FIEL m .± NA Zoology NEW SERIES, NO. 39 STUDIES IN NEOTROPICAL MAMMALOGY Essays in Honor of Philip Hershkovitz Edited by Bruce D. Patterson and Robert M. Timm December 31, 1987 Publication 1382 PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY Information for Contributors to Fieldiana . and researc n.anala'il'is tVoin : ^pace permits. The Journal Ccirries a page charge ot $63 iiaction ihcivot. L oiiinnuiions iioiu suuT, research associales. and invited authors will be con- uion regardless of ahiliix to pav paee charges, but the full charge is mandatory for nonaffiliated authors of unsohcued manuscripts. Pav ; 'f page charges qualifies a paper for expedited p. mg. which reduces the publication time. Manuscripts should be submitted to Dr. Timothy Plowman. Scientific Editor, Fieldiana, Field Museum of Natural History, Chicago, Illinois 60605-2496, USA. Three complete copies of the text (including title page and abstract) and i.t the illustrations should be submitted (one original copy plus two review copies which may be machine copies). No manuscripts will be consideicd for publication or submitted to reviewers before all materials are complete ar hands of the Scientific Editor. Text: ^Manuscripts must be typewritten double-spaced on standard-weight. mch paper on all foursides. Forpapers longer than 100 manuscript pages, authors are requested lo submit a 'Table ol Coiucais. a •List of Illustrations." and a "Ust of Tables." In most cases, the text should be preceded by an "Abstract" and should conclude with "Acknowledgments" (if any) and "Literature Cited." All measurements should be in the metric The fonnai and style of headings should follow those of recent issues oi Fieldiana. For more detailed style imoiuui- won, see The Chicu.'o Manual ofStxlc (l.-^th cd. ), published by The University of Chicago Press, and also recent issues '.A Fieldiatui In '•Literature Cilcd." authors arc encouraged lo give journal and book uiics m lun. >micil a.u.iLMauw,,, .■.,. .ksirablc (c i: . in citation of synonymies), authors consistenily should follow Botanico-Periodicum-Huntianum and terauire by F A. Stafieu & R. S. Cowan (1976 et if^.) (botanical papers) ox Serial Source lu.'Ms uata nuM ^ 1983) published by the BioScicn'— ^nu^ynv.,u,^n S.n vie o. References should be typed in the following fo: rado Island. Stanford University Press, Stanford, Calif, 943 pp CiKUJii. K J.. J. R. Uuvi). .\ND T. U. Fknnington. 1963. A comparison of monta— - ' • i ■- .....-.,. •orest structure, physiognomy, and tloristics. Journal of Ecology, 51: 567-60! , H. J. M. ' among the Siona: Cultural pauems in visions, pp. fc.^-Si [/., eds., Sp: lis. and Stars. Mouton Publishers, The Hague, Nethedands 1. 1946. The historic tnbes of Ecuador, pp. 785-821. /// Steward. J. H.. ed.. Handbook ol South Aiucncan laaians llv Aiuk-an Cis ili/.aioiK. Bulletin !4.^. Rureau of American i-ihnolngy, Smithsonian Institution. Washington, D C 'Part II. Pol> Fieldiana: Botan\ Illustrations: Illustrations are relencu to m ihc text as "figures" uwt a^ -plates"). Figures must nc accompanicu nx some indication of scale, normally a reference bai. Statements in figure captions alone, such as " x 0.8." are not acceptable. Captions should be typed double-spa, nsecutivel - at issues oi r icticable, be 8 Illustrations should be mounted on boards m the anangemeni sou wi.sii lu obtain m liK ,, OIK 1 1! -t should be suitable for transmission to the printer as follows: Pen and ink drawings may iials (pre! otostats; shaded drawings should be originals, but within the s; >n; and photo- ,,i,id be li lack and V ^ All illustrations should be marKcu on ine reverse with ,, ,„^,. i, -top." Or rations will be returned to Uie author upon publication \uihors who wish to publish tigi juire costly special p;'. ith the Scientific Editor. s a two-step correction s vill norma ns, and changes can be made and queries answered. Unl> onc^clol page I IV n,^ nnis; h.- nv.u]r or 'Iw" sinL-Ic sol o\' wvzc pfoofs. Chanecs in page STUDIES IN NEOTROPICAL MAMMALOGY Essays in Honor of Philip Hershkovitz Phiup Hershkovttz FIELDIANA Zoology NEW SERIES, NO. 39 STUDIES IN NEOTROPICAL MAMMALOGY Essays in Honor of Philip Hershkovitz Edited by Bruce D. Patterson Division of Mammals Field Museum of Natural History Chicago, Illinois 60605-2496 Robert M. Timm Division of Mammals Field Museum of Natural History Present address: Museum of Natural History Department of Systematics and Ecology University of Kansas Lawrence, Kansas 66045 Accepted for publication July 30, 1985 December 31, 1987 Publication 1382 PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY HELDIANA: ZOOLOGY New Series, No. 39 Studies in Neotropical Mammalogy: Essays in Honor of Philip Hershkovitz Bruce D. Patterson and Robert M. Timm, Editors © 1987 Field Museum of Natural History Library of Congress Catalog Card Number: 87-82549 ISSN 0015-0754 PRINTED IN THE UNITED STATES OF AMERICA Table of Contents Preface vii A Biographical Sketch of Philip Hershkovitz, with a Complete Scientific Bibliography 1 Bruce D. Patterson A History of the Recent Mammalogy of the Neotropical Region from 1492 to 1850 11 Philip Hershkovitz A New Superfamily in the Extensive Radiation of South American Paleogene Marsupials 99 Rosendo Pascual and Alfredo A. Carlini An Additional 14-Chromosome Karyotype and Sex-Chromosome Mosaicism in South American Marsupials 1 1 1 Milton H. Gallardo and Bruce D. Patterson Notes on the Black-Shouldered Opossum, Caluromysiops irrupta 117 Robert J. Izor and Ronald H. Pine Feeding Habits of the Opossum {Didelphis marsupialis) in Northern Venezuela 125 Gerardo A. Cordero R. and Ruben A. Nicolas B. Notes on Distribution of Some Bats from Southwestern Colombia 133 Michael S. Alberico Distributional Records of Bats from the Caribbean Lowlands of Belize and Adjacent Guatemala and Mexico 137 Timothy J. McCarthy New Species of Mammals from Northern South America: Fruit-Eating Bats, Genus Artibeus Leach 163 Charles O. Handley. Jr. Seasonality of Reproduction in Peruvian Bats 173 Gary L. Graham Tent Construction by Bats of the Genera Artibeus and Uroderma 187 Robert M. Timm Comparative Ultrastructure and Evolutionary Patterns of Acinar Secretory Product of Parotid Salivary Glands in Neotropical Bats 213 Carleton J. Phillips, Toshikazu Nagato, and Bernard Tandler Distribution of the Species and Subspecies of Cebids in Venezuela 231 Roberta Bodini and Roger Perez- Hernandez Host Associations and Coevolutionary Relationships of Astigmatid Mite Parasites of New World Primates. I. Families Psoroptidae and Audycoptidae 245 Barry M. OConnor Notes on Bolivian Mammals. 2. Taxonomy and Distribution of Rice Rats of the Subgenus Oli- goryzomys 261 Nancy Olds and Sydney Anderson New Records and Current Status of Euneomys (Cricetidae) in Southern South America 283 Jose L. Ydhez, Juan C Torres-Mura. Jaime R. Rau, and Luis C. Contreras Morphological Variation, Karyology, and Systematic Relationships of Heteromys gaumeri (Ro- dentia: Heteromyidae) 289 Mark D. Engstrom. Hugh H. Genoways, and Priscilla K. Tucker Species Groups of Spiny Rats, Genus Proechimys (Rodentia: Echimyidae) 305 James L. Patton An Assessment of the Systematics and Evolution of the Akodontini, with the Description of New Fossil Species of Akodon (Cricetidae: Sigmodontinae) 347 Osvaldo A. Reig V Biogeography of Octodontid Rodents: An Eco-Evolutionary Hypothesis 40 1 Luis C. Contreras, Juan C. Torres-Mura, and Jose L. Ydnez Population Dynamics and Ecology of Small Mammals in the Northern Chilean Semiarid Region 413 Peter L. Meserve and Eric Le Boulenge Demography and Reproduction of the Silky Desert Mouse (Eligmodontia) in Argentina 433 Oliver Pearson. Susana Martin, and Javier Bellati Baculum of the Lesser Andean Coati, Nasuella olivacea (Gray), and of the Larger Grison, Galictis