the Amazon Basin. Nevertheless, it is possible that the height of the Andes was sufficient to cause increased precipitation at high elevations when the air from the south- east that picked up moisture and caused aridity in the lowlands reached them (see also Fairbridge, 1972). In the most recent discussion of Andean Pleistocene climatology, Dollf us ( 1976 ) stated that the climatic conditions on the two sides of the Equator differed during glacial maxima. North of the Equator, cold periods were asso- ciated with dry conditions during peniglacial phases, as proposed by Fairbridge (1972); south of the Equator, peniglacial phases were cold and wet. Dollfus postulated that during major glacial advances, the equatorial high pressure system off western South America "weakened" and the onshore winds were re- duced in force. The strength of the Humboldt Current was correspondingly reduced, and the northern Andes of Peru received increased moisture in the form of showers. This hypoth- esis conflicts with that of Garner ( 1959 ) . One point emphasized by Dollfus is that local conditions would have altered major climatic patterns and led to anomalies during Pleisto- cene glacial periods. The aridity of the Alti- plano, which he dated as glacial, would have been a result of such local phenomena. He based this conclusion on the fact that, under present conditions, the Altiplano is driest when the Amazon Basin receives abnormally high amounts of rainfall. It is evident that Dollfus' hypothesis conflicts with the findings that during glacial periods the Amazon Basin was dry, not wet, and with the studies of 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 183 WESTERN CORDILLERA EASTERN CORDILLERA CORRESPONDING PEDIPLAIN ARROYO THROUGH FLOWING RIVER ARID MINOR HUMIDITY ARIDITY MAJOR HUMIDITY SEVERE ARIDITY MAJOR HUMIDITY SEVERE ARIDITY HUMIDITY LEVELS PRESENT l ALTIPLANO HUMIDITY SLIGHT GLACIAL SUBARID HUMID GLACIAL SEVERE ARIDITY HUMID SEVERE GLACIAL SEVERE ARIDITY MAJOR HUMIDITY SEVERE ARIDITY Fig. 7:8. Geomorphological sequences indicating Pleistocene climatic changes on the eastern and western slopes of the Peruvian Andes after Garner (1956). The eastern sequence is from the Urubamba Valley, Cordillera Oriental, Peru and the western sequence from southern Peru in the Cordillera Occidental west of Arequipa. Shading indicates aggredation accumulated during times of aridity. In humid periods, strongly flowing rivers and streams incise the rubble previously deposited in stream beds. Numbers are assigned to level of the same presumed age. Non-synchrony of arid and humid phases on the two Cordilleras is indicated by the divergent geomorphological processes at the same level. Secuencias geomorfologicas indicando los cambios climdticos del Pleistoceno en las vertientes oriental y occidental de los Andes peruanos de acuerdo a Garner (1956). La secuencia oriental corresponde al Valle de Urubamba, Cordillera Oriental, Peru, y la sceuencia oeste corresponde al sur del Peru, en la Cordillera Occi- dental al oeste de Arequipa. El sombreado indica deposiciones acuniuiadas durante los tiempos de aridez. En periodos humedos, fuertes flujos fluviales y de arroyos carcomieron la coma de piedra previamente depositada en la eorriente. Los numeros corresponden a niveles de una misma presupuesta (clad. La ausencia de sinconi- sidad de las fuses humedos y dridas en las dos Cordilleras estd indicada por los procesos geomorfologicos diver- gentes al misma nivel. 184 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Kessler (1963) that showed increased pre- cipitation and lake formation across the Altiplano during glacial periods. Obviously there is still disagreement about the time of Pleistocene humid phases during or after glacial cycles throughout South America. Nevertheless, the various theories about Pleistocene climatology show the im- portance of taking local factors into account and the dangers of overgeneralization from data derived from a limited area. As empha- sized by Beurlen (1970), more palynological data and absolute datings are necessary be- fore a truly meaningful and comprehensive picture of South American Quaternary clima- tology can be drawn. ACKNOWLEDGMENTS Thanks are due J. Haffer and J. L. Neff for their helpful suggestions on the manu- script and to A. Tangerini and S. Yankowski for the excution of the drawings. RESUMEN Las regiones altas (sobre 2000 m) de Sudamerica ocurren principalmente en la parte oeste del continente donde la Cordil- lera de los Andes corre paralela a la costa por mas de 8000 km. Otras regiones altas con elevaciones por sobre los 1000 m se encuen- tran en el sudeste del Brasil y a lo largo de la frontera venezolana-guayanesa-brasilera. Los areas por encima de los 2000 m en estas dos regiones, sin embargo, son bastante pequenas. Las regiones montanosas del sudeste brasileiio son conocidas colcctivamente como las Al- turas Brasilenas del Sudeste, y aquellos del norte como Alturas Guayanesas. Geologicamente, el sistema andino comple- to es muy joven, teniendo casi todas sus por- ciones bajo el nivel del mar hasta al final del Cretaceo. Aiin cuando hay diferencias en detalles cronologicos, la mayoria de sus uni- dades geologicas esperimentaron pulsos de levantamiento a traves del Terciario y del Cuaternario. La mayoria de las unidades emergieron al final del Cretaceo o al comienzo del Eoceno, experimentaron un levantamiento mayor en el Mioceno y estuvieron sujetos a un levantemiento final e importante al tiempo del Plio-pleistoceno. Por el contrario, las Al- turas Brasilenas del Sudeste y las Alturas Guayanesas han sido masas de tierras conti- nentales desde tiempos pre-Cambricos. Ele- vaciones hasta 1000 m o mas probablemente ocurrieron desde el Cretaceo. Sin embargo, como una consecuencia de las diferencias temporales para guarecer plantas y animales, las alturas del Este tienen una biota mas an- tigua y endemica que la de los altos Andes. Esto es porque todos aquellos plantas y ani- males que viven por sobre 2000-3000 m de elevacion en los Andes han sufrido inmigra- cion, colonization, y diferenciaeion solo desde el termino del Terciario. Los eventos cli- maticos y geologicos del Cuaternario fueron, entonces, una parte integral de la formation de esta biota. En las alturas del Este, los elementos del Cuaternario constituyeron mer- amente una etapa mas de una larga historia evolutiva. Dada la extension latitudinale de la Cor- dillera de los Andes, clima y vegetation difier- en grandamente de norte a sur. Del mismo modo, los efectos de los eventos climaticos del Cuaternario fueron diferentes de acuerdo a la position latitudinal. En general, las tem- peraturas disminuyeron a traves de la Cor- dillera en forma sincronica con las expan- siones mundiales de los hielos glaciales (de- nominados aqui "periodos glaciales"). A lo largo de los Andes, la linea de las nieves tam- bien bajo. El algunas areas, como en los Andes del norte, porciones de los ciclos gla- ciales parecen haber sido mas secos que lo que ellos son hoy en dia. En otros areas, como en el Altiplano y en las regiones tem- pladas del sur, las condiciones glaciales parecen haber sido mas humedas que al presente. El numero de ciclos de expansion de los hielos glaciales a lo largo de la Cor- dillera varia desde un episodic tardio en algunos recientemente elevados picos en el extreme norte, masta al menos cuatro en la region sureria chileno-argentina y donde el primero de los cuales ocurrio en el Pliocene En las Alturas Brasilenas del Sudeste la ocurreneia de una glaciacion no ha sido pro- vada ciertamente aim cuando restos geomor- fologicos sugieren periodos de extremo frio. Datos geomorfologicos adicionales implican 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 185 ciclos fn'os sincronicos con periodos de aridez. Evidencias inferidas desde areas contiguas sugieren que las Alturas Guayanesas fueron tambien mas fn'as durante los periodos gla- ciales y probablemente experimentaron con- diciones mas secas durante partes de estos ciclos que aquellas presentas boy. Los afectos de los ciclos glaciales sobre la biota variaron consecuentemente en las dis- tintas regiones montanosas. En areas como los Andes del norte donde los picos emer- gentes contienen expansiones aisladas de paramos y los Andes del sur donde picos disjuntos estan cubiertos con ambientes al- pinos, los cambios climaticos glaciales per- miticron expansiones de los rangos de ele- mentos de altura y facilitaron la dispersion. Reduction del tamano poblacional, aislami- ento, y diferenciacion ocurrieron primaria- mente durante los periodos interglaciales. Para los elementos de la puna, especialmente aqucllos del Altiplano, la presencia de hielos glaciales y la formation de lagos glaciales causaron restricciones y aislamiento de pobla- ciones. En estos organismos, expansiones de rango y reunification de poblaciones que pa- recen haber sufrido solo una diferenciacion debil y expansiones recientes han producido los complejos patrones biologicos de sobre- montaje secundario e hibridizacion. En el sudeste del Rrasil y en las Alturas Guayanesas, los periodos glaciales fueron sin duda oportunidades en que los ambientes de altura se expandieron y, consecuentemente, durante las cuales las migraciones dentro de los sistemas montaiiosos se facilitaron. En el caso de las montarias del sudeste brasileno, una inmigracion desde los altos Andes y/o desde los areas temperados de altas latitudes tambien ocurrio durante los ciclos glaciales. Aiin cuando la migration de aves desde los Andes a las Alturas Guayanesas pudiera haber sido similarmente promovida durante los pe- riodos frios, un analisis de la flora de esta region demuestra que pocas, si algunas, plantas colonizaron hacia el este durante el Pleistoceno. Varias teorias han sido elaboradas para explicar los diversos regimenes de humedad/ temperatura durante las diversas fases del Pleistoceno en diferentes partes del conti- nente. Algunas de estas teorias explican todos los cambios climaticos en terminos de una depresion de temperatura solamente. Otros proponen varios cambios en la ubicacion de los sistemas de presion atmosferica. Sin em- bargo, ninguna de estos hipotesis explica com- pletamente todo el panorama continental. Se necesita mas trabajos palinologicos y geomor- fologicos combinados con determinaciones de edades mas precisos para ofrecer una vision completa de los cambios climaticos del Cua- ternario. LITERATURE CITED Ahlfeld, F. 1970. Zur Tektonic des Andinen Bo- livien. Geo]. Rundsch. 59:1124-1140. Ahlfeld. F., Bramsva, L. 1960. Geologia de Bo- livia. Bosco, La Paz. 245 p. Almeida, F. F. M. de, Amaral, G., Cordaxi, U. G., Ka washita, K. 1973. The Precambriam evolu- tion of the South American cratonic margin south of the Amazon River, pp. 411-446 in Nairn, A. E. M., Stehli, F. G. (eds.). The Ocean Basins and Margins. Vol. 1. The South Atlantic. Ple- num, New York, 584 p. Auer, V. 1970. The Pleistocene of Fuego-Patagonia. Part V. Quaternary problems of southern South America. Ann. Acad. Sci. Fenn. Ser. A3 100: 1-195. Beurle.v, K. 1970. Geologie von Brazilien. Gebruder Borntraeger, Berlin, 144 p. [Vol. 9 of Beitrage zur regionalen Geologie der Erde. H. J. Martini (ed.).] Bigahella, J. J., Moushino, M. R. 1966. Slope de- velopment in southeastern and southern Brazil. Z. Geomorphol. 10:150-160. Bihot, P. 1970. Esquisse comparative des differents types de modele en roches cristallines dans la zone temperee chaude de l'Amerique du Sud. Rev. Geogr. Phys. Geol. Dynam. 4:299-312. Bonatti, E., Gartner, S., Jr. 1973. Caribbean climate during the Pleistocene ice ages. Science 244:563-565. Brade, A. C. 1956. A flora do Parque Nacional de Itatiaia. Bol. Parque Nac. Itatiaia, Brazil. Rio de Janeiro. 5:1-92. Bruggen, J. 1950. Fundamentos de la Geologia de Chile. Inst. Geogr. Militar, Santiago, Chile, 374 p. Budel, J. 1951. Die Klimazonen des Eiszeitalters. Eiszeitalter 1:16-26. Burcl, H. 1961. Historia geologica de Colombia. Rev. Acad. Colombiana Cien. Exactas Fis. Nat. 11:137-191. Cabrera, A. L., Willink, A. 1973. Biogeografla de America Latina. General Secretariat of the Or- ganization of the American States, Washington, D.C. Monografia 13. Ser. Biologia, 120 p. Caldenius, C. C. 1932. Las glaciaciones Cuater- 186 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 narias en la Patagonia y Tierra del Fuego. Geogr. Annaler 14:1-164. Carlquist, S. 1957. Anatomy of Guayana Mutisieae. Mem. N.Y. Bot. Gard. 9:441-476. Carlquist, S. 1974. Island Biology. Columbia Univ. Press, N.Y., 660 p. Clapperton, C. M. 1972. The Pleistocene moraine stages of west-central Peru. J. Glaciol. 11:255- 263. Cobbing, E. J., Ozard, J. M., Snelling, N. J. 1977. Reconnaissance geochronology of the crystalline basement rocks of the Coastal Cordillera of Peru. Geol. Soc. Amer. Bull. 88:241-246. Cook, R. E. 1974. Origin of the highland avifauna of southern Venezuela. Syst. Zool. 23:257-264. Cuatrecasas, J. 1968. Paramo vegetation and its life forms, pp. 163-186 in Troll, C. (ed.). Geo- ecology of the mountainous regions of the tropical Americas. Colloq. Geog., Geogr. Inst. Univ. Bonn. 9:1-223. Dalmayrac, B., Lanceolot, J. R., Layreloup, A. 1977. Two-billion-year granulites in the Late Precambrian metamorphic basement along the southern Peruvian coast. Science 198:49-51. Dessanti, R. N. 1972. Andes Patagonicos septen- trionales, pp. 65.5-687 in Leanza, A. F. (ed. ). Geologia regional Argentina. Acad. Nac. Cien. Cordoba, 869 p. Dollfus, O. 1959/1960. Presentation de la struc- ture des Andes centrales peruviennes. Trav. Inst. Francais Etud. Andines 8:53-64. Dollfus, O. 1976. Les changements climatiques Holocenes dans les hautes Andes tropicales. Bull. Assoc. Geogr. Franc. 433:95-103. Dott, R. H., Jr., Winn, R. D., Jr., DeWit, M. J., Bruhn, R. L. 1977. Tectonic and sedimentary significance of Cretaceous Tekenika Beds of Ti- erra del Fuego. Nature 266:620-622. Ebert, H. 1960. Novas observacoes sobre a glaciacao Pleistocenica na Serra do Itatiaia. An. Acad. Bras. Ciena 32:51-73. Fairbridge, R. W. 1972. Climatology of a glacial cycle. Quat. Res. (N.Y.) 2:283-302. Freitas, R. O. de. 1951. Ensaio sobre a tectonica moderna do Brazil. Univ. Sao Paulo Fac. Filos. Cienc. Let. Bol. Geol. 130:1-120. Gansser, A. 1954. The Guiana Shield (S. America). Eclogae Geol. Helv. 47:77-112. Gansser, A. 1955. Ein Beitrage zur Geologie and Petrographie der Sierra Nevada de Santa Malta (Kolumbien, Siidamerika). Schweiz. Mineral. Petrogr. Mitt. 35:209-279. Gansser, A. 1973. Facts and theories on the Andes. J. Geol. Soc. London 129:93-131. Garner, H. F. 1959. Strati graphic-sedimentary sig- nificance of contemporary climate and relief in four regions of the Andes mountains. Geol. Soc. Amer. Bull. 70:1327-1368. Garner, H. F. 1975. Radiometric evidence for pre- Wisconsin glaciation in the northern Andes: Com- ment. Geology 3:230-231. Haffer, J. 1970a. Geologic-climatic history and zoogeographic significance of the Uraba region in northwestern Colombia. Caldasia 10:603-636. Haffer, J. 1970b. Entstehung und Ausbreitung nord-Andiner Bergvogel. Zool. Jahrb. Abt. Syst. Oekol. Geogr. Tiere 97:301-337. Haffer, J. 1974. Avian speciation in tropical South America. Publ. Nuttall Ornithological Club, Cam- bridge, Mass., 14, 390 p. Hammond, A. L. 1976. Paleoclimate: Ice age earth was cool and dry. Science 191:455. Harrington, H. J. 1956. Main morphostructural re- gions of South America, pp. xii-xviii in Jenks, W. F. (ed.). Handbook of South American Geol- ogy. Geol. Soc. Amer. Mem. 65, 378 p. Hastenrath, S. I. 1967. Observations on the snow line in the Peruvian Andes. I. Glaciol. 6:541- 550. Herd, D. G., Naeser, C. W. 1974. Radiometric evi- dence for pre-Wisconsin glaciation in the northern Andes. Geology 2:603-604. Heusser, C. J. 1974. Vegetation and climate of the southern Chilean Lake District during and since the last interglaciation. Quat. Res. (N.Y.) 4: 290-315. Heusser, C. J., Flint, R. F. 1977. Quaternary gla- ciations and environments of northern Isla Chiloe, Chile. Geology 5:305-308. Hoffstetter, R., Paskoff, R. 1966. Presence des genres Macrauchcnia et Hippidon dans la faune Pleistocene du Chili. Bull. Mus. Natl. Hist. Nat. Zool. Paris (2)38:476-490. Hueck, K. 1966. Die Walder Sudamerikas. Gustav Fischer Verlag. Stuttgart, 422 p. Hueck, K., Seibert, P. 1972. Vegetationskarte von Siidamerika. Mapa de la Vegetacion de America del Sur. Ibid. 69 pp. Hurley, P. M„ Rand, J. R. 1973. Outline of Pre- cambrian chronology in lands bordering the South Atlantic, exclusive of Brazil, pp. 391-^410 in Nairn, A. E. M., Stehli, F. G. (eds. ). The ocean basins and margins. 1. The South Atlantic, Ple- num Press, N.Y., 583 p. James, D. E. 1971. Plate tectonic model for the evolution ol the central Andes. Geol. Soc. Amer. Bull. 82:332.5-3346. James, D. E. 1973. The evolution of the Andes. Sci. Amer. 229(2) :60-70. Johnson, A. M. 1976. The Climate of Peru, Bolivia and Ecuador, pp. 174-218 in Schwerdtfeger, W. (ed.). Climates of Central and South America. Elsevier, N.Y., 532 p. [Vol. 12 of World Survey of Climatology. H. E. Landsberg, ed. in Chief]. Kessler, A. 1963. Uber Klima und Wasserhaushalt des Altiplano (Bolivien, Peru) wahrend des Hochstandes der lezten Vereissung. Erdkunde 17:16.5-173. Klein, R. M. 1975. Southern Brazilian phytogeo- graphic features and the probable influence of Upper Quaternary climatic change in the Holistic distributions, pp. 67-8S in Bigarella, J. J., Becker, R. D. (eds.). Internat. Symp. on the Quaternary (Southern Brazil, July 1975). Bol. Parana. Geocienc. 33. Koepcke, H.-W. 1961. Synokologische Studien an 1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 187 der Westseite der peruanischen Anden. Bonn. Geogr. Abhandl. 29:1-320. Kummebow, J., Matte, V., Schlegel. F. 1961. Zum problem der Nebelwalder and der zentral- chilenischen Kiiste. Ber. Dsch. Bot. Ges. 74:135- 145. Labouriau, M. S. L., Schubert, C. 1977. Glacial and palynological studies in die Venezuelan Andes and late Quaternary instability. I\' Symp. Int. Ecol. Tropical, Panama, 7-11 March, 1977. p. 103 [Abstract]. Lexico Estratigraftco de Venezuela (ed. 2). 1970. Bol. Geol.. Publ. Especial 4. Minist. Minas Hidrocarburos., Dir. Geol., 758 p. MacArthur, R. H., Wilson-, E. O. 1967. Theory of Island Biogeography. Princeton Univ. Press, Princeton, N.J., 203 p. Macuire, B. 1970. On the flora of the Guayana Highland. Biotropica 2:85-100. Mares, M. A. 1976. Convergent evolution of desert rodents: Multivariate analysis and zoogeographic implications. Paleobiology 2:39—63. Mavd, E., Phelps, W. H.. Jr. 1967. The origin of the bird fauna of the south Venezuelan highlands. Bull. Amer. Mus. Nat. Hist. 136:269-328. Mercer, J. H. 1976. Glacial history of southernmost South America. Quat. Res. (N.Y.) 6:125-166. Mercer, J. H., Palacios M., O. 1977. Radiocarbon dating of the last glaciation in Peru. Geology 5:600-604. Miller. A. 1976. The climate of Chile, pp. 113- 146 in Sciiwedtfecer, \V. (ed.). Climates of Central and South America. Elsevier, N.Y., 532 p. [Vol. 12 of World Survey of Climatology, H. E. Landsberg, ed. in Chief]. Muller, P. 1968. Die Herpetofauna der Insel von Sao Sebastiao (Brasilien). Saarbrucker Zeitung Verlag, 68 p. Nogami. M. 1972. The snowline and climate during the last glacial period in the Andes mountains. Quat. Res. (Tokyo) 11:71-81. Okada, A. 1971. On the neotectonics of the Atacama fault zone region. — Preliminary notes on late Cenozoic faulting and geomorphic development of the Coast Range of northern Chile. Bull. Dep. Geogr. Univ. Tokyo. 3:47-65. Oliveira, A. I. de., Leonardos, O. H. 1943. Geol- ogia do Brazil. Min. Agricultura, Rio de Janeiro, 813 p. Paskoff, R. P. 1977. Quaternary of Chile: The state of research. Quat. Res. -(N.Y.) 8:2-31. Paynter, R. A., Jr. 1972. Biology and evolution of the Atlapctes schistaceus species group (Aves: Emberizinae). Bull. Mus. Comp. Zool. Harvard Univ. 143:297-320. Prohaska, F. 1976. The climate of Argentina, Para- guay and Uruguay, pp. 13-112 in Schwerdtfe- cer, W. (ed.). Climates of Central and South America. Elsevier, N.Y., 532 p. [Vol. 12 of World Survey of Climatology, H. E. Landsberg, ed. in Chief]. Putzer, H. 1968. Uberbliek uber die geologische Entwicklung Sudamerikas, pp. 1-24 in Fittkau, E. J., Illies, J., Klinge, H., Schabe, G. H, Soli, H. (eds.). Biogeography and Ecology in South America. Junk. The Hague. Vol. 1. 445 p. Saver, W. 1965. Geologia del Ecuador. Min. Edu- cation, Quito, Ecuador, 384 p. Sauer, W. 1971. Geologie von Ecuador. Gebriider Borntraeger, Berlin, 316 p. [Vol. 11 of Beitrage zur regionalen Geologie der Erde. H. I. Martini (ed.)]. Schubert, C. 1974a. Late Pleistocene Merida gla- ciation, Venezuelan Andes. Boreas 3:147-152. Schubert, C. 1974b. Late Pleistocene glaciation of Paramo de la Culata. north-central Andes. Geol. Rundsch. 63:516-538. Schwerdtfeger, W. (ed.). 1976. World Survey of Climatology. Vol. 12. Climates of Central and South America. Elsevier, Amsterdam, 532 p. Shagam, R. 1975. The northern termination of the Andes, pp. 325—420 in Nairn, A. E. M.. Stehli, F. G. (eds.). The Ocean Basins and Margins, Vol. 3. The Gulf of Mexico and the Caribbean. Plenum Press, N.Y., 583 p. Sillitoe, R. H. 1974. Tectonic segmentation of the Andes: Implications for magnetism and metallog- eny. Nature 250:542-545. Simpson, B. B. 1973. Contrasting modes of evolu- tion in two groups nf Perezia ( Mutisieae; Com- positae) of southern South America. Taxon 22: 525-536. Simpson, B. B. 1975. Pleistocene changes in the flora of the high tropical Andes. Paleobiology 1:273-294. Simpson, B. B. In press. A revision of the genus Polijkpis (Rosaceae: Sanguisorbeae). Smithsoni- an Contrib. Bot. Simpson, B. B., Vervoorst, F. 1977. Physiographic settings of the study sites, pp. 16-25 in Orians, G. II., Solbric. O. T. (eds.). Convergent Evo- lution in Warm Deserts. Dowden, Hutchinson and Ross, Stroudsburg, Pa., 333 p. Smith, L. B. 1962. Origins of the flora of southern Brazil. Contrib. U.S. Natl. Herb. 35:21.5-249. Snelling, N. J. 1963. Age of the Roraima Forma- tion. British Guiana. Nature 198:1079-1080. Snow, J. W. 1976. The climate of northern South America, pp. 295-^404 in Schwerdtfegfr, W. (ed.). Climates of Central and South America. Elsevier, Amsterdam, 532 p. [Vol. 12 of World Survey of Climatology. H. E. Landsberg, ed. in Chief]. Steinmann, G. 1930. Geologia del Peru. Carl Win- ters, Heidelberg, 448 p. Tosi, J. A., Jr. 1960. Zonas de vida natural en el Peru. Inst. Interam. Cienc. Agric. OEA (Organ. Estados Unidos) Cent. Trop. Invest. Ensenanza Grad. Inf. Zona Andina. Bol. Tech. 5:1-271. Trewartha, G. T. 1961. The Earth's Problem Cli- mates. Univ. Wisconsin Press, Madison, 334 p. Troll, K. 1927. Yon Titikakasee zum Pooposee und zum Solar von Coipasa. Petennanns Mitt. 73: 218-222. 188 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Troll, K. 1959. Die tropischen Giberge. Bonn. Geogr. Abhandl. 25:1-93. Troll, K. 1965. Seasonal climates of the earth, pp. 19-25 in Rodenwaldt, E., Jusatz, H. J. (eds.). 1965 World Maps of Climatology. Springer- Ver- lag, New York, 2nd ed., 28 p. Turner, J. C. M. 1972. Cordillera Oriental, pp. 117- 142 in Leanza, A. F. (ed.). Geologia Regional Argentina. Acad. Nac. Cien., Cordoba, 869 p. van- der Hammen, T. 1974. The Pleistocene changes of vegetation and climate in tropical South Amer- ica. J. Biogeogr. 1:3-26. van der Hammen, T., Werner, J. H., Dommelen, H. van. 1973. Palynological record of the up- heaval of the northern Andes: a study of the Pliocene and lower Quaternary of the Colombian Eastern Cordillera and the early evolution of its high-Andean biota. Rev. Palaeobot. Palynol. 16: 1-122. Vuilleumier, B. S. 1971. Pleistocene changes in the fauna and flora of South America. Science 173: 771-780. Weberbauer, A. 1945. EI Mundo Vegetal de los Andes Peruanos. Minist. Agric, Lima, Peru, 766 p. Yrigoyen, M. R. 1972. Cordillera Principal, pp. 345-364 in Leanza, A. F. (ed.). Geologia Re- gional Argentina. Acad. Nac. Cien., Cordoba, 869 p. 8. The Amphibians of the Lowland Tropical Forests John D. Lynch School of Life Sciences University of Nebraska Lincoln, Nebraska 68508 USA Prior to the separation of South America from Gondwanaland and through much of the Cretaceous, most of the continent sup- ported a lowland tropical flora with a temper- ate flora to the extreme south (Savage, 1973). During the Cenozoic, tropical lowland floras were restricted because of the emergence of the Andean ranges and the expansion of arid and semiarid regions in western South Amer- ica (Axelrod, 1960). The Andes separated lowland forests of extreme northwestern South America from those of central South America, and the developing arid zones separated those of central South America from those of the southeast. As used here, "lowland tropical forests" re- fers to those forested environments from sea level to 1000 meters elevation. Amphibians distributed within the savannas occurring in forests are not considered forest elements. Thus, the tree frogs — Hyla crepitans, H. par- tialis, and H. raniceps — are ignored here, al- though each occasionally invades one or more of the major South American rainforests. On the other hand, the marine toad, Bnfo mari- nus, is relatively common in nonforested en- vironments in Central America and northern South America and also is distributed through the northern forests penetrating even those aseasonal forests with very high annual pre- cipitation (Choco and Napo regions). Although I have not always segregated the evergreen forests from those that are season- ally dry, I have pointed out the subdivision within each major forest. (The subdivisions usually are correlated with seasonal or asea- sonal forests.) Those forests that have pro- nounced and prolonged dry seasons are not included here. As a general rule, such forests are inhabited by nearly the same suite of species found in adjacent nonforest environ- ments (cerrado, llanos). The greatest diffi- culty (and least confidence) was to extract estimates of approximate altitudinal distribu- tions for the many amphibians closely asso- ciated with the major mountain systems in South America (Andes, Guiana Shield, Bra- silian Shield, and coastal ranges). Published data are especially scant for the diverse fauna of southeastern Brasil. Several previous biogeographic essays (e.g., Parker, 1935; Lutz, 1972; Lescure, 1975) on the Amazonian herpetofauna have dealt with members of both forested and nonfor- ested assemblies in preliminary fashions. The current effort suffers from some of the same limitations imposed on previous efforts, name- ly, incomplete distributional data (but prob- ably not so incomplete as suggested by Heyer, 1976) and indefinite phylogenetic constructs. The South American rainforests exist as four or five disjunct elements (Fig. 8:1). The most southern, the Austral Forests, is tem- perate and very distinct faunistically, sharing only one genus (Bafo) and no species with the other forests. The northern tropical for- ests are tenuously connected, chiefly by gal- lery forests and forest islands distributed through the nonforest areas separating them. The Trans-Andean Forests are isolated from the vast Central Cis-Andean Forests by the northern Andes but are connected weakly via the Northern Forests of Colombia and Vene- zuela; the latter are not entirely discrete (ten- uous connections through the forests fringing the Merida Andes and Venezuelan llanos). The largest of the forests ( approximately 80$ of all tropical South American forests) is the Central Cis-Andean forest ( Amazonian or Hijlaea) drained by the Rio Amazonas and its tributaries. The central forests become drier southeastwardly and grade into the cer- rado on the northwestern face of the Bra- silian Shield. A less marked dry belt ( Reinke's Corridor) is described by numerous, isolated and generally small savannas extending from IV) 190 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 NORTHERN so TRANS- /JV/T A N D E A N 3 7/L_^\^ 17 j^^, <\ Z^^/^ATLANTIC AUSTRAL*] r' VIP Fic. 8:1. Rainforests of South America. The fine- stippled forests are tropical lowland forests. The coarse-suppled forest is temperate ( after Hueck, 1966, and Schmithusen, 1968). Pluviosclvas sudamericanas. El acliurado fino muestra a las selvas tropicales de tierras bajas. El acliurado gmeso a las selvas temperadas (de Hucck, 1966, y Schmithusen, 1968). Amazonian Venezuela southeastward to the mouth of the Rio Amazonas. The central forests are tenuously connected to the At- lantic Forests of Brasil by gallery forests and forest islands in southern Brasil. The four tropical lowland forests (Trans- Andean, Northern, Central Cis-Andean, and Atlantic) harbor 530 species of forest Am- phibia, 95 percent of which are restricted to one of the four forests (Fig. 8:2). Only 26 species are shared by two or more forests and only five (all frogs) — Bufo typhonius (Bufonidae), Hyla rubra and Phrynohyas venulosa (Hylidae), Leptodactylus wagneri (Lcptodactylidae), and Rami palmipes (Ran- idae) — occur in all four forests. Of the 126 species found in Trans-Andean Forests, 88 percent are endemic; 56 percent of the 39 species found in Northern Forests are en- demic; 90 percent of the 225 species found in Central Cis-Andean Forests are endemic; and 92 percent of the 183 species of the At- lantic Forests are endemic. Ill J) f® I22 J^=> 203 of t 0 |/ 53-17 109-65 Fie. 8:9. Amphibian species densities and en- demicities in subunits of the Atlantic Forests. The hyphenated numbers refer to Total amphibians — Endemic amphibians. Densidad de las especics anfibias y endemismos en las subunidades de las selvas atldnticas. Los nu- meros equivalen a Anfibios total — Anfibios endemicos. even in the event that these five species are deleted, the endemism in the Sao Paulo and Atlantic forests remains at the same order of magnitude. A large measure of the endem- ism of the Sao Paulo Unit can be ascribed to the distributions of the many montane rain- forest species found on the several isolated serras of southeastern Brasil at elevations be- tween 400 and 1200 m (Lutz, 1973). The contiguous lowland forest (below 400 m) acts as a barrier to broader distributions, and vicars occur on narrowly separated serras, thus amplifying the faunistic distinction of the region (Lutz, 1973). Therefore, amphib- ians tend not to support Muller's (1973) sub- division of these forests into three subcenters. DISCUSSION Perusal of the distributional data for the tropical lowland forest amphibians reveals a pattern by which allied species are allopatri- cially (or parapatrically ) distributed. The theme seems to be applicable to many, if not most, of the groups of forest amphibians (Fig. 8:10). Such a pattern may be ecologi- cally and/or historically mediated. Haffer (1969, 1974, this volume) suggested that the pattern may be a function of repeated con- traction of forests into forest islands during the cyclic climatic phases of the Pleistocene. Duellman (1972), Duellman and Crump (1974), and Silverstone (1975) employed the model to explain distributional patterns in two groups of hylid frogs and some dendro- batid frogs, respectively. Clearly, Pleistocene variations in moisture patterns can be argued to have had dramatic effects on forest environments, and if those forest environments expand and contract alternatively, the forest amphibians will be subject to waves of dispersal and retreat with concomitant likelihoods of swamping and iso- lation. The model can be employed readily to explain high species densities in some of the forests [operating in essentially the same fashion as Pianka's (1969) model for lizard diversities]. An Ecological Explanation Rainforests require as little as 1,800 mm of annual precipitation if that precipitation is unform over time (Richards, 1952), or, if some months are dry, more rainfall is re- quired. The so-called aseasonal forests usual- ly receive 3000 mm annual precipitation (or more), whereas rainforests existing within monsoon climates have pronounced dry sea- sons (one to three months with 0-60 mm rainfall ) . These extremes present dramatically different environments to forest-adapted am- phibians. Amphibians exhibit an impressive array of reproductive modes in tropical South American forests (Table 8:5). Crump (1974) listed ten modes for frogs at Santa Cecilia in eastern Ecuador (Fig. 8:11). These ten modes range from the most common ( among amphibians as a whole, Mode 1) wherein eggs are deposited in aquatic situations where the larvae undergo development, to the less common and more spectacular ( Modes 8, 9, 10), in which development is direct (no tad- pole ) and the eggs terrestrial or carried about on specialized regions of the back of the adult 1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 203 Fig. 8:10. Parapatric distribution patterns. A. Bufo guttatus group (3 species): Chocoan B. blombergi, Napoan B. glaberrimus, and Guianan B. guttatus. B. Leptodactylus fuscus group (5 species): Amazonian L. sp. "A," Guianan L. longirostris, Pernambucan L. mystaceus, Paulistan L. sp. "N," and Chocoan L. vcntrimac- ulatus. C. Eleutherodactylus fitzingeri group (7 species): Chocoan E. achatinus, Guianan E. chiastonotus, Na- poan E. conspicillatus (open hatching), Madeiran E. fenestrates (fine hatching), Ucayalian E. peruvianus (coarse stipple), Venezuelan E. terraebolivaris and supra-Amazonian E. vilarsi (fine stipple). D. Hatching (Eleutherodactylus sulcatus group, 2 species): Napo- Ucayalian E. sulcatus; Venezuelan E. maussi. Stipple (Dendrobates tinctorius group, 4 species): Chocoan D. auratus, Nechian D. truncatus (coarse stipple), Guia- nan D. tinctorius, and Paran D. galactonotus. Sources: A.) Cei (1972); B.) W. R. Heyer (pers. comm.); C.) Lynch (in prep.); D.) Lynch (1975), Silverstone (1975). Patrones dc distribucion parapdtricas. A. Grupo de Bufo guttatus (3 especies): B. blombergi del Choco, B. glaberrimus del Napo, B. guttatus guy ones. B. Grupo de Leptodactylus fuscus (5 especies): L. sp. "A" del Amazonas, L. longirostris de Guyana, L. mystaceus dc Pemamhuco, L. sp. "N" Paulistano, y L. ventrimacu- latus del Choco. C. Grupo Eleutherodactylus fitzingeri (7 especies): E. achatinus del Choco, E. chiastonotus de Guyana, E. conspicillatus dc Napo (lincado amplio), E. fenestratus de Madeira (lineado fino), E. peruvianus del Ucayali (achurado grueso), E. terraebolivaris de Venezuela y E. vilarsi del Alto Amazonas (achurado fino). D. Lineado (Grupo de Eleutherodactylus sulcatus, 2 especies): Napo-Ucayali E. sulcatus; Venezuela, E. maussi. Achurado (Grupo de Dendrobates tinctorius, 4 especies): Choco, D. auratus; Ncchi, D. truncatus (achurado grueso): Guyana, D. tinctorius; y Para, D. galactonotus. Fuentes: A.) Cci (1972); B.) W. R. Heyer (pers. comm.); C.) Lynch (en prep.); D) Lynch (1975), Silverstone (1975). 204 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Table 8:5. — Distribution of Reproductive Modes Among Tropical Lowland Forest Amphibians. Reproductive Mode 8 9 10 Plethodontidae B rachy cephalidae Bufonidae Centrolenidae _ Dendrobatidae __ Amphignathodontinae Hemiphractinae Hylinae Phyllomedusinae Batrachylini Ceratophryinae Eleutherodactylini Elosiinae Grypiscini Leptodactylinae Odontophynini _ Microhylidae __ Pipidae _. Ranidae X X1 X X X X X X X X X X X X X .... X .... .... X X2 X On land or in water ( Lutz, 1931 ). ! On land, live in decomposing jelly mass (Lutz, 1929). (hemiphractine hylids and pipids). The modes are as follows (from Crump 1974:9- 10): "(1) eggs deposited in ditches, puddles, swamps, ponds, lakes, and streams, with free- swimming aquatic larvae; (2) eggs deposited in tree cavity above ground, with free-swim- ming aquatic larvae; (3) eggs deposited in basin constructed on ground by male, with free-swimming aquatic larvae; (4) eggs de- posited on vegetation above water, with free- swimming aquatic larvae (tadpoles hatch and fall into water ) ; ( 5 ) eggs deposited in foam nest on or near water, with free-swimming aquatic larvae; (6) eggs deposited on land, with free-swimming aquatic larvae (tad- poles carried to water on dorsum of adult ) ; (7) eggs deposited in foam nest on land and larvae develop within foam; (8) eggs de- posited out of water, with direct develop- ment; (9) eggs carried in depressions on dor- sum of aquatic female, with direct develop- ment; (10) eggs carried in depressions on dorsum of terrestrial female, with direct de- velopment." Sedation of the modes by Crump was along a dimension attempting to measure parental investment. If the modes are re- ordered along a dimension attempting to de- scribe dependence on forest-mediated envi- ronments, the order is as follows: least dependent -> most dependent (1,2,3,9) (5) (4) (7) (6) (8, 10). Those depositing eggs directly in water (with subsequent development occurring there as well ) are less dependent on the high humidity provided by a forest than are those depositing eggs in terrestrial environments (where all needed water is extracted from the air). The truth and rationale of this generalization is supported by casual inspection of the range of forested and nonforested environments in which amphibians having various reproduc- tive modes occur. Amphibians of Mode 1 oc- cur throughout the range of habitats occupied by amphibians (nonforests to aseasonal for- ests). Modes 2 and 3 are exhibited by so few amphibian species that no correlation can be advocated; I chose to treat each as minor variations of Mode 1. Likewise, Mode 9 is found in so few species as to challenge any generalization but, because these species occur in water, they might not be considered restricted to forest environments per se. Frogs exhibiting Mode 5 are perhaps more charac- teristic of nonforested habitats than forested ones but range throughout the habitat pro- file. Heyer (1969) examined some of the variations on the theme of Mode 5. Seem- 1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 205 /FOREST MODES FOREST MODES NON-FOREST MODES Fig. 8:11. Reproductive modes among rainforest anurans. Mode 9 (Pipa with direct development) and Mode 10 ( amphignathodontine and hemiphractine hylids, direct development) not shown. Modes 5 and 7 involve foam nests; eggs are laid on vegetation in Mode 4 (after Crump, 1974). Modos reproductivos de los anuros de la pluviselva. A los Modo 9 (Pipa con desarrollo directo) y 10 (hyli- dos amphignathodontinos y hemiphractinos, desarrollo directo) no se los muestra. Modos 5 y 7 tienen nidos de espuma; los huevos son puestos en la vegetation en el Modo 4 (de Crump, 1974). ingly greater risks (and concomitantly, greater dependency) are taken by those am- phibians depositing their eggs out of water (Modes 4-8, 10). The larvae of amphibians having Modes 5 and 7 presumably secure some protection from the foam. Those spe- cies employing Mode 4 (eggs on vegetation over water, larvae in water) have a more obvious association with wet environments (and therefore, more frequently forests) but also occur in nonforested habitats (llanos, scrub ) . Anurans characterized by Modes 6, 7, 8, and 10 employ terrestrial eggs of various durations. Mode 10 constitutes a special case in that the embryos are carried about on the parent's back and moisture requirements may apply more to the adult than the embryos. Species with Mode 6 have terrestrial eggs which, upon hatching, yield tadpoles that pass some time on the parent's back but eventually are transported to bodies of water for "normal'' development. Terrestrial eggs and embryos resulting in miniature replicas of the adult produced without submersion in water characterize Modes 7 and 8. If the asserted sequence is correct, we should expect to discover different patterns of distribution ( and patterns of endemism if Haffer's model is real) among different re- productive modes. Amphibians exhibiting Mode 1 should exhibit less endemism than those modes characterizing higher forest- fidelity (e.g., 6 and 8). Modes less dependent 206 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 13/8 1/1 16/15) Fig. 8:12. Species densities and endemicities of am American rainforest units. The first number is the rami lowing the solidus is the number of endemic species of 5, Leptodactylidae (except Adenomera). C. Mode 4, Dendrobatidae ( the numbers in parentheses in southeast dactylini ( the numbers in southeastern Brasil refer to Densidad dc especies y endemismos de varios atrftb osas de Sud America. El primer numero es el numero la linca diagonal representa el numero de especies ende Modo 5, Leptodactylidae (exccpto Adenomera). C. M Modo 6, Dendrobatidae (los numeros cntre parentesis 8, Eleuthcrodactylini (los numeros en parentesis en el on special attributes of forest environments (high humidity) exhibit lower percentages of endemism than do forest-adapted modes over the same terrain (Fig. 8:12). Conversely, if the model is more realistic, we ought to expect the opposite pattern among the nonendemic taxa of amphibians found in South American lowland tropical forests (Table 8:1). We find departure from the expected numbers of species of each re- productive mode in that certain reproductive modes are "over-represented" or "under-repre- sented" among the nonendemic taxa (Table phibians having various reproductive modes in South ber of species of the group in the area; the number fol- the group in the area. A. Mode 1, Hylinae. B. Mode Centrolenidae, Hylinae, Phyllomedusinae. D. Mode 6, ern Brasil refer to Elosiinae). E. Mode 8, Eleuthero- Grypiscini). ios con diversos modos reproductivos en las selvas lluvi- i/c especies del grupo en el area; el numero siguiente a micas del grupo en el area. A. Modo 1, Hylinae. B. odo 4, Centrolenidae, Hylinae, Phyllomedusinae. D. en el sudestc del Brasil refieren a Elosiinae). E. Modo sudestc del Brasil refieren a Grypiscini). 8:6). As might be predicted, modes with high-forest fidelity (6-8, 10) are collectively "under-represented," whereas the modes of low-fidelity ( 1-5 ) are "over-represented." Such "under-representation" and "over- representation" are also consistent with the notion that r-selected species tend to be better colonizers than K-selected species. Crump's (1974) seriation of modes tends to describe one dimension of the r- and K-continuum; in general, frogs of lower numbered modes are more frequently r-selected species and those of high numbered modes K-selected species 1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 207 Table 8:6. — Frequencies of Frogs and Salamanders of Tropical Lowland Forests Exhibiting 10 Repro- ductive Modes Compared to Frequencies for Non- endemic Species. Number Species Number c if Nonendemic Species Mode Observed Expected" Deviation 1 184 12 9.2 + 2.8 2 7 0 0.4 -0.4 3 3 1 0.2 + 0.8 4 65 4 3.2 + 0.8 5 30 5 1.5 + 3.5 6 49 + 16" 1 3.2 -2.2 7 6 1 0.3 + 0.7 8 91 + 13c 0 5.2 -5.2 9 5 0 0.2 -0.2 10 22 0 1.1 -1.1 * 5% of the 480 species are nonendemic. "Expected" values were computed by multiplying Number of species of a mode by 5%. b Dendrobabds + elosiine leptodactylids. If elosiines are included in Mode 1 (see Table 5), the expected value for Mode 6 is 2.4 and the deviation —1.4; in that case, the expected for Mode 1 would be 10.0 and the deviation +2.0. c Eleutherodactylines and plethodontids + grypiscine leptodactylids. (as judged from clutch size and/or egg size; Salthe and Duellman, 1973). Historical Explanations The patterns of distribution and endemism among the South American, lowland forest amphibians seem explicable on an ecological basis. An historical component is implied by noting that the centers of endemism corre- spond well with the refugia postulated in Haffer's model. That species distributions cor- respond with a Pleistocene model does not require that the implied speciation be Pleis- tocene in age, only that at least part of the existing distribution pattern is a product of relatively recent perturbations. Savage ( 1973 ) envisioned the late Meso- zoic-Early Cenozoic amphibian fauna of South America as consisting of three tropical groups (caecilians, microhylid frogs, and pipid frogs ) and two south temperate groups (ascaphid and leptodactyloid frogs). By the Paleocene, bufonid, hylid, and leptodactylid frogs had invaded the tropical zone (Estes and Reig, 1973; Savage, 1973). Subsequent to the Ncogenc emergence of the Panamanian Isthmus, bolitoglossine salamanders and ranid frogs invaded South America from the north. The stem leptodactylids all radiate from southern South America (Lynch, 1971, 1978; Savage, 1973), and most do not range appre- ciably north (Lynch, 1971, 1978). The roles of the old highland areas in the diversification of the amphibian fauna are obscure at best. The occurrence of a batrachyline leptodacty- lid frog (Thoropa), elosiine leptodactylid frogs, and grypiscine leptodactylid frogs asso- ciated with the Brasilian Highlands suggest the once important role of that highland mass in the Cenozoic radiation of leptodactylids on the continent. The advanced groups (dendro- batid and eleutherodactyline leptodactylid frogs) are not represented in southern South America and only poorly represented in the area of the Brasilian highlands, but they are dominant amphibian groups in the northern Andes and the adjacent lowland tropical forests. The general absence of stem groups of amphibians associated with the Brasilian and/ or the Guiana highlands and the youth of the Andes, compels most observers to con- clude that the complex fauna was lowland in origin. Because the lowlands were once more equable than they are today, and because few groups persist in the temperate regions of southern South America, it may be concluded that most amphibian groups originated in the lowland rainforests. The following apparently originated and/or diversified in the lowland rainforests of South America: Caeciliidae Rhinatrematidae0 Typhlonectidae Brachycephalidae Dendrobatidae" Ilemiphractinae Hylinae Eleutherodactylini° Elosiinae Grypiscini Microhylidae Pipinae Those marked with an asterisk were greatly affected by the uplift of the northern Andes. The following invaded the lowland tropi- cal forests of South America from adjacent areas (Central America or nonforested en- vironments ) : Plethodontidae Bntonidae Batrachylini Leptodaetylinae Odontophrynini Ceratophryinae Ranidae Amphignathodontine and phyllomedusine hy- 208 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 lids and centrolenid frogs (and possibly cae- ciliid and hyline stocks ) do not fall into either pattern. Amphignathodontines and centrolen- ids are recent invaders of Central America (Savage, 1973) but are usually found in up- lifted areas (discontinuous distributions). Phyllomcdusines are much better represented in Central America (Agalychnis and Pachy- medusa) and are essentially replaced in South America by Phyllomedusa; the replace- ment is suggestive of geographic separation of the two forest stocks during much of the Tertiary. Some of these amphibian groups were iso- lated and radiating in Central America during much of the Tertiary; Savage (1973) cited bufonids, hylids (hylines and phyllomedu- sines), microhylids, and eleutherodactyline leptodactylids. Savage and Wake (1972) identified the Middle American caeciliids iso- lated in Central America during the same time period. Most of these isolates are recog- nizable in that the species groups or genera are essentially endemic to Central America and extend only into the Choco in South America. Savage ( 1973 ) suggested that Eleu- therodactijlus and some derived genera under- went an Eocene-Miocene radiation in Middle America and subsequent to the uplift of the Panamanian Portal radiated into South Amer- ica. My study of Eleutherodactylus does not confirm such a conviction. Ancestral relicts (Hylactophryne and Ischnocnema) persist at opposite ends of the center of Eleutherodac- tylus diversity and the wealth of species are distributed in the northern Andes. The spe- cies groups in eastern and southeastern Brasil are apparently primitive stocks, as is the Eleutherodactylus discoidalis group (distrib- uated along the eastern edge of the Andes from northern Argentina to southern Colom- bia). The Middle American Eleutherodacty- lus include representatives of two groups, the endemic alfredi group in Mexico and Guatemala and the fitzingeri group, which is distributed nearly as widely as are mainland Eleutherodactylus (Mexico to Bolivia and eastern Brasil), as well as South American groups in lower Central America (hiporcatus and unistrigatus groups; Lynch, 1976). Some eleutherodactyline frogs certainly invaded Middle America prior to the Neogene uplift of the Panamanian Isthmus, but most of the radiation of the group occurred in the West Indies and in northern South America. The development of biogeographic con- structs (historical biogeography) without some defensible phylogenetic constructs is not science (see Rosen, 1975, for philosophy and methodology). Unfortunately, the sorts of phylogenetic input required to generate ro- bust biogeographic constructs are not avail- able for most lowland Amphibia. In spite of the impressive strides made in the past decade in tabulating systematic data for some of the larger amphibian groups in South America, our knowledge of the relationships within those groups remain regretfully nascent. This is less a critique of previous work than a statement of what remains to be done for most groups of tropical amphibians and rep- tiles. I do not believe that we have the systematic data necessary to generate formu- lations of how the tropical lowland forest amphibian groups developed in time and space. ACKNOWLEDGMENTS Most of the distributional data presented here represents summaries of unpublished data collected by colleagues. My thanks go to William E. Duellman, M. J. Fouquette, Jr., W. Ronald Heyer, Marinus S. Hoogmoed, and Jean Lescure for sharing data. Gathering distributional data for eleutherodactyline frogs was facilitated by provision of working space by the curators at the American Mu- seum of Natural History, British Museum (Natural History), National Museum of Na- tural History, The University of Kansas Mu- seum of Natural History, and The University of Michigan Museum of Zoology, and finan- cial support from The University of Nebraska Research Council. I owe a major debt to several colleagues for serving as sources of stimulation and criti- cism; special thanks are extended to William E. Duellman, Linda Trueb, and Charles F. Walker. I defer to identify the many others to avoid failing to recall some. 1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 209 RESUMEN Las selvas tropicales de tierras bajas (plu- viselvas, bajo 1000 m) de Sud America con- forman cuatro unidades — 1) Las selvas transandinas (Choco, Nechi, y Magdalena); 2) Las selvas cisandinas centrales (Amazonia o Hylaea); 3) Las selvas nortinas (Santa Marta, Cuenca de Maracaibo, y la costa de Venezuela ) ; y 4 ) Las selvas atlanticas ( Este y sudeste del Brasil). Estas cuatro regiones forestales contienen 530 especies de anfibios pertenecientes a 13 familias (una de salaman- dras, tres de cecilios, y nueve de anuros). El endemismo es pronunciado; solo 26 especies (4.9%) son compartidas por dos o mas regiones. Las selvas nortinas son las mas pobres en especies (39) y las menos distintas (56% de endemismo). Las selvas atlanticas contienen 183 especies (92% de endemismo), las selvas centrales 225 (90% de endemismo), y las selvas transandinas 126 (88% de ende- mismo). Las ranas hylidas son dominantes en las cuatro regiones selvaticas (atlantica, 38%, cis- andina central 37%, nortina 26%, y transandina 25%). Los leptodactylidos de las selvas cen- trales, nortinas, y transandinas son especial- mente eleutherodactylinos y leptodactylinos mientras que los de las selvas adanticas son primariamente elosiinos, grypiscinos, lepto- dactylinos, y odontophryninos. Ranas den- drobatidas son componentes prominentes de las selvas centrales, nortinas, y transandinas pero estan sustancialmente ausentes en las selvas atlanticas. Cuatro areas de un marcado endemismo se reconocen — 1) Choco sudamericano, 2) la cuenca del Amazonas superior (drenajes del Napo-Ucayali ) , 3) las Guayanas, y 4) la costa brasileira (area de Rio de Janeiro-Sao Paulo). Un cuarenta y cuatro por ciento (233) de todas las especies forestales de la fauna an- fibia sudamericana estan restringidas a estas cuatro regiones. Varios grupos de anfibios presentan un modelo distribucional congruente con la teoria de refugios forestales de Haffer. Tal modelo consistente en oleadas de avance y retroceso de las selvas, puede ser utilizado para explicar la alta densidad de especies ob- servada en algunos grupos. Sin embargo, los anfibios estan pobremente representados en algunos de los refugios propuestos. Esta baja representation puede reflejar profundas difer- encias entre selvas separadas por una esta- cionalidad en el regimen pluvial. A diferen- cia de los amniotas o vertebrados acuaticos, la mayoria de los anfibios tratan de ser ter- restrial sin tener los mecanismos adecuados para conservar agua. La falta de estos me- canismos se hace mas notoria en las estrate- gias reproductivas de los anfibios. La alta densidad de especies (y endemis- mo) de los anfibios se correlaciona bien con las areas de una alta y bien distribuida lluvia, sin estacionalidad. En regiones forestales, tales lluvias proveen con una serie de micro- habitats de elevada humedad y de larga dura- tion, ambas caracteristicas necesarias para las mas fragiles modalidades reproductivas pre- sentes en los anfibios ( desarrollo directo, hue- vos y larvas terrestres). Tales ambientes no reducen necesariamente el exito de los mas comunes modos reproductivos como son los huevos y larvas desarrollandose en agua. La importancia de la modalidad reproduc- tiva en sentido de proveer restricciones o aumentos de la capacidad de dispersion de los anfibios selvaticos esta demonstrada por la observation de que una mayor proporcion de especies con modo 1 y modo 5 son no- endemicas que lo que pudiera predecirse y que una menor proporcion de especies con modo 6 y modo 8 son no-endemicas que lo que pudiera predecirse si la ausencia de endemismo y el modo reproductivo no estu- viasen relacionados. Si la sensibilidad a la estacionalidad de las lluvias es un indice apropriado de fidelidad forestal, entones aquellos grupos que poseen los modos repro- ductivos mas sensibles adquieren importancia para la inferencia de eventos climaticos y biogeograficos del pasado; estos incluyen a las salamandras bolitoglossinas, y ranas den- drobatidads y eleutherodactylinas. Las areas de tierra altas han tenido solo un limitado papel en el origen y dispersion de los anfibios de selvas tropicales de tierras bajas. Las alturas del sudeste brasileno pare- cer haber servido como un islote de paso a la radiation de leptodactylidos fuera de la 210 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Patagonia. Los Andes del norte han contri- buido a una fragmentation de ciertos grupos (ranas dendrobatidads y eleutherodactylinas, y cecilios rhinatrematidos ) . Los cecilios rhinatrematidos y typhlonec- tidos y los anuros amphignathodontinos, brachycephalidos, centralenidos, dendroba- tidos, la mayoria de los eleutherodactylinos, elosiinos, grypiscinos, hemiphractinos, y pipi- dos se originaron y diferenciaron al parecer en las selvas tropicales de tierras bajas de Sud America. Ciertos otros grnpos (cecilios caeciliidos, algnnos bufonidos, hylinos, micro- hylidos, y phyllomedusinos, y posiblemente algunos eleutherodactylinos) se originaron en estas selvas pero irradiation en Centro Amer- ica tambien. Las salamandras plethodontidos y los anuros ranidos son invasores recientes desde las selvas de Centro America, y los anuros batrachylinos, la mayoria de los bufo- nidos, ceratophryinos, leptodactylinos, y odontophryinos son invasores de ambientes surefios temperados y subtropicales. LITERATURE CITED Axelrod, D. I. 1960. The evolution of flowering plants, pp. 227-305 Ml Tax, S. (ed.). Evolution after Darwin. I. Univ. Chicago Press, 629 p. Blair, W. F. (ed.). 1972. Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Brame, A. H., Jr. 1968. Systematics and evolution of the Mesoamerican salamander genus Oedipina. J. Herpetol. 2:1-64. Brame, A. H., Jr., Wake, D. B. 1963. The sala- manders of South America. Nat. Hist. Mus. Los Angeles Cry. Contrib. Sci. (69): 1-72. Brame, A. H., Jr., Wake, D. B. 1972. New species of salamanders ( genus Bolitoglossa ) from Co- lombia, Ecuador, and Panama. Ibid. (219): 1-34. Carvalho, A. L. de. 1954. A preliminary synopsis of the genera of American microhylid frogs. Occas. Pap. Mus. Zool. Univ. Michigan (555): 1-20. Cei, J. M. 1972. Bufo of South America, pp. 82-92 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Cochran, D. M. 1955. Frogs of southeastern Brazil. Bull. U.S. Natl. Mus. (206): 1-423. Cochran, D. M., Goin, C. J. 1970. Frogs of Co- lombia. Ibid. (288): 1-655. Crump, M. L. 1974. Reproductive strategies in a tropical anuran community. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (61): 1-68. Duellman, W. E. 1970. The hylid frogs of Middle America. Univ. Kansas Mus. Nat. Hist. Monogr. (l):l-753. Duellman, W. E. 1971a. A taxonomic review of South American hylid frogs, genus Phrynohyas. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (4): 1-21. Duellman, W. E. 1971b. The identities of some Ecuadorian hylid frogs. Herpetologica 27:212- 227. Duellman, W. E. 1972. South American frogs of the Hyla rostrata group ( Amphibia, Anura, Hy- lidae). Zool Meded. Rijks Mus. Nat. Hist. Leiden 47:177-192. Duellman, W. E. 1973. Frogs of the Hyla geo- graphica group. Copeia 1973 (3):515-533. Duellman, W. E. 1974. A reassessment of the taxo- nomic status of some Neotropical hylid frogs. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (27): 1-27. Duellman, W. E. 1977. Liste der Rezenten Am- phibien und Reptilien. Hylidae, Centrolenidae, Pseudidae. Das Tierreich 95:1-225. Duellman, W. E., Crump, M. L. 1974. Speciation in frogs of the Hyla parviceps group in the upper Amazon Basin. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (23): 1-40. Dunn, E. R. 1948. American frogs of the family Pipidae. Amer. Mus. Novit. (1384): 1-13. Dunn, E. R. 1949. Notes on South American frogs of the family Microhylidae. Ibid. (1419):1-21. Estes, R„ Reig, O. A. 1973. The early fossil record of frogs: A review of the evidence, pp. 11-63 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Gallardo, J. M. 1961. On the species of Pseudidae (Amphibia, Anura). Bull. Mus. Comp. Zool. Har- vard Univ. 125:111-134. Haffer, J. 1969. Speciation in Amazonian forest birds. Science 165:131-137. Haffer, J. 1974. Avian speciation in tropical South America with a systematic survey of the Toucans ( Ramphastidae ) and Jacamars (Galbulidae). Publ. Nuttall Omithol. Club, Cambridge (14): 1-390. Heyer, W. R. 1969. The adaptive ecology of the species groups of the genus Leptodactylus (Am- phibia, Leptodactylidae). Evolution 23:421—128. Heyer, W. R. 1970. Studies on the frogs of the genus Leptodactylus (Amphibia: Leptodactyli- dae). VI. Biosystematics of the melanonotus group. Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (191): 1-48. Heyer, W. R. 1973. Systematics of the marmoratus group of the frog genus Leptodactylus (Amphibia, Leptodactylidae). Ibid. (251): 1-50. Heyer, W. R. 1975. A preliminary analysis of the intergeneric relationships of the frog family Lep- todactvlidae. Smithsonian Contrib. Zool. (199): 1-55. ' Heyer, W. R. 1976. Notes on the frog fauna of the Amazon Basin. Acta Amazonica 6:369—378. 1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 211 Heyer, W. R., Liem, D. S. 1976. Analysis of the intergeneric relationships of the Australian frog family Mvobatrachidae. Smithsonian Contrib. Zool. (233): 1-29. Hueck, K. 1966. Die Walder Siidamerikas. Fischer Verlag, Stuttgart, 422 p. Izecksohx, E. 1971. Novo genero e nova especie de Brachycephalidae do Estado do Rio de Janeiro, Brasil. Bol. Mus. Nac. Rio de Janeiro Zool. (280): 1-12. Izecksohx, E. 1976a. O status sistematieo de Phryn- iscus proboscideus Boulenger ( Amphibia, Anura, Bufonidae). Rev. Bras. Biol. 36:341-345. Izecksohx, E. 1976b. Una nova especie de Pipa, do estado do Amazonas, Brasil (Amphibia, Anura, Pipidae). Ibid. 36:507-510. Lescure, J. 1973. Notes biogeographiques sur quel- ques Amphibiens du bassin superior du Maroni. C. R. Seances Soc. Biogeogr. (439):58-63. Lescure, J. 1975. Biogeographie et ecologie des Amphibiens de Guyane Francaise. Ibid. (440): 68-82. Lescure, J. 1976. Contribution a l'etude des Am- phibiens de Guyane francaise. VI. Liste pre- liminaire des Anoures. Bull. Mus. Natl. Hist. Nat. Zool. Paris (265): 475-525. Lutz, A. 1929. Taxonomia e biologia do genero Cyclorhamphus. Mem. Inst. Oswaldo Cruz 22(1): 1-25. Lutz, A. 1931. Observacoes sobre batrachios brasi- leiros. Taxonomia e biologia des elosiinas. Ibid. 24(4): 195-222. Lutz, B. 1972. Geographical and ecological notes on Cisandine to Platine frogs. J. Herpetol. 6: 83-100. Lutz, B. 1973. Brazilian species of Hyla. Univ. Texas Press, Austin, 265 p. Lynch, J. D. 1971. Evolutionary relationships, oste- ology, and zoogeography of leptodactyloid frogs. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): 1-238. Lynch, J. D. 1973. The transition from archaic to advanced frogs, pp. 133-182 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contempo- rary research on major problems. Univ. Missouri Press, Columbia, 470 p. Lynch, J. D. 1975. The identity of the frog Eleu- therodactylus conspicillatus (Gunther), with de- scriptions of two related species from northwest- ern South America. Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (272): 1-19. Lynch, J. D. 1976. The species groups of the South American frogs of the genus Eleutherodactylus ( Leptodactylidae ) . Univ. Kansas Mus. Nat. Hist. Occas. Pap. (61): 1-24. Lynch, J. D. 1978. A reassessment of the telmato- biine leptodactylid frogs of Patagonia. Ibid. (72): 1-57. Lynch, J. D., Hoocmoed, M. S. 1977. Two new species of Eleuthodactylus (Amphibia: Lepto- dactylidae) from northeastern South America. Proc. Biol. Soc. Washington 90:424-439. Lynch, J. D., Myers, C. W. [in preparation]. The frogs of the fitzingeri group of Eleutherodactylus in the Chocoan lowlands. McDiarmid, R. W. 1971. Comparative morphology and evolution of frogs of the Neotropical genera Atelopus, Dendrophryniscus, Melanophryniscus, and Oreophrvnella. Nat. Hist. Mus. Los Angeles Cty. Sci. Bull. (12): 1-66. Moore, J. A. 1944. Geographic variation in Rana pipiens Schreber of eastern North America. Bull. Amer. Mus. Nat. Hist. 82:345-370. Moore, J. A. 1975. Rana pipiens — the changing par- adigm. Amer. Zool. 15:837-849. Muller, P. 1973. Dispersal centres of terrestrial vertebrates in the Neotropical Realm. Biogeo- graphica, 2, Junk, The Hague, 244 p. Muller, P. 1974. Aspects of zoogeography. Junk. The Hague, 208 p. Noble, G. K. 1931. The biology of the Amphibia. McGraw-Hill Book Co., New York, 577 p. Nussbaum, R. A. 1977. Rhinatrematidae: A new family of caecilians (Amphibia: Gymnophiona). Occas. Pap. Mus. Zool. Univ. Michigan (682): 1-30. Pace, A. E. 1974. Systematic and biological studies of the leopard frogs (Rana pipiens complex) of the United States. Misc. Publ. Mus. Zool. Univ. Michigan (148): 1-140. Parker, H. W. 1934. A monograph of the frogs of the family Microhylidae. Brit. Mus. (Nat. Hist.), London, 208 p. Parker, H. W. 1935. The frogs, lizards, and snakes of British Guiana. Proc. Zool. Soc. London 1935 (3):505-530. Peters, J. A. 1973. The frog genus Atelopus in Ecuador (Anura: Bufonidae). Smithsonian Con- trib. Zool. (145): 1-49. Pianka, E. R. 1969. Habitat specificity, speciation, and species diversity in Australian desert lizards. Ecology 50:498-502. Richards, P. W. 1952. The tropical rain forest. Cambridge Univ. Press., Cambridge, 450 p. Rivero, J. A. 1961. Salientia of Venezuela. Bull. Mus. Comp. Zool. Harvard Univ. 126:1-207. Rosen, D. E. 1975. A vicariance model of Caribbean biogeography. Syst. Zool. 24:431—464. Salthe, S. N., Duellman, W. E. 1973. Quantitative constraints associated with reproductive modes in anurans, pp. 229-249 in Vial, J. L. (ed.). Evo- lutionary biology in the anurans: Contemporary research on major problems. Univ. Missouri Press, Columbia, 470 p. Savage, J. M. 1968. The dendrobatid frogs of Cen- tral America. Copeia 1968(4) : 745-776. Savage, J. M. 1973. The geographical distribution of frogs: Patterns and predictions, pp. 351—145 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Savage, J. M., Wake, M. 1972. Geographic varia- tion and systematica of the Middle American cae- cilians, genera Dermophis and Gymnopis. Copeia 1972(4) : 680-695. 212 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Schmithusen, J. 1968. Allgemeine Vegetationsgeo- graphie. Walter de Gruyter, Berlin, 436 p. Silverstone, P. A. 1975. A revision of the poison- arrow frogs of the genus Dendrobates Wagler. Nat. Hist. Mus. Los Angeles Cty., Sci. Bull. (21): 1-55. Silverstone, P. A. 1976. A revision of the poison- arrow frogs of the genus Phyllobates Bibron in Sagra (family Dendrobatidae ) . Ibid. (27): 1-53. Taylor, E. H. 1968. The caecilians of the world. A taxonomic review. Univ. Kansas Press, Law- rence, 848 p. Truer, L. 1971. Phylogenetic relationships of cer- tain Neotropical toads with the description of a new genus (Anura: Bufonidae). Nat. Hist. Mus. Los Angeles Cty. Contrib. Sci. (216): 1-40. Truer, L. 1974. Systematic relationships of Neo- tropical horned frogs, genus Hemiphractus (Anura: Hylidae). Univ. Kansas Mus. Nat. Hist. Occas. Pap. (29): 1-60. Trueb, L., Duellman, W. E. 1971. A synopsis of Neotropical hylid frogs, genus Ostoecephalus. Ibid. (l):l-47. Wake, D. B. 1966. Comparative osteology and evo- lution of the lungless salamanders, family Pletho- dontidae. Mem. South. California Acad. Sci. 4: 1-111. Wake, D. B„ Lynch, J. F. 1976. The distribution, ecology, and evolutionary history of plethodontid salamanders in tropical America. Nat. Hist. Mus. Los Angeles Cty. Sci. Bull. (25): 1-65. Appendlx Nonendemic taxa (i.e., found in one denoting the lowland forests (T = Salamanders Oedipina complex Oedipina parvipes Bolitoglossa biseriata Bolitoglossa chica Bolitoglossa medemi Bolitoglossa phalarosoma Bolitoglossa silverstonei Bolitoglossa sima Bolitoglossa taijlori ( expected ) Caecilians Epicrionops marmoratus Epicrionops parkeri Caecilia antioquiaensis Caecilia caribea Caecilia guntheri Caecilia leucocephala Caecilia nigricans Caecilia perdita Caecilia subnigricans "" Caecilia tentaculata °c Caecilia tliompsoni Dennopliis parviceps Oscaecilia och rocephala Oscaecilia polyzona Parvicaccilia nicefori Parvicaecilia pricei Typhlonectes natans Anurans Atelopus balios Atclopus elegans Atelopus glyphus Atelopus longibrachius Atelopus longirostris Atelopus mindoensis Atelopus spurrelli Bufo blombergi Bufo coniferus Bufo huematiticus APPENDICES 8:1. — Amphibians of the Trans-Andean Forests. or more other tropical lowland forests) are noted by an asterisk and letters Trans- Andean, N = Northern, C = Central Cis-Andean, A = Atlantic). Barchyolos pulcher Leptodactylus bolivianus °NC Leptodactylus melanonotus Leptodactylus pentadactylus SAC Leptodactylus poccilochilus "" Leptodactylus ventrimaculatus Leptodactylus wagneri °NCA Physalaemus pustulosus *" Eleutherodactylus sp. "A" Eleutherodactylus acliatinus Eleutherodactylus anomalus Eleutherodactylus areolatus Eleutherodactylus biporcatus Eleutherodactylus bufoniformis Eleutherodacttjlus caryopliyllaceus Eleutherodactylus caprifer Eleutherodactylus cruentus Eleutherodactylus diastema Eleutherodactylus fitzingeri Eleutherodactylus gaigeae Eleutherodactylus gularis Eleutherodactylus latidiscus Eleutherodactylus longirostris Eleutherodactylus mow Eleutherodactylus ornatissimus Eleutherodactylus raniformis Eleutherodactylus ridcns Eleutherodactylus roseus Eleutherodactylus subsigillatus Eleutherodactylus tacniatus Eleutherodactylus walkeri Eleutherodactylus sp. "Z" Colostethus chocoensis Colostethus imbricolus Colostethus latinasus Colostcthtis nubicola Colostethus pratti Colostethus talamancae Dendrobates auratus Dendrobates fulguritus Dendrobates histrionicus Dendrobates truncatus Dendrobates viridis Phyllobates anihonyi Phyllobates aurotaenia Phyllobates bicolor Phyllobates boulengeri Phyllobates espinosai Agalychnis calcarifer Agalychnis callidryas (expected) Agalychnis litodryas Agalychnis spurrelli Phyllomcdusa sp. Phyllomedusa venusta Hemiphractus fasciatus Gastrotheca angustifrons Gastrotheca cornuta Gastrotheca nicefori Hyla boons *NC Hyla boulengeri Hyla ebraccata Hyla gryllata Hyla miliaria Hyla pellucens Hyla picturata Hyla phlebodes Hyla quinquefasciata Hyla rosenbcrgi Hyla rubra °NCA Hyla ruhracyla Hyla subocularis Hyla sugillata Phrynohyas venulosa •NCA Smilisca phacota Smilisca sila Trachyccphalus jordani Centrolenclla flcischmanni *N0 Centrolenella ilex Centrolenclla ocetlifera Centrolenella prosoblcpon Centrolenella spinosa Glossostoma aterrim um 1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 213 Bnfo hypomelas Bufo marinus 0NC Bufo typhonius *NCA Dcndrobatcs minutus Dcndrobatcs occultator Rclictivomer pearsei "" Rana palmipes *"'CA Appendix 8:2. — Amphibians of the Central Cis-Andean Lowland Forests. Salamanders Bolitoglossa altamazonica Bolitoglossa equatoriana Bolitoglossa peruviana Caecilians Epicrinops nigrus Rhinatrema bicolor Brasilotyphlus braziliensis Caccilia hokcnnanni Caecilia disossea Caecilia dunni Caecilia gracilis Caecilia pressula Caccilia tentaculata 0T Microcaecilia albiceps Microcaecilia rabei Microcaecilia unicolor Oscaecilia bassleri Oscaecilia zweifcli Siphonops annulatus °A Ncctocaccilia ladigesi Nectocaecilia petersi Potornotyphlus haupii Typhlonectes compressicaudus Typhlonectes obesus Anurans Pipa aspers Pipa arrabali Pipa snethlageae Pipa pipa Atelopus franciscus Atelopus flavescens Atelopus palmatus Atelopus pulcher Bufo ccratophrys Bufo dapsilis Bufo glaberrimus Bufo guttatus Bufo manicorensis Bufo marinus °TN Bufo typhonius °TNA Dcndrophryniscus minutus Rhamphopluyne festae Ceratoplirys cornuta Adenomera andreae Adenomera hylaedactyla'^A. Adenomera martinezi Edalorhina perezi Hydrolaetare schmidti Leptodactylus sp. "A" Leptodactylus bolivianus *TN Leptodactylus dantasi Leptodactylus kxmdseni Leptodactylus longirostris Leptodactylus pentadactylus °T Eleuthcrodactylus gutturalis Eleuthcrodactylus inguinalis Eleuthcrodactylus lacrimosus Eleuthcrodactylus lanthanites Eleuthcrodactylus malkini Eleuthcrodactylus marmoratus Eleuthcrodactylus martiae Eleuthcrodactylus nigrovittatus Eleuthcrodactylus ockendeni Eleuthcrodactylus orphnolaimus Eleuthcrodactylus paululus Eleuthcrodactylus pcruvianus Eleuthcrodactylus platydactylus Eleutherodactylus pseudoacuminatus Eleuthcrodactylus quaquaversus Eleutherodactylus sulcatus Eleutherodactylus t rachyblcpharis Eleutherodactylus variabilis Eleuthcrodactylus vcntrimarmoratus Eleutherodactylus vilarsi Eleutherodactylus zeuctotylus Euparkerclla myrmecoides Ischnocnema quixensis Colostethus beebei Colostcthus brunneus Colostethus degranvillei Colostcthus fuliginosus Colostethus intermedins Colostethus marchcsiamis Colostethus sauli Dendrobates azureus Dcndrobatcs galactonotus Dendrobates leucomelas Dendrobates quinquevittatus Dendrobates steyermarki Dendrobates iinctorius Phyllobatcs bassleri Phyllobatcs bolivianus Phyllobatcs femoralis Phyllobatcs ingeri Phyllobatcs parvulus Phxjllobatcs pictus OA Phyllobatcs pulchritectus Phyllobatcs petersi Phyllobatcs smaragdinus Phyllobatcs trivittatus Phyllobatcs zaparo Agalychnis craspedopus Phyllomedusa bicolor Phyllomedusa buckleyi Phyllonwdusa palliata Phyllomedusa tarsius Phyllomedusa tomopterna Phyllomedusa trinitatus *N Phyllomedusa vaillunti Hemiphractus bubalus llcmiphractus johnsoni Hyla cruentomma Hyla dentci Hyla egleri OA Hyla epacrorhina Hyla fasciata Hyla favosa Hyla fuentei Hyla funerea Hyla garbei Hyla geographica *A Hyla goinorum Hyla grandisonac Hyla granosa Hyla haraldschultzi Hyla helenae Hyla hypselops Hyla imitator Hyla inframaculata Hyla lanciformis Hyla leucophyllata OA Hyla luteocellata ON Hyla marmorata Hyla minima Hyla minuta °A Hyla multifasciata Hyla ornatissima Hyla parviceps Hyla proboscidca Hyla punctata Hyla rhodopcpla Hyla riveroi Hyla rossalleni Hyla roeschmanni Hyla rubra OTNA Hyla sarayacuensis Hyla schubarti Hyla scnicula OA Hyla steinbachi Hyla surinamensis Hyla tintinnabulum Hyla triangulum Hyla tuberculosa Nyctimantis rugiceps Osteocephalus buckleyi Osteoccphalus leprieurii Osteocephalus taurinus Osteocephalus pearsoni Osteocephalus verrucigerus Phrynohyas coriacea Phrynohyas vcnulosa °T'VV Phyllodytes auratus Sphaenorhynchus carneus Sphaenorhynchus dorisae Sphaenorhynchus curhostos Centrolenella fleischmanni *' Centrolenella geijskesi Centrolenella medemi 214 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Leptodactylus rhodomystax Leptodactylus rhodostigma Leptodactylus rugosus Leptodactylus stenodema Leptodactylus vilarsi Leptodactylus wagneri *TNA Litlwdytes lineatus Physalaemus pctersi Vanzolinus discodactylus Disch idodactylus duidensis Elcuthcrodactyhis acuminatus Elcuthcrodactyhis altamazonicus Elcuthcrodactyhis carvalhoi Elcuthcrodactyhis chiastonotus Elcuthcrodactyhis conspicillatus Elcuthcrodactyhis croccoinguinis Elcuthcrodactyhis diadematus Elcuthcrodactyhis fencstratus Elcuthcrodactyhis granulosus Hcmiphractus proboscidcus Hemiph ractus scutatus Flcctonotus fitzgcraldi Gastrotheca longipes Gastrotheca testudinca Stefania evansi Sfefauia goini Stefania marahuaquensis Allophrync ruthveni Aparasphcnodon vcnezolana Hyla alhoguttata Hyla alhomarginata *NA Hyla haumgardneri Hyla henitczi Hyla hifurca Hyla hocscmani Hyla hokennauni Hyla boons 0T'N Hyla brevifrons Hyla calcarata Centrolenella midas Ccntrolenella munozorum Centrolenella oyampiensis Centrolenella pulidoi Centrolenella ritae Centrolenella taylori Chiasmocleis anatipes Chiasmocleis bassleri Chiasmocleis h udsoni Chiasmocleis sh udikarensis Chiasmocleis vcntrimaculata Ctenophrync gayi Hamptophrync boliviano Otophrync robust a Syncope antenori Syncope carvalhoi Synapturanus mirandaribciroi Synapturanus rabus Synapturanus salscri Rana palmipes °TNA Appendix 8:3. — Amphibians of the Caecilians Caccilia flavopunctata Caecilia subnigricans °T Nectocaccilia haydee Typhlonectes venezuclense Anurans Atetopus cruciger Bufo marinus °T0 Bufo stcrnosignatus Bufo typhonius °TCA Ceratophrys calcarata Adenomera hylacdactyla°CA Leptodactylus bolivianus OTC Northern Forests ( Santa Marta, M Range, and Bases of Merida Andes Leptodactylus poecilochilus °T Leptodactylus wagneri °TCA Physalaemus pustulosus OT Elcuthcrodactyhis bicumulus Elcuthcrodactyhis maussi Elcuthcrodactyhis rozei Elcuthcrodactyhis tcrracbolivaris Colostcthus dunni Colostethus hermani Colostcthus rivcroi Colostethus sp. Ph yllomedusa medinae Phyllomedusa trinitatus °c Gastrotheca williamsoni aracaibo Basin, Falcon District, Coastal )• Flcctonotus pygmaeus Hyla alhomarginata °CA Hyla alemani Hyla battcrsbyi Hyla boons °TC Hyla luteocellata "' Hyla rubra fTCA Phrynohyas vcnulosa °TCA Centrolenella antisthenesi Centrolenella flcischmanni 0TC Centrolenella orient alis Centrolenella orocostalis Relictivomer pearsci 0T Rana palmipes OTCA Appendix 8:4. — Amphibia of Atlantic Forests. Caecilians Mimosiphonops vermiculatus Oscaecilia h ypereu m eces Siphonops annulatus °c Siphonops confusionis Siphonops hardyi Siphonops insulanus S iph onops Icucoderus Siphonops paulcnsis Chthoncrpcton bracstrapi Anurans Hcmipipa carvalhoi Brachycephahis ephippium Psyllophrync didactyla Atclopus pernambucensis Dcndrophryniscus brcvipollicat us Dcndrophryniscus leucomystax Rhamphophryne proboscideus Bufo crucifcr Bufo ictericus Cycloramphus elcuthcrodactyhis Cycloramphus fulginosus Cycloramphus granulosus Cycloramphus neglcctus Cycloramphus ohausi Scythrophrys sawayae Zachaenus parvulus Zachaenus sanctaccatharinac Proceratophrys appendiculatus Proceratophrys boici Proccratoph rys cristiceps Proceratophrys fryi Macrogenioglottus alipioi Elcuthcrodactyhis bilincatus Elciithcrodactylus binotatus Elcuthcrodactyhis gualtcri Elcuthcrodactyhis gucuthcri Elcuthcrodactyhis hensclii Elcuthcrodactyhis hoehnei Elcuthcrodactyhis nasutus Elcuthcrodactyhis octavioi Hyla cuspidata Hyla cymbalum Hyla duartci Hyla egleri "'' Hyla faber Hyla fiavoguttata Hyla fuscomarginata Hyla geographica "' Hyla hayii Hyla humilis Hyla langci Hyla leucophyllata " Hyla limai Hyla marginata Hyla microps Hyla minuta "' Hyla nuhdcrvri Hyla oliverrai Hyla pachychrus Hyla prasina Hyla rizibilis 1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 215 Bufo typhonius "TNC Bufo ocellatus Ceratophrys aurita Crossodactylus aeneus Crossodactylus dispar Crossodactylus gaudichaudii Crossodactylus grandis Crossodactylus schmidti Crossodactylus trachystoma Hylodcs aspera Hylodes glabrus Hylodcs latcristrigatus Hylodcs magalhaesi Hylodcs nicrtcnsi Hylodcs meridionalis Hylodcs nasus Hylodcs ornata Hylodcs pcrplicatus Megaclosia gocldi Adenomcra bokermanni Adenomcra hylaedactyla°KC Adcnomcra marmorata Leptodactylus mystaceus Lcptodactylus mystacinus Leptodactylus sp. "N" Lcptodactylus pentadactylus 01 Leptodactylus troglodytes Leptodactylus wagneri *TN0 Physalacmus maculiventris Physalaemus moreirae Physalacmus nanus Physalacmus olfersi Physalacm us signiferus Thoropa miliaris Thoropa lutzi Thoropa pctropolitana Crossodactylodcs pintoi Cycloramphus asper Cycloramphus boulengeri Cycloramph us diringshoefensi Cycloramphus dubuis Eleutherodactylus paulodutrai Eleutherodactylus parvus Eleutherodactylus ramagii Eleutherodactylus venancioi Eleutherodactylus vinhai Euparkcrella hrasiliensis Ischnocnema verrucosa Colosteth us alagoanus Colostcthus capixaba Colostcthus carioca Colosteth us olfersioides Phyllobates pictus "c Phyllomedusa aspera Phyllomcdusa aycaye Phyllomedusa bahiana Phyllomedusa centralis Phyllomcdusa distincta Phyllomcdusa fimbriata Phyllomcdusa guttata Phyllomedusa marginata Phyllomcdusa rohdei Flectonotus fissilis Fritziana gocldii Fritiziana ohausi Gastrotheca ernestoi Gastrotheca fissipes Gastrotheca microdisca Gastrotheca viridis Aparasphcnodon brunoi Hyla albicans Hijla albofrcnata Hyla albolincata Hyla albomarginata eNC Hyla albosignata Hyla astartca Hyla aurata Hyla ariadne Hyla argyreornata Hyla catharinae Hyla claresignata Hyla crospedospila Hyla rubra °TNr Hyla secedem Hyla scnicula °c Hyla strigilata Ostcoccphalus langsdorffii Phyllodytes acuminatus Phyllodytes lutcolus Phrynohyas imitatrix Phrynohyas mesophaea Phrynohyas venulosa °TNC Sphacnorhynchus bromclicola Sphacnorhynchus orophilus Sphacnorhynchus palustris Sphaenorhynchus pauloalvini Sphacnorhynchus planicola Sphacnorhynchus prasinus Sphacnorhynchus surdus Trachycephalus altlas Trachycephalus nigromaculatus Centrolcnclla albotunica Centrolcnella bokermanni Centrolcnclla divaricans Centrolcnclla dubia Centrolcnclla curygnatha Centrolcnclla lutzorum Centrolcnella parvula Centrolcnclla pctropolitana Centrolcnclla surda Centrolcnella uranoscopa Centrolcnella vanzolinii Arcovomer passarellii Dasypops schirchi Chiasmoclcis albopunctata Chiasmoclcis bicegoi Chiasmoclcis centralis Chiasmoclcis schubarti Chiasmoclcis urbanae Myersiclla microps Hyophryne histrio Stcreocyclops incrassatus Rana palmipes OTVC 9. Origin and Distribution of Reptiles in Lowland Tropical Rainforests of South America James R. Dixon Department of Wildlife and Fisheries Sciences Texas AirM University College Station, Texas 77843 USA The task of summarizing the distribution, natural history and evolution of approximate- ly half of the known species of reptiles of South America is a nearly futile exercise be- cause the taxonomy and distribution of many important groups are poorly known. Never- theless, there are some redeeming values in any exercise — in this case, a summary of the state of our knowledge of the distribution and ecology of the South American reptilian fauna. As a matter of convenience throughout this paper, I refer to rainforest reptiles, rather than "tropical lowland rainforest reptiles." Rainforest reptiles include those genera and species that occur below 1000 m elevation and within the environs of rainforest (i.e., yearly temperature varying between 23° and 28°C; more or less continuous rainfall be- tween 150 and 600 cm annually). In the con- text of this paper, any reptile that occurs within the conditions mentioned above quali- fies as a rainforest species, although it may not be one ecologically. There are about 14 genera of savanna relicts within the confines of the rainforest environment, an additional 22 genera are aquatic, and about 30 genera are fossorial (Table 9:1). Several genera are wide ranging, occupying many vegetation zones, including rainforests. Therefore, the task of assigning a genus and/ or its attendant species as rainforest endemics, frequently is difficult without some basic knowledge of their ecology. There is a limited understanding of the evolution and distribution of rainforest rep- tiles in South America because only a few individuals (Vanzolini, 1972, 1974, 1976; Van- zolini and Reboucas-Spieker, 1973; Duellman, 1978; Fitch, 1970; Myers, 1974; Crump, 1971; Dixon and Soini, 1975, 1977; Rand and Humphrey, 1968) have investigated these topics. Vanzolini ( 1967 ) discussed the problems of application of ecological principles to the Amazonian biota without having a firm under- standing of the taxonomy and evolution of its constituent groups. Only through detailed taxonomic studies of various lowland tropical rainforest genera and species of reptiles (Uz- zell, 1966; Oliver, 1948; Duellman, 1958; My- ers, 1974; Savage, 1960; Kluge, 1969; Peters, 1960; Gans, 1971; Dixon, 1973, 1974a,b; Roze, 1967; Ruibal, 1952) and geographic and/ or ecological information ( Rand and Humphrey, 1968; Crump, 1971; Reebe, 1944a-c, 1945; Vanzolini, 1968, 1972, 1974, 1976; Vanzolini and Williams, 1970; Test, et al., 1966; Duell- man, 1978) will the reptilian rainforest fauna ever be fully understood. The following information obtained for forest associations and their included reptile species is only a rough estimate and, without doubt, will be subject to much future revision. We are raising questions that cannot be answered firmly without further studies on fossil histories, taxonomy, and ecology of par- ticular reptile groups. Time is short, for the rainforests are rapidly being altered by man (Richards, 1973), and data concerning the reptile fauna soon may be impossible to obtain. Without detailed natural history observa- tions on the majority of the species, little can be accomplished concerning the relationships among rainforest reptiles. For example, the taxonomy and distribution of Anolis is so poorly known (Williams, 1976) that the large number of described South American taxa (80) is enough to confuse and mask the rela- 217 218 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Table 9:1. — Reptilian Genera that Contain: A) Species that Are Savanna Relicts; B) Species that Are Aquatic or Semi-aquatic; C) Species that Are Semi-fossorial within Lowland Rainforest Situations. Amciva Anolis Cnemidophorus Crotalus Drymarchon Gymnophthalmus Iguana Kcntropyx Leptodeira Mabuya Mastigodryas Pscudoboa Phimophis Trctioscincus Tropidurus Tnpinainbis Caiman Chelus Chdydra Chrysemys Crocodihuus Crocodylus Dracaena Eunectes Hclicops Hydromedusa Hydrops Hydrodynastes Kinostemon Melanosuchus Neusticurus Palcosuchus Ph rynops Plat any s Podocncmis Pscudocryx Rliittoclemys Trctanorhinus Amphisbaena Anilius Anomalepis Apostolepis Atractus Aulura Bachia Bronia Diaphorolcpis Elapomorphus Emmochliophis Enulius Gcophis Leposternon Lcptotyphlops Liotyphlops Mesobaena Micrurus Ninia Notliopsis Opliiodes Rhadinaea Saphcnophis Sordellina Stenorrhina Synophis Taniilla Typldophis Typhlops tionships of other forest genera in any zoo- geographical analyses. I estimated that there are 31+ species of anoles within the limits of the Chocoan rainforest (based upon all literature available, other than Williams, 1976). Williams' (1976) list of species groups reduces my list of Chocoan anoles to IS spe- cies, while the number of species in the Amazon and Atlantic forests remain about the same 17(14) and 5(5), respectively. The disparity in the number of anole species rec- ognized in the Chocoan forest represents a 1 percent error when comparing forest reptile species, 3 percent comparing only lizards, 9 percent comparing arboreal lizards, and 33 percent comparing species of Anolis. Frequently a common and speciose genus of reptiles is also poorly understood (e.g., Anolis, Gonatodes, Amphisbaena, Atractus, Bothrops, Oxyrhopus, Micrurus, Liophis, Lei- madophis) . However, some students of South American herpetology (e.g., Gans, 1962; Wil- liams, 1976; Thomas, 1976; Wiest, in prep.; Peters, I960; Savage, 1960; Roze, 1966; Van- zolini, 1951, 1957, 1967; Vanzolini and Valen- cia, 1965; Oftedal, 1974) are making progress in systematic revisions of lowland rainforest reptile genera and species that eventually will lead to an improved understanding of the distribution of rainforest reptile species. HISTORY OF THE TROPICAL RAINFORESTS AND THEIR REPTILIAN INHABITANTS The Geological Evidence There are many congruent facts that imply that the major uplift of the Andes occurred in the Late Cenozoic, thus beginning the for- mation of the modern Amazon Basin. The geologic evidence of Harrington (1962), Loh- mann (1970), James (1971), Jenks (1956), Haffer (1974), Berry (1938), Herrero-Du- cloxc (1963), Rutland, et al. (1965) suggests that the older formations of the eastern basins were formed by the Andean uplift. Harrington ( 1962 ) suggested that part of the basin was covered by as much as 700 m of Tertiary deposits. The basins in eastern Peru and Ecuador accumulated fluvial sediments throughout much of the Tertiary and consist of both freshwater and brackish deposits. Several marine ingressions from the Pacific Ocean occurred through the middle of the Oligocene, followed by the union of the Ecua- dorian and Peruvian Andes (at the Huan- cabamba Deflection), closing the last portal through the Andes to the Pacific ( Harrington, 1962). Occasional marine ingressions of the Atlantic Ocean into the slowly rising Amazon Basin, via the portal between the Guiana and 1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 219 Rrasilian shields, the Orinoco (Llanos) Basin of Venezuela, or the Parana Basin of Para- guay, is indicated by the presence of brack- ish and saltwater fossil deposits found in eastern Peru and Ecuador (Haffer, 1974). It has been suggested that during the Late Cenozoic the upper Amazon Basin was a huge inland sea, in which sediments were constantly being deposited through torrential rains upon the Andes. These runoffs created vast flood plains that were crossed by numer- ous meandering streams that now form the major river systems of the Amazon Basin. Although the narrow straits between the Guiana and Brasilian shields may have been blocked at times (with subsequent flow through the Orinoco or Parana basins) an almost continuous connection with the At- lantic Ocean seems likely. Haffer's (1974) review of other studies indicates that Marajo Island, at the mouth of the Amazon, is under- lain by over 4000 m of fossiliferous marine and brackish, Cretaceous and Tertiary sedi- ments. The evidence suggests that continen- tal fluvial sediments were transported both from east and west to the sub-Andean basin of the upper Amazon. Permanent, easterly directed flow of sediments probably followed the final major uplift of the Andes in late Pliocene and early Pleistocene. Raven and Axelrod (1975) showed that the angiosperm flora of South America had a definite relationship to Africa through the Eocene, but as the plates ( continents ) drifted apart, each continent developed a unique flora. The process was enhanced by the Mio- cene orogenies in eastern Africa and western South America, changes in wind and tempera- ture patterns associated with those orogenies, and the influence of angiosperms from tropi- cal Asia that immigrated to the approaching African continent via sweepstakes dispersal. Much of the diversity of the tropical flora of Africa was diminished by the orogenies of the eastern plateau of Africa and associated wind, temperature, and hydrologic pattern shifts, but South America apparently maintained several centers of floral refugia (Haffer, 1974; Vanzolini, 1973; Midler, 1973; Vuilleumier, 1971) and thus retained a greater diversity of angiosperms. Historically, it seems that the rainforests were continuous geographically throughout much of the Early and Middle Cenozoic, subjected to disruption during the Miocene and Pliocene orogenies (Emiliani, 1956; Dorf, 1959; Tanner, 1968), and severely restricted to relatively small refugia by wind, temperature, and rainfall patterns during and following the Pleistocene (Flint, 1971; Sioli, 1975; Simpson, 1975; Raven and Axelrod, 1975; Emiliani, 1972; Haffer, 1974; Vuilleu- mier, 1971; Damuth and Fairbridge, 1970). Most modern reptilian genera with wide distributions that also occur in rainforests lack fossil records in South America (Amphis- baena, Ameiva, Iguana, Mabuya, Clelia, Epi- crates, Erythrolamprus, Lachesis, Leptodeira, Leptophis, Mastigodryas, Spilotes, Tantilla, Tripanurgos) . However, their current distri- butions imply that they may have evolved concurrently with those widespread genera with recorded fossil histories in the Miocene and Pliocene (Dracaena, Tupinambis, Teius, Diplolaemus and Crotalus). Some of these genera are also xeric adapted; thus, it is im- possible to postulate their origins without fossils or other evolutionary evidence. Estes and Price (1973) identified Pristiguana from the Upper Cretaceous in Brasil as the earliest known iguanid and suggested a Gondwanan origin of the family based on morphological, paleontological, and zoogeographical consid- erations. The Fossil Evidence The fossil history of rainforest reptiles is relatively good for those groups with large skeletal features (e.g., turtles and crocodil- ians) but extremely poor for amphisbaenians, lizards, and snakes. The fossil record of am- phibians and reptiles in South America has been documented by Baez and Gasparini (this volume). Only a few comments espe- cially pertinent to rainforest groups are in- cluded here. The presence of relatively modern gavials (Langston, 1965) in the Rio Magdalena of Colombia during the upper Oligocene and in the Pliocene in Argentina suggests that the lowland tropical vegetation was extensive throughout much of South America for long periods of time. This is further enhanced by 220 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 the probable replacement of gavials as pisciv- orous carnivores by longirostrine crocodiles, first in northern South America and later in the south. Langston (1965) suggested that longirostrine crocodiles were more adaptive and were probably replacing the gavials ( tak- ing over the fish-eating niche) throughout the Cenozoic Era. Langston pointed out that the significance of the record is that longirostrine crocodiles were assuming this role, not alli- gatorids. However, the fossil record indicates that the alligatorids were evolving at the same time that longirostrine crocodiles pre- sumably were competing with gavials. Cli- matic changes brought about by Mio-PIiocene orogenies may have severely limited the habi- tat and tolerances of longirostrine crocodiles and they in turn, were competively excluded by alligatorids. The presence of modern crocodiles occupying parapatric geographic ranges in South America suggests that they (C. acutus, C. intermedins) are relicts of some ancient stock or populations that were excluded in earlier competition, were (or became) salt tolerant, survived in island re- fugia, and are recent invaders. The fossil history of South American tur- tles is even more confusing, perhaps owing to the conservative nature of the morphologies of the groups and the widespread distribution of the pleurodires in South America and the rest of the world. Many early fossils of pleu- rodires were misidentified and, only recently, have workers (Auffenberg, 1974; Wood, 1972, 1976; Wood and Gamero, 1971) attempted to unravel the myriad of names assigned to fos- sils discovered in the past two centuries. The presence of Podocnemis and Chelus in the Pliocene in Estado Falcon, Venezuela, Chelus in the Miocene in the upper Rio Magdalena Valley, Colombia, Podocnemis in the Late Cretaceous in Brasil and in the Eocene/ Plio- cene in Peru, suggests that turtles and croco- dilians were occupying similar habitats dur- ing comparable periods of their evolution in South America. The orogenies of the late Miocene and throughout much of the Pliocene disrupted the general distributions of many turtle and crocodile species and apparently caused the demise of four genera of turtles in South America (Schwoeboemys, Parahijdraspis, Stu- pendemys, Trionyx and possibly Taphros- phys) and fourteen genera of crocodilians (Gavialis, Sebecus, Charactosuchus, Rham- phostomopsis, Nettosuchus, Eocaiman, Bra- chygnathosuchus, Gryposuchus, Proalligator, Leptorrhamphus, Purassaunis, Ilchunaia, No- tocaiman, Necrosuchus) . The majority of the species of those genera that survived the climatic, edaphic, and hydrologic shifts of Miocene/ Pliocene eras (Paleosuchus, Mela- nosuchus, Caiman, Crocodyhis, Podocnemis, Chelus, Phrynops, Kinosternon, Rhinoclem- mys, Chrysemys, Geochelone) are those that exist today in Neotropical rainforests (Wood, 1976; Wood and Gamero, 1971; Wood and Patterson, 1973; Romer, 1956; Auffenberg, 1974; Langston, 1965; Medina, 1976). The presence of the lizard genera Dra- caena, Tupinambis, and cf. Polychrus (Estes, 1961) in the Miocene deposits of the upper Rio Magdalena of Colombia strongly suggests that many contemporary lizard genera were widespread rainforest occupants. Although few fossil South American lizard and snake genera are currently known, many modern Cis-Andean and Trans-Andean genera were probably well differentiated prior to the orogeny of the Andes (Estes, 1961). Although we have little evidence that tropical environments have remained, in part, from the Mesozoic to modern times, the fossil evidence of tropical reptiles in South America suggests persistent tropical environments in most areas of South America prior to the up- lift of the Andes (Harrington, 1962). Historical Relationships of Currently Disjunct Rainforests The distribution of rainforest reptiles is dependent ultimately upon the history of en- vironmental factors that have affected the habitat and the reptilian gene pool. The increased vagility of one group compared with another, the inherent, fixed attributes of a particular species versus those with highly variable attributes, the physical and ecolog- ical barriers between different and/or equal habitat areas through time, and diffuse com- petition from other organisms within the niche hypervolume all play major and/or minor roles in formulating the distribution of various groups of reptiles. 1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 221 The relationship of the herpetofauna of the Chocoan Forest with those of other north- ern forests and the Amazon Forest is best understood by examining a series of geolog- ical events. Based upon present day distribu- tion of rainforest reptiles, I suggest that there was a more or less continuous connection be- tween rainforests prior to the major orogeny of the Andes. Simpson (1975) summarized several papers on the geology of the Andes that suggested the orogeny of the Andes maintained several centers of activity throughout the Cenozoic, with the Venezue- lan Andes rising in the Paleocene, the Colom- bian and Ecuadorian Andes in middle Eo- cene, followed by later Tertiary erosion. The Huancabama Deflection remained open until the upper Pliocene, when the Cordillera of Peru and Chile was uplifted further. The Andean orogeny proceeded from east to west across the South American continent. First the rainforests of Venezuela were isolated from those in Colombia. Subsequently, the Colombian rainforests were isolated from the Amazonian rainforests, and finally, the Cho- coan rainforests were separated from the Amazonian forests by the closure of the Huancabama Deflection. Based on fossil and present distributions of some reptiles, there are indications that a lowland rainforest cor- ridor may have existed between the Chocoan and Amazonian forests through the Huan- cabama Deflection during the last major in- terglacial period. The fossil record reveals several genera that are geologically and genet- ically older than the Andes and that occur now, or occurred in the past, on both sides of the mountains. The ranges of several spe- cies of snakes belonging to three genera (Chironius, Leimadophis, Leptophis) seem to have been connected recently, probably through the Huancabama Deflection. The water gaps between Central Amer- ica and South America may have been nar- row enough to allow some interchange of faunas between the two continents during the Cenozoic ( Hershkovitz, 1966; Savage, 1974). Land connections (not necessarily continu- ous) between those regions beginning in the Pliocene (Haffer, 1974) certainly allowed faunal interchange on a larger scale. Rain- forest faunas probably were exchanged when local environmental conditions were appropri- ate after the closure of the Panamanian Portal. The present reptilian fauna of the Cho- coan forest seems to be influenced more by a Central American faunal element than by an Amazonian one. Perhaps this results from the proximity of, and continuous connection to, the Central American faunal element after closure of the Panamanian Portal (Table 9:2). This also suggests that there has been a longer period of faunal interchange between the Chocoan Forest and Central American rain- forests than between those of Amazonia and the Choco, especially following the orogeny of the Colombian and Ecuadorian Andes. This is not to deny that the Chocoan rain- forest probably was isolated from the Central American and Amazonian rainforests (espe- cially from the latter beginning in at least mid-Pliocene) for periods of time sufficient to develop moderate amount of endemism. Connection of the Amazonian and Atlantic rainforests seems to have occurred several times. Haffer ( 1974 ) , Damuth and Fairbridge (1970), Vuilleumier (1971), Simpson (1975) and Raven and Axelrod (1975) have garnered evidence concerning the Quaternary history, especially that dealing with arid/ humid ( glacial/ interglacial) phases from the Pleisto- cene to 2500 years ago. Most data place the end of the Wisconsin glaciation at 10,000 to 15,000 y.b.p., with two short arid cycles 4,000 and 2,500 y.b.p. These may have caused the present hiatus between the Amazon and At- lantic forests. Three to four glacial periods are postulated for the Andes and most other regions, and there are five terrace levels along the Rio Amazonas resulting from cyclic ero- sion and aggradation (Haffer, 1974, this vol- ume). I assume there was at least one pre- glacial period of long duration prior to the Pleistocene, that was followed by four inter- glacial periods of various lengths of time after each of the four major glacral periods of the Pleistocene. Because the Amazonian forest is relatively young— 800,000 to 1,800,000 years and has been affected variously by flooding and/ or cool arid cycles, I assume that the five, recognized terrace levels are the result of preglacial and interglacial flooding. If we assume that each arid ( glacial ) phase created broad savanna regions and restricted the rain- 222 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Table 9:2. — A) North and Central American and/or Endemic Reptilian Genera that ical Rainforest Species that Reach Their Southern Distributions in the Northern and Endemic to the Choco Forest). B) South American Reptilian Genera that Contain forest Species that Reach Their Northern Limits in the Northern and Choco forests Relict). Contain Lowland Trop- Chocoan forests ( " = Lowland Tropical Rain- ( f Indicates a Savanna Anomalepis Basiliscus Chelydra Chnjsemijs Conioplianes Corythophancs Crocodylus Diaphorolepis'' Echinosaura* Emmochliophis* Enulius Ceopliis Lampropcltis Masticophis Scaphiodontophis Sibon Sphaerodactylus Stenorrhina Trachyboa' Tretanorhinus Ungaliophis Alopoglossus Amphisbaena Atractus Enyalioides Geochelone Helicops Liophis Morunasaurus Ophryoessoides Phimophis\ Prionodactylus Pseudoboa] Siphlophis Tretioscincus^ Tripanurgos Tropidophis Tropidurus\ Tupinambis\ forest to relatively small refugia (Fig. 9:1), the union of the restricted Atlantic and Ama- zonian rainforests may have occurred an equal number of times during interglacials, probably along the northeastern part of Bra- sil. Paleobotanic and paleoedaphic evidence ( Haffer, 1974; Damuth and Fairbridge, 1974 ) suggests that the "Diagonal of Open Forma- tions" (Vanzolini, 1963, 1974) has been pres- ent from Late Tertiary to present with only minor alterations. Therefore, a belt of Ama- zonian forest may have reached the Atlantic forest as indicated above, or via the Rio Madeira-Rio Parana Drainage around the west and south sides of the Brasilian Shield. A filter corridor through the Brasilian Shield along gallery forests also may have been pos- sible, but these routes probably were restric- tive. Many Amazonian reptiles have reached only the northern part of the Atlantic forest (e.g., Anolis punctatus, A. ortoni, A. fusco- auratus, Kcntropyx calcaratus, Bothrops bi- lineatus, Chironius carinatus, Coleodactylus meridionalis, Polychnis marmoratus, Tripan- urgos compressus, Mastigochyas boddaerti, Dipsas indica), thereby suggesting that the former route is the most plausible. Most Amazonian forest species that have reached the Atlantic forest show little or no geograph- ic variation, suggesting that a large number of species may have reached the Atlantic forest during the last Amazon-Atlantic con- nection (probably 3,000-5,000 y.b.p.). A speciation model. — From Wiest's (in prep.) work on the colubrid snake genus Chironius, it is feasible to postulate the num- ber of times this genus extended its distribu- tion into the Atlantic forest during interglacial periods with alternate glacial periods of spe- ciation in both the Amazonian and Atlantic forests. In the case of Cliironius, there seems to be a progressive intrusion of recently-differ- entiated Amazonian stocks resulting from the repeated contraction of the forest into refugia during at least four glacial periods. Each interglacial expansion of the Amazonian forest that reached the Atlantic forest seemed to carry with it the more primitive members of each evolving gene pool. Without fossil evi- dence, the model is purely hypothetical, but Wiest ( pers. comm. ) constructed an inde- pendent model, and we differ only on minor points. If we assume that the primitive stock of the gene pool is peripheral to the de- veloping core and the set of characters defin- ing the primitive conditions for Chironius are correct, the following sequence of events Distribucion propuesta para el Plcistoceno dc los tipos de vegetation Suramericana durante los pcriodos de maxima glacial. D = Desicrto; GI = Hielo Glacial; GL = Pastizales; M = Monte; NE = Bosque de Nothofagus; P = Paramo (Alpino); PS = Matorral semi-desertico dc Patagonia; S = Savana; TF/S = M atonal dc bosque espinoso; TM = Bosque tropical montano; TR = Bosque tropical lluvioso; TS-E/D = Bosque tropical semi-siempreverde/detiduo. La zona punteada y con Uncus diagonal cs en la cuenca ama- zonica puede habcr sido un bosque lu'imcdo tropical. 1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 223 Fie. 9:1. Proposed Pleistocene distribution of South American vegetation types during maximum glacial periods. D = Desert; GI = Glacial Ice; GL = Grassland; M = Monte; NF = Nothofagus Forest; P = Paramo (Alpine); PS = Patagonian (Semi-desert Scrub); S = Savanna; TF/S = Thorn Forest/Scrub; TM = Trop- ical montane; TR = Tropical rainforest; TS-E/D = Tropical semi-evergreen/deciduous Forest. Dashed and diagonal lined area in Amazon Basin may have been an additional isolated tropical rainforest. 224 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 should represent the various invasions of Chironius stock (and possibly other reptiles) into the Atlantic forest. 1. Froto-Chironius stock invaded the At- lantic forest during a pre-Pleistocene period when the two lowland tropical rainforests were connected along a coastal corridor, from the mouth of the Amazon eastward and southward along the northeastern part of Brasil. 2. During the first major Pleistocene gla- cial period, the forests were discontin- uous refugia with C. multiventris and C. exoletus as developing taxa in Ama- zonian refugia and C. bicarinatus in Atlantic refugia. Two additional spe- cies, C. quadricarinatus and flavolinea- tus, were speciating in gallery forests within savannas (the "Cerrado-Caatin- ga Diagonal" of Vanzolini, 1963, 1974) between the two rainforests. 3. The first major Pleistocene interglacial period allowed expansion of the forest refugia and eventual reconnection of the Atlantic and Amazonian forests. Chironius quadricarinatus and C. fla- volineatus evolved into full, biological species and extended their ranges within the diagonal of open formations in Brasil where they have remained to the present time. In Amazonia, C. multiventris and exoletus had devel- oped into biological species and now invaded the Atlantic forest, coming into contact with C. bicarinatus, a species with a different diet and probably not a competitor. 4. The second, major glacial period again reduced the forests to discrete refugia. The C. exoletus stock remained in the Atlantic refugia with some contact with C. exoletus populations in Amazonia along gallery forests, while the Atlantic population of C. multiventris became isolated and formed a vicariant bio- logical species (C. foveatus). Chiron- ius bicarinatus remained undifferenti- ated in Atlantic refugia while C. multi- ventris did likewise in the Amazonia refugia. Chironius scurrulus developed into a biological species in one of the several Amazonian forest refuges. 5. The second interglacial period allowed a reunion of the lowland rainforests. Chironius scurrulus invaded the ex- panded Atlantic forest for the first time, while C. multiventris and exoletus did so for the second time. Chironius fo- veatus had developed in the Atlantic forest as an ecological equivalent of C. multiventris, and the two species formed competitive populations. Chir- onius bicarinatus did not differentiate. Previously allopatric populations of C. exoletus interbred along zones of con- tact. 6. The third glacial period reduced the forests to discrete refugia. Chironius exoletus, multiventris and scurrulus re- mained undifferentiated populations in Amazonian refugia. Chironius carina- tus became a developing biological spe- cies from an isolated C. exoletus stock in the Guianan Forest Refugium. Chi- ronius bicarinatus and foveatus re- mained relatively undifferentiated in Atlantic forest refugia, while the com- peting Atlantic populations of C. multi- ventris became extinct. The semi-iso- lated Atlantic population of C. exoletus differentiated subspecifically (C. e. pyrrhopogon), and an isolated Atlantic population of C. scurrulus diverged into C. s. laevicollis. 7. During the third interglacial period, a reunion of the two forests allowed for the reinvasion of C. exoletus, multiven- tris, scurrulus and an initial invasion of C. carinatus into the Atlantic Forest. Chironius bicarinatus and foveatus re- mained intact, although the latter again became competitive with the invading C. multiventris. The once isolated pop- ulations of C. exoletus-scurrulus formed intergrading complexes. 8. The fourth glacial period reduced the forests into discrete refugia once more. Chironius exoletus, scurrulus, multi- ventris, and carinatus remained undif- ferentiated in Amazonian forest refugia, while an isolated population of Ama- zonian C. scurrulus developed into a biological species (C. fuscus). The isolated Atlantic Forest population of 1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 225 C. multiventris was again outcompeted by C. foveatus and became extinct. The Atlantic populations of C. exoletus and scurrulus became isolated from their Amazonian parental stock and be- came increasingly isolated reproduc- tively into C. pyrrhopogon and laevi- collis, respectively. The Atlantic C. bicarinatus remained undifferentiated, and the Atlantic population of C. cari- natus was restricted to small pockets of forest along the northeastern coast of Rrasil owing to competition with C. bicarinatus and foveatus, and may have become extinct there at that time. 9. The fourth interglacial period (possibly the past 10,000 years) brings us to the current distribution of species of Chi- ronius in the Amazonian and Atlantic forests. Chironius exoletus, multiven- tris, scurrulus, carinatus, and fuscus are sympatric and/ or parapatric in Ama- zonia. The first four species may have briefly invaded the Atlantic forest dur- ing a connection of the two forests somewhere between 10,000 and 4,500 y.b.p., and the latter species invaded for the first time. Atlantic populations of C. carinatus and fuscus currently are undifferentiated from Amazonian popu- lations. Atlantic populations of C. scur- rulus have differentiated into an allo- patric, recognizable race (C. s. laevi- collis), whereas Atlantic populations of C. exoletus are weakly differentiated into a recognized subspecies, C. e. pyrrhopogon, that intergrades with the nominate race over a broad zone. Chi- ronius bicarinatus and foveatus seem to be undifferentiated and distributed throughout the Atlantic Forest. COMPOSITION OF THE HERPETOFAUNA Faunal Similarity At the rate taxonomists are describing new taxa of reptiles from the Neotropics ( 10-30 species per year), it will be a long time be- fore we obtain sufficient knowledge to allow ,«sPr,ere RePf'/e r Fig. 9:2. The percentage of the 335 genera of reptiles in the Western Hemisphere that are endemic to, or shared with, each of the two major continents and the Central American corridor. El porcentaje de los 335 generos de reptiles en el Hemisferio Occidental que son endemicos para o compartidos con cada uno de los continentcs mayores y con el corredor centroamcricano. us to comprehend the complex ecological structure of reptilian communities. Any num- ber I mention is subject to immediate change, and should be accepted with a broad margin for error. Of approximately 1,100 species and 203 genera of mainland South American reptiles, 550 species ( 149 genera ) occur in the rain- forests. This represents 73.4 percent of the total number of South American reptile gen- era and 49.5 percent of the species. It also represents 44.5 percent of all genera and 17.8 percent of the species in the Western Hemisphere (Fig. 9:2). Of the 149 reptile genera, 21 (14.1%) are shared between the Choco and the Amazon; 17 (10.7%) between the Atlantic and the Ama- zon; 3 (2.7%) between the Choco and the Atlantic; 38 (25.5%) among all major Neo- tropical rainforests (Fig. 9:3, Table 9:3). Of the 550 species, 23 (4.3%) are shared between 226 MONOGRAPH MUSEUM OF NATURAL HISTORY GENERA SPECIES NO. 7 □ : Amazon ;;;!; Choco Atlantic LI Shared Fig. 9:3. Percentage of reptilian genera endemic to each of the major tropical rainforests and those shared between two or more of the forests. El porcentaje de generos de reptiles endemicos en cada uno de los bosques humedos tropicales maijores ij de aquellos comunes a dos o mas bosques. the Choco and the Amazon; 43 (8.0%) be- tween the Atlantic and Amazon; 1 (0.02%) between the Choco and the Atlantic; 18 (3.1%) among all major Neotropical forests (Fig. 9:4, Appendix 9:1). The smaller trans- Andean forests (Magdalena, Santa Marta, Maracaibo, and zuela) show little endemism ( and those closest to the Choco are essentially identical in their reptile faunas to the latter. Each of the more isolated of these forests (Santa Marta, Maracaibo, coastal Venezuela) contains one lowland rainforest endemic spe- cies. The coastal Venezuelan Forest shows closer affinities to the Amazon Forest, whereas the Maracaibo and Santa Marta forests con- tain a number of species that are broadly distributed in both Central and South Amer- ica (see Appendix 9:1). Of the 149 genera of rainforest reptiles, 72 (48.3%) are snakes, 63 (42.3%) are lizards [includes 5 (3.4%) amphisbaenids], 10 (6.7%) turtles and 4 (2.7%) crocodilians; of the spe- Amazon SS Choco Atlantic U Shared Sinu, Nechi, coastal Vene- tian 4%) Fie. 9:4. Percentage of reptilian species endemic to each of the major tropical rainforests and those shared between two or more of the forests. El porcentaje de especies de reptiles endemicos en cada uno de los bosques tropicales humedos mayorcs u de aqucllas comunes a dos o mds bosques. cies, 284 (51.4%) are snakes, 235 (43.0%) are lizards [includes 30 (5.5%) amphisbaenids], 24 (4.4%) turtles and 6 (1.1%) crocodilians. Of 95 lizard and 94 snake genera in South America, 63 (66.3%) and 72 (76.6%), respec- tively, occur in rainforests. I agree with Burt (1958) and Udvardy ( 1969 ) that any formula chosen to obtain indices for similarities or differences between faunal elements depends upon the accuracy of the basic data set, the relative size and/or equivalent nature of each of the faunal areas and elements, and the taxonomic category chosen for the comparison (Fig. 9:5A). Perhaps one of the best ways to compare faunal elements is to obtain subsets of data from equivalent areas within any given, major faunal region. The investigator thus gains an understanding of the variation within and be- tween geographic subsets of data. It seems obvious that an increase in the number of species of any particular group from a given area will decrease the faunal similarity index 1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 227 Table 9:3. — Genera Shared among the Three Major Tropical Lowland Rainforests. Amazon/ Atlantic Apostolepis Cercosaura Coleodactylus Colobosaura Eny alius Hemidactylus Kentropyx Lcpostcmon Palcosuchus Pantodactylus Philodryas Plirynops Platemys Podocnemis Thamnodynastcs Tropidurus Tupinamhis Xenopholis Choco/Atlantic Diploglossus Liotyphlops Lygophis Amazon/Choco Alopoglossus Bach ia Boa Cnemidophorus Corallus Dendrophidion Drymarchon Drymobius Enyalioidcs Gonatodes Gymnophthalmus Imantodes Lepidoblepharis Morunasaurus Ninia Oxybelis Pliocercus Prionodactylus Ptychoglossus Rhinobothryum Synophis Thccadactylus Amazon/Choco/Atlantic Ameiva Amphisbaena Anolis Atractus Bothrops Caiman Chironius Clelia Dipsas Epicrates Erythrolamprus Geoclielone(?) Hclicops Iguana Kinostemon Lachesis Leimadophis Leposoma Lcptodcira Leptophis Leptotyphlops Liophis Mabuya Mastigodryas Micrurus Oxyrhopus Polyclinic Pscudoboa Pscustes Rhadinaca Spilotcs Siphlophis T ant ilia Tripanurgos Tropidophis( ?) Tupinamhis Typhlops Xenodon (FSI) value proportionally (Fig. 9:5B). It is also obvious that combination of different groups (e.g., lizards, snakes, turtles, croco- dilians, or any other vertebrate groups) to arrive at a FSI value tends to mask the true relationship of each of the subsets (i.e., lizards versus lizards). Each group (species, genus, family) has its own independent history, and only broad intei'pretations can be made when one major group (Reptilia) of one faunal area is compared to that of another (Fig. 9:15c-e). Although the Amazonian and Atlantic for- ests tend to show a closer relationship to each other than to the Choco Forest, there are 18 species and 38 genera that are common to all three forests. When the common species are removed from consideration and attention is focused upon the probable origin of the spe- cies and/or genera, a more meaningful rela- tionship can be observed between various South American ecophysiographic regions (Fig. 9:6). South American rainforests have a high degree of endemism among reptiles. How- ever, when one considers the land area occu- pied by rainforests in South America (40%) and the proportion of South American rep- tiles that are endemic to rainforests (17%), it is apparent that there is high endemism elsewhere in South America and/ or many widely distributed species. Area and Diversity In order to appreciate the potential biotic diversity, one must first visualize the mag- nitude of the rainforests of South America (Fig. 9:7). Those forests are the largest in the world, about 4.5 million sq km2 before modern timber harvesting techniques were employed (Richards, 1973), as compared to those of Asia (approximately 2.8 million km2) or Africa (about 2 million km2). The variation in size of independent rain- forests does not seem to dictate the number of silvicolous lizards. The predictors of di- versity are the same for each forest — 1) low- land tropical rainforest; 2) latitudes; 3) cli- mate; 4) structural heterogeneity; and 5) de- gree of trophic specialization. The only in- dependent variable is total land area occupied by the rainforests. Assuming that the pre- dictors for each species in each forest are basically the same, the number of silvicolous- food-generalist species can be predicted for a given forest (e.g., Choco, 0.24; Amazon, 4.0; Congo, 1.8; Thailand, 0.5 million km2). Thus, if lizards respond to a set of environmental parameters within the constraints of the phys- iognomy of the forest (e.g., type of bark, roots, branching, diameter, height, canopy, epiphytes), the number of species that are habitat specialists and food generalists should 228 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Amazon (1071 Moiopon 1341 Atlantic (77) CenlroUmin (28) Amazon (318) Atlantic (150) Atlantic (62) Choco(72) Amazon (166) Amazon (531 Amazon 11121 Atlantic 1141 Amazon (41) Amazon 126) Atlantic 1111 1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 229 be relatively constant, providing that the age, evolution, and environmental conditions with- in the forests are similar. My knowledge of silvicolous lizards suggests that the Choco contains about 38 species, Amazon 44, Congo 39, and Thailand 40. Thus, the predicted similarities in numbers of species between forests, regardless of size, apparently are realistic. Therefore, if we have knowledge of certain evolutionary and ecological parameters concerning the habitat and its attendant spe- cies, we should be able to predict the number of species that may occupy that habitat re- gardless of their generic or familial status or size of the forest. Iguanids and teiids basi- cally fill the silvicolous-food-generalist role in the Choco and Amazon forests, gekkonids and chamaelionids in the Congo, and gekkonids and agamids in Thailand. The constraints placed upon a terrestrial lizard are much more difficult to discern than those of a silvicolous arboreal one. Compari- son of arboreal, eurytrophic lizards is pos- sible between rainforests of similar latitudes, temperatures, and general physiognomy. However, comparison of terrestrial lizards is more difficult owing to the patchiness of the forest floor. Adequate knowledge of the habits of the animals in question is the only recourse for quality information on faunal relationships (Pianka, 1974). For example, the numbers of lizards inhabiting the forest floor of the Choco, Amazon, and Atlantic, and, in parentheses, those that are true leaf litter inhabitants of primary rainforest are 22(18) in the Choco, 43(35) in the Amazon, and 19(18) in the Atlantic forest. These figures indicate that forest size may be important for terrestrial lizards; almost twice as many species of terrestrial lizards occur in the Amazon than in the other two forests. Among several possible explanations for the influence of forest size on the number of Amazonian lizards are the following. 1) Various seg- ments of the Amazonian forest may have had different evolutionary histories owing to cli- matic alterations during the Quaternary, whereas the Choco and Atlantic forests may have remained as single units (Fig. 9:1). 2 ) Physical attributes of the forest floor ( sand hills, lowland swamps, different soil proper- ties, slope exposure, depth of leaf litter) may have affected speciation over time. 3) A longer ecotone between savanna and forest affected heterogeneity. I favor the first al- ternative because the distribution of reptiles coincides with postulated Quaternary refugia (Figs. 9:8-9). Why such refugia would pro- Fig. 9:5. (A) The three tropical lowland rainforests of South America compared for reptilian faunal similar- ities based on genera and on species. Simpson's formula (C/N,) X 100 is used because it indicates the percentage of common taxa in the smaller of the two faunas, regardless of the size of the larger one. How- ever, the larger fauna must also be compared to obtain similar data in the opposite direction. The solid lines of the triangle represent the percent of the common taxa in the larger of the two faunas (direction of arrow is from larger to smaller fauna), whereas the inner dashed line shows percentage of common taxa in the smaller of the two faunas (direction of arrow from smaller to larger fauna). Numbers in parentheses following names of forests represent number of taxa within that forest. (B) A Simpson formula (C/Ni) X 100 com- parison of species of lizards and snakes from three local rainforest sites within the Iquitos region, Peru. When faunas of equal size are compared (number of snakes between Centra Union and Mishana), then Duellman's (1965) modified formula (2C/Ni + N=) X 100 is used because the resulting Faunal Resemblance Factor compares Simpson's formula in both directions with a resultant average of the two percentages. (C, D, and E) Simpson's Formula (C/N,) X 100 comparing various reptilian species components in the three tropical lowland rainforests of South America. (A) Comparacion de los tres bosques tropicales de tierra baja de Sur America, en base a similaridad de la fauna reptiliana en cuanto a generos y especies. Se uso la formula de Simpson (C/Ni) X 100 porque indica el porcentajc de taxones coryiunes in la mas pequena de las dos faunas, sin importar el tamano de la mayor. Sin embargo, la fauna mayor debe de ser tambien comparada para obtener datos similares en direccion opuesta. Las lineas solidas del tridngido representan el porecntaje de laxones comunes en la mas grande de las dos faunas (la direccion de la flecha va de la mayor a la menor fauna), mientras que la linea quebrada interior muestra el porcentaje de taxones comunes en la mas pequena de las dos faunas (la direccion de la flecha va de la menor a la mayor fauna). (B) Comparacion de las especies de la lagartos y serpientes de tres locali- dades de bosque amazonico dentro de la region de Iquitos, Peru, usando la formula de taxa (como las de Ccntro Union y Mishana), entonces, usamos la formula modifieado por Duellman (1965) (2C/N, + N:) X 100 porque el factor de similaridad resultante compara en dmbas direcciones con una resultante de la media de los dos porcentajes. (C, D, and E ) Comparacion de varias especies componentes de tres bosques tropicales de tierra baja en Sur America usando la formula de Simpson. 230 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fie. 9:6. Percentage contribution of nonendemic reptilian species to a particular rainforest. Number within a circle represents approximate percent of endemism of reptilian species for a particular forest. Contribution porcentual de especies no endemicas de reptiles a un basque tropical particular. El numero dentro del circuit) representa el porcentaje aproximado de endemismo de especies de reptiles para un bosque en particular. 1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 231 75 6£ . 55 15 fJf1^~ <5D g£ •10 #^pr|T. mf 1 lb. 3 5 ***M If : t % »^~ — i ■0 ^E&r^^l / %i > \^ , / . *■;■■■ ^ •10 j^ •> • r \ ) J ■ I > / \ / i i \ ■v-4 <--, /' . A Fig. 9:7. Distribution of lowland tropical rainforests of South America (slightly modified from Hueck and Seibert, 1972). Distribution de los bosques de tierra baja de Sur America (algo modificada de Hucck y Seibert, 1972). duce more terrestrial species than arboreal species is difficult to answer, but I suggest that arboreal habitats may be more homogeneous than terrestrial ones. Otherwise, the number of species should be approximately equal in each region. If the postulated refugia of each forested region had similar physical and bio- logical properties (e.g., topography, moisture, temperature, and habitat heterogeneity), the Choco, western Amazon, eastern Amazon, and Atlantic forests should contain about the same number of terrestrial, rainforest lizard species. They do, with 18, 18, 17, and 18 species, respectively. The most diverse group of lowland rain- forest reptiles is the snakes (282+ species), but most species apparently have low popu- lation densities. Ecological data are lacking for many species (Dixon and Soini, 1977). Rainforest lizards seem to be habitat special- ists and food generalists, whereas sympatric snakes seem to be habitat generalists and food specialists (Duellman, 1978). Shine ( 1977 ) found that six species of elapids in a 232 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fie. 9:8. Generalized ranges of rainforest lizards that show restricted distributions. In addition, there are 18 species of lizards that are distributed throughout the Amazonian forest. Distribuciones generalizadas de saurios de bosque mk (18) Diaphorolepis ( 2 ) ... Drepanoides ( 1 ) __ Drymarchon ( 1 ) Drymohius ( 2 ) Drymoluber ( 1 ) Epicratcs ( 1 ) Elapomorphus (3) . Emmochliophis (1) . Enulius (2) Erythrolamprus (4) Eunectes (3) Geophis ( 1 ) Helicops (11) Helminthophis ( 1 ) . Hydrodynastes ( 1 ) . Hydrops ( 2 ) Imantodes (3) Lachesis ( 1 ) Lampropeltis ( 1 ) _. Leimadophis (8) Leptodeira ( 2 ) 1 3 2 2 10 5 1 1 „ 1 6 1 2 .. 2 3 1 ._ __ 1? 1 1 4 1 1 _ 1? 2 „ .. 1 1 .. 1 .. .. 2 __ 2 _ __ 2 „ 1 1 1 1 1 1 1 1 __ 1 1 1 - -- -- 2 - 1 __ __ 1° 1 1 1° r 2° 2° __ 1* go 1" i' 1? 2 1 4 3 2 6 1 1 ] 1 1 3 7 4 2 1 4 7 3 1 1 1 2 1 2 2 !• 1 2 4 - 5 5 2 1 -- 1 -- 1 1 1 2 i 1 .. 1 1 1 i 3 ._ 1 _ 2 __ 1 1 3 2 2 1 -- 1 2 -- 2 2 2 1 1 - - -- 1 2 2 .. 1 1 1 1 1 .. 1 1 1 2 1 2 1 1 1 2 1 240 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Appendix 9:1 (concluded) Genera Leptophis (4) Leptotyphlops (12) _ Liotyphlops (5) Liophis (9) Lygophis (2) Lystrophis ( 1 ) Mastigodryas ( 3 ) Micrurus (23) Ninia (2) Nothopsis ( 1 ) Oxybelis ( 4 ) Oxyrhopus (5) Philodryas ( 4 ) Phimophis (1) Pliocercus ( 1 ) Pseudoboa (3) Pseudoeryx ( 1 ) Pseustes (4) Ptychophis ( 1 ) Rhadinaea (10) Rhinobothyrum (2) _ Saplienophis ( 1 ) Scaphiodontophis ( 1 ) Sibon (1) Sibynomorphus ( 1 ) _ Siphlophis (4) Sordellina ( 1 ) Spilotes (1) Stenorrliina ( 1 ) Sy nophis (2) Tantilla (4) Thamnodynastes (3) Trachyboa (1) Trctanorhinus ( 1 ) .... Tripanurgos ( 1 ) Tropidodryas (2) Tropidophis (3) Typhlophis ( 1 ) Typhlops ( 4 ) Ungaliophis ( 1 ) Xenoboa ( 1 ) Xenodon (6) Xenopholis ( 1 ) Eastern Amazon 2 5 Western Middle Widespread Amazon Amazon Amazon .. 1 7 - 3 2 Atlantic Forest 1 2 1 3 1 1 2 3 1 1? 