vittata (Schreber) 447 Edgardo Mondolfi Origin, Diversification, and Zoogeography of the South American Canidae 455 Annalisa Bert a Comparative Cytogenetics of South American Deer 473 Angel E. Spot or no. Nadir Brum, and Mariela Di Tomaso Faunal Representation in Museum Collections of Mammals: Osgood's Mammals of Chile 485 Bruce D. Patterson and Clare E. Feigl Taxonomic Index 497 Subject Index 505 VI Preface In the early 1 980s, we discussed the possibility of a testimonial volume for Philip Hershkoviiz with Larry Marshall, then with the Department of Geology, Field Museum. As the senior mammal- ogist at Field Museum and a student of South American mammalogy for almost half a century, Hershkovitz had generously provided invaluable advice and assistance to each of us in the early stages of our careers. We felt a Festschrift in his honor might repay a portion of our debts to him and, at the same time, serve as an independent, lasting tribute to his life-work. In the entire history of Field Museum, only three testimonial volumes had been produced in honor of museum scientists. Each recognized the contributions of men who were both preeminent scientists and museum administrators: Wilfred H. Osgood, Chief Curator of Zoology, 1921-1941; Karl P. Schmidt, Chief Curator of Zoology, 1941- 1956; and Rainer Zangerl, Chief Curator and Chairman of Geology, 1962-1974. Although Hershkovitz has never served in an upper-level administrative capacity, his contributions to the museum through distinguished and continuing re- search clearly qualified him for this honor. However, plans for a testimonial volume in Fieldiana: Zoology did not materialize until No- vember 1 983. By that time, Marshall had assumed a new position at the University of Arizona and Hershkovitz had just celebrated his 74th birthday. Given realistic editing and publication schedules, we were faced with the prospect of producing the volume nearly midway between traditionally cel- ebrated anniversary dates. Nevertheless, such tim- ing is somehow fitting: Hershkovitz the man is both extemporaneous and unconventional. Another notable departure from the Festschrift tradition is evident from the table of contents: Hershkovitz himself is a contributor! On many occasions Hershkovitz had lamented the lack of a historical review of South American mammalogy. During the present information explosion, scien- tists are hard-pressed to keep up with current de- velopments of direct relevance to their research; much less are they afforded the occasion to amble through historical records in Latin, German, French, Spanish, and Portuguese, even though these records are full of interesting and relevant infor- mation. As a result of his 50 years in the discipline, Hershkovitz may be unique in his broad knowl- edge of both historical literature and current re- search on Neotropical mammals. The editors therefore prevailed upon him to write such a his- torical survey to complement and enhance this volume. We convinced him that, by assembling a historical analysis of the subject, he would provide a tremendous service to younger workers. Other contributions to the volume came from friends and colleagues of Hershkovitz. All share an interest in the distribution, taxonomy, and nat- ural history of Neotropical mammals, and each one was inspired to honor Hershkovitz with their contribution. Each of the contributions focuses on those fields of Neotropical mammalogy to which Hershkovitz has contributed most significantly. We owe thanks to numerous persons connected with this volume. First and foremost, Tanisse Be- zin, Managing Editor of Field Museum Press, de- serves recognition. Her keen eye for grammar and style eliminated numerous editorial inconsisten- cies forwarded by the volume editors. Graham Harles, Field Museum Press copy editor, provided skillful editing and proofreading. The Scientific Editor for Fieldiana, Timothy Plowman, endured countless interruptions during production of this volume and served as corresponding editor for our own papers. Translations of abstracts into Spanish and Portuguese were kindly provided by Myriam Ibarra (an Ecuadorean ichthyologist) and Debra Moskovits (a Brazilian ecologist), who offered these as their own tributes to Hershkovitz. Assistance in assembling the indices was provided by Mary Anne Rogers. Finally, we are enormously indebted to a ded- icated body of reviewers, who critically evaluated papers in this volume. Their constructive advice and recommendations made editorial tasks far lighter. The editors gratefully thank: W. T. Atyeo, P. V. August, K. Benirschke, W. A. Clemens, J. A. Davis, W. B. Davis, M. R. Dawson, M. D. Engstrom, J. Fooden, G. L. Forman, M. H. Ga- llardo, A. L. Gardner, H. H. Genoways, W. E. Glanz, M. S. Hafner, D. Hunsaker II, R. J. Izor, J. A. W. Kirsch, K. F. Koopman, M. A. Mares, R. E. Martin, T. J. McCarthy, G. G. Musser, P. Myers, J. L. Patton, O. P. Pearson, R. H. Pine, W. B. Quay, L. Radinsky, O. J. Reichman, D. S. Rog- ers, R. W. Thorington, Jr., W. D. Tumbull, J. H. Wahlert, S. D. Webb, C. Wemmer, J. O. Whitaker, Jr., D. E. Wilson, R. G. Wolff, and A. E. Wood, in addition to anonymous reviewers of our own papers. B. D. Patterson R. M. TiMM Chicago, Illinois A Biographical Sketch of Philip Hershkovitz, with a Complete Scientific Bibliography Bruce D. Patterson Philip Hershkovitz was bom October 12, 1909, in Pittsburgh, Pennsylvania, to Aba Hershkovitz and Bertha Halpem. He was the second of four children and their only son. He attended primary and secondary schools in Pittsburgh, graduating from Schenley High School in February 1927. In 1929 he enrolled at the University of Pittsburgh where he majored in zoology, serving as an Un- dergraduate Assistant in that department during 1 930-1 931. Having exhausted Pittsburgh's course offerings in zoology and seeking to pursue a career in mammalogy, he was advised to transfer to another school with an expanded curriculum (Har- vard University, University of Michigan, or Uni- versity of California, Berkeley). In his junior year (1931), he transferred to the University of Mich- igan at Ann Arbor because of its proximity to his home. There he became an Undergraduate Assis- tant in the Museum of Zoology, working under the supervision of Professor and Curator Lee R. Dice during 1931-1932. He supplemented the meager earnings of this position with taxidermy jobs, which supported him during the early years of the Great Depression. His first fieldwork was undertaken during the summer of 1932. He went to the San Marcos re- gion of Texas to collect blind cave salamanders {Typhlomolge rathbuni) for Professor Uhlenhuth of the University of Maryland Medical School. Having a principal interest in mammals, he want- ed to collect small mammals in areas surrounding the caves, but Dice could spare no traps for him and told him to purchase some in Texas. While hitchhiking from Ann Arbor to Texas, Hershkovitz stopped to visit friends in Chicago. There, a chance visit to Field Museum of Nat- ural History secured him the traps and supplies he needed and seemingly set the course of his later career. Colin Sanborn, then Curator of Mammals during Wilfred Osgood's tenure as Chief Curator of Zoology (1921-1941), befriended Hershkovitz and loaned him the necessary supplies. As a con- sequence, the mammals that Hershkovitz collect- ed in Texas that first of many field seasons were deposited in the Field Museum collections. He now maintains that his chance visit to Field Museum in 1932 indelibly fixed that institution as the place at which to pursue his career goals. Hershkovitz's formal education was delayed by the worsening economic situation during 1 933. No longer able to afford tuition, he sought advice on subsistence during the Depression, and was told that Ecuador and