5 Choco Forest 3 3 4 Northern Forests 1 2 10. The Herpetofauna of the Guianan Region Marinus S. Hoogmoed Rijksmuscum van Natuurlijkc Historic Postbus 9517 2300 RA Leiden, The Netherlands Although this paper deals with a highland fauna, it is not limited to the reptiles and amphibians that occur at elevations of more than 1000 m. One of the main reasons for this is that "highlands" above 1000 m in the Guiana area are few and occupy only a very small part of the total area of the Guiana Shield. Another reason is that our knowledge of the herpetofauna at higher elevations in Guiana is still very fragmentary. These facts prompted me to deal with the herpetofauna of the entire Guiana Shield. The coast of the Guianas was discovered in 1499 by Alonso de Ojeda and Amerigo Vespucci. After the discovery of the so-called "Spanish Main" or "Wild Coast," numerous expeditions tried to explore the interior in search of the fabulous El Dorado. Most famous of these adventurers was Sir Walter Raleigh, who undertook several expeditions into the interior of Guiana. Zoologically, these expeditions were of no importance whatsoever. From about the beginning of the 18th Century, zoological specimens, main- ly from Surinam, started to reach Europe, and an important percentage of the species of reptiles and amphibians described by Lin- naeus in 1758 originated from the Guianas. One of the first scientific explorers of the interior was Von Humboldt, who in 1801 visited the Rio Orinoco and the Cassiquiare Canal (Gleason, 1931). In 1835 Sir Robert Schomburgk started his explorations in Guy- ana and adjacent countries in order to settle the frontiers. During these explorations, zoo- logical collections were made that supplied a wealth of new data. Since Schomburgk's travels, an increasing number of scientific expeditions penetrated into the interior of lowland Guiana and it would lead us too far astray to try to deal with them here in any detail. I make an exception for the expedi- tions exploring the tepuis in Venezuela and Guyana, several of which even at the present day remain unvisited. The most renowned of these tepuis is Roraima with an altitude of 2810 m, discovered in 1838 by Robert Schom- burgk and climbed for the first time in 1884 by Im Thurn and Perkins. The first zoological collection ever made near any tepui was assembled there in 1842 by Richard Schomburgk. Other collec- tions were made at the foot in the 1880's. The first herpetological specimens from the summit were secured by Quelch and McCon- nell in 1894. In 1898 they made a second expedition to the summit plateau. The ma- terial of these expeditions contained several new species. Boulenger (1895, 1900) studied them and described the frogs OreophryneUa quelchii, O. macconneUi, Otophryne robusta and Hylodes marmoratus, and the lizards Neusticurus rndis and EuspondyJas leucostic- tus. Of these species, only the first and the last came from the summit of the mountain; the other species were collected at the base. The next zoological expedition, on a much larger scale and under the auspices of the American Museum of Natural History, visited Roraima in 1927-28, spending two weeks on the summit and about two months at the base (Tate, 1928, 1930a,b, 1932, 1939; Chap- man, 1931). Among the material collected were several reptiles and amphibians. In 1971, 1973 and 1974 Roraima was visited again, this time by parties with herpetologists as members (Warren, 1973). Their collec- tions contained many novelties. In 1928 Mount Duida, at the western end of the series of tepuis, was explored zoo- logically. The expedition was the first that succeeded in climbing the mountain and spent three months at the summit. Among the herpetological material collected were the types of the teiid lizards Pantodactylus tyleri and Arthrosaura tatei, of the hylid frog Stef- 241 242 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 ania goini, and of the leptodactylid frog Elosia duidensis. In 19.37-38 Auyantepui was zoologically explored. From 1938 until the present, considerable exploration took place in the tepui region, mostly ornithological and botanical, but as a by-product many herpe- tological specimens were collected, some of which at the time being new to science were described by Roze ( 1958a,b; Auyantepui, Chimantatepui), Rivero (1961, 1965, 1966, 1967a,b, 1968a-d, 1970, 1971; Duida, Mara- huaca, Chimantatepui, La Escalera region and other tepuis) and Lancini (1968; Cerro Jaua). The most recent biological explora- tion of some large tepuis were the expeditions to Cerro Jaua and Sarisarinama in 1974 and to Cerro Yapacana in 1978. These were some of the rare expeditions in which herpetologists participated (Nott, 1975; Orejas-Miranda and Quesada, 1976). The herpetological results of these expeditions have not yet been pub- lished. DELIMITATION AND DESCRIPTION OF GUIANA Guiana is the area bordered by the Rio Orinoco, the Cassiquiare Canal (connecting the Orinoco and Amazon drainages), and the Rio Negro in the west, by the Rio Amazonas in the south and by the Atlantic Ocean in the north and the east. The area comprises three political units in their entirety, namely, Guy- ana, Surinam and French Guiana. Of Vene- zuela it comprises the Estado Bolivar and the Territorio Federal Amazonas, known under the common denomer Guayana. Of Brasil it comprises the Territorio do Amapa, the Territorio de Roraima and those parts of the states of Para and Amazonas that are situated north of the Rio Amazonas and Rio Negro. Recently Lescure ( 1977 ) and Des- camps et al. (1978) defined Guiana as the area bordered in the west by the Rio Barama (Venezuela) and in the southeast by the Rio Araguari (Brasil). The southern border would be formed by the watershed between rivers emptying directly into the Atlantic Ocean and rivers belonging to the Amazonian drainage. In my opinion, this definition of Guiana is artificial and not in accordance with the biogeographical and geographical data (Fig. 10:1). The Serra Acarai and the Tumuc Humac Mountains, forming the divide between the French authors' Guiana and Amazonia apparently do form a geographical barrier for a number of endemic species (mainly frogs), but this is too small a pro- portion of the entire fauna to justify the definition of the Guiana area as they do. Far more species are spread on both sides of the divide and occur both in the Orinoco and Amazon Basin (Haffer, 1974; Miiller, 1973). In a discussion of the Guiana herpetofauna I think it is better to take into consideration biogeographical data of the majority of the (herpeto) fauna being studied, rather than rely only on those of some endemic frogs. In that way the biogeographical definition of Guiana, as accepted here, agrees closely with the geographical, geological and climatolog- ical data. However, there are good grounds for considering the Guiana of the French authors as a subregion of the Guiana as here defined. Geologically, this area is a unit known as the Guiana Shield (Gansser, 1954; Fittkau, 1974), of which small parts are situated west of the area as here delimited (Fig. 10:2). Along the edges, notably in the north, the east and the south, there are belts of alluvial deposits; however, the core is made up of pre-Cambrian metamorphic and igneous rocks. Together with the Brasilian Shield, it can be considered as part of the geological foundation of South America. Since Paleozoic times these shields have not been submerged. During the Mesozoic both shields were con- nected, for the Amazon was not yet present. During the Late Cretaceous the area was slightly uplifted and the first signs of the present Amazon Basin became visible. In the Tertiary there was a further uplift ( Haffer, 1974). The higher, central parts of the Guiana Shield are covered with sandstone remnants of the Roraima Formation. Deposition of this sandstone took place in Proterozoic time, 1600-1800 m.y.b.p., as stream and delta de- posits laid down in continental to epiconti- nental environments (Priem et al., 1973). After uplift, this formation covered the Gui- ana Shield as an extensive sandstone plateau or tableland, on which the early Guianan flora 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 243 Fig. 10:1. Map of Guiana, showing the borders of defined by Descamps et al. (1978) and by Lescure are gray and indicated by numbers: 1 = Imeri Refuge, Refuges. The line of fine dots (in this and the follow Mapa de la Guayana, mostrando los limites del terri finido por Descamps et al. (1978) y por Lescure (1977) en gris, indicados con numeros: 1 = Refugio de Imeri, 4 = Refugios de Tepuyes. La linea punteada f\na (en torno de 200 m. the area as here defined (heavy broken line) and as (1977) (heavy dotted line). Presumed forest refugia 2 = Guiana Refuge, 3 = Imataca Refuge, 4 = Tepui ing maps of Guiana) represents the 200 m contour line. torio definido aqui (linea cntrecortada gruesa) y el de- (linea gruesa punteada). Supuestos refugios forestales 2 = Refugio de Guayana, 3 = Refugio de Imataca, este y los siguientes mapas) representa la linea de con- developed (Maguire, 1970). During the Cre- taceous and Tertiary uplift of the area, ero- sion shaped the present-day table mountains or tepuis. These mountains consist of layered, unfossiliferous, pink sandstones, with dolerite dikes and sills, reaching a maximum thickness of about 2400 m in Auyantepui in southeastern Venezuela and decreasing to 700 m in the Tafelberg in central Surinam (Haffer, 1974). At present, the Roraima Formation covers an area of about 450,000 km2 and is spread over a total area of 1,200,000 km2 in Venezuela, Guyana, Brasil and Surinam (Priem et al., 1973). The greater part of the Roraima sand- stone is concentrated in the Gran Sabana region of Venezuela and the adjacent parts of Brasil and Guyana, with many isolated remnants in the western part of the Estado Bolivar and in the Territorio Federal Ama- zonas and two outlying remnants in eastern Guyana and in central Surinam (Bisschops, 1969; Priem et al., 1973) (Fig. 10:2). Some geologists (e.g., Priem et al., 1973:1677) are of the opinion that "It is impossible to decide whether occurrences represent erosional rem- nants of a once-continuous cover or sediments deposited in a number of isolated basins." However, most geologists and biologists re- gard the present-day sandstone mountains as remnants of a once-continuous sandstone cover. Also, one could imagine a combination of the possibilities, in which the western 244 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fig. 10:2. Map of Guiana showing the extent of the Guiana Shield (gray area) and of the Roraima sandstone formation (black). The white areas in the sandstone represent gabbro (after Bisschops, 1969; Gansser, 1954; Priem et al., 1973). Mapa de la Guayana mostrando la extension del Escudo Guayani (gris) y de la formation Roraima (negro). Las areas blancas en el gres de Roraima representan gabbro (segiin Bisschops, 1969; Gansser, 1954; Priem et al, 1973). Roraima Formation once formed a continu- ous cover and the two outlying areas in De- merara and Surinam could have been de- posited in isolated basins. However, decisions on this subject should be reached by geolo- gists, although perhaps biologists may con- tribute to the solution. In this paper I adhere to the view that the tepuis are remnants of a once-continuous formation. It seems useful to state that the arch of sandstone tepuis in southern Colombia, west of the Rio Orin- oco, and ending quite close to the Andes in the Sierra de Macarena, is not of the same as the Roraima Formation (Lescure, 1977). This sandstone is much younger and probably represents a deposition of erosional products of the Roraima Formation (Haffer, 1974; Paba-Silva and Van der Hammen, 1960). These tepuis probably arose by "Block fault- ing in conjunction with the Andean uplift toward the end of the Tertiary and at the beginning of the Pleistocene" (Haffer, 1974). The Guiana Shield consists of an elevated portion in the west, rising from sea level to well over 1000 m in relatively extensive areas. This portion bears the sandstone tepuis of which the highest attain a height of 2810 m (Mount Roraima) and 3014 m (Serra de Neblina). These tepuis are mostly flattopped, with perpendicular cliffs several hundred meters high separating the plateau summits from the talus formed by the accumulation of erosional products at the base of the cliffs (Figs. 10:3-4). The western part is separated from an eastern elevated part by a depression formed by the river systems of the Rio Branco and the Essequibo River, which may be connected 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 245 Fig. 10:3. The Tafelberg in Surinam, the easternmost remnant of the Roraima formation. Note the flat top, the steep, bare upper reaches of the flanks and the sloping talus covered with forest. La Montana Tafelberg en Surinam, el residuo mas oriental de la formaeion Roraima. Observase la cumbre aplanada, las flancos escarpados y rasos en su parte superior y el tabid inclinado y cubierto de selva. Fig. 10:4. View of several tepuis south of El Manteco, Estado Bolivar, Venezuela. Vista de algunos tepuyes al sur de El Manteco, Estado Bolivar, Venezuela. 246 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fig. 10:5. Map of Guiana with the contour lines of 200, 1000 and 1500 m (after Mayr and Phelps, 1967). Mapa de la Guayana con lineas de contorno de 200, 1000 y 1500 m (segi'm Mayr y Phelps, 1967). during the rainy season when large areas are inundated. The eastern part is much lower than the western one, reaching a maximum height of 1280 m (Julianatop) in central Surinam. From the divide between the Ama- zon Basin and rivers flowing north to the Atlantic Ocean (nowhere over 1000 m high), the country gradually slopes down to sea level. Thus, although the topography of east- ern Guiana may be rather rugged, with many mountain ranges and valleys separating them, the area hardly ever exceeds 1000 m. In western Guiana, the topography is more or less the same, but on a higher level, with the consequence that more extensive areas are over 1000 m. However, superimposed on the Guiana Shield in this region are sandstone tepuis that may reach elevations of almost 3000 m (Fig. 10:5). The Guiana Highlands are also known as Pantepui. The greater part of the area is covered by tropical rainforest, but savannas also play an important role. In the western and north- western portion of the shield there are sa- vannas more or less continuous with the llanos of central Venezuela. In the three Guianas there is a band of coastal savannas on white sand, reaching from Georgetown in the west to Cayenne in the east. East of Cayenne and in Amapa the white sand is absent and some extensive swamps in that region are dry savannas in the dry season. In Amapa this coastal belt is bordered on the west by a belt of cerradrj — savanna with isolated trees. Iso- lated, extensive savanna complexes of the cer- rado type are present (Hills, 1969) in south- western Guyana (Rupununi), in southeastern Venezuela (Gran Sabana) and on the border between Surinam and Brasil (Sipaliwini/Paru savannas). Smaller, isolated savannas occur in Surinam and in Venezuela both on the Roraima sandstone and on other substrates (Fig. 10:6). On the higher points, starting at about 800 to 1000 m, cloud forest occurs with 1979 HOOGMOED: HERPETOFAUNA OF GUI AN AN REGION 247 Fig. 10:6. Map of Guiana showing the distribution of forest and savannas. Forested areas white, savannas gray, inundated savannas hatched. The zone with lower rainfall (cf. Fig. 10:7) has been indicated with heavy broken lines (after Hills, 1969; Muller, 1973; Oldenburger et al., 1973; Prance, 1973; Romariz, 1974 and personal field data). Mapa de la Guayana mostrando la distribution de selva y sabana. Selva en bianco, cerrado en gris, campo rayado. La zona menos lluviosa (cf. Fig. 10:7) se ha indicado con una linea cntrecortada gruesa (segun Hills, 1969; Muller, 1973; Oldenburger et al., 1973; Prance, 1973; Romariz, 1974 y observacioncs personales). thick layers of mosses covering the trees, shrubs and the ground. This is especially so on the talus of many of the tepuis. The pla- teau summits of the smaller tepuis have only a shallow layer of soil, which is insufficient to support forest; thus, the vegetation is low, often savannalike. The plateau summits of the larger (more extensive) tepuis is more diver- sified, and in some places a sufficiently deep layer of soil has accumulated to support mod- erately high forest; however, in other places there is only sparse vegetation (Chapman, 1931; Gleason, 1931; Maguire, 1945, 1955, 1970; Mayr and Phelps, 1967; Tate, 1928, 1930a,b, 1932, 1938a,b, 1939; Tate and Hitch- cock, 1930). The climate of the region under discussion is characterized by two dry and two rainy seasons per year. Their duration and the period of the year in which they fall are some- what variable, and at higher elevations the distinction between dry and rainy seasons may be hardly evident, but in general this division holds true for the greater part of the area. Within the area a wide zone with dis- tinctly lower rainfall extends northwest-south- east connecting the llanos of Venezuela with the caatinga and cerrado region of cen- tral Brasil (Figs. 10:6-7). Within this zone, which roughly covers the extreme southwest- ern part of Surinam, southern Guyana, south- eastern Venezuela and the Guianan part of Para, the annual rainfall is 2000 mm or less. To the northeast and to the southwest the 248 MONOGRAPH MUSEUM OF NATURAL HISTORY 70 60 50 NO. 7 >3000 ::S^™s:SS^S '00 0- : 15 00 H 2500- 3000 (1000 2000 — 2500 1500 — 2000 mountains above 2000 Fig. 10:7. Rainfall (mm) distribution in northern South America (after Prance, 1973; Reinke, 1962). Distribution dc la lluvia (mm) que cae anualmente en la parte norte de la America del Sur (segun Prance, 1973; Reinke, 1962). annual rainfall increases, reaching maxima of over 3000 mm in northeastern French Guiana and coastal Amapa and of some 2500 to 3000 mm in the upper Orinoco region (Reinke, 1962; Prance, 1973). Mean annual tempera- tures are between 24° and 27°C in the low- lands and decrease with increasing altitude. During the last few years it has become increasingly clear that Pleistocene and Holo- cene climatic changes had a profound influ- ence on the vegetation of northern South America, especially in Amazonia and adjacent regions. It is presumed that during dry cli- matic phases the rainforest disappeared from large stretches of the Amazon Basin and was restricted to refuges, mostly along its periph- ery (Brown et al., 1974; Haffer, 1969, 1974, this volume; Vanzolini, 1970a). Inversely, during the wet climatic phases the rainforest spread again from the refugia and the sa- vanna vegetation and fauna retreated into refuges. Of importance in this connection are the Guiana, Tepui and Imeri forest refuges of Haffer (1969, 1974); the Guiana (forest), Pantepui (montane forest) and Roraima (sa- vanna) centers of Midler (1973); and the Guiana, Imataca and Imeri refuges of Prance ( 1973 ) , all of which are situated within the limits of the area considered here (Fig. 10:1). The aforementioned belt with a lower pre- 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 249 cipitation played an important role in the distribution of plants and animals during the different climatic periods. At present, the savanna complexes of interior Guiana are situ- ated in this belt (Fig. 10:6). However, during dry climatic phases probably much larger areas of it were covered with savanna, thereby providing a dispersal route for savanna in- habitants, either to the north or to the south, and at the same time forming a barrier to east-west dispersal of forest inhabitants. On the other hand, both areas with higher rain- fall, adjacent to this dry belt, are thought to be the areas where forest refuges were situ- ated during arid phases — to the northeast the Guiana refuge, to the southwest the Imeri refuge. During wet climatic phases, the forest spread from these refuges and invaded the savanna belt, fragmenting it into several iso- lated savanna complexes, as is the case today (Fig. 10:6). The montane forests covering the slopes of the tepuis in southern Venezuela can be regarded as isolated occurrences of rainforest on places with favorable climatic conditions (high elevation, high rainfall) gen- erally having unfavorable climatic, and pos- sibly edaphic, conditions (Gran Sabana area). These forests, which are different from the tropical lowland rainforests, probably were only connected with the lowland forests dur- ing very wet climatic phases. Although Mid- ler's concept of the Guiana center is much wider ( and based on several different groups ) than Haffer's, Prance's and others' Guiana refuge, I think we can synonymize the two without problems; the same is true for the Pantepui center and the Tepui refuge. There is no parallel in Midler's concepts of Haffer's and Prance's Imeri refuge. The Imataca ref- uge, which was postulated by Prance ( 1973 ) for plants is only substantiated further by data from butterflies (Brown et al., 1974) (Fig. 10:8). HERPETOFAUNA Although since 1894 quite a substantial number of reptiles and amphibians has been collected from the sandstone tepuis, only a small part of it was collected by herpetolo- gists. This partly explains our scant and frag- mentary knowledge of these groups. Thor- ough herpetological exploration of the tepui region, starting with the now easily accessible La Escalera region in eastern Venezuela, probably will provide us with many interest- ing finds. Because our present knowledge is so fragmentary, it is often difficult to decide whether a certain species is really restricted to one tepui or not. The available data permit some zoogeographical conclusions, but those regarding the so-called endemics certainly have to be drawn with much reserve. Presently a total of 408 species of reptiles and amphibians is known to occur in the Guiana region (Table 10:1, Appendix 10:1). Seventy-six species are represented by 108 subspecies, which raises the number of spe- cies-group taxa for the region to 440. The herpetofauna of Guiana can be allocated to eight groups, which in turn can be partly subdivided. 1. Endemic in Guiana region: A. Highland (over 1000 m)— 18 am- phibians, 9 reptiles. B. Lowland (below 1000 m) — 74 am- phibians, 50 reptiles. 2. Amazonian: A. Periferal along western and northern margin of basin — 10 amphibians, 19 reptiles. B. With disjunct populations in upper Amazonia and near the mouth of the Amazon — 2 amphibians, 1 reptile. C. Species of Amazon Basin occurring on southern edge of Guiana and along eastern margin, where they may reach French Guiana — 3 am- phibians, 11 reptiles. D. Widespread Amazonian, occurring throughout greater part of Guiana — 39 amphibians, 62 reptiles. 3. Widespread species (distribution ex- tending from Mexico or Central Amer- ica over entire cis-Andean tropical South America): 12 amphibians, 35 reptiles. 4. Species reaching their eastern distribu- tion limit on the Guiana Shield, from Central America, northwestern South America or upper Amazonia: 11 am- phibians, 17 reptiles. 250 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 haffer,i969 , 1974 vanzolini,1970 prance.1973 brown et al.,1974 Fig. 10:8. Location of forest refugia during arid periods according to several authors; numbers as in Fig. 10:1. Situation de los refugios de selva durante los periodos secos segtin algunos autores; los numeros como en la Fig. 10:1. 5. Species from southeastern or central Brasil reaching Guiana, mostly not farther than French Guiana, some reaching Surinam or even Venezuela: 8 amphibians, 13 reptiles. 6. Cosmopolitan species: 0 amphibians, 6 reptiles. 7. Species imported from the Caribbean region: 1 amphibian, 4 reptiles. 8. Species with limited or uncertain dis- tributions that may occur in the region: 0 amphibians, 3 reptiles. The last three groups in the tabulation above are of no importance in the following considerations. The five cosmopolitan species of sea turtles and one species of cosmopolitan gecko are of no importance here. It is evident that the imported species do not need further attention. Of the three species in the last group, it has not been established beyond doubt that they occur in the Guiana region. Thus, there remain five important groups, totaling 177 amphibians and 217 reptiles, that reflect the complicated history of the Guiana herpetofauna and that are dealt with in detail later. Considerable differences exist between the percentages of reptiles and amphibians in five different groups and subgroups (Fig. 10:9, Table 10:2). These groups are highland (1A) and lowland (IB) endemics, disjunct (2B) and widespread ( 2D ) Amazonian, and wide- 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 251 Table 10:1. — Composition of the Guianan Herpeto- fauna. The columns contain the total numbers of species in a given family that inhabit the Guianas and the numbers of species that are endemic to the Guianas. Species Family Total Endemic Anurans Pipidae 2 15 1 38 17 75 2 7 8 165 2 2 9 13 178 4 1 1 2 1 4 5 18 1 4 5 2 8 4 1 5 6 91 9 6 132 13 18 1 33 65 10 10 230 408 1 Dendrobatidae ... 9 Ranidae Leptodactylidae . 12 Bufonidae Hylidae ... 8 43 Pseudidae ._ . _. Centrolenidae 6 Microhylidae Total Anurans 4 83 Caecilians Rhinatrematidae Typhlonectidae 2 1 Caeciliidae Total Caecilians Total Amphibians Chelonians Cheloniidae Dermochelyidae 6 9 92 Kinosternidae ... .. Testudinae .. . .. Emydidae Peloniedusidae Chelidae Total Chelonians Crocodilians Crocodylidae .... .. .... — — Alligatoridae Total Crocodilians Snakes Anomalepidae Leptotyphlopidae Typhlopidae Aniliidae ... Boidae Dipsadidae Colubridae 20 Elapidae . Crotalidae 2 Total Snakes . Lizards Gekkonidae Iguanidae 29 4 3 Scincidae Tei idae Total Lizards ... Amphisbaenians Amphisbaenidae __ . Total Amphisbaenians Total Reptiles 16 23 6 6 59 Total Amphibians and Reptiles 151 Fig. 10:9. Proportion of the total numbers of species accounted for by each group; numbers of groups as in Table 10.2. A = amphibians, B = reptiles. Porcentaje que representa cada grupo del numero total de especies; numeros de gfupos como en la Tabla 10:2. A = anfibios, B = reptiles. spread (3). Both in the widespread Amazo- nian and in the generally widespread species the percentage of reptiles is distinctly higher than that of amphibians; moreover, in the species reaching their eastern distribution limits in Guiana (4), and in the species com- 252 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Table 10:2. — Composition of the Guianan Herpetofauna. The columns contain the total number of species having a certain distribution and the percentage this group forms of the total number of species in the Guianan Region. Type of distribution Species Amphibians Reptiles Total Number Percentage Number Percentage Number Percentage 1. Endemics 18 74 10 10.17 41.81 51.98 5.64 1.12 1.69 22.03 30.48 6.77 6.21 4.51 99.95 9 50 19 1 11 62 35 17 13 217 6 4 3 230 4.15 23.04 27.19 8.75 0.46 5.06 28.57 42.84 16.12 7.83 5.99 99.97 27 124 29 3 14 101 47 28 21 394 6 5 3 408 6.85 2. B. Lowland Amazonian A. Periferal 31.47 38.32 7.36 Tt Disjnnrt n RnsiTi 2 3 0.76 3.55 D. Widespread 39 25.63 3. 4 Widespread Reaching eastern limit 12 11 37.30 11.92 7.10 5. From SE. or C. Brasil __ .._. 8 5.32 Subtotal Guiana Region _ Cosmopolitan 177 99.96 fi. 7. 8. Imported from Caribbean Uncertain distribution „ .. 1 178 ing from the southeast (5), the percentage of reptiles is slightly higher than that of am- phibians. In my opinion, this is the clue to the explanation of the differences observed. It is a reflection of the greater mobility of reptiles, as compared with the more sedentary habits of amphibians, which are restricted by their mode of reproduction. As there is a strict dependence on the kind of water ( stand- ing or running, large or small body of water), which for most species is very specific, this further restricts the possibilities for amphibian dispersal. Also species that have direct de- velopment on land are still dependent on water in the form of a high humidity. This explains the high endemism of this group in Guiana and also the higher percentage of disjunct Amazonian species of amphibians (Lynch, this volume). In all cases, the am- phibians did not have the chance to expand their ranges far beyond the region of origin or from that of isolation during one of the climatic phases. Reptiles, given the same time and being independent of water for their reproduction, had a much greater rate of dis- persal and either spread beyond the borders of Guiana (and thus ceased to be endemics of that region) or closed the gap between dis- junct populations of one species. I consider those species having distribu- tions that do not, or hardly, exceed the bor- ders of Guiana to be endemics. This is a fairly large area. Among the endemics several subdivisions can be recognized; the one be- tween highland and lowland endemics will be discussed later. The other subdivision is be- tween local and wide-ranging endemics, but this is partly artificial and mainly reflects our fragmentary knowledge of the species con- sidered to be local. Altitudinal Distribution Of the indubitably native species only 55 ( 31% of total amphibians ) species of frogs and 38 (18% of total reptiles) species of reptiles (lizards and snakes) (Appendix 10:1) occur at elevations of more than 1000 m. No caecil- ians, crocodilians, chelonians or amphisbae- nians are known from above 1000 m. Of the species occurring over 1000 m, 37 frogs and 29 reptiles also occur below 1000 m, which leaves 18 frogs (10% of total) and 9 reptiles (4% of total) restricted to elevations of more than 1000 m. All of these are highland en- demics, restricted to the western part of the Guiana Shield. Highland endemics. — Most of these spe- 1979 HOOGMOED: HERPETOFAUNA OF GUI AN AN REGION 253 cies have restricted distributions, usually con- sisting only of the summit or talus slopes of one or a few adjacent tepuis (Figs. 10:10-11). As stated before, this either reflects our frag- mentary knowledge of the herpetofauna of the Guiana Highlands, or these distributions are real and the comparable habitat on other tepuis is occupied by a related species. How- ever, this has only been documented (and poorly so) for the endemic frog genus Stefania. The bufonid genus Oreophrijnella from Roraima and Auyantepui is considered to be a specialized derivative from the general ate- lopodid stock and to have evolved in isola- tion since the Early Tertiary or the Cretaceous (McDiarmid, 1971). The same is true for the microhylid frog Otophryne (not an altitudinal endemic), composed only by O. robusta with two subspecies — one restricted to high eleva- tions on Chimantatepui, the other occurring in the greater part of interior Guiana at eleva- tions of 200 to 1666 m. Like Oreophrijnella, Otophryne also shows a combination of prim- itive, derived and unique characters. This is most easily explained by assuming that these frogs were subject to a long evolution in isolation on the sandstone formation, prob- ably since the Cretaceous or Early Tertiary; the invasion of tropical lowland Guiana by Otophryne may be considered as secondary. According to Lynch (pers. coram.), the lepto- dactylid frog Hylodes duidensis belongs to an undescribed genus of the tribe Eleutherodac- tylini. Its relations are not clear, but it may have developed on the Guiana Shield as a highland derivative of the eleutherodaetyline stock. Stefania is an endemic, egg-brooding hylid frog genus clearly related to the north- ern Andean Cryptohatrachus. According to Rivero (1970), these frogs can be divided into the Stefania goini group, with two spe- cies, and the Stefania evansi group with five species (and three undescribed ones). One member of the goini group occurs on Mount Duida in the west, the other on Chimantatepui in the east. One member of the evansi group occurs on Cerro Marahuaca and the others on the eastern part of the Roraima Formation. The distribution of the members of these species groups can be explained most easily by assuming that the genus Stefania arose from hylid stock in the Guiana Highlands, probably prior to the Oligocene. Initially, the stock split into two groups, which during the most recent uplift of the area in Mio-Pliocene times became isolated on several tepuis and since differentiated into the several species now composing the two species groups. The occurrence of Stefania evansi in lowland areas may be regarded, as in Otophryne, as being secondarily, induced by the Pleistocene cli- matic changes, which lowered the general temperature of the area by about 3°C (Van der Hammen, 1974). Species showing a slight degree of Andean relationships are members of the frog genera Centrolenella and Eleutherodactylus, and of the colubrid snake genus Atractus; all three genera probably evolved in or near the Andes, either in the foothills or in the lowlands, and subsequently spread to the east. However, the endemic altitudinal species belonging to these genera have no direct relations with Andean species and probably are altitudinal forms derived from lowland species. The matter is slightly different for the species of Euspondylus, a genus of teiid lizards of Andean origin, members of which live at medium to high altitudes in the Andes from Peru to Venezuela; two species reached the higher altitudes of the Guiana Shield, pos- sibly during a time of Pleistocene climatic depression. The altitudinal endemics of the tree frog genus Hyla all apparently are re- lated to lowland species groups. Riolama, a monotypic, endemic teiid lizard genus restricted to the summit of Mount Roraima, is known only from the type speci- men. Presumably it is related to Leposoma and its relatives, but its history is not clear. It may have evolved from lowland microteiids by isolation on a sandstone tableland prior to the Oligocene, as was probably the case in Stefania. The colubrid snakes Liophis and Thamnodynastes occur in lowland Amazonia and Guiana, but they seem to be of southern Brasilian origin and to have evolved into sev- eral altitudinal species in Guiana. The near- est relative of the iguanid lizard, Tropidurus bogerti, is T. torquatus hispidus (R. Ether- idge, pers. comm.), a member of a species or 254 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fig. 10:10. Distribution of some endemic species within Guiana. Distribution de algunas espeties endemicas en la Guayana. 1 = Hyla mtdtifasciata. 2 = Aparasplicnodon venezolanus. 3 = Bufo nasicus, Hijla sibleszi, Hyla lemai. 4 = Hyla ornatissima. 5 = Otophryne robusta. 6 = Allophrync ruthveni. 7 = Hyla ginesi, Stefania goini, Stcfania marahuaquensis, "Hyludes" duidensis. species complex, which may be of southeast- ern Brasilian origin. The few altitudinal en- demic subspecies all have evolved by isolation at higher altitudes from lowland relatives of different origins. Attempts to explain the origin of the fauna of Pantcpui have been based on the distribu- tion of birds (Chapman, 1931; Haffer, 1974; Mayr and Phelps, 1967), mammals (Tate, 1939), frogs (Rivero, 1965) and snails (Haas, 1957). Because of different dispersal abilities and different geological ages of the groups concerned, these studies came to different con- clusions. For instance, birds supposedly were able to reach Guiana from the Andes by (simply stated) flying from one mountain with suitable climate to the next. This pos- sibility doesn't exist for the other groups. The distribution of the endemic Guianan herpeto- fauna can be explained with the aid of the following theories. 1. The Mountain Bridge Theory as pre- sented by several authors (Todd and Carriker, 1922; Haas, 1957) apparently is useless, because there is no geological evidence for a connection of southern Venezuela with the Andes. As has been pointed out, the sandstone mountains (Sierra de Macarena) in southern Co- lombia are not the remnants of such a bridge. 2. The Plateau Theory, starting from the assumption that "a more extensive tableland probably did exist on the Guayana shield during the Mesozoic and Tertiary, prior to an intensive ero- sional dissection" (Haffer, 1974:163) is useful to explain the presence of several 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 255 Fig. 10:11. Distribution of some endemic species within Guiana. Distribution de algunas especies endemicas en la Guayana. 1 = Hyla proboscidea. 2 = Dcndrobatcs tinctorius. 3 = Dendrobates leucomelas. 4 = Hyla rodriguezi, Stcfania scalac, Eleutherodactylus pulvinatus. 5 = Dendrobates azurcus. 6 = Pliyllobatcs pulchripectus, Ama- pasaurus tetradactylus. relicts, such as the frogs Oreophrynella and Otophryne. A slightly modified version, starting with the assumption that the Roraima Formation underwent orogenic movements that shaped it into a mountain range before erosion graded it into a plateau, which in turn was uplifted and eroded into its present shape, serves well to explain the distri- bution of the genus Stefania ( Rivero, 1970). The Modified Cool Climate Theory de- parts from the assumption that during the glacial periods of the Pleistocene, the lowlands between the Andes and Guiana, and within the Amazonian basin had a cooler climate. This indeed was true, the temperature of the low- lands having been about 3°C lower than at present (Van der Hammen, 1974), but this was not sufficient to make the lowlands subtropical instead of tropical, as had been assumed for- merly (Chapman, 1931; Tate, 1939). However, it may have facilitated the dispersal of certain organisms, because all life zones on mountains shifted to lower altitudes, thus creating suitable habitats for subtropical organisms in places where they were formerly ab- sent. These still widely-separated, sub- tropical habitats could have been of importance for birds. The distribution of amphibians and reptiles apparently related to Andean taxa could not have gone only via those "stepping stones" but most likely through the lowlands at times of cooler temperatures. 256 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 4. The Habitat Shift Theory assumes that part of the fauna of Pantepui was de- rived from tropical lowland elements that changed their habitat preference. It serves to explain the distribution and occurrence of the majority of the taxa living at higher elevations. They either differentiated in situ after invasion of the highlands by a lowland ancestor (most highland endemics) or are themselves widely distributed in the lowlands surrounding Pantepui and ap- parently have wide ecological ampli- tude. 5. The Distance Dispersal Theory, which assumes that the Guiana Highlands were colonized from distant sources by island hopping, is of no use in explain- ing the distribution of the herpeto- fauna, although it seems to be useful for partly explaining the distribution of flying organisms (mainly vertebrates) (Mayr and Phelps, 1967). Lowland endemics. — The lowland endem- ics are a rather mixed group, containing spe- cies restricted to elevations below 1000 m and species occurring from sea level to well above 1000 m. Several species occur from about sea level to a maximum of 2400 m. In a number of cases [Neusticarns tatei, N. racenisi, N. rudis (all three teiid lizards), Stefania evansi (hylid frog), Otophryne robusta (microhylid frog)] they clearly evolved on part of the sandstone plateau and secondarily invaded the tropical lowlands. Others, like Dendro- bates steyermarki (poison-arrow frog), Hyla ginesi, H. benitezi, H. kanaima, H. lemai, H. sibleszi, Stefania marahuaquensis, S. woodlcyi (hylid frogs) and Eiiparkerella sp. "A" (lep- todactylid frog), have narrower elevational distributions, occurring only from about 600 to 1500 m. They also probably evolved at higher altitudes and secondarily invaded the adjacent lowlands, but apparently their eco- logical tolerance is not so great as that of the species in the first group. The remaining spe- cies occurring above 1000 m are actually low- land species, having arisen in tropical low- lands and from there extended their range by moving up onto the sandstone plateau, often to the base of the tepuis, sometimes even to the summit. Dendrobates steyermarki known from an isolated sandstone mountain in western Vene- zuelan Guiana is most closely related to Andean species of the Dendrobates minutus group (Silverstone, 1975). This is the only Guianan lowland species showing such a link and probably this is a relict of a formerly more widespread group, which became iso- lated from the main body of the group when temperatures increased during one of the Pleistocene climatic phases. Euparkerella, with recent representatives living at low to high elevations in areas per- iferal to the Amazon Basin (Roraima, Andes of Ecuador and Peru, southeastern Brasil), is represented by one endemic species. Its distribution may be explained by assuming that the presently known species are the sur- vivors of a genus that once occupied a more extensive range, covering the entire Amazon region and adjacent territories. When the range of the genus became discontinuous is not clear, but tentatively we may place that event in the early Pleistocene. It was prob- ably caused by the evolution in the Amazon Basin of new groups of litter-adapted frogs. There are five ( monotypic ) lowland en- demic genera [AUophryne (hylid frog), Rhinatrema (caecilian), Peltoccphalus (pelo- medusid turtle), Amapasaurtis (teiid lizard) and Mesobaena (amphisbaenian)], of which only AUophryne, Rhinatrema and Peltocepha- lus have more or less extensive ranges. Ama- pasaurtis is restricted to a small area in east- ern Guiana and Mesobaena to western Gui- ana. The ranges of the first four genera and of many endemic species coincide with that of the postulated Guiana Forest Refuge ( Haf- fer, 1969, 1974; Lescure, 1975, 1977) or with parts of it (Figs. 10:10-11). Therefore, it seems possible that these genera arose in this refuge during the early Pleistocene. The same holds true for most of the other lowland endemic species, but here we might date the specific diversification as late Pleistocene. Endemic subspecies of species not en- demic to Guiana probably arose during one of the more recent (late Pleistocene or Holo- 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 257 cene) dry or wet climatic phases occurring in northern South America. A few of the lowland endemics occurring in western Guiana, mainly around the head- waters of the Rio Orinoco, seem to strengthen Haffer's view of an Imeri forest refuge. These endemics include one monotypic genus (Mesobaena, amphisbaenian), eight species [Aparasphenodon venezolanus (tree frog), Dendrobates leucomelas and D. steyermarki (poison-arrow frogs), Atractus insipidus, Hel- icops hogei, Liophis canaima (all colubrid snakes), PhyUodactylus dixoni (gekkonid lizard), Crocodijlus intermedins (crocodile)] and three subspecies [Hydrops triangularis venezuelensis, Leptophis ahaetulla copei (both colubrid snakes), Micrurus surinamen- sis nattereri (elapid snake)]. A similar situa- tion is known in birds, with one endemic genus and nine endemic species (Haffer, 1974). Assuming a similar divergence rate for the organisms involved, this seems to point to at least three arid phases during which the forest fauna was isolated in this Imeri forest refuge. Different patterns of distribution exist in Guiana. The endemic species are not evenly distributed throughout the area. As has been noted in the section on altitudinal distribu- tion, all altitudinal endemics are restricted to the western part of the Guiana Shield, the area west of the Essequibo-Rio Rranco De- pression. The ranges of most of the species that supposedly originated on the higher parts of the sandstone area do not extend far be- yond; only a few reach the Essequibo River in the east. Exceptions, like the microhylid frog, Otophryne robusta, and the teiid lizard, Neusticurus rudis, extend their ranges beyond the Essequibo River. The Essequibo-Rio Rranco Depression seems to have been a bar- rier to the eastward distribution of a number of species, mainly Pantepui species. On the other hand, it was a barrier to the westward distribution of a number of species. The ef- fect of this barrier is evident from the ranges of lowland endemics (Fig. 10:11). Of the 74 endemic species of lowland amphibians (Table 10:2, Appendix 10:1) 18 (24%) occur on both sides of the Essequibo-Rio Branco Depression, 32 (43%) only occur east of the depression, and 24 (32%) only occur west of it. Of the 50 endemic species of lowland rep- tiles (Table 10:2, Appendix 10:1) IS (36%) occur on both sides of the Essequibo-Rio Branco Depression, 17 (34%) only occur east of the depression, and 15 (30%) only occur west of it. The picture changes distinctly when the altitudinal endemics also are con- sidered. In that case the number of amphib- ians restricted to the western part of Guiana becomes 42 and the corresponding number of reptiles 24. The percentages change ac- cordingly, for amphibians respectively 18 (22%), 32 (39%) and 42 (51%); for reptiles respectively 18 (31%), 17 (29%) and 24 (41%). Among the widespread endemics the propor- tion of reptiles is considerably higher than that of amphibians; in both the western and eastern endemics the proportion of amphib- ians is higher than that of reptiles, reflecting the greater mobility of reptiles. When only the lowland endemics are considered, the per- centage of amphibian species restricted to the east is distinctly higher than that of species restricted to the west, in reptiles it is only slightly higher. This probably reflects the greater importance of the Guiana Refuge for amphibians, as compared to the importance of the Imeri Refuge. For reptiles, both refuges apparently were equally important. Why the Guiana Refuge was more important for amphibians than for reptiles remains a matter of conjecture. However, possibly it results from the greater dependence of am- phibians on water and moist habitats. Thus, isolation in different refuges was more severe for amphibians than for reptiles; reptiles re- stricted to different forest refuges probably came into contact earlier than the amphibians, thus diminishing the possibilities of having attained reproductive incompatibility. Maybe it was simply a matter of size, the Guiana Refuge having been larger (and therefore possibly harboring more species) than the Imeri Refuge. Perhaps both factors played a role. The Essequibo-Rio Branco Depression also served as a route for lowland Amazonian species invading the northern part of Guiana. 258 120 MONOGRAPH MUSEUM OF NATURAL HISTORY 100 80 60 40 NO. 7 Fig. 10:12. Distribution of species belonging to the groups 2-5 (Table 10:2, Appendix 10:1). Distribution de especies pertenecientas a los grupos 2-5 (Tabla 10:2, Apendicc 10:1). 1 = Pseudopaludicola pusilla (group 4). 2 = Htjla geographica (group 2d). 3 = Lcpidoblepharis festae (group 2b). 4 = Crocodilurus lacertinus (group 2c). Amazonian Species I do not treat the other groups (Figs. 10: 12-13) in the detail that I have done for the Guiana endemics, because they are dealt with by Dixon, Lynch, and Rivero-Blanco and Dixon (this volume). A few species reach parts of Guiana be- cause of certain hydrological features. The occurrence in Guyana of the aquatic Ama- zonian species Melanosuchus niger (the black caiman) and Chehis fimbriatus (the matama- ta) apparently is the result of the rainy sea- son connection between the Rio Branco and the Essequibo River via the flooded Rupununi Savanna. The occurrence of these species in eastern French Guiana can be explained in a similar way, because the extensive coastal swamps and inundated savannas in Amapa, during the rainy season form an unbroken connection between the Amazon and the Oyapoc, Approuague and Mahury basins. In 1979 120 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 100 80 60 40 259 Fig. 10:13. Distribution of species belong to the groups 2-5 (Table 10:2, Appendix 10:1). Distribution de espccies pertenecientas a los grupos 2—5 (Tabla 10:2, Apendice 10:1). 1 = Lysapsus limellus (group 5). 2 = Lysapsus limellus laevis (group 5). 3 = Hyla senicula melan- argyrea (group 5). 4 = Leptodactylus rhodomystax (group 2a). 5 = Phrynohyas venulosa (group 3). Surinam no such connections occur between the Corantijn or Marowijne river systems and the Amazon Basin; this explains the absence of these two species in that country. Other species of the Amazon Valley appar- ently succeeded in reaching eastern French Guiana but did not penetrate farther west. The distribution of a few species with dis- junct populations in upper Amazonia and near the mouth of the Rio Amazonas (Table 10:2, Appendix 10:1, Fig. 10:12) is correlated with areas of high rainfall (over 2500 mm) (Fig. 10:7) and may have been caused by the most recent arid phase, which apparently ended 2000 years ago and caused a separation of the upper and lower Amazonian forests (and the animals living in them) (Haffer, 1974). A number of these species are dis- tributed in an arciform area from Bolivia along the eastern foot of the Andes to the Guianas. This arc can be termed the Ama- zonian Arc. Lescure ( 1977 ) called the north- 260 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 ern part of this arc ( Serra do Navio to Loreto, Peru) the Roraima Arc, because he believed that all sandstone in this region belonged to the Roraima Formation. As pointed out be- fore, this is not the case. However, the exist- ence of an arciform distribution pattern in sev- eral reptiles and amphibians seems to be real (Fig. 10:13). For at least one species, the toad Bufo guttatus, this pattern apparently is caused by its being saxicolous. It is nearly always found associated with rocks; the geo- logical nature of these rocks apparently is not important, as it may consist of either granite or sandstone. The absence of this species in central Amazonia is understandable, because in that area rocks are absent; only alluvial material is present. A number of Guianan en- demics [Otophnjne robusta (Fig. 10:10), Leptodactyhis rugosus], which formerly were thought to be restricted to the northern part of the arc because of their close association with sandstone, now have been found asso- ciated with other types of rocks as well. For most of the species having the periferal or Amazonian Arc distribution it is not pos- sible to explain simply their absence from central Amazonia. Possibly the presence of close relatives or other ecological competitors there is the most important reason. Widespread Species Most of the species in this group appar- ently had their origin in Amazonian South America; from there they dispersed into southern Central America (Fig. 10:13); a few are of Central American origin and dispersed into South America. One example of this last subgroup is the teiid lizard Cnemidophorus 1. lemniscatus, occurring from Honduras to the mouth of the Amazon. This species occurs only along the coast in Guiana. The fact that this species is still extending its range along the lower Amazon (Vanzolini, 1970b) and that it does not occur in the far interior of the Guianas indicate that it is a recent immigrant from the northwest. The presence of forests in southern Surinam and French Guiana, was a barrier to the dispersal of C. I. lem- niscatus ( a savanna inhabitant ) into the large inland, edaphic savannas in the Sipaliwini/ Paru area. Fluctuation of the size of the forests, thereby at times forming a barrier between the inland and coastal savannas, was responsible for the isolation of the inland sa- vannas; lizards living there are distinctly dif- ferent from the populations of the same spe- cies in savannas farther north (Hoogmoed, 1973). Species Reaching the Eastern Limit of Their Distributions on the Guiana Shield Some of these species are of Central Amer- ican origin, others of upper Amazonian, or coastal Venezuelan origin. A number of them are savanna inhabitants that just reach the western part of the Guiana Shield, where the llanos extend east of the Rio Orinoco (the leptodactylid frog Ceratophrys calcarata). A few leptodactylid frogs (Tlujsalaemus pustu- losus, Plenrodema brachyops) reach the Rupununi Savanna, one leptodactylid frog (Pseudopahidicola pusilla) just reaches the Sipaliwini Savanna in Surinam (Fig. 10:12), and one tree frog (Hyla rostrata) so far has only been found in the vicinity of El Dorado (Venezuela) and possibly near Cayenne (French Guiana). All of these species have been dealt with in other sections. Species From Southeastern or Central Brasil Reaching Guiana Most of the species from southeastern or central Brasil reaching Guiana do not extend farther west than French Guiana or Surinam (Fig. 10:13); only seven [Hyla x-signata (tree frog), Leptodactyhis fuscus (leptodactylid frog), Pseudis paradoxus (pseudid frog), Phrynops geoffroanus (chelid turtle), Liophis miliaris (colubrid snake), Coleodactylus mer- idionalis (gekkonid lizard), and Tropidurus torquatus (iguanid lizard)] reach Venezuela. The majority, if not all, of these species are inhabitants of savannas or open swamps, and their distributions are closely associated with those habitats. Apparently these species are recent immigrants from the southeast that either used the savanna corridor (central and northeastern Brasil to southeastern Vene- zuela) during the last arid phase (about 2000-3500 years ago), when the greater part 1979 HOOGMOED: HERPETOFAUNA OF GUI AN AN REGION 261 of this area was covered with a cerradohke Table 10:3. — Comparison of Rainforest Frog Faunas vegetation, or they used the open swampy of Different Regions in Northeastern South America. coastal area of Amapa. Most of these species Species in Common have not differentiated and when they have, Brasilian the subspecies occurring in the Guianas is „ „ „ Western Eastern part •j i- l -ii ^ ■ j t> -i t ft? Guiana Guiana Guiana Belem identical with the one in northeastern Brasil. ... . ^—. =s — — 7-, ^ — — ■=-. — „, , , . ., ,, 7 , Western Guiana ...... 76 41 31 14 I he exceptions are the bufomd Melanophnj- Eastern Guiana .. 0.51 83 42 22 niscus moreirae, and the pseudid frogs, Pseu- Brasilian part Guiana 0.52 0.67 43 19 dis paradoxus and Lysapsus limellus, all of Belem °-28 °-40 0-58 23 which have endemic subspecies in Guiana. The last two species may have reached Gui- T,A*L* 10:4.-Comparison of Savanna Frog Faunas j . i. j i i ill oi Different Regions in Northeastern South America. ana during an earlier dry phase and probably ^=^^=^=^=^=^=^^=^=:^^= along a different route (from Rio Tapajos Species in Common via Rio Negro and Rio Branco to the north). Western Eastem Br^ Furthermore, PseudlS paradoxus reaches the FRF Guiana Guiana Guiana Belem western part of Guiana, whereas in Guiana Western Guiana . 3(5 23 20 13 Lysapsus limellus is only in the western part Eastern Guiana 0.69 31 19 13 / pjrr 10 13) Brasilian part Guiana 0.69 0.72 22 12 K &' ' '' Belem 0.51 0.46 0.65 15 ANALYSIS OF GEOGRAPHIC DISTRIBUTION For three groups (frogs, lizards and snakes) data are sufficient to permit an at- tempt of comparison with localities outside the Guianan Region. However, data were scarce and comparisons for frogs only could be made with the Belem region (Crump, 1971), for lizards with Belem (Crump, 1971; Da Cunha, 1961) and Iquitos (Dixon and Soini, 1975), and for snakes only with Iquitos (Dixon and Soini, 1977 ) . All data have been compiled in Tables 10:3-8. In these tables the total num- ber of species in each locality is on the di- agonal from upper left to lower right. The number of species common to each combina- tion of regions is to the right and above the diagonal with the totals. To the left and be- low the diagonal are the Faunal Resemblance Factors (FRF) as computed for each com- bination of regions, using the formula ( Duell- man, 1965, 1966): Table 10:5. — Comparison of Rainforest Snake Faunas of Different Regions in Northern South America. Species in Common Western FRF Guiana Western Guiana 80 Eastern Guiana 0.78 Brasilian part Guiana 0.75 Iquitos 0.65 Brasilian Eastern part Guiana Guiana Iquitos 65 53 54 85 54 53 0.74 60 46 0.62 0.63 84 Table 10:6. — Comparison of Open Formation Snake Faunas of Different Regions in Northern South America. Species in Common Brasilian Western Eastern part FRF Guiana Guiana Guiana Iquitos Western Guiana ... 19 14 11 4 Eastern Guiana .. 0.75 18 12 5 Brasilian part Guiana 0.68 0.77 13 4 Iquitos __ 0.33 0.43 0.44 5 FRF 2C Ni + N2 where Ni and N2 are the numbers of species occurring in any two given regions and C is the number of species common to the two regions compared. The computations were made both for forest- and savanna-inhabiting species. The data for frogs (Tables 10:3-4) show that among forest inhabitants there is a dis- tinctly higher resemblance between the anu- ran faunas of eastern Guiana (Guyana east of the Essequibo River, Surinam, French Guiana, and northern part of Amapa) and of the Brasilian part of Guiana (Guiana Re- gion south of divide) than between both of 262 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. Table 10:7. — Comparison of Rainforest Lizard Faunas of Several Regions in Northern South America. Species in Common FRF o ■___ ■fi 4> G C a 'a C 3 O _. _ o as __ c_ o __ S 4-1 a p. c _ c_ -_. ca s 3 O 6 0) c_ a -3 6 g o a £ p. o _J < _H ca CQ m O _j Paramaribo 20 17 13 11 16 18 17 20 20 20 Cayenne 0.77 24 14 10 15 18 19 22 24 22 Lely Mountains .. . 0.62 0.61 22 14 16 18 16 21 22 22 Alto Marara (ISfi 0.47 0.68 19 14 13 14 19 19 18 Iquitos ..... 0.51 0.45 0.49 0.45 43 19 23 25 25 24 Rolivar 0.69 0.64 0.67 0.51 0 51 32 18 24 27 32 25 Relem . 0.68 0.70 0.62 0.57 0.63 0.55 30 28 28 Rrasilian part Guiana .. 0.66 0.68 0.67 0.64 0.60 0.66 0 80 40 34 35 Eastern Guiana . 0.59 0.73 0.69 0.63 0.59 0.73 0.78 0.83 41 37 Western Guiana _____ 0.59 0.61 0.63 0.54 0.53 0.81 0.64 0.80 0.84 47 Table 10:8. — Comparison of Savanna Lizard Faunas of Seve •ral Rej Sons in Northern South America. Species in Common FRF o ___> •a v. G '3 a 3 aj o « __ 4-> p. c C •a o a c ca ■3 O e c__ 1 _- c c a. o o __ •a _ a _, c __ _ s c_ O ►J < _? CQ 0J ___ 3S w tl_ ^ Paramaribo .. 3 0 0 0 3 2 3 3 3 1.0 3 0 0 0 3 2 3 3 3 T.ely Mountains 0 0 0 0 0 0 0 0 0 0 Alto Marara 0 0 0 0 0 0 0 0 0 0 Iquitos 0 0 0 0 0 0 0 0 0 0 Rolivar 0.35 0.35 0 0 0 14 2 4 6 14 Relem _ 0.80 0.85 0.80 0.85 0 0 0 0 0 0 0.25 0.44 2 0.66 2 4 2 4 2 Rrasilian part Guiana 4 Eastern Cniana 0.66 0.66 0 0 0 0.60 0.50 0.80 6 6 Western Guiana 0.35 0.35 0 0 0 1.0 0.25 0.44 0.60 14 those areas and western Guiana (Guyana west of Essequibo River, Venezuelan Guay- ana). This could be explained by the barrier function of the Essequibo-Rio Bianco De- pression. On the other hand, the resemblance between the forest anuran fauna of Belem as compared to the other regions, shows a steady decrease towards the west. We see quite a different picture when comparing the savanna anuran faunas. Here the resemblance between eastern Guiana and the Brasilian part of Guiana is only negligibly higher than that between both of those areas and western Guiana. The resemblance between Belem and the Brasilian part of Guiana is in the same category as that between the three areas of the Guiana Region among themselves. However, the resemblance between Belem and both eastern and western Guiana is dis- tinctly lower. Although 13 of the 15 species known from Belem occur throughout western and eastern Guiana, the resemblance factor is low because these species only are a fraction of the much larger savanna anuran fauna there, which contains a fairly high proportion of local endemics (which still may prove to be more widespread) and a number of species reaching their eastern distribution limits in Guiana. Thus, it can be concluded that for savanna-inhabiting frogs there are no barriers within Guiana and, for that matter, scarcely any barriers towards areas surrounding Gui- 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 263 ana. For forest-inhabiting frogs there is a distinct barrier within Guiana, formed by the Essequibo-Rio Bianco Depression and also, the forest anuran fauna is distinctly separated from that to the southeast. However, these conclusions are based only on data from four areas (three of which chosen with a certain bias) and therefore should be treated with much reserve, although they do confirm the picture that emerged from a first study of distribution maps. These findings for the anuran faunas can be most easily explained by assuming that the Essequibo-Rio Branco Depression not only served as a connection (and dispersal route for aquatic species) between Guiana and Amazonia, but also was the area that retained its savanna vegetation longest, as still indicated by the presence of large savanna areas in the border region of Guyana and Brasil and in the coastal area near the mouth of the Essequibo River. Thus, this area formed an efficient barrier to the dispersal of forest frogs; at the same time, it formed a dispersal route for savanna frogs. This situation apparently lasted until fairly recently, until under the influence of an in- creasingly wet climate the forests in the Guiana and Imeri refuges started to expand and met in the Essequibo-Rio Branco De- pression. This explains why many forest spe- cies have their eastern or western distribution limits at the Essequibo-Rio Branco Depres- sion. Several species that apparently were associated with one of the refuges in the re- gion succeeded in crossing the depression, but this could have taken place only recently when the savanna vegetation was substituted by forest. Comparison of the snake and lizard faunas of several localities gives quite a different picture. The data for forest-inhabiting snakes (Table 10:5) show that there is a great re- semblance between different parts of the Guiana Region and that the resemblance with the forest snake fauna of Iquitos is fairly good, but distinctly lower than within Guiana. The data suggest a gradual transition within Guiana from west to east and also from Iqui- tos to Guiana, but owing to lack of data from intermediate localities this last hypothesis cannot be proved. Only five snakes inhabiting open formations are found in the Iquitos area; all of these are either associated with open aquatic or edge situations. Real savanna spe- cies are absent, because no suitable habitat is available in the region (Dixon and Soini, 1977). When comparing these snakes (Table 10:6) with the open formation species of Guiana, it is clear that the resemblance be- tween Iquitos and the three parts of Guiana is small. Within Guiana there is a lower degree of resemblance between the snake faunas of western Guiana and the Brasilian part of Guiana, but this is caused by the presence in western Guiana of several species reaching their eastern distribution limits there and in the Brasilian part of Guiana of species reach- ing their northern distribution limits there, and of species that are known only from the Amazon Basin. However, the data for snakes again are based only on four areas. The lizards and amphisbaenians ("sauri- ans") appeared to offer the best possibilities for a faunal analysis, because there were sev- eral places from which representative samples seemed to be present (Tables 10:7-8). How- ever, upon closer examination, it soon turned out that the data were not very reliable. This holds true for the forest lizards and amphis- baenians of Lely Mountains, Paramaribo, and Alto Maraca (Amapa). When compared with the entire region of which they are part, re- spectively eastern Guiana (twice) and the Brasilian part of Guiana, they show resem- blance factors of only 0.69, 0.59 and 0.64, re- spectively. These are hardly more, or even lower, than their respective resemblance fac- tors with the lizard fauna of Belem (0.62, 0.68 and 0.57). For Cayenne the situation seems to be better; when compared with east- ern Guiana it shows a resemblance factor of 0.73, but here we should keep in mind that the total number of species reported from this locality also contains old records that pos- sibly refer to specimens that were shipped from Cayenne but actually did not occur there. From the remaining data it is clear that there is a diminishing resemblance westward between the rainforest lizard faunas of Belem and Guianan areas. Within Guiana the re- semblance is high, and nowhere is a clear break apparent. The resemblance factor be- 264 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 tween Belem and Iquitos is slightly higher than that between Iquitos and the different parts of Guiana. Resemblance between Iqui- tos and western Guiana is smaller than that with the Brasilian part of Guiana. This ap- parent reversal of expected resemblances is the result of the influence of Amazonian spe- cies in the Brasilian part of Guiana and in eastern Guiana. Apparently there is a fairly well-developed barrier southwest of Guiana, separating the lizard faunas of Guiana and upper Amazonia. Again, our knowledge of areas intermediate between Iquitos and Gui- ana is poor, and the conclusions must be regarded as preliminary. When comparing the savanna lizard faunas of different regions (Table 10:8) we get a very different picture. There is no resemblance with Iquitos, where this category of lizards is completely absent. The agreement between Belem and the dif- ferent Guianan areas diminishes westward, and there seems to be a break between west- ern Guiana on the one hand and eastern Guiana and the Brasilian part of Guiana on the other. Upon closer examination, this ap- parent break is caused completely by the presence in western Guiana of a number of altitudinal endemics and of western species just reaching their eastern limits in Guiana. When these species (forming 605? of the savanna lizard fauna) are excluded, there are no breaks for the remaining general savanna lizards within Guiana, neither with Belem. The only break is between Iquitos and Gui- ana, and this can be completely explained by the absence of suitable savanna habitat in upper Amazonia. As for savanna-inhabiting frogs, the Essequibo-Rio Branco Depression formed no barrier to that part of the savanna inhabitants that had been in the area rela- tively long. For a number of local savanna endemics it seems to act as such, mainly be- cause those endemics did not have the chance to expand their area of distribution. On the basis of the data presented in Tables 10:. 3-8 it can be concluded that for forest inhabitants Guiana seems to be a real herpetogeographic entity, well separated from surrounding areas to the southwest and the southeast. Within the area, the Essequibo- Rio Branco Depression forms a barrier for the distribution of a number of eastern forest species to the west and of western forest spe- cies to the east. No such function is present for savanna inhabitants that, with the excep- tion of local endemics, are spread throughout the area. There is no separation to the south- east for savanna-inhabiting species, which consequently show a great resemblance with the savanna fauna of northeastern Brasil. The data do not present any evidence for the recognition of a Guiana Region as de- fined by Lescure (1977); distinct breaks be- tween eastern Guiana and the Brasilian part of Guiana are nowhere evident. CONCLUSIONS The herpetofauna of Guiana, as it is known at present, is a composite fauna with a complex history. A number of endemic spe- cies belong to old genera (endemic or with disjunct, relict distributions) that apparently inhabited certain parts of the area since the Cretaceous. Other endemics probably origi- nated in the region during periods of isola- tion in forest or savanna refuges, which are assumed to have existed during arid and wet phases in the Pleistocene-Holocene, respec- tively. The most important forest refuge was the Guiana Refuge on the northern slopes of the Tumuc-Humac and Acarai mountains. A less important role was played by the Imeri Refuge in the region of Serra Imeri and Serra da Neblina. The species restricted to higher altitudes survived the arid phases in disjunct forests on the higher slopes of the tepuis, col- lectively known as Tepui Refuges. During arid phases of the Pleistocene and Holocene, the species isolated in the refuges underwent differentiation and, depending on the time of their arrival in the area and also on their rates of evolution, they differentiated into endemic genera, species, or subspecies. Although some species show relationships to Andean species, these are not direct and only indicate that both Andean and Guianan species evolved from the same or related low- land species. The Guianan species of Atrac- tus (colubrid snakes), Eleutherodactylus ( leptodactylid frogs) and Centrolenella 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 265 (glass frogs) are in this category. The pres- ence of two species of the Andean teiid lizard genus Euspondylus in the Guiana Highlands is the only evidence for a direct link with the Andes. As these species are poorly known from only a few individuals each, their taxo- nomic position remains uncertain. Therefore, it would be premature to conclude on the basis of this meager evidence that there would have been any invasions of Guiana from the Andes. A number of species invaded Guiana from the south via a wide belt of cerradolike vege- tation, connecting central and northeastern Rrasil with southeastern Venezuela, during the last arid phase. When the climate became more humid and the forests expanded, these species were left stranded on the isolated savannas of Guiana, most of them in the east. During this same period a number of savanna- inhabitants from northwestern South America invaded the western part of Guiana, where they exist today as representatives of the llanos fauna. Within Guiana there are dif- ferences between the western part, where sandstone tepuis are present, and the eastern part, which generally has a much lower ele- vation. A number of species (most of them endemic) are restricted to the sandstone re- gion; others (mostly invaders from southern and central Brasil or from the Amazon Val- ley) occur only in the east. Apparently Gui- ana has been, and still is being, invaded from the northwest and from the southeast. Within Guiana the Essequibo-Rio Bianco Depression seems to have acted as a barrier to the eastern dispersal of western elements and to a lesser extent for the dispersal of eastern ele- ments to the west. The notable exceptions are some of the species from central and southeastern Brasil. Also this depression ( and the low coastal area of Amapa) served as corridors into northern Guiana for a number of Amazonian species. Endemism in the entire region is high in amphibians (52%) but much lower in reptiles (27%). At elevations above 1000 m, only frogs, lizards and snakes occur; endemism for frogs there is 33 percent, for reptiles 24 per- cent. Endemism for amphibians below 1000 m is 47 percent (frogs only 40%, caecilians only 69% ) , for reptiles 24 percent. From these data it is clear that although frogs have a high de- gree of endemism at higher elevations, the amount of endemism in the lowlands is even higher. However, part of this probably results from our still scanty knowledge of this group; of the 83 endemic frogs, 29 (35%) have yet to be named. ACKNOWLEDGMENTS Fieldwork in Surinam and Venezuela was supported by grants W956-2, W87-78 and WR87-131 from the Netherlands Foundation for the Advancement of Tropical Research (WOTRO), and by grants from the Royal Dutch Academy of Sciences ( Melchior Treub Foundation) and the Treub Society. The photographs were made by Mr. E. L. M. van Esch of the Rijksmuseum van Natuurlijke His- toric Leiden, from color slides taken by me. The drawings were made by Mr. J. J. A. M. Wessendorp, also of the Rijksmuseum van Natuurlijke Historic The Spanish text was checked by Dr. F. Carrasquer of the Depart- ment of Spanish Language of the University of Leiden. Any mistakes are completely my responsibility. RESUMEN Limitado por el Orinoco, Brazo Cassiqui- are, Rio Negro, el Amazonas y el Atlantico, la Guayana es geologicamente uno de los territorios mas antiguos de America del Sur. Su mayor parte viene formado por el escudo guayani precambrico, cuya arenisca de Ro- raima cubre partes del sur de Venezuela y de la Guyana occidental, con restos aislados en la Guyana oriental y en el Surinam central. La altura media no pasa de los 1000 m, pero los restos areniscos pueden alcanzar hasta 3000 m. Debido a su position aislada y a su elevation relativamente considerable, estas montafias (tepuyes) representan ser como islas subtropicales en un mar o llanura tropi- cal. La exploration biologica de esta alti- planicie guayani se emprendio a mediados del siglo pasado y se prosigue aiin hoy, sin haber explorado mas que una pequefia parte 266 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 de los tepuyes y de una manera suficiente tan solo su aspecto herpetologico. Dado este es- tado de cosas, por fuerza hemos de limitarnos a interpretar los datos obtenidos de la herpe- tofauna de la altiplanicie de la Guayana. Actualmente se conocen en la Guayana 178 especies de anfibios y 230 reptiles, totali- zando asi las 408 especies heipetologicas. No se conoce las salamandras. En dicho numero se comprenden: cinco tortugas marinas, un chacon (Gekkonidae) cosmopolita, cinco especies importadas de las Antillas y tres cuya presencia en la Guayana es de origen dudoso. El residuo puede clasificarse en cinco grandes grupos — endemicos (38%), amazonicos (37%), de vasta extension (12%), llegando el limite oriental de su extension hasta el escudo gua- yani (7%) y a la Guayana del Brasil central y sudeste (5%). Si se consideran estos datos aisladamente, referidos a los reptiles y a los anfibios por separado, constatamos diferencias importantes. De los anfibios, el 52 porciento son endemicos, el 30 porciento son amazonicos y el 7 porciento de vasta extension; mientras que para los reptiles tenemos los siguientes porcentajes — 27, 42 y 16, respectivamente. Estas diferencias se explican por el hecho de que los anfibios necesitan para su reproduc- tion agua y esta dependencia los hace mas limitados que los reptiles en su capacidad de dispersion. Noventa y dos especies de anfibios y 59 de reptiles son endemicos de la Guayana, con una pequeiia portion fuera de los limites de la region que estamos describiendo. Las espe- cies de los otros grupos pueden formar sub- especies endemicas en este territorio (39 sub- especies pertenecen a 29 especies ) . La mayor parte de las especies endemicas se concentran en la parte occidental de la Guayana y pueden clasificarse en endemicos de llanura y de al- tura, segiin se hallen por debajo o por encima de los 1000 m de altitud. Aproximadamente el 19 porciento de los anfibios endemicos y el 15 porciento de los reptiles endemicos son de altura, ateniendonos solo a la parte occi- dental. Generos endemicos tales como Oto- phryne y Oreophrynella parecen representar reliquias del joven Terciario, Stefania parece representar una radiation reciente y la posi- tion de Riolama no queda muy clara. Los generos endemicos residuales (AUophnjne, Rliinatrema, Pekocephalus, Amapasaurus, Mesobaena) son endemicos de llanura y su historial esta probablemente asociado con el de los refugios de Guayana y de Imeri. Una especie endemica de Euparkerella apunta tener alguna relation con las montanas del sudeste brasileno. La "Hylodes" duidensis que hasta hace poco se creia estaba emparen- tada con formas del sudeste brasileno, resulta ser bastante diferente y mas bien representa una derivation de los eleutherodactylini de llanura. La mayor parte de los endemicos de altura son derivaciones subtropicales de pari- entes de la llanura tropical. Los endemicos de altura tienen una extension limitada a uno o varios tepuyes. El origen de la mayor parte de los endemicos de llanura se explica prob- ablemente por la formation en el pasado de refugios forestales a traves de los cambios climaticos del periodo cuatemario. De esos supuestos refugios en la region es el de la Guayana el mas importante, siendo de menor importancia el de Imeri. Estos refugios, sep- arados por la sabana, han procurado en su dia una especificacion alopatrica en todo un territorio donde hasta hace poco se creia sin barreras ecologicas de importancia. Las especies amazonicas las tratan otros autores en otros articulos. En todo caso se dividen en cuatro subgrupos. Algunas de estas especies llegan hasta la parte septen- trional de la Guayana por la depresion del Essequibo-Rio Branco, o siguiendo las forma- ciones abiertas del Amapa costero y de la Guayana francesa septentrional. La mayor parte de las especies de vasta extension son de origen sudamericano, y una especie de origen centroamericano refuerza la hipotesis de los refugios forestales. Las especies que llegan al limite oriental del escudo guayani son de origen mixto y vienen tratadas en otros capitulos. Las especies del Brasil central y sudeste que llegan a la Guayana estan por lo general presente en la parte oriental y muchas de ellas asociadas a las vegetaciones abiertas. Su ex- tension es probablemente correlativa a la extension de las vegetaciones abiertas del ul- 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 267 timo periodo arido, como lo prueba su exis- tencia en las sabanas actualmente aisladas de la Amazonia. No parece que haya en la Guayana bar- reras geograficas de importancia, aunque la depresion del Essequibo-Rio Bianco ha de- bido de hacer de barrera a la extension de ciertas especies de llanura y a no pocas formas de derivation de las de altura. La extension caracteristica en formas de llanura se debe probablemente, en la mayoria de los casos a la competition ecologia u a otras condiciones particulares del medio ambiente. LITERATURE CITED Bisschops, J. H. 1969. The Roraima Formation in Surinam. Verh. K. Ned. Geol. Mijnbouwkd. Genoot. Geol. Ser. 27:109-117. Boulencer, G. A. 1895. Description of a new ba- trachian ( Oreophryne Quelchii ) discovered by Messrs. J. J. Quelch and F. McConnell on the summit of Mt. Roraima. Ann. Mag. Nat. Hist. (6)15:521-522. Boulencer, G. A. 1900. Reptiles: 53-54, Batraehi- ans:55-56, in Lankster, E. R., 1900. Report on a collection made by Messrs. F. V. McConnell and J. J. Quelch at Mount Roraima in British Guiana. Trans. Linn. Soc. London Zool. (2)8(2): 51-76. Brown, K. S., Sheppard, P. M., Turner, J. R. G. 1974. Quaternary refugia in tropical America: Evidence from race formation in Heliconius but- terflies. Proc. Roy. Soc. London, B, 187:369-378. Chapman, F. M. 1931. The upper zonal bird-life of Mts. Roraima and Duida. Bull. Amer. Mus. Nat. Hist. 63:1-135. Crump, M. L. 1971. Quantitative analysis of the ecological distribution of a tropical herpetofauna. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (3): 1-62. Cunha, O. R. da. 1961. II. Lacertilios da Ama- zonia. Os lagartos da Amazonia brasileira, com especial referenda aos representados na colecao do Museo Goeldi. Bol. Mus. Paraense Emilio Goeldi Nova Ser. Zool. (39): 1-189. Descamps, M., Gasc, J. P., Lescure, J., Sastre, C. 1978. Etude des ecosystemes quyanais. II. Don- nees biogeographiques sur la partie orientale des Guyanes. C. R. Seances Soc. Biogeogr. 467: 55-82. Dixon, J. R„ Soini, P. 1975. The reptiles of the upper Amazon Basin, Iquitos Region, Peru. I. Lizards and amphisbaenians. Contrib. Biol. Geol. Milwaukee Publ. Mus. (4):l-58. Dkon, J. R., Soini, P. 1977. The reptiles of the upper Amazon Basin, Iquitos Region, Peru. II. Crocodilians, Turtles and snakes. Ibid. (12): 1-91. Duellman, W. E. 1965. A biogeographic account of the herpetofauna of Michoacan, Mexico. Univ. Kansas Mus. Nat. Hist. Misc. Publ. 15:627-709. Duellman, W. E. 1966. The Central American herpetofauna: An ecological perspective. Copeia 1966(4):700-719. Fittkau, E. J. 1974. Zur okologischen Gliederung Amazoniens. I. Die erdgeschichtliche Entwick- lung Amazoniens. Amazoniana 5:77-134. Gansser, A. 1954. The Guiana Shield (South Amer- ica). Eclogae Geol. Helv. 47:77-117. Gleason, H. A. 1931. Botanical results of the Tyler- Duida Expedition. Bull. Torrey Bot. Club 58: 277-506. Haas, F. 1957. Zur Tiergeographie von Amazonien und dem Guayana Schild. Mitt. Naturforsch. Ges. Bern 14:59-64. Haffer, J. 1969. Speciation in Amazonian forest birds. Science 165:131-137. Haffer, J. 1974. Avian speciation in tropical South America. With a systematic survey of the tou- cans ( Rhamphastidae ) and jacamars ( Galbuli- dae). Publ. Nuttal Ornithol. Club, Cambridge (14): 1-390. Hills, T. L. 1969. The savanna landscapes of the Amazon Basin. Savanna Res. Ser. Dep. Geogr. McGill Univ. Montreal, (14): 1-38. Hoocmoed, M. S. 1973. Notes on the herpetofauna of Surinam IV. The lizards and amphisbaenians of Surinam. Biogeographica 4:1-419. Lancini, A. R. 1968. El genero Euspondijhts (Sau- ria, Teiidae) en Venezuela. Publ. Occas. Mus. Cienc. Nat. Zool. (12): 1-8. Lescure, J. 1975. Biogeographie et ecologie des Amphibiens de Guyane Francaise. C. R. Seances Soc. Biogeogr. 440:68-82. Lescure, J. 1977. Diversite des origines biogeo- graphiques chez les Amphibiens de la region guyanaise. Publ. Lab. Zool. ficole Normale Super. 9:53-65. Maguire, B. 1945. Notes on the geology and geog- raphy of Tafelberg, Suriname. Geogr. Rev. 35: 563-579. Maguire, B. 1955. Cerro de la Neblina, Amazonas, Venezuela. A newly discovered sandstone moun- tain. Ibid. 45:27-51. Macuire, B. 1970. On the flora of the Guayana highland. Biotropica 2:85-100. Mayr, E., Phelps, W. H., Jr. 1967. The origin of the bird fauna of the South Venezuelan High- lands. Bull. Amer. Mus. Nat. Hist. 136:269-328. McDiarmid, R. W. 1971. Comparative morphology and evolution of frogs of the Neotropical genera Atelopus, Dendrophryniscus, Melanophryniscus, and OreophryneUa. Nat. Hist. Mus. Los Angeles Cty. Sci. Bull. (12): 1-66. Muller, P. 1973. The dispersal centres of terrestrial vertebrates in the Neotropical realm. A study in the evolution of the Neotropical biota and its native landscapes. Biogeographica 2:1-244. Nott, D. 1975. Into the lost world. A descent into prehistoric time. Prentice-Hall, Englewood Cliffs, N.J., 186 p. 268 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Oldenburger, F. H. F., Norde, R., Riezebos, H. T. 1973. Ecological investigations on the vegetation of the Sipaliwini-savanna area ( Southern Suri- nam): 1-51. (Mimeographed) Orejas-Miranda, B., Quesada, A. 1976. Ecosis- temas fragiles. Ciencia Interamericana 17:9-15. Paba-Silva, F., van der Hammen, T. 1960. Sobre la geologia de la parte sur de la Macarena. Bol. Geol. Bogota 6:7-30. Prance, G. T. 1973. Phytogeographic support for the theory of Pleistocene forest refuges in the Amazon Basin, based on evidence from distribu- tion patterns in Caryocaraceae, Chrysobalanaceae, Dichapetalaceae and Lecythidaceae. Acta Ama- zonica 3:5-28. Priem, H. N. A., Boelrijk, N., Hebeda, E., Ver- durmen, E., Verschure, R. 1973. Age of the Precambrian Roraima Formation in northeastern South America: Evidence from isotopic dating of Roraima pyroclastic rocks in Surinam. Geol. Soc. Amer. Bull. 84:1677-1684. Reinke, R. 1962. Das Klima Amazoniens. PhD Dissert. Univ. Tubingen, 101 p. Rivero, J. A. 1961. Salientia of Venezuela. Bull. Mus. Comp. Zool. Harvard Univ. 126:1-207. Rivero, J. A. 1965 (1964). The distribution of Venezuelan frogs. V. The Venezuelan Guayana. Caribb. J. Sci. 4:411-420. Rivero, J. A. 1966. Notes on the genus Cnjpto- batrachus (Amphibia, Salientia) with the de- scription of a new race and four new species of a new genus of hylid frogs. Ibid. 6:137-149. Rivero, J. A. 1967a. Anfibios colleccionados por la expedicion Franco- Venezolana al Alto Orinoco 1951-1952. Ibid. 7:145-154. Rivero, J. A. 1967b. A new race of Otophryne ro- busta Boulunger (Amphibia, Salientia) from the Chimanta-Tepui of Venezuela. Ibid. 7:155-158. Rivero, J. A. 1968a. A new species of Elosia (Am- phibia, Salientia) from Mt. Duida, Venezuela. Amer. Mus. Novit. (2334): 1-9. Rivero, J. A. 1968b. Los centrolenidos de Venezuela (Amphibia, Salientia). Mem. Soc. Cienc. Nat. La Salle 28:301-334. Rivero, J. A. 1968c. A new species of Eleuthero- dactylus (Amphibia, Salientia) from the Guyana region, Edo. Bolivar, Venezuela. Breviora (306): 1-11. Rivero, J. A. 1968d. A new species of Htjla (Am- phibia, Salientia) from the Venezuelan Guyana. Ibid. (307): 1-5. Rivero, J. A. 1970. On the origin, endemism and distribution of the genus Stefania Rivero (Am- phibia, Salientia) with a description of a new species from southeastern Venezuela. Bol. Soc. Venezolana Cienc. Nat. 28:456-481. Rivero, J. A. 1971. Notas sobre los anfibios de Venezuela. I. Sobre los hilidos de la guayana venezolana. Caribb. J. Sci. 11:181-193. Romariz, D. A. 1974. Aspectos da vegetacao do Brasil. Instituto Brasileiro de Geografia e Esta- tistica, Diretoria tecnica, Rio de Janeiro, 126 p. Roze, J. A. 1958a. Resultados zoologicos de la expedicion de la Universidad Central de Vene- zuela a la region del Auyantepui en la Guayana venezolana, Abril de 1956. 5. Los reptiles del Auyantepui, Venezuela, basandose en las colec- ciones de las expediciones de Phelps-Tate, del American Museum of Natural History, 1937-1938, y de la Universidad Central de Venezuela, 1956. Acta Biol. Venez. 2:243-270. Roze, J. A. 1958b. Los reptiles del Chimanta tepui (Estado Bolivar, Venezuela) colectados por la expedicion botanica del Chicago Natural History Museum. Ibid. 2:299-314. Silverstone, P. A. 1975. A revision of the poison- arrow frogs of the genus Dcndrobatcs Wagler. Nat. Hist. Mus. Los Angeles Cty. Sci. Bull. 21: 1-55. Tate, G. H. H. 1928. The "Lost World" of Mount Roraima. The account of an expedition to a strange and little known flat-topped mountain in the heart of the South American jungle. Nat. Hist. 28:318-328. Tate, G. H. H. 1930a. Notes on the Mount Roraima region. Geogr. Rev. 20:53-68. Tate, G. H. H. 1930b. Through Brazil to the sum- mit of Mount Roraima. Natl. Geogr. Mag. 58: 584-605. Tate, G. H. H. 1932. Life zones at Mount Roraima. Ecology 13:235-257. Tate, G. H. H. 1938a. Auyantepui. Notes on the Phelps Venezuelan expedition. Geogr. Rev. 28: 452^174. Tate, G. H. H. 1938b. A new "Lost World." Nat. Hist. 42:107-120, 153. Tate, G. H. H. 1939. The mammals of the Guiana Region. Bull. Amer. Mus. Nat. Hist. 76:151-229. Tate, G. H. H., Hitchcock, C. B. 1930. The Cerro Duida Region of Venezuela. Geogr. Rev. 20: 31-52. Todd, W. E. C, Cahriker, M. A., Jr. 1922. The birds of the Santa Marta region of Colombia: A study in altitudinal distribution. Ann. Carnegie Mus. 14:3-611. van der Hammen, T. 1974. The Pleistocene changes of vegetation and climate in tropical South America. J. Biogeogr. 1:3-26. Vanzolini, P. E. 1970a. Zoologia systematica, geo- grafia e a origem das especies. Univ. Sao Paulo, Inst. Geografia Tese Monogr. 3:1-56. Vanzolini, P. E. 1970b. Unisexual Cnemidophorus lemniscatus in the Amazonas valley: a prelim- inary note (Sauria, Teiidae). Pap. Avul. Zool. (Sao Paulo) 23:63-68. Warren, A. N. 1973. Roraima. Report of the 1971 British expedition to Mount Roraima in Guyana, South America. Private publication, Oxford, 152 p. 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 269 | E 8 - £ tS ■» II cl sji _ 5 bg § -3a 8" II 1 1 II So «; . c •43 ft c ^ g °is s ■a a c H > _o « ■ -J e *1 !.|d Iq s 3 ii II m CJCO C 4j 3 a Q'g § z«s Oh - C Sill -C ft ft 5" - a 5W ■fl 3 . o ~ a « §*J 5 £ c II (3 g ca C o & § in to ft o is a a s In •3-3 ^ CO - O >* *°^ O _-T3 • -< 3 ni 3 (U O 3 .2 5 ^ I « . CO. S.S5 o .a 2 > oXS 3 tS O £ c ^ a 2'K o Bu a- w tS || P. K < a „_, o o a "■'■ B t-1 v. ceo < pa N C-3 2 .2 £ C CPL, N C 'S rt ro cs o o N CO 3 S P ■^ S § §.2 E "3 a. 1. ~C H 3 "K c r N Ssg» 1*8 | d at 3 vi -"- -i.CGajn)3(uajajaQj!DrtrtftaicD P5P5 :cic!>4.| dooo2: Scici05 HIOHO -— i 5 Co &o W5 io Co c -S--5-Fo "o "o "o ^ a a .2! cj- « a s i ^ « « ^^ jx 2. i ^ *3 ~'^: Ci) Qj ^j ^; ^j i o o CJ 2-° i : i : i ! : ! I | 1 ! : i i ! | 20 0 2* ■E *9 « 2 -C 3 ! en o ^r ^c: s <3 ~ 5 § 2 2 ol '. O 9*3* ?'S3 £ V, 3 :-2 = 2P& 5-5-H.a 3 "5 s 'a - >-a "c"c "c "«"Q"C "C ~5 ~ "X *- q o "X"t^"?i"X"^:'^i,^t ^"^l"X ft a a a -o -a -a o o o "S "S "2 "S "S a == a K OO Q Q QQ QS ^ c ^ ^ ^ a o ~ oooooooo "ct: sj cj ^. ^tiooooooio o ooookmoo oconoiooim^ coo o 2 *- w -i- p^ <£» ^J "2; •*- 3 O a. a. ~ iC tj .^ 1) 5i< ?2 ^^^Siii^^ ^ r^l rn^Ti «^> *^> *ii — .1 ~J ■£] •£! (31 •-) ~) *-' C>3 «* p- ni r**r**f"*r-*r'-*r'-,,-*i—*r*rh*t1-*s-H T"* w a ^ o a a <3 a a a a goo "a "a "a "a ts "o "c a a a a a a a w 3) Q a a a a. a. a a. £?3 ^ ^ ^ ^ ^ ^ ^ kj O >J hj kJ kJ ^] a a a a a •B-CBt! o o o o o 5 § S g "g y-ft : j? = ~ . a. a 2 2 O. a = — j- = -Sa>c cjo a"a o !C 4) Co CO a a a a o -^ a a a a a a a a a~a ^ a^ .JcJ kj Kj Cj - s-3 S ■-J a, 'S S S S a. a ~ ^ a 5 ° 5 a: c* a 60 G£) o a < « ■~ Si o s; s :=? 2 s "£"<:?- ** n °" « co ™ a • L) a S o C-_~ u K u "I'S 'S -2 -S -2 £ a» ca "^ cj ^ *~ a a a ^ "a £"' ? ? 5> a a t;<:-i;- i o c ted ■3 ftO 0) 3 6 g ii =3 ^ '3 W J< j<<::?< © ~* © ©el OO OlO h o CD © O © tn .ot-oiiooo-r-^o — 00-H =>or-,-i-"i ^H 0 © © 0 o o o < 00 5SS ~r 'Sj S 8 ^j \u O O <-=> -a -a -a -o 53 _53 ja ja a a % s ■*->'-*-■«■!- ^ CY, 1- . __ 5, ,53 ^ 60 &fi M M Csc Cu c a .g « ^ o cj S 2 '3 S ~ a c ^ a -S : X : 00 : : 00 IX : en ! I ! pS IX I 00 CO 00 ! co 00 : i : 00 icoSoopS ; co : : ico : : ico >S :b= iw iS ,W 00 00 ; >5 i 00 2S co co ™ co PS PS co co : 00 00 PS 1 X 00 1 : pS 00 OO O CD lOOCt 00 O O O Of-clOoOOOOTOT.-sr— n. CO 1C CO © c? -T S co °,' OOOOOOOOOOO © Ol O •-! O CN CM 05 s _2 a> o a. 3 C3 Cy ~ ~ 1, ^> ~ ■2 *> « :^ 5 5 5 = O ^ C "C3 ^ ^ C ~ a o 9 a ^j cs ° ? ? r 2 < n b Q g): - - : ? d, a d d i-; c/i c« « wj -S-S-S-S-S-S-^-S^-S-S-S.S^^^S^^^^0 aGQtaaaQac ceoaaaoaaoa S -2 - s: as a: £ a; as as "15'* ^'S'^'^'* * * * *^'* * * -,' *' a;' as' aj" a:' a-' a:' as' a;' as' * a;' as' a;' as' as' as' 272 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 CD c s o c ^ .2 11 N C d cd . — G G Pi J-. o — - J N C » ft a ra c3 1-, c C < W ctj b .2 E "« o ■* -^ d — I in , oo in COOOOOOWO --ocinionoioof, o« oooodoooo«^^2^^aa CO 0> in °* ' I ! I Z< «i -2 = •«. o o S .i=-S " 3 -= S £ a 5 ' •s Co CO Co Co r^r* r* *s 5^ V « « r, « SuCsooaooii ~:s?.<«b = o o a a s S 3 ft CO a ft CO l a i a; a; a; S; a; a; a; a;* -S -s as a; O CJ CJ Cj o o o p •Si cu Cy Cy Cy Cy "9 "53 "9"a-~ o cy a ? 5 ■ a a a a a a a o o o o o o o s> i» s 3; ? L? l?^^,^,^=^-~ -Ji -Jt >i •*--•*-. i- *t- •*- 1- § s = i: i^ i: ^ ™ « ™ ^?"^: ijcjdi^^^^^^jj J) Co ^ co o O" So o fc S -Si to c^ .'o fc idoxu a flei a gei) a oya 8.5 ^ a C3 Pj Cy CU <4> 5j ft C C = c c Cy Cy CD "4) CU /•£ 'o "o "o O O ^: k~ ^ t. *- k. g = c c S cy ca cy c c (a ca Bh UUO oo 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 273 O '-3 || E-< to 4! => „-g a 9| . < K -ti a a E eft, a - s * co PS en PS PS PS PS co PS moopS c/3 co ^ o OOOOCOOOO-hO „Of OfOM'CBIM'J Sooooooioo "(NO S K :< IPS o-< :<« :< ig IPS : : : :X IPS i : i :X PS I : I :X i •< co PS n- ! < : •< :<*- :< PS en i IPS IRS : 2 = o i: ^j "S "^ o 2 ' H c3 o^-^ . to « -a to to -a bj kJ Q *4 E-, ft, CC -a a ft to to a to to CO O S" & ■ rT 71 »3 *; ~ - U 3 — o o o o "S 5 5j O O O "^ -c -c &h a, a. &h O On Gh 274 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 O a ft s -3 a Pi a o e u o o a 6 C.S 8.1 §1 oj rt oj J-2 g - o» T3 3 s 0) C U £ -<- O CD q | 3 3 a eo au .£ CO w6 w ft c << < a. << << << <£ CO C/3 c/5 PS SCffi en en en co ;X ; X la. : PS la. : pS g PS g W < PS I PS X X ; :W :B= << :<< :< C/2 C/3 C/3 Cn ! pj o. pej : o- ! PS W pi W W < pS :«o: i PS PS PS pS c/3 en C/3 C/3 :pS | a. i(£iWrtUW ~ o 23 «= ©©o in i c cS CM PS PS PS PS >o©oo©oo©© -,Hf OOOfOOlOOOOlQ UMMrHf>]'HtMMCl'HOlMCOW OOOi ©oei© ©©©©© eo © © CO © -* — , ~ ©o©* rr 10 r-H © CN T3 3 2 <3 .2 ! ! ^-- r- Ci) Cj ft Ci; Co &5 Co u o ? s; *- c; o a *- co = a g « £ a §-g u 3 — .5 ^* co Co " a a » M I. -J - -■ 5| d) (0 tO 'O (C 1 5 co , O O ° -S -S -2 *S -2 -^ a. .a "3 "3 yphlo hloph totypi a 5,3 » „"G"c: g g~"oft3~~~o a. a. a. =l a a. 5 ^ b ' .? a^ 51 31 o o o o o o o ^ ^4 ^J >< S o o 5^ ft, », O U O o. ft, <. ^ o S- ?■ a a. a. a. a. a a. a c lJ h. lJ kJ tj j >j c c c 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 275 -a a a a S OS & o o '.C < s as 3<; 3 S3 o) "a "5 C C o « «> c > .2 c c ~ -F .3 M ftO 5 £ C M It c i « wo w iX pstd XX XX : XX :X :X : :X 'PC Ed PC w PC : W PC Ed S I :X<*> UN , 'OCK ' ips : = S-2o3g|o°g 0 0-49 0-23 0-24 0-24 0-50 Ol (N (N icU i c^i^ii o o O (N in cq l o ' -si O 2 O o- O n- "' IO T1 TT iUJi o O OlO o I ! ! I I I as •- 5 -g a U£i '■< i= a.~ o ~3 ■ ■s oj a 2 _5 = c -a = 2 ^^ Si«H^HC5C)(o co^osococs&sG's^w as -s^; o = S CO 5 S 2-2 S 3 s. c ft =~-~~ UOO^JOOOOOOOOOO <.< a a a a a a a . 9 .5 3 o i a ,o "ts ^_= g g o.~ c a. Co Co a. n. c o OQQQQQtd -2 -2 .2 -2 -2 ?3 "S ^ "S "^ 0 w CO Si tj«S;5a;a;a S;!: ~ 2 ^-2 •a 2 1 *■?■ ^tr ^1 o •S &c o -C c a a .5 ^ 4S .2 .2 a i» ^ ^ ■*- to co ^ ^r co co ->» -^ O O a. a. § g"c"a o g -a i a a "c "5 § § J .§ 5; a; ^ .5 hj ^j 276 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 .a a o "3 | I -a a ft cd & o c <~ .2 O B H^3 < « 2QJ S n C ■ « « a 9 i.s u K-rt i = JCPS 0) a) a> £ a CO 3 ftO CO - — W . en oo OO "WW : : W * 05 05 < :X (/) co : oo tnU : :WtfXX 05 05 ^05 : :o5 : < g g 05 05 05 g g WWwWa; :W 05050505* < Ul,,l/1(« CO W ™ W W p5 : W 05 * 05 05 -< 0O CO 05 go5Wl 00 05 CO :Wo5XX 05 : co a5 <3 ifi W W :WW W 05p5o5o5O5 '05 " :05co<< WW 05 « 050505"" :XX05 X X w«ww 2*05 05 05 oo : W < 0O CO ,Xoo :g iP5go5XX I : i 05 : : 05 : I g g 05 ! « g g en . to oo W W W Wp- O5O5 05C5 00 ... 00 CO W w W W 05 « 05 05 CO ,, CO CO w £ w w pj 05 b5 o5 CO CO :cooo Woo :Wo5XX ; co 05 05 05 co oo ;oooo W co : W : g :05oo ;WWrt „;WW ,c&t" 0505^ ,0;O5O5 I co 05 , W . .WW ' 05 05 co w '3 'Z§ c° co , rl :oooog : : g 05 05 g X i co 05 : l oo < g g o5 I 05 WW 05 05 CO W r , co CO OO I gggg 'KS^j XX oo : co oo W oo 05 O5 05g CO . n oo oo WSWWj; 05 * 05 05 o _ o © io © © lO tt cq Tf o co : W CO X oo oo W 05 ;ooo5 :oo pfioo<;g g 05 I 05 g g oo oo oo 05 ooo5ooo5coooo5oo^ggo5o5o5gg : oo oo 05 © o in in o o © o) o oo © , © ^ ^ S *■ £ <3 CO « « a-~ ■"* *« a -^ -^ « a. a. fi .2 ^3 ^3 "ti a. 2 2 o o Cj Ci> ca (}> coo I 2» 59 a a Ji g -O 60^2 ~ U) c -ocju-S^a.*"2 co co co co co co co CO aa.a.s5.a.s.e.B, .o p o o o o o o 3 t :■ o i I I I I i | •2 5..' S'S'S 2 60 ^ 8 .*! ~ ° E * ! I 60 .£ o a a o ■2 2i2i 3)Jo Q 6i) 60 S tu -S » -s c 3^:~ = +> _•! 3 r- q Cij c c , ^ jj o o o^:^:^:« ^^^^►J^^J^J^JkJ^kJ ; -a 3 2 3 33aa-~-~00r>~- H o 3 -a 5j cu ca _« ^ ^; O -S: H H H H Szoooo . "a o o -S ^^ £ S3 o-~~~caoo S1 h ''='=- '= 4> *> » « 3 £-2 u' u u Q; (u &) %) : x >< x x x >-j : a ftflS ESSE E'v^ Ui <<<<<<< <>> IPS ipSXW IPS : : PS l£ ipS^WpSPSpS :p- rt ipsxgpspsps :ps ;*s PS IPSX Wps :03 : PS PS XPSpSPSX WpSpSPS IPS PS (75 X ! : : : W : i : oo : PS PS : IX Wpsps :wpS PS XPSpSPSXWpSpScnoopS PS IPSPS O cd in (N o O S"T«205M-1<1102(M(N in y ^ ob ' o O w . NM CO OM" CO (N Aeii^ciiiScici I I ! I I : -a !•* :- 2 t3 a IsS. — -S a. w Z a -a .a cocoCoCoCocoCoCoCo '"a 5 C n U G U fl O Q'O u &rts fls E a cs rt PS w PS PS . Ph . 'PS PS :PS :g IPS w PS ' w :pS IPS : . W IPS B "* PS 'PS ' ' W ,ps 'a 'PS W , , PL, , ^ , . , w PS • ■ w ■ « ■ ' ■ p; ^ '« : ; ^wg PS :psg : :& : PS : : ,g :c/5ps PS ' icnps W w. ! cn PS W PS BSPSPSH :^S| :|«WgpH^| gX • W. w u PS :tf£X :3X : IPS :g i oo PS g : Ph 0)OMO(l.OOl50nlN cDooait^cocoincsi§"j oooo oooo in o o in CM ^H o o S o fflnOO —I =o *^ ^ b^ -^! "? '^l "S "J^ "^ c\, a a. «5 «* as «pa co Co oo _ ^) Cii Qj ■a "a "c o o 2 o h ? P a a o a C9 ^ ^j j- j- j- j- •S "o "o 5 o o o ►JUOOUOU ■x 2 co 2~ 5 g CO -Q O 1- Cj U, to go a ■n 6 o » S a u -= • 2 2 ° ^-5 <:<^ 278 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 A ■^ Pi -. r, o CD 60 B „_, O o B E-i co -3 c m c'5 .2 fc CPh JJ N C ■ Pi'5' w S '3 o ■ < O CD CD us > 03 « 2 ftO i 5 CO • — CD 3 a rt £6 w CCc8^C2CCJftfl)(BU^C5)|v«caj)ci)DO,c:-vJ _Q CD CD CD C CD G CD CD CD CD JJJJ C CD CD CD < J W W en :^PS oSW OS wg^w PS C/3 ww ^ososw os ps gg r|eSPSW os iX en < : W C/5 w OS ' os os wos os CO W OS OS ! W os CO : W OS en iW os C/3 Wen pS OS OS w os i w os OS '03 w w OS . w OS w '* ;S PS : Wen OS en : Wen OS IX ;cn PS :2cn WW OS OS ji n en en I^PSOSW icnpS iWa^jgu : pS oS OS OS W < K^OS H CO n «i< < : :WpS igcnXpSOS^X w a PL.a^'OSW'I'npSXS W p; ; WenpSpSDS PS§enenen^ i X X W as en W tn : pS OS £j X £j X PS 3g :|nWw .«„; .wS^ - * g os ^ K os en : W cS PS PS PS Ch !S-«1 oo in o o m o oi o ooo mioocdoo© in co e>)00ot^roo-rooCv. noiOnHoejHHHO-OrtO OJr-^COcJJCO^Ol-Hin (M^fflSHOeDCJ-i«ngnricJ 3o o ooo oo o o o o o .« to CJ 2 1 1 | i i 0 O ! CO o 1 =o » 3 £ -~ ej ! ooo°ooooooo§ooo i I WBSenWen :*£« :£ t~ t- o o in o o ~h OO CD noo(Mcoi' rtSeCi00HH°cp7-HHHH 2 e» » t) 3t3 ■ft ° -2-c 5 a. a eo£ | j* 3 3 0 <* •2 5 1 co 3 a co c? •= ; | e u j '<0 i 3 ! ! *- ? ? *H o o oooooo o o —I T 2 -a ii ^2 ~ ^ 2 M ~ Cti g u — 0 c ■< 3 3 CO a 5 CO ^ s 0 0 ■2 cy ea o s - CO o 2 CO 5 a I 3 0 ^5. -a a ea co 0 3 0 0 3 CD c .9 60 •2 S i o 2 ■"— .2 w S ~, <0 =0 Q ■2 g 3 s s 60 2 -2 2 § ei o S, 5 -» u c, K H K H - ~- "S cj c-J u o <<<. cci-oacaatj!: CD g ;(i;he-.boo<;fl;<<^ _ O O O p 60 B. C3 C3 3 co co eo co 3=2 O U to O ^.-.^CCCCOO Hj tjj cu C^j t^J Ci) cr 14 14 it « »i -J ^^2;^ 1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 279 — a & ^ o O '43 c c o « « S"> o c -o c c < W 'ar ! § £ c : O J > PQ ■KB a a £0 wo 5 S a | CO C 3 fa 5* h o 0. c i- ^^ H -o ° s .5 o fa£ 3 c a o c o 'So g M OS c n o '3 0 O 0) 'So £ si 2 0 in c > cs in a 3 £Q C 00 rt s CS O fa U 3 + + + — + — + — — + — — — — — — + + + + — — + + + + + + + + + + - + + + + + + + — — — — _ _ + + — — — + — — + + + — — — + — — + — + + + + + — + + + + + + — — + + + + + + + + — — — + + + — — — — — — — _ _ + + + — — — — — — + + + + + + + - - - + + - - + + - - - + - + + + — — — + + — — + + — — — + + — + + + + + + + + + — — — — — — + + + + - - - + + + + - + - - - - - + + — — _ — _ — + + — — — + + — — — + + — — — — — + + + + + — — — + + + + — — — + — — + + + — — — + — — — + + + + + + — — — + + — — — + + - - - + - + + + + — + — + — + + — — — — + + + + + + — + + + + + + + + — — — — + — + + + — — — — — + + + + + + + + + + + + + + + — — — + — — + Tretioscincus bifasciatus Tropidurus torquatus _ Tupinambis teguixin Snakes Boa constrictor Bothrops lansbergi Clclia clelia Chironius carinatus ._ Corallus cnhydris Crotahis durissus Crotalus vcgrandis Drymarchon corais Drymobius margaritiferus Enulius flavortorques Epicrates cenchria Euncctcs murtnus Hclicops angulatus Helicops danieli Helicops scalaris Helminthopus flavotcrminatus . Hydrops triangularis Imantodes cenchoa Leimadophis melanotus Leimadophis reginae Leimadophis typhlus Leptodeira anmdata Leptodeira sepentrionalis Leptophis ahactulla Leptotyphlops macrolepis Leptotyphlops dimidiatus Liotyphlops albirostris Lygophis lineatus Masticophis mentovarius Mastigodryas bifossatus _ Mastigodryas boddaerti Mastigodryas pleei Micrurus carinicauda Micrurus circinalis Micrurus collaris Micrurus dissoleucus Micrurus isozonus Micrurus lemniscatus Micrurus psyches _ Oxybelis aeneus Oxybelis fulgidus Oxyrhopus petola _ Pliimophis guianensis Pseudoboa coronata ._ Pseudoboa neuwiedii 298 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Appendix 11:1 (Concluded). Species Locality Zone 03 CO o ft o o H o *H . H o H en u o 5 2^3 05 o 3 U-i 3 >, .S o Q en O" >, O" t c o oj X X D o — ^3 ir1 £B o •> mo Sfe SB + — — — + — — — + — — — + — — — + + — — + — — — + + + — + — — — + — — — + — — — Vegetation Zone § •a m o ■a C CO 03 ' s " o qj 'So S £Q C "m Jl 03 O 3 Pseustes poccilonotus Rhinobothryum bovallii . Sibon nebulata Spilotes pullatus Tantilla mclanocephala _ Tantilla semicinctum Thamnodynastes strigilis .. Typhlops lehneri Typhlops reticulatus Xenodon rhabdocephalus Xenodon severus + + + + + + + + + + + + + + + + + + + + 12. Composition, Distribution y Origen de la Herpetofauna Chaqueiia Jose M. Gallardo Museo Argentino de Ciencias Naturales Avenida Angel Gallardo 470 Buenos Aires, Argentina Las primeras referencias sobre la herpeto- fauna chaqueiia se tienen a traves de los re- lators y libros de los sacerdotes misioneros de la epoca colonial (Siglos XVII-XVIII). Los Padres Ruiz de Montoya, Lozano, Paucke, Juarez, Sanchez Labrador (vease referencias en Gallardo, 1961b) tienen descripciones y a veces grabados sobre diversas especies cha- queiias de ofidios, saurios, cocodrilos, y tor- tugas, con informaciones por observation directa y otras referencias de los indigenas del "Gran Chaco Gualamba." Algunos natur- alistas viajeros como d'Orbigny (1847) tam- bien se refieren a algunos de los anfibios y reptiles de la region chaqueiia. Como he in- dicado en un trabajo anterior (Gallardo, 1966c), aiios despues Cope (1862) describe especies colectadas por la expedition del Cap- itan T. Page; luego sucesivamente: Stein- dachner (1864), Boettger (1885), Boulenger (1889, 1894, 1898), Peracca (1895), Budgett ( 1899 ) , Mehely ( 1904 ) se ocupan de la fauna chaqueiia. Mas recientemente deben men- cionarse los trabajos de Miiller y Hellmich (1936), Vellard (1948), Cei (1948, 1949, 1950, 1955), Barrio (1965), y Gallardo (1951, 1957, 1959, 1961a, 1962, 1964a-e, 1965a, 1966 a-c, 1968a,b, 1969a, 1971). DESCRIPCION DE LA REGION CHAQUENA La region chaqueiia abarca una enorme area geografica, que va desde el sur de San Jose de Chiquitos y las Serranias de Santiago en Bolivia hasta el norte de la Provincia de Cordoba en Argentina. Posee una longitud de norte al sur de 1280 km. Su ancho maximo en la Argentina, oscila entre 350-400 km, desde el este de Salta hasta la mitad de Formosa, o entre el este de Tucuman y la mitad de Chaco. Cubre los departamentos de Boque- ron y Olimpo de Paraguay, al oeste del Rio Paraguay. En la Argentina abarca toda la Pro- vincia de Santiago del Estero y penetra en los valles del este de Jujuy, el este de Catamarca y de La Rioja y el norte de San Luis (Fig. 12:1). Presenta areas ecotonales o de transi- tion con otras faunas penetrando en el Para- guay hacia el este. En Brasil algunos elemen- tos faunisticos del Chaco llegan al Mato Grosso, al sudeste y al nordeste a favor del Cerrado y la Caatinga; Uegando aun hasta las Guayanas, y algunos elementos alcanzan el Uruguay. En la Argentina se superpone con otras formas — en el este de Formosa y Chaco, el oeste de Corrientes, el noroeste de Entre Rios y el norte de Santa Fe. Ciertos elementos llegan a Mendoza, San Juan, La Pampa, y Rio Negro. La fauna chaqueiia propiamente dicha se extiende por la llanura chaco-bonariense en su parte norte, la cual tiene una elevation entre 100 a 500 m. Por debajo de los 100 m existe la zona de transicion faunistica con la fauna litoral-mesopotamica, y por arriba de los 500 m con la fauna subandina. Algunos elementos chaqueiios avanzan por las serran- ias bajas del oeste. Por la presencia o ausen- cia de rios la fauna se distribuye de acuerdo a las limitaciones de su dispersion. En la region chaqueiia las lluvias maximas anuales se producen en verano ( lo mismo que en la region subandina); en el area litoral- mesopotamica se producen en primavera y otono, y en el area de transicion faunistica las epocas son intermedias. Los promedios anuales chaqueiios van de 550 a 760 mm (en la Argentina); en el area de transicion son practicamente el doble, entre 1000 a 1460 299 300 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fig. 12:1. En el mapa se ha indicado el area abarcada por la herpetofauna chaqueha y la zona dc transi- tion con la fauna litoral-mesopotdmica. Se indican las ciudadcs capitales dentro del area y los rios principalcs. Map showing the area occupied by the Chacoan herpetofauna and the transition zone with the Litoral- Mesopotamian fauna. Capital cities and principal rivers are shown. mm, y en el area subandina son menos de la mitad que en la chaquena (160-325 mm). De modo que existe un escalon pluviometrico coincidente en general con el topografico. En cuanto a las temperaturas anuales hay diferencia entre el area ocupada por la fauna de transition y la ocupada por la litoral-meso- potamica, de la mitad de Santa Fe hacia el sur. Asi en la primavera hay cinco meses de posibles temperaturas bajo cero (entre mayo y septiembre inclusive), con temperaturas minimas de hasta —5° y — 6°C. En la mitad sur de Santa Fe hay seis o siete meses de posibles temperaturas bajo cero (de abril o mayo a octubre inclusive), lo que amplia considerablemente el esquema anterior, ade- mas las minimas pueden alcanzar los — 10°C. Esto se halla ligado a un cambio faunistico, que coincide con la desaparacion hacia el sur de la fauna herpetologica chaquena. En la region considerada domina el par- que chaqueno, con bosques y arboles aislados (Acacia, Prosopis, ScJtinopsis, Aspklospcrma) en pastizales de gramineas (Setaria, Digi- taria, Trichloris) ; a veces se trata de palmares ( Tritrinax, Copernicia ) o de matorrales espi- nosos (en especial, estos ultimos, en areas degradadas), otras son areas de vegetation halofila o falta la vegetation casi totalmente en detenninados espacios. Corresponde basi- camente a la Provincia Chaquena del Dominio Chaquerio (Cabrera, 1971), la herpetofauna chaquena tambien puede extenderse por las provincias del Monte y del Espinal. No existe, en cambio, en ambientes puramente herbaceos como los del sur de Santa Fe y de Cordoba, y la Provincia de Buenos Aires, que corresponden a la llamada "Pampa humeda." Es alii donde vemos que existe una com- ponente climatica y fitogeografica que no favorece el avance de la herpetofauna cha- quena hacia el sur. 1979 GALLARDO: HERPETOFAUNA CHAQUENA 301 COMPOSICI6N HERPETOFAUN1STICA La herpetofauna chaquena se compone con cinco familias de anfibios ( 14 generos, 30 especies) y 13 familias de reptiles (35 generos, 49 especies). Entre los anfibios hay una marcada abundancia de leptodactylidos (6 generos, 16 especies) y en segundo ter- mino los hylidos ( 3 generos, 7 especies ) . Los saurios tienen un marcado predominio de iguanidos (6 generos, 7 especies) y en se- gundo termino de teiidos (5 generos, 6 espe- cies). De las 79 especies de anfibios y rep- tiles que habitan la region chaquena, 40 especies (21 anfibios, 19 reptiles) son basica- mente endemicas de esa region (Tabla 12:1). ECOLOGIA Y ADAPTACIONES En las especies de anfibios y reptiles se notan interesantes adaptaciones a las condi- ciones ambientales. Adaptaciones de los Anfibios Las epocas de reproduction de los anfibios chaquenos se hallan en coincidencia con las epocas de lluvia, lo que resulta de vital im- portancia en un area relativamente seca. Pero dentro de ese contexto hay adaptaciones pro- pias de algunas especies. Construction de nidos. — Hay varios tipos de nidos. 1. Nidos de espuma en cuevas en el sue- lo. Las ranas en epocas de lluvia hacen cuevas en el barro humedo; es el macho que se encarga de esta construccion y el que asomado a la cueva canta Ilaman- do a la hembra. Esto es seguido por el amplexo, construccion del nido de espuma por la pareja, puesta y fecun- dation; la pareja luego abandona la cueva. El desarrollo embrionario y la primera parte de la vida larval ocurren dentro de la cueva. Una ulterior lluvia causa la inundation de la cueva y la salida de los renacuajos del nido; ellos completan rapidamente su desarrollo y metamorfosis. Este comportamiento se da en Leptodactylus anceps, L. bufo- nius, y L. gualambensis. 2. Nidos de espuma flotantes. No hay construccion de cuevas, por parte de las ranas, los nidos son construidos por la pareja. Esto se da en las ranas Lep- todactylus chaquensis y L. laticeps, y las especies de Physalaemus y Pleuro- dema. 3. Nidos en las hojas. Son construidos en hojas sobre ramas de arboles, sobre cuerpos de agua, adonde luego caen los renacuajos para completar su desarrollo larval. Tal sucede en las especies de PhyUomedusa, ranas arboricolas. Desarrollo larval. — Varias estrategias son usadas por los renacuajos. 1. Desarrollo rapido de unas dos semanas de duracion, se da en las ranas con- structoras de cuevas en el suelo y nidos de espuma (vease las especies men- cionadas anteriormente ) . 2. Desarrollo lento de una duracion de uno o dos meses en los sapos del genero Bufo, algunas ranas del genero Lepto- dactylus, y varios hylidos. Adaptaciones contra el desecamiento. — En las epocas de receso reproductive y de falta de lluvias, varias adaptaciones permiten la supervivencia en estos meses desfavorables. 1. Formas cavicolas. En particular las especies de Ceratophrys, Lepidobatra- chus, Odontophrynus, y Pleurodema poseen tuberculos tarsales o metatar- sals cavadores que facilitan su oculta- miento bajo tierra, donde permanecen hasta que se produzcan las lluvias. 2. Formas comensales. Algunas especies, como Leptodactylus laticeps, L. cha- quensis, L. bufonius, y Bufo arenarum, se han adaptado a vivir en cuevas de roedores, especialmente de vizcacha (Lagostomus maximus), lo que les permite soportar largos periodos de sequias. Otros, como Dermatonotus mulleri y Elachistocleis bicolor habitan termiteros. 3. Acumulacion de mudas de piel. Otra forma de defensa consiste en la accu- mulation de la muda, que se da en especial en las especies de Ceratophrys (Gallardo, 1953) y Lepidobatrachus (McClanahan, Shoemaker, y Ruibal, 302 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Tabla 12:1. — Lista de los Anfibios y Reptiles Chaquenos y Sus Distribuciones en Regiones Ultracbaqueiias. ( + = presente; — = ausente; X = representado por otra subespeeie; * = basicamente endemica de Chaco) Especies Litoral- Mesopotamica Pampas Patagonia Region Subandina Batracios Ceratophrys pierotti" Ceratophrys sp.° Lepidobatrachus asper" Lepidobatrachus llanensis" ._ Lepidobatrachus salinicola" ... Leptodactylus anceps" Leptodactylus bujonius" Leptodactylus chaquensis" ... Leptodactylus gualambcnsis" Leptodactylus laticeps" . Odontophrynus americanus ._ Physalaemus albonotatus" Physalaemus biligonigerus Pleurodema borellii" Pleurodema guayapae Pleurodema tucumana" Bufo arenarum Bufo granulosus ~ Bufo paracnemis _ Melanophryniscus stelzneri ... Pseudis paradoxus Hyla acuminata'' Hyla fuscovaria" Hyla raniceps Hyla x-signata Phrynohyas venulosa — Phyllomedusa hypocondrialis Phyllomedusa sauvagii" Elachistocleis bicolor Dermatonotus mulleri" Saurios Homonota horrida" Phyllopezus pollicaris" Leiosaurus paronae" Liolaemus chacoensis" Ophryoessoides caducus" Pristidactylus vautieri" Proctotretus doellojuradoi" ... Tropidurus spinulosus Tropidurus sp.° Cnemidophorus Icachi" Gymnophthalmus rubricauda" Kentropyx lagartija" Kcntropyx viridistriga" Teius teyou Tupinambis rufescens Ophiodes intermedins Mabuya frenata Anfisbenios Amphisbacna camura" Anops kingii Lcpostcrnon microccphalum . Ofidios Leptotyphlops albipuncta" ... Leptotypldops unguirostris Leptotyphlops weyrauchi" Constrictor constrictor Epicrates cenchria _ + + X X X X + + X X X X + + + + + + + + + + + + + + 1979 GALLARDO: HERPETOFAUNA CHAQUENA 303 E species Litoral- Mesopotamica Pampas Patagonia Region Subandina Eunectes notaeus Clclia clclia CIclia occipitolutea . Elapomorpltus tricolor Leimadophis sagittifer Ltjsirophis dorhignyi _ Lystrophis semicinctus Oxyrhopus rhombifer Philodryas aestivus _ Philodryas baroni Philodryas patagonicnsis ... Philodryas psammophideus Phimophis vittatus ... Pseudotomodon trigonatus . Sibynomorphus turgidus ... Waglerophis merremii Micrurus frontalis . Boihrops altcrnatus Bothrops ncuwiedii Crotalus durissus _ Quelonios Kinosternon scorpioides Gcochelone petersi" Cocodrilos Caiman latirostris 1- + + — + — + + + + + + + + + — + + + - + — + — + + + — + + + + + + + + + + + + + + + + + + 1976). Ademas los hylidos arboricolas Phyllo medusa sauvagii y P. hypocon- drialis cubren el cuerpo entero con la secretion de lipidos de las glandulas cutaneas alveolares distribuida con las manos y los pies (Blaylock, Ruibal, y Piatt- Aloia, 1976). 4. Refugios en huecos de arboles. El fondo de estos huecos es impermeabili- zado por secreciones cutaneas, donde se protegen las ranas durante periodos de sequias. Este comportamiento es pro- pio de un hylido, Phrynohyas venulosa. Otras Adaptaciones Alimentation. — En los anfibios el con- sumo de presas es principalmente de artro- podos (insectos, aracnidos, y miriapodos), moluscos, y pequefios vertebrados. Los mas grandes entre ellos (Leptoclactylus chaquen- sis, Bufo paracnemis, Ceratophrys, y Lepido- batrachus) pueden capturar otros anfibios y roedores. En los saurios la dieta es principal- mente de artropodos, salvo los de gran ta- mafio, como Tupinambis, donde ademas de la dieta vegetal, se agregan pequefios vertebra- dos y moluscos. En los ofidios se da la dieta de pequefios vertebrados, en especial anfibios y roedores, a veces saurios, otros ofidios, o aves. En quelonios, Geochelone es basica- mente vegetariana (cactos, frutos, y hojas); Kinosternon es animalivoro. En cocodrilos. Caiman varia la dieta con la edad; los juve- niles se alimentan de insectos, moluscos, y anfibios, ya adultos capturan peces y verte- brados superiores. Coloraciones en funcion defensiva. — En varios anfibios y reptiles se dan las colora- ciones cripticas. Mientras que otros como Lcptodactylus laticeps y Micrurus poseen col- oraciones aposematicas en relation con secre- ciones de action toxica o de veneno. Utilization del habitat. — Los anfibios salen de los refugios solamente en las epocas de lluvias. Varios saurios usan diferentes partes de su habitat estructural — bosques, pastizales, piedras, y el suelo (Fig. 12:2). Comportamiento territorial. — Se ha com- probado en algunos anfibios del genero Lep- todactylus y de Phyllomedusa, ademas de los saurios del genero Tropidurus. 304 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fig. 12:2. Esquema de la utilization del habitat de algunos saurios y anfisbenios en la zona norte de Cordoba y sur de Santiago de Estero, Argentina. Sketch of the habitat utilization by some lizards and amphisbaenians in the area north of Cordoba and south of Santiago del Estero, Argentina. 1. Tropidurus spinulosus, Tropidurus sp., 2. Homonota horrida, 3. Leiosaurus paronae, 4. Pristidactylus vautieri, 5. Ophiodes intermedins, 6. Amphisbaena camura, Anops kingii, 7. Tupinambis rufescens, Teius teyou, Cnemidophorus leachi, 8. Proctotretus docllojuradoi, 9. Liolaemus chacoensis. ORIGEN Y RELACIONES DE LA HERPETOFAUNA Con respecto a los anfibios y en grado menor a los saurios, se nota un paralelismo o aparente repetition de faunas con especies y subespecies vicariantes o cercanas entre si, en las faunas chaquefia y litoral-mesopotamica (Tabla 12:2). Pareceria tratarse de dos faunas de un mismo origen guayano-brasileno, luego difer- enciadas por aislamiento y nuevamente en contacto en algunos puntos o areas mas o menos amplias. Aqui cabe aplicar la teoria de los refugios, en periodos climaticos des- favorables, para luego producirse el repobla- miento en los periodos favorables. En cuanto a los anfibios el papel de los sistemas hidrograficos ha sido fundamental para la distribution de las especies; esta ac- tion continua tambien en nuestros dias. La teoria de los refugios y la de la fidelidad a los sistemas hidrograficos se complementan al explicar la subespeciacion de Bufo granulosus Tabla 12:2.- Mesopotamia. -Especies y Subespecies Vicariantes o Cercanas Entre si en las Faunas Chaquefia y Litoral- Chaqi Litoral-Mesopotamia Ceratophrys sp. Leptodactylus anceps Leptodactylus bufonius Leptodactylus chaquensis Leptodactylus gualambcnsis Bufo arcnarum chaguar Bufo granulosus major Pseudis paradoxus occidcntalis Liolaemus chacoensis Cnemidophorus leachi Teius teyou cyanogaster Tupinambis rufescens Ceratophrys ornata Leptodactylus prognathus Lep todactylus m y statin us Leptodactylus ocellatus Leptodactylus gracilis Bufo arcnarum platensis Bufo granulosus fernandczac Pseudis paradoxus platensis Liolacm us wiegmannii Cnemidophorus lacertoides Teius teyou teyou Tupinambis tequixin 1979 GALLARDO: HERPETOFAUNA CHAQUElVA 305 Fig. 12:3. Mapa de las areas ocupadas por las subespecies de Bufo granulosus, coincidentes con los llamados refugios (adaptado de Gallardo, 1965b). Map of the areas inhabited by the subspecies of Bufo granulosus, coinciding with the so-called refuges (adapted from Gallardo, 1965b). 1. B. g. humboldti, 2. barbouri, 3. beebei, 4. merianae, 5. goeldi, 6. mini, 7. mirandaribeiroi, 8. lutzi, 9. granulosus, 10. major, 11. azarae, 12. fernandezae, 13. pygmaeus, 14. dorbignyi. y de Pseudis paradoxus. Asi en dos trabajos anteriores (Gallardo, 1965b, 1969b) se de- scriben 14 subespecies para Bufo granulosus (Fig. 12:3), correspondientes basicamente a los refugios que otros autores han asignado para diversas especies de animales (Haffer, 1969, 1974). Sin embargo, al mismo tiempo esas subespecies se extienden por los sistemas hidrograficos correspondientes; a la fauna chaquena corresponde Bufo granulosus major. En otro trabajo sobre Pseudis paradoxus (Gallardo, 1961c) tambien se encuentra una coincidencia de este tipo, correspondiendo a la fauna chaquena, Pseudis paradoxus occi- dentalis. Los ofidios son de mas amplia distribution y en general no responden tanto al modelo aplicable a los anfibios y saurios. Asi de las 26 especies de ofidios citados para la fauna chaquena, 8 tambien habitan la Provincia de Buenos Aires; mientras que los anfibios solo dos a tres coinciden, sobre un total de 30 especies citadas para la fauna chaquena. Entre los saurios chaqueiios, Leiosaurus pa- ronae tiene sus vicariantes en la fauna sub- andina y en la Patagonia — L. catamarcensis y L. bellii, respectivamente (Gallardo, 1961a). Otro tanto puede decirse con respecto a Lio- laemus chacoensis y las especies subandinas y patagonicas del genero Liolaemus y de Homonota horrida con respecto a sus vicari- antes H. borellii y H. darwinii. Por lo que podemos suponer un parentesco andino-pata- gonico para algunos saurios chaqueiios. Los cocodrilos se han distribuido a traves de los rios y han alcanzado localidades muy al oeste en la region; en esto coinciden con la distri- bucion de los peces. Hay constancias paleontologicas de un avance mucho mas hacia el oeste de faunas similares a las del Rio Parana, lo que habria coincidido con un avance de una fauna orig- inalmente de un ambiente mas humedo, que dio origen a la fauna herpetologica chaquena, adaptada a condiciones climaticas mas rigu- rosas. Asi surge de los trabajos de Fernandez ( 1976 ) , quien sefiala el hallazgo de esta fauna acuatica fosil, correspondiente al Eoceno (Formacion Lumbreras), en la actual puna jujefia. Por otra parte el ambiente chaqueno se habria extendido muy hacia el sur, alcan- zando el sudeste de la Provincia de Buenos Aires (Monte Hermoso), de acuerdo a los trabajos paleontologicos de Tonni (1974) y de Chani (1977), pues segun dichos autores esto se deberia a que habria habido fluctua- ciones paleoecologicas de climas secos y climas humedos entre el Plioceno superior y la actualidad. Vease Baez y Scillato Yane (este volumen) para una exposition com- pleta de los paleoclimas de la region. En general se nota actualmente un em- pobrecimiento faunistico hacia el sur, con re- specto a la fauna de anfibios del sudeste de Brasil y la Provincia de Misiones, Argentina, en la fauna litoral-mesopotamica. Es asi como en la Provincia de Buenos Aires de 22 especies 306 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 existentes en el nordeste de la provincia que- dan solamente seis en el sur. En zonas inter- medias, como la Provincia de Corrientes, hay una duplication de faunas al coexistir en parte de la provincia las faunas chaquenas y litoral-mesopotamia (33 anfibios sefialados para Corrientes). SUMMARY The Chacoan herpetofauna occupies the northern part of the Chaco-Bonariensian Plain having an elevation of 100 to 500 m. Below 100 m there is a transition to the Litoral-Meso- potamian fauna; above 500 m there is a tran- sition to the subandean fauna. In the Chaco the rainfall gradient generally coincides with the topography. The vegetation principally is a Chacoan park type with monte and isolated trees interspersed in grassland; thorn bushes and halophytic vegetation are present locally. Rivers influence faunal distributions. The herpetofauna is composed of five fam- ilies of amphibians, five of lizards, one of amphisbaenians, five of snakes, two of tur- tles, and one of crocodilians. There is a total of 52 genera and 79 species (Table 12:1). The amphibians and to a lesser degree the lizards have a parallel origin with the Litoral- Mesopotamian fauna. Both apparently orig- inated from the same Guiana-Brasilian stock, which differentiated later through isolation and reinvasion of some areas. This view fits the refugia theory for survival during unfav- orable climatic periods with subsequent re- invasion of areas when favorable climates returned. River systems have been fundamen- tal to the distribution of amphibians. Snakes have broader distributions and generally do not follow the patterns of amphibians and lizards. Turtles and crocodilians reach farth- est west through the rivers; their distributions tend to coincide with those of the fish fauna. The Chacoan herpetofauna seems to have originated from the Parana River fauna, which in the past extended to the west, where it subsequently adapted to the more rigorous climatic conditions. Among the species of the Chacoan herpetofauna are several that show reproductive, dietary, and/ or ethologi- cal adaptations denoting their adjustment to the conditions present on the Chaco-Bonar- iensian Plain. BIBLIOGRAFIA Barrio, A. 1965. Afinidades del canto nuptial de la especies cavicolas del genero Lcptodactylus (Anura, Leptodactylidae) . Physis 25:401-410. Blaylock, L. A., Ruibal, R., Platt-Aloia, K. 1976. Skin structure and wiping behaviour of phyllome- dusine frogs. Copeia 1976(2) :283-295. Boettger, O. 1885. Liste von Reptilien und Bat- rachiern aus Paraguay. Z. Naturforsch. 58:213- 248. Boulencer, G. A. 1889. On a collection of batra- chians made by Prof. Charles Spegazzini at Co- lonia Resistencia, South Chaco, Argentine Repub- lic. Ann. Mus. Civ. Stor. Nat. Genova 7:246-249. Boulenger, G. A. 1894. List of reptiles and batra- chians, collected by Dr. J. Bohls near Asuncion, Paraguay. Ann. Mag. Nat. Hist. (6)13:342-348. Boulenger, G. A. 1898. List of reptiles, batrachians and fishes collected by Cav. Guido Boggiani in the northern Chaco. Ann. Mus. Civ. Stor. Nat. Genova (9)39:125-126. Budgett, J. S. 1899. Notes on the batrachians of the Paraguayan Chaco, with observation upon their breeding habits and development, especially with regard Phyllomedusa hypochondrialis Cope. Also description of new genus. Q. J. Microsc. Sci. 42:305-333. Cabrera, A. L. 1971. Fitogeografia de la Republica Argentina. Bol. Soc. Argentina Bot. 14:1-42. Cei, J. M. 1948. El ritmo estacional en los feno- menos ciclicos endocrinosexuales de la rana cri- olla (Lcptodactylus ocellatus (L.)) del Norte Argentino. Acta. Zool. Lilloana 6:283-331. Cei, J. M. 1949. Generalidades sobre el ciclo sexual y el predominio de la espermatogenesis anual continua en varios batracios de la region cha- quena. Ibid. 7:527-544. Cei, J. M. 1950. Lcptodactylus chaqucnsis n. sp. y el valor sistemiitico real de la especie linneana Lcptodactylus ocellatus en la Argentina. Ibid. 9: 395-423. Cei, I. M. 1955. Chacoan batrachian in central Ar- gentina. Copeia 1955(4):291-293. Cham, J. M. 1977. Relaciones de un nuevo Teiidae (Lacertilia) fosil del Plioceno Superior de Ar- gentina, Callopistes bicuspidatus n.sp. Rev. Inst. Miguel Lillo Publ. Espec: 133-153. Cope, E. D. 1862. Catalogue of reptiles obtained during the exploration of the Parana-Paraguay- Bermejo and Uruguay rivers by Cap. Thos. J. Page. Proc. Acad. Nat. Sci. Philadelphia 14:346- 359. D'Orbicny, A. 1847. Voyage dans l'Amerique Mer- idionale. Iere. Partie, Reptiles 5:5-12. Duellman, W. E., Veloso, M. A. 1977. Phylogeny of Pleurodcma (Anura, Leptodactylidae). A bio- geographic model. Univ. Kansas Mus. Nat. Hist. Occas. Pap. ( 64 ) : 1-46. 1979 GALLARDO: HERPETOFAUNA CHAQUEftA 307 Fernandez, J. 1976. Hallazgo de peces pulmonados fosiles en la puna jujeria. Ann. Soc. Cient. Argen- tina 201:1.3-18. Gallardo, J. M. 1951. Sobre un Teiidae (Reptilia, Sauna ) poco conocido para la fauna argentina. Com. Mus. Argent. Cienc. Nat. Bernardino Riva- davia Inst. Nac. Invest. Cienc. Nat. Zool. 2:1-8. Gallardo, J. M. 1953. El escuerzo como animal de terrario. Ichthys 1:75-79. Gallardo, J. M. 1957. Las subespecies argentinas de Bufo granulosus Spix. Rev. Mus. Argent. Cienc. Nat. Bernardino Rivadavia Inst. Nac. Invest. Cienc. Nat. Zool. 3:337-374. Gallardo, J. M. 1959. Sobre un Iguamdae del noroeste argentino, Leiocephalus caducus ( Cope ) . Acta. Zool. Lilloana 17:485-497. Gallardo, J. M. 1961a. Estudio zoogeografico del genero Lciosaurus (Reptilia, Sauria). Physis 22: 113-118. Gallardo, J. M. 1961b. Panorama zoologico argen- tino: Batracios y Reptiles. Ibid. 22:171-180. Gallardo, J. M. 1961c. On the species of Pseudidae (Amphibia, Anura). Bull. Mus. Comp. Zool. Harvard Univ. 125:111-134. Gallardo, J. M. 1962. El genero Kentropyx (Sauria, Teiidae) en la Republica Argentina. Acta Zool. Lilloana 18:243-250. Gallardo, J. M. 1964a. Los Anfibios de la Provin- cia de Entre Rios, Argentina y algunas notas sobre su distribucion geografica y ecologia. Neo- tropica 10:23-28. Gallardo, J. M. 1964b. Consideraciones sobre Lep- todactylus occllatus (L.) (Amphibia, Anura) y especies aisladas. Physis 24:373-384. Gallardo, J. M. 1964c. Leptodactylus gracilis (D. et B.) y especies aisladas (Amphibia, Leptodac- tylidae). Rev. Mus. Argent. Cienc. Nat. Bernar- dino Rivadavia Inst. Nac. Invest. Cienc. Nat. Zool. 9:37-57. Gallardo, J. M. 1964d. Una nueva forma de Pseu- didae (Amphibia, Anura) y algunas considera- ciones sobre las especies argentinas de esta fami- lia. Acta Zool. Lilloana 20:193-209. Gallardo, J. M. 1964e. Leptodactylus prognathus Boul. y L. mystacinus (Burm.) con sus respectivas especies aliadas (Amphibia, Leptodactylidae del grupo Cavicola). Rev. Mus. Argent. Cienc. Nat. Bernardino Rivadavia Inst. Nac. Invest. Cienc. Nat. Zool. 9:91-121. Gallardo, J. M. 1965a. Una nueva subespecie cha- queria Bufo arenarum cliaguar (Amphibia, Bufo- nidae). Neotropica 11:84-88. Gallardo, J. M. 1965b. The species Bufo granu- losus Spix (Salientia, Bufonidae) and its geo- graphic variation. Bull. Mus. Comp. Zool. Har- vard Univ. 134:107-138. Gallardo, J. M. 1966a. Liolaemus lentus nov. sp. (Iguanidae) de La Pampa y algunas observa- ciones sobre Ios saurios de dicha provincia argen- tina y del oeste de Buenos Aires. Neotropica 12: 13-29. Gallardo, J. M. 1966b. Las especies argentinas del genero Ophiodes Wagler. Rev. Mus. Argent. Cienc. Nat. Bernardino Rivadavia Inst. Nac. In- vest. Cienc. Nat. Zool. 9:123-146. Gallardo, J. M. 1966c. Zoogeografia de los anfibios chaquenos. Physis 26:67-81. Gallardo, J. M. 1968a. Relaciones zoogeograficas de la fauna batracologica del oeste de la Provin- cia de Santa Fe (Argentina). Com. Mus. Argent. Cienc. Nat. Bernardino Rivadavia Inst. Nac. In- vest. Cienc. Nat. Ecol. 1:1-13. Gallardo, J. M. 1968b. Las especies argentinas del genero Mabuya Fitzinger ( Scincidae, Sauria). Rev. Mus. Argent. Cienc. Nats. Bernardino Riva- davia Inst. Nac. Invest. Cienc. Nat. Zool. 9:177- 196. Gallardo, J. M. 1968c. Sobre la validez de algunas especies argentinas de Pleurodcma ( Anura, Lepto- dactylidae). Physis 28:135-144. Gallardo, J. M. 1969a. Especies de saurios (Rep- tilia) de la Provincia de Santa Fe, Argentina y consideraciones sobre su ecologia y zoogeografia. Neotropica 15:73-81. Gallardo, J. M. 1969b. La distribucion de las sub- especies de Bufo granulosus Spix: Su fidelidad a los sistemas hidrografieos Sudamericanos. Cien. Invest. 25:406-416. Gallardo, J. M. 1971. Composition faunistica de los Saurios de la Provincia de La Pampa, Repub- lica Argentina. Neotropica 17:44—48. Haffer, J. 1969. Speciation in Amazonian forest birds. Science 165:131-137. Haffer, ]. 1974. Avian speciation in tropical South America. Publ. Nuttall Ornithol. Club, Cam- bridge 14:1-390. McClanahan, L. L., Jr., Shoemaker, V. H., Rui- bal, R. 1976. Structure and Function of the Cocoon of a Ceratophryid Frog. Copeia 1976 (1):179-185. Mehely, L. 1904. Investigations on Paraguayan batrachians. Ann. Mus. Nat. Hungary 2:207-231. Muller, L., Hellmich, W. 1936. Amphibien und Reptilien. I Teil: Amphibia, Chelonia, Loricata. Wiss. Ergeb. Deutsch. Gran Chaco-Expedition. Stuttgart, 120 p. Peracca, M. G. 1895. Viaggio del dott. Alfredo Borelli nella Rep. Argentina e nel Paraguay rettili ed anfibi. Boll. Mus. Zool. Torino 10( 195): 1-32. Steindachner, F. 1864. Batrachologische Mitteil- ungen. Verh. Zool. Bot. Ges. Wien. 14:239-288. Tonni, E. P. 1974. Un nuevo Cariamido (Aves, Gruiformes) del Plioceno Superior de la Provin- cia de Buenos Aires. Ameghiniana 11:366-372. Vellard, J. 1948. Batracios del Chaco Argentino. Acta Zool. Lilloana 5:137-174. 13. The Patagonian Herpetofauna Jose M. Cei Institute) de Biologia Animal Universidad National de Cuyo Casilla Correo 327 Mendoza, Argentina The word Patagonia is derived from the term "Patagones," meaning big-legged men, applied to the tall Tehuelche Indians of southernmost South America by Ferdinand Magellan in 1520. Subsequently, this pic- turesque name came to be applied to a con- spicuous continental region and to its biota. Biologically, Patagonia can be defined as that region east of the Andes and extending southward to the Straits of Magellan and eastward to the Atlantic Ocean. The northern boundary is not so clear cut. Elements of the Pampean biota penetrate southward along the coast between the Rio Colorado and the Rio Negro (Fig. 13:1). Also, in the west Pata- gonian landscapes and biota enter the vol- canic regions of southern Mendoza, almost reaching the Rio Atuel Basin. The Pata- gonian region has a wide ecotonal zone with the Chacoan region (Gallardo, this volume). The monte vegetation (Morello, 1958) with its several formations containing numerous subtropical elements extends south to the Peninsula de Valdes; the monte enters the Rio Chubut drainage and extends westward to the Rio Neuquen, Rio Agrio, and Rio Limay valleys. South of the Rio Negro, the monte associations exist in a system of saline low- lands (bajos) and reach irregular spurs of the Meseta de Somuncura, a typical Patagon- ian environment (Cei, 1969a,b; Ruiz Leal, 1972). Nevertheless, there is a general, some- times remarkable, agreement between the phytogeographic boundaries of the Monte- Pampean and the Patagonian regions and the distribution patterns of their herpetofaunas. Herein I emphasize the biota of the Cis- Andean steppe to the near exclusion of the Trans-Andean austral forest ecosystems treated by Formas (this volume). Patagonia is a region of sedimentary rocks and soils, mostly tablelands subjected to pro- longed erosion. Scattered through the region are extensive areas of extrusive basaltic rocks. The open landscape is dissected by transverse rivers descending from the snowy Andean cordillera; drainage is poor near the Atlantic coast. Patagonia is subjected to severe sea- sonal drought with about five cold winter months and a cool dry summer, infrequently interrupted by irregular rains and floods. HISTORY OF THE PATAGONIAN BIOTA In contrast to the present, almost uniform steppe associations in Rio Negro, Chubut, and Santa Cruz provinces, during Oligocene and Miocene times tropical and subtropical vegetation occurred along with xerophytic woodlands with luxuriant mesophytic gallery forests. A comparison of the rich Miocene flora of Pichi Leufu, Rio Negro (Berry, 1938) with analagous associations from Mirhoja, Chubut; Valcheta, Rio Negro; and Rio Chalia, Santa Cruz, shows a mixture of mesic tropical elements (Ficns, Fagara, Nectandra, Tabe- buia, Mijristica, Sterculia, tree ferns, Erythro- xylon, Oreopanax, Maytenus), including climbers (Buettneria, Banisteria, Bignonia, Cissus, Paullinia, Sapindus, Strychnos), to- gether with nontropical genera (Araucaria, Azora, Berberis, Ginkgo, Laurelia, Emboth- rium, Fitzroya, Libocedrus, Podocarpus, Lo- matia, Peumus, Myrceugenia, Drimys). Most of the latter are characteristic components of the present temperate Valdivian forest. Nev- ertheless, xeric areas in the Middle Tertiary of Patagonia are suggested by certain paleo- floras containing Schinopsis, Schinus, and Cu- pania. The former is a significant genus of trees in the subtropical Chacoan region. Nothofagus forests were widespread in Patagonia in the Eocene and Oligocene, but 309 310 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 &0, /Canquel Meseta V SOUTHERN TEMPERATE PHYTOGEOGRAPHICAL REGIONS Greaf~Central ^V ESPINAL C. Chacoan Province MONTE S. Chacoan Province \5L ( PATAGONIAN Steppe Region ~^JW gg w AUSTRAL Forest Belt ^-^ \ 1 500 KM i i — i 1 Fig. 13:1. Phytogeographic regions of austral South America. Regiones fitogeogrdficas tie Sutl America austral. 1979 CEI: PATAGONIAN HERPETOFAUNA 311 400 KM TERTIARY PATAGONIAN FLORAS AND HERPETOFAUNA Miocene Patagonian Tropical and Valdivian Floral Remains Temperate Austral Forest Belt /Valdivian Flora Present Lower Limit of Caudiverbera (to 30° S. I at.) Oligocene Eupsophus and Neoprocoela [Scarritt Pocket) Miocene Ceratophrynid Wawelia Oligo-Miocene I Patagonian Caudiverbera Fig. 13:2. Paleontological records of the lower Tertiary Patagonian flora and of leptodactylid frogs. Hallazgos paleontologicos de flora patagdnica del Terciario inferior y de anuros leptodactylidos. 312 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 these associations decreased and retreated southward and westward in the Middle and Late Tertiary. Chusquea bamboo groves oc- cur in Cenozoic deposits at Laguna Hunco, Chubut. Further evidence of paleoclimatic conditions in Patagonia is derived from the extensive paleo-mammal faunas ( Raez and Scillato Yane, this volume) and more limited paleo-herpetofaunas (Raez and Gasparini, this volume). Primitive leptodactylid, ceratophrynine, and bufonid frogs have been recorded by Schaeffer (1949), Chaffee (1952), and Casa- miquela (1963) from the Deseadan, early Oligocene Scarritt Pocket Formation (Can- quel, Chubut) and from the upper Miocene of Rio Negro (Fig. 13:2). The living Chilean frog Caudiverbera caudiverbera is almost identical to the fossil frogs of the same genus. A fossil frog from the Oligocene of Chubut referred to Eupsophus by Schaeffer (1949) has been considered the same as living E. roseus, a species characteristic of the Notho- fagus forests of southern Chile (Rogart, 1970). These fossils clearly establish the pres- ence of telmatobiine frogs in Patagonia in the Oligocene and Miocene. The Oligocene Neoprocoela was provision- ally referred to a Batrachophrynus or Telma- tobius-like leptodactylid stock by Schaeffer (1949). Tihen (1962) considered it to be a species of Bufo in the Palearctic Bnfo cala- mita group. The fossil was again associated with the telmatobiine genera Tehnatobufo and Batrachophrynus by Lynch (1971). New material from the same formation supports the inclusion of Neoprocoela in Bufo (Estcs, pers. comm. ) . The placement of Neoprocoela in the Bufo calamita group has interesting biogeographical implications. Serological evi- dence (Cei, 1977) supports a relationship be- tween the European Bufo calamita and the small B. variegatus presently restricted to the austral Nothofagus forests of Argentina and Chile (Gallardo, 1962). The presence of a ceratophrynine frog (Wawelia gerholdi) in the Miocene provides herpetological evidence for the southward ex- tent of tropical elements in the Middle Ter- tiary. Reptilian remains substantiate the long history of tropical elements in Patagonia. The fossil snake Dinilysia patagonica from the Upper Cretaceous of Neuquen is related to boids and aniliids that are widespread in trop- ical South America. Furthermore, boid snakes (Madtsoia), crocodilians (Necrosuchus, Se- becus, Eocaiman), and meiolaniid and pelo- medusid turtles from Paleocene-Eocene de- posits in Chubut are indicative of tropical environments (Gasparini and Raez, 1975). Iguanid (Erichosaurus debilis) and teiid (Di- asemosaurus occidentalis) lizards lived in southern Santa Cruz in the Miocene. PATAGONIAN FAUNAL REGIONS Two major faunal regions (habitats) can be defined in Patagonia. These are the north- ern or ancient region and the southern or Santa Cruz region; the border between these regions is approximately at the Rio Chubut at 45°S (Fig. 13:3). These habitats corre- spond to ancient physiographic areas, the Patagonian Massif and the Deseado Massif, respectively (Figs. 13:4-5). These massifs are ancient structural continental units known as nesocratons (Harrington, 1962). In spite of its less marked subpositive tendency in comparison with the Pampean Massif, the whole region of the Patagonian Massif has been a site of almost uninterrupted accumu- lation of continental deposits. More rarely it received shallow marine deposits peripherally at times of oceanic transgressions in the Eocene-early Oligocene, middle Oligocene, and middle Miocene. The smaller Deseado Massif had a subpositive tendency even less marked than the Patagonian Massif; accord- ingly, its relief was often depressed, and dur- ing prolonged subsidences it became a sedi- mentary area like the adjacent pericratonic basins (Harrington, 1962). The northern or ancient Patagonian re- gion extends through Neuquen, Rio Negro, and Chubut provinces (Fig. 13:6). The sub- cordilleran area in Neuquen is drained by the Rio Agrio and Rio Neuquen, which flow into the Rio Limay, a tributary of the Rio Negro. Extra-cordilleran mesetas include the large Meseta de Somuncura ( 1000-1700 m eleva- 1979 CEI: PATAGONIAN HERPETOFAUNA 313 PATAGONIAN MAJOR ANIMAL HABITATS Northern or Ancient Patagonian Major Animal Habitat Southern or Santa Cruz Major Animal Habitat 500 KM Fig. 13:3. Major herpetofaunal regions of Patagonia. Regiones herpetojaunisticas fundamentals de Patagonia. 314 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Shield Shield Francisco sin tan Deseado "x Massif CONTINENTAL GEOTECTONIC UNITS Cratons and Nesocratons Geosynclines Fig. 13:4. General tectonic structure of South America (after Harrington, 1962). The area in the box is enlarged in figure 5. Estructura geotectonica dc Sud America (segihi Harrington, 1962). El area en el recorte aparece aumentada en la figura 5. 1979 CEI: PATAGONIAN HERPETOFAUNA 315 NORTHERN OR ANCIENT MAJOR ANIMAL HABITAT SOUTHERN OR SANTA CRUZ MAJOR ANIMAL HABITAT 800 KM Fig. 13:5. Location of the southern massifs and their relation to the major Patagonian herpetofaunal regions. Ubication de los macizos aust rales y su relation con las regiones herpctofaunisticas fundamentales patagonicas. 