Paraguay were undoubtedly the least expensive countries in this hemisphere in which to live. Transportation costs decided the issue, and in 1933 he set sail via the Grace Line from New York to Guayaquil, Ecuador for the whopping sum of $600, one-way. He stayed in Ecuador until 1937. During this period, he mastered Spanish and learned how to live off the land in the Neotropics. His boots dis- integrated after six months' time and thereafter he went barefoot. He assembled a fine collection of Ecuadorean mammals for the Museum of Zool- ogy, University of Michigan, supporting his activ- ities in part by selling horses bought on the Pe- ruvian frontier. He then returned to the University of Michigan where he again enrolled as an undergraduate, grad- uating in 1938 with a Bachelor of Science degree. By this time. Dice had moved from the Museum of Zoology to the Laboratory of Vertebrate Ge- netics, and William H. Burt had assumed the cu- ratorship in the Museum. Hershkovitz spent the years 1938-1941 as a graduate student enrolled at the University of Michigan, working on his Ecua- PATTERSON: BIOGRAPHY OF PHILIP HERSHKOVITZ dorean collection under Burt's direction. From 1939-1941, he was supported in this work by a Graduate Assistantship. In 1940 he received his Master of Science degree and immediately entered the doctoral program. Two years before the expected completion of his doctoral program, the Curator of Reptiles and Amphibians at the Museum of Zoology, Helen Gage, told Hershkovitz about the Walter Rath- bone Bacon Travelling Scholarship of the United States National Museum. This program was cus- tomarily reserved for postdoctoral support, but Mrs. Gage strongly urged him to apply immedi- ately. Thus encouraged, Hershkovitz submitted a brief proposal for work in the Santa Marta region of northern Colombia; his compliance with Mrs. Gage's wishes in this matter was so perfunctory that he failed to include a map of the proposed itinerary. But Remington Kellogg at the National Museum had long wished to obtain a Bacon Schol- ar for the Mammal Division and asked Hersh- kovitz to send the omitted material. Much to his surprise, Hershkovitz was awarded the scholar- ship and left Ann Arbor immediately for Wash- ington. He spent two months there studying the then very poor collection of Neotropical mam- mals. Afterward he spent two years in Colombia (1941-1943) collecting mammals, other verte- brates, and ectoparasites. The resulting collection was the National Museum's first large and repre- sentative Neotropical mammal accession. In 1943 Hershkovitz's work was interrupted by World War II, and he returned to Ann Arbor to enlist in the Armed Services. He was assigned to the Office of Strategic Services (OSS) and served from 1943-1946 in the European Theater. While serving in France, he met Anne Marie Pierrette, whom he married in 1946. The two returned to the United States, where in 1946 and 1947 he continued his Bacon Scholarship studies of Co- lombian mammals in Washington. The first of three children (Francine, Michael, and Mark) was bom in 1947. About this time, he was contacted regarding the opening of a curatorial position at Field Museum in Chicago, an opportunity he eagerly hailed for several reasons: ( 1 ) The comprehensive collections of Neotropical mammals at Field Museum would be a tremendous resource for what he had already decided would be his life's work; (2) he had the highest regard for W. H. Osgood, who as a prin- cipal authority on South American mammals would be a great personal resource on which to draw; (3) the press of family responsibibties made continuation of his graduate studies untenable; and (4) aspirations to a curatorial position had been the raison d'etre of his graduate program; a cur- atorial position made the graduate degree sujjer- fluous. Thus he jumped at the offer of employment at Field Museum, knowing full well that it marked the end of his graduate program at Michigan. Like many similar institutions, Michigan had a final year-in-residence degree requirement. Unfortu- nately, Osgood died in June of 1947, and what might have been a remarkably productive ap- prenticeship under Osgood never came to pass. Upon his arrival at Field Museum, Hershkovitz found an uncurated backlog of some four or five years of accessions. Nevertheless, he wasted little time in returning to the field, prompted in part by postwar housing shortages in Chicago. (One can almost hear him now, telling the Museum's Di- rector Clifford Gregg that the nearest affordable housing was in Bogota!) In 1 948 he and his family moved to Colombia where he resumed his inven- tory of the mammals of that country. He remained in Colombia until the press of curatorial duties and a gently delivered ultimatum from Sanborn finally recalled him to Chicago in 1952. The collections he made in Colombia, first for the National Museum, then for Field Museum, were to be the heart of all his subsequent research. But unlike others studying the mammal faunas of specific geographical regions, Hershkovitz found it unsatisfying to assess the systematics of Colom- bian mammals without following them across na- tional boundaries. Studies of a species or species group in Colombia led him to evaluate its context within genera, families, and even orders; and the remarkable diversity of Colombia's mammal fau- na led him into most major groups and most Neo- tropical subregions. In the course of his career, he has published dozens of generic, tribal, and fa- milial revisions, covering all 1 2 orders of Neo- tropical mammals. Few spatial and temporal boundaries have withstood the onslaught of his studies of Neotropical mammals. As examples one can point to the cosmopolitan Catalog of Living Whales {\9()6)—2iiXtr all, most cetaceans do occur in South American waters— and studies of Oli- gocene and later fossils (1974, 1982). One senses that the Department of Zoology dur- ing Hershkovitz's early years was a stimulating, harmonious one. Chief Curator Karl P. Schmidt took an almost paternal interest in junior staff and served as a confidant on the most personal of mat- ters. In addition to Colin Sanborn, who was most considerate of his junior curator's interests and HELDIANA: ZOOLOGY talents, Hershkovitz shared mammalogical prob- lems and topics with Dwight Davis, Curator of Anatomy, and Bryan Patterson, Curator of Ver- tebrate Paleontology. During the early and mid- 1950s, Hershkovitz established a vigorous and productive research program and participated in all aspects of departmental affairs. However, upon Schmidt's retirement in 1957, Austin S. Rand became Chief Curator of Zoology, and neither Rand nor Hershkovitz did much to disguise their antipathy for one another. Over the ensuing years, Hershkovitz increasingly detached himself from museum operations, culminating with Joseph Moore's appointment as Curator of Mam- mals in 1961, and Hershkovitz's appointment that year to Research Curator. No one before or since has held this title at Field Museum. Hershkovitz formally retired in 1971, although his work has continued unabated as Curator Emeritus. During his career, he assisted countless students in mam- mal projects, but has served on only a single grad- uate committee, that of Jack Fooden, now himself a renowned biologist and primate specialist at Field Museum. Few scientists can claim the independence in research that is indicated in Hershkovitz's bibli- ography. Of his approximately 300 scientific, pop- ular, and encyclopedia articles, only three repre- sent collaborative efforts. The first, with William P. Harris, an important benefactor of the Museum of Zoology of the University of Michigan, was suggested by Burt in recognition of Harris's inter- ests in squirrels and in token repayment for his patronage of the museum. The