316 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 tion) and the Meseta de Canquel, plus small- er mesetas of the same lower and middle Tertiary age along the Andean front; their small lakes are close to the last eastern patches of Nothofagus and Araucaria forests. Along the base of the Andes there is an Austral- Patagonian ecotonal zone; however, it is much narrower than the Monte-Patagonian ecotone extending from the Rio Colorado to Valcheta, onto the northern spurs of the Me- seta de Somuncura, and southward to near Bahia Camarones on the Atlantic coast (Ca- brera, 1951). Floristically, the ancient Patagonian re- gion is characterized by a predominantly steppe vegetation with scattered green bushes (Mulinum spinosum), several grasses, low herbaceous plants, some spiny plants, and a variety of low bushes; additional kinds of plants are present in saline environments and riparian situations (Table 13:1). The most characteristic element of the monte forma- tions is the creosote bush (Larrea), which is represented by five sympatric species in the ecotonal zone at Valcheta (Rio Negro). In typical Patagonian landscapes only the low Larrea ameghinoi is present. The cool Patagonian steppes dominated by Mulinum and Stipa, with scattered creeping, cushionlike plants, exist in western Neuquen, southwestern Rio Negro, and in most of Chu- but, provinces. In these areas the steppes are commonly associated with basaltic landscapes resulting from the rampant Cenozoic volcanic activity. The steppes are discontinuous in northern Neuquen and Rio Negro provinces, where they occur mostly at elevations of more than 900 to 1000 m. The southern or Santa Cruz Faunal Re- gion extends from about 45°S to the Straits of Magellan (Fig. 13:7). This region encom- passes some distinct physiographic areas. The arid valley of the Rio Deseado borders the northern limits of the large Altiplanicie Cen- tral, a dead volcanic landscape with scattered clay basins and petrified early Cenozoic trees (Auracarites). South of the great plateau the drainage basins of the rios Chalia, Santa Cruz, Coyle, and Gallegos provide more moist low- lands extending to the Straits of Magellan. These rivers drain the glacial valleys of the Table 13:1. — Characteristic Types of Vegetation in the Ancient Patagonian Region. Herbs Stipa Festuca Poa Senecio filaginoides Grindelia chilocnsis Verbena ligustrina Acaena caesiritosa Shrubs Mulinum spinosum Colliguaja integerrima Bcrbcris cuncata Lijcium tenuispinosurn Anarthrophijllum rigidum Anarthrophyllum desideratum Trcvoa spinifer Prosopis patagonica Larrea ameghinoi Halophyllic Plants Atriplex Frankenia Spartina Spiny or Sclerotic Plants Chuquiraga Nassauvia Ephedra Styllingia Verbena Pantacantlia Adesmia Austrocactus Hydrophyllic Plants Juncus Carex Ranunculus Hi/psela Plagioboth rys Acaena macrostemon Caltha Cortaderia AzoreUa rugged southernmost Andean cordillera. The Andes are commonly bordered by sharp- edged basaltic mesetas having elevations of 1000 to 1500 m. Phytogeographically, the Santa Cruz Faunal Region agrees with Cabrera's ( 1951 ) Patagonian districts — Patagonico Subandino, Patagonico Central, and Golfo de San Jorge. The Patagonico Subandino includes the ba- saltic mesetas (e.g., Meseta Vizcachas, Me- seta Asador, Meseta de la Muerte, and the Meseta de Lago del Sello) and the southern humid lowlands.1 In these areas open steppe associations of Festuca monticola, Bromus macranthus, Hordeum cornosum, Agropyron magellanicum, Poa sp., and Dcschampsia sp. predominate, but some shrubs (Bcrberis cuncata, Nassauvia aculeata, or Mulinum spi- nosum) are present. Phytogeographic differences between the Sub-Andean district and the central and San Jorge districts are evident by the monotonous grasslands of Stipa humilis in the latter. The grasses are interrupted by the broad circular bushes of the blackish "mata negra" (Verbena 1 Although the Sub-Andean District is considered to be a single physiographic unit (Fig. 13:7), for pur- poses of herpetofaunal analysis, I distinguish the Humid Southern Lowlands. 1979 CEI: PATAGONIAN HERPETOFAUNA 317 Table 13:2. — Comparison of the Herpetofaunas in Ten Districts in Patagonia. ( Numbers of species is a given district are in boldface; numbers of species in common to two districts are in Roman, and the italics are Faunal Resemblance Factors [N, + N2 /2C (Duellman, 1966)]. a a o o o W c a •a o bo a ffl On o & u CO c a J3 B0 £ -OS M 3 O c 3 s o CO <1> i-i C O 0 CO 3 on c o J c o 3 O co c aj T3 _o "3 en 0) -a o BO as J 0 c o 2 c o s o BO ct! as Oh c a) jo "3 > a! J2 < a! as O u 3 < "5 CO as Monte „ . 17 0.69 17 32 2 7 3 7 4 1 5 1 4 2 1 Monte-Patagonian Ecotone Patagonian Steppe 0.12 0.29 17 12 7 8 7 3 3 Volcanic Highlands 0.14 0.2.5 0.57 25 8 6 5 3 1 Meseta de Somuncura 0.19 0.50 0.44 11 5 4 2 1 Altiplanicie Central 0.08 0.24 0.62 0.35 0.50 9 8 3 4 Coastal District 0.08 0.20 0.56 0.30 0.42 0.94 8 2 4 Humid Southern Lowlands 0.11 0.26 0.19 0.24 0.40 0.29 6 1 2 Sub-Andean Area 0.06 0.29 0.07 0.13 0.62 0.67 0.20 4 Meseta del Lago de Sello 0.44 3 tridens). Where shrubby formations occur, dominant plants are Prosopis patagonica, Lij- cium ameghinoi, Berberis cuneata, Chuquir- aga aurea and avellaneclae, Brachyclados cae- spitosus, Acantholippia seriphioides, Pleuro- fora patagonica, Ameghinoa sp., and Euphor- bia sp. The small trees, Trevoa patagonica, around the Golfo de San Jorge and in the arid valley of the Rio Deseado are the most conspicuous plants in the southern region. COMPOSITION OF THE HERPETOFAUNA The herpetofauna of the Patagonian steppe is composed of 60 species, six of which have two or more subspecies in Patagonia; the entire herpetofauna consists of 70 spe- cies and subspecies with a noticeable degree of endemism. The fauna is made up of 14 species of anurans (23.3%), one turtle (1.7%), 34 lizards (56.7%), and 11 snakes (18.3%). For purposes of discussion, the herpetofauna has been divided according to the two major faunal regions. Of the 60 Patagonian species, 56 occur in the northern or ancient Pata- gonian Region, and 13 occur in the southern or Santa Cruz Region; nine species are com- mon to the two regions (Table 13:2, Appendix 13:1). Although lizards are dominant in both regions, they comprise a much higher per- centage of the herpetofauna in the southern region (Fig. 13:8). Northern Patagonian Herpetofauna For purposes of analysis, the region has been divided into five ecophysiographic areas (Fig. 13:6) — 1) Monte associations, 2) Monte-Patagonian ecotone, 3) Patagonian steppe, 4) Volcanic highlands, and 5) Meseta de Somuncura. The distributions of the spe- cies and subspecies of amphibians and rep- tiles in these five areas are tabulated in Ap- pendix 13:1. The sole Patagonian turtle, Geo- chelone donosobarrosi, and all of the Pata- gonian anurans are in the northern region, although one species, Pleurodemu bufonina, is widely distributed in the southern region. Likewise, the single amphisbaenian and all of the snakes are in the northern region, al- though Bothrops ammodytoides enters the southern region. All of the 17 species occurring in the monte associations are among the 32 species in the Monte-Patagonian ecotone. Included in these areas are several species characteristic of, or related to species in, the more northern regions — Pampas and Chaco; this is true of 318 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 400 KM ■ ■ I HERPETOFAUNAL AND PHYSIOGRAPHIC DISTRICTS Ancient Patagonian Steppe Climax Western Volcanic Tablelands with Closed Drainages Patagonian Phytogeographical Range Austral Forest Belt Monte Fig. 13:6. Herpetofaunal and physiographic districts of the ancient Patagonian region. Distritos herpetofaunisticos ij fisiogrdficos de hi region patagdnica antigua. 1979 CEI: PATAGONIAN HERPETOFAUNA 319 Fig. 13:7. Herpetofaunal and physiographic districts of the southern Patagonian region. Distritos herpetofaunisticos y fisiograficos de la region sur-patagonica. 320 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 PERCENTAGE HERPETOLOGICAL ELEMENTS IN FAUNAS z < E Z < l (A |U < z _« K E 3 SI z < E Z < u W l|J 2 < z .'i,;'.'V.V; ■ • ■■ ANCIENT PATAGONIAN MAJOR ANIMAL HABITAT SOUTHERN MAJOR ANIMAL HABITAT Fig. 13:8. Herpetofaunal composition of the Patagonian regions. Composition herpetofaunistica de las regiones patagdnicas. the anurans Bufo arenarum and Leptodacty- lus ocellatus, the lizard Mdbuya frenata, and the snakes Elapomorphus bilineatiis, Lystro- phis semicinctus, and Micrurus frontalis ( Fig. 13:9). The analysis of latitudinal distribu- tions reveals that Geochelone donosobarrosi is a true member of the xerophytic scrub asso- ciation south of the Rio Colorado and only slightly penetrates the neighboring flats of the Pampean region. The distributions of the snake Philodryas patagoniensis and the lizard Proctotretus pectinatus extend eastward into the Pampean region (Figs. 13:10-11). On the contrary, the frog Pleurodema bufonina and the lizards Homonota danvinii and Liolaemus bibronii are characteristic Patagonian ele- ments and only enter the monte peripherally; these species ascend the Andean slopes north of the Rio Barrancas. In the Patagonian steppes, extrusive basal- tic rocks provide shelter for numerous lizards. Liolaemus elongatus is a conspicuous species in rocky areas, some of which also are in- habited by more cryptic lizards — Liolaemus ceii and L. kriegi. Isolated populations of Phymaturus patagonicus are subspecifically distinct — P. p. patagonicus in the valley of the Rio Chubut and P. p. indistinctus in the Sierra de San Bernardo (Cei and Castro, 1973) (Fig. 13:12). Liolaemus fitzingeri can- queli inhabits the rocky slopes of the Meseta de Canquel and extends eastward through the salt flats to the coast, where it meets L. f. fitzingeri, the subspecies that is common in southern Chubut and Santa Cruz. Diplolae- mus bibronii, more characteristic of the south- ern region, reaches the northern limits of its distribution at the edge of the Meseta de Somuncura (Cei, 1971b). Clay soils in the region hold water in the form of temporary ponds during the brief rainy season. These ponds and intermittent and permanent streams are the habitats and/ or breeding sites for several species of anurans, especially leptodactylids (Barrio, 1973; Cei, 1969a,b, 1970b, 1972b; Cei and Roig, 1966, 1968; Gallardo, 1970). Some spe- cies have restricted ranges; for example, Ate- lognathus solitarius is known only from Ar- royo Las Bayas, south of Pilcaniyeu, Rio Negro. Twenty-four species are known to inhabit the volcanic plateaus and extra-Andean high- lands in western Neuquen and Rio Negro provinces. Among them are four species of telmatobiine leptodactylid frogs (Fig. 13:3). The aquatic Atelognathus patagonicus is con- fined to the Laguna Blanca Basin. The semi- terrestrial Atelognathus praebasalticus is com- posed of four geographically isolated sub- species— A. p. praebasalticus at Laguna Blan- ca, A. p. agilis at Laguna Casa de Piedra, A. p. luisi at Laguna Catan Lil, and A. p. dobeslawi at Barda de Santo Tomas (Cei, 1972b). Atelognathus nitoi and Alsodes gar- gola gargola have different distributional traits. In western Rio Negro these frogs oc- cur in small Andean lakes — Laguna Verde near Cerro Blanco at about 1450 m elevation (Barrio, 1973), and Laguna Tonchek and Laguna Schmoll at 1700 to 1750 m on the slopes of Cerro Catedral near Bariloche ( Gal- lardo, 1970). Together with Alsodes gargola 1979 CEI: PATAGONIAN HERPETOFAUNA 321 Fie. 13:9. Southern limits of distribution of taxa in the monte formation in northern Patagonia. Areas of sympatry of Pleurodema bufonina and P. thaul are indicated. Limites mcridionales de distribution de taxa en la formation del monte en el norte patagonico. Se indi- ca las areas de simpatria de Pleurodema bufonina y P. thaul. neuquensis from the thermal brooks on the sandy Meseta Lonco Luan, 1500 m elevation in Neuquen (Cei, 1976), they belong to a transitional herpetofauna of the austral-Pata- gonian ecotone. Other transitional species are the frog Pleurodema thaul and two lizards, Liolaemus tenuis and Diplolaemus leopardin- us, all characteristic inhabitants of Araucaria forests (Cei, 1970a, 1974b). On the xeric Meseta de Lonco Luan, dead patches of Nothofagus and Chusquea exist near the bor- der of the Araucaria forests. Liolaemus lineo- maculatus occurs on the Meseta de Lonco Luan. In other ecotonal and subandean areas of Neuquen, Pleurodema thaul, Liolaemus chilensis, and L. buergeri are found. The lat- ter occurs sympatrically with the typical Pata- gonian lizards Liolaemus elongatus and L. kriegi. Likewise, the characteristic lizards of rocky Patagonian communities, Phymaturus palluma and P. patagonicus zapalensis occur on the basaltic plateaus. Phymaturus palluma extends northward in the Andes at elevations of 3000 to 3500 m to La Rioja and San Juan, and P. patagonicus has distinct populations (P. p. payuniae and P. p. nevadoi) to the 322 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Tentative Boundary of So. Chacoan Province or Monte *1 Tentative Distribution of . Liolaemus darwlni Tentative So. Limit of Chelonia and Colubrid Snakes of Patagonian Region 500 KM Fig. 13:10. Distribution of Liolaemus darwinii, a characteristic species of the monte formation. The south- ern edge of this formation is the southern limits of distribution of Geochelone and of colubrid snakes in north- ern Patagonia. Distribucion de Liolaemus darwinii, especie caracter islica dc la formation del monte. El limite meridional de esta formacion es el limite meridional de Geochelone y de los ofidios colubridos en la Patagonia septentrional. 1979 CEI: PATAGONIAN HERPETOFAUNA 323 casuhatiensis B Tentative Oistributio of Diplolaemus Fig. 13:11. Patterns of distribution of tropidurine iguanid lizards in Patagonia. Distribution de los saurios igudnidos tropidurinos en Patagonia. 324 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 vflfcl P. palagontcus ssp 1 pataqonicus 2 indigtinplug 3 somuncurensis 4 2apalensis 5 payuniae 6 nevadoi Tentative Distribution ot Leiosaurus belli! 500 KM Fig. 13:12. Patterns of distribution of tropidurine iguanid lizards in Patagonia. Distribution de los saurios igudnidos tropidurinos en Patagonia. 1979 CEI: PATAGONIAN HERPETOFAUNA 325 EXTRA-ANDEAN TELMATOBIINE FROGS NORTHERN MAJOR ANIMAL HABITAT LOCALITIES AND TENTATIVE DISTRIBUTIONS Ateloqnathus pataqonicus A nitoi A solitarius A reverberii Alsodes sp Somuncuria somuncurensis Fig. 13:13. Distribution of extra- Andean telmatobiine frogs in northern Patagonia. Distribution de los anuros telmatobiinos extra-andinos en la Patagonia septentrional. north in southern Mendoza Province (Cei and Castro, 1973; Cei and Roig, 1975) (Fig. 13:12). The isolated Meseta de Somuncura (150 X 80 km), with elevations to 1700 m, has some peculiar habitats. Between 800 and 1700 m, above the Monte-Patagonian ecotone, 11 species of amphibians and reptiles are known. Six of the species are widespread in Patagonia — Pleurodema bufonina, Homonota darwinii, Diplolaemus darwinii, Liolaemus bibronii, L. bendengeri, and L. rothi. Three other species and two subspecies are endemic — Phymaturus patagonicus somuncurensis and Liolaemus elongatus petrophilus are the en- demic subspecies of lizards. The endemic Liolaemus ruizleali inhabits rocky summits of the meseta at 1200 to 1700 m. The two en- demic, telmatobiine leptodactylid frogs have unique morphological and ecological traits. The monotypic Somuncuria somuncurensis lives in streams at 800 to 1000 m issuing from thermal springs on the northeastern slopes of the meseta. Inhabiting the same streams having a temperature of about 18°C is the endemic characid fish Gymnochacinus bergi. Conversely, Atelognathus reverberii, a nearly fossorial frog, inhabits the arid plateau at elevations of more than 1000 m and breeds in small temporary pools. 326 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Southern Patagonian Herpetofauna The distributions of the 11 lizards, one snake, and one frog are examined with re- spect to five ecophysiographic areas (Fig. 13:7)— 1) Altiplano Central, 2) Coastal Dis- trict, 3) Humid Southern Lowlands, 4) Sub- Andean Area, and 5) Meseta de Lago del Sello. The distributions of the species and subspecies of amphibians and reptiles in these five areas are tabulated in Appendix 13:1, and the distributional relationships between the areas are analyzed in Table 13:2. Nine species occur in the Altiplano Cen- tral; all but the frog Pleurodema bufonina also occur in the coastal district. Only six spe- cies occur in the humid southern lowlands; the southernmost frog, Pleurodema bufonina, is rather rare there. The lizard Liolaemas magellanicus, the only herpetofaunal species on Tierra del Fuego, is common on the south- ern end of the mainland. Only four species ( all lizards ) occur in the sub-Andean region. The Meseta de Lago del Sello at 47°S is nearly 50 km in diameter. On the top of the plateau, low grasses (Festuca, Poa, and Stipa) are dominant with thorny plants or cushion plants (Benthamiella azorella. Ver- bena, Senecio, Nassauvia); lichens are abun- dant in rocky areas. Owing to the proximity of the great Patagonian Ice Field, cold winds whip the plateau, even in summer. Only three iguanid lizards have been found on the pla- teau ( 1200-1600 m ) ; the two species of Lio- laemus are widespread in southern Patagonia, whereas the monotypic Vilcunia sylvanae is endemic. No amphibians have been found on the plateau; the widespread Pleurodema bufonina ascends the slopes to only 900 m. ORIGIN OF THE HERPETOFAUNA The Patagonian herpetofauna is distinc- tive in the diversity of telmatobiine leptodac- tylid frogs and tropidurine iguanid lizards. Most other species inhabiting Patagonia are members of groups that are mainly distrib- uted to the north of Patagonia. Thus, Geo- chelone, Cnemidophorus, Mabuya, Hornono- ta, Leptodactylus, Odontophrynus, Bufo are- narum, and all of the genera of snakes are more northern groups. Bufo spinidosus and Alsodes are primarily Andean groups. The former enters Patagonia in many disjunct valleys; the species does not occur on major basaltic mesetas. In Pata- gonia, Alsodes occurs only in the volcanic highlands adjacent to the Andes. Pleurodema is a primitive leptodactyline that may have originated in, and dispersed from, the austral forests (Duellman and Ve- loso, 1977); the genus has dispersed north- ward in nonforested habitats to the Carib- bean. Two species are peripheral in Pata- gonia— P. thaul in the austral forest — Pata- gonian steppe ecotone and P. nebulosa in the monte. Pleurodema bufonina is a widespread species endemic to Patagonian habitats. Al- though P. thaul and bufonina are distinctive in their morphology and behavior, popula- tions intermediate between the species exist in high valleys in western Neuquen. Lynch (1978) provided an hypothesis for the evolution of lower telmatobiine frogs in Patagonia. The peculiar monotypic Somun- curia is endemic to the Meseta de Somun- cura, whereas five species of Atelognathus occur in isolated basaltic areas in Patagonia. The only extra-Patagonian Atelognathus is A. grandisonae from Puerto Eden in extreme southern Chile. Among the iguanid lizards, the monotypic Vilcunia has characters of both Proctotretus and Liolaemus and is endemic to southern Patagonia (Donoso Barros and Cei, 1971). Diplolaemus and Phymaturus are fundamen- tally austral genera (Figs. 13:11-12). Cten- oblcpharis, Leiosaurus, Proctotretus, and Pris- tidactijlus are widely distributed to the north of Patagonia (Cei, 1973c,d). Two species of Proctotretus are distributed in temperate areas in southern Brasil, Uruguay, and cen- tral Argentina; P. pectinatus occurs in the monte-Patagonian ecotone. Pristidactylus fas- ciatus is primarily Patagonian, but congeners occur in disjunct Andean areas (P. scapula- tus) and in isolated extra- Andean massifs of central Argentina — Sierra Grande de Cordoba (P. achalensis) and Sierra de la Ventana (P. casuhatiensis) (Gallardo, 1964, 1968; Cei and Castro, 1975). 1979 CEI: PATAGONIAN HERPETOFAUNA 327 Three major points relative to the origin and evolution of the Patagonian heipetofauna need to be emphasized. 1. There has been a radiation of primitive leptodactylid frogs, remnants of Gond- wanan elements. Vuilleumier ( 196S ) and Lynch (1971, 1978) noted the aus- tral center of radiation of telmatobiine leptodactylids in Patagonia and the austral forests; Formas (this volume) discussed the biogeography and ecol- ogy of the telmatobiines of the austral forests — Abodes, Batrachyla, Caudi- verbera, Eupsophus, Hylorina, Insueto- phrynus, and Telmatobnfo, some of which also occur in Argentina (Cei, 1978). The limited paleontological evidence supports the austral center of radiation (Schaeffer, 1949; Chaffee, 1952; Casamiquela, 1963; Estes and Reig, 1973). 2. An austral South American center of evolution and adaptive radiation of an ancestral stock of iguanid lizards is evi- dent (Cei, 1973c,d, 1975c; Cei and Castro, 1975 ) . Fourteen genera are aus- tral in Argentina and Chile; seven of these are Patagonian. There are some 20 genera of iguanids in tropical South America and another ten genera in the Sonoran region of North America. With the exception of Anolis and Sceloporus, no other iguanid genus displays such an impressive adaptive radiation as does Liolaemus, the dominant lizards in any Patagonian community. In most of these same communities there exist rep- resentatives of the other Patagonian iguanid genera — Diplolaemus, Leiosau- rus, Phymaturus, Pristidactylus, Procto- tretus, and Vilcunia. 3. An impressive post-Pleistocene adaptive radiation has taken place in four groups of Patagonian Liolaemus (see following section ) . Thus, far from being a totally barren re- gion biologically, irregularly colonized by ele- ments from neighboring biotas, Patagonia has been, and apparently still is, a center of active speciation of several hcrpetofaunal elements. The old radiations are supported by the scat- tered relicts, unique witnesses to some of the most ancient steps in the history of continental vertebrates, whereas the Recent speciation of Liolaemus attests to the continued evolu- tionary activity in the region. Evolutionary Radiation of Patagonian Liolaemus Lizards of the genus Liolaemus are wide- spread in temperate South America. Three species range into southern Rrasil, and several species occur in the Andes, two extending northward to central Peru (Duellman, this volume). Twenty-six taxa are Patagonian or Andean-Patagonian. Four major evolutionary units can be rec- ognized among the Patagonian Liolaemus as follows: 1) L. fitzingeri complex, 2) L. elon- gatus-kriegi complex, 3) L. kingii-archeforus complex, and 4) L. magellanicus-lineomacu- latus complex (Cei, 1971a, 1972a, 1973a,b 1974a, 1975a,b, 1975d,e; Cei and Scolaro, 1977; Scolaro and Cei, 1977). The lizards that are not members of these groups are pri- marily peripheral to Patagonia and/ or are ecotonal elements. Some of these are mem- bers of transcordilleran groups — Liolaemus altissimus, chilensis, cyanogaster, lemniscatus, pictus, and tenuis; L. bibronii is related to the Chilean L. fuscus. Liolaemus boidengeri, darwinii, and gracilis are members of the more northern monte fauna and are primarily peripheral in Patagonia (Fig. 13:10). Liolaemus fitzingeri complex. — This group is characterized by 1) patch of enlarged scales on posterior surfaces of thighs, espe- cially well developed in males (Fig. 13:14); 2) high number (52-82) of blunt, slightly keeled scales around body; 3) high number (7-11) of preanal pores; 4) stout body and relatively long tail, 1.5 times length of body; 5) tendency to have black venters and dark humeral collars; 6) predominate dorsal color patterns consisting of wide transverse dark blotches, bordered posteriorly by white, but spotted erythristic, and melanistic variations not uncommon. Content: L. fitzingeri canqueli, L. fitzin- geri fitzingeri, L. fitzingeri melanops, L. rothi, ?L. ruizleali. Liolaemus f. fitzingeri is the 328 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 frtzingeri Complex \UiMW 52-82 scales at the middle of the body 7-11 anal pores EVOLUTIONARY RADIATION OF PATAGONIAN Liolaemus U Fig. 13:14. Morphological traits of lizards of the Liolaemus fitzingeri complex. Caracteristicas morfologicas dc los saurios del confunto Liolaemus fitzingeri. 1979 CEI: PATAGONIAN HERPETOFAUNA 329 TENTATIVE DISTRIBUTION OF PATAGONIAN LIOLAEMUS FITZINGERI COMPLEX Tentative N. and S. Boundaries Liolaemus rothi Liolaemus ruizleali ?r£^ Endemic to Somuncura Mts. Fig. 13:15. Distribution of the Liolaemus fitzingeri complex in Patagonia. Distribution del conjunto Liolaemus fitzingeri en Patagonia. 330 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 southernmost member of the group, inhabit- ing sandy Patagonian steppe in Santa Cruz and Chubut; it meets L. f. canqueli in the vicinity of Trelew and Dos Pozos on the coast of Chubut (Cei and Scolaro, 1977). The latter subspecies is characteristic of the arid vol- canic massif of Canquel south of the Rio Chubut. Liolaemus f. melanops is distributed in coastal areas north of the Rio Chubut to the Rio Negro and thence inland through Men- doza to San Juan and La Rioja. Liolaemus rothi is endemic to nordiern Patagonia, as is L. ruizleali, known only from the Meseta de Somuncura (Fig. 13:15). Members of the Liolaemus fitzingeri com- plex are mostly stout, polymorphic lizards that are psammophilous or fossorial. The polychromatism in these lizards has been a source of confusion. Liolaemus melanops was considered to be a color variety of L. fitzin- geri by Donoso-Barros ( 1966 ) and Peters and Donoso-Barros ( 1970 ) , and as a northern sub- species of L. fitzingeri by Cei (1975d). How- ever, careful analyses of morphological and serological attributes may suggest that L. melanops probably is a distinct species. Popu- lations of L. f. melanops near Puerto Madryn are highly variable; some individuals are mor- phologically indistinguishable from L. goet- schi (Cei, 1975a), a monomorphic lizard ex- tending from the Rio Colorado north to San Juan and La Rioja. Serological analysis shows that populations formerly assigned to L. melanops and L. goetschi are conspecific; thus, only one taxon (L. melanops) is rec- ognized (Cei and Scolaro, 1977). The serological distance between L. dar- winii and members of the Liolaemus fitzin- geri complex suggests that L. darivinii di- verged early from the ancestral stock of that group. Although juveniles and females of L. darivinii and L. boulengeri are strikingly simi- lar, a noticeable serological distance exists between the species, whereas L. boulengeri is serologically closer to L. /. melanops. Liolaemus rothi has morphological char- acters that ally it with the Liolaemus fitzingeri group, but serologically it is not so distant from other Patagonian complexes of Liolae- mus as are the other members of the Liolae- mus fitzingeri group. Liolaemus rothi could be considered as a primitive, ecologically gen- eralized species of the Liolaemus fitzingeri complex. The poorly known L. ruizleali is morphologically close to L. rothi, except that the postfemoral enlarged scales are absent in L. ruizleali. Enlarged scales on the posterior surfaces of the thighs are characteristic of some other extra-Patagonian Liolaemus. Such is the case in the Andean L. ornatus and L. mocquardi, which morphologically are similar to L. dar- winii. The character also is present in L. wiegmannii in southeastern Brasil, Uruguay, and the Argentine pampas, and in L. multi- maculatus from the Atlantic coast of Buenos Aires. Because of the many differences dis- played by these two lizards from one another and from members of the Liolaemus fitzingeri complex, the enlarged postfemoral scales are considered to be independently evolved char- acters in these three lines. Liolaemus elongatus-kriegi complex. — This group is characterized by 1) no patch of enlarged scales, but a row of projecting scales on posterior surface of thigh; 2) high number (72-120) of acuminate, keeled scales around body (Fig. 13:16); 3) few (1-4) preanal pores; 4) slender body with very long tail; 5) absence of ventral melanism and dark nuchal collar; 6) dorsal pattern of blackish irregular stripes, not bordered by white, and confluent into vertebral and lateral bands. Content: L. austromendocinus, L. buer- geri, L. ceii, L. elongatus elongatus, L. elon- gatus petrophilus, L. kriegi. Liolaemus elon- gatus is widespread in rocky habitats in Chubut, Rio Negro, and Neuquen, and north- ward in the precordillera in Mendoza (Fig. 13:17). It is a highly variable species, and notable serological distances have been found among scattered, isolated populations (Cei, 1974a); only the population of the Meseta de Somuncura has been recognized taxonomi- cally — L. e. petrophilus. Liolaemus austro- mendocinus occurs in arid habitats below 1500 m in volcanic regions in southern Men- doza and in the Rio Neuquen and Rio Colo- rado basins. Liolaemus kriegi occupies ba- saltic areas in Neuquen and Rio Negro, where it occurs sympatrically with L. ceii. Liolae- mus buergeri occurs sympatrically with L. 1979 CEI: PATAGONIAN HERPETOFAUNA 331 elongatus -kriegi Complex 72-120 scales at the middle of the body anal pores i EVOLUTIONARY RADIATION OF PATAGONIAN Liolaemus II Fig. 13:16. Morphological traits of lizards of the Liolaemus elongatus-hriegi complex. Caracteristicas morfologicas de los saurios del conjunto Liolaemus elongatus-kriegi. 332 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 500 KM TENTATIVE DISTRIBUTION OF PATAGONIAN LIOLAEMUS SPECIES GROUPS pr-vy.'j; ;•,;;:' elongatus-krieqi Complex archeforus-kingii Complex maoellanicus-lineomaculalus Complex Fie. 13:17. Distributions of three Liolacmus species complexes in Patagonia. Distribution dc trcs conjuntos especificos de Liolaemus en Patagonia. 1979 CEI: PATAGONIAN HERPETOFAUNA 333 archeforus-kingjj Complex 58-84 scales at the middle of the body 6-10 anal pores EVOLUTIONARY RADIATION OF PATAGONIAN Liolaemus III Fig. 13:18. Morphological traits of lizards of the Liolaemus archeforus-kingii complex. Caracteristicas morfologicas de los saurios del conjunto Liolaemus archeforus-kingii. 334 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 elongatus or austromendodnus in Patagonian associations in southern Mendoza and with L. elongatus or kriegi in valleys in Neuquen. Liolaemus buergeri and kriegi occur in Pata- gonian habitats in Chile; presumably the transcordilleran migration was via Paso Pe- huenche (2500 m, 36°S). Several immunological cross-reactions among allopatric and sympatric populations of L. austromendocinus and elongatus show equally good specific levels of differentiation as among L. elongatus, kriegi, and buergeri (Cei, 1974a, 1975b). Liolaemus kingii-archeforus complex. — This group is characterized by 1) no patch of enlarged scales on the posterior surface of the thigh (Fig. 13:18); 2) moderately high number (58-84) of faintly keeled scales around body; 3) high number (6-10) of pre- anal pores; 4) short legs and relatively short tail, only slightly longer than body; 5) ven- ter with dark spots; absence of dark nuchal collar; 6) a series of yellowish or whitish transverse bars on the dark dorsal ground color. Contents: L. archeforus archcforus, L. archcforus sarmientoi, L. kingii. This group is endemic to the southern faunal region (Fig. 13:17). Liolaemus a. archcforus occurs on the isolated Meseta de Lago del Sello; it is replaced by L. a. sarmientoi at lower eleva- tions eastward in the Patagonian steppe be- tween the Rio Coyle and the Rio Gallegos. Liolaemus kingii, which lies at a moderate serological distance from L. archcforus, is a rather stout, apparent ecological generalist inhabiting ravines and open bushy habitats in most of Santa Cruz. It reaches the Atlantic coast, and in the western part of its range is broadly sympatric with L. archeforus. Liolaemus magellanicus-lineomaculatus complex. — This group is characterized by 1) no patch of enlarged scales on the pos- terior surface of thigh; 2) low number (40- 70 ) of large, mucronate, acuminate ( dorsally ) scales around body; 3) moderate number (3-8) of preanal pores; 4) very short limbs and tail (Fig. 13:19); 5) absence of ventral melanism and dark nuchal collar; 6) dorsum irregularly spotted with black and having whitish longitudinal lines. Content: L. lineomaculatus, L. magcllani- cus. The latter occurs in the humid southern lowlands on the mainland and in isolated populations on Tierra del Fuego, whereas L. lineomaculatus is found in the volcanic high- lands in the ancient Patagonian region, where it inhabits open formations and Araucaria woodlands (Fig. 13:17). Occasional immacu- late individuals of L. lineomaculatus are known (Cei, 1971a). These last two complexes of Liolaemus are limited to austral Patagonia and are conserva- tive in their diversity, as compared to the Liolaemus jitzingcri and elongatus-kriegi complexes, both of which apparently have undergone recent (post-Pleistocene) specia- tion. The results of these radiations are nu- merous morphologically similar species differ- ing from one another biochemically and eco- logically. ACKNOWLEDGMENTS In acknowledge the special interest and efforts of William E. Duellman in rewriting the original English draft of this manuscript and in his critical analysis of the distribu- tional data. I also thank John D. Lynch for information about his research on the mor- phology and relationships of telmatobiine frogs in the Patagonian Region. RESUMEN Riologicamente, la Patagonia se define como la region al este de los Andes, extendi- endose hasta el Oceano Atlantico, hacia el sur hasta el Estrecho de Magallanes; en el norte hay una zona de transicion entre la biota patagonica y las del norte, entre los rios Negro y Colorado. La Patagonia cs una region de suelos de rocas sedimentarias y mesetas de rocas efu- sivas, presentando severas sequias estacionales con cinco meses de invierno frio, veranos usualmente secos y clima fresco. En contraste con las asociaciones de estepa uniformes que cxisten alii actualmentc, una vegctacion tropical y subtropical occurio al mismo tiempo que bosques xerofiticos y bos- 1979 CEI: PATAGONIAN HERPETOFAUNA 335 magellanicus-lineomaculatus Complex 40-70 scales at the middle of the body LJK anal pores EVOLUTIONARY RADIATION OF PATAGONIAN Liolaemus IV plex. Fig. 13:19. Morphological traits of lizards of the Liolaemus magellanicus-lineomaculatus com- Caracteristicas morfologicas de los saurios del conjunto Liolaemus magellanicus-lineomaculatus. ques mesofiticos de galena durante el Oligo- ceno y el Mioceno. Bosques de Nothofagus existian durante el Eoceno y el Oligoceno. A mediados del Terciario los climas se vol- vieron mas secos dando lugar a la expansion de la vegetation xerofitica. En los depositos del Oligoceno y del Mio- ceno telmatobidos primitivos, ceratofrinidos, y bufonidos son conocidos, asi como boideos primitivos, cocodrilos, y tortugas meiolanidas y pelomedusidas estan representados en los depositos del Cretaceo superior y del Ceno- 336 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 zoico inferior. Los fosiles existentes, la pre- serve distribucion y los patrones de especia- cion indican que la Patagonia ha sido una region importante para la evolucion de la herpetofauna austral. Dos regiones faunisticas se reconocen en la Patagonia — 1) la region del norte, o Pata- gonia antigua, 2) la region del sur, o de Santa Cruz. El limite entre estas dos regiones se encuentra aproximadamente en el Rio Chu- but, 45°S (Fig. 13:3). La herpetofauna patagonica esta compu- esta de 60 especies: 14 anuros, 1 tortuga, 34 saurios, 11 ofidios. Hay un grado notable de endemismo. De las 60 especies, 56 ocurren el la region del norte, 13 especies en la region del sur, y nueve especies estan representadas en ambas regiones. En la region del norte se encuentra la unica tortuga y el unico am- phisbenido y todas las especies de ofidios y anuros, excepto una especie de rana (Plcuro- dema bufonina) y una especie de serpiente (Bothrops ammodijtoides) que entran en la region del sur. El grado de endemismo es alto, especial- mente en los altiplanos volcanicos, en las estribaciones de los Andes y en las mesetas aisladas. Los generos monotipicos Somun- curia y Vdcunia son endemicos de las mesetas asi como las especies y/o subespecies de Al- sodes, Atelognathus, Liolaemus, y Phijma- turus. En la Patagonia ha habido temprana ex- pansion adaptativa de los telmatobidos de la familia Leptodactylidae. La region tambien fue el centro de evolucion de las diversas lineas de iguanidos, y actualmente es un cen- tro de especiaci6n de Liolaemus. Cuatro gru- pos de especies de Liolaemus tienen su centro de dispersion en la Patagonia. Estos han sufrido especiacion post-pleistocenica y suces- iva dispersion en la region, con el resultado de muchas lineas de especies afines bio- quimica y ecologicainente bien definidas. LITERATURE CITED Bakhio, A. 1973. Una nueva especie de Telmatobius (Anura, Leptodactylidae) procedente del dominio austral cordillerano Argentino. Physis 32C:207- 213. Berry, E. W. 1938. Tertiary flora from the Rio Pichi Leufu, Argentina. Geol. Soc. Amer. Spec. Pap. (12):1-149. Bogart, J. P. 1970. Systematic problems in the am- phibians family Leptodactylidae (Anura) as indi- cated by karyotypic analysis. Cytogenetics 9: 369-383. Cabrera, A. L. 1951. Territorios fitogeognificos de la Republiea Argentina. Bol. Soc. Argentina Bot. 4:21-65. Casamiquela, R. M. 1963. Sobre un par de anuros del Mioceno de Rio Negro (Patagonia). Wawelia gerholdi n. gen. et sp. (Ceratophrydidae) y Gi- gantobatrachus parodi (Leptodactylidae). Ame- ghiniana 3:141-160. Cei, J. M. 1969a. La meseta basaltiea de Somun- cura, Rio Negro. Herpetofauna endemica y sus peculiares equilibrios biocenoticos. Physis 28: 257-271. Cei, J. M. 1969b. The Patagonian telmatobiid fauna of the volcanic Somuncura Plateau. J. Herpetol. 3:1-18. Cei, J. M. 1970a. Fluctuaciones biocenoticas y re- Iictos herpetologicos de la planicie de Lonco- Luan (Neuquen). Acta Zool. Lilloana 27:193- 200. Cei, J. M. 1970b. Telmatobius solitarius n. sp., a new rare telmatobiid frog from the highland Patagonian territories ( Rio Negro, Argentina ) . Herpetologica 26:18-23. Cei, J. M. 1971a. Herpetologia Patagonica — I. Lio- laemus del grupo magellanicus. Caracteristicas taxonomicas y geneticas. Physis 30:417—424. Cei, J. M. 1971b. Herpetologia Patagonica — -II. Notas sobre la distribucion geografica del genero Diplolaemus. Ibid. 30:471-474. Cei, J. M. 1972a. Herpetologia Patagonica — III. Re- laciones de afinidad seroproteinicas y fileticas en el genero Liolaemus. Ibid. 31:411-422. Cei, J. M. 1972b. Herpetologia Patagonica — V. Las especies extracordilleranas alto-Patagonicas del genero Telmatobius. Ibid. 31:431-449. Cei, J. M. 1973a. Herpetologia Patagonica — VI. Los Liolaemus del grupo fitzingcri en Santa Cruz y Chubut (Sauria, Iguanidae). Ibid. 32C:447- 458. Cei, J. M. 1973b. Herpetologia Patagonica — VII. Notas ecologicas y morfologicas sobre Liolaemus bibroni y L. boulengcri. Ibid. 32C: 459-469. Cei, J. M. 1973c. Comentarios sobre algunos generos de iguanidos: Diplolaemus, Leiosaurus, Apero- pristis y Cupriguanus. Ibid. 32C:269-276. Cei, J. M. 1973d. Distribucion geografica y carac- teres poblacionales de Cupriguanus fasciatus (D'Orbigny) (Sauria, Iguanidae). Ibid. 32C: 255-262. Cei, J. M. 1974a. Revision of the Patagonian lizards of the Liolaemus elongatus complex. J. Herpetol. 8:219-229. Cei, J. M. 1974b. Herpetologia Patagonica — VIII. La altiplanicie entre Primeros Pinos y Rio Kilka, Neuquen. Physis 33C: 183-185. Cei, J. M. 1975a. Herpetologia Patagonica — IX. Lio- laemus goetschi y el conjunto Liolaemus darwini- boulengeri. Ibid. 34C: 199-202. 1979 CEI: PATAGONIAN HERPETOFAUNA 337 Cei, J. M. 1975b. Herpetologia Patagonica — X. El conjunto evolutivo de Liolaemus elongatus: ana- lisis serologico. Ibid. 34C:203-208. Cei, J. M. 1975c. Herpetologia Patagonica — XI. Diferenciacion serologica de Diplolaemus dar- wirti y Diplolaemus hibroni en poblaciones alo- simpatridas. Ibid. 34C:209-210. Cei, J. M. 1975d. Liolaemus melanops Bunneister and the subspecific status of the Liolaemus fitz- ingeri group ( Sauria-Iguanidae). J. Herpetol. 9: 217-222. Cei, J. M. 1975e. Southern Patagonian lizards of the Liolaemus kingi group. Herpetologica 31: 109-116. Cei, J. M. 1976. Remarks on some Neotropical am- phibians of the genus Abodes from southern Argentina (Anura, Leptodactylidae ) . Atti Soc. Italia Sci. Nat. Mus. Civ. Stor. Nat. Milano 117: 79-84. Cei, J. M. 1977. Serological relationships of the Patagonian toad Bufo variegatus (Gunther). Serol. Mus. Bull. 52:2. Cei, J. M. 1979. Amphibians of Argentina. Monit. Zool. Italiano Monog. Zool. (in press). Cei, J. M., Castro, L. P. 1973. Taxonomic and sero- logic researches on the Phymalurus patagonicus complex. J. Herpetol. 7:237-247. Cei, J. M., Castro, L. P. 1975. A serological con- tribution to the taxonomic status of Cupriguanus, a South American genus of iguanid lizards. Serol. Mus. Bull. 51:5-6. Cei, J. M„ Roic, V. G. 1966. Caracteres biocenoticos de las lagunas basalticas del oeste de Neuquen. Bol. Est. Geog. Univ. Nac. Cuyo 13:182-201. Cei, J. M., Roic, V. G. 1968. Telmatobiinos de las lagunas basalticas de Neuquen (Anura, Lepto- dactylidae). Physis 27:265-284. Cei, J. M., Roic, V. G. 1975. A new lizard from the Sierra del Nevado Mountains, central Argentina. J. Herpetol. 9:256. Cei, J. M., Scolaro, J. A. 1977. Herpetologia Pata- gonica— XIII. La identidad de Liolaemus goet- schi y de la forma melanops del grupo Liolaemus fitzingeri, en Rio Negro y Chubut. Physis 36C: 225-226. Chaffee, R. G. 1952. The Deseadan vertebrate fauna of Scarritt Pocket, Patagonia. Bull. Amer. Mus. Nat. Hist. 98:509-562. Donoso-Barros, R. 1966. Reptiles de Chile. Ed. Univ. Chile, Santiago, 458 p. Donoso-Barros, R., Cei, J. M. 1971. New lizards from Patagonian volcanic tablelands of Argentina. J. Herpetol. 5:89-95. Duellman, W. E. 1966. The Central American herpetofauna: An ecological perspective. Copeia 1966(4) :700-719. Duellman, W. E., Veloso M., A. 1977. Phylogeny of Plcurodema (Anura: Leptodactylidae): A biogeographic model. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (64): 1-46. Estes, R., Reig, O. A. 1973. The early fossil record of frogs: A review of the evidence, pp. 11-63 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary research on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Gallardo, J. M. 1962. A proposito de Bufo varie- gatus (Gunther) sapo del bosque humedo Ant- artandico, y las otras especies de Bufo neotropi- cales. Physis 23:93-102. Gallardo, J. M. 1964. Los generos Urostrophus D. & B. y Cupriguanus gen. n. (Sauria, Iguanidae) y sus especies. Neotropica 10:125-136. Gallardo, J. M. 1968. Dos nuevas especies de Iguanidae (Sauria) de la Argentina. Ibid. 14:1-8. Gallardo, J. M. 1970. A proposito de los Telma- tobiinae (Anura, Leptodactvlidae ) patagonicos. Ibid. 16:73-85. Gasparint, Z. B., Baez, A. M. 1975. Aportes al cono- cimiento de la herpetofauna terciaria de la Argen- tina. Acta I Congr. Argentino Paleontol. Biostrat. 2:377-413. Harrington, H. J. 1962. Paleogeographic develop- ment of South America. Bull. Amer. Assoc. Petrol. Geol. 46:1773-1814. Lynch, J. D. 1971. Evolutionary relationships, oste- ology, and zoogeography of leptodactyloid frogs. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): 1-238. Lynch, J. D. 1978. A re-assessment of the telma- tobiine leptodactylid frogs of Patagonia. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (72): 1-57. Morello, J. 1958. La provincia fitogeognifica del monte. Opera Lilloana 2:5-115. Peters, J. A., Donoso-Barros, R. 1970. Catalogue of the Neotropical Squamata II. Lizards and am- phisbaenians. Bull. U.S. Natl. Mus. (297): 1-293. Ruiz Leal, A. 1972. Los confines boreal y austral de las provincias patagonicas y central respectiva- mente. Bol. Soc. Argentina Bot. 13 (Supple- ment) .89-118. Schaeffer, R. 1949. Anurans from the early Terti- ary of Patagonia. Bull. Amer. Mus. Nat. Hist. 93:47-68. Scolaro, J. A., Cei, J. M. 1977. Herpetologia Pata- gonica— XII. Los iguanidos del grupo Liolaevius fitzingeri en Chubut: datos serologicos y position taxonomica. Physis 36C:219-223. Tihen, J. A. 1962. A review of the New World fos- sil bufonids. Amer. Midi. Nat. 68:1-50. Vuilleumier, F. 1968. Origin of frogs of Pata- gonian forest. Nature 219:87-90. 338 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 APPENDIX Appendix 13:1. — Distribution of species and subspecies of amphibians and reptiles in ten districts of the two major faunal regions in Patagonia. Northern Region c B 8 w Southern Region -3 Taxon ■2 o a « ft* a a ■5 o bo -a q X o 'S o > G U o ►J 3 O C/3 o5 O O -a c < 3 t/3 Anurans Alsodcs gargola gargola ... Abodes gargola neuquensis Atelognathus nitoi _ Atelognathus patagonicus Atelognathus praebasalticus praebasalticus Atelognathus praebasalticus agilis Atelognathus praebasalticus dobeslawi Atelognathus praebasalticus luisi Atelognathus reverberii Atelognathus solitarius Leptodacttjlus ocellatus _ Odontophnjnus occidentalis Pleurodema bufonina Pleurodema nebulosa Pleurodema thaul Somuncuria somuncuriensis Bufo arenarum Bufo spimdosus Lizards Homonota darwinii ... Homonota horrida Ctenoblepharis donosobarrosi Diplolacmus bibronii Diplolacmus darwinii .. _ Diplolacmus leopardinus . Leiosaurus bellii _ — Liolaemus archeforus archeforus Liolaemus archeforus sarmientoi . Liolaemus uustromendocinus Liolaemus bibronii Liolaemus boulengeri Liolaemus buergeri ._ Liolaemus ceii Liolaemus chilensis Liolaem us darwinii Liolaemus elongatus elongatus Liolaemus elongatus petrophilus ... Liolaemus fitzingeri fitzingeri Liolaemus fitzingeri canqueli Liolaemus fitzingeri melanops Liolaemus gracilis Liolaemus kingii _ _ Liolaemus kricgi - Liolaemus lineomaculatus Liolaemus magellanicus Liolaemus rothi + - - - + - - - - + - - - - + - - - - + - - - - + - - - + - + — + + + — + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - + + + + + - . - + _ + _ + + - - + + — — — — — + + - — — — + — + - + + - + + + + + + + + - + + - - - + + + + + - - + - - + + + + + - + - - - + - + - - + - - - - + + + 1979 CEI: PATAGONIAN HERPETOFAUNA 339 Appendix 13:1 (Concluded). Northern Region m c o Southern Region Taxon c o W c a •3 o 60 a o 0) T) ft c ft RJ aj -a Z> 60 B o o a, > o O 3 X -a c < Liolaemus ruizlcali Liolaemus tenuis Phymaturus palluma _ _ Phymaturus patagonicus patagonicus Phymaturus patagonicus indistinctus Phymaturus patagonicus somuncurensis Phymaturus patagonicus zapalensis Pristidactylus fasciatus Proctotretus pectinatus Vilcunia sylvanae Mabuya frcnata Cnemidophorus longicauda Amphisbaenians Amphisbaena angustifrons Snakes Leptotyphlops australis Leptotyphlops borrichiana Elapomorphus bilineatus Leimadophis sagittifer Lystrophis semicinctus Philodryas burmeisteri Philodryas patagoniensis Philodryas psammophideus Pseudotomodon trigonatus Micrurus frontalis Bothrops ammodytoides Turtles Geochelone donosobarrosi Total Taxa Total Species - - - + - - - + - - + - - - + - - - - + + + + + - + - - + + + + + + 17 17 + 32 32 + + + - - + - - + - - + + + + + + + + + + + + + 18 17 + 29 25 11 11 + + + 14. La Herpetofauna de los Bosques Temperados de Sudamerica J. Ramon Formas Instituto de Zoologia Universidad Austral de Chile Casilla 567 Valdivia, Chile Los bosques temperados de Sudamerica, ubicados en el extremo sur de Chile y partes adyaeentes de Argentina, se caracterizan por tener pocos taxa de anfibios (Vellard, 1957; Cei, 1962a; Darlington, 1965; Vuilleumier, 196S) y reptiles (Heimlich, 1934, 1937; Do- noso-Barros, 1960). Estos ambientes boscosos temperados, aislados en el norte por la estepa semiarida de Acacia caven y por este por la estepa fria patagonica, presentan once gen- eros de anuros, dos de saurios (Liolaemus y Pristidactylus) y dos de serpientes (Alsophis y Tachymenis). Entre los anuros se encuen- tran muchos endemismos ( Caudiverbera, Tel- matobufo, Hylorina, Eupsophus, Batrachyla, Insiietophrynus y Rhinoderma) y solamente Alsodes, Atelognathus, Pleurodema y Bufo exceden los limites del bosque. Existe aqui una familia monotipica (Rhinodermatidae) y tres generos con una sola especie (Caudiver- bera, Hylorina e Insiietophrynus). La may- oria de los generos de anfibios poseen dos o tres especies (Telmatobufo, Batrachyla) y solamente los saurios del genero Liolaemus son las que presentan la mayor diversificaeion (cinco especies) en el area. Algunas de las especies existentes en el bosque temperado austral muestran notables adaptaciones a este biotopo, las cuales se ob- servan especialmente durante la reproduction y el desarrollo. Entre los anuros, destacan el cuidado parenteral de Rhinoderma, las pos- turas en terreno vegetal humedo de las espe- cies de Batrachyla y los renacuajos de los arroyos de montana de Telmatobufo australis. La viviparidad aparece como la adaptation reproductiva mas frecuente entre los reptiles (Liolaemus cyanogaster, L. pictus y Tachy- menis chilensis). Desde el punto de vista historico, algunos anuros (Caudiverbera y Eupsophus) tienen una antigiiedad que se remonta hasta el Ter- ciario (Shaeffer, 1949). Los endemismos, la pobreza de especies, la escasa diversificaeion de los generos, las adaptaciones reproductivas y la antigiiedad de algunos taxa, han sugerido diversas inter- pretaciones sobre el origen de los batracios en el bosque temperado sudamericano. Dar- lington (1965) considera a esta batracofauna como empobrecida y derivada de otras de amplia distribucion en Sudamerica. Vellard (1957), Cei (1962a) y Vuilleumier (1968) proponen que la fauna de batracios australes esta compuesta por generos endemicos de probable origen Terciario y otros secundaria- mente emigrados a la region. Para los rep- tiles, Donoso-Barros ( 1966 ) postula generos de origen septentrional (Liolaemus) y relictos de las selvas del Terciario (Pristidactylus). En base a los antecedentes ecologicos e historicos de la region y de la sistematica, ecologia y distribucion de la herpetofauna se propone una hipotesis acerca del origen de los anfibios y reptiles que habitan los bosques temperados de Sudamerica. CARACTERISTICAS DEL AREA Los bosques temperados de Sudamerica se ubican especialmente al suroeste de la Cordillera de los Andes ocupando una franja de territorio chileno comprendida entre los 37° y 55°S de latitud sur ( Cerceau-Larrival, 1968). Entre los paralelos 35 y 37, el bosque se desplaza levemente hacia el oriente pene- trando en Argentina. Desde el punto de vista ecologico, estos biotopos boscosos estan ais- 341 342 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 lados del resto del continente sudaniericano por estepas aridas o semiaridas. En el Valle Central chileno, al norte del paralelo 37, existe una estepa de marcadas condiciones xero- fiticas en la cual predominan los matorrales de Acacia caven ( Papilonaceae ) (Mann, 1960). Este ambiente semiarido es una zona de transition entre los desiertos costeros de Sudamerica y los bosques temperados aus- trales. Por el oriente, los biotopos boscosos limitan con la estepa fria de la Patagonia. Alii predominan las asociaciones de gra- mineas (Stipa, Poa), compuestas (Chuqui- raga, CoIUguaja) y matorrales con plantas del genero Trevoa (Rhamnaceae) (Solbrig, 1976). En la figura 14:1 se muestra la ubica- cion del bosque temperado en el continente sudaniericano y los biotopos que lo circundan. En el sur de Chile existen tres caracteres fisiograficos : la Cordillera de los Andes, la Cordillera de la Costa y el Valle Central. Estos dos ultimos caracteres se aprecian mar- cadamente hasta el paralelo 41; desde alii al sur tienden a desaparecer siendo reempla- zados por una intrincada geografia compuesta de islas, arehipielagos, peninsulas y fiordos (Region de los Canales). En la figura 14:2 se indican las caracteristicas fisiograficas del area cubierta por los bosques australes tem- perados en el sur de Chile. El factor mas relevante de los Andes de esta region, es el vulcanismo extrusivo del Cuaternario (Rriiggen, 1950) y las alturas aqui predominantes son los volcanes. Estos nunca bajan de los 2000 m y en algunos casos sobrepasan los 3000 m. Al sur del paralelo 37, limite norte de los bosques temperados, la altura de la Cordillera de los Andes decrece en relation con los sectores del centro y norte de Chile. Es asi que en esas areas, alcanza alturas promedios de 5000 m mientras que en el sur nunca sube de los 3000 m. Entre los paralelos 37 y 42 la actividad volcanica cuaternaria origino rocas igneas tales como basaltos, andesitas y andesitas basalticas. Tambien se pueden encontrar alii rocas sedi- mentarias correspondientes al Terciario y Cretacico continental, Jurasico Triasico y Paleozoico (Fuenzalida, 1965a; Murioz Cris- ti, 1973). Al sur del paralelo 42 predominan 76 Estepa de Acacia caven < 3" . •c •» * 0) tv a e -*2 ^l 0) «4I 3 0) ig o tTv CO M Fig. 14:1. Ubicacion del bosque temperado (ne- gro) y biotopos que lo eircundan. Map of the temperate austral forests (black) and neighboring biotopes. 1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 343 Fie. 14:2. Caracteres fisiogrdficos del area cubierta por los bosques australcs de Sudamerica. Physiography of the area covered by the austral forest of South America. granites, dioritas y granidioritas las cuales se originaron a craves del vulcanismo intrusivo (Ruiz et al., 1965). La action dc los hielos ha sido un factor muy importance en el mo- delado del macizo andino de esta region. Hoy existen gran cantidad de glaciates, los cuales en la mayoria de los casos no salen de la Cordillera de los Andes, pero al sur de los 45°S algunos llegan hasta el nivel del mar (San Rafael; 46°40'S) (Lliboutry, 1956). Fuera de los glaciares existen dos grandes masas de hielo continental dcpositadas en la Cordillera de los Andes de las provincias de Aysen y Magallanes. La primera de ellas se ubica en los 47°S y la segunda de mayor longitud cubre una distancia comprendida entre los 48° 10' v los 52°30'S (Lliboutry, 1956). El Valle Central, ubicado entre la Cor- dillera de los Andes y la Cordillera de la Costa, es un rasgo fisiografico del centro y sur de Chile. Esta larga depresion ubicada entre los 37° y 42°S tiene origen tectonico y se formo durante el Plioceno (Briiggen, 1950). La superficie de este gran valle longitudinal, que no alcanza mas de 250 m de altera y 90 kms de ancho promedio, ha sido rellenada por depositos de origen glaciar, fluvial y la- custre (Briiggen, 1950; Fuenzalida, 1965a). El Valle Central llega por el sur hasta el paralelo 41, alii se hunde en el mar para aparecer nuevamente en todo el sector occi- dental de la Isla de Chiloe. Hacia el sur de esta isla, desaparece definitivamente bajo el oceano en la region del Archipielago de los Chonos. La Cordillera de la Costa es un caracter fisiografico que se encuentra solamente en el territorio de Chile. En la region de los bosques temperados, este macizo costero ti- ende a presentarse fragmentado no alcan- zando alturas superiores a los 1500 m (Cor- dillera de Nahuelbuta). Al sur de la ciudad de Valdivia (40°S) la cordillera costera se levanta sobre los 1000 m y constituye alii la llamada Cordillera Pelada que envia sus cor- dones hasta la ciudad de Maullin (41°30'S). Desde alii hacia el sur desaparece bajo el Canal de Chacao para reaparecer en la Isla de Chiloe. El macizo costero desaparece al sur de esta gran isla, pero sus ultimos restos 344 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Tadla 14:1. — Temperatura Media, Humedad Relativa y Precipitaciones de Diferentes Puntos del sur de Chile (segun Hajek y Di Castri, 1975). Lugar Ubicacion Punta Lavapie 37°08'S-37°35'W Contulmo 38°02'S-73°12'W Valdivia 39°48'S-73° 14'W Pto. Montt 41°28'S-72°57'W Castro 42°29'S-73°48'W Melinka 43°54'S-73°46'W Pto. Aysen 45°24'S-72°42'W San Pedro 47°43'S-74°55'W Pto. Eden 49°08'S-74°25'W San Isidro 53°47'S-70°58'W Navarino 55° 10'S-67°30'W Temperatura Media °C Humedad Relativa Precipita- 13.3 82 803.9 12.6 82 1896.0 11.9 83 2348.7 11.2 85 2341.8 11.6 82 1598.5 10.0 3137.7 9.0 86 2940.0 8.2 91 4266.3 7.2 84 2343.1 5.9 81 848.5 5.9 84 540.8 se aprecian con claridad en la peninsula de Taitao (46°30'S). Las rocas metamorficas constituyen parte importante de la Cordillera de la Costa y es asi que se encuentran mica- citas, cuarcitas y filitas ( Munoz Cristi, 1973). Toda la region de los bosques temperados esta atravesada por rios medianos de caracter exorreico, que se originan en el derretimiento de las nieves de los Andes (Bio-Bio, Tolten, Imperial) o tienen un regimen mixto (nieve y lluvia). Dentro de este ultimo tipo se en- cuentran las hoyas de los rios Valdivia, Bueno y Maullin, que incluyen en su recorrido la entrada y salida por grandes sistemas lacus- tres (Fuenzalida, 1965b). Fuera de los rios, existe un gran sistema de lagos entre los paralelos 39 y 41. La mayoria de ellos son de tipo oligotrofico ( Thomasson, 1963 ) y en muchos casos ocupan cuencas excavadas por los glaciares (Arenas, 1972). Este autor cita la presencia de morre- nas terminales en los sectores occidentales de los lagos Calafquen, Rinihue y Panguipulli. Cinturones morrenicos han sido descritos para el lago Llanquihue (Bruggen, 1950) y Ranco ( Mercer y Laugenie, 1973 ) . El mayor de los lagos de esta region es el Llanquihue (Pro- vincia de Llanquihue) con 351 km- de super- ficie y uno de los menores es el Caburga ( Pro- vincia de Cautin) con 53 km- (ENDESA, 1972). Algunos de estos euerpos de agua son muy profundos y Arenas ( 1972 ) detecto 320 mctros para el lago Rinihue (Provincia de Valdivia ) . El clima de la region cubierta por los bosques temperados se caracteriza por ser frio y humedo. En la Tabla 14:1 se muestran las caracteristicas climaticas de diferentes puntos del sur de Chile. En general se observa un decremento de la temperatura en direction al sur y un aumento de las precipitaciones en el mismo sentido. La region de los bosques australes sudamericanos es azotada por fuer- tes tormentas las cuales son muy frecuentes en invierno. Los vientos trios y humedos del oeste, originados en el anticiclon del Pacifico (35°S; 100°W), son los causantes de la lluvia y la humedad del sur de Chile (Fuenzalida, 1965c). La influencia de algunos caracteres fisiograficos locales especialmente la Cordil- lera de la Costa, determinan en el Valle Cen- tral algunas condiciones de mediterraneidad. Por otro lado, no se debe olvidar que ningun lugar del sur de Chile se encuentra muy lejos del mar y que por lo tanto el oceano tiene in- fluencia en el clima. Di Castri (1968) indica que el sur de Chile tiene un clima con influ- encia mediterranea y maritima. Fuenzalida ( 1965c ) usando el sistema de Koppen ( 194S ) divide el sur de Chile en diferentes zonas climaticas templadas. En el Valle Central se presenta un clima templado de verano seco y corta estacion de sequia (Csb2), que se extiende entre los paralelos 35 y 39. Al sur de este punto, y en el mismo Valle Central, hay un clima templado humedo de verano fresco y tendencia a seco (Cfsb) que se ex- tiende hasta el paralelo 42. Toda la zona costera comprendida entre Conception y la Isla de Chiloe se caracteriza por tener un clima templado humedo de verano fresco (Cfb). Estas condiciones climaticas se ex- 1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 345 tienden a Chiloe continental, archipielago de los Chonos y Peninsula de Taitao. A] sur del golfo de Penas (47°S) y en toda la region de los canales, hasta el paralelo 53, existe una zona de clima templado humedo de verano fresco o frio (Cfc). Los basques temperados del sur de Chile, que cubren toda el area anteriormente de- scrita, se caracterizan por tener rasgos higro- morficos. Estos se acentuan a partir del pa- ralelo 38 y alcanzan un maximo desarrollo en los 45°S. Desde alii hacia el sur, hay una marcada tendencia al xeromorfismo debido a las bajas temperaturas y a los vientos pre- dominantes del oeste. Los bosques australes son densos, siempre verdes y alcanzan alturas que superan los 40 m. En el, hay varios estra- tos vegetacionales con un tupido sotobosque y un piso rico en vegetacion. En este bosque es posible encontrar arbustos con hojas an- chas, ya sea de tipo magnolia (Drimys) o laurel ( Lanrelia ) . Los troncos de los grandes arboles estan cubiertos de enredaderas, mus- gos, helechos y liquenes. La abundancia de vegetacion determina que la obscuridad sea un caracter predominante dentro del bosque. La flora del bosque austral tiene varias especies endemicas entre las cuales destacan los arboles del genero Nothofagus (Faga- ceae). Es frecuente tambien encontrar taxa monotipicos, ya sea a nivel familiar o gene- rico. Entre los primeros destaca la familia Aextocicaceae (Aextoxicum punctatus) y en- tre los segundos los generos Guevina (Pro- teaceae), Tepitaha (Mirtaceae), Fitzroya (Cupressaceae) y Myzodemdrum (Myzoden- draceae). Floristicamente los bosques tem- perados de Sudamerica tienen un origen doble: austral y tropical (Reiche, 1937; Menendez, 1969). Como tipicos elementos australes destacan Nothofagus, Fitzroya y Araucaria y como componentes tropicales Drimys, Fuchsia y Chusquea. A pesar que el bosque austral sudameri- cano muestra cierto grado de uniformidad, es posible encontrar ciertas variaciones locales. Entre ellas, la mas notable es el llamado "Bosque Valdiviano," el cual representa la region mas caracteristica de los bosques aus- trales. Aqui se da una breve descripcion de el en base a los trabajos de Reiche ( 1934 ) , Pisano (1956), Oberdorfer (1960), Fuenza- lida (1965d) y Quintanilla (1974). El bosque valdiviano comienza en la Cor- dillera de los Andes a partir del paralelo 39, en la Cordillera de la Costa en el paralelo 40 y en el Valle Central en el paralelo 41. El limite sur no esta claramente dcfinido, pero se le puede situar entre los paralelos 43 y 44. En este bosque la humedad relativa es muy alta (84%) y la temperatura promedio anual es de 10.5°C. El verano es medianamente calido y las lluvias tienen una distribucion homogenea a traves de todo el afio. La pluvi- osidad anual fluctiia entre los 2000 y 2500 mm. La abundancia de precipitaciones, la existen- cia de suelos bien drenados con una capa freatica profunda, la gran humedad y la alta temperatura en verano, permiten el desarrollo de un bosque rico en especies. Los troncos de los arboles estan cubiertos por liquenes ( Usnea ) , musgos epifitos, enredaderas ( Sar- mentia y Luzuriaga) y lianas (Hydrangea y Cissus). Aqui existe un sotobosque denso en el cual hay bambues (Chusquea quila), ar- bustos (Lomatia, Fuchsia) y helechos (Blech- num, Lophosoria). El piso del bosque es rico en liquenes y helechos (Dryopteris, Adi- antum). Ties son los arboles mas caracter- isticos del bosque valdiviano: Nothofagus dombeyi, Eucryphia cordifolia y Aextoxicum punctatus. Otras especies importantes son aqui las coniferas, entre las cuales se pueden citar a Fitzroya, Saxogotea, Podocarpus y Pil- gerodendron. Fitzroya cupressoides es el mas alto do los arboles chilenos y alcanza alturas sobre los 55 m. Su diametro puede alcanzar a los 5 m y se la han calculado edades sobre los 2000 anos. Entre las especies secundarias se encuentran Laurelia serrata, Drimys tain- ted, Weismania trichosperma y Persea lingue. Al norte del bosque valdiviano existe un bosque caducifolio en el cual destacan como especies mas relevantes Nothofagus obliqua y Guevina avellano. En la cordillera de Na- huelbuta (Cordillera de la Costa) y en los Andes, entre los paralelos 37 y 40, se desarro- llan bosques de Araucaria araucaria. Estas formaciones boscosas, ubicadas entre los 1300 y 2000 m, presentan tambien Nothofagus pu- milio y Nothofagus antarctica. Al sur del bosque valdiviano hay una selva 346 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 norpatagonica que se extiende hasta el para- lelo 48. Alii destacan como arboles mas im- portantes Podocarpus salignus y Pilgeroden- dron uviferum. AI oeste de estas formaciones boscosas hay pantanos, los cuales se ubican especialmente en las islas. Aqui hay Perneth- ija (Ericaceae), Gleichnia ( Pteridofita ) , pero tambien Nothofagus betuloides y Nothofagus antarctica. HISTORIA DEL AREA El Terciario sudamericano se caracteriza especialmente por el levantaniiento de la Cor- dillera de los Andes. En el Cretacico Superior se hicieron sentir, en el oeste del continente, los primeros movimientos orogenicos del 11a- mado ciclo Andino (Harrington, 1962) que originaron las diversas partes de la Cordillera de los Andes. Durante el Eoceno, se aprecia un aceleramiento de los procesos orogeneticos que alcanzan gran desarrollo en el Mioceno; seguidos en el Plioceno de movimientos que dieron origen a la forma actual del macizo andino (Harrington, 1962; Haffer, 1970). Al termino del Terciario Inferior (Eo- ceno), y posiblemente en el Oligoceno, el ter- ritorio de Chile fue un area inestable. Du- rante el Oligoceno se produjeron procesos de deformacion y plegamiento en varias regiones del pais, al final de las cuales el territorio adquirio las caracteristicas de una region estable, en el que aparecieron sistemas de montaiias de poco relieve (Fuenzalida, 1965a). Durante el Mioceno hubo una gran transgresion marina que cubrio extensas areas del sur de Chile (Cecioni, 1970). A fines del Plioceno o comienzos de la epoca siguiente, Pleistoceno, el territorio chileno fue profunda- mente modificado por un fuerte tectonismo. Este trajo como consecuencia el levantami- ento de la Cordillera de los Andes, de la Cordillera de la Costa y la formacion del Valle Central (Briiggen, 1950; Fuenzalida, 1965a). Durante el Pleistoceno, ocurrieron en el extremo sur de Sudamerica fuertes pro- cesos glaciares ( Vuilleumier, 1971). La inva- sion de estas masas de hielo trajo como con- secuencia fuertes modificaciones en el clima y en la fisiografia. Las islas, archipielagos y fiordos de la region de los canales, se forma- ron en gran medida por la accion del hielo glaciar (Briiggen, 1950; Fuenzalida, 1965a). Los bosques australes sudamericanos, con sus elementos tipicos (Nothofagus, Araucaria y Laurelia), han existido desde el Terciario (Jeannelle, 1967; Cerceau-Larrival, 196S) y se acepta que tuvieron una distribution gond- wanica (Couper, 1960). En el extremo sur de Sudamerica alcanzaron un rango de exten- sion mucho mas amplio que el que tienen hoy, llegando hasta la actual Patagonia (Menen- dez, 1969 ) . Durante el Eoceno muchos de los elementos de la flora tropical penetraron hacia el sur y aparecen en los estratos fosiliferos de Rio Turbio, Argentina, mezclados con ele- mentos australes (Menendez, 1969). Aqui las capas mas inferiores muestran elementos tipicamente surenos (Nothofagus) los cuales son reemplazados en los estratos superiores por elementos tropicales (Persea, Psidun). La coexistencia de una flora austral con una flora tropical se explica debido a que estos ultimos elementos ocupaban las partes bajas (valles) y los australes las partes superiores de las montanas (Briiggen, 1950; Menendez, 1969). Durante el Mioceno y Oligoceno la flora tropical retrocedio hacia el norte y simul- taneamente se produjo un avance de la flora austral en la misma direction, hasta los limites actuales del bosque temperado ( Solbrig, 1976; Menendez, 1969). La extincion de los bosques australes en la actual Patagonia se debe, en gran medida, al efecto que causo la Cordillera de los Andes al impedir la pasada de los vientos frios y humedos del oeste. Durante el Paleoceno y el Eoceno el macizo andino estaba poco le- vantado y los vientos del Pacifico llegaban hasta los sectores orientales del extremo sud- americano. Durante el Oligoceno la barrera de los Andes llego a ser un obstaculo para ellos, los cuales se vieron definitivamente fre- nados en el Mioceno. Al no haber lluvias ni humedad en el sector oriental del macizo andino, las formaeiones boscosas desaparecie- ron, dando origen a la estepa semiarida pata- gonica (Solbrig, 1976). Fuera de los cambios del relieve y la vegetation ocurrieron varia- ciones simultaneas en las condicioncs clima- ticas que afectaron el extremo austral de Sud- 1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 347 america. Durante el Paleoceno el clima del continente fue mas calido que hoy (Solbrig, 1976) y despues del Eoceno se aprecia un gradual enfriamiento y desecacion (Axelrod y Bailey, 1969; Wolfe, 1971) el cual eulmina en el Pleistoceno durante las etapas glaciares. Fuera de los cambios anteriorniente re- feridos, es posible que la transgresion marina del Mioceno y las glaciaciones del Pleistoceno hayan afectado la distribucion de la heipeto- fauna austral. La entrada miocenica del mar, afecto el sur de Chile entre los paralelos 37 y 41 (Briiggen, 1950; Cecioni, 1970; lilies, 1970; Auboin et al., 1973). Los estratos pro- dueidos por esta invasion del mar se eneuen- tran en la region de Santiago (Navidad) y hacia el sur, en las areas de Coneepcion (Ranquil); Temuco (Pilmahue); Osorno (Cheuqueno) y Chiloe. Segun Briiggen ( 1950 ) , en estos estratos hay areniscas ar- cillosas de grano fino y color gris claro que se distinguen por tener una abundancia de fosiles marinos. lilies (1970) indica que la transgresion marina tuvo baja profundidad y que como consecuencia de ella se produjeron una gran cantidad de islas y bahias que semejan los archipielagos e islas de la costa del extremo sur occidental de Sudamerica. Durante el Pleistoceno el hielo ocupo en el sur de Chile una amplia extension cubri- endo el area comprendida entre los 41° y 55°S (Vuilleumier, 1971). Sin embargo Briiggen (1948) para explicar la expansion del bosque de Nothofagus, al sur de paralelo 41, en la epoca post-glacial, propone que du- rante los periodos glaciales quedaron refugios boscosos en los faldeos de la costa del Pacifico sobre los glaciares. Condiciones parecidas a las supuestas por Briiggen se encuentran hoy en el glaciar de San Rafael. Aqui se sucedi- eron tres o cuatro glaciaciones (Briiggen, 1948; 1950; Auer, I960; Vuilleumier, 1971, Simpson; este volumen) que penetraron en el Valle Central hasta la latitud de la ciudad de Santiago (Briiggen, 1950). Este autor ha de- scrito sistemas de morrenas terminales cerca del Rio Maipo, al sur de Santiago (33°30'S), al norte de Curico (35°S) y en la vecindad de Puerto Montt (40°30'S). La presencia de morrenas en el Valle Cen- tral y en el sector oriental de la Cordillera de la Costa hacen presumir que las pendientes occidentales de este macizo costero no tu- vieron influencia glaciar. Heusser (1966) e lilies (1970) indican que la Cordillera de la Costa permanecio fuera de la action de estas masas de hielo. COMPOSICI6N DE LA HERPETOFAUNA La herpetofauna de los bosques australes esta compuesta de 28 especies, 20 de las cuales (71.4%) son anuros, seis son saurios (21.4%) y dos seqjientes (7.1%). Los anuros (sapos y ranas) pertenecen a tres familia distintas: Bufonidae, Leptodactylidae y Rhinodermatidae. Las serpientes pertenecen a la familia Colubridae y los saurios se ubi- can en la familia Iguanidae. En la Tabla 14:2 se muestra la composition herpetofaun- istica de los bosques temperados de Sud- america. DISTRIBUCION DE LA HERPETOFAUNA La herpetofauna de los bosques tempera- dos sudamericanos presenta patrones de dis- tribucion caracteristicos. Al norte del para- lelo 44 existe la mayor concentracion de gen- eros de anfibios y reptiles, los cuales a partir de esta latitud comienzan a disminuir grad- ualmente hacia el sur. La figura 14:3 muestra los patrones de distribucion generica de los anfibios y reptiles del bosque austral. En la region costera del area comprendida entre los 39°30'S y los 40°20'S existe la mayor concentracion de generos de anuros. La zona con menor concentracion de anfibios es la que se encuentra al sur del paralelo 50; lle- gando hasta alii solamente los anuros del genera Bufo. Algunos de las especies de anfibios pre- sentes en el bosque tienen amplia distribucion en el. Dentro de esta categoria se pueden in- cluir a Rhinoderma darwinii, Batrachyla lep- topus, Batrachyla taeniata, Eupsophus roseus, Eupsophus vittatus^ CaucJiverbera caudiver- bera y Pleurodema thanl. Otras especies ocu- pan rangos medianos (Hylorina sylvatica, Al- sodes monticola, Bufo variegatus, Bufo rubro- 348 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Tabla 14:2. — Composicion de la Herpetofauna del Bosque Temperado de Sudamerica. Familia Generos Espeeie y Subespecie Leptodactylidae Rhinodei matidae Bufonidae Iguanidae Colubridae Alsodes Atelognathus Batrachyla Caudivcrbera Eupsophus Insuetophrynus Pleurodema Telmatobufo Rhinodcrma Bufo Liolaemus Pristidactylus Alsophis Tachymenis monticola grandisonae antartandica leptopus taeniata caudivcrbera roseus vanzolinii vittatus(= E. vcrtebralis) migueli acarpicus thaul australis venustus(= T. bullocki) darwinii rufum chilensis ruhropunctatus variegatus chilensis cyanogaster cyanogaster cyanogaster brattstroemi monticola villaricensis /rictus pictus l>ictus cliilocnsis pictus major pictus talcanensis tenuis tenuis tenuis punctatissimus torquatus(= Cupriguanus) chamissonis chilensis (= T. peruviana) punctatus, Batrachyla antartandica, Rhinoder- ma rufum y Bufo chilensis) y unas pocas estan restringidas a ambitos muy pequenos (Ate- lognathus grandisonae, Insuetophrynus acar- picus, Telmatobufo australis, Telmatobufo venustus, Eupsophus vanzolinii y Eupsophus migueli). Las figuras 14:4-10 muestran los rangos de distribution de todas las especies de anuros presentes en el bosque temperado. Areas de simpatria han sido encontradas para algunas especies de batracios. Rhino- derma darwinii y Rhinoderma rufum super- ponen su distribucion en Chiguayante (Pro- vincia de Conception) (Formas et al., 1975). Silva et al. (1968) encontraron poblaciones simpatricas de Bufo variegatus y Bufo rubro- punctatus en la Cordillera de los Andes de la Provincia de Llanquihue. Batrachyla lepto- pus y Batrachyla antartandica tienen pobla- ciones que se superponen en el cerro Mirador (Cordillera Pelada, Provincia de Valdivia), Puerto Blest y Lago Frias (Nahuel Huapi, Argentina ) y en EI Correntoso ( Puerto Montt, Chile) (Barrio, 1967a). Batrachyla taeniata y Batrachyla leptopus viven en condiciones de simpatria en los alrededores de la ciudad de Valdivia. Eupsophus vittatus y Eupsophus roseus son simpatricas en un area muy amplia que cubre todo el rango de distribucion de Eupsophus vittatus. Las figuras 14:11-15 muestran los rangos distribucionales de los reptiles del bosque temperado de Sudamerica. Pristidactylus tor- quatus alcanza alturas que fluctuan entre los 50 m (Catamutun, Provincia de Valdivia) y los 1400 m (Cordillera dc Nahuclbuta ) . Lio- laemus monticola villaricensis tiene rangos de distribucion altitudinal que fluctua entre los 1000 m y los 1400 m (Hellmich. 1950). Liolaemus chilensis ha sido colectada en al- turas que varian entre los 100 m y 1200 m (Hellmich, 1950). Liolaemus pictus se ubica entre los 100 m y 800 in; mientras que Lio- laemus cyanogaster lo hace entre los 10 m y 1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 349 40 34 »8 52 I • • • 1 • • • I OOO" 2 ■III! 4 SW® 5 ♦♦♦++ c + + + ♦ + o 7 i 8 I 9 ho in 75 67 73 65 Fig. 14:3. Patrones de distribution latitudinal de los Renews de anfibios y reptiles. Las areas en bianco (11) corresponden a hielo continental o regiones carentes de anfibios o reptiles. Los numeros indican la den- sidad generica. Latitudinal patterns of distribution of the genera of amphibians and reptiles. The white areas (11) repre- sent ice-covered areas or areas free of amphibians or reptiles. The numbers indicate the generic density. 350 m. Ambas cspecies son simpatricas en una amplia area. Liolaemus pictus presenta tres subespecies (L. p. chiloensis, L. p. major y L. p. talcanensis) que se distribuyen en el archipielago de Chiloe ( Donoso-Barros, 1966; Urbina y Ziiniga, 1977). Liolaemus cijano- gaster tiene una subespecie (L. c. bratt- stroemi) que se distribuye en la Isla Grande de Chiloe (Donoso-Barros, 1966). Liolaemus tenuis ocupa alturas que van desde el nivel del mar hasta los 1000 m y en la region co- stera es reemplazada por una subespecie, la 350 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 75 73 71 1 1 Caudiverbera /• 1 caudiver bera f |' Telmatobufo v-J \* l venustus IU • N > ^~ ft t : ( i \ • \ * ' \ "' • < i \ 1 Telmatobufo f aus tra 1 is ^*^ ////// K t&b.^ 200 km - 37 - 41 Fig. 14:4. Patron distributional de las especies de Telmatobufo y Caudiverbera caudiverbera en el sur de Cltile. Distribution pattern of Telmatobufo and Caudi- verbera caudiverbera in southern Chile. cual Donoso-Barros (1966) llama Liolaemus tenuis punctatissimus. Tachymenis chilensis esta desde el nivel del mar hasta los valles de la Cordillera de los Andes (Hellmich, 1937). Alsophis chamis- sonis ocupa el Valle Central y puede ascender en la Cordillera de los Andes hasta los 1500 m de altura. Desde el comienzo del bosque temperado, en el paralelo 37, estan presentes los cuatro generos de reptiles existentes en el area, los cuales llegan juntos hasta el paralelo 40. Los sauries del genero Liolaemus son las que alcanzan el limite mas austral de distribution, ya que en esta area penetran hasta el para- lelo 45. Es posible que la gradiente de dis- minucion termica que existe en direction norte-sur, la cual se muestra en la Tabla 14:1, sea la responsable de este patron distribu- tional que afecta tanto a anfibios como a reptiles. Eupsophus van zo 1 1 n i i - 37 41 45 Fig. 14:5. Patron de distribucion de las especies del genero Eupsophus en el sur de Chile. Distribution pattern of the speeies of Eupsophus in southern Chile. Altitudinalmente los anfibios llegan hasta los 1000 m; sin embargo Pleuroclema thaul, Bufo variegatus y Bufo chilensis pueden al- canzar hasta los 2000 m. Es posible tambien que la temperatura sea un factor limitante en la distribucion altitudinal de los anfibios. La figura 14:16 muestra el patron de distribu- cion altitudinal de los anfibios del bosque temperado. 1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 351 75 73 71 1 I II I I ,' 1 ( Rhi noder ma r~ <=^A rufum {••• S: -^- V, - ^^^"^™ ~_ /• 1 ^-^i . X — ft' ^•v» X--~^jU • • • ? \« >. • r ~ V V 1 s Rhinoder ma da r w i n ii i ^ — ^ -1 ■^7 :\ 1 nsuet ophr y 1USF A ■ A ^ aca r pi c us I* ' I* • / ^ • > .J • | - V*/" .3d 0 ( • \ \ r X ° ]J^V)oc5 J ' T *j)SD I ' 200 km N / j i / r u ) O •• • • • / / - 37 Fig. 14:6. Patrones de distribution de las espe- cies de los generos Rhinoderma y Insuetophrynus en el sur de Chile. Distribution patterns of species of Rhinoderma and Insuetophrynus in southern Chile. ECOLOGIA DE LA HERPETOFAUNA La mayoria de las especies de anuros vive en el piso del bosque, ya sea entre la vegeta- tion (Rhinoderma, Batrachyla), la hojarasca (Eupsophus) y bajo troncos en descomposi- cion o piedras (Bujo, Abodes). Telmatobufo e Insuetophrynus esta asociados a ambientes acuaticos de tipo lotico mientras Caudiver- bera esta en cuerpos de agua de tipo lentico. A pesar de la abnndante vegetation que existe en el bosque austral no hay especies arboreas; sin embargo en forma ocasional se han encontrado a algunos individuos de Hy- lo r ina I vatica ///// 200 km 38 42 - 46 Fic. 14:7. Patrones de distribucion de Hylorina sylvatica y Alsodes monticola en el sur de Chile. Distribution patterns of Hylorina sylvatica and Alsodes monticola in southern Chile. lorina sylvatica y Batrachyla leptopus ( Busse, 1971) sobre ramas o troncos. Pleurodema thaul vive bajo troncos o piedras y tambien en lugares con fuerte intervention humana. La mayoria de los saurios del genero Lio- laemus tiene habitos trepadores y se encuen- tran especialmente en los arbustos del soto- bosque (Liolaemus pictus, Liolaemus cyano- gaster, Liolaemus chilensis y Liolaemus ten- uis). Liolaemus monticola villaricensis vive preferentement en las rocas y campos de lava de la Cordillera de los Andes (Hellmich, 1934). Pristidactylus torquatus es segun 352 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 75 Batrachyla taeniata Batrachyla leptopus T^37 Batrachy la antartandica t>£ • { 200 Km 41 Fig. 14:8. Patrones de distribution del genero Batrachyla en el sur de Chile. Distribution patterns of Batrachyla in southern Chile. Donoso-Barros (1966) un lagarto que trepa en los troncos de los arboles del genero No- thofagus. Alsophis chamissonis y Tachymenis chilensis son dos serpientes que viven en los lugares mas secos del bosque; sin embargo la ultima especie puede tambien ser encon- trada en lugares con mucha humedad. No existen antecedentes suficientes para trazar un cuadro detallado sobre la alimenta- tion de la herpetofauna del bosque austral. Los pocos datos disponibles permitcn decir solamente que no hay animales altamente especializados en la alimentation. Rhinoderma darwinii se alimenta de in- sectos (Schneider, 1930) lo mismo que Tel- matobufo venustus (Schmidt, 1952). Rybertt y Daniel (1976) determinaron que Eupso- phus vittatus y Eupsophus roseus se alimen- 75 73 71 1 1 1/! 1 1 .yJ <** [f\ \ . Pleurodema thaul C >^~ ^ ff /• / /% s\ — w jf y «i M / •» ^£jt£ 5> "V • i ft . tjo 6 K\* •• • • • • • J*. r~' t 2r\ " • > • CI •#)o<^ ^> • • / A^ ^2, ^< * • 1 / r J •V s o A • • 7t f\ ^er-^ 200 km »• °^\ of* • r • • / I i **&$:; f . — >-* - 37 - 41 Fig. 14:9. Rango de distribution de Pleurodema thaul i>n el sur de Chile. Range of distribution of Pleurodema thaul in southern Chile. tan de insectos, larvas, acaros, caracoles, es- corpiones, pseudoescorpiones y oligoquetos. Candiverbera caudiverbera come especial- mente peces, batracios, larvas de insectos, crustaceos (Aegla) y hasta pajaros y peque- fios mamiferos (Lira, 1946; Cei, 1962a). La mayoria de los lagartos que habitan el bosque temperado tiene habitos insectivoros, pero en Liolaemus monticola villaricensis se ban detactado habitos de herbivoria ( Donoso- Barros, 1966). Alsophis chamissonis se ali- menta especialmente de lagartijas de genero Liolaemus y roedores (Octodon degus y Mus 1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 353 75 73 71 ••; (1 ■ * 1 f * * liolaemus t. (r M* ■\* punctatissimus ■• r — ^v ■* / 1* \w 1» ^^ • • 1* u /■ " ■ » y — •" ■•/ • — L 1* < <^9 • 5' •i •i •> •t •V * •>> ) •' S •< \ -\ Li o 1 a em u s t. fl 1* ^, ■^-L • tenuis 11 I* F / ■* A" Li— J# to\ /♦ * n ■• ^— ^» * - I*M / — *) 200 km £ -jfa^ro > 0-* en ■» - 35 ca ca ca s s s i* u !- _ o o o O fa fa fa ca c3 c c o o b* (- OJ 0J ft ft c c -o -o ca ca __^ > > CO CD q> a; t — t"- f w OS CO -O- _Q — i ^h O O O O O O - _ fa 4) OJ ca ca D OJ 3 3 c 9 4) oj , 000 0)00 c ■ ■" °i 7 t T •- i "H l/j es fa 05 CM >. EC ; o> ca ' oj i w s 3 CD o i £9, fe- OJ C CD 1.81*00 C •-> < t o o i (N — I 1> o> ca ca cd t*- t~ •OhCDO) OJ „--H i-H OJ OJ - fa U"ca"c« B1 « I oo oo qo co 00 O O -h 2 a*S 3=55 S 5 2 S 5 2 *> ex s Is g»i = 5 2 2 a a a a~: o,-b eo -0 >~ ~ CiJ "-. ^^ ^ •^ ex ox .£; ^ -c -a 8 a a g a s> a a. q S ^ t- fc olialella gatji ( Rana chilena). Biologica 36:43-53. Jorquera, B., Pugin, E., Goicoechea, O. 1972. Tabla de desarrollo normal de Rhinoderma dar- wini. Arch. Med. Veter. 4:1-15. Jorquera, B., Pugin, E., Goicoechea, O. 1974. Tabla de desarrollo normal de Rhinoderma dar- wini ( Conception ) . Bol. Soc. Biol. Concepcion 48:127-146. Koppen, W. 1948. Climatologia. Fondo Cult. Econ. Mexico, 340 p. Kuramoto, M. 1972. Karyotypes of three leptodac- tylid frogs from Chile, with a discussion on the chromosomes numbers of the family Leptodacty- lidae. Bull. Fukuoka Univ. Education 21:133- 141. Lira, E. 1946. Limite de saciedad y sensation de replecion en Cahjptoceplialus gayi. Biologica 5: 31-42. Lynch, J. D. 1971. Evolutionary relationships, oste- ology, and zoogeography of leptodactyloid frogs. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): 1-238. Lynch, J. D. 1972. Generic partitioning of the South American leptodactylid frog genus Eupsoplius Fitzinger, 1843 (sensu lato). Bull. South. Cali- fornia Acad. Sci. 71:2-11. Lynch, J. D. 1975a. A new Chilean frog of the extra-Andean assemblage of Tclmatobius ( Am- phibia: Leptodactylidae). Ibid. 74:160-161. Lynch, J. D. 1975b. A review of the Andean lep- todactylid frog genus Phrynopus. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (35): 1-51. Lynch, J. D. 1976. The systematic significance of variation of the coccyx in some South American leptodactylid frogs. Herpetol. Rev. 7:90. (Ab- stract) Lynch, J. D. 1978. A re-assessment of the telma- tobiine leptodactylid frogs of Patagonia. Univ. Kansas Mus. Nat. Hist. Occas. Pap. (72): 1-57. Lliboutry, L. 1956. Nieves y glaciares de Chile. Ed. Univ. Chile, Santiago, 471 p. Maglio, V. J. 1970. West Indian xenodontine colu- brid snakes: Their probable origin, phylogeny, and zoogeography. Bull. Mus. Comp. Zool. Har- vard Univ. 141:1-54. Mann, G. 1960. Regiones biogeograficas de Chile. Invest. Zool. Chil. 6:15-49. Martin, R. F. 1972. Evidence from osteology, pp. 37-70 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Menendez, C. A. 1969. Die fossilen floren Sudamer- ikas, pp. 519-561 in Fitkau, E. J., Illies, J., Klinge, H, Schawe, C, Sioli, H. (eds.). Bio- geography and ecology in South America, 2. A. Junk., The Hague, 946 p. Mercer, J. H., Laugenie, C. 1973. Glacier in Chile ended a major readvancement about 36000 years ago: some global comparisons. Science 182:1017-1019. Morescalchi, A. 1973. Amphibia, pp. 233-238 in Chairelli, A. B., Capanna, E. (eds.). Cytotax- onomy and vertebrate evolution. Acad. Press. London, New York, 783 p. Muller, P. 1973. The dispersal centres of terrestrial vertebrates in the Neotropical Realm. A. Junk, The Hague, 224 p. 1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 369 Munoz-Cristi, J. 1973. Geologia de Chile. Ed. Andres Bello, Santiago, 209 p. Noble, G. K. 1931. The biology of the Amphibia. McGraw-Hill Book Co., New York, 577 p. Oberdobfer, E. 1960. Pflanzensoziologisehe Studien in Chile, ein Vergleieh mit Europa. Verlag J. Kramer, Weinheim, 208 p. Orton, G. L. 1953. The systematics of vertebrae larvae. Syst. Zool. 2:63-75. Pefaub, J. 1971. Nota sobre Tehnatobufo bullocki Schmidt (Anura, Leptodactylidae ) . Mus. Nac. Hist. Nat. (Santiago de Chile) Bol. 32:215-225. Peters, J. A., Donoso-Barros, R. 1970. Catalogue of the Neotropical Squamata: Part II. Lizards and amphisbaenians. Bull. U. S. Natl. Mus. 297: 1-293. Petebs, J., Obejas-Miranda, B. 1970. Catalogue of the Neotropical Squamata: Part I. Snakes. Ibid. 1-347. Pisano, E. 1956. Esquema de clasificacion de las comunidades vegetales de Chile. Rev. Agron. Santiago 2:30-33. Quintanilla, V. 1974. La representation carto- gnifiea preliminar de la vegetation chilena: un ensayo fitoecologico del sur de Chile. Ed. Univ. Valparaiso, 73 p. Reiche, K. 1934. Geografia botanica de Chile. Imp. Univ. Santiago, 1:1-245. Reiche, K. 1937. Geografia botanica de Chile. Imp. Univ. Santiago, 2:1-146. Reig, O. A. 1960. Las relaciones genericas del anuro chileno Calyptocephaella gayi (Dum. & Bibr.). Actas Trab. Cong. Sudamer. Zool. 4:113-131. Reig, O. A. 1972. Macrogenioglotus and the South American bunonid toads, pp. 14—36 in Blair, W. F. (ed.). Evolution in the genus Bufo. Univ. Texas Press, Austin, 459 p. Ruiz, C, Corvalan, J., Aguirre, L. 1965. Geologia. In Geografia Economica de Chile, Corfo, Santiago: 35-92. Rybertt, G., Daniel, M. V. 1976. Rol de las pobla- ciones de anfibios en la subtrama trofica del suelo en el bosque San Martin, Valdivia-Chile. Tesis Univ. Austral Chile, 28 p. Savace, J. M. 1973. The geographic distribution of frogs: Patterns and predictions, pp. 351-455 in Vial, J. L. (ed.). Evolutionary biology of the anurans: Contemporary reserch on major prob- lems. Univ. Missouri Press, Columbia, 470 p. Schaeffer, B. 1949. Anurans from the early Terti- ary of Patagonia. Bull. Amer. Mus. Nat. Hist. 93:47-68. Schmidt, K. P. 1952. A new leptodactvlid frog from Chile. Fieldiana Zool. 31:11-15. Schneider, C. O. 1930. Observaciones sobre batra- cios chilenos. Rev. Chil. Hist. Nat. 34:220-223. Silva, F., Veloso, A., Solervicens, J., Ortiz, J. C. 1968. Investigaciones Zoologicas en el Parque Nacional Vicente Perez Rosales y Zona de Pargua. Mus. Nac. Hist. Nat. (Santiago de Chile) Bol. 148:1-12. Solbrig, O. T. 1976. The origin and floristic affini- ties of the South American temperate desert and semidesert regions, pp. 7—49 in Goodall, D. W. ( ed. ) . Evolution of desert biota. Univ. Texas Press, Austin, 244 p. Tihen, J. A. 1962. Osteological observations on New World Bufo. Amer. Midi. Nat. 67:157-183. Tihen, J. A. 1965. Evolutionary trends in frogs. Amer. Zool. 5:309-318. Thomasson, K. 1963. Araucanian lakes. Acta Phy- togeogr. Suecia 47:1-139. Urbina, M., Zuniga, O. 1977. Liolaemus pictus tal- canensis nob. subsp. (Squamata-Iguanidae). Nue- vo reptil para el Archipielago de Chiloe. An. Mus. Hist. Nat. Valparaiso 10:69-74. Vellard, J. 1957. Repartition des batracien dans les Andes au sud de l'Equateur. Trav. Inst. Fran- cais Etud. Andines, Lima, 5:141-161. Venegas, W. 1975. Los cromosomas de Aruncus venustus (Philippi) 1899 (= Tehnatobufo bul- locki Schmidt, 1952) Amphibia, Anura. Bol. Soc. Biol. Concepcion 49:71-77. Wilhelm, O. G. 1927. La Rhinoderma darwinii D. & B. Ibid. 1-2:166-170. Wilhelm, O. G. 1932. Nuevas demostraciones acer- ca de la neomelia de la Rhinoderma dancini. Rev. Chilena Hist. Nat. 36:166-170. Wolfe, J. A. 1971. Tertiary climatic fluctuations and methods of analysis of Tertiary floras. Paleo- geogr. Paleoelimatol. Paleocol. 9:27-57. Vuilleumier, B. S. 1971. Pleistocene changes in the fauna and flora of South America. Science 173:771-780. Vuilleumier, F. 1968. Origin of frogs of Patagonian forests. Nature 219:87-89. 15. The Herpetofauna of the Andes: Patterns of Distribution, Origin, Differentiation, and Present Communities William E. Duellman Museum of Natural History and Department of Systematics and Ecology The University of Kansas Lawrence, Kansas 66045 USA The Andes — the longest mountain chain in the world — extend nearly -8,000 km along the northern and western edges of South America. This young mountain chain contains many active volcanoes and innumerable dormant ones. More than a dozen peaks reach heights of more than 6,000 m; only the Himalayas and Pamirs in Asia have peaks that are higher. Frequent earthquakes attest to continuing tectonic activity. Spanning 66° of latitude through the tropics and southern temperate zone and reaching to within 1,300 km of the Antarctic Circle, the Andes are a major fac- tor in the formation of climates in western South America. Blocking both easterly and westerly moisture-laden winds, the massive mountain range creates immense rain-sha- dows west of the Andes between 5° and 35°S and east of the Andes between 28° and 38°S. The eastern face of the Andes in the tropics and the western face north of the Equator and south of 37°S receive abundant rainfall. At high elevations daily temperatures vary as much as 20°C; in many areas freezing tem- peratures are a nightly occurrence. Thus, the climates and environments of the Andes are highly diverse. At lower lati- tudes, the slopes receiving moisture-laden winds are covered with lush tropical forests, which give way at higher elevations to an elfin forest of stunted trees heavily laden with thick growths of mosses. Above tree line a variety of composites, including frailejones and cushion plants are dominant life forms in the paramos. In drier areas, vegetation may be nearly absent on the slopes and pres- ent only in valleys where bunch grasses form the puna vegetation. In the extreme south the austral forests extend nearly to snow line. Permanent snow and glaciers exist on the higher peaks throughout the Andes, and the Cordillera Real extending for about 300 km in Bolivia and the Cordillera Blanca about 400 km in length in Peru, are nearly continu- ous snow-covered ranges. The melting snow and glaciers provide water for countless An- dean lakes, many of which are trapped in glacial cirques and reach gigantic proportions in Lago Titicaca (177 X 56 km, 3812 m), and myriads of small streams, some of which fed by heavy precipitation on the Andean slopes, grow and merge to form the giant tributaries of the Rio Amazonas. For more general in- formation on the Andes and excellent photo- graphs, the reader is referred to the works by Morrison ( 1974, 1975 ) ; a real appreciation for early exploration in the Andes can be gained from Whymper (1892). The complex topography and variety of environments resulting from tectonic events and climatic fluctuations in the Pleistocene and continuing to the present provide an array of habitats for a diverse Andean fauna that is far richer than one might expect. More than 700 species of amphibians and reptiles are known to inhabit the Andes. The purposes of this paper are to 1) describe the distributional patterns of the Andean herpe- tofauna, 2 ) determine the origin of the fauna, 3) hypothesize geological and climatic changes that influenced the differentiation and dispersal of the Andean herpetofauna, and 4) examine the existing herpetofaunal com- munities in the Andes. 371 372 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 METHODOLOGY The Andean herpetofauna never has been reviewed or summarized in its entirety. Some faunistic studies contributed much basic data. Thus, Ruthven's (1922) report on the herpe- tofauna of the Sierra Nevada de Santa Marta, Rivero's ( 1961 ) account of the frogs of Vene- zuela, Cei's (1962) study of Chilean am- phibians, Rivero's (1963a) summary of the distribution of Venezuelan Andean frogs, Donoso-Barros' (1966) account of the reptiles of Chile, Roze's (1966) summary of Venezue- lan snakes, and Cochran and Goin's (1970) account of Colombian frogs have been useful sources of information, as have been the more limited papers on the Loja Basin in Ecuador by Parker (1932, 1934, 1938) and on the Titicaca Basin by Parker (1940). The cata- logues of Neotropical squamates by Peters and Orejas-Miranda (1970) and Peters and Donoso-Barros (1970) were a primary source for taxonomic literature on snakes and lizards, as were the checklists of leptodactylid frogs by Gorham (1966), Eleutherodactijlus by Lynch (1976b), and hylid and centrolenid frogs by Duellman ( 1977 ) . Substantial distributional data are in- corporated in numerous systematic studies, as follow: Brame and Wake (1963) on sala- manders of the genus Bolitoglossa; Cei ( 1971, 1973, 1974a,b), and Cei and Castro (1973) on iguanid lizards; Cei (1972) on Bufo; Duellman (1972) on Hyla; Duellman (1974), and Duellman and Fritts (1972) on Gastro- theca; Duellman and Veloso ( 1977 ) on Pleu- rodema; Edwards (1974) on the frogs of the genus Colostethus; Fritts (1974) on lizards of the genus Stenocercus; Lynch (1975a-c, 1976b) on frogs of the genera Eleutherodac- tijlus and Phrynopus; Montanucci (1973) on lizards of the genus Pholidobolus; Oftcdal (1974) on lizards of the genus Anadia; Ruiz and Hernandez (1976) on Colombian mon- tane bufonids; Taylor (1968) on caecilians; Trueb (1971, 1974, 1979) on frogs of the genera Rhamphophryne, Hemiphractus, and Telmatobius; Uzzell (1970, 1973) on micro- teiid lizards; Vellard (1951-1960) on Peru- vian frogs; and Veloso and Trueb (1976) on frogs of the genus Telmatobius. Much of the distributional data used here has not been published previously. Some of the data were obtained from museum collec- tions in the United States, Europe, and South America, but much of it is from the extensive Andean collections in the Museum of Natural History at The University of Kansas. Distri- butional data are provided for many unnamed species that are designated solely by letters. The 727 species of amphibians and reptiles known to occur in the Andes were tabulated for 1) altitudinal ranges, 2) major habitats occupied, and 3) physiographic regions in- habited (Appendices 15:1-3). Only species occurring above 1,000 m are included. Many species primarily inhabiting lowlands, and only peripherally inhabiting Andean slopes, were excluded. The 27 physiographic regions are classified and defined in six units, as follow. These units are arbitrary groupings and do not necessarily correspond to geomorphological regions. A. Venezuelan Andes 1. Serrania de Paria. — Easternmost high- lands on the Peninsula de Paria. 2. Serrania de Turumiquire. — Isolated highland mass in northeastern Vene- zuela. 3. Cordillera de la Costa. — The coastal ranges of northern Venezuela. 4. Merida Andes. — The eastern spur of the Andes in western Venezuela. B. Sierra Nevada de Santa Marta 5. Sierra Nevada de Santa Marta. — Iso- lated range in northern Colombia. C. Northern Andes 6. Cordillera Occidental in Colombia. — The Andes west of the Rio Cauca Val- ley. 7. Cordillera Central in Colombia. — The Andean range between the Rio Cauca and Rio Magdalena valleys. 8. Cordillera Oriental in Colombia. — The Andes east of the Rio Magdalena Val- ley. 9. Nudo de Pasto.— The highland mass in southern Colombia and extreme north- 1979 DUELLMAN: HERPETO FAUNA OF ANDES 373 ern Ecuador from which the Colombian and Ecuadorian cordilleras diverge. 10. Cordillera Occidental in Ecuador. — The western Andean range. 11. Cordillera Oriental in Ecuador. — The eastern Andean range. 12. Inter-Andean Basins in Ecuador. — The high valleys lying between the eastern and western ranges. D. Huancabamba Depression 13. Huancabamba Depression. — The low ranges and basins in northern Peru and southern Ecuador. E. Central Andes 14. Cordillera Central in Perii. — The An- dean range in northern Peru between Rio Maranon and Rio Huallaga val- leys. 15. Cordillera Oriental in northern Pent. — The eastern range of the Andes to the east of the Rio Huallaga and Rio Mantaro valleys. 16. Cordillera Oriental in southern Peru. — The eastern range of the Andes east and north of the Rio Apurimac and the Altiplano. 17. Cordillera Occidental in northern Peru. — The western range of the An- des north of Lima. 18. Cordillera Occidental in southern Perii. — The western range of the An- des south of Lima. 19. Maranon Valley. — The upper valley of the Rio Maranon between the Cordil- lera Occidental and Cordillera Cen- tral in northern Peru. 20. Huallaga Valley. — The upper valley of the Rio Huallaga between the Cordil- lera Central and Cordillera Oriental in central Peru. 21. Mantaro- Apurimac Valleys. — The in- termontane valleys of the Rio Mantaro and Rio Apurimac in central Peru. 22. Cordillera Oriental in Bolivia. — The Andes north and east of the Altiplano in Rolivia. F. Southern Andes 23. Altiplano. — The high Andean plateau in Bolivia, southern Peru, and north- ern Argentina. 24. Andes in northern Argentina. — The ranges east of the Altiplano to 27°S Lat. 25. Andes in northern Chile. — The ranges west of the Altiplano to 27°S Lat. 26. Andes in southern Ctule. — The west- ern slopes of the Andes south of 27°S Lat. 27. Andes in southern Argentina. — The eastern slopes of the Andes south of 27°S Lat. Many of the regions are definitive physio- graphically, whereas other divisions are ones of convenience for analyzing distributions. The Cordillera Central in Colombia is con- tinuous with the Cordillera Oriental in Ecua- dor. The Cordillera Oriental in Peru is con- tinuous with the Cordillera Oriental in Bolivia and the Andes in Argentina. The Cordillera Occidental in Peru is continuous with the Chilean Andes, the southern ranges of which are solely the western slopes of the Andes of southern Argentina. The ten habitat types are defined, as fol- low (see Simpson, this volume, for more extensive descriptions and discussion of vege- tation ) : 1. Arid. — Sparse, xeric-adapted vegetation on the western cordilleras in Peru and northern and central Chile and on the eastern cordilleras in central Argentina characterized locally by legumes (Aca- cia, Adesmia, Prosopis), cactus, and terrestrial bromeliads (Puya) (Fig. 15:1). 2. Cloud forest. — The humid lush forests on the windward slopes from Venezuela to Bolivia are termed variously humid montane forest, upper montane forest, or ceja (Fig. 15:2). The lower limits of this forest are usually no less than 1000 m, whereas the upper limit varies locally from 2600 to 3800 m. Charac- terized by a diversity of woody plants, especially various species of Podocar- pus, the cloud forest has many tree ferns and epiphytes. 3. Dry forest. — Xerophytic scrub forest 374 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 dominated by legumes, principally Acacia and Prosopis, and a variety of cacti, is characteristic of the Andean slopes in southwestern Ecuador, the Huancabamba Depression, some in- terior valleys in Peru, Bolivia, and northern Argentina (Fig. 15:3). 4. Nothofagus forest. — This term applies to the austral cool temperate forests in southern Argentina and Chile domi- nated by various species of southern beech (Nothofagus) and conifers (Araucaria, Cupressoides, Fitzroya); see Formas (this volume) for a de- tailed discussion (Fig. 15:4). 5. Paramo. — The vegetation above tree line (2600-3S00 m) in the northern Andes and Merida Andes generally is composed of low ( < 1 m ) herbaceous vegetation with some woody bushes (especially Baccharis), cushion plants (Distichia), grasses (especially Fes- tuca), and in northern Ecuador, Co- lombia, and Venezuela the character- istic composites Espeletia (Fig. 15:5). 6. Patagonian scrub. — This cold-adapted, xerophilic vegetation formation charac- teristic of Patagonia ascends the eastern slopes of the Andes in central Argentina to elevations of about 3500 m and exists in local areas in Chile west of Andean passes (Fig. 15:6). Bushes (Mulinum, Bcrberis, and others) are mixed with grasses (Festuca, Poa, Stipa), herbs (Senecio, Acaena, etc.), and some low spiny sclerophylls (Ephedra, Adesmia, etc.). 7. Puna. — The montane habitat above tree line that is drier than paramo and ex- tends from southern Ecuador to north- ern Argentina is called puna. The puna is dominated by bunch grasses (Fes- tuca, Poa, and especially Stipa); in many extensive areas, grasses are the only evident vegetation (Fig. 15:7). Composites, such as Baccharis, Lepi- dophyllum, and Senecio, are wide- spread, whereas low trees (Polylepis) and cushion plants (principally the umbellifer Azorella) are local in their distributions. In many areas the puna grasses are grazed by domestic herds of sheep, llamas, and alpacas (Fig. 15:8). 8. Rainforest. — The lowland and lower montane ( < 1000 m ) rainforests are not part of the Andean vegetation; notation of the occurrence of a species in this habitat indicates that it inhabits rainforest in addition to some Andean habitat, usually cloud forest. 9. Subpdramo. — A localized ecotone be- tween cloud forest and paramo occurs sporadically at elevations of 2700 to 3500 m along the eastern Andean front from Colombia to Bolivia. Usually the vegetation consists of stunted, closely packed trees (Polylepis) or bushes (Baccharis) heavily laden with mosses and in some areas supporting many bromeliads (Fig. 15:9). The bamboo (Chusquea) usually is present. 10. Valley vegetation. — -This term is ap- plied to the vegetation of the high inter-Andean valleys, which for cen- turies have been modified by man so that remaining grasses have been grazed, and fields are devoted to crops, principally wheat and potatoes. The numerous rock fences and irrigation ditches provide suitable habitats for many kinds of amphibians and reptiles (Fig. 15:10). The taxonomy of many Andean amphib- ians and reptiles is known inadequately. Thorough taxonomic studies are needed for frogs of the genus Tehnatobius (currently being studied by Linda Trueb), lizards of the genus Proctoporus (currently being studied by Thomas H. Fritts), and snakes of the genus Atr actus. Taxonomic problems still remain in the large iguanid lizard genus Liolaemus and the hylid frog genus Gastrotheca. Based on the recent rate of acquisition of new species of frogs of the genera Centrolenella, Colo- stethus, and Eleutheroclactylus, many more species remain to be discovered in unexplored and poorly collected ranges and valleys. At the present time the Ecuadorian Andes, Me- rida Andes, Cordillera de la Costa in Vene- zuela, and southern Chilean Andes probably are the best-known regions, whereas the Cor- dillera Central in Peru, Cordillera Oriental 1979 DUELLMAN: HERPETOFAUNA OF ANDES 375 "r -•iJ^*wii flF Chile. Fig. 15:1. Arid slopes on the west face of the Andes, 6 km E Lo Valdes, 2250 m, Provincia de Santiago, le. Ladcras dridas en el hdo oeste de los Andes, 6 km E Lo Valdes, 2250 tn, Provincia de Santiago, Chile. Fie. 15:2. Cloud forest in Cordillera de la Costa, Rancho Grande, 1100 m, Estado de Aragua, Venezuela. Selva de neblina en la Cordillera de la Costa, Rancho Grande, 1100 m, Estado de Aragua, Venezuela. 376 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Fig. 15:3. Dry forest in an eastern valley, 15 km S Quiroga, 1750 m, Departamento de Cochabamba, Bolivia. Selva seca en un valle del este, 15 km S Quiroga, 1750 m, Departamento de Cochabamba, Bolivia. Fig. 15:4. Nothofagus forest at Lago de Huechulaf quen, 900 in, Provincia de Neuquen, Argentina. Selvas de Nothofagus en el Lago de Huechulafquen, 9(H) m, Provincia de Neuquen, Argentina. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 377 Fig. 15:5. Paramo dominated by Espeletia in Paramo El Angel, 14 km SW Tulcan, 3340 m, Provincia dt Carchi, Ecuador. Paramo dominado por Espeletia en el Paramo El Angel, 14 km SO Tulcan, 3340 m, Provincia de Carchi Ecuador. Fig. 15:6. Patagonian scrub, with scattered Adesmia, on south slope of Paso El Choique, 1950 m, Provincia de Mendoza, Argentina. Matorral patagonico, con Adesmia dispersas, en la ladera sur del Paso El Choique, 1950 m, Provincia de Mendoza, Argentina. 378 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 '*m%hto Fig. 15:7. Bunch grass puna, Pampas de Ramoscruz, 31 km W Orcos, 4120 m, Departamento de Ayacucho, Peru. Puna con pasto en champas, Pampas de Ramoscruz, 31 km O Orcos, 4120 m, Departamento de Ayacucho, Peru. Fie. 15:8. Rocky puna, Altiplano, 38 km W Challa, 4300 m, Departamento de Cochabamba, Bolivia. Puna rocosa, Altiplano, 38 km O Challa, 4300 m, Departamento de Cochabamba, Bolivia. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 379 ~850 m) northern (seaward) slopes of the cordillera support luxuriant cloud forest, which dis- integrates into dry forest on most of the lee- ward slopes. Cloud forest occurs at elevations above 1500 m on Cerro Turumiquire and as low as 600 m in the Serrania de Paria. In addition to the general references cited, distributional data for the Venezuelan high- lands were obtained from Donoso-Barros (1968), Rivero ( 1963a,b, 1964, 1968, 1972, 1974), Rivero and Mayorga (1973), Test, Sex- ton and Heatwole (1966), Williams (1974), and Williams, et al. (1970). Published data were supplemented by information provided by Scott J. Maness and by material collected by me in 1974. Forty-eight amphibians and 32 reptiles oc- cur principally at elevations of more than 1000 m in the Venezuelan highlands.1 Of 'Rivero ("1976" [1978]) named three additional species of frogs (Colostethus) from the Merida Andes. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 381 Fig. 15:11. The Venezuelan Andes. A. Merida Andes; B. Cordillera de la Costa; C. Cerro Turumiquire; D. Serrania de Paria. Los Andes venezolanos. A. Los Andes de Merida. B. Cordillera de la Costa. C. Cerro de Turumiquire. D. Serrania de Paria. these, only six are confined to elevations of more than 2500 m. All six occur in the Merida Andes and include four frogs (Eleutherodac- tylus boconoensis, E. ginesi, E. lancinii, Atelo- pus mucubajiensis) and two lizards (Anadia bitaeniata, A. brevirostris) . Only three other species exceed 2500 m — the salamander, Bo- Utoglossa orestes, and the frog, CentroleneUa buckleyi, in the Merida Andes, and the frog, Colostethus mandelorum, on Cerro Turumi- quire. Only four species are known from the Serrania de Paria and seven from Cerro Tu- rumiquire, each having three and five en- demic species, respectively.2 Of the 46 species in the Cordillera de la Costa, 34 are endemic. Thus, in these three areas, specific endemism is 71-75 percent. Only 18 (56%) of the 32 species in the Merida Andes are endemic. Despite the low (600 m) separation of the Merida Andes from the Cordillera Oriental in Colombia, seven species occur in both cor- 2 S. J. Gorzula (pers. comm. ) collected six additional species of frogs (2 Eleutherodacttjlus, 3 Centrole- neUa, and Flectonotus) in the Serrania de Paria in 1978. dilleras. Two widespread frogs (Hyla labialis and CentroleneUa buckleyi) inhabit subpara- mo and paramo from 2000 to 2700 m in the Merida Andes and similar habitats at 2400 to 3000 m (Hyla) and 2100 to 3400 m (Centro- leneUa) in Colombia. The other species — Gastrotheca nicefori ( 1575 m ) , Anolis nigro- punctatus (1200 m), Chironius monticola (1000-1600 m), Leimadophis bimaculatus (1400-2500 m), and Micrurus mipartitus (800-2000 m) inhabit cloud forest on both sides of the Depresion de San Cristobal. Other species in the Merida Andes have af- finities with species in the Colombian Andes. Atelopus oxyrhynchus and A. mucubajiensis are members of the Atelopus ignescens group, which is speciose in Colombia and Ecuador and occurs in the Sierra Nevada de Santa Marta (Rivero, 1963b). Hyla platydactyla is a member of the Hyla bogotensis group con- taining four species in the main Andean Cor- dillera. One salamander ( Bolitoglossa savagei, 1000-2000 m) and one snake (Micrurus mi- partitus, 1600-2000 m) occur in the Merida Andes and the Sierra Nevada de Santa Marta in northern Colombia. 382 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 The major faunal affinities between the Merida Andes and Cordillera de la Costa are among six snakes — Atractus badius, Chironius monticola, Dendrophidion percarinatus, Lam- propeltis triangulum, Leimadophis zweifeli, and Micrurus mipartitus, all but two of which have lower distributional limits of less than 1000 m. No frogs or lizards are shared by the two Cordilleras. Two species of frogs (Colostethus her- minae and Eleutherodactylus urichi) are shared between the Cordillera de la Costa and Cerro Turumiquire; the Eleutherodacty- lus also occurs in the Serrania de Paria and on Trinidad. According to John D. Lynch (pers. comm.), many of the Eleutherodacty- lus in the Cordillera de la Costa have affinities with West Indian species rather than with those in the Andes of Colombia and Ecuador. This faunal relationship also appears in the hylid frog Flectonotus pygmaeus that inhabits the Cordillera de la Costa and Isla Tobago north of Trinidad. Two other species are in- cluded in the genus — F. fitzgeraldi on Trini- dad and F. fissilis in the highlands of south- eastern Brasil. Five inhabitants of cloud forest in the Venezuelan highlands (2 frogs — Gastrotheca nicefori, Centrolenella fleischmanni; 3 snakes — Dendrophidion percarinatus, Lampropeltis triangulum, Micrurus mipartitus) also occur in cloud forests in lower Central America. The hylid frog, Phyllomedusa medinae, be- longs to a group having a species in lower Central America (P. lemur), one on the east- ern slopes of the Andes in Ecuador ( P. buck- leyi) , and an unnamed species on the Pacific slopes in Colombia (Duellman, 1970). Two monotypic genera of snakes ( Umbri- vaga mertensi and "Urotheca" williamsi) are endemic to the Cordillera de la Costa. No species are shared with the Guiana Highlands. Thus, each of the four regions of the Venezuelan highlands has endemic species of amphibians and reptiles; those of the two small highland areas (Cerro Turumiquire and Serrania de Paria) seem to have been derived from the Cordillera de la Costa, which shares few species with the Merida Andes (Fig. 15: 12). Thirty-two species occur in the Merida Andes, as compared with 103 species in the Cordillera Oriental in Colombia; only seven species are in common. The most speciose genera in the Venezuelan highlands are Eleu- therodactylus, Colostethus, Centrolenella, Anolis, Anadia, and Atractus, all of which are widespread and diverse in humid lowland and foothill habitats. Therefore, it seems most likely that the species of these genera that are endemic to the highlands were derived from lowland ancestral stocks. A minor per- centage of the fauna of the Merida Andes ap- parently was derived from highland stocks in the Cordillera Oriental of Colombia. In the latter group are Bolitoglossa orestes, B. sa- vagei, Atelopus mucubajiensis, A. oxyrhyn- chus, Gastrotheca nicefori, Hyla labialis, Hyla platydactyla, Centrolenella bucklcyi, and An- olis nigropunctatus. Certainly in contrast to the Colombian and Ecuadorian ranges of the Andes, the fauna of the Merida Andes is depauperate; this suggests that the Depresion de San Cristobal has been an effective barrier to the dispersal of most highland groups. Furthermore, the recency of elevation of the Merida Andes, combined with late Pleisto- cene glaciation of the small areas now sup- porting paramo, may be partly responsible for the few high montane species of amphib- ians and reptiles. Sierra Nevada de Santa Marta An isolated volcanic range, the Sierra Nevada de Santa Marta, consists of an area of only about 16,000 sq km. The highest ele- vations, such as Pico Cristobal Colon at 5775 m, are perpetually covered with snow. The sierra rises abruptly from the Caribbean coastal plain and is narrowly separated by arid lowlands at elevations of less than 500 m from the northern part of the Cordillera Ori- ental of the Andes; the northern part of the cordillera along the Colombian-Venezuelan border is the Siena de Perija, which attains elevations of 3750 m. The first uplift of the Sierra Nevada de Santa Marta was in the Miocene, but the final uplift did not occur until the end of the Pleistocene (Gansser, 1955). According to Carriker (1922), cloud forest occurs at elevations of 1385 to 2200 m, but I have observed cloud forest on Cerro San Lorenzo (= Cerro Kennedy) at 2700 m. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 383 — -1 — -____ TO SM 21/16 ^-\— ^_-- -^*" SP 4/3 TR CC 46/34 ■* 1 — » ^2 * I MA 32/18 <*T ^2 ro-r «-7-> *S CT 7/5 IC )3/ 36 Fig. 15:12. Herpetotaunal comparisons of the regions of the Venezulan Andes and adjacent areas. Numbers in blocks are number of species/number of endemic species; numbers of shared species are within arrows. CC = Cordillera de la Costa; COrC = Cordillera Oriental de Colombia; CT = Cerro Turumiquire; MA = Merida Andes; SM = Sierra Nevada de Santa Maria; SP = Serrania de Paria; TO = Tobago; TR = Trinidad. Comparaciones faunisticas de la herpetofauna de las regiones de los Andes venezolanos y areas adyacentes. Ni'imeros dentro de hloqucs representor! numero de especies/ numero de especies endemicas; numero de especies en comi'tn estdn en las flechas. Paramo exists above tree line to the lower limits of snow at about 4900 m (Carriker, 1922). The basis for a discussion of the herpeto- fauna of the Sierra Nevada de Santa Marta is Ruthven's ( 1922 ) account, supplemented by my own field work in June 1974. Ruthven's taxonomy was modified by Brame and Wake (1963), ' Cochran and Goin (1970), Lynch (1975b, 1978a), Oftedal (1974), and Rivero (1963b). Of the 21 species of amphibians and rep- tiles known from the paramo and cloud forest on the Sierra Nevada de Santa Marta, 16 are endemic. Four of the nonendemic species also occur in the Cordillera Oriental of the Andes. Of these, one frog, Eleutherodactylus prolixodiscus, and a teiid lizard, Anadia pul- chella, occur at elevations of 2100 to 2700 m in the Cordillera Oriental. The frog Amhhj- phrynus ingeri inhabits cloud forests at 1720 to 1980 m in the Cordillera Oriental and the Cordillera Central. The snake Micrnrus mi- partitus occurs at elevations of less than 2000 m in the Colombian Andes, Sierra Nevada de Santa Marta, Merida Andes, Cordillera de la Costa, and highlands in lower Central Amer- ica. The salamander, BoJitoglossa savagei, in- habiting cloud forest at 1000 to 2100 m in the Sierra Nevada de Santa Marta also occurs at 2000 m in the Merida Andes (Brame and Wake, 1963). Of the 16 endemic species, three frogs (Atelopus carrikeri, A. ivalkeri, CentroleneJla sp. "P") have relatives in the Cordillera Oriental and in the Merida Andes. Two frogs (Colostethus sp. "A," Cryptobatrachus bou- lengeri) have relatives in the Colombian An- des. The relationships of the enigmatic frog Geobatrachtis icalkeri are unknown. The rep- tiles (Pseudogonatodes, Anolis, Atractas) and two frogs that are members of the Eleuthero- dactylus fitzingeri group (E. carmelitae and E. insignitus) have many congeners in the lowlands. The other species of Eleutherodac- tylus are members of the Eleutherodactylus unistrigatus group and have relationships with species in the Cordillera de la Costa in Venezuela (John D. Lynch, pers. comm.). The apparent absence of tree frogs of the Hyla bogotensis group and Gastrotheca is noteworthy; both are represented by numer- ous species in the cloud forests in the Co- lombian Andes. On the basis of the limited information available, it seems that some of the herpet- ofauna of the Sierra Nevada de Santa Marta might have been derived from the Cordillera 384 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 03/36 cc 46/34 Fig. 15:13. Herpetofaunal comparisons of the Sierra Nevada de Santa Marta with other highlands. Numbers in blocks are number of species/number of endemic species; numbers of shared species are within arrows. CC = Cordillera de la Costa; CCC = Cordillera Central de Colombia; COcC = Cordillera Occidental de Colombia; COrC = Cordillera Oriental de Colombia; MA = Merida Andes; SM = Sierra Nevada de Santa Marta. Comparaciones faunisticas de la herpetofauna de la Sierra Nevada de Santa Marta con otras tierras de alturas. Nitnieros dcntro de bloques representan numero de especies/ numero de especies endemicas; numero de especies en comun estdn en las flechas. de la Costa in Venezuela, some from the Merida Andes, some from the Cordillera Ori- ental, and some from the surrounding low- lands (Fig. 15:13). With the exception of the widespread Micrurus mipartitus, no spe- cies are shared with the Central American highlands. Much exploration remains to be done in the Sierra Nevada de Santa Marta, a region that probably has many more species than known at present. Also, the geographi- cally important but biologically unexplored Sierra de Perija may hold the key to under- standing the fauna! relationships of the Sierra Nevada de Santa Marta. Northern Andes The northern Andes are comprised of five major north-south ranges diverging from the Nudo de Pasto and a series of high intermon- tane basins in Ecuador; the entire northern Andes extend for about 1800 km from the Caribbean lowlands at 10°50'N to the Huan- cabamba Depression at 4°30'S. Central to the physiography of the northern Andes is the high massif of the Nudo de Pasto in southern Colombia and northern Ecuador (Fig. 15: 14). The Nudo de Pasto encompasses a north-south extent of about 110 km and a breadth of about 130 km. The nudo is bor- dered to the northwest by the Rio Patia, to the northeast by the headwaters of the Rio Caqueta, and to the south by the Rio Chota. Much of the nudo is above 3000 m with two peaks, Volcan Chiles and Volcan Cumbal, reaching 4760 m. The Andes north of the Nudo de Pasto form three distinct ranges. The western range, the Cordillera Occidental extends for about 650 km between the Pacific lowlands and the valley of the Rio Cauca. The southern border is the Rio Patia; the dry upper valley of the river at about 1200 m separates the Cordillera Occidental from the Nudo de Pasto. The Cordillera Occidental is narrow ( <50 km) and lacks continuous high ridges; the two highest peaks are Cerro Tamana (4200 m) and Pico Frontino (4080 m). The Cordillera Central extends 750 km north from the Nudo de Pasto; this range about 100 km in width is bordered on the west by the valley of the Rio Cauca and on the east by the valley of the Rio Magdalena. Extensive areas are above 3000 m, and four peaks with permanent snow exceed 5000 m — Nevado del Huila (5760 m), Nevado del Quindio (5400 m), Nevado del Ruiz (5400 m), and Nevado del Tolima (5215 m). The Cordillera Oriental is narrowly sep- 1979 DUELLMAN: HERPETOFAUNA OF ANDES 385 io° - Fig. 15:14. The northern Andes. A. Sierra Nevada de Santa Marta; B. Cordillera Occidental de Colombia; C. Cordillera Central de Colombia; D. Cordillera Oriental de Colombia; E. Nudo de Pasto; F. Cordillera Occidental de Ecuador; G. Cordillera Oriental de Ecuador; H. Huancabamba Depres- sion. The inter-Andean Basins lie between the Cordillera Occidental and the Cordillera Oriental in Ecuador. Los Andes del norte. Las hoijas interandinas estdn entre la Cordillera Occidental y la Cordillera Oriental en Ecuador. 386 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 arated from the Nudo de Pasto by the upper Rio Caqueta Valley at about 1200 m and extends about 1200 km north-northeastward to about 10°50'N, where the northern part of the Cordillera closely approximates the dis- junct Sierra Nevada de Santa Marta. The Cordillera Oriental reaches a width of 200 km where the topography is a complex array of ranges and basins. With the exception of the northern one-fourth of the cordillera, there are large areas at elevations above 3000 m; the highest peak is the Nevado del Cocuy ( 5493 m ) . Lying to the east of the Cordillera Oriental is the isolated Serrania de la Maca- rena which reaches elevations of more than 2000 m. South of the Nudo de Pasto, the Cordillera Occidental of Ecuador extends southward for about 500 km. The Cordillera Occidental is limited in the north by the dry valley of the Rio Chota at about 1500 m and in the south by the dry valley of the Rio Jubones at about 900 m. The elevation of the entire southern two-thirds of the cordillera exceeds 3000 in. Many volcanoes in the cordillera exceed 4500 m; the highest is the majestic Volcan Chim- borazo (6310 m), capped with snow and gla- ciers and the highest peak in the Andes north of Peru. The Cordillera Oriental of Ecuador is con- tinuous with the Nudo de Pasto and extends southward for about 620 km to the Huanca- bamba Depression. Unlike the western cor- dillera, the high elevations of the Cordillera Oriental are interrupted by the valleys of the Rio Pastaza, Rio Paute, and Rio Zamora. The eastern cordillera has many volcanoes, the highest of which, Volcan Cayambe, reaches 5790 in; three volcanoes over 5000 m are active — Volcan Cotopaxi, Volcan Sangay, and Volcan Tungurahua. Whereas the Cor- dillera Occidental drops precipitously to the Pacific lowlands, the Cordillera Oriental slopes much more gradually into the Amazon Basin. Three disjunct highland areas rise from the foothills of the Cordillera Oriental — Cerro Sumaco (3900 m), Cordillera de Cu- tucu (2200 m), and Cordillera del Condor (2450 m). Between the eastern and western Cordil- leras in Ecuador are 10 basins (cuencas or Fie. 15:15. The Inter-Andean Basins of Ecuador. A. Tulcan; B. Ibarra; C. Otavalo; D. Quito; E. Latacunga; F. Riobamba; G. Alausi; H. Cuenca; I. Saraguro; J. Loja. Area between 3000 and 5000 m is shaded. Las hoyas interandinas de Ecuador. Las areas cntrc 3000 y 5000 m estdn sombreadas. hoyas) that are separated by transverse ridges (mulcts) completely or partially separating the basins and in most cases connecting the eastern and western Cordilleras. The basins have elevations ranging from 2000 to 3100 m (Fig. 15:15). 1979 DUELLMAN: HERPETOFAUNA OF ANDES 387 The northern Andes have received con- siderable geological study (Biirgl, 1961; Herd and Naeser, 1974; Sauer, 1965, 1971; Shagam, 1975) and intensive palynological investiga- tion (van der Hammen, et al., 1973; van der Hammen, 1974). The conclusions of these workers and Simpson (this volume) indicate that the northern Andes probably had few areas over 1000 m above sea level at early Pliocene time. The major orogeny occurred at the end of the Pliocene with the uplift of the eastern cordilleras taking place before that of the western cordilleras. The absence of evidence of glaciation on some high peaks suggests that their final uplift occurred after the last major glaciation. Nonetheless, most areas above 3700 m were glaciated; climatic depression was in the magnitude of 6-7°C with a downward shift of environments of about 1000 to 1200 m (van der Hammen, 1974). The extensive areas above tree line are humid and cool with annual precipitation of 1000 to 2000 mm and little seasonal fluctua- tion in temperature, but daily variation of 10°C or more ( Cuatrecasas, 1968). These areas of paramo have grasses (Festuca), rosette herbs (Espeletia and Senecio), cushion plants (Distichia), and low bushes (Baccha- ris). The western slopes of the Cordillera Occidental and the eastern slopes of the Cor- dillera Oriental from the Depresion de San Cristobal southward support luxuriant cloud forests. These humid montane forests also occur locally in the Cordillera Central, espe- cially in the northern part. Subparamo is common but localized in the eastern and western cordilleras. The inter-Andean basins in Ecuador possibly were subparamo prior to human modification into cultivated fields and grazing of livestock. In addition to the general publications al- ready cited, I have drawn information from the works on anurans by Duellman ( 1972, 1973), Duellman and Altig (1978), Duellman and Simmons (1977), Lynch ( 1975a,b, 1976a, 1979), Lynch and Duellman (1973, 1979), Myers and Daly (1976a,b). and Peters (1973). The works Myers (1973, 1974) and Savage (1960) on colubrid snakes also were used. Much of the distributional data is based on the extensive collections in the Mu- seum of Natural Histoiy at The University of Kansas and in the National Museum of Natural History. Herpetologically, the northern Andes have the richest fauna in the continent; 415 (57$?) of the 727 Andean species occur in this region. Of these, 345 species (837c) are endemic to the northern Andes. The taxonomic disposi- tion of the 415 species (number of endemics in parentheses ) is caecilians 15 ( 12 ) , sala- manders 11 (11), frogs 262 (225),:! lizards 54 (40), snakes 73 (57). Seven genera (Am- phignathodon, Centrolene, Osornophryne, Phenacosaurus, Pholidobohis, Saphenophis, Synophis) are endemic to the northern Andes, and one (Cryptobatrachus) is endemic save for one species in the Sierra Nevada de Santa Marta. Of the 70 species having ranges extending beyond the limits of the northern Andes, 43 also are present in the adjacent lowlands. Some of these also occur in the central Andes, especially species that inhabit rainforest and cloud forest. Thus, 15 species are shared be- tween the Cordillera Oriental in Ecuador and the Cordillera Central in Peru, and 20 are shared between the former and the Cordillera Oriental in Peru. Fourteen species on the Pacific slopes of the Cordillera Occidental in Colombia and Ecuador, and /or the northern parts of the Colombian cordilleras are shared with the highlands in lower Central America; these include six frogs (Gastrotheca nicefori, Hemiphractus fasciatus, Centrolenella fleisch- manni, C. griffithsi, C. prosoblepon, C. vaJ- erioi), five lizards (Anolis antonii, A. chloris, Basiliscus galeritus, Polychrus gutturosus, Pri- onodactylus uertebralis), and three snakes (Dendrophidion percarinatus, Micrurus mi- partitus, Bothrops schlegeli). With the ex- ception of Gastrotheca nicefori and Centrole- nella griffithsi, all of these species range well below 1000 m. As noted previously, four species are shared with the Sierra Nevada de 3 Not included in these figures or in Appendix 15:2 are Atelopus carauta from 1300 m in the Cordillera Occidental of Colombia ( Ruiz and Hernandez (1978) or Colostethus abditaurantius from 1450 m in the Cordillera Central of Colombia ( Silverstone, 1975). 388 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Santa Marta, and seven are shared with the Merida Andes. If only those species that do not occur below 2500 m are considered, 72 of the 73 species are endemic to the northern Andes. The exception is the boid snake Tropidophis taczanowskyi, which inhabits the Cordillera Oriental in Ecuador and the Cordillera Cen- tral in Peru. Within the northern Andes, only two spe- cies, Centrolenella buckleyi (2100-3400 m in subparamo and paramo) and Eleuthero- dactylus w-nigrum ( 1230-2800 m in cloud forest and subparamo) occur in all seven re- gions. Eleutherodactylus vertebralis (2340- 3500 m in subparamo and paramo) occurs in the three Colombian cordilleras, the Nudo de Pasto, and the Cordillera Occidental in Ecuador. Four species of frogs (Eleuthero- dactylus buckleyi, E. unistrigatus, Atelopus ignescens, and Gastrotheca riobambae) dis- tributed mostly above 2500 m occur in four regions; three snakes (Chironius monticola, Rhadinaea latehstriga, and Micrurus miparti- tus) distributed below 2000 m also occur in four regions. Of the remaining 405 species in the northern Andes, 155 species occur in two or three regions, and 250 are endemic to a given region, with the largest number of en- demics in the Cordillera Oriental in Ecuador (74) and the Cordillera Occidental in Ecua- dor (65), but with the highest percent of endemism in the Cordillera Central in Co- lombia (54%). The highest faunal similarities among the seven regions in the northern Andes are be- tween the eastern cordilleras in Colombia and Ecuador (45 species in common) and the western cordilleras in Colombia and Ecuador (33 species in common) (Fig. 15:16). The low number of species (23) and endemics (1 lizard, Proctoporus laevis) in the Nudo de Pasto reflects a bias in the analysis; only high elevations ( > 2500 m ) were assigned to the nudo. The herpetofauna of the inter-Andean basins is composed mostly of species also in- habiting the adjacent cordilleras. Only three species (2 lizards — Pholidobolus montium, Proctoporus occidatus; 1 snake — Atractus leh- manni) are restricted to the basins. The southernmost basin ( Loja ) is considered to be part of the Huancabamba Depression. Only considering those 73 species that do not occur below 2500 m, a much different picture is evident (Fig. 15:17). Endemism in each region ranges from 20 percent in the Nudo de Pasto to 100 percent in the Cordil- lera Occidental in Colombia. Whereas the eastern cordilleras in Colombia and Ecuador and the western cordilleras in Colombia and Ecuador shared the greatest numbers of spe- cies when the entire fauna was considered, in an analysis of only the high montane species, they have no species in common. The greatest species richness and highest percentage of endemism in the northern An- des is amongst the frogs (especially Colo- stethus, Eleutherodactylus, and Centrole- nella), which form 60 percent of the entire Andean herpetofauna, but 65 percent in the northern Andes and 71 percent in the Cor- dillera Oriental in Ecuador. Although the entire fauna in the equatorial cordilleras is large, local communities are much smaller. Altitudinal and latitudinal changes in com- munity composition result in localized faun- ules in the cloud forest, subparamo, and par- amo, with generally decreasing numbers of species at higher altitudes. Equatorial tran- sects in the Cordillera Occidental and Cor- dillera Oriental reveal the presence of 62 and 79 species, respectively (Fig. 15:18). On the eastern slopes there is a dimunition of species at about 2000 m; this is especially evident upon examining the altitudinal distribution of individual species (Fig. 15:19). Analysis of broad latitudinal distributions of species on the eastern slopes of the Cordillera Oriental shows that even at lower elevations ( 1000- 1500 m) more than one-third of the species have limited distributions, whereas this per- centage nearly doubles at elevations above 2500 m (Table 15:1). High species richness, especially for anu- rans, in the equatorial cordilleras can be ascribed to the equable conditions with mod- erate to cool temperatures and high humidity relatively constant throughout the year. Local endemism, especially in the Cordillera Oriental in Ecuador, most likely is due to the discontinuous highlands. 