second, with Paul Rode, came about one afternoon in the Museum National d'Histoire Naturelle in Paris when Hershkovitz offhandedly suggested that designat- ing a lectotype might solve a nomenclatural prob- lem that Rode had encountered in his research. Rode insisted that Hershkovitz share authorship on the resulting paper. Later, after further study in the United States, Hershkovitz arrived at a con- trary opinion and wrote a paper, with Rode as coauthor, correcting their earlier one. Independent thought is also exemplified by the sometimes heated debates in which Hershkovitz has participated over the years. His published re- views and the discussion sections of many of his papers record his clearly enunciated views on such topics as the role of penial morphology in rodent taxonomy, the age and derivation of the South American fauna, panbiogeography, evolution of pelage coloration, and the systematic position of certain species (e.g., Dolichocebus). While such firmly held views brand him as something other than conciliatory or diplomatic, they accurately reflect his abiding passion and zest in science. Un- fortunately, some acerbic exchanges had the effect of stifling the scientific dialogue to which they were offered (e.g., penial morphology). Hershkovitz has focused his research on Neo- tropical mammals, their origin, evolution, dis- persal, classification, nomenclature, and system- atics. Specialists in these fields are well aware of his impact. However, he is perhaps most widely known for his work on three general topics of Neotropical mammalogy: faunal origins, meta- chromism, and New World monkeys. It would be folly to attempt to review all of his research, and more definitive appraisals on selected topics can be found scattered throughout this volume. How- ever, some comments on these general issues seem in order. As late as his revision of phyllotine rodents ( 1 962), Hershkovitz adhered to traditional notions of the derivation of certain South American taxa, notably "cricetid" rodents, from North and Mid- dle American stocks. This hypothesis of origins has been advocated most articulately by George G. Simpson, Bryan Patterson, and Rosendo Pas- cual, and more recently by Larry G. Marshall and S. David Webb. However, in the early 1960s, Hershkovitz was approached by Rupert Wenzel, Curator of Insects at Field Museum, who ques- tioned him on the evidence for Plio-Pleistocene origins of the Neotropical cricetids. Wenzel's stud- ies of the ectoparasites of Panamanian mammals suggested much earlier. South American origins. His interest piqued, Hershkovitz reviewed avail- able evidence, synthesizing continental drift (which was then becoming established in geological cir- cles) and neontological studies of mammals (es- pecially those of Hooper and Musser, which showed a relatively sharp dichotomy between sim- ple and complex penis-types of cricetids). He con- cluded that continental drift permitted a much greater role for paleotropical stocks in South American faunal origins than was allowed by the Simpsonian school, which in turn pointed to a much greater time period for independent evo- lution. Interestingly, and perhaps even character- istically, Hershkovitz concluded that South Amer- ican rodents were not only not derived from North American stocks, but instead gave rise to them. These views were published in 1966, 1969, and 1972. Hershkovitz's theory of metachromism, or de- terministic evolution of pelage coloration through PATTERSON: BIOGRAPHY OF PHILIP HERSHKOVITZ the loss of one or the other or both classes of hair pigments (eumelanins and phaeomelanins), was first pubHshed in 1968. Since then he has used it repeatedly in describing geographic variation in platyrrhine monkeys (e.g., 1977). However, the origins of this concept stem from his earlier work on the Sciunds granatensis group in northern Co- lombia where populations of squirrels thoroughly isolated from one another show similar progres- sions of pelage patterns. Few workers other than Neotropical primatologists (and not all of these) have accepted his interpretations, although the theory is potentially applicable to a variety of oth- er, mostly diurnal taxa showing pelage pattern variations. While Timothy Lawlor detailed some theoretical misgivings with the theory in a 1969 paper in Evolution (rebutted by Hershkovitz in 1970), to my knowledge it has not been substan- tially refuted. The theory is eminently testable: refutation would simply entail showing that pelage pattern variation of taxa arranged by metachro- mism is not congruent with well-established phy- letic patterns. Finally, some explanation seems warranted for Hershkovitz's current devotion to primates. In- deed, many recent workers unschooled in mam- malian systematics think of him as a primatolo- gist. Nothing could be further from the truth, as he hastens to point out. He had published several articles on primates in the course of working up his Colombian collections, but gave these taxa no special attention until the 1960s. Then govern- ment funding for primate studies soared, largely because of interest in biomedical applications, es- pecially for the complex and taxonomically con- fused Callitrichidae. For almost 20 years, Hersh- kovitz has focused first on the Callitrichidae and Callimiconidae, now on the Cebidae. His slower progress through these groups is attributable to the vast body of current knowledge about them; his 1977 and subsequent works serve as model syntheses of skin and skull morphology with bio- chemistry, karyology, ethology, serology, and ep- idemiology. By his own estimation, monkeys do not culminate his studies of Neotropical mam- mals, but rather represent a large and complex group to be covered in his attempt to treat all South American mammals. After seven years of work on Volume II of his primate monograph, he has near- ly completed generic revisions of cebids lacking prehensile tails and is beginning comparative stud- ies of their organ systems. In 1 984 he submitted another grant proposal for this work, totaling one- half million dollars in direct costs. His is not a modest work; it has been described by Pine ( 1 982; Vol. 6, Spec. Publ. Ser., Pymatuning Lab. Ecol.) as "the most heroically monumental revisionary monograph ever devoted to a Neotropical group." In 1984, Hershkovitz turned 75 years old. The 14 years he spent in the field in South America have served him well, for he seems younger than many men 15 years his junior. His tireless energy is best indicated by his habitual use of stairs rather than elevators (even his two divisional offices are three floors apart), a continuing program of field- work (most recently in Brazil during 1986 and 1987), and a museum workday that extends from 9 a.m. to 6 p.m., uninterrupted by coffee breaks or even lunch. Visitors to his home, now within walking distance of the Museum, know of his office there which relieves the chronic insomnia of ad- vancing years. He is an outstanding cook, a genial host, a trusted and valued friend, and an awe- somely productive scientist. Publications of Philip Hershkovitz 1938 1. A new caecilian from Ecuador. Occasional Papers, Museum of Zoology, University of Michigan, 370:1-3. 2. Two new squirrels fi-om Ecuador. Occasion- al Papers, Museum of Zoology, University of Michigan, 391:1-6 (with W. P. Harris). 3. A review of the rabbits of the andinus group and their distribution in Ecuador. Occasion- al Papers, Museum of Zoology, University of Michigan, 393:1-15. 1940 4. Four new oryzomyine rodents from Ecua- dor. Journal of Mammalogy, 21:78-84. 5. Notes on the distribution of the akodont ro- dent, Akodon mollis, in Ecuador with a de- scription of a new race. Occasional Papers, Museum of Zoology, University of Michi- gan, 418:1-3. 