1979 Table 15:1. — DUELLMAN: HERPETOFAUNA OF Latitudinal Distribution of 79 Species of Amphibians and Ecuador. ANDES Reptiles on the Andean 389 Slopes of Elevation N Ecuador Only Ecuador and Colombia Ecuador and Peru Colombia Ecuador Peril 1000-1500 m 1500-2500 m 44 43 17 (38.6%) 25 (58.6%) 11 (64.7%) 15 (34.1%) 10 (22.7%) 6 (35.3%) 6 (13.6%) 6 (13.656) 0 (00.0%) 6 (13.6%) 2 (04.5%) >2500 m 17 0 (00.0%) Huancabamba Depression Along the entire length of the main range of the Andes there is only one pass below tree line. This is in the complex system of low ranges and basins collectively referred to as the Huancabamba Depression in northern Peru and extreme southern Ecuador. Here, the northern extremity of the Cordillera Occi- dental of Peru is breached by the Abra de Porculla at 2145 m. In the Huancabamba Depression, the major cordilleras either termi- nate or fragment into isolated ranges usually less than 3500 m high and separated by val- leys mostly between 1000 and 2000 m above sea level (Fig. 15:20). Several small rivers drain the Pacific slopes, but east of the con- tinental divide, all streams eventually flow into the Rio Marafion. The interior basins are dry and support dry forest dominated by legumes and cacti. Except for the eastern front ranges, the east slopes are also dry, whereas the tops of ridges above 3000 m and the upper western slopes have a low cloud forest with many bromeliads. At the Huancabamba Depression there is a structural deflection of the Andean faults. There were extensive marine transgressions through this area in the Cretaceous ( Ham and Herrera, 1963 ) . The northern Peruvian Andes were uplifted only moderately prior to the Pliocene; the present elevations and drainage patterns were probably attained in the Pleis- tocene (Steinmann, 1930; Harrington, 1956; Gansser, 1973). Few papers have been published dealing with the fauna of the Huancabamba Depres- sion, but information on the distribution of some of the taxa can be found in Rarbour and Noble (1920b), Noble (1921), Vellard (1959), Duellman and Fritts (1972), Duell- man (1974), Fritts (1974), Trueb and Duell- man (1978), and Lynch (1979). Forty-three species of amphibians and rep- tiles, exclusive of predominantly lowland taxa, are known from the Huancabamba Depres- sion; 29 species, including the monotypic gen- era Polychroides and Macropholidus, are en- demic to the region.4 The endemics include 10 frogs, 10 snakes, and 9 lizards, six of the latter are members of the genus Stenocercus, most species of which occur southward in Peru. Five of the frogs are Eleutherodactylus and two are aquatic Telmatobius. Six species occur in the depression and in the Andes to the north, and two species occur in the Andes to the south and in the depres- sion. Two species ( 1 frog — Eleutherodactylus cojamarcensis; 1 snake — Philodryas simonsii) occur in the depression and in the Andes to the north and south; 21 species occur on the Andean slopes to the north and south but not in the depression (Fig. 15:21). Those species in the latter group are primarily inhabitants of cloud forests and rainforest on the eastern slopes of the Andes; 10 of the species are widely distributed at elevations below 1000 m, and only one snake (Tropidophis tacza- nowskyi) is not known from below 2000 m. Just as the Huancabamba Depression is a barrier to north-south dispersal of Andean species, it is a dispersal route between the arid Pacific lowlands and the Rio Marafion Valley. Several predominantly lowland taxa (Phyllodactylus, Tropidurus, Leptodeira) are so distributed. Central Andes The Central Andes compose the most mas- sive part of the Andean highlands (Fig. 15: 22). The Cordillera Occidental forms the backbone of the Andes south of the Huan- 1 The recently described colubrid snake, Sibynomor- phus oneilli, also is endemic (Rossman and Thomas, 1979). 390 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 COcC 66/29 10 > CCC 78/42 < 24 > COrC 103/36 33 45 COcE 137/65 < 15 * 22/3 10- -21- COrE 164/74 Fig. 15:16. Herpetofaunal comparisons of the regions of the northern Andes. Numbers in blocks are numbers of species/number of endemic species; numbers of shared species are within arrows. CCC = Cordillera Central de Colombia; COcC = Cordillera Occidental de Colombia; COcE = Cordillera Occidental de Ecuador; COrC = Cordillera Oriental de Colombia; COrE = Cordillera Oriental de Ecuador; IAB = Inter-Andean Basins; NP = Nudo de Pasto. One species is in common between CCC and IAB. Comparaciones faunisticas de la herpetofauna de las regimes dc los Andes del norte. Numcros dentro de bloques rcpresentan numero de cspecies /ni'imero de cspecies cndemicas; el numero de cspecies en comun estdn en las flechas. Una cspecies cs compartida por CCC y IAB. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 391 Fig. 15:17. Herpetofaunal comparisons of the regions of the northern Andes using only species that do not occur below 2500 m. Numbers and abbreviations are same as in figure 16. Comparaciones faunisticas de la herpetofauna de las regiones de los Andes del norte, comprendiendo solo aquellas especies que no bajan de los 2500 m. Numeros y abreviaciones igual que en la figura 16. 392 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 PACFC F - 3000 - - 2500 - F AMAZON AN I I 1 1 E 1 1 1 i f n Tl i . i r i i h s : 1 i ...i 1 1: ,: 1 1 ! 1:.. 1 1 :■'•'■! 1 DUU " Elev.(m) I . ;i 1 :.:-: 1 I 1 :::l 1 t |. ..::■;.:.: i 1 1 1 t. 1 25 20 15 10 5 0 0 5 NUMBERS OF SPECIES 10 20 25 30 35 Fig. 15:18. Species abundance at different elevations along equatorial transects of the Andes in Ecuador; reptiles are shaded and amphibians are open symbols. Abundancia de especies a diferentes alturas a lo largo de transects ecuatoriales en los Andes de Ecuador. Reptiles en simbolos sombreados y anfibios en simbolos claros. UJ -3000 2500 2000 -1500 1 1 1 1 1 i Amphibians Reptiles 1 N H 4 N 1 I I II I I I I I II I Mill ■ INI Fig. 15:19. Altitudinal distribution of amphibians (solid lines) and reptiles (broken lines) along an equa- torial transect of the Cordillera Oriental in Ecuador. Distribucion altitudinal de anfibios (en lineas continuas) y reptiles (en lineas discontinue) a lo largo de transect ecuatorial en la Cordillera Oriental en Ecuador. cabamba Depression. Originating at about 6°S Lat., the western Cordillera reaches ele- vations in excess of 4000 m at 8°S Lat.; from that point only one pass exists below 4000 m for a distance of about 2800 km to 31°S Lat. The highest mountains in Peru are in the Cordillera Occidental; the Cordillera Blanca is nearly 400 km in length and is mostly above 5000 m. The highest peak is Nevado Huascaran (6745 m). Although the Cordi- llera Occidental is a continuous range, for the purposes of analysis, I have arbitrarily divided it into northern and southern sections in Peru; the point of division is at about the latitude of Lima, inland from which the high central Nudo de Pasco forms a high connec- tion between the eastern and western Cor- dilleras. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 393 3000- 2000- 1000 M Fig. 15:20. Profile of the Huancabamba Depression at 5°15'S Lat. Perfil de la Depresion de Huancabamba a 5°15'Lat. S. ECUADOR HUANCABAMBA DEPRESSION PERU SPECIES 2 6 6 29 21 Fig. 15:21. Distribution patterns of Andean amphibians and reptiles in the Huancabamba Depression. Pat rones de distribution de anfibios u reptiles andinos en la Depresion de Huancabamba. Whereas the western Cordillera is a con- tinuous highland range, the eastern Cordil- leras in Peru and Bolivia are made up of many high ranges separated by long north- south valleys, the major rivers of which break through the cordilleras and drop into the Amazon Basin. For purposes of analysis, I recognize the Cordillera Central in northern Peru (6°-10°S Lat.) bordered to the west by the Rio Marafion Valley, the east by the Rio Huallaga Valley, and to the south separated from the Nudo de Pasco by the Rio Huertas. Although large areas of the Cordillera Central are over 4000 m, no peaks have permanent snow. The northern part of the Cordillera Oriental extending from 10°S to 12°S Lat. includes the Nudo de Pasco and cordilleras south to the Rio Mantaro Valley, which also forms the eastern border of the region. In the northern part of the Cordillera Oriental, extensive areas are above 4000 m, and the Nevado Hueyta Pailana has permanent snow above 5000 m. The southern part of the Cordillera Orien- tal in Peru begins at the Rio Tambo Valley and is continuous with the eastern cordilleras of Bolivia. In Peru the eastern cordillera is bordered on the west by the Rio Apurimac Valley and consists of many high ranges ( Cor- dillera de Vilcabamba, Cordillera de Vilca- 394 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 16 Kilometers AREA ABOVE 3000 METERS 82 Fig. 15:22. The central Andes. A. Cordillera Occidental North; B. Upper Maranon Valley; C. Cordillera Central; D. Upper Huallaga Valley; E. Cordillera Oriental North; F. Cordillera Occidental South; G. Man- taro-Apurimac Valley; H. Cordillera Oriental South; I. Cordillera Oriental de Bolivia; J. Altiplano. Los Andes centrales. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 395 6000 5000 4000 3000 2000 1000 M SOUTHWEST NORTHEAST Fig. 15:23. Profile of the Cordillera Oriental in southern Peru. Perfil de la Cordillera Oriental en el sur del Peru. nota, and Cordillera de Carabaya) having many peaks with permanent snow, the high- est of which is the Nevado Salcantaya (6271 m). Lower front ranges (Cadena de Pau- cartambo, Cadena de Pantiacolla) do not ex- ceed 4500 m; the deep valleys separating the various ranges give a relief of 2000 to 3000 m to the cordillera (Fig. 15:23). The major rivers dissecting the mountains are the Rio Apurimac, Rio Urubamba, and Rio Vilcanota. In Bolivia the eastern cordillera consists of a single range, the Cordillera Real. In the northwest and to the southeast there are two ranges — the Cordillera Central separated by the Rio Caipe from the outer range, the Cor- dillera Oriental. The highest peaks are in the snow-covered Cordillera Real, where four peaks exceed 6000 m, and the highest is Cerro Illimani (6460 m). The Cordillera Real drops precipitously into the Amazon Basin; the steep slopes dissected by deep ravines are known as the Yungas. Rising from the lowlands of the upper .Amazon Basin, several mountain ranges reaching above 2500 m are isolated from the main Andean cordillera. These ranges in Peru are, from north to south, Cerros de Otanahui, Cordillera Azul, Cerro de la Sal, and Serrania de Sira. The major montane valleys separating the principal cordilleras are those of the Rio Maranon, and Rio Huallaga flowing north- ward in northern Peru, and the Rio Mantaro flowing southeastward in central Peru, and the Rio Apurimac flowing northwestward in southern Peru. The Mantaro and Apurimac converge to form the Rio Ene; thus, these two valleys are placed together for the purposes of analysis. The initial uplift of the central Andes was 396 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 in the Miocene (Harrington, 1962; Aubodin, et al., 1973) with final major uplift completed by the end of the Pliocene and some addi- tional elevation in the Pleistocene (Petersen, 1958; Dollfus, 1960; Rutland, et al., 1965; Ahlfeld, 1970; James, 1971, 1973; Gansser, 1973). Considerable Pleistocene and Recent glaciation in the Peruvian Cordilleras de- pressed snow lines as much as 1500 m during at least two glaciations (Hastenrath, 1967; Kinzl, 1968; Simpson, this volume); the last major glaciation in the Cordillera de Vilca- nota has been dated as 28,000 to 14,000 years b.p. (Mercer and Palacios, 1977). The Cordillera Occidental is arid through- out its length with low xerophytic vegetation on the high Pacific slopes and puna on the high eastern slopes. The eastern and northern slopes of the easternmost ranges of the eastern cordilleras support lush cloud forest, the crests of these outer ranges have subparamo and/or wet puna habitats. The western and southern slopes of the outer ranges are dry with puna and low xerophytic vegetation. The high ridges of the interior ranges of the eastern cordilleras are drier than the outer ranges and have extensive areas of puna. The deep valleys between the ranges are dry with low sclerophytic vegetation at higher eleva- tions and dry scrub forest at lower elevations. The high montane valleys are extensively cultivated and also support puna, much of which is grazed. The diverse herptofauna of the central Andes has never been summarized. In addi- tion to the general works cited previously, the following works are pertinent to the system- atics and distribution of amphibians and rep- tiles of the central Andes: Barbour and Noble (1920a) on southern Peruvian taxa, Dixon and Huey (1970), Dixon and Wright (1975), Fritts (1974), and Uzzell (1969, 1970) on lizards; Schmidt and Walker (1943) and Walker (1945) on snakes; Duellman (1976, 1978a-c), Duellman and Fritts (1972), Duell- man and Toft (1979), Gallardo (1961), Ma- cedo (1960), Schmidt (1954), Silverstone (1975, 1976), and Vellard (1951-1960) on frogs. The herpetofauna of the central Andes, as presently known, consists of 159 species (75, or 48%, endemic to the regions) — 5 caecilians (1), 89 frogs (52), 44 lizards (11), 21 snakes (11). One genus of frogs (Batrachophrynus) and one of lizards (Opipeuter) are endemic to the region. Several genera are highly speciose in the central Andes — frogs of the genera Phrijnopus (Lynch, 1975a), Telmatobius (Macedo, 1960), and Gastrotheca (Duellman and Fritts, 1972); lizards of the genera Eu- spondylus (Uzzell, 1973) and Stenocercus (Fritts,' 1974). Of the 84 species that occur in the central Andes and elsewhere, 13 also inhabit tropical forests east of the Andes; some of these and some others comprise the 23 species that also occur in humid montane habitats in the northern Andes. Eight species also occur on the dry Pacific lowlands, and one of these inhabits dry forest in the northern Andes. Eight species (including 2 of 23 noted above) occur in the central Andes and the Huancabamba Depression. Only 12 species are shared between the central Andes and the southern Andes (including the Altiplano); one of these also ranges into the cis-Andean lowlands, two occur on the Pacific lowlands, and nine are restricted to high ( > 2500 m ) elevations in the central Andes and at least in the northern parts of the southern Andes. No species of amphibian or reptile occurs in all nine of the regions within the central Andes. The most widespread species are the frogs Pleurodema marmorata, Telmatobius marmoratus, Bufo spinulosus, the lizard Lio- laemus multiformis, and the snake Tachy- menis peruviana. These are the only species that occur in eastern and western cordilleras and intermontane valleys. The major simi- larities in the cordilleras are among the east- ern ranges, which individually share 9 to 14 species, with three species occurring in all four regions (Fig. 15:24). Only three species are shared between the northern and southern parts of the Cordillera Occidental, and no more than five species are common to any part of the Cordillera Occidental and any range in the eastern cordilleras. The similari- ties of the eastern ranges are principally in those species that inhabit the cloud forests between 1000 and 2000 m; the three species with distributions including all four regions of the eastern cordilleras are forest dwellers 1979 DUELLMAN: HERPETOFAUNA OF ANDES 397 Fie. 15:24. Herpetofaunal comparisons of the Andean ranges in the central Andes. Numbers in blocks are number of species/number of endemic species; numbers of shared species are within arrows. CC = Cordillera Central; COcN = Cordillera Occidental North; COcS = Cordillera Occidental South; COrB = Cordillera Ori- ental de Bolivia; COrN = Cordillera Oriental North; COrS = Cordillera Oriental South. Comparaciones faunisticas de la herpetofauna de las Cordilleras andinas de los Andes centrales. Numeros dentro de bloques representan numero de especies /numero de especies endemicas; numero de especies en comun estdn en las flcchas. — two lizards (Prionodactijlus argulus and P. manicatus) and one snake (Chironius monti- cola). The faunal list from the Cordillera Central is unrealistic; the area has not been studied adequately. The species richness in the southern part of the Cordillera Oriental reflects the complex topography and diverse habitats in that region (Fig. 15:25). Elevational changes in one inter-Andean valley result in striking differ- 398 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 -4000 2000 h < -I000M o E o Amazon Lowlands Fig. 15:25. Transect from the upper Rio Vilcanota Valley across the Cordillera de Vilcanota ( Abra Huancarane), Rio Yavare Valley ( Paucartambo), Ca- dena de Paucartambo (Abra Acanacu), and down the Rio Cosnipata Valley to the Amazon lowlands. Each line is the distribution of one species; solid lines are amphibians and broken lines are reptiles. Transect desde el vallc del alto Rio Vilcanota (Abra Huancarane), vallc del Rio Yavare (Paucar- tamho), Cadena de Paucartambo (Abra Acanacu), y hacia abajo del valle del Rio Cosnipata hasta las tierras bajas del Amazonas. Cada lima representa la distribution de una especie; las Uncus continuas rep- resentor* a anfibios y las lineas discontinuas a reptiles. ences in habitats and changes in species com- position. For example, in the Rio Urubamba Valley, three species of lizards of the genus Proctoporus replace one another along the length of the valley between the Cordillera de Vilcabamba and the Cordillera de Vil- canota. Likewise, in the same valley and Cordillera de Vilcanota, marsupial frogs have essentially parapatric distributions — Gastro- theca tnarsupiata is on the valley floor, G. ochoai inhabits bromeliads on the cliffs of the cordillera, and G. excubitor lives on the high parts of the cordillera ( Duellman and Fritts, 1972). The high montane valleys in Peru have a depauperate herpetofauna with a total of 20 species (7 endemic); otherwise, the species are shared with the neighboring cordilleras and to a lesser extent with the other valleys. The toad Bufo trifoliam is the only species occurring in the Huallaga, Maranon, and Mantaro-Apurimac valleys, and the frog, Gas- trotheca peruana occurs in the Huallaga and Maranon valleys. The most notable endemism is the monotypic frog genus Batrachophrynus restricted to Lago Junin and streams in the upper Rio Mantaro Valley. Five species in the Mantaro-Apurimac Valley are shared with the Altiplano in southern Peru and Bolivia; all of these are widespread highland species. Southern Andes Included in the southern Andes are the cordilleras in Argentina and Chile and the Altiplano from southern Peru through Bolivia to Argentina. As noted previously, the Cor- dillera Occidental in Peru is continuous with the Andes of Chile and western Bolivia; the Cordillera Central in Bolivia is continuous with the Andes of northern Argentina. The Chilean and Argentinean ranges unite south of the Altiplano at about 27°S Lat. (Fig. 15:26). The Altiplano, in its broadest sense, ex- tends about 1400 km north-south; its greatest width is about 300 km in Bolivia. Elevations of this high plateau range from 3400 to 4000 m; drainage is centripetal, forming lakes or salt basins. Precipitation in the form of rain or snow is mostly in the summer and de- 1979 DUELLMAN: HERPETOFAUNA OF ANDES 399 70 \r"t 1000 M Above 3000M Altiplano ■ ■ ■-"— 500 Km 20 30 50 60 Fig. 15:26. The southern Andes. A. Northern Chile; B. Altiplano; C. Northern Argentina; D. Southern Chile; E. Southern Argentina. Los Andes del sur. creases from about 500 mm annually in the north to essentially zero in the south. The vegetation of the northern and eastern parts of the Altiplano is puna dominated by bunch grasses, principally Festuca, but also Poa and Stipa, low (<1 m) shrubs of Adesmia and Parastrephia (Troll, 1959; Cabrera, 1968). To the south and west even bunch grasses be- come sparse and eventually absent in exten- sive "salares." The Andes of northern Argentina have ex- tensive areas over 4500 m and some peaks exceeding 6000 m, the highest being Cerro Bonete (6872 m). North- and south-flowing rivers separate lower eastern front ranges from the higher major cordillera which descends westward to the Altiplano. The Andes in northern Chile from a high main cordillera with many snow-covered peaks exceeding 6000 m; the highest in the extreme north is Cerro Parinacota (6330 m), whereas farther south near the southern end of the Altiplano, Nevado Ojos del Salado reaches 6880 m. The Andean precordillera exceeds 4500 m in most areas and chops precipitously to the narrow, xeric coastal strip. South of 27°S Lat. the single Andean cordillera continues for 3000 km to the tip of the continent. Whereas many peaks in the northern part of the range exceed 6000 m, including Cerro Aconcagua at 6959 m (the highest mountain in the New World), to the south there are few peaks over 4000 m. Equally important biologically are the eleva- tions of passes between the eastern and west- ern slopes; north of 31°S Lat. there are no passes below 4000 m. Passes between 2000 and 3000 m exist between 35° and 37CS Lat., south of which are found the only passes below 2000 m. Of course, passes at higher latitudes are correspondingly higher biologically. Al- though the western slopes of the Andes de- scend rapidly to the Pacific coast or to the Valle Longitudinal separated from the Pacific by the Cordillera de la Costa, the eastern slopes are much more complex with numerous precordilleran ranges, some of which, such as the Sierra Grande and Sierra de San Luis, are completely separated from the principal cor- dillera. In the south, the Andes descend only to the Patagonian plateaus at 600 to 1300 m. The Pacific slopes of the Andes in northern Chile (south to about 27°S Lat.) are ex- tremely arid with little vegetation, which in some places consists of only scattered cacti; at elevations above 4000 m, puna grasses are present. In a transition area between the desert and the austral forests (30°-38°S Lat.) low, sparse matorral is present on the Pacific slopes (Simpson, this volume). The Andes in northern Argentina have puna at high elevations and forest at lower elevations. From north to south (about 28°S Lat.) there is a change from cloud forest to 400 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 evergreen forest and deciduous broadleaf for- est. At about 26°S Lat. to 38°S Lat., the eastern slopes of the Andes are arid with de- ciduous forest existing in river valleys and Patagonian scrub infiltrating the lower slopes to 2000 m ( Roig, 1972 ) ; the Patagonian scrub crosses the continental divide at some low passes. South of 38°S Lat. on the Pacific slopes and 36°S Lat. on the eastern slopes are the austral forests characterized by a diversity of Nothofagus and in places dominated by Araucaria or Fitzroya (Formas, this volume). Within tins area is the so-called lake region of the southern Andes where cold streams cas- cade down from glaciers. The Andes in southern Chile and Argen- tina were uplifted nearly to their present heights by the end of the Miocene (Dott, et al., 1977). At that time, the initial uplift of the northern part of the Argentine, Chilean, and Bolivian Andes and the Altiplano took place (Petersen, 1958; Rutland, et al., 1965). The final, major orogeny of the Altiplano and the principal Cordilleras in northern Argen- tina and Chile was completed by the end of the Pliocene (Turner, 1972; Yrigoyen, 1972; James, 1971, 1973), whereas at least some of the extra-cordilleran ranges in Argentina were elevated later (Simpson and Vervoorst, 1977), and the coastal cordillera of Chile was uplifted earlier ( Okada, 1971 ) . The southern Andes were extensively glaciated during the Pleistocene, with large montane glaciers per- sisting to the present; at the height of glacia- tion, all of the Andes south of 30°S Lat. were entirely glaciated (Patterson and Lanning, 1967; Heusser, 1974). The herpetofauna of Chile has been re- viewed thoroughly by Cei ( 1962 ) and Donoso-Barros (1966, 1970). Cei (1979) re- viewed the amphibians of Argentina, but the reptiles have not been summarized. The only significant paper on the Altiplano is Parker's (1940). Important works dealing with frogs are, as follow: genus Telmatobius — Vellard (1946), Callardo (1962), Laurent (1970, 1973, 1977), Veloso and Trueb (1976), and Cei (1977); genus Gastrotheca — Laurent (1967, 1969a,b, 1976); genus Bufo— Gallardo (1967), Cei (1968, 1972). Barrio (1965) dis- cussed the Hijla pulchella complex; Duellman and Veloso (1977) reviewed Pleurodema, and Lynch (1978b) summarized data on Alsodes. The iguanid lizards have been studied by Cei (1971, 1973, 1974a,b), Cei and Castro (1973), and Donoso-Barros (1972). The herpetofauna of the southern Andes consists of 64 species (30 frogs, 31 lizards, 3 snakes). No genera are endemic to the southern Andes, but austral endemics such as Alsodes are shared with the lowland forests; Diplolaemus and Phymaturus are shared with Patagonia, and Garthia is shared with the Pacific coastal deserts. Of the 64 species, 36 are endemic. Of the 28 nonendemic species, 12 also occur in arid habitats east of the Andes, principally in Patagonia (e.g., the iguanid lizards Diplolaemus leopardinus, Lio- laemus bibronii, L. elongatus). Six others oc- cur in lowland Nothofagus forests (e.g., frogs such as Alsodes nodosus and Bufo variega- tus ) ; five occur on the arid Pacific lowlands (e.g., lizards such as PhyUodactylus gerrhopy- gus and Tropidurus peruvianus) . The faunal similarities between the central and southern Andes have already been discussed. Within the southern Andes, only two spe- cies, the toad Bufo spinulosus and the snake Tachymenis peruviana, occur in all five re- gions. Despite the high Andean divides, the southern Andes of Chile and those in Argen- tina share more species (10) than any other two regions (Fig. 15:27). Most of the en- demics are frogs of the genera Alsodes and Telmatobius restricted to separate stream drainages and highland lizards of the genus Liolaemus. The distribution of amphibians and rep- tiles on either side of the Andes in relation to passes through the high cordillera suggests that available structural habitat (or perhaps food) may limit their distributions instead of altitude and the associated climatic stresses. For example, at 33°S Lat. the cordillera is breached by Puerto Bermejo at 3883 m. In the immediate vicinity of the pass two lizards, Liolaemus altissimus and L. fxtzgeraldi, reach elevations of 3500 m and 2800 m, respectively, and occur on both sides of the Andes (Fig. 15:28). In the same area four other species of lizards occur only on the eastern slopes — Liolaemus elongatus (up to 2800 m), L. rui- 1979 DUELLMAN: HERPETO FAUNA OF ANDES 401 Fig. 15:27. Herpetofaunal comparisons of regions in the southern Andes. Numbers in blocks are numbers of species/number of endemic species; numbers of shared species are within arrows. AAN = Andes of Argentina north; AAS = Andes of Argentina south; ACN = Andes of Chile north; ACS = Andes of Chile south; ALT = Altiplano. Comparaciones faunisticas de la herpetofauna de las regiones de los Andes del sur. Numeros dentro de bloques representan numero de especies/ numero de especies endemicas; numero de species en comun estdn en las flechas. ball (2900 m), Phymaturus palluma (3500 m), and Pristidachjlus scapulatus (2900 m). Likewise, two species are restricted to the western slope — Liolaemus leopardinus (3000 m) and L. nigroviridis (3000 m). On the other hand, farther south there is continuity of habitat through much lower passes. For example, Patagonian scrub continues through Puerto de Buta Malin (37°30'S, 1800 m) onto the western side of the cordillera to Laguna 402 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 o UJ 2 ^ ?; & § 1 0. CHILE METERS 4000 - 3000 - 2000 ARGENTINA 1000 Fig. 15:28. Distribution of iguanid lizards on eastern and western slopes of the Andes in the vicin- ity of Puerto Bermejo, 33°S Lat. Distribution de saurios iguanidos en las laderas este y ocste de los Andes en la vecindad de Puerto Bermejo, 33°Lat. S. de La Laja, where such typical Patagonian species as Pleuroclema bufonina, Liolaemus kriegi, and Plujmaturus palluma occur. ANALYSIS OF DISTRIBUTION PATTERNS Once the major patterns of distribution in the various physiographic regions of the Andes have been described and documented, it is desirable to analyze the total Andean herpetofauna. The testing of the a priori divi- sion of the Andes into six major units shows that the faunal resemblance factors are all less than 0.1, except between the central Andes and the Southern Andes (Table 15:2). Thus, the recognition of the central and southern Andes as distinct major units is not so realistic as the distinction of the other units. Although the Huancabamba Depression has some en- demics, it shares eight species with the northern and eight with the central Andes. A cluster analysis of all 727 species in 27 regions emphasizes the close similarity be- tween the Cordillera Occidental in southern Peru and in northern Chile (Fig. 15:29). This analysis also shows the relatively close similarity of the eastern and western slopes of the Andes in southern Argentina and Chile, the similarity between the Altiplano and the Mantaro-Apurimac Valley and the similarity of the Cordillera Occidental in northern Peru with the upper Maranon and Huallaga val- leys. Likewise, the distinctness of the Vene- zuelan highlands. Sierra Nevada de Santa Marta, and Huancabamba Depression are evident. A second analysis involved only those 147 species having distributions above 2500 m (Fig. 15:30): this analysis eliminated four re- gions— Serrania de Paria, Cerro Turumiquire, Cordillera de la Costa, and Sierra Nevada de Santa Marta. The Merida Andes and the Cordillera Occidental in Colombia are distinc- tive in sharing no taxa with any other region. The other regions of the northern Andes cluster together and are weakly linked with the central and southern Andes and the Huan- cabamba Depression. The Andes of southern Chile and southern Argentina each has two endemic species and no species shared with any other region; thus, each region is distinct from all of the others. These analyses and knowledge of the physiography and environments of the Andes allow a general interpretation of the kinds and effectiveness of barriers to herpetofaunal Table 15:2. — Faunal Resemblance of the Herpetofauna in Six Major Andean Regions. Numbers of species in a given region are in boldface; numbers of species in common to two regions are in Roman, and the faunal resemblance factors t2C/(N, + Ns) = FRF] are in italics. Sierra Venezuelan Nevada de Northern Huancabamba Central Southern Andes Santa Marta Andes Depression Andes Andes Venezuelan Andes . 80 0.040 0.028 0.000 0.008 0.000 Sierra Nevada de Santa Marta ... __ 2 21 0.018 0.0(H) 0.000 0.000 Northern Andes 7 4 415 0.035 0.008 0.000 Huancabamba Depression 0 0 8 43 0.079 0.000 Central Andes 1 0 23 8 159 0.116 Southern Andes . 0 0 0 0 13 64 1979 DUELLMAN: HERPETOFAUNA OF ANDES 403 00 01 0.2 0.3 SIMILARITY 04 £ -C Serroni'a de Pana Cerro Turumiquire Cordillera de la Costa Merida Andes Sierra Nevada de Santa Marta Huancabamba Depression Cordillera Occidental, Colombia Cordillera Occidental, Ecuador Inter-Andean Basins, Ecuador Nudo de Pasto, Colombia Cordillera Central, Colombia Cordillera Oriental, Colombia Cordillera Oriental, Ecuador Cordillera Central, Peru Cordillera Onental North, Peru Cordillera Oriental South, Peru Cordillera Occidental North, Peru Upper Marandn Valley, Peru Upper Huallaga Valley, Peru Cordillera Onental, Bolivia Andes North, Argentina Mantaro-Apurimac Valleys, Peru Altiplano Cordillera Occidental South, Peru Andes North, Chile Andes South, Chile Andes South, Argentina 05 Fig. 15:29. Cluster analysis of 727 species of amphibians and reptiles in 27 physiographic regions of the Andes; analysis is by the unweighted pair-group method using arithmetic means. Andlisis de agrupaeion de 727 especies de anfibios y reptiles en 27 regiones fisiogrdficas de los Andes. El andlisis usa el metodo de grupos de parejas no compensadas en sus promedios aritmeticos. dispersal in the Andes (Fig. 15:31). The physiographic barriers in the northern part of the Andes coincide with those demonstrated for birds by F. Vuilleumier ( 1977 ) . The ma- jor physiographic barriers separate the major Andean regions — Merida Andes, Sierra Ne- vada de Santa Marta, northern Andes and central Andes — from one another, whereas minor barriers are within major regions — be- tween the Merida Andes and the Cordillera de la Costa, and between the Cordillera Occi- dental in Colombia and the rest of the north- ern Andes. The major ecological barrier is the drastic change from cloud forest to rela- tively dry puna, which follows the upper reaches of the outer ranges of the eastern cor- 404 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 £ Menda Andes Cordillera Occidental, Colombia Cordillera Central, Colombia Cordillera Oriental, Colombia Cordillera Oriental, Ecuador Nudo de Pasto, Colombia Cordillera Occidental, Ecuador Inter-Andean Basins, Ecuador Huancabamba Depression Cordillera Oriental North, Peru Cordillera Central, Peru Cordillera Occidental North, Peru Upper Marafi6n Valley, Peru Upper Huallaga Valley, Peru Mantaro-Apunmac Valleys, Peru Cordillera Oriental South, Peru Cordillera Occidental South, Peru Altiplano Cordillera Oriental, Bolivia Andes North, Chile Andes North, Argentina Andes South, Chile Andes South, Argentina 00 0 1 0.2 03 04 05 06 07 08 09 10 SIMILARITY Fie. 15:30. Cluster analysis of 147 species of amphibians and reptiles occurring only above 2500 m in 23 physiographic regions of the Andes; analysis is by the unweighted pair-group method using arithmetic means. Andlisis de agrupacion de 147 especies de anfibios ij reptiles que habitan solamente por encima de los 2500 m en 23 regiones fisiogrdficas de los Andes. Metodo al igual que figura 29. dillera in the central Andes. Comparatively broad latitudinal transition zones exist be- tween the cloud forest and the deciduous for- ests on the eastern slopes in northern Argen- tina, and between the Nothofagus forests of the southern Andes and the arid slopes to the north. Faunal comparisons of these eight ma- jor ecogeographic regions in the Andes reveal that the regions separated by major geo- graphical barriers or ecological differences have faunal resemblance factors of less than 0.1, whereas those separated by minor physio- graphic barriers have factors greater than 0.1 (Table 15:3). It is obvious that species richness is high- est in those tropical regions supporting both cloud forest and equable habitats above tree line — Cordillera Oriental in Ecuador ( 164 species), Cordillera Occidental in Ecuador (137 species), Cordillera Oriental in Colom- bia (103 species). Between the Equator and 24 °S Lat. there is a dramatic decline in spe- cies richness, most notable in amphibians; farther south there is little change (Fig. 15:32). Endemism is as high as 76 percent in the Sierra Nevada de Santa Marta, whereas the average percentage of endemic species in any 1979 DUELLMAN: HERPETOFAUNA OF ANDES T 405 Fig. 15:31. Diagrammatic map of the Andes showing barriers to herpetofaunal dispersal. Two solid lines = major physiographic barriers; single solid lines = minor physiographic barriers; broken lines = major ecological barriers; shaded areas = transition zones between ecologically different regions. A. Cordillera de la Costa; B. Merida Andes; C. Sierra Nevada de Santa Marta; D. Cordillera Occidental de Colombia; E. Northern Andes; F. Eastern slopes of central Andes; G. Nonforested central and southern Andes; H. Forested southern Andes. DEGREES SOUTH LATITUDE Fig. 15:32. Herpetofaunal species richness at dif- ferent latitudes in the Andes. Data were accumu- lated for species within one degree north and south of the latitude given; only species occurring above 1000 m are included. Riqucza de especie de la herpetofauna a diferentes latitudes en los Andes. Information fue acumulada para las especies dentro de un grado al norte y al stir de la latitud dada; solo las especies que ocurren por sobre los 1000 m estdn incluidas. Mapa diagramdtico de los Andes mostrando ba- rreras para la dispersion de la herpetofauna. Dos lineas continual = barreras fisigrdficas mayores; una linea continua = barrera fisiogrdfica menor; lineas discontinuas = barreras ecologicas mayores; areas sombreadas = zonas de transition entre regiones diferentes ecologicamente. 406 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Table 15:3. — Faunal Resemblance of the Herpetofauna in Eight Major Eco-geographic Regions of the Andes. Numbers of species in a given region are in boldface; numbers of species in common to two regions are in Roman, and the faunal resemblance factors [2C/(N, + N-) = FRF] are in italics. O 13 o O 2 C S O I o O 13 3 J3 c u O fa 6 V =3 +^ c O « 3 o > o A U 3 a* W 2 < 5 Bolitoglossa __.. 1 Elcutherodactijlus ... 1 Phnjnopus 1 Tclmatobius - Colostcthus 1 Atelopus .. Osornophryne „ .. Gastrothcca .. Hula 1 Centrolenella ._ Stenocercus _ Anadia 1 Pholidobolus .. Prionodactylus ._ Proctoporus Total Species 6 no species in common with sites in the other Cordilleras (Tahle 15:12). In the relatively simple paramo communi- ties, the differential utilization of resources was measured with respect to 1) distance from water, 2) utilization of rock cover, 3) diel activity, and 4) snout- vent length (larg- est adult male). The ratio of diurnal to noc- turnal species varies from 1:3 to 5:3; all of the reptiles and frogs of the genera Atelopus and Colostethus are diurnal. Using the Par- amo de Vigajual as an example, it can be seen that the seven species in the community are distributed throughout the spectrum of 430 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Table 15:12. — Comparison of the Herpetofaunas in Twelve Communities in Northern Andean Paramos. (Numbers of species in a given community are in boldface; numbers of species in common to two communi- ties are in Roman, and the Coefficients of Community are in italics. ) > XI "c5 "2 a o •o o OS o "3 bo o •o 3 ■2 Vi 00 O C 0) a H Oh H a! Oh ^ 3 O fC > O X £ Z w W CM ^ Oh Guantiva Rusia Vigajual Choachi Hermosas Purace NE Pasto E Pasto El Angel Papallacta .._. Mulalo Palmira 6 1.00 0.92 0.55 6 6 0.92 0.55 6 6 7 0.50 0.14 0.14 1 7 0.43 0.13 0.13 0.29 0.15 1 3 7 0.13 0.53 0.43 0.15 1 1 2 1 1 4 3 1 8 3 3 1 1 1 0.3S 8 4 2 1 1 0.40 0.53 7 3 2 2 0.14 0.29 0.46 6 2 2 0.14 0.14 0.31 0.33 6 5 0.14 0.14 0.31 0.33 0.83 6 resources; among the four small species of frogs (left side of Fig. 15:43), two are noc- turnal and two are diurnal. CENTRAL AND SOUTHERN ANDEAN COMMUNITIES Herpetological communities in the high central and southern Andes have only 3-7 species; this number increases to 10 in com- munities at lower elevations in the Patagonian transition zone in the Andean foothills in southern Argentina. Of the 21 communities analyzed (Table 15:13), four are on the east- ern ridges of the Cordillera Oriental ( Tapuna, "Malaga, Amaparaes, Acanacu) and have 60- 80 (x = 75.5) percent of the fauna composed of anurans. Another 12 sites are in the drier central Andes, Cordillera Occidental, and Alti- plano, where anurans comprise 33-75 (x = 57.5) percent of the fauna. Five sites are on the arid slopes of the southern Andes; 0-50 □ Fig. 15:43. Three dimensional plot of the multivariate means of seven species in a paramo community at the Paramo de Vigajual, Colombia. Axis I and III are the same as in figure 42; Axis II is association with rock, increasing from top to bottom. Solid symbols are nocturnal species, open ones are diurnal; circles are amphib- ians, squares are reptiles. Distribution tridimensional de los promedios multioariados de siete especies en mm comunidad tic pdramo en el Paramo de Vigajual, Colombia. Ejcs 1 tj 111 igual quo on la figura 12; l.jo 11 es asociacion con rocas, incrementdndose tie abajo hacia arriba. Los simbolos llenos representor! especies noctumas. los simbolos claros diurnas; circulos son anfibios, cuadrados son reptiles. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 431 Table 15:13. — Herpetofaunal Composition of Twenty-one Communities in the Central and Southern Andes. Genus o o 3 J3 o OS £ a 3 N IS 3 3 S o C 3 a en ^5 CO CD « cti ft E 3 o g ctl CO ct) O O _o "o o o CD CJ o u o 01 1- CD § at cj § 3 a) CO o « as c o c o s ctl '3 Sf ft ct) _C3 a .c to C ct) 0 o a! 'a! ft to -CD "c3 > 0 CD 3 O- O U ct! 'c? J a! s Hi 'X U H 'M < < H w w en Oh H U U A D J w J CQ Atehgnathus _ __ .. _ _ _ __ _ _ _. __ Phrynopus . _ _ _ 1 1 2 _ Pleurodema _ 1 1 1 1 1 1 2 2 1 1 1 1 _. __ _ 1 Telmatobius . 1 1 1 2 2 1 1 1 1 1 1 .. Bufo 1 1 _ _ _ _ __ 1 1 1 1 1 .. 1 1 ._ 1 Gastrntheca 1 1 1 2 1 1 2 1 1 1 1 1 Homonota 1 _ Diplclaemus .. 1 _ Liolaemm 1 1 1 2 1 1 2 2 2 1 3 3 4 1 4 Phymaturus .. _ .. _ .. _ __ __ .. _ 1 __ _ 1 1 Pristidactylus _ _ _ _ 1 1 .. _. Stenocercus ... ?, 1 1 _ Prnr.tnpnni.i 1 1 1 1 1 1 1 Leptotyphlops 1 Philodryas 1 _ .. Tachymenis _ 1 1 _ ._ Total Species 4 5 5 5 5 5 8 5 5 5 7 7 4 5 3 4 5 4 7 4 10 (x = 21) percent of the fauna is anurans. Lizards of the genus Liolaemus are conspicu- ous members of these communities, except in northern Peru and on most of the humid eastern ridges in central and southern Peru. Elsewhere in the central and southern Andes, 1^4 species of Liolaemus are present and ac- count for 14-75 (x = 37.8) percent of the species within each community. Within the central and southern Andes, comparative species composition of communi- ties apparently is a function of habitat and distance of sites from one another (Table 15:14). Some species, such as the frogs Phry- nopus cophites and Gastrotheca excubitor oc- cur only in the more humid sites on the east- ernmost ridges of the Andes, whereas toads (Bufo) and lizards of the genus Liolaemus are absent at these sites. Farther south in Argentina and Chile, the high uninhabitable backbone of the Andes is an absolute barrier to amphibians and reptiles; thus, species com- positions of sites at the same latitude but on opposite sides of the Andes are very different. However, in southern Argentina and Chile, where low passes exist in the Andes, species composition on the two sides of the Andes is more alike. Resource utilization was analyzed in the same manner as in paramo communities; again, it is noteworthy that species utilize a broad spectrum of resources within a given community. For example, the community at Santa Rosa on the Altiplano has seven spe- cies— five frogs, one lizard, and one snake (Fig. 15:44). All of the amphibians are noc- turnal, and the reptiles are diurnal. Among the five frogs, Telmatobius marmoratus is aquatic; Bufo spinulosus and Pleurodema marmorata deposit eggs in shallow temporary pools, whereas Pleurodema cinerea constructs a foam nest in ponds, and Gastrotheca mar- supiata broods its eggs in a pouch and sub- sequently releases its tadpoles into ponds. At the southernmost site (Laguna Blan- ca), three species of amphibians are closely associated with the lake (Fig. 15:45). Of the seven lizards, Homonota darwinii is noctur- nal. Two species of lizards (Liolaemus bib- ronii and L. darwinii) are associated with bunch grass, and the other four are associated with rocks. Of these, the herbivorous Phy- maturus patagonicus seeks shelter in crevices in extrusive basaltic rocks. The other three species are similar in their size and habitat (clumped at right of Fig. 15:45). Of these, Diplolaemus darwinii is carnivorous, Liolae- mus elongatus is primarily insectivorous, and L. kriegi is omnivorous, with more than 50 percent of its diet consisting of plants. 432 .S~ ° « o £• w .3. 'C S rt 3 O o IE > oonipEuitfnjj i— l i — l ■ — l i — l . — l ! h N H lO rt O co ■<* -* co co cm -** O M M o « n N d CO CO CO CO 1(5 © >o d O ■- co in ID CO N odd CM y— | rH l-H CM ^H »-H ^H io >o ■» s O O O O q >o ic ^ n i-i d d d d NHHHNNIMB SN O) 0")N CO CO 00 CO t^ CO © © o o © © co co co co CO CO CO CO M< CO CO CM CM CM CM © © © d © d © >-i © CO CO cm © lO CM CO CO CM CM CM CM o o o o o o MW^WC'lNNNNNlCtON CC(CCO(N(N^"HM»H ©©'©©©©©©d rt^USCMOOOt^t^CMOJOCM CO CO CO CM i-h •— < CM CM CM CM ooooooo'o'o'o r-l i-H 1» Ot^oOOt^c^CMO"MCM C0t--CoCOCM^H'"- also occur in Andes of Colombia, "" also occur in Andes of Colombia and Sierra Nevada de Santa Marta. Merida Cordillera Cerro Serrania Species Elevation Andes de la Costa Turumiquire de Paria Salamanders Bolitoglossa borburata ... 1000-1200 C Bolitoglossa orestes 2000-3500 SP Bolitoglossa savagei" - 2000 S Frogs Eleutherodactylus anotis — 950-1300 C Eleutherodactylus bimuculus 900-1200 C Eleutherodactylus boconoensis 2900 P Eleutherodactylus briceni 1620 C Eleutherodactylus ginesi 2800-1000 P Eleutherodactylus lancinii 2800-3000 P Eleutherodactylus mausii 100-1150 RC Eleutherodactylus orocostalis 1900-2100 — - C Eleutherodactylus racenisi 1900-2100 C Eleutherodactylus reticulatus 1275 C Eleutherodactylus rozei 1000 C Eleutherodactylus stcnodiscus .... — 1275 C Eleutherodactylus terraebolivaris 650-1800 RC Eleutherodactylus turumiquirensis .... 1675 C Eleutherodactylus urichi'"' 1000-2450 C C C Eleutherodactylus williamsi 1900-2100 C Atelopus cruciger . 200-1100 RC Atelopus mucubajiensis 2900-3100 Atelopus oxyrhynchus 2010-3500 CP Bufo stemosignatus 200-1800 RC Colostethus alboguttatus 1600-2000 C Colostethus bromelicola 1200 C Colostethus collaris 1500-1600 C Colostethus dunni 370-1650 RC Colostethus herminae 150-1650 RC C Colostethus mandelorum 2400-2630 C Colostethus meridensis 1600-1700 C Colostethus sp. "1"' 600 C * Species of Colostethus designated by numbers are being named in a manuscript by S. R. Edwards. 444 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Species Elevation Colostethus riveroi 600 Flectonotus pygmaeus'" 1075-1200 Gastrotheca nicefori*™ 1575 Gastrotheca ovifera _ 1000-1800 Gastrotheca sp. "F" 650-1100 Hyla battcrsbyi 1000 Hyla labialis"' _ _ 2000 Hyla platydactyla __ 1600-2500 Phyllomcdusa medinae ... 1100 Centrolenella altitudinalis 2400 Centrolenella andina 840-1050 Centrolenella antisthenesi _ _. 240-1100 Centrolenella buckteyi*"* 2000-2700 Centrolenella estevesi 1300-2400 Centrolenella flcischmanni' ° " — 1800 Centrolenella orientalis _. 1200 Centrolenella orocostalis 240-1200 Lizards Gonatodes ceciliae . 600 Gonatodes taniae 650-1100 Anolis jacare 1500-1800 Anolis nigropunctatus — 1200 Anolis squamulatus 200-1100 Anolis tigrinus 1100 Anadia bitaeniata 2500-3050 Anadia bhhei 1520-1830 Anadia brevifrontalis .. 2900-3600 Anadia rnarmorata 1100-2200 Euspondyhis acutirostris 1100 Proctoporus achlyens 1100 Proctoporus luctuosus 1100 Snakes Liotyphlops caracasensis 800-1100 Leptotyphlops affinis .. 1100-2000 Leptotyphlops macrolepis 200—1800 Atractus badius 400-2000 Atractus fuliginosus ± 2000 Atractus lancinii — . 1700 Atractus univittatus 800-1100 Atractus ventrimaculatus 1200-2000 Atractus vittatus 800-1800 Chironius monticola'" 1100-1600 Dendrophidion percarinatus 600-1600 Lampropeltis triangulum 1300-1600 Leimadophis bimaculatus" 00 1400-2500 Leimadophis zweifeli — 600-1700 BJiadinaea multilineata __ 800-2000 Umbrivaga mertensi .. 1000-1200 "Urotheca" williamsi .. 1400-2000 Micrurus mipartatus' " "' ..._ 800-2000 Bothrops medusa 1400-2000 Total Amphibians (48) Total Reptiles ( 32 ) Total Species (80) Merida Cordillera Cerro Serrania Andes de la Costa Turuniiquire de Paria c .... RC X .... .... S CP .... .... .... c c RC RC SP .... C c .... .... C .... RC .... .... .... RC .... c c .... c RC .... p C .... .... c P .... C C .... C .... C .... — RC cs .... RC .... cs RC c C .... RC cs RC c C c C .... c C cs c RC RC .... -*»- C .... C cs C .... C 18 25 5 3 14 21 2 1 32 46 7 4 1979 DUELLMAN: HERPETOFAUNA OF ANDES 445 Appendix 15:2. — Altitudinal and geographic distribution of amphibians and reptiles in the Andes of Colombia and Ecuador, and the eastern Cordilleras of Peru, Bolivia, and northern Argentina; all highland species in the Huancabamba Depression are included. Habitats: C = cloud forest, D = dry forest; P = paramo or puna, R = rainforest, S = subparamo, X = unknown. Species noted by ' are those diat are restricted to outlying ranges east of the Andes — Sierra de Macarena, Cordillera de Cutucii, Cordillera del Condor, Serrania de Sira. Species noted by " occur in the Cordillera Occidental in northern Peni. 6 "3 > c 2 -3 3 E 75 be u O w £ « < Species Elevation TJ 5 3 £ w„ "c o f ^ 7. ?. « * flj CJ (D a 5 T3 M 3 S o 2 F flj W U o C 3 3 2 O 1/5 c a , ^ -o -a -a a o o c o o o o O M 5 ' & 5 2 S '5 ,$ cc c C C ,3 H •C 'C m | o O OOO gg ^ IV ^ *U W -*? .4 ■-0 *- ■? 12 i? T3 ooo 3oo o o 5 •£ .3 OOO ZOO O O O £ X Caecilians Caecilia abitaguae 1100-1280 __ _ _ __ _ C .._ Caecilia antioquiensis 1980 C Caecilia attcnuata _ 1900 _ _ __ __ — C C Caecilia crassisquama 1400-1800 .... — .... __ — __ C Caecilia degenerata 1800-2200 _ C __ __ Caecilia occidentals 1200-1700 __ C _ — . Caecilia orientalis .... 1660-1935 _ . __ __ __ — .... C Caecilia pachynema __ __ 1200-2000 _ C C .... _ _ -. ~~ Caecilia parvipes 1650 __ __ — C Caecilia subdermalis ± 2500 _ — _ C Caecilia sp. "A" 1860-2150 _ C _ .._ _ _ _ Caecilia sp. "B" 2100 .- D _- .... .... -. ..., _ Epicrionops bicolor 1200-1400 __ C .... ... . _- — .... C Epicrionops peruvianus 1400 — _ — — . — - — - — C Epicrionops petersi .... 1150-1900 __ — — — . — — - C C Epicrionops sp. "A" 1980 __ —. _ — — C Salamanders Bolitoglossa adspersa 2500-3000 _ P — — BolUoglossa capitana 1780 — — . — — C Bolitoglossa hypacra 3610 — — P — — — - — - Bolitoglossa nicefori 1500 _ C — Bolitoglossa palmata 2000 .— _ — — C Bolitoglossa pandi 1300 — — — -— — - C ... . — Bolitoglossa phalarosoma 1500 — — C Bolitoglossa ramosi 1930 — — . .... C Bolitoglossa savagei 1000-2100 C _._. _ __ _ _ Bolitoglossa vallecula 2300-2700 .... _ _ S .... __ .... __ Bolitoglossa walkeri 1980-2050 — .- C .... .... .... _ Bolitoglossa sp. "A" . 2240 _ _ C .... __ .- — . _ Frogs Amblyphrynus helonotus .__ 1200 C __ _ _ ..- — Amblyphrymts ingeri _ 1720-1980 C _ .... C .... C _ __ Eleutherodactylus sp. "1"' 2660 — - - — — C ... Eleutherodactylus actites 2400 .— C Eleutherodactylus achatinus 140-1460 __ RC RC Eleutherodactylus affinis 2800 P Eleutherodactylus appendiculatus 1960-2010 .... C Eleutherodactylus areolatus 130-2010 _ RC RC .... .... _ .... _ Eleutherodactylus atratus 2200-1850 .... .... .... ... .... CS Eleutherodactylus balionotus 2700-2800 .... .... ... . — . S Species of Eleutherodactylus designated by numbers are being named by Lynch and Duellman (1979). 446 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Appendix 15:2 (Continued). 5 -5 ^ -5 5 S T3 Species Elevation § 5 3 u co d d _- t3 3 3 Js s u w ft, cj Is 2 "3 q~ CJ OJ Qj s £ s O T3 c > 3 O o < W on (1) H d .a n 0> z c 53 U ^ ^ ^ O O o U o 3 o 3 T3 c3 O w m O o _QJ J1J CL» JL> Q o U o O -d o O U ^ 3 Eleutherodactylus viridicans __ 2170-2680 Eleutherodactylus walked 220-1270 Eleutherodactylus w-nigrum 1230-2800 Eleutherodactylus sp. "A" . __ 1850-2160 Eleutherodactylus sp. "B" ._ 2100-2750 Eleutherodactylus sp. "C" 3200 Eleutherodactylus sp. "D" 3750 Eleutherodactylus sp. "E" . 2900-3000 Eleutherodactylus sp. "F" 3400 Eleutherodactylus sp. "G" 1230 Eleutherodactylus sp. "H" 1230-2010 Eleutherodactylus sp. "I" 1230 Eleutherodactylus sp. "J" 1540 Eleutherodactylus sp. "K" 1540 Eleutherodactylus sp. "L" 1960-2150 Eleutherodactylus sp. "M" .. ..... 1540-2010 Eleutherodactylus sp. "N" 2400 Eleutherodactylus sp. "O" 2010-2700 Eleutherodactylus sp. "P" 2700 Eleutherodactylus sp. "Q" 2700 Eleutherodactylus sp. "R" 2700 Eleutherodactylus sp. "S" ..... 1960-2700 Eleutherodactylus sp. "T" 2100 Eleutherodactylus sp. "U" _ 1460-1960 Eleutherodactylus sp. "V" 2120 Eleutherodactylus sp. "W" 2010-2580 Eleutherodactylus sp. "X" 1500-2580 Eleutherodactylus sp. "Y" 1740 Eleutherodactylus sp. "Z" 1740-2130 Eleutherodactylus sp. "AA" ..... 1740-2130 Eleutherodactylus sp. "BB" 2130 Eleutherodactylus sp. "CC" _ 1890 Euparkcrella lochitcs" 1550 Ischnocnema simmonsi" 1830 Leptodactylus labrosus 1000-1700 Leptodactylus wagncri 200-1820 Phrynopus brunncus .. 3000-3200 Phrynopus columbianus 1000-1300 Phrynopus cophites 3400-4100 Phrynopus flavomaculatus 2460-3100 Phrynopus laplacai .. 3400 Phrynopus nwntium 3400 Phrynopus nanus 2640-3400 Phrynopus parkeri" 2700-3100 Phrynopus peraccae 3100-3350 Phrynopus pereger .. 2400-3700 Phrynopus peruanus 3600 Phrynopus pcruvianus . . 2400-3700 RC CS CS CS s p s I' c c c c c c c c c c c c c c c c c c S CS CS C ~ D D D D C C C CD RC RC RC _ S CS SP I' S SP p CP 1979 DUELLMAN: HERPETOFAUNA OF ANDES 449 Appendix 15:2 (Continued). O T3 IS rt a! a c o 1 3 o W o "o O ^c3 S o 0 CIS o S o 6 o « . 3 w s .3 > "o 03 c < T3 3 O W I {/I 1* Species Elevation C CD 0> In c o U o" "c3 "c3 c ~C3 e aj "ffl c CIS .3

cs ca as Ch (« a) rt S3 z JD o 33 JZ T3 _4> jj Jj 0) 5 ^3 c« cd ■3 -a 3 o -0 •B -a T3 -5 -3 c OS o o O o o o o o -1-1 3 c/> U U U Z U U u U U X Phrynopus shnotisii' " 3050-3500 Phrynopus wettsteini 2000 c Tchnatobius ignavus 2000-2770 c Telmatobius latirostris 2000 c Tchnatobius niger . 2400-3500 SP s s Telmatobius vcllardi 2760-3050 - .... .... — — s s — — — s Telmatobius sp. "A" 2700-2850 Geobatrachus walheri ..... 1550-2870 CP Atelopus arthuri 2800 s Atelopus bomolochus 2500-2800 cs Atelopus boulengeri 900-1830 c Atelopus carrikeri .... 2350-4400 p Atelopus ebenoides 2660-3600 SP SP Atelopus elegans 300-1140 RC RC Atelopus erythropus ..... 1800-2500 c Atelopus halihelos -+- 1500 c Atelopus ignescens ..- 2900-4500 P P p s Atelopus longibrachus 800-1200 RC Atelopus longirostris ..... 500-2500 RC RC Atelopus mindoensis ... 900-2010 RC Atelopus nepiozomus 2000-3400 SP Atelopus nicefori 1800 C Atelopus pachydermia 2700-3100 s Atelopus palmatus ..... 1150-1740 c Atelopus pedimarmoratus ..... 2600-3100 S s Atelopus planispinus" -H2000 c Atelopus rugulosus .. 2100-2500 c Atelopus tricolor ..... 1700-2100 c Atelopus walkeri 1850-2160 c Atelopus sp. "A" 1950 c Atelopus sp. "B" .__ 2800 s Atelopus sp. "C" 2900 c Bufo chanchanensis 200-1460 RC Bufo fissipes ... 1800 c Bufo gnustae ... +2000 D Bufo inca . ... 1500-2000 c Bufo leptoscelis 1950 c Bufo limensis ..... 100-2200 D Bufo nesiotes" .... 1100-1280 c Bufo poeppigii 800-1670 RC Bufo quechua 2200-2600 c Bufo typhonius ..... 100-1840 RC RC RC Bufo veraguensis . 1300-1900 c c Osomophryne bufoniformis . 2700-3700 P SP P Osornophryne percrassa 3750 p Rhamphophryne macrorhina . ..... 1890-2130 c Rliamphophryne nicefori 2670 c PJiamphophryne rostrata 1890 _. _ c 450 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Appendix 15:2 (Continued). o -o 3 2 13 £ J= o U — f _r o « O _o Species Elevation is -g z ,i jj; ja a o o O as ^- 0) c 3 2 3 o u w "3 c S 3 oj rt ^ ■a S o o o o O o -o 3 Z o O u u o o o O W O) 0* 0; 1* at n 3 Colostethus sp. "2"" 1100-2400 Colostethus anthracinus 2500-3500 Colostethus sp. "3" _ 1410-1740 Colostethus sp. "4" 3700 Colostethus bolivianus 800-1200 Colostethus sp. "5" 900-1190 Colostethus elachyhistus .. 800-2800 Colostethus sp. "6" 1800-2900 Colostethus fraterdanieli 1900 Colostethus fuliginosus 480-1740 Colostethus sp. "7" 1900-2070 Colostethus infraguttatus ... 340-1270 Colostethus kingsburyi 1150-1410 Colostethus lehmanni 1900 Colostethus mertensi 2170-2680 Colostethus sp. "8" 1900-2350 Colostethus sp. "9" 2300-3500 Colostethus palmatus 900-2500 Colostethus sp. "10" .... 1500-1700 Colostethus sp. "11"" 1830 Colostethus ramosi 1240 Colostethus sp. "12" 1150-2790 Colostethus sp. "13" 2100-3500 Colostethus subpunctatus .. .. 2100-3300 Colostethus sylvatica .. . 1580-2970 Colostethus taeniatus _ 1740-2970 Colostethus sp. "14",> 1830 Colostethus vertebralis .. 2500-3200 Colostethus whymperi .. 1460-2120 Colostethus sp. "IS"00 1800 Colostethus sp. "A" .. 670-2130 Dendrobates abditus 1650-1700 Dendrohates lehmanni 850-1200 Dendrobates opisthomelas 1160-2200 Dendrobates viridis 200-1200 PhyUobates anthonyi .. 150-1690 Phyllobates bicolor . 25-1525 PhyUobates bolivianus 800-1200 Phyllobates tricolor _ . 1250-1770 Phyllobates sp. "A" _ 1300-1600 Amphignathodon guentheri 1200-2010 Cryptobatraehus boulengeri 1230-1700 Cryptobatraehus fuhrmanni . 1600-2550 Cryptobatraehus nicefori .. 1000-1200 Gastrotheca andaquiensis 2000 Gastrothcca argenteovirens 2850-3300 Gastrotheca aureomaculata 2300-2700 RC S RC SP DC CS C RC C C P C DC C C C .... c c RC RC RC SP SP RC S P SP SP c c c SP c c c c s cs cs c .... p RC D D D Species of Colostethus designated by numbers are being named in a manuscript by S. R. Edwards. 1979 DUELLMAN: HERPETOFAUNA OF ANDES 451 Appendix 15:2 (Continued). O 2 2 S-H a a! as 3 O w ■a o "3 U 2 'J3 J3 "o U 2 2 s o '3 S o 3 1-1 o a! 3 o w s 0-, ■2 > a be < O 08 o W a .9 W c a o T "c3 .1 D, Species Elevation -a "2 'o o 3 1 o" ■c c c c .Si c Q a) o o 0 as O o o O o aj > « OS rt P* a a) a) a rt a) Z Jlj 1- ED =3 jj JU it JJ J£ £ 3 a! fc -o -a o -0 -o T3 TJ T) 1-1 0) C as O O o 3 o O O o O C/5 U U U Z U U u U U X Gastrotheca cavia 2000-2890 s 2600 .... .... — .... .... — - — D D — Gastrothcca ehrysosticta 1530 Gastrotheca cornuta 200-1300 RC RC .... .... .... .... P .... .... .... Gastrotheca excubitor 3100-3550 Gastrotheca galeata 1740-2130 .... D Gastrotheca gracilis 1500-2000 .... .... .... D .... Gastrotheca griswoldi .... 3200-3800 P Gastrotheca helenae 3400 P Gastrotheca lojana .... 2100-2350 .... D Gastrotheca marsupiata 2760-4360 SP SP .... Gastrotheca medemi* 1140-1500 c Gastrotheca mertensi 2600-2900 s s s Gastrotheca monticola 1600-2500 s S D Gastrotheca nicefori 800-2100 c c Gastrotheca ochoai 2760-2800 D Gastrotheca peruana'"' .... _..... 2300-4600 Gastrotheca plumbea . 1300-2800 cs Gastrotheca psychrophila 2750-2850 s Gastrotheca riobambac 1800-4135 SP .... SP SP .... s Gastrotheca testudinea 1100-1840 .... .... c — c c c C c .... .... .... Gastrotheca weinlandii 1100-1800 Gastrotheca sp. "A" 1750-2600 cs .... .... Gastrotheca sp. "B" - 2170-2540 C .... .... .... Gastrotheca sp. "C" 1100-1600 c c .... .... Gastrotheca sp. "D" 2400 .... .... .... .... — c p — — .... Gastrotheca sp. "E" 3500-3800 Hemiphractus bubalus 300-1740 RC RC .... Hemiphractus fasciatus 300-1600 RC RC Hemiphractus johnsoni . 300-1910 c c C RC C Hemiphractus snutatus 250-1800 — .... RC — .... .... RC RC .... .... .... Hula albopu nctulata 300-1230 Hula ahitolulax 800-1540 c C Hula armata .. 1700-2400 C c Hyla balzani 1200-1840 .... C .... Hijla bogotensis .. 2500-2900 SP SP Hyla callipl euro _ 500-1800 .... .... c c Hyla carnifi jx 900-2010 c C c .... Hyla colu m biana 1700-2000 c .... .... Hyla denticulenta ... 1400-2400 c — - .... Hyla labialis .... 2400-3000 SP .... Hyla larinopygion 1900-2660 c c c Hyla lascinia .. 1700-2850 SP Hyla lindac 2600-2660 c .... .... Hyla phyllognatha 610-1740 c RC RC __ Hyla pulchella 500-3300 DS DS Hyla torrenticola .. 1400-1500 c C Hyla sp. "A 1750 c 452 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Appendix 15:2 (Continued). W Species Elevation § 2 1 Z 2, e E O O ^3 O U £ o "3 O O 3 z 2 2 O o 3 o W g o o w i: .B x; ■« a « O O *- *" t» hM M _ ID <0 V V JO jp o o O U o o O .a s 3 Hy/a sp. "B" 1750 Hyla sp. "C" 2300 Osteocephalus pearsoni 300-1620 Osteocephalus vcrrucigerus ... 500-1840 Phyllomedusa bakea° ___. 1280 Phyllomedusa boliviano 500-2000 Phyllomcdusa perinesos 1410-1490 Centrolene geckoideum 1920-2150 Centrolenella anomala 1740 Centrolenella antioquiensis 1890-2475 Centrolcnella audax 1490-1700 Centrolenella buckleyi .. ..... 2100-3400 Centrolenella cochranae . 1100-1410 Centrolenella flavopunctata 720-1800 Centrolenella fleisehmanni 200-1410 Centrolenella grandisonae 1460-2000 Centrolenella griffithsi .. 1200-2170 Centrolenella johnelsi 2500 Centrolenella mariae' 1550 Centrolenella medemi .. 1100-1490 Centrolenella megacheira 1490-1740 Centrolenella ocellata .. 1200-1820 Centrolenella ocellifera 700-2300 Centrolenella pellucida 1740 Centrolenella perkticta .. 1410-1460 Centrolenella pipilata ..... 1700-1740 Centrolenella prosoblepon 200-1410 Centrolenella siren .. 1200-1950 Centrolenella spiculata 1000-1700 Centrolenella truebae 1700 Centrolenella valeroi .. 200-1100 Centrolenella sp. "A" 1410-1490 Centrolenella sp. "B" .. 2660 Centrolenella sp. "C" .. 1140-1540 Centrolenella sp. "D" 1140 Centrolenella sp. "E" 1540 Centrolenella sp. "F" .. 1960-2150 Centrolenella sp. "G" 1960-2010 Centrolenella sp. "H" .. 1950 Centrolenella sp. "1" 1230 Centrolenella sp. "J" 1230 Centrolenella sp. "K" . 1230 Centrolenella sp. "L" 1750 Centrolenella sp. "M" 1750 Centrolenella sp. "N" 1750 Centrolenella sp. "O" 1050 Centrolenella sp. "P" 1200-1850 Glossostoma aequatoriale 2500-3615 c c SP SP SP RC RC C C C C '. RC C .... RC RC RC RC C C C C C C C C C SP .... .... .... RC RC RC RC .... .... C Z Z c Z Z Z c Z ..'.'. c c Z SP SP SP ... .... c ... .... ... c .... c c c c c c c c c c c c D c c c c c 1979 DUELLMAN: HERPETOFAUNA OF ANDES 453 Appendix 15:2 (Continued). Species O _Q as u eg 5 13 as 3 6 "o 3 E a! £ £ o 13 as S 03 Oh .5 > ■B O 13 M 3 3 3 -a « a W U i5 o u 3 U 3 O W o bo H <; w CD Elevation c s CD "2 'o 3 o O o" cd 3 CD c3 as 3 as 3 CD .3 pa as o O o O a) o O Eh o o o cfl ^1 > a PL| a! a OS as as a! z CD =3 13 CD 13 jd JSJ — JD a 3 J3 a! O a « 13 O 13 -a 13 13 -3 13 "3 M a 6 c a a> 60 H C _a> "C o O 13 rt 3 o W .3 rt & ft a 0) S O 3 1- 3 -a 1 -3 u 1 ■-3 c8 u C a o o 0 o o O o o £ C/5 o U U ^ U U u U U X Euspondylus spinalis _ 1200-2840 Euspondylus stenolcpis 2200 Macropholidus ruthveni 3100 Neusticurus cochranae . ~ 1150-1300 Neusticurus ccpleopus ... 200-1700 Neusticurus strangulatus . 1100-1410 Opipeuter xestus _ 1000-3000 Pholidobolus affinis .. 1800-3050 Pholidobolus annectens 2150-2335 Pholidobolus macbrydei - 2315-3960 Pholidobolus montium 2000-3190 Pholidobolus prefrontalis 2295-2885 Prionodactylus argus 100-1600 Prionodactylus dicrus — 600-1800 Prionodactylus manicatus 300-1750 Prionodactylus vcrtcbralis .. 700-3020 Proctoporus bolivianus 3300—1080 Proctoporus columbianus 2300-3000 Proctoporus guentheri 1000-3200 Proctoporus hypostictus ? Proctoporus lacvis 3000 Proctoporus mcleagris 2200-3150 Proctoporus oculatus 2640-3300 Proctoporus pachyurus 2900-3800 Proctoporus sirnopterus 2500-3000 Proctoporus striatus 2200-2500 Proctoporus unicolor 2800-3100 Proctoporus ventrimaculatus __ 2200-2700 Proctoporus sp. "A" .. 2300-2500 Proctoporus sp. "B" ..._ 2300-2500 Proctoporus sp. "C" .. 2400-3000 Ptychoglossus bicolor . 1700 Ptychoglosstis brevifrontalis .... ._ 200-1450 Snakes Leptotyphlops anthracinus 1100-1850 Leptotyphlops joshua 1700-1970 Leptotyphlops nicefori .... 1750 Leptotyphlops peruvianas 1400-1600 Leptotyphlops teaguei 2350-2700 Tropidophis taczanowskyi .. . 2500-2900 Atractus bocttgeri ... ? Atractus carrioni 2100-2275 Atractus craasicaudatus .._ 1700-2SOO Atractus ecuadorensis ... ? Atractus ernmeli _ - 2400 Atraclus indistinctus 1200 Atractus lasallei .... 2500 Atractus lelunanni 2600 P .... SP ... s .... RC RC X .... cs cs s .... RC RC RC ... C C SP . RC RC RC RC RC RC C C P P ... CS CS ... .... — CS CS s .... s ... .... D C C DC RC RC C C C C s X C D D I) D 1979 DUELLMAN: HERPETOFAUNA OF ANDES 455 Appendix 15:2 (Continued). 2 w u a 2 3 -2 3-3 Species Elevation 0) n 13 d o o C 13 O O > CO CJ C3 y o 0) ID T3 13 T1 IH o O O m U U U 3 h O c c o 13 « ti 13 _ 3 o O o 3 bo < W ■si 0) o O CO rt « as 1 c •£ a c c3 Q o a* o u a> o CQ s O r O 6 6 ■r 6 cd 0) V a> "o .3 3 3 _» 2 J2 11 Z "c3 "« ~d w 60 60 1 d s CIS 3 c c ■a Sh Ph 3 s a 3 < < u u 0) o cj ID ■c 3 < O V O) (U c V. 13 > "3 > -a -0 T3 X) O O O d O CO O c < c < ^ % 0) o s aj 60 6 U 1 3 W ti 3 B ti ■=3 u a O 3 o -3 o 0 1 o O -3 M O U a 3 -2 13. < o U ■3 o Z ■5 3 O C/2 -S •c 3 a < 2 X o 1) ^z — U - 13 -d T) T3 0 < 3 3 < < e e 8 s 0) > "3 3 5 3 0) J£ ~ "fi •a w 60 . 3 3 WJ o M 13 13 Tl T3 O O > o 3 < 3 3 < 3 0) c 1 )-. 60 6 1 3 )-H at H E 6 E T3 o o T3 13 i- o U ca a! 3 13. < o U ■5 o Z rS 3 O C/3 EL) -3 3 O C/3 4) O Z Stenocercus marmoratus 2750-3350 __ _ Stenocercus melanopygus __ 2700-3250 __ __ _ __ __ Stenocercus ochoai __ 2000-3000 _ __ A __ Stenocercus orientalis _ 2340 _ __ _ Stenocercus ornatissimus _. 2000-3400 _ A __ Stenocercus variabilis . _ 1200-3000 .... . .. AV Tropidurus peruvianas 0-2600 .... ... . — . ... . A Tropidurus tigris 0-2800 . .... A A Snakes Philodryas chamissonis 0-1800 — . ._. — . _ -— Philodryas taclujmenoid.es 0-3000 __ — . _ — . A Tachymenis affinis 2200-2400 .... __ A . Tachymenis attenuata .... 2500 _ ._. — . ... . Tachymenis peruviana 70^4570 ... . P P ..