6. A new spiny mouse of the genus Neacomys from eastern Ecuador. Occasional Papers, Museum of Zoology, University of Michi- gan, 419:1-4. 1941 7. The South American harvest mice of the ge- nus Reithrodontomys. Occasional Papers, Museum of Zoology, University of Michi- gan, 441:1-7. FIELDIANA: ZOOLOGY 1944 8. A systematic review of the Neotropical water rats of the genus Nectomys (Cricetinae). Mis- cellaneous Publications, Museum of Zool- ogy, University of Michigan, 58:1-88. 1945 9. Designation d'un lectotype de Callithrix penicillatus (E. Geoffroy). Bulletin du Mu- seum National d'Histoire Naturelle, Paris 17(3):22 1-222 (with P. Rode). 1947 10. A correction. Journal of Mammalogy, 28(1): 68 (with P. Rode). 1 1 . Mammals of northern Colombia. Prelimi- nary report no. 1 : Squirrels (Sciuridae). Pro- ceedings of the United States National Mu- seum, 97:1-46. 1948 12. Mammals of northern Colombia. Prelimi- nary report no. 2: Spiny rats (Echimyidae), with supplemental notes on related forms. Proceedings of the United States National Museum, 97:125-140. 13. Mammals of northern Colombia. Prelimi- nary report no. 3: Water rats (genus Necto- mys), with supplemental notes on related forms. Proceedings of the United States Na- tional Museum, 98:49-56. 1 4. The technical name of the Virginia deer with a list of the South American forms. Pro- ceedings of the Biological Society of Wash- ington, 61:41-48. 1 5. Names of mammals dated from Frisch, 1 775, and Zimmermann, 1777. Journal of Mam- malogy, 29(3):272-277. 1949 1 6. Technical names for the fallow deer and Vir- ginia deer. Journal of Mammalogy, 30(1): 94. 1 7. Generic names of the four-eyed pouch opos- sum and the woolly opossum (Didelphidae). Proceedings of the Biological Society of Washington, 62:11-12. 18. Technical names of the African muishond (genus Zorilla) and the Colombian hog-nosed skunk (genus Conepatus). Proceedings of the Biological Society of Washington, 62: 13-16. 1 9. Mammals of northern Colombia. Prelimi- nary report no. 4: Monkeys (Primates), with taxonomic revisions of some forms. Pro- ceedings of the United States National Mu- seum, 98:323-427. 20. Mammals of northern Colombia. Prelimi- nary report no. 5: Bats (Chiroptera). Pro- ceedings of the United States National Mu- seum, 99:429-454. 21. Status of names credited to Oken, 1816. Journal of Mammalogy, 30(3):289-301. 22. Tapirs: Strange mammals native to Asia and tropical America from Mexico south. Chi- cago Natural History Museum Bulletin, 20(9):6-7. 1950 23. Mammals of northern Colombia. Prelimi- nary report no. 6: Rabbits (Leporidae), with notes on the classification and distribution of the South American forms. Proceedings of the United States National Museum, 100: 327-375. 1951 24. Mammals from British Honduras, Mexico, Jamaica and Haiti. Fieldiana: Zoology, 31(47):547-569. 1953 25. Zorilla I. Geoffroy and Spilogale Gray, ge- neric names for African and American pole- cats, respectively. Journal of Mammalogy, 34(3):378-382. 26. Four years on a zoological expedition in Co- lombia. Chicago Natural History Museum Bulletin, 24(l):3-4. 27. The reindeer— Important to man in fact and fancy. Chicago Natural History Museum Bulletin, 24(12):3-4. 1954 28. Mammals of northern Colombia, Prelimi- nary report no. 7: Tapirs (genus Tapirus), with a systematic review of American species. Proceedings of the United States National Museum, 103:465-496. 29. What the groundhog undergoes to make a "holiday." Chicago Natural History Mu- seum Bulletin, 25(2):3-4. 30. Who's a cow? Chicago Natural History Mu- seum Bulletin, 25(7):5. PATTERSON: BIOGRAPHY OF PHILIP HERSHKOVITZ 3 1 . Some ecological aspects of natural versus ar- tificial rehabilitation of a water basin area in Bogota, Colombia. Boletin del Instituto de U Salle, Bogota, 41(193/194):80-83. 1955 32. South American marsh rats genus Holochi- lus with a summary of sigmodont rodents. Fieldiana: Zoology, 37:639-673. 33. [Review] Mammals, a guide to familiar American species. Chicago Natural History Museum Bulletin, 26(7):7. 34. Notes on American monkeys of the genus Cebus. Journal of Mammalogy, 36:449-452. 35. Status of the generic name Zorilla (Mam- malia): Nomenclature by rule or by caprice. Proceedings of the Biological Society of Washington, 68:185-192. 36. On the cheek pouches of the tropical Amer- ican paca. Agouti paca (Linnaeus, 1766). Saiigetierkundliche Mitteilungen, 3(2):67-70. 37. Know your rabbits. Sports Afield, 134(6): 36-41,88. 1956 38. Comments on Galerella Gray, Herpestes II- liger, Leucomitra Howell, Icticyon Lund, Lutreola Wagner, Oryctogale Merriam, Paracynictis Pocock. Opinion 384 Interna- tional Commission of Zoological Nomen- clature, 12(5):71-190(intext). 39. Critical remarks on the status of names in Oken's "Lehrbuch." Opinion 417, Interna- tional Commission on Zoological Nomen- clature, 14(l):33-35. 1957 40. Comments on Canis dingo Meyer. Opinion 451, International Commission on Zoolog- ical Nomenclature, 15(17):335-336. 41. Comments on the validation of Muntiacus Rafinesque. Opinion 460, International Commission on Zoological Nomenclature, 15(26):467-468. 42. Comments on the generic name Mormoops Leach. Opinion 462, International Com- mission on Zoological Nomenclature, 16(1): 8-9. 43. Comments on Sciurus gambianus. Opinion 464, International Commission on Zoolog- ical Nomenclature, 16(3):36-39. 44. Comments on the validation of silvestris Schreber, 1777 [Felis {catus) silvestris]. Opinion 465, International Commission on Zoological Nomenclature, 16(4):49. 45. Comments on the validation of the name Phacochoerus Cuvier. Opinion 466, Inter- national Commission on Zoological No- menclature, 16(5):67-68. 46. Comments on the validation of the name Odobenus Brisson. Opinion 467, Interna- tional Commission on Zoological Nomen- clature, 16(6): 84-8 5. 47. The systematic position of the marmoset, Simia leonina Humboldt (Primates). Pro- ceedings of the Biological Society of Wash- ington, 70: 1 7-20. 48. The type locality of Bison bison Linnaeus. Proceedings of the Biological Society of Washington, 70:31-32. 49. A synopsis of the wild dogs of Colombia. Novedades Colombianas Museo de Historia Naturale Universidad del Cauca (Popayan), no. 3:157-161. 50. On the possible occurrence of the spectacled bear, Tremarctos ornatus(F. Cuvier, 1825), in Panama. Saugetierkundliche Mitteilun- gen, 5(3): 122-1 23. 1958 5 1 . [Review] Biological investigations in the Sel- va Lacandona, Chiapas, Mexico. Quarterly Review of Biology, 33(1):67. 52. [Review] Mammals of the Anglo-Egyptian Sudan, by Henry Setzer. Quarterly Review of Biology, 33:82. 53. Technical names of the South American marsh deer and pampas deer. Proceedings of the Biological Society of Washington, 71: 13-16. 54. Type localities and nomenclature of some American Primates, with remarks on sec- ondary homonyms. Proceedings of the Bi- ological Society of Washington, 71:53-56. 55. Stabilization of zoological nomenclature by a "Law of prescription." Bulletin of Zoolog- ical Nomenclature, 15B(20/21):630-632. 56. A critique of Professor Chester Bradley's "Principle of conservation." Bulletin of Zoo- logical Nomenclature, 15B(25/28):9 11-913. 57. The status of secondary homonyms and the concept of permanent rejection. Bulletin of Zoological Nomenclature, 15B(37/38):1242- 1243. 58. A geographic classification of Neotropical mammals. Fieldiana: Zoology, 36(6):583- 620. FIELDIANA: ZOOLOGY 59. The metatarsal glands in white-tailed deer and related forms of the Neotropical region. Mammalia, 22(4): 5 3 7-546. 1959 60. The scientific names of the species of ca- puchin monkeys (Cebus Erxleben). Proceed- ings of the Biological Society of Washington, 72:1-4. 6 1 . Two new genera of South American rodents (Cricetinae). Proceedings of the Biological Society of Washington, 72:5-10. 