- AP Tachymenis tarmensis 3000 P Total Amphibians (45) 10 5 4 9 Total Reptiles (60) - 0 2 11 8 10 Total Species (105) 1 2 16 12 19 A .... A ~ .... NA _ .... _ A N N AP 8 9 5 2 3 7 8 4 12 12 16 4 11 15 13 12 20 21 21 P 8 ? P 16. Refugia, Refuges and Minimum Critical Size: Problems in the Conservation of the Neotropical Herpetofauna Thomas E. Lovejoy World Wildlife Fund 1601 Connecticut Avenue, N.W. Washington, D.C. 20009 USA I divide my remarks into a consideration of the contributions biogeography can make to conservation of the South American herpe- tofauna and a discussion of what constitutes the minimum critical size of a reptile com- munity such that it can maintain its structural integrity over reasonably long periods of time. I also note that conservation contains ques- tions of intrinsic scientific interest. The term refuge is used in the sense of a wildlife refuge — a conservation area or unit, whereas refugia is used sensu Haffer (1969) to indicate rem- nant forest patches that persisted when rain forest was turned to a savanna-like environ- ment during the drier portions of the Pleisto- cene. First, I raise the question as to whether the biogeography relevant for conservation purposes needs to include the traditional his- torical component? Isn't it enough just to know which species occur where? If one ignores history and does not con- sider the refugia, then one could look at con- tours of endemism, although both are defined by the same set of distributional data. Yet one could also consider contours of species diversity. Taking care not to leave out any species, one could choose the necessary num- ber of areas of high species diversity so as to conserve the herpetofauna. Based on a very simple model of the refugia concept in which all species are in refugia, one could argue for the approach of contours of species diversity. For if one chose the region where two refugia had come into contact, one would end up with a reserve con- taining the species of both. But as Haffer (this volume) showed, organisms seem to have varying abilities to disperse from re- fugia, and this results in a gradient of ende- mism that decreases away from the center of the refugium. Clearly this complicates the task of getting all of the herpetofauna into reserves unless one follows contours of en- demism as opposed to diversity. In an ex- treme case one might have to draw reserve boundaries so large as to include the foci of both refugia. More importantly, we need to recognize that in refuges encompassing refugia and in those encompassing secondary contact areas, or what Bemington ( 1968 ) termed suture zones, we are conserving two distinct kinds of biological communities. To the extent we be- lieve that the fundamental purposes of con- servation include conserving future oppor- tunities for biological research and maintain- ing evolutionary opportunities, it becomes important to conserve both. It is also important to note in setting up refuges of either sort we are in essence creat- ing refugia to the extent we allow remaining areas to become highly modified by man; and man then becomes a force on evolution of a magnitude similar to Pleistocene climatic changes. This really becomes a very serious question in terms of our own species manag- ing its future without knowledge of the con- sequences, and argues more strongly than ever before for development of wise land use policies. We need to recognize, too, that with all the best intentions and best design, reserves are not immune to influences originating out- side the reserves. Toxic substances immedi- ately come to mind. Moreover, in an area such as the Amazon, massive conversion to pasture and agriculture (more than one-third 461 462 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 of the primary forest has been cut) could have an important drying effect on climate,1 an effect that climatological science seems unprepared or unable to predict. If Brasil's plan to protect five percent of the Amazon in conservation units is thus imperiled, it well could mean that man's environmental effect would be yet greater than the magnitude of the Pleistocene climatic changes. The failure of paleoecologists to locate palynological evidence of the whereabouts of the deciduous forests of the northern United States during Pleistocene glaciations suggests that this forest may not be the highly co- evolved unit it was for so long thought to be. As a consequence we must exercise care not to leave government and nonscientific con- servationists with the impression that because we talk of our centers of endemism as refugia they are indeed likely to be refugia in times of future climatic change. It should be force- fully pointed out that refugia, at least those not associated with features of geological re- lief, may well have been in different places in different periods, and that dry periods varied in duration and perhaps even in temperature. The deciduous forest of the northeast or the Amazonian hyalea did not necessarily move around as a unit. Yet with respect to the Amazon forests one wonders how so many one to one relationships and intricate systems like that of the oropendolas — giant cowbirds — PhiJornis flies (Smith, 1969) could evolve if such were the case, or whether this then says something interesting about evolutionary rates in the tropics. Having clearly indicated that the histor- ical component of biogeography is important to conservation, we now need to consider how to choose conservation areas when refugia for different groups do not coincide. In many cases this will be because the size of centers of endemism varies with the taxon. For in- stance. Brown (1977), in a recent analysis of Nymphalid butterflies, proposed 38 refugia plus an additional 20 subrefugia, whereas Salati et al. (1978) and Villa Nova et al. (1976) estimated that slightly over half of the precipitation in the Amazon basin is forest-generated. Sufficient forest conversion presumably could initiate a drying trend. Vanzolini's (1973) reptilian refugia num- bered about six. Such differences raise inter- esting questions about the size of the grain of the environment for different kinds of organisms, and says to the extent that non- overlap merely reflects the abundance of refugia of a particular taxon, the job is merely to create more refuges. But suppose a taxon with few but large refugia, such as reptiles, has a refugium which does not overlap with a refugium from a second taxon with numerous, smaller refugia. Given continued belief in the refugia hypothesis, the lack of coincidence of refugia most likely indicates our data base is inadequate, or may suggest interesting dif- ferences in speciation in the different groups. The Brasilian government has taken the refugia approach in its recent document "Uma Analise de Prioridades em Conservacao da Natureza na Amazonia" (Wetterberg et al., 1977). choosing priority areas from overlays of refugia for heliconian butterflies, four plant families, birds, and lizards; highest priority has been assigned to areas with the greatest coincidence of refugia. The document has been included in the national development plan, and further was adopted by the six (non-Guianan) Amazonian nations at their second technical meeting on conservation in Brasilia in July 1977. The Brasilian govern- ment is to be applauded for taking biogeog- raphy into account, but it is incumbent upon us to keep them aware of all refinements in our knowledge. The problem of size of reserves has been considered recently by island biogeographers primarily using data dealing with birds. Their primary contribution to date has been to alert us to the problem that reserves can be too small to maintain their original species com- position over time. We really know very little about the subject, even though it is involved very fundamentally with the structure and function of ecosystems. Long before MacArthur and Wilson (1963, 1967) formally presented the theory of island biogeography, it was known that the number of species increases with area, and this relationship alone tells us we can make a conservation area too small. That there is a minimum critical size of an ecosystem is per- 1979 LOVEJOY: CONSERVATION OF HERPETOFAUNA 463 haps not so apparent. However, loss of spe- cies from too small an area obviously is inevi- table if only because low density species will occur in numbers too low to reproduce or to withstand stochastic fluctuations. Undoubt- edly there are other processes involved in the decay process of an ecosystem with a species number too large for the area left intact by man, but to date they have not been explored. An important question to test is one advanced by Lovejoy and Oren (1979), namely, whether this decay process leads to a predictable species composition of the im- poverished community. There has been sharp controversy as to whether more species can be preserved by a series of small reserves rather than a large one (e.g.. Diamond, 1975; Si'mberloff and Abele, 1976; Terborgh, 1976). Arguments that concentrate on simple numbers of species can be faulted for having tended to treat all species as equal, but more importantly for ignoring that what we really want to preserve are functioning ecosystems, not some kind of glorified zoos.2 Ecosystem protection will re- quire large areas or at least ones greater than the minimum critical size. Terborgh (1974) has estimated that for tropical rain forest birds, an area of about 1,000 square kilometers is needed in order to keep extinctions down to one percent of the original species complement per century. This size is probably in the right order of magnitude, but obviously minimum critical size will vary considerably with the taxon and ecosystem concerned. In designing reserves with ecosystem pres- ervation in mind, the minimum critical size to be taken is that which is the largest of those of all the taxonomic components. Yet there may be occasions when a reserve should be designed primarily for its herpetofaunal community. Further, there are probably in- teresting things to be learned by consideration of factors relating to minimum critical size in different kinds of organisms. Almost all the work on minimum critical ! This is not meant to denigrate in any fashion the important role zoos can play in education and cap- tive propagation. size of reptile communities has been carried out by Rruce Wilcox, who has kindly made his data available. His studies are primarily concerned with the lizard faunas of islands in the Gulf of California (Wilcox, 1978). These land-bridge islands can be dated as to age of isolation on the basis of the depth of the water separating them from the mainland and a knowledge of rates of sea level rises at the end of the Pleistocene. Consequently, these islands provide the first time sequence evidence of the decay process of super- saturated faunas of any sort. Wilcox also was able to estimate immigra- tion and extinction rates for these lizard island communities and found the former so low as to be negligible in his calculations. Ameiva ameiva and Mabuya mabuija not- withstanding, this would seem to make sense because lizards probably do not have the dispersal facility of birds. That the extinction rate is lower for lizards than birds may be reflective of a slower pulse to the dynamics of the poikilothermic system, or that poor dis- perses may be good persisters (Oren, pers. comm. ) . If immigration rates are low and oppor- tunities for dispersal to a conservation unit few, large areas are in order to maintain the original lizard biota. However, Wilcox (in preparation) suggests that the lower energy demands of poikilotherms should make higher biomass and densities possible, which, all other things being equal, would indicate that the area need not be as large as for a com- parable avian community. I am not aware of data on minimum criti- cal size for amphibian ecosystems, although I would guess it to be yet smaller and frequent- ly involving special features of the environ- ment. In any case, this leaves plenty for herpetologists to do, as long as our species doesn't reduce the possible future directions in which biological knowledge can grow by doing away with Neotropical ecosystems in the name of supposed progress. ACKNOWLEDGMENTS I acknowledge with gratitude the help of Jiirgen Haffer, Philip S. Humphrey, Ghillean 464 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 T. Prance, David C. Oren, Rruce A. Wilcox, and Richard O. Bierregaard. RESUMEN Gran parte de la herpetofauna asi como otras biotas sudamericanas solo sobreviviran en parques nacionales y reservas biologicas. Las bases biogeograficas para el disefio de dichas areas se nava en base a la teoria de los refugios. Las reservas designadas para rep- tiles y anfibios podran ser mas pequenas que aquellas para aves. LITERATURE CITED Brown, K. S., Jr. 1977. Centres de evolucao, re- fugios quaternaries e conservacao de patrimonios geneticos na regiao neotropical: padroes de difer- enciaeao em Ithomiinae ( Lepidoptera: Numpha- Iidae). Acta Amazoniea 7:75-137. Diamond, J. M. 1975. The island dilemma: lessons for the design of natural reserves. Biol. Conser. 7:129-146. Haffeh, J. 1969. Speciation in Amazonian forest birds. Science 165:131-137. Lovejoy, T. E., Oren, D. C. 1979. Minimum criti- cal size of ecosystems, in Burgess, R. L., Sharpe, D. M. (eds. ). Forest island dynamics in Man- dominated landscapes. Springer- Verlag. New York ( In press). MacArthur, R. H., Wilson, E. O. 1963. An equi- librium theory of insular biogeography. Evol. 17:373-387. MacArthur, R. H., Wilson, E. O. 1967. The Theory of Island Biogeography. Princeton Uni- versity Press. Princeton, New Jersey. 203 p. Remington, C. L. 1968. Suture-zones of hybrid in- teraction between recently joined biotas, pp. 321- 428 in Dobzhansky, T., Hecht, M. K., Steere, W. C. (eds.). Evolutionary Biology. Vol. 2. 452 p. Salati, E., Marques, J., Molion, L.-C. B. 1978. Origem e distribucao das chuvas na Amazonia. Interciencia 3:200-205. Simberloff, D. S., Abele, L. G. 1976. Biogeog- raphy theory and conservation practice. Science 191:285-286. Smith, N. G. 1969. The importance of being para- sitized. Nature 219:690-694. Terborgh, J. 1974. Preservation of natural diver- sity: The problem of extinction prone species. Bioscience 24:715-722. Terborgh, J. 1976. Island biogeography and con- servation: Strategy and limitations. Science 193: 1028-1029. Vanzolini, P. E. 1973. Paleoclimates, relief, and species multiplication in equatorial forests, pp. 255-258 in Meggers, B. J., Ayensu, E. S., Duck- worth, E. D. (eds.). Tropical forest ecosystems in Africa and South America. A comparative re- view. Smithsonian Institution Press, Washington, D.C., 350 p. Villa Nova, N." A., Salati, E., Matsui, E. 1976. Estimativa de evapotranspiracao na bacia Ama- zoniea. Acta Amazoniea 6:215-228. Wetterberg, G. B., Jorge Padua, M. T., Soares de Castro, C, Vasconcelos, J. M. C. 1977. Uma analise de prioridades em conservacao da natureza na Amazonia. Projeto de Desenvolvimento e Pesquisa Florestal. Instituto Brasileiro do Desen- volvimento Florestal, Ser. Teen. (8), PNUD/ FAO/IBDF/BRA-545. Brasilia, 62 p. Wilcox, B. A. 1978. Supersaturated island faunas: A species-age relationship for lizards on post- Pleistocene land-bridge islands in the Gulf of California. Science 199:996-998. Wilcox, B. A. In prep. Comparative biogeography of vertebrates. 1979 INDEX SUBJECT INDEX 465 Roman numerals refer to text, italics to maps and photographs. Adaptive types arboreal colubrid snakes 64 austral forest amphibians 361 lowland amphibians 203, 301 lowland reptiles 303 Africa faunal similarity with Australia 14 faunal similarity with South America 14, 55, 57 fossil record 57 herpetofauna 57 origin of herpetofauna 64 separation from South America 29, 55, 219 species/area 15 Altiplano 373, 378, 398, 399 climate 165 herpetofauna 400 tectonics 161, 400 vegetation 170, 399 Altitudinal distribution Central Andes 398 Guiana Highlands 252 Northern Andes 388 Southern Andes 350, 401 Andes 23, 158, 371 barriers to dispersal 165, 166, 382, 389, 402, 405 Central Andes 373, 394 climate 158, 164, 387 description 23, 157, 341, 371, 380, 381, 385, 398, 399 dispersal routes 412, 420 displacement of vegetation zones 111, 172, 177 distribution patterns 380, 387, 389, 396, 400, 402, 414, 415-416, 421-423, 425, 443 ecophvsiographic regions 402, 405 endemism 381, 383, 387, 389, 404 faunal similarities within 383, 388, 397, 401 faunal similarities without 406 geological history 157, 346, 387, 398 glacial climates 173 glaciation 151, 172, 400 habitats 373, 375-379 herpetofaunal communities 426 human modifications 374, 434 inter-Andean basins 373, 386 origin of herpetofauna 409 Quaternary history 172, 416 species richness 387, 397, 404 vegetation 169, 373 Areas of congruence 409 Arid Andean habitats 373, 375, 399 Atacama Desert 22, 23, 417 Aunt Arctica 73 Austral forests 20, 310, 341, 342, 374, 376 climate 344 distribution patterns 347, 349-354, 356 endemism 363 herpetofauna 347 origin of herpetofauna 355 physiography 342, 343 Quaternary history 178, 343, 346 vegetation 171, 309, 345 Australia 74 connection with New Zealand 100 endemism 98 faunal similarity with .Africa 14 faunal similarity with South America 14, 73, 100 fossil record 88, 97 herpetofauna 73 history of continent 74 origin of herpetofauna 77, 83, 91 species/area 15 Australopapuan fauna distribution patterns 77, 80, 84 origins 90 Avifauna distribution patterns 122, 124, 129 paramo islands 417 Barriers 166 Andean 165, 402, 405 Austral forests 360 Depresion de San Cristobal 382 Essequibo-Rio Branco Depression 264 Huancabamba Depression 389 inter-Andean valleys 164 Marafion Valley 167 Beringia 17 Biogeographical theories 55, 114, 254, 417 Brasilian Highlands age 163 climate 167 glaciation 180 Quaternary history 180 tectonics 163 vegetation 171 Caatinga 22, 299 Center of evolution 1 17 Centers of origin 70 Central America 16, 383, 407 connection with South America 16, 173 faunal exchange with South America 17, 221, 414 radiation of amphibians 208 Central Andes 373, 394 herpetofauna 389 physiographv 396 Cerrados 22, 299 ' Chaco 299, 300 Chaco-Pampean Plain 141, 142, 149 Chilean Archipelago 342 Chilean Lake District 344 Climate Altiplano 165 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Andes 158, 164 Austral forests 344, 347 Australia 75 Brasilian Highlands 167 Chaco 299 Cretaceous 30 desiccation of 20, 75, 114, 143, 151, 347 Early Tertiary 30, 44, 143 Guiana Highlands 168, 247 history of 182 Late Tertiary 20, 143 Llanos 281 Patagonia 151 Pleistocene 20, 108, 113, 150, 172, 347, 417 Postglacial 114 Quaternary 113, 144 Tierra del Fuego 165 Cloud forest description 373, 375 dispersal in 420 distribution in Andes 373 Colonizers 66, 77, 181, 206 Communities 426 Andean 428 cloud forest 427 Patagonian 431 rainforest 227 Conservation 23, 434, 461 Cordillera de la Costa 372, 381 herpetofauna 381 physiography 380 Depresion de San Cristobal 380, 382 Depresion de Unare 380 Dispersal center 120 Dispersal model 70 Dispersal routes Andean 412 Guianan 257 Huancabamba Depression 221, 389 lowland forests 202, 221 paramo 418 trans- Andean 121, 420 Distribution center 117, 118, 122 Distribution patterns Amazonian arc 198 Andes 380, 387, 396, 400, 402, 414, 415-416, 421-423, 425, 443 Anolis 126 Austral forests 347, 349-354, 356 birds 127, 130 butterflies 126 Chaco 302 continental 19, 21 Dendrobates 126 Guianan herpetofauna 249, 254-255, 258-259 Huancabamba Depression 389 languages 127 Llanos 290 monkeys 127 Oriental-Pacific 77, 80, 84 Patagonian 317, 321-325, 329, 332, 338 plants 125, 127 rainforest amphibians 187-188, 200-201, 203 rainforest reptiles 230, 232, 233 western Pacific islands 80, 82 Dry Andean forests 373, 376 East Indies dispersal routes 80 distribution patterns 77 Ecophysiographic regions 6 Andean 402, 405 distribution of families in 5 Endemism Andes 381, 383, 387, 404 Atlantic forest amphibians 201 Austral forests 341, 362 Australia 98 Central cis-Andean amphibians 199 Guianan region 249, 265 Huancabamba Depression 389 lowland tropical reptiles 225 Napo-Ucayali amphibians 198 Patagonia 325 trans-Andean forest amphibians 196 with respect to reproductive modes 206 Espinal 309, 310 Essequibo-Rio Branco Depression 264 Evolutionary radiation leptodactylid frogs 207 Liolaemus 327 telmatobiine frogs 327 tropidurine lizards 327 Evolutionary rates 20, 116, 131 Exploration of South America 1, 241 Extra-Andean mesetas 312, 326 Faunal exchange Africa with South America 56, 64 Central America with South America 17 West Indies with South America 18 Faunal origins Andes 409 Australia 77 Central America 17 Chaco 304 Guiana 254 Holarctic 56 India 56 Laurasia 64 lowland forest 207, 221 Oriental 77, 100 Patagonia 326 West Indies 19 Faunal similarity Africa 14, 55, 57 Andes 407 Atacama Desert 408 Australia 14, 73, 100 Central America 14, 407 Guiana 261 highland regions 406 Huancabamba Depression 389 North America 16, 44 Patagonia 317, 408 Sierra Nevada de Santa Marta 384 1979 INDEX 467 within Andes 383, 388, 397, 401 within tropical forests 190, 221, 229, 291 Fossil record Africa 57, 61, 63 amphisbaenians 62 Australia 88, 89, 97 birds 114 caecilians 57 crocodilians 61, 90 Europe 62 frogs 58, 89 Laurasia 61 mammals 114, 146, 148, 151 salamanders 57 snakes 63, 89 South American amphibians 312, 356 South American reptiles 219, 312 South American taxa, list of 51 turtles 61, 88, 97 Fossil sites 34, 34-38, 311 Galapagos Islands 17, 22, 364 Geological interpretations Amazon Basin 113, 287 Geomorphological interpretations 111, 183 Geomorphological units 160, 314-315 Glacial climatology, theory 181 Glacial phases ages 113, 173, 347 climate 108, 113, 147, 150, 173, 179 correlation with northern 146, 174, 178 number 146, 173 Glaciers Andean 151, 172, 400 Brasilian Highlands 180 Chilean 178, 343, 347, 359 Patagonian 146, 178 Gondwanaland breakup of 29, 55, 75, 219 distributions in 14, 56, 65 historical components of 16 Guiana Shield age 163 climate 168, 247 delimitation 242, 243-244, 246-248 erosion 164 geology 163, 242 herpetofauna 249 Quaternary history 180 vegetation 172, 247 Habitats in Andes 373 Habitat utilization amphibians 351 lizards 304 reptiles 351 Herpetofauna, South American geographic origins of families 4 fossil record 29, 51 review of families 2 taxonomic composition 3 Historical biogeographic analysis, method 118 Huallaga Valley 373, 393 Huancabamba Depression 22, 218, 221, 373, 389 Human modifications in Andes 374, 37.9, 434 Humboldt Current 20, 22, 99, 182, 417 Immunological distance and continental divergence 93 Immunological evidence of relationships hylid frogs 92 phyllomedusine frogs 8 Inabresia 56, 58, 64 Insect distributions and speciation 123 Inter-Andean Basins 373, 386 herpetofauna 388 physiography 386 Interglacial phases climate 147, 150, 173, 179 Karyological evidence of relationships hylid frogs 91 leptodactylid frogs 358, 360, 362 microhylid frogs 60, 85, 87 Lago Titicaca 175, 371, 399 Lost Pacific continent 100 Lithostratigraphic units 144 Llanos 281, 282, 284, 286 Mammal ages 144 Mammal distributions and speciation Amazon Basin 128 Mandibular musculature, frogs 81, 84, 86, 91 Mantaro-Apurimac Valley 373, 393 Maraiion Valley 22, 373, 393 Megafaunal extinction 114, 145 Merida Andes 372, 381 herpetofauna 382 physiography 380 Mesopotamia 141, 142, 148, 299 Montane rainforest 374 Monte 309, 310 Natural reserves 23, 434 forest refugia 132 size 462 New Guinea 74 connection with Australia 76 herpetofauna 81 New Zealand 74 connection with Australia 100 herpetofauna 99 Nonforests 22, 282 Northern Andes 372, 385 herpetofauna 387 physiography 384 Nudo de Pasto 372, 384 Oceanic dispersal 99 Paleofloras Maslin Bay 75 Patagonian 311 Tertiary-Chaco 19, 309, 414 Valdivian 311 West Gondwanan 19 Palynological interpretations 110, 163, 172 468 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Pampas 309 Pampean Ranges 161 Panamanian Portal 173, 221 Pantepui 246 Paramo 377 distribution 170, 374 floral composition 168 fluctuation in size 418 Paramo islands 417, 419 Paranense Sea 143, 149 Patagonia 141, 142, 147, 309 distribution patterns 317, 321-325, 329, 332 faunal regions 312, 313, 315, 318-319 herpetofauna 317 physiographic districts 318-319 steppe 20, 310, 316, 374, 377 vegetation 316 Plate tectonics Antarctica 29, 75 Antarctic-Australian suture 75 Australian-Oriental Plate collision 76, 89 Caribbean Plate 18, 29 Indian Plate 60, 70 Nazca-South American Plate collision 161 opening of South Atlantic 29, 55, 219 Pleistocene-Holocene boundary 144 Plio-Pleistocene boundary 144 Prehistoric man 128 Proto- Antilles 16 Puna 378, 399 distribution 170, 374 floral composition 170 Quaternary history Andes 172, 416 Austral forests 178, 346 Brasilian Highlands 180 Ethiopian Region 132 Guiana Highlands 180, 248 Neotropical Region 107, 219 temperate lowlands 141, 146 tropical lowlands 107, 219 Radar images 108, 112 Rainfall patterns 109, 248 Refuges 117, 120, 461 Refugia forest 116, 123, 124, 132, 139, 248, 359 location during arid phases 129 natural preserves 461 nonforest 116 speciation in 131, 202 Reinke's Corridor 189, 198, 260 Reproductive modes, amphibians 202, 206, 301, 353 Savanna gallery forest in 282, 283 grassland 108, 282, 282, 286 Guianan 246, 247 palm 282, 283 relicts in 218 varsea campos 108 woodland 108 Sea level changes, Pleistocene 110, 151 Secondary contact zones 118, 122 Serological evidence of relationships, Liolacmus 330 Serra da Mantiqueira 161 Serro do Mar 161 Serrania de Paria 372, 381 Serrania de Turumiquire 372, 381 Sierra Nevada de Santa Marta 372, 382 herpetofauna 383 tectonics 157, 382 Southern Andes 372, 399 herpetofauna 400 physiography 398 Speciation Andes 406, 424 Chaco 304 forest refugia 131, 202 Patagonian lizards 327 Speciation model 222 Species/area Africa 15 Australia 15 South America 15 tropical forest reptiles 227 Species richness Andes, southern Peru 397 latitudinal gradient in Andes 404 lowland rainforests 206, 231 northern Andes 387 Sub-Andean region 141, 142, 148, 299 Subparamo 379 description 374 distribution 374 isolation of 420 Symposium participants vii Taxonomic diversity Africa 15 Australia 15 South America 15 West Indies 18 Tepuis 244, 245 Tetrapods comparison of numbers between South America and world 3 Tierra del Fuego climate 165 herpetofauna 326 vegetation 171 Trans-Atlantic migration 56, 64 Trans-Pacific migration 11,99 Tropical dry forests 282, 287-288, 300 Tropical rainforests 20, 22, 108, J 09, 189, 190, 231, 374, 408 Amazonian 108, 227 amphibian faunas 190 Atlantic 190, 198 Central cis-Andean 189, 197 Chocoan 327 distribution of 22, 109, 231 Guianan 246 northern 190, 196 origin of fauna 207, 221 reptilian faunas 217, 225 trans-Andean 190, 195 1979 INDEX 469 Valle Central de Chile 342 Vegetation Andes 169, 374 Austral forests 345 Brasilian Highlands 171 Chaco 300 Guiana Highlands 172, 246, 247 Llanos 281 northern lowlands 281 Patagonia 316 Pleistocene 223 vertical shifts in Andes 111, 172, 177 Vicariance model 18, 70, 117, 409 West Indies endemism 18 origin of herpetofauna 19 origin of islands 18 taxonomic diversity 18 Western Pacific island faunas 99 dispersal routes 80 distribution patterns 82 470 MONOGRAPH MUSEUM OF NATURAL HISTORY TAXONOMIC INDEX NO. 7 All scientific names of families and lower taxonomic levels are indexed, except those trivial names appearing in appendices. Roman numbers refer to text, italics refer to figures and boldface to appendices. Acacia 171, 287, 300, 373, 374 caven 341, 342, 364 Acacioxtjlon 148 Acaena 173, 374 caespitosa 316 macrostemon 316 Acantholippia seriphioides 317 Achropogon 172 Acontinae 87 Acris 93 Acrochordidae 13, 88 Acrochordus 88 Adenomera 190, 206, 213, 214, 215, 269 andreae 197, 201 bokermanni 201 hylaedactyla 191, 197, 200, 201 marmorata 201 Adesmia 171, 316, 373, 374, 377, 399 horrida 170 Adiantum 345 Aegla 352 Aextoxicum punctatus 345 Afrixalns 61 Agahjchnis 8, 208, 212, 213 Agamidae 14, 15, 55, 57, 62, 65, 66, 71, 87 Agave 434 Agropyron magellanicum 316 Alligatoridae 31, 32, 33, 43, 44, 51, 52, 53, 54, 65, 220, 251 Allophryne 194, 214, 256, 270 nithveni 254 Alnus 163, 169, 173, 175 jorullensis 169 Alopoglossus 222, 227, 238, 278, 408, 410, 453 Alouatta senicuhts 128 villosa 128 Abodes 20, 325, 326, 327, 338, 341, 351, 361, 363, 365, 400, 407, 409, 410, 411, 412, 420, 457 gargola 361 gargola 320 neuquensis 321 monticola 347, 348, 351, 355, 361 nodosits 361, 400 pehuenche 361 verrucosus 361 Alsophis 12, 19, 22, 341 angustilineatus 364 cantherigenus 364 chamissonis 348, 350, 352, 356, 364 tachymenoid.es 364 Amapasaurus 238, 256, 278 tetradactylus 255 Amblyphrynus 383, 410, 411, 413, 445 ingeri 383 Amblyrhynchus 10, 17 Amcghinoa 317 Ameiva 18, 19, 123, 218, 219, 227, 238, 278, 296 amciva 289, 290, 291, 463 bifrontata 289 Amphibolurus 87 Amphichelydia 61 Amphignathodon 387, 410, 411, 413, 450 Amphignathodontinae 5, 8, 17, 61, 194, 204, 205, 208 Amphisbaena 11, 18, 218, 219, 222, 227, 238, 279, 296, 339 alba 291 camura 302, 304 fuliginosa 291 Amphisbaenidae 3, 4, 5, 11, 14, 17, 32, 54, 56, 63, 65, 251 Anadia 238, 372, 382, 407, 410, 411, 414, 420, 429, 444, 453 bitacniata 381 brevirostris 381 pulchella 383 Anaea 125 AnarthrophyUum desideratum 316 rigidum 316 Anemone 180 Anguidae 3,4,5,11.17,32,56,65 Aniliidae 3, 4, 5, 12, 32, 42, 51, 53, 251 Anilius 42, 218, 239, 274 Anisolcpis 238 Anodontolujla 87 Anolis 2, 19, 22, 218, 227, 238, 277, 278, 296, 327, 382, 383, 407, 408, 410, 413, 444, 453 annectens 289, 291 antonii 387 auratus 291 chloris 387 chrysolepis 123, 126 chrysolepis group 123, 124 fuscoauratus 222 nigropunctatus 381, 382 onca 289, 291 ortoni 222 punctatus 222 Anomalepidae 3, 4, 5, 11, 17, 18, 251 Anomalepis 11, 218, 222, 239 Anops kingii 302, 304 Aparasphcnodon 214, 215, 271 vcnczolanus 254, 257 Apodichelys htcianoi 51 Apodops 56, 57 /(nee! 51 Aporophis 274 Apmtolepis 13, 218, 227, 239, 274 Apuleia fcrra 171 Araucarta 151, 171, 180, 309, 316, 321, 334, 345, 316,374, 400, 11 1 angustifolia 171 araucana 345 Arcotorner 87, 194, 215 1979 INDEX 471 Areophryne rotunda 95 Aristelliger 19 Arracacia 174 Arthrosaura 238, 278 tatei 241 Ascaphus 3, 58, 100 Aspidites 89 Aspidosperma 300 Aster 169 Asterophryinae 79, 81, 82, 84 Asterophrys 87 Ateles fusciceps 128 geoffroyi 128 Atehgnathus 326, 338, 341, 361, 363, 431 grandisonae 326, 348, 353, 360, 361 nitoi 320, 325, 361 patagonicus 320, 325, 361 praebasalticus 320, 325, 361 agiZis 320 dobeslawi 320 /nisi 320 praebasalticus 320 rcvcrbcrii 325 solitarius 320, 325, 361 Atelopus 7, 192, 212, 213, 214, 270, 407, 410, 413, 416, 420, 429, 443, 449 carauta 387 carrikcri 383 crucigcr 197 ebenoides 428 ignescens 388 ignescens group 381, 411, 413, 418 mucubajiensis 381, 382 oxyrhynchus 381, 382 pulcher 198 varius group 18 walkcri 383 Atlapetes schistaceous 174 Atrarfiis 18, 64, 218, 222, 227, 239, 253, 264, 274, 275, 374, 382, 383, 407, 408, 411, 413, 444, 454, 455 /jarfius 382 insipidus 257 lehmanni 388 Atriplcx 316 AuZura 218, 238 Auracarites 316 AustTalobatrachus 7 ilius 89 Australocrinia 95 Austrocactus 316 Azara 180 Azemiops 64 Azora 309 Azorclla 170, 173, 316, 374 Baccharis 169, 170, 175, 374, 387 Bac/ita 18, 218, 227, 238, 278, 296 /wo/or 289 guiancnsis 289 tafpa 289 Balanerodus 43, 52 Baniiteria 309 «<;rtv/(7 173 Barycholos 194. 212 Basiliscus 10, 222, 238, 296, 408, 410, 453 galeritus 387 Batrachophrynus 312. 358, 361, 362, 396, 398, 409, 410.411, 412, 424,434,457 brachydactylus 358 macrostomus 358 Batrachyla 20, 2i, 96, 327, 341, 351, 352, 361 362 363, 365 antartandica 348, 3.55, 358, 360, 361 leptopus 347, 348, 351, 355, 358, 361, 362 taeniata 347, 348, 355, 358, 360, 361 Batrachylini 194, 204, 207, 361 Berberis 173, 180, 309, 374 cuneata 316, 317 Bignonia 309 Bipes 11 Blanus 11 Blechnum 345 Boa 18, 227, 239, 274, 297 constrictor 290, 291 Bogertia 62 Boidae 3, 4, 5, 12, 14, 31, 32, 42, 51, 52, 53, 54, 63, 66, 88, 100, 251 Boiginae 64 Boiga 64 Boinae 55, 63 Boini 56, 63 Bolitoglossa 2, 22, 191, 212, 213, 372, 407, 410, 414, 420, 421, 429, 443, 445 adspersa group 418 orcstcs 381, 382 savagei 381, 382, 383, 421 Bolitoglossini 56, 207 Boopis 180 Bothrernydidae 61 Bothrops 13, 14, 18, 218, 227, 239, 277, 297, 339, 407,408,411,444,456 altematus 303 ammodijtoides 317 bilineatus 222 lansbergi 289 ncuwiedii 303 schlegeli 387 Boulengerula 57 Bowdichia 284 Brachycephalidae 3, 4, 5, 7, 30, 60, 191, 192, 194, 204, 207 Brachycephalus 194, 214 Brachyclados caespitosus 317 Brachygnathosuchus 53, 220 Brachylophus 11,98 brevicephalus 98 fasciatus 98 Brasilotyphlus 194, 213, 273 Breviceps 87 Brifoa 62 Brevicipinae 60 Bromelia humilis 285 Bromus macranthus 316 />7r.m« 218,238,279 Brunellia 169 Bucttncria 309 472 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Bufo 4, 7, 39, 53, 54, 58, 189, 190, 192, 212, 213, 214, 215, 270, 295, 301, 338, 341, 351, 353, 362, 363, 400, 407, 410, 412, 426, 431, 434, 443, 449, 457 arenarum 59, 301, 302, 320, 326 chaguar 304 platensh 304 blombcrgi 59, 203 boreas group 412 calamita 312 cahmita group 4, 39, 312, 362, 363 chilensis 348, 350, 355, 358, 362, 363 coniferus 17 crucifer 199 glaberrimus 203 granulosus 23, 289, 290, 302, 303, 305 fernandezae 304, 305 major 304, 305 guttatus 203, 260, 290 guttatus group 203 haematiticus 18, 196 haenmtiticus group 59 ictericus 200, 201 marinus 4, 17, 33, 191, 195, 197, 201, 289, 290 marinus group 39 nasicus 254 paracnemis 148, 302, 303 regularis 59 rubropunctatus 348, 362, 363 spinulosis 20, 326, 362, 396, 400, 415, 431 spinulosis group 7, 407, 409, 411, 412, 420 stemosignatus 197 superciliaris 59 trifolium 398 hjphonius 18, 190, 191, 195, 197, 200 typhonius group 414 valliccps group 17 variegatm 312, 347, 348, 350, 355, 362, 363, 400, 412 viridis group 412 Bufonidae 3, 4, 5, 14, 15, 17, 30, 39, 44, 51, 53, 54, 56, 58, 59, 60, 65, 192, 204, 207, 251 Bijrsonima 284 Caecilia 9, 18, 57, 190, 192, 212, 213, 214, 273, 407, 408, 410, 445 thompsoni 196 tentaculata 191 subnigricans 191, 195 Caeciliidae 3, 4, 5, 9, 17, 30, 32, 51, 56, 57, 192, 207, 208 Caesalpinia echinata 171 Caiman 14, 33, 43, 52, 53, 218, 220, 227, 238, 274, 295, 303 crocodylns 289 latirostris 33, 43, 303 yacare 43 Calamagrostis 169 Calamelaps 64 Callicebus moloch 128 torquatus 128 Callicorini 125 Callopistes 22, 33, 41, 53 Callopsis 54 Calhiella 81, 82 Caltha 316 Calyptahyla 7, 18 Calyptocephalellinae 356 Calyptocephalellini 194, 357, 360 Campylorhynchas griseus 130 Candoia 89 Capparis 285 Cardamine 171 Caretta 273 Carettochelyidae 88, 101 Carettochelys insculpta 88 Carex 316 Cariamidae 148 Caryocar 127 amygdaliferum 127 amygdaliforme 127 dentatum 127 edule 127 gracile 127 nuciferum 127 pallidum 127 Caryocaraceae 125 Caudiverbera 20, 32, 33, 38, 52, 53, 311, 327, 341, 351, 356, 357, 358, 359, 360, 361, 363, 365, 412 casamayorensis 52, 356, 360 caudiverbera 52, 53, 312, 347, 348, 350, 352, 354, 355, 356, 358 Cebidae 143 Cebus albifrons 128 capucinus 128 griseus 128 Cecropia 171 Centrolene 387, 410, 411, 413, 452 Centrolenella 2, 17, 21, 190, 192, 196, 212, 213, 214, 215, 253, 264, 272, 273, 374, 381, 382, 383, 388, 406, 407, 410, 411, 413, 420, 427, 428, 434, 444, 452 albotunica 201 buckleyi 381, 382, 388, 420 divaricans 201 dubia 201 eurygnatha 201 fleischmanni 191, 382, 387 griffilhsi 387 prosoblepon 387 valerioi 387 vanzolinii 201 Centrolenidae 3, 4, 5, 8, 17, 22, 30, 56, 60, 192, 204, 206, 208, 251 Ceratophryinae 4, 5, 38, 96, 193, 194, 204, 207, 357 Ceratophrys 38, 53, 54, 213, 214, 215, 269, 295, 301, 302, 303, 304 aurila 39, 201 calcarata 197, 260 comuta 198, 201 ornata 38, 304 pierotti 302 stolzmanni 195 Cereidiu m 285 1979 INDEX 473 Cercosaura 227, 238, 278, 296, 408, 410, 453 Chaetophractus vellerosus 148 Chamaeleolis 19 Chamaeleontidae 55, 57, 62, 65, 66 Chamaelinorops 19 Changlosaurus 63 Chaperina 60 Characoidae 128 Charactosuchus 53, 220 fieldsi 43 Chelidae 3, 4, 5, 9, 14, 31, 32, 44, 52, 53, 97, 100, 251 Chelodina 97 Chelonia 273 Cheloniidae 251 Chelonoides 40, 53 Chelus 40, 52, 53, 218, 220, 238, 273, 296 colonibianus 40 fimbriatus 258 lewisi 40 Chelyearapookidae 98 Chehjcarapookus 98 arcuatus 98, 100 Chelydra 9, 218, 222, 238 Chelydridae 3, 4, 5, 9, 17, 18, 31 Chersydrus 88 Chiasmocleis 60, 214, 215, 273 Chimantea 172 Chiromantis 60 Chhonius 13, 18, 64, 220, 222, 227, 239, 275, 297, 408,411,444,455 bicarinatus 224, 225 carinatus 222, 225, 291 cxoletus 224, 225 pijrrhopogon 225 flavolineatus 224 foveatus 224, 225 /uscus 224, 225 monticola 381, 382, 388, 397, 420, 423 multivcntris 224, 225 quadricarinatus 224 scurruhis 224, 225 laevicollis 224, 225 Chondropython 89 Chrysemys 9, 18, 218, 220, 222, 296 scripts 289 Chrysobalanceae 125 Chtiwnerpeton 192, 194, 214 Chuquiraga 170, 316, 342 aurea 317 avellanedae 317 Chusquea 312, 321, 345, 374 penifolia 172 guj'Za 345 Cichlidae 128 Ctssus 309, 345 C/e/m 18, 219, 227, 239, 275, 297, 408, 411, 455 clelia 290, 303 occipitolutea 303 Clemantis 180 CZuria 169 Cnemidophorus 11, 18, 23, 218, 227, 238, 278, 296, 326, 339 lacertoides 304 leachi 302, 304 lemniscatus 289, 290, 291 lemniscatus 260 Cocus 171 Coleodactylus 10, 227, 277 meridionalis 222, 260 Colliguaja 342 integerrima 316 Colobodactylus 238 Colobosaura 227, 238 Colombophis 53 portai 42 Colostethus 2, 22, 192, 212, 213, 214, 215, 269, 372, 374, 380, 382, 383, 388, 406, 407, 410, 411, 414, 420, 427, 429, 443, 444, 450 abditaurantius 387 fuliginosus group 197 lierminae 382 mandelorum 381 Colubridae 3, 4, 12, 14, 32, 33, 42, 53, 54, 63 88 228, 251 Colubrinae 5, 12, 13, 17, 18, 56 Columbia corensis 130 picazuro 130 Coniophanes 18, 222 Conolophus 10, 17 Constrictor constrictor 302 Copernicia 282, 300 tectorum 282, 284 Cophixalus 83, 84, 87 Cophylinae 60 Corallus 18, 227, 239, 274, 297 enhydris 291 Cordicephalus 58 Cordylidae 14, 15, 57, 62, 66 Cornufer 78 Cortaderia 316 modesta 172 Cori/p/jospingus cucullatus 130 pileatus 130 Corythophanes 222, 238 Cremolobus 175 Crepidophryne 7 Cricetinae 145 Cricosaura 19 Crinia 95 haswelli 95 Crocodylidae 3, 4, 5, 14, 17, 31, 32, 33, 42, 43, 51, 53, 56, 61, 66, 100, 101, 251 Crocodilurus 218, 238, 278 lacertinus 258 Crocodylus 14, 18, 43, 98, 218, 220, 222, 238, 274, 295, 296 acutus 18, 220 intermedins 220, 257, 289, 291 johnsoni 98 novaeguinae 98 porosus 98 Crossochclys 52 Crossodactylodes 194, 215 Crossodactylus 194, 215 dispar 199, 200 474 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Crotalidae 251 Crotalinae 5, 17, 18, 42, 55, 56, 65, 66 Crotalus 14, 218, 219, 239, 277, 297 durissus 289, 291, 303 vegrandis 289, 291 Crotaphopeltis 64 Cnjtobatrachus 23, 253, 387, 408, 411, 413, 420, 450 boulengeri 383 Cryptodira 61 C 'rythiosaurus 63 Ctenoblepharis 326, 338, 407, 409, 410, 411, 415, 458 Ctenophryne 194, 214, 273 geayi 197 Ctenosaura 18, 19 Cupania 309 Cuprcssoides 374 Cupriguanus 348 Curatella 284, 286 Cyclagras 275 Cyclorarnphus 194, 214, 215 asper 200 Cyclorana 7, 91, 92, 93, 94, 95, 96 alboguttatus 93, 94, 96 australis 94 brevipes 94 dahlii 93 inermis 93 Cycloraninae 95, 96 Cyclura 10, 19 Cylindrophis 42, 88 Cynognathus 55, 64 Dasypeltinae 57 Dasyphyllum 172 Dasypodidae 149 Dasypops 195, 215 Dcndrobatcs 126, 192, 197, 212, 213, 269, 295, 408, 410, 450 auratus 195, 203 azureus 255 galactonotus 126, 203 histrionicus 126 leucomelas 126, 255, 257, 290 minutus 256 quinquevittatus 198 steyermarki 256, 257 tinctorius 126, 203, 255 tinctorius group 203 trunctatus 195, 203 Dendrobatidae 3, 5, 7, 17, 22, 30, 39, 56, 60, 192, 204, 206, 207, 251 Dendrophidion 13, 227, 239, 275, 408, 411, 444, 455 percarinatus 382, 387 Dendrophryniscus 7, 192, 213, 214, 270 minutus 198 Dcrmatonotus 60, 87, 193 muZ/en 301, 302 Dermochelyidae 251 Dermochelys 273 Dermophiinae 56 Dermophis 9, 192, 194, 212 Deschampsia 316 Diaphorolepis 18, 218, 222, 239, 407, 411, 413, 455 Diasemosaurus 52 occidentalis 312 Dichapetalaceae 125 Dicrodon 41, 54 Didynamipus 60 Digit aria 300 Dinilysia 12, 42, 63 patagonica 42, 312 Dinodon 88 Dipholostcphium 170 Diplodactylidae 96 Diplodactylinae 97, 100 Diplodactylus vittatus 97 Diploglossus 11, 19, 227, 238 Diplolacmus 219, 323, 326, 327, 338, 400, 408, 410, 431, 458 bibronii 320 darwinii 325, 431 leopardinus 321, 400, 420 Diplostephium 169, 171, 173 Dipsadoboa 64 Dipsadidae 251 D/p.sas 227, 234, 239, 274, 407, 408, 411, 455 indie a 233. Dischidodactyhts 194, 214 Discoglossidae 56, 59 Discoglossus 59, 60 Dispholidinae 64 Distichia 374, 387 Dolichotis 149, 150 DraZ>a 173 Dracaena 33, 41, 53, 218, 219, 220, 234, 238, 278 Drapetcs muscosa 178 Drepanoides 239, 275 Drimi/s 169, 171, 175, 176, 309, 345 brasiliensis 171 winteri 171, 345 Dromicus 12, 22 Drosophila 123 Drymarchon 13, 218, 227, 239, 275, 297 corai's 290, 291 Drymobius 13, 227, 239, 275, 297 Drymoluber 13, 239, 275 Dryoptcris 345 Duidea 172 Dusycion australis 148 Dyrosauridae 61 Dyscophinae 60, 81 Dyscophus 60, 87 Echimyidae 143, 148, 149 Echinosaura 18, 222, 238 Ecpleopus 238, 414 Edalorhina 194, 213 Elachistocleis 18, 60, 87, 193, 273, 295 fcico/or 301, 302 ot'o/is 290 Elapidae 14, 57, 66, 71, 88, 228, 251 Elapomorphus 13, 64, 218, 239, 275, 339 bilineatus 320 tricolor 303 Eleutherodactylini 4, 17, 194, 204, 206, 207, 208 Eleutherodactylus 2, 17, 18, 19, 22, 208, 212, 213, 214, 215, 253, 264, 269, 372, 374, 382, 383, 1979 INDEX 475 388, 389, 406, 407, 410, 416, 420, 427, 428, 429, 443, 445, 446, 447, 448 achatinus 203 acuminatum 200 affinis 418 alfrcdi group 208 biporcatus group 208 bicumulus 196 binotatus 200 binotatus group 201 boconoensis 381 bogotensis 418 bogotensis group 418 buckleyi 388 cajamarccnsis 389 carmelitae 383 chiastonotus 201, 203 conspicillatus 191, 203 curtipes complex 418 discoidalis 208 elegans 418 elegans complex 418 fenestratus 203 fitzingeri 196 fttzingeri group 201, 203, 208, 383, 413, 420 ginesi 381 inoptatus group 19 insignitus 383 lacrimosus 198 lancinii 381 longirostris 196 lynchi 418 maussi 197, 203 nicefori 418 nigrovittatus 201 ockendeni 200 peruvianus 203 prolixodiscus 383 pulvinatus 255 rani for mis 195, 197 rozci 196 sulcatus 203 sulcatus group 203, 413 terraebolivaris 196, 203 unistrigatus 208, 388 unistrigatus group 383, 411, 414, 418 uric/ii 382 vertebraK? 388 t'l'/ars! 203 w-nigrum 388, 420 zeuctotylus 201 EZosi'a duidensis 242, 253, 254 Elosiinae 4, 5, 193, 194, 204, 206, 207 Elseija 97, 98 Embothrium 309 Emmochliophis 218, 222, 239 Emydidae 3, 4, 5, 9, 31, 44, 53, 54, 61, 251 Emijdura 97 macquari 98 Engystomatidae 59 Emdius 218, 222, 239, 297 Enyalioides 18, 22, 222, 227, 238, 408, 410, 453 Enyalius 227, 238 Eocaiman 52, 53, 220, 312 cavernensis 43 Eoxenopoides 58 Eophractus 356 Ephedra 143, 316, 374 Epicrates 219, 227, 239, 274, 297 cenchria 290, 291, 302 Epicrionops 9, 192, 212, 213, 273, 408, 410, 445 Epilobium 171 Equu.s 175, 178 Erethizontidae 148 Erctmochphjs 273 Erichosaurus debilis 312 Erycinae 63 Erymnochelys 61 Erythrolamprtis 219, 227, 239, 275 Erythroxylon 309 Escallonia 169, 172, 176, 180 Espeletia 169, 170, 173, 374, 377, 387 Eublepharidae 96 Eublepharinae 62 Eucryphia cordifolia 345 Eugenia 169, 171 Eulychnia 171 Eumeces 11, 86, 87 Euncctcs 33, 42, 53, 218, 239, 274, 297 murinus 289 notaeus 303 Euparkerella 213, 215, 256, 269, 408, 410, 448 myrmecoides 201 Euphorbia 317 Euphractus 148, 149, 150 Eupsophus 20, 33, 38, 52, 3ii, 312, 327, 341, 350, 351, 360, 361, 362, 363, 365, 412 migueli 348 roseus 312, 347, 348, 352, 355, 358, 360 vanzolinii 348 vertebralis 348 uittafus 347, 348, 352, 355, 358 Euspondylus 23, 238, 253, 265. 278, 396, 407, 410, 411,413,414,444,453,454 leucostictus 241 Eusuchia 31, 33, 61 Euterpe 171 Fagara 309 Ferfuca 170, 316, 326, 374, 387, 399 monticola 316 Feylininae 87 Ficus 309 Fitzroya 309, 345, 374, 400 cupressoides 171, 345 Fleetonotus 23, 214, 215, 381, 408, 410, 413, 444 fissilis 200, 382 fitzgeraldi 382 pygmaeus 196, 382 F/i/ui'co/a nengeta 130 pica 130 Frankcnia 316 Franseria 171 Fritziana 195, 215, 413 Fuchsia 345 476 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Galbula cyanescens 127 galbula 127 melanogenia 127 pastazae 127 mficauda 127 ruficauda 127 rufoviridis 127 tombacea 127 Garthia 10, 22, 400, 408, 410, 458 Gastrophryne 8, 60, 87 Gastrophrynoides 60 Gastrotheca 8, 18, 23, 212, 214, 215, 372, 374, 383, 396, 400, 407, 408, 410, 411, 413, 416, 420, 424, 425, 426, 431, 434, 444, 450, 451 argenteovirens group 425 aureomaculata group 418 excubitor 398, 431 fissipes 199 mamipiata 398, 431 marsupiata group 418, 425 nicefori 381, 382, 387 ochoai 398 peruana 398 plumbea group 418, 425 riobambae 388 viridis 200 williamsi 96 Gaultheria 169 Gavialidae 31, 33, 43, 52, 53, 56, 61, 65 Gavialis 43, 52, 220 Gekkonidae 3, 4, 10, 14, 15, 32, 41, 51, 56, 62, 65, 96, 99, 100, 251 Gekkoninae 5, 17, 62, 96, 97 Gcntiana 173 Genyophrync 79, 81 Geobatrachus 8, 9, 410, 411, 413, 420, 449 walkeri 383 Geochelone 9, 18, 33, 40, 41, 52, 53, 54, 61, 220, 222, 227, 238, 296, 303, 322, 326, 339 carbonaria 33, 289 chilensis 33 denticulata 33 donosobarrosi 317, 320 gallardoi 40 gringorum 40 hesterna 33, 40 petersi 303 Geocrinia laevis 90 Geoemyda 40, 54 Geonoma 171 Geophis 218, 222, 239 Geotrypetes 56, 57 Geranium 169, 173 Cigantobatrachus 356 Ginkgo 309 Gkichnia 346 Glossarion 172 Glossostoma 18, 60, 87, 194, 212, 408, 410, 452 aequatoriale 193 Gochnatia 170 Gonatodes 18, 218, 227, 238, 277, 296, 408, 410, 444,453 vittatus 289 Grindclia chilocnsis 316 Grypiscini 4, 194, 204, 206, 207 Gryposuchus 53, 220 Guevina 345 avellano 345 Gunnera 172, 176 Gymnochacinus bergi 325 Gynmodactylus 10, 238 Gymnophthalmus 18, 218, 227, 278, 296 rubricauda 302 speciosus 289, 291 Gymnopis 9, 57 Gynograma 180 Gynoxi/x 169 Hamptophryne 87, 194, 214, 273 boliviano 198 Hegetotheriidae 149 Heleioporus 95 Heleophryne 4, 96 Heleophryninae 56, 57, 96 Heliconius 123, i24, 125, 131 erafo i26 Helicops 64, 218, 222, 227, 239, 275, 297 anguhtus 290 rfanie/t 289, 291 /iogei 257 scalaris 289 Helminthophk 11,297 Hemidactylus 10, 18, 62, 227, 238, 277, 296 brooki 62 mabouia 62 palaichthus 289, 291 Hemiphractinae 5, 8, 17, 193, 194, 204, 205, 207 Hemiphractus 8, 18, 212, 213, 214, 372, 407, 410, 411, 413, 420, 422, 451 bubalus 421 fasciatus 196, 387, 42i Hemipipa 195, 214 Herpelinae 56, 57 Heterodactylus 238 Homalopsinae 13 Homonota 62, 326, 338, 431 fcoreZ/ii 305 darwinii 305, 320, 325, 431 horrida 302, 304, 305 Hoplophryninae 60 Hordcum comosum 316 Hydracthiops 64 Hydrangea 345 Hydrodynastes 218, 239, 275 Hydrolaetare 194, 213, 270 schmidt i 197 Hydromedusa 32, 40, 52, 218, 238 Hydrops 218, 239, 275, 297 triangularis 257 tfy/a 7, 8, 18, 22, 39, 90, 92, 93, 190, 212, 213, 214, 215, 253, 271, 272, 295, 372, 381, 407, 410, 444, 451 acuminata 302 albomarginata 8, 191, 200 albopunctata 201 bcnitezi 256 boons 8, 191, 195, 197, 201 1979 INDEX 477 bogotensis 8, 381 bogotensis group 383, 407, 4 1 1 , 4 13, 420 brevifrons 198 catharinae 200 chinensis 91 crepitans 189, 289, 290 cbraccata 17 eg/en 191, 200 faber 199, 201 fasicata 198 fuscomarginata 199 fuscovaria 302 geographica 191, 197, 199, 200, 258 ginest 254, 256 /tayu 200 kanaima 256 /aiiafo 381, 382, 420 labialis group 411,413 hmciformis 198, 201 larinopygkm group 407, 411, 413 /cmai 254, 256 leucophijllata 8, 191, 197, 199, 200 leucophyllata group 61 licteocellata 191 microcephala 17, 289, 290 microps 200 mintiscula 289 minuta 191, 197 multifasciata 198, 201, 254 omatissima 254 pardalis 189 parviceps group 2i phlcbodes 195 platydactyla 381, 382 proboscidea 255 pulchella complex 400 raniceps 189, 302 rhodopepla 198 rodriguezi 255 rostrata 260, 289, 290 rostrata group 197 rubra 8, 190, 191, 195, 197, 200, 289 senicula 199, 200 mclanargyrea 259 sibleszi 254, 256 wandae 289, 290 x-signata 260, 302 Hylactophryne 208 Hylidae 3, 4, 7, 14, 15, 30, 32, 39, 55, 56, 60, 65, 73, 75, 89, 90, 91, 92, 93, 100, 192, 193, 194, 207, 251 Hvlinae 5, 17, 193, 194, 204, 206, 207, 208 Hylodes 195, 215, 270 duidensis 253, 254 gollmeri 191 marmoratus 241 Hyhphorbus 81, 84 Hylorina 20, 327, 341, 361, 363, 365 sylvatica 347, 351, 355, 358, 360 Hymenochirus 58 Hyophryne 195, 215 Hypericum 169, 173 Hyperoliidae 14, 55, 57, 60, 65 Hyperoliini 61 Hypopachus 8, 87 Hypothyris ninonia llypscla 316 726 Ichthyophiidae 57 Iguana 18, 41, 54, 218, 219, 227, 238, 278, 296 iguana 289, 291 Iguanidae 3, 4, 10, 15, 17, 32, 41, 51, 52, 53, 54, 55, 56, 57, 62, 65, 99, 123, 228, 251 Iguaninae 5, 17 Ikanogavialis 53 Ilchunaia 52, 220 Ilex 169 paraguariensis 171 Imantodes 227, 239, 275, 297 cenchoa 290, 291 Indobatrachus 70 Insuetophrymts 327, 341, 351, 361, 363, 365 acarpicus 348, 355 Iphisa 238, 278 Ischnocnema 208, 213, 215, 408, 410, 448 Jamesonia 180 Jtiglans 169 australis 169 Juncits 316 Kalophrynus 60, 83 pleurostigma 83 Kaloula 60, 87 Kankanophryne 95 Kentropyx 18, 218, 227, 238, 278, 296 calcaratus 222 lagartija 302 striates 291 riridistriga 302 Kinosternidae 3, 4, 5, 9, 17, 18, 31, 40, 251 Kinosternon 9, 218, 220, 227, 238, 273, 296, 303 fcauri 18 scorpioides 289 Kyarranus 96 Lacertidae 15, 55, 57, 62, 65, 66 Lachesis 219, 227, 239, 277 mi/ fa 13 Lagostomus maximus 301 Lampropeltis 222, 239, 408, 411, 444 triangulum 382 Laparrentophis 63 Larrca 316 ameghinoi 316 Laurclia 171, 309, 345, 346, 414 serrata 345 Leguminosae 125 Leimadophis 12, 218, 221, 227, 239, 275, 276, 297, 339,407, 408,411,444,455 bimaculatm 381 melanotus 291 regina 291 sagittifer 303 h/pWus 291 zweifeli 382 Leiocephatus 19 Lciopehna 3, 58, 100 478 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 archeyi 100 hamiltoni 100 hochstetteri 100 Leiopelmatidae 3, 4, 38, 44, 57, 58, 64 Leiosaurus 326, 327, 338 bcllii 41, 305, 324 catamarcensk 305 paronae 302, 304, 305 Lemaireocerus 285 Lepichina 180 Lepidobatrachus 301, 303 asper 302 llanensis 302 salinicola 302 Lepidoblepharis 10, 18, 227, 239, 277, 296, 408, 410, 453 ferfae 258 sanctaemartac 289, 291 Lepidochelys 273 Lcpidophyllum 170, 175, 374 Leposoma 227, 239, 253, 278, 296 Lepostemon 54, 63, 218, 227, 238 microceplialum 302 Leptodactylidae 3, 14, 15, 22, 30, 31, 38, 39, 44, 51, 52, 53, 54, 55, 56, 60, 65, 73, 90, 91, 92, 95, 96, 101, 192, 193, 194, 198, 207, 208, 251, 361 Leptodactvlinae 4, 5, 17, 193, 194, 204, 206, 207, 362 Leptodactylus 4, 17, 18, 39, 53, 54, 190, 200, 201, 203, 212, 213, 214, 215, 270, 295, 301, 303, 326, 338, 408, 410, 448 anceps 301, 302, 304 bolivianus 191, 195, 289, 290 bufonius 301, 302, 304 chaquensis 301, 302, 303, 304 fragilis 289, 290 fuscus 60, 260, 289, 290 fuscus group 203 gracilis 304 gualambensis 301, 302, 304 labrosus 195 laticeps 301, 302, 303 longirostris 203 macrostemum 289, 290 mystaceus 199, 201, 203 mystacinus 199, 304 ocellatus 304, 320 pentadactylus 190, 191, 196, 197, 199, 200, 289 poecihchilus 191, 195, 289 prognathic 304 rhodomystax 259 rugosus 260, 289 troglodytes 199 ventrimaculatufi 203 wagncri 190, 191, 195, 197, 199, 200, 289, 290 Leplodeira 64, 218, 219, 227, 239, 276, 297, 389 annulata 290, 291 Leptomicrurus 13, 277 Leptophis 13, 64, 219, 221, 227, 240, 276, 297, 408, 111,455 ahaetulla 290, 291 copci 257 Leptorrhampus 220 Leptotyphlopidae 3, 4, 5, 12, 14, 17, 56, 63, 65, 251 Leptotyphlops 19, 218, 227, 240, 274, 297, 339, 407, 408,411,431,444,454 albipuncta 302 dimidiatus 289 unguirostris 302 weyrauchi 302 Leucheria 171 Liasis 89 Liboccdrus 309 Libycosuchidae 61 Llerasia 174 Linmodynastinae 96 Limnodynastes 89 dumerili 90 tasmaniensis 90 Limnomedusa 361 Limnophis 64 Liolaemus 19, 20, 21, 131, 239, 305, 326, 327, 330, 338, 341, 350, 351, 352, 353, 363, 364, 374, 400, 406, 407, 409, 410, 411, 415, 420, 431, 434, 458 altissimus 327, 364, 400 archeforus complex 327, 332, 333, 334 archeforus 334 sarmicntoi 334 austromendocinus 330, 334 bibronii 320, 325, 327, 364, 400, 431 boulengeri 325, 327, 330 buergeri 321, 330, 334, 364 ceii 320, 330 chacoensis 302, 304, 305 chilensis 321, 327, 348, 351, 354, 364 cyunogaster 327, 341, 348, 349, 351, 364 brattstrocmi 348, 349, 354 cyanogaster 348, 354 darwinii 322, 327, 330, 431 elongatus 320, 321, 327, 330, 331, 332, 334, 400, 431 elongatus 330 pctrophilus 325, 330 fitzgcraldi 400 fxtzingeri 330, 334, 363, 420 canqucli 320, 327, 330 fitzingcri 320, 327 melanops 327, 330 fitzingeri complex 327, 32S, 329, 330 fuscus 327 goctschi 330 gracilis 327 Xraicnhorstii 364 /cm^ii 327, 332, 333, 334 feriegi 320, 321, 330, 334, 402, 420, 431 fcriegi complex 327, 33J, 332, 334 lemniscatus 327 leopardinus 364, 401 lineomaculatus 321, 334 lineomaculatus complex 327, 332, 334, 335 lorenzmiilleri 364 magellanicus 326, 334 magellanicus complex 327, 332, 334, 335 melanops 330 1979 INDEX 479 mocquardi 330 monticola 364 chiUanensis 364 villariccnsis 348, 351, 352, 354, 364 multiformis 396, 420 multimaculatus 330 nigroviridis 364, 401 ornatus 330 picftw 327, 341, 348, 349, 351, 364 pictus 348, 356 chilocnsis 348, 349, 356 ma/or 348, 349, 356 talcanensis 348, 349, 356 rothi 325, 327, 330 ruibali 401 ruizleali 325, 327, 330 schroderi 364 tenuis 321, 327, 349, 351, 364 tenuis 348, 353 punctatissimus 348, 350, 353 wiegmannii 304, 330 Liop/iis 218, 222, 227, 240, 253, 276, 408, 455 canaima 257 miliar is 260 Liotijphlops 11, 218, 227, 240, 274, 297, 408, 411, 444 Lithodijtcs 194, 214, 270 lineatus 198 Lftoria 7, 89, 90, 91, 92, 93, 94 alboguttata 93, 94, 96 amboinensis 91 atirea 79, 92, 94 caerulea 75, 89, 92 dahlii 93 darlingtoni 91 everetti 91 ewingi 90 freycineti 92 inermis 93 infrafrenata 91 peroni 91 pcroni group 91 raniformis 94 rothii 91 rubella 75 Livistona mariae 75 Lomatia 309, 345 Lophosoria 345 Loricaria 169, 170, 173 Loveridgchps 88 Lupinus 169, 170 Luzuriaga 345 Lycium amcghinoi 317 tenuispinosum 316 Lycodontinae 12, 57, 64, 66 Lycophidion 64 Lycopodium 172 fuegiana 178 Lygodactylus 10, 62, 65 Lygophis 12, 227, 240, 276, 297, 407 lineatus 291 Lygosominae 62, 87 Lyncodon 148, 150 Lysapsus 272 /ime/Zus 259, 261 /aeui's 259 Lysipomia 173 Lystrophis 240, 339 dorbignyi 303 semicinctus 303, 320 Mabuya 11, 18, 63, 65, 218, 219, 227, 239, 278, 296, 326, 339 /rcnafa 302, 320 mabuya 463 Machacrium 171 Macrauchcnia 175 Macrogenioglottus 194, 214 a/i'pi'oi 200 Macropholidus 389, 410, 411, 413, 454 Madtsoia 12, 42, 52, 63, 312 fca/ 52, 89 madagascariensis 89 Madtsoiinae 42, 100 Mammallaria 285 Mantipus 87 Masticophis 13, 222, 297 mcntovarius 291 Mastigodryas 13, 18, 218, 219, 227, 240, 276, 297, 408, 411, 456 bifossatus 290, 291 boddaerti 222, 290 ;;/eei 289 Mastodon 178 Mauritia minor 282, 2S5 Maytenus 309 Megaclosia 195, 215 Megalania 85 prisca 85 Megalonychidae 143, 145, 149 Megatherioidae 143 Megatherium 175 Meiolaniidae 4, 9, 31, 40, 51, 52 Melanerpes Candidas 130 Mclanobatrachus 60 Melanophryniscus 7, 192, 270, 413 moreirae 261 stelzneri 302 Melanosuchus 14, 33, 43, 53, 218, 220, 238, 274 niger 258 Melocactus 285 Memecyleae 125 Afesofcsena 218, 238, 256, 257, 279 Mesosuchia 43, 61 Mesotheriidae 149 Miconia 169, 172 Microcaecilia 213, 273 Microcavia 148, 150 Microhyla 60, 87 Microhylidae 3, 4, 5, 8, 14, 17, 30, 56, 58, 59, 60, 65, 71, 79, 83, 84, 86, 101, 192, 193, 204, 207, 208, 251 Microhylinae 60 Mucruridae 3, 4, 5, 13, 17, 18 Micrurinae 56, 64, 66 Micruroides 13 480 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Micrurus 13, 218, 227, 240, 277, 297, 303, 339, 408, 411, 444, 456 circinalis 289 dissoleucas 289 frontalis 303, 320 isozonus 289, 291 mipartitus 381, 382, 383, 384, 387, 388 surinamensis nattereri 257 Mimosiphonops 194, 214 Mixophys 94 Morelia 89 Morpho 125 Morunasaurus 222, 227, 239 Mourasuchus 53 Mouriri 125 Muehlenbergia 173 Mulinum 316, 374 spinosum 316 Sfus musculus 352 Mi/fma 170, 174, 175 Myersiella 195, 215 Mylodon 146, 178 Mylodontidae 145 Myobatrachidae 56, 65, 70, 73, 94, 95, 96 Myobatrachinae 95, 96 Myobatrachus 95 Myrceugenella 171 Myrceugenia 176, 309 Myrica 172 Myristica 309 Myrmecophagidae 143, 148 Myzodemdrum 345 Nassauvia 171, 316, 326 aculcata 316 Natricinae 13 Mrfra fasciata 18, 19 NebZfrui 172 Nectocaccilia 213, 214 Necrosuchus 43, 220, 312 tonensts 42, 51 Nectandra 309 Nectophrynoides 59 Neobatrachus 95 Neochelys 61 Neoprocoela 4, 38, 39, 52, 32 i, 312, 363 edentata 363 Nettosuchidae 31, 33, 43, 53 Nettosuchus 220 Neusticurus 218, 234, 239, 278, 408, 410, 454 racenisi 256 rudfa 241, 256, 257 totei 256 Nigerophis 63 Ninia 218, 227, 240, 276 Niolamia patagonica 51 Niphogeton 174 Nordophyllum 170 Nothofagus 75, 170, 176, 178, 223, 309, 312, 316, 321, 345, 346, 347, 352, 359, 360, 361, 362, 364, 365, 374, 376, 400, 404, 414 antarctica 171,345,346 betuloides 171, 346 dombeyi 171, 345 /usca 75 menziesi 75 obliqua 171, 345 procera 171 pumilio 171, 345 Nothopsis 218, 240 Nothrotheriinae 143, 149 Notobatrachus 3, 44, 58, 100 Notocaiman 220 stromeri 43 Notosuchidae 61 Nyctimantis 194, 213 Nyctimystes 7, 90, 93 Nyctimystinae 90, 93 Octodon degus 352 Octodontidae 149 Odatria 85 Odontophrynini 4, 194, 204, 207 Odontophrynus 193, 301, 326, 338 americanus 302 Oedipina 2, 192, 194, 212 complex 195 Ogmodon 88 Omoiotyphlops 63 Ophiodes 11,218,239 intermedins 302, 304 Ophryoessoides 222, 239, 296, 409 caducus 302 erythrogaster 289, 291 Opipeuter 396, 410, 411, 413, 414, 454 Opuntia 285, 287 Oreopanax 169, 309 Oreophrync 83, 84 Oreophrynella 7, 253, 255, 270 jnacconnelli 241 quelchii 241 Oscaecilia 9, 18, 57, 192, 212, 213, 214, 273 Osornophryne 192, 387, 410, 411, 413, 426, 418, 429, 449 Osteocephalus 21, 22, 213, 215, 272, 408, 410, 452 langsdorffi 200, 201 taurinus 197, 201 Osteopilus 7, 18 Otophrync 60, 87, 194, 214, 253, 255, 273 rofcusfa 241, 253, 254, 256, 257, 260 Oxybelis 64, 227, 240, 276, 297 aenci/s 290, 291 Oxyrhopus 64, 218, 227, 240, 276, 297, 408, 411, 456 pefo/a 291 rhombifer 303 Pachymedusa 208 Palacolama 151 Palaeophidae 63 Palaeobatrachidae 58 Paleosuchus 14, 218, 220, 227, 238, 274, 296 Palorchestes 98 a^aeZ 98 Pantacantha 316 Pantodactylus 227, 239 (i//en 241 Paracophyla 87 Paracrinia 95 1979 INDEX 481 Parahydraspis 220 Parastcphia 170, 399 Parlioplophrtjne 60 Parities 125 Parkia pcndula 171 Parvicaccilia 194, 212 Paspahim 284, 290 fasciculatum 284, 286 Paullinia 309 Pelobatidae 56, 59 Pclochclys bibwni 88 Pelodryadidae 7, 65, 73, 90, 92, 93 Pelomedusidae 3, 4, 5, 9, 14, 31, 32, 39, 44, 51, 52, 53, 56, 61, 64, 65, 66, 251 Pyocephalus 256, 273 Perezia 171, 172, 175 Pemethya 346 Persea 346 lingue 345 Peumus 309 Phenacosaunis 387, 410, 411, 413, 426, 418, 453 Plulodryas 227, 240, 276, 339, 408, 409, 411, 431, 456, 459 acstivus 303 baroni 303 patagoniensis 303, 320 psammophideus 303 simonsii 389 Philoria 96 Philornis 462 Philothamninae 64 Phimophis 218, 222, 240, 276, 297 guianensis 23, 289, 291 vittatus 303 Pholidobolus 372, 387, 410, 411, 413, 426, 418, 429, 434, 454 montium 388 Pholidosauridae 61 Phrynohyas 8, 212, 213, 215, 272, 295 coriacea 198 imitatrix 200 mesophaea 200, 201 oenufosa 190, 191, 195, 197, 200, 201, 259, 289, 290, 302 Phrynomantis 87 stictogaster 85, 87 Phrynomerinae 60 Phrynomerus 60, 87 Phrynops 218, 220, 227, 238, 273, 274, 296 geoffroanus 260 Phrtjnopus 372, 396, 410, 411, 413, 4i6, 418, 424, 429, 431, 448, 449 cophites 431 Phycoides 125 Phyllobates 190, 193, 212, 213, 215, 269, 408, 410, 450 femoralis 198 pirti/.s 191, 197, 200 pulchripectus 255 Phyllodactylus 10, 18, 22, 23, 97, 277, 296, 389, 408, 410, 458 dixoni 257, 289, 291 gerrhopygus 400 ventralis 291 Phyllodytes 213, 215 luteolus 199 Phyllomedusa 8, 190, 208, 212, 213, 214, 215, 272, 290, 295, 301, 303, 408, 410, 444, 452 bicolor 198 buckleyi group 18 fimbriata 200 hypocondrialis 302, 303 lemur 382 medinae 196, 382 sauvagii 302, 303 trinitatus 191 vaillanti 198 Phyllomedusinae 5, 8, 17, 90, 193, 194, 204, 206, 208 Phyllopezus 21 pollicaris 302 Phymaturus 326, 327, 339, 400, 408, 410, 431, 458 palluma 321, 324, 401, 402, 420 patagonicus 320, 321, 324, 364, 431 indistinctus 320, 324 nevadoi 321, 324 patagonicus 320, 324 payuniae 321, 324 somuncurensis 324, 325 zapalensis 321, 324 Physalaemus 4, 17, 190, 212, 214, 215, 270, 295, 301 albonotatus 302 biligonigertis 60, 302 enesefae 289, 290 pustulatus 195 pustulosis 191, 195, 260, 290 Pilgirodendrum 345 uviferurn 346 Pinguicula 171 Pijw 3, 18, 58, 194, 205, 213, 269, 295 parwz 3, 193, 289, 291 pipa 197, 290 Pipidae 3, 4, 5, 14, 17, 30, 32, 38, 39, 51, 52, 56, 58, 59, 60, 64, 65, 192, 193, 204, 207, 251 Pipinae 38, 207 Piptadenia peregrina 171 Pithccolobium 285 Placosoma 238, 414 Plagiobothrys 316 Planocrania datangensis 98 Platemys 98, 218, 227, 238 Platychcloides 61 Platychelys 61 Platemys 274 Platyhyla 87 Platymantis 77, 78, 79, SO batantae 77, 79 cheesmanae 79 gilliardi 77 meyeri 77 mimicus 77 myersi 77 papuensis 79, SI punctata 77, 79 vitianus 98 vitiensis 99 Platypelis 87 Platyplectron 95 482 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Plethodontidae 2, 4, 5, 17, 57, 191, 192, 204, 207 Pletlwdontohyla 87 Pleurodema 4, 18, 20, 21, 23, 270, 295, 301, 326, 338, 341, 362, 363, 372, 400, 407, 409, 410, 411, 412, 415, 431, 434, 457 horellii 302 brachyops 21, 23, 260, 290 bufonina 317, 320, 321, 325, 326, 364, 402, 420 cinerea 431 guayapae 302 marmorata 396, 420 nebulosa 326 thaul 321, 326, 347, 348, 350, 351, 352, 355, 358, 362 tucumena 302 Pleurodira 52 Pleurofora patagonica 317 Plica 234, 239, 278 Pliocercus 227, 240 Poa 170, 316, 326, 342, 374, 399 humilis 170 Podocarpus 75, 169, 171, 175, 178, 180, 309, 345, 373 andinus 178 lambertii 171 montanus 169 nubigenus 169, 171 oleifolius 169 parleteoreii 169 rospiglossii 169 salignus 346 Podocneminae 61, 65 Podocnemis 9, 31, 39, 41, 51, 53, 218, 220, 227, 238, 273, 285, 296 argentinetisis 52 bassleri 53 brasiliensis 51 elegans 39, 51 expansa 33, 39 uogfc 289, 291 Polychroides 389, 410, 411, 413, 414, 453 Polychrus 220, 227, 239, 278, 296, 408, 410, 413, 414 gutturosus 387 marmoratus 222 Polylepis 163, 170, 173, 174, 175, 374 australis 169 cocuyensis 169 sericea 169 Primula 171 Potomotyphlus 194, 213, 273 Prionodactylus 222, 227, 239, 279, 407, 411, 414, 429, 454 arguhis 397 manicatus 397 vertebrate 387 Pristidactylus 326, 327, 341, 364, 408, 410, 431, 458 achalensis 323, 326, 364 alvaroi 364 casuhatiensis 323, 326, 364 fasciatus 323, 326, 364 pectinatus 326 scapuhtus 323, 326, 364, 401 torquatus 348, 351, 354, 364 valeriae 364 vautieri 302, 304 Pristiguana 62, 219 brasiliensis 51 Proalligator 220 Proceratophrys 195, 214 fcoiei' 200 appendicular 200 cristiceps 199 Proctoporus 374, 398, 411, 413, 414, 424, 429, 431 434, 444, 454 /aet-i's 388 occulatus 388 Proctotretus 323, 326, 327, 339, 409 doellojuradoi 302, 304 pectinatus 320 Procyonidae 145 Proganochelydia 61 Prosopis 170, 171, 285, 300, 373, 374 /crox 170 patagonica 316, 317 Proterochersis 61 Proustia 171 Pseudemydura 97 Pseudemys 9, 238 Pseudhymenochirus 58 Pseudidae 3, 4, 5, 7, 30, 56, 60, 193, 251 PseuA.s 272, 295 paradoxus 260, 261, 289, 290, 302, 305 occidentalis 304, 305 plutcnsis 304 Pseudoboa 18, 218, 222, 227, 240, 276, 297 neuwiedi 289, 291 Pseudoeryx 218, 240, 276 Pseudogonatodes 10, 238, 277, 296, 383, 408, 453 Pseudonaja 88 nuchalis 88 Pseudopaludicola 270, 295 /wsiHa 25S, 260 Pscudophryne 95 occidentalis 95 Pseudotomodon 339 trigonatus 303 Pseusfcs 227, 240, 276, 298 Psidum 346 Psyllophryne 195, 214 Ptychadena 61 Ptychoglossus 227, 239, 279, 408, 411, 454 Ptychophis 240 Purussaurus 53, 220 Puya 170, 171, 173, 373 Pycnophyllum 170 Pygopodidae 14, 98, 100 Pi/f/ion 89 Pythonini 55, 57, 63, 66, 71 Quclchia 172 Quercus 163, 173, 175 Ramanella 60, 87 Ranfl 79, 82, 213. 214, 215, 269, 295 palmipes 8, 58, 190, 191, 193, 197, 200 pipicns 190 1979 INDEX 483 Ranidae 3, 4, 5, 8, 14, 15, 17, 30, 39, 55, 56, 58, 59, 60, 65, 77, 78, 101, 192, 193, 204, 207, 251 Ranidclla signifera 90 Ranunculus 171, 180, 316 Relictivomer IS, 190, 213, 214, 273, 295 pearsei 191, 195 Rliacophoridae 90 Rhadinaea 218, 227, 240, 276, 408, 411, 444, 456 lateristriga 388 Rliamnophis 64 Rhamphophryne 7, 18, 192, 213, 214, 372, 407, 410, 411,413,418,420,449 Rhamphostomopsis 33, 53, 220 Rlicobatrachus 96 silus 95 Rhinatrema 192, 194, 213, 256, 273 Rhinatrematidae 3, 4, 5, 9, 56, 57, 192, 207, 251 Rhineura 63 Rhinobothryum 227, 240, 276, 298 Rhinoclemys 9, 218, 220, 238, 273, 296 Rhinodcrma 7, 341, 351, 354, 359, 362, 363, 365 darwinii 347, 348, 352, 354, 355, 358, 362 rufum 348, 354, 355, 358, 362 Rhinodermatidae 3, 4, 5, 7, 30, 39, 60, 341 Rhinophrynidae 17 Rhizocephalum 173 Rhynchocephalia 99 Riolama 253, 279 Roupahi 171, 284 Roxoclielys 39 vilavdensis 51 wanderleyi 51 Rubiaceae 125 Saguinus 127, 128 inustus 127, 128 leucopus 127 midas 127, 128 ocdipus geoffroyi 127, 128 oedipus 127, 128 Salamandridae 56 Salomonclaps 88 Saltenia 58 ibanczi 38, 51 Sambucus 169 peruviana 169 Saphenophis 218, 240, 387, 411, 413, 456 Sapindus 309 Sarcosuchus 61 Sarmentia 345 Sajro£,'o(ea 345 Scaphiodontophis 222, 240 Scaphiophryninae 59, 60 Scelidontherium 175 Sccloporus 327 Sc/nnopsts 300, 309 Schinus 309 Schwoeboemys 61, 220 Scincella 1 1 Scincidae 3, 4, 5, 11, 14, 15, 17, 32, 56, 62, 66, 85, 101, 251 Scincinae 86, 87 Scincomorpha 62 Scolecomorphidae 14, 57 Scythrophrys 195, 214 Sebecidae 31,32,33,43,51,52,53 Sebecosuchia 43 Sebecus 51, 52, 53, 220, 312 icaeorhinus 98 Senec/o 169, 170, 172, 175, 326, 374, 387 filaginoides 316 Setaria 300 Sliclania 58 Sifcon 222, 240, 274, 298, 408, 411, 456 nebulata 290 Sibynomorphus 240, 408, 411, 456 oneilli 389 turgidis 303 Siphlophis 222, 227, 240, 276 Siphonops 190, 213, 214, 273 annulatus 191, 197, 199 paulensis 199 Smilisca 8, 194, 212 phaeota 195 stfa 195 Sminthittus 19 Somuncuria 326, 338, 361 somuncurensis 325 Sordellina 218, 240 Spartina 316 Sphoenorhynchus 213, 215, 272, 295 Sphaerodactylidae 96 Sphaerodactylinae 5, 17, 56 Sphaerodactylus 10, 19, 222, 239, 277, 296 Sphenodon punctatus 98 Sphenophryne 83 Sphcnophrvninae 60, 79, 81, 82, 84 Spitofes 13, 219, 227, 240, 276, 298 pullatus 290, 291 Stefania 23, 194, 214, 253, 255, 272, 408, 413 evansi 256 euansi group 253 goini 242, 253 goini group 253 marahuaguensis 254, 256 scalae 255 woodleiji 256 Stcgonotus 88 Stenocerctw 239, 372, 389, 396, 407, 409, 410, 411, 424, 429, 431, 434, 453, 458, 459 Stcnoglossa 59 Stenolepis 239 Stenopadus 172 Stenorrhina 218, 222, 240 Steradia 309 Stcreocyclops 87, 195, 215 Sti>o 170, 316, 326, 342, 374, 399 humilis 316 ichu 170 StrobUums 239 Strychnos 309 Stupendemys 53, 220 geographicm 40 Styllingia 316 Swallenochloa 169 Synapturanus 194, 214, 273 mirandaribeiroi 198 484 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7 Syncope 194, 214 carvalhoi 201 antenori 201 Synovitis 218, 227, 240, 387, 411, 413, 456 Syrrhophus 19 Tabebuia 309 Tachymenis 341, 365, 407, 409, 411, 415, 431, 459 affinis 365 attenuata 365 chilensis 341, 348, 350, 352, 353, 356, 364, 365 peruviana 20, 348, 365, 396, 400, 420 surinamensis 365 tarmensis 365 Tantilk 218, 219, 227, 240, 276, 298, 408, 411, 456 melanocephala 290 semicincta 289 Taphrosphys 40, 52, 220 Tapirus bairdii 128 terrestris 128 Tarentola 18, 62 Taudactylus 95 Tecoma 171 Teiidae 44, 62, 65, 228, 251 Teius 219 feyou 302, 304 cyanogaster 304 teyou 304 Tebnatobiinae 4, 5, 38, 60, 193, 194, 357, 361 Telmatobiini 4, 194, 361 Telmatobius 95, 312, 357, 358, 361, 372, 374, 389, 396, 400, 409, 410, 411, 412, 475, 420, 424, 429, 431, 434, 449, 457 culeus 434 marmoratus 396, 431 Telmatobufo 20, 312, 327, 341, 350, 351, 357, 358, 359, 360, 363, 365, 412 australis 341, 348, 354, 358, 360 hullocki 348 venustus 348, 352, 355, 358, 360 Tepualia 345 Testudinidae 3, 4, 5, 9, 14, 17, 31, 40, 44, 52, 54, 56, 61,251 Testudo 273 Thamnodynastes 227, 240, 253, 276, 298 strigilis 289, 291 Thamnophis 13 Thecadactylus 10, 18, 227, 239, 277, 296 rapicaudus 290 Thelotomis 64 Thoruciliacus 58 Thoropa 195, 207, 215 miliaria 200 Thrasops 64 Tibouchina 172 TilUmdsia 170 Todirostrum 123 Tomodactylus 19 Tomoptema delalandii 60 Trachipogon 282, 284 Trachyboa 12, 18, 222, 240 Trachycephalus 212, 215 nigromaculatus 200 Tretanorhinus 18, 218, 222, 240 Tretioscincus 218, 222, 2.39, 279, 297 bifasciatus 289, 291 Treuoa 342 patagonica 317 spinifer 316 Trichloris 300 Trichocereus 171 pascana 170 terscheckii 170 Trichocline 112 Trionychidae 4, 10, 31, 40, 53, 61, 88 Trionyx 10, 220 australiensis 97 Tripanurgos 219, 222, 227, 240, 276 compressus 222 Tritrinax 300 Tropidodryas 240 Tropidophiidae 3, 4, 5, 12, 17, 18 Tropidophis 12, 19, 222, 227, 240, 408, 411, 454 taczanowskyi 388, 389 Tropidurinae 5 Tropidurus 21, 22, 23, 218, 222, 227, 239, 278, 297, 302, 303, 304, 389, 408, 410, 459 bogerti 253 peruvianus 400 spinulosis 302, 304 torquatus 260, 289, 291 hispidus 253 Tupinambis 33, 41, 52, 53, 54, 62, 85, 218, 219, 220, 222, 227, 239, 279, 297, 303 rufescens 302, 304 tegumn 33, 290, 291, 304 Typhlina 12, 88 Typhlonectcs 212, 213, 214, 273 Typhlonectidae 3, 4, 5, 9, 56, 57, 192, 207, 251 Typhlophis 218, 240 Typhlopidae 3, 4, 5, 12, 14, 17, 56, 63, 65, 88, 101, 251 Typhlops 12, 19, 88, 218, 227, 234, 240, 274, 298 lehneri 289 Umbrivaga 411,413,444 mertensi 382 Ungaliophis 12, 222, 240 Uperodon 60, 87 Uracentron 234. 239, 278 Uraeotyphlus 57 Uranoscodon 239, 278 Uropeltidac 70, 88 Urostrophus 239 Urothcca 411,413,444 williamsi 382 Usnea 345 Valeriana 173 Vanzofcnus 194, 214 Varanidae 14, 57, 62, 71, 85, 101 Varanus 85 Verbena 316, 321, 326 ligustrinia 316 tridens 316 Vicraella 3, 44, 58, 100 1979 INDEX 485 Vilcunia 323, 326, 327, 339 sylvanac 326 Viperidae 3, 4, 13, 14, 32, 42, 54, 64, 228 Viperinae 55, 57, 65, 66 Waglerophis merremii 303 Wawelia 53,311 gerholdi 312 Weinmannia 169, 172 fagaroides 169 jahnii 169 microphylla 169 Weismania trichosperma 345 Werneria 59, 170 Wetmorena 19 Wonambi 12, 89, 100 naracoortensis 89 Xcnoboa 240 Xenodon 227, 240, 277, 298 Xenodontinae 5, 12, 13, 17, 18, 56, 64, 66 Xenopholis 227, 240 Xenopus 3, 38, 58, 95 pascuali 52 romeri 51 Zachaenus Zuccagnia 195, 214 170 OL657.A1 S68 1979 Hi! South \m i hcrpctofi ., i Harvard MCZ Library \l\lln 3 2044 062 372 826 Date Due Cfe^ A£R 1 5 1996 -m 3 1 ZQOZ WAY 2 0 2002- Contributors: Ana Maria Baez Jose M. Cei James R. Dixon » William E. Duellman • J. Ramon Formas • JoseM. Gallardo • Zulma B. de Gasparini • Jiirgen Haffer * Marinus S. Hoogmoed • Raymond F. Laurent • Thomas E. Lovejoy • John D. Lynch • Carlos Rivero-Blanco • Gustavo Ju in Scillato Yane • Beryl B. Simpson • Michael J. Tyler Cover design: Linda Trueb.