62. A new species of South American brocket, genus Mazama (Cervidae). Proceedings of the Biological Society of Washington, 72: 45-54. 63. A new race of red brocket deer {Mazama americana) from Colombia. Proceedings of the Biological Society of Washington, 72: 93-96. 64. The type locality of Felix concolor concolor Linnaeus. Proceedings of the Biological So- ciety of Washington, 72:97-100. 65. Nomenclature and taxonomy of the Neo- tropical mammals described by Olfers, 1818. Journal of Mammalogy, 40(3):337-353. 1960 66. Supposed ape-man or "missing link" of South America. Chicago Natural History Museum Bulletin, 31(4):6-7. 67. [Review] The Mammals of North America. Chicago Natural History Museum Bulletin, 31(5):6-7. 68. Publication dates for names of the Anubis baboon. Journal of Mammalogy, 41 (3):402- 403. 69. Mammals of northern Colombia. Prelimi- nary report no. 8: Arboreal rice rats, a sys- tematic revision of the subgenus Oecomys, genus Oryzomys. Proceedings of the United States National Museum, 1 10:513-568. 1961 70. On the South American small-eared zorro Atelocynus microtis Sclater (Canidae). Field- iana: Zoology, 39(44):505-523. 71. On the nomenclature of certain whales. Fieldiana: Zoology, 39(49):547-565. 72. "This is a mammal." Chicago Natural His- tory Museum Bulletin, 3 2(6): 3. 1962 73. Suriname zoological expedition. Chicago Natural History Museum Bulletin, 33(4):3, 7-8. 74. Bats and their menus. Chicago Natural His- tory Museum Bulletin, 33(8):2-3, 5-8. 75. Evolution of Neotropical cricetine rodents (Muridae) with special reference to the phyl- lotine group. Fieldiana: Zoology, 46:1-524. 1963 76. A systematic and zoogeographic account of the monkeys of the genus Callicebus (Cebi- dae) of the Amazonas and Orinoco River basins. Mammalia, 27(l):3-79. 77. [Review] Primates. Comparative Anatomy and Taxonomy. Vol. V, Cebidae, part B., A Monograph; Edinburgh University Press. American Journal of Physical Anthropolo- gy, 21(l):92-98. 78. [Review] Primates. Comparative Anatomy and Taxonomy. Vol. V, Cebidae, part B., A Monograph; Edinburgh University Press. American Journal of Physical Anthropolo- gy, 2 1(3):39 1-398. 79. Notes on South American dolphins of the genera Inia, Sotalia and Tursiops. Journal of Mammalogy, 44(1):98-103. 80. The nomenclature of South American pec- caries. Proceedings of the Biological Society of Washington, 76:85-88. 81. The Recent mammals of South America. Proceedings of the XVI International Con- gress of Zoology, Washington, D.C., Aug. 20-27, 1963. 82. Comments on the proposed suppression of Zorilla I. Geoffroy, 1826. Z.N.(S.) 758. Bul- letin of Zoological Nomenclature, 20(4):242- 244. 1965 83. Primate research and systematics. Science, 147(3662):1 156-1 157. 84. The importance of taxonomy in primate re- search and care. Illinois Society for Medical Research— Chicago Branch— Animal Care Panel Bulletin, 39:2 pp. 1966 85. Catalog of living whales. Bulletin of the United States National Museum, 246: 1-259. PATTERSON: BIOGRAPHY OF PHILIP HERSHKOVITZ 86. Taxonomic notes on tamarins, genus Sa- guinus (Callithricidae, Primates), with de- scriptions of four new forms. Folia Prima- tologica, 4:381-395. 87. On the identification of some marmosets, family Callithricidae (Primates). Mamma- lia, 30(2):327-332. 88. What ever happened to hairy man? Science, 153:362. 89. Comments on the proposal for conservation oi Pan Oken, 1816, and Panthera Oken, 1816. Bulletin of Zoological Nomenclature, 23(2/3):67-69. 90. [Review] Evolutionary and Genetic Biology of Primates, vol. II; Academic Press. Amer- ican Biology Teacher, 28(7):564-565. 91. Comments on the proposed suppression of Meles montanus Richardson, 1829, and M. jeffersonii Harlan, 1825. Z.N.(S.) 1639. Bul- letin of Zoological Nomenclature, 22(5/6): 336-337. 92. On the status of Procyon brachyurus Wieg- mann and P. obscurus Wiegmann. Z.N.(S.) 1640. Bulletin of Zoological Nomenclature, 22(5/6):338. 93. South American swamp and fossorial rats of the Scapteromyine group (Cricetinae, Mu- ridae) with comments on the glans penis in murid taxonomy. Zeitschrift fiir Saugetier- kunde, 31(2):81-149. 94. Status of the black-footed ferret in Wyo- ming. Journal of Mammalogy, 47(2):346- 347. 95. Comments on the proposal on Zorilla by Dr. Van Gelder and the counter proposal by Dr. China. Z.N.(S.) 758. Bulletin of Zoological Nomenclature, 2 3(2/3): 74-7 5. 96. Museum taxonomy serves medical research. Bulletin of the Field Museum of Natural History, 37(9):4-7. 97. Mice, land bridges and Latin American fau- nal interchange, pp. 725-751. In Wenzel, R. L., and V. J. Tipton, eds.. Ectoparasites of Panama. Field Museum of Natural History, Chicago. 1967 98. (Review] Evolutionary and Genetic Biology of Primates, vol. I; Academic Press. Amer- ican Biology Teacher, Nov. 1967:665. 99. Reply to Mayr's comment on the proposed preservation oi Pan from Oken, 1816. Z.N.(S.) 482. Bulletin of Zoological Nomen- clature 24(5): 1 p. 1 00. Dynamics of rodent molar evolution: A study based on New World Cricetinae, family Mu- ridae. Journal of Dental Research, Suppl. to 46(5):829-842. 1968 101. Metachromism or the principle of evolu- tionary change in mammalian tegumentary colors. Evolution, 22(3):556-575. 102. [Review] Dynamics of rodent molar evolu- tion: A study based on New World Cricet- inae, family Muridae. Oral Research Ab- stracts, May 1968. 1969 103. Comments on Cynocephalus Boddaert ver- sus Galeopithecus Pallas. Z.N.(S.) 1 792. Bul- letin of Zoological Nomenclature, 25(6):202- 203. 1 04. The evolution of mammals on southern con- tinents. VI. The Recent mammals of the Neotropical Region: A zoogeographic and ecological review. Quarterly Review of Bi- ology, 44(1): 1-70. 1970 105. The decorative chin. Field Museum of Nat- ural History Bulletin, 41(5):7-10. 106. Dental and periodontal diseases and abnor- malities in wild-caught marmosets (Pri- mates—Callithricidae). American Journal of Physical Anthropology, 32(3):377-392. 107. [Review] Taxonomy and Evolution of the Monkeys of Celebes (Primates: Cercopithe- cidae); Bibliotheca Primatologica, No. 10; Karger. Folia Primatologica, 13(l):75-76. 108. Metachromism like it is. Evolution, 24(3): 644-648. 1 09. Notes on Tertiary platyrrhine monkeys and description of a new genus for the Late Mio- cene of Colombia. Folia Primatologica, 12: 1-37. 110. Errata: Notes on Tertiary platyrrhine mon- keys and description of a new genus for the Late Miocene of Colombia. Foha Primato- logica, 12:1-37(1970). 111. Cerebral fissural patterns in platyrrhine monkeys. Folia Primatologica, 13:213-240. 112. [Review] The Squirrel Monkey; Academic Press. Journal of Mammalogy, 51(4):836- 839. 113. Supplementary notes on Neotropical Ory- zomys dimidiatus and Oryzomys hammondi (Cricetinae). Journal of Mammalogy, 51(4): 789-794. FIELDIANA: ZOOLOGY 1971 1977 1 14. Stapedial processes in tympanic cavities of capuchin monkeys (Cebus). Journal of Mammalogy, 52(3):607-609. 115. Basic crown patterns and cusp homologies of mammalian teeth, pp. 95-150. In Dahl- berg, A. A., ed., Dental Morphology and Evolution. University of Chicago Press, Chi- cago. 116. A new rice rat of the Oryzomys palustris group (Cricetinae, Muridae) from north- western Colombia, with remarks on distri- bution. Journal of Mammalogy, 52(4):700- 709. 126. Comment: Pan and Panthera or Oken's Lehrbuch? Z.N.(S.) 482. Bulletin of Zoolog- ical Nomenclature, 33(3/4): 135-1 36. 127. [Review] Catalogue of Primates in the Brit- ish Museum (Natural History). I. Families Callitrichidae and Cebidae; British Museum (Natural History). Folia Primatologica, 28: 315. 128. Living New World Monkeys (Platyrrhini). With an Introduction to Primates. Volume I. University of Chicago Press, Chicago, xiv +1117 pp. 1972 1 1 7. Notes on New World monkeys. Internation- al Zoo Yearbook, 12:3-12. 118. The Recent mammals of the Neotropical Region: A zoogeographic and ecological re- view, pp. 31 1-431. In Keast, A., F. C. Erk, and B. Glass, eds.. Evolution, Mammals and Southern Continents. State University of New York Press, Albany. 1974 119. The ectotym panic bone and origin of higher primates. Folia Primatologica, 22:237-242. 120. A new genus of Late Oligocene monkey (Cebidae, Platyrrhini) with notes on post- orbital closure and platyrrhine evolution. Folia Primatologica, 21:1-35. 1975 121. [Review] Taxonomic Atlas of Living Pri- mates; Academic Press. American Journal of Physical Anthropology, 41:155-156. 122. The scientific name of the \tsstv Noctilio (Chiroptera), with notes on the chauve-sou- ris de la Vallee d'Ylo (Peru). Journal of Mammalogy, 56(l):242-247. 123. Comments on the taxonomy of Brazilian marmosets (Callithrix, Callitrichidae). Folia Primatologica, 24:137-172. 1976 124. The taxonomic status of """Noctilio ruber Rengger." Mammalia, 40(1): 164-166. 125. Comments on generic names of four-eyed opossums (family Didelphidae). Proceed- ings of the Biological Society of Washington, 89(23):295-304. 1979 1 29. Races of the emperor tamarin, Saguinus im- perator Goeldi (Callitrichidae, Primates). Primates, 20(2):277-287. 1 30. The species of sakis, genus Pithecia (Cebi- dae, Primates), with notes on sexual dichro- matism. Folia Primatologica, 31:1-22. 1981 131. Comparative anatomy of platyrrhine man- dibular cheek teeth dpm4, pm4, ml with particular reference to those oT Homunculus (Cebidae), and comments on platyrrhine origins. Folia Primatologica, 35:179-217. 132. Philander and four-eyed opossums once again. Proceedings of the Biological Society of Washington, 93(4):943-946. 1982 133. Supposed squirrel monkey affinities of the late Oligocene Dolichocebus gaimanensis. Nature, 298(5870):20 1-202. 134. Subspecies and geographic distribution of black-mantle tamarins Saguinus nigricollis Spix (Primates: Callitrichidae). Proceedings of the Biological Society of Washington, 95(4):647-656. 135. Neotropical deer (Cervidae). Part I. Pudus, genus Pudu Gray. Fieldiana: Zoology, n.s., 11:1-86. 136. The staggered marsupial lower third incisor (I3), pp. 191-200. In Buffetaut, E., P. Jan- vier, J. C. Rage, and P. Tassy, eds., Phylo- genie et Paleobiogeographie. Livre jubiliare en I'honneur de Robert Hoffstetter. Geobios, memoire special 6, Lyon. PATTERSON: BIOGRAPHY OF PHILIP HERSHKOVITZ 1983 137. Two new species of night monkeys, genus Aotus (Cebidae, Platyrrhini): A preliminary report on Aott4s taxonomy. American Jour- nal of Primatology, 4:209-243. 138. On the validity of the family-group name Callitrichidae (Platyrrhini, Primates). Mam- malia, 48:153. 1 39. Taxonomy of squirrel monkeys, genus Sai- miri (Cebidae. Platyrrhini): A preliminary report with description of a hitherto un- named form. American Journal of Prima- tology, 6:257-312. 140. [Review] Mammalian Biology in South America. M. A. Mares and H. H. Genoways, eds. Ecology, 65(6): 1944-1 945. 1985 141. 1986 142. A preliminary taxonomic review of the South American bearded saki monkeys, genus Chi- roptes (Cebidae, Platyrrhini), with the de- scription of a new subspecies. Fieldiana: Zo- ology, n.s., 27:1-46. [Review] Handbook of Squirrel Monkey Re- search. L. A. Rosenblum and C. L. Coe, eds. Quarterly Review of Biology, 61:286-287. 143. The piebald saki. Field Museum of Natural History Bulletin, 57(2):coverplate + 24-25. 1987 144. Uacaries, New World monkeys of the genus Cacajao (Cebidae, Platyrrhini): A prelimi- nary taxonomic review with the description of a new subspecies. American Journal of Primatology, 12:1-53. 1 45. First South American record of Coue's marsh rice rat, Oryzomys couesi. Journal of Mam- malogy, 68(1): 152-1 54. 146. The titi. Field Museum of Natural History Bulletin, 58(6): 11-15. 147. The taxonomy of South American sakis, ge- nus Pithecia (Cebidae, Platyrrhini): A pre- liminary report and critical review with the description of a new species and a new sub- species. American Journal of Primatology, 12:387-468. In Press 148. More on the Homunculus Dpm4 and ml and comparisons with Alouatta and Stirto- nia (Primates, Platyrrhini, Cebidae). Amer- ican Journal of Primatology. 10 HELDIANA: ZOOLOGY A History of the Recent Mammalogy of the Neotropical Region from 1492 to 1850 Philip Hershkovitz ABSTRACTS The history of Neotropical mammalogy began with the first voyage of Christopher Colum- bus in 1492. The earliest notices were purely anecdotal, recorded by Spanish chroniclers from the mouths of the sailors on their return from voyages of discovery during the 1 5th and 1 6th centuries. Colonization of the Guianan and Brazilian coasts during the 1 7th century provided opportunities for inventories and descriptions of the mammals by trained European naturalists and physicians. The systematization and scientific naming of the known Brazilian species by Carolus Linnaeus in 1758 were based primarily on the mammals described in the 17th century monograph of Brazilian biota by Georg Marcgraf The actual collection and preservation of mammals for study, however, began in the 18th century with the Brazilian-bom Alexandre Rodrigues Ferreira. The 18th and first half of the 19th century was an explosive and romantic period of independently or govemmentally sponsored scientific expeditions for field observa- tions, collections, preservations, and taxonomic studies of the specimens shipped to European museums and private collectors. Outstanding among the naturalists who made significant con- tributions to mammalogy during this period are Alexander von Humboldt, Johann Baptist Ritter von Spix (Brazil), Maximilian Prinz Wied-Neuwied (Brazil), Johann Natterer (Brazil), Sir Robert Herman Schomburgk and Richard Schomburgk (Guyana), Claudio Gay (Chile), Johann Jakob von Tschudi (Peru), Felix de Azara (Paraguay), Rudolph Rengger (Paraguay), Alcide Charles Victor d'Orbigny (Argentina, Bolivia), and Charles Robert Darwin (Patagonia and Galapagos). Their itineraries, collections of mammals, taxonomies, and some field notes are included in the accounts of these and other noteworthy naturalists. By the middle of the 1 9th century, the mammalian fauna of South America became the best known of any continent with exception of the western European part of Eurasia. The problems of origins and distribution of Neotropical mammals intrigued scholars from among the earliest chroniclers down to pre- evolutionary Darwin. Their concepts on these subjects are briefly discussed. La historia de mastozoolo^a neotropical empieza con el primer viaje transatlantico de Cris- tobal Colon en 1492. Poco despues de desembarcarse de sus viajes de regreso los descrubridores y conquistadores del Mundo Nuevo en los siglos quince y diez y seis contaron a los cronistas espaiioles de las cosas curiosas que encontraron. Colonizacion de las costas guyanas y brasileras durante el siglo diez y siete ofrecio oportunidades a los naturalistas y medicos europeos residentes para le van tar inventarios de los mamiferos y anotar y hacer informes sobre sus observaciones. La sistematizacion y el nombramiento cientifico de las especies brasileras conocidas por Carolus Linnaeus en 1758 fueron basadas primariamente sobre los mamiferos descritos y figurados por Jorge Marcgraf en su monografia del siglo diez y siete. La coleccion y preservacion efectiva de mamiferos para el estudio empezo en comienzos del siglo diez y ocho con el "Viajem Filosofica" de Alejandro Rodriguez Ferreira, brasilero de nacimiento. From the Division of Mammals, Department of Zo- ology, Field Museum of Natural History, Chicago, Illi- nois 60605-2496. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 1 1 El siglo diez y ocho y las primeras decadas del siglo diez y nueve senalaron un periodo explosive y romantico de expediciones cientificas fomentadas por gobiemos europeos, o por naturalistas particulares con los objectos de hacer observaciones sobre la fauna, tomar notas de campo, y recoger y preservar ejemplares para estudios taxonomicos en los museos extranjeros. Entre los naturalistas europeos que hicieron contribuciones de consequencia a la mastozoologia neo- tropical en este epoca, se cuentan Alejandro von Humboldt, Juan Baptista Ritter von Spix (Brasil), Maximiliano Principe de Wied-Neuwied (Brasil), Juan Natterer (Brasil), Lord Roberto Herman Schomburgk y Ricardo Schomburgk (Guyana), Claudio Gay (Chile), Juan Jacobo von Tschudi (Peru), Felix de Azara (Paraguay), Rodolpho Rengger (Paraguay), Alcidio Carlos Victor d'Orbigny (Argentina, Bolivia), y Carlos Roberto Darwin (Patagonia y Galapagos). Compren- dido en este informe son los itinerarios, listas de mamiferos coleccionados y observados, taxonomias, y algunas experiencias de campo de los naturalistas mentados, y de otros digno de atencion. A mediados del siglo diez y nueve, la fauna mamifera de Sud America llego a ser la mejor conocida de todos los continentes del mundo menos Europa. Problemas de origen y reparticion geografica de los mamiferos del Mundo Nuevo estimularon la imaginacion de sabios desde los primeros cronistas del Descubrimiento hasta el joven Darwin pre-evolutionario. Los conceptos sobre estos temas son brevemente discutidos. A historia da mastozoologia neotropical, come90u com a primeira viagem transatlantica de Cristovao Colombo, em 1492. Os primeiros relatorios, puramente anedotais, foram registrados pelos cronistas espanhois, logo apos o regresso das viagens de descobrimento durante os seculos XV e XVI. As colonizacoes da costa Guianense e Brasileira durante o seculo XVII, ofereceram amplas oportunidades a naturalistas e medicos, de treinamento Europeu, para inventoriar e descrever os mamiferos encontrados. A sistematica e a nomenclatura cientifica das especies Brasileiras conhecidas por Carolus Linnaeus em 1758 basearam-se primariamente nos ma- miferos descritos por Georg Marcgraf, em sua monografia do seculo XVII. No entanto, as colecoes e preservacoes de mamiferos para estudos come^aram, efetivamente, no seculo XVIII, com a "Viajem filosofica" do Brasileiro, Alexandre Rodrigues Ferreira. Marcaram o seculo XVIII, e as primeiros decadas do seculo XIX, um periodo explosivo e romantico nas expedicoes cientificas. Estas foram patrocinadas tanto por naturalistas indepen- dentes, como por govemos Europeus, a fim de realizarem observacoes sobre a fauna e colecoes para estudos taxonomicos nos museus Europeus. Entres os naturalistas Europeus que distin- guiram-ce em suas contribuicoes aos estudos de mamiferos neotropicais durante esta epoca, sobressaem: Alexandre von Humboldt, Juan Baptista Ritter von Spix (Brasil), Maximilian Principe de Wied-Neuwied (Brasil), Johan Natterer (Brasil), Sir Robert Herman Schomburgk e Richard Schomburgk (Guiana), Claudio Gay (Chile), Johan Jakob von Tschudi (Peru), Felix de Azara (Paraguai), Rudolph Rengger (Paraguai), Alcides Charles Victor d'Orbigny (Argentina, Bolivia) e Charles Robert Darwin (Patagonia e Galapagos). Os itinerarios, as listas de mamiferos observados e colecionados, as taxonomias, e algumas notas de campo encontram-se incluidos nos relatorios aqui apresentados sobre estes e outros naturalistas importantes. Nas meadas do seculo XIX, a fauna mamifera sul-americana tomouse a melhor conhecida de todos OS continentes, exceto a da Europa. Os problemas de origem e da distribuicao geografica dos mamiferos neotropicais estimularam a imaginacao de varios estudiosos, desde os primeiros cronistas ate o pre-evolucionario Darwin. Seus conceitos sobre estes temas sao brevemente discutidos. Organization II. Voyages of Discovery: 1 5th and 16th Centuries 14 I. Introduction 13 III. Spanish Chroniclers of New The Neotropical Region Defined . . 14 World Discoveries 14 12 HELDIANA: ZOOLOGY IV. First Mammals: Anecdotal Period 16 Island Mammals of the Discoverers 16 Mainland Mammals of the Discoverers 18 V. Brazil: Mammalogy Through 1 8th Century 21 Andre Thevet (1503-1592) 21 Georg Marcgraf (1610-1644) 21 Alexandre Rodrigues Ferreira (1756-1815) 21 VI. Brazil: Mammalogy to Middle of 1 9th Century 27 Introduction 27 Johann Baptist Ritter von Spix (1781-1826) and Carl Friedrich von Martius (1794-1866) 27 Maximilian Prinz von Wied-Neu- wied (1782-1867) 31 Johann Natterer (1787-1843) .... 34 VII. GuiANAs: Mammalogy to End of 1 8th Century 38 Pierre Barrere (1690-1755) 38 Jose Gumilla (d. 1750) 38 Jacques Nicolas Bellin (1703-1772) 38 Edward Bancroft (1744-1821) .... 38 Philippe Fermin (1720-1790) .... 39 Monsieur Bajon (1763?) 40 John Gabriel Stedman (1744-1797) 40 VIII. GuiANAs: Mammalogy of First Half of 19th Century 43 Sir Robert Herman Schomburgk (1804-1865) and Richard Schomburgk (181 1-1891) 43 IX. Alexander von Humboldt ( 1769- 1859) AND AlME BONPLAND (1773-1858) 51 X. Paraguay 57 Felix de Azara (1746-181 1) 59 Johann Rudolph Rengger (1795-1832) 64 XI. Chile 64 Giovanni Ignazio Molina (1737-1829) 64 Eduard Friedrich Poeppig (1798-1868) 65 Claudio Gay (1800-1873) 65 XII. Peru 65 Johann Jacob von Tschudi (1818-1889) 65 XIII. Patagonia 71 Alcide Charles Victor d'Orbigny (1802-1857) 71 Charles Robert Darwin (1809-1882) 77 XIV. Georges Louis Leclerc de Buffon (1707-1788) 87 XV. Faunal Origins and Distribution 87 Jose de Acosta (1539-1600) 87 Antonio Vazquez de Espinosa (1560/1575-1630) 90 Carolus Linnaeus (1707-1778) ... 90 Georges Louis Leclerc de Buffon (1707-1788) 90 Johann Andreas Wagner (1797-1861) 91 Maximilian Prinz von Wied- Neuwied (1782-1867) 91 Johann Jacob von Tschudi (1818-1889) 91 Charles Robert Darwin (1809-1882) 91 XVI. Inventories to Middle of 1 9th Century 91 System Naturce of Linnaeus, 1758, 1766 91 Histoire Naturelle of Buffon, 1750-1789 92 Synopsis Mammalium of Schinz, 1844 92 XVII. Summary 92 XVIII. Acknowledgments 94 XIX. Literature Cited 94 I. Introduction The gradual accumulation of knowledge of Neotropical mammals is recorded here from the time of the first voyage of discovery by Christo- pher Columbus in 1492 to the middle of the 19th century, or just before the Darwinian revolution in biological thought. The knowledge was mainly of species or kinds, the numbers of kinds, their behavior, habitat, geographic distribution, and re- lationship to man. Early voyagers to the New World followed by naturalist-travelers gathered the data used later by philosophers and scientists for the development of biological principles. Only the most important and better-known contributors are discussed here. At least as many more personages could be included in a more extended account. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 13 The Neotropical Region Defined The Neotropical Region, as defined by its mam- malian fauna, includes all South America, Middle America except the dry and temperate zones of Mexico, continental islands of coastal Middle and South America, and the oceanic Bahamas, West Indies, Galapagos, and Falklands (Hershkovitz, 1972, p. 326). With few exceptions, modem names for Neo- tropical countries and geographic features are used throughout the text. The map (fig. 1) shows the South America of the colonial period with colonial or precolonial names for political subdivisions and geographic features. II. Voyages of Discovery: 15th and 16th Centuries The inhabited islands found by Columbus on his first voyage across the Atlantic Ocean in 1492 were thought to be near the mainland of China or India. The islanders welcomed the ships' crews with food and drink, but the great stores of pre- cious metals, stones, and artifacts the travelers ex- pected to find were not seen. Nevertheless, the voyagers claimed the islands for the Spanish crown and returned with accounts told to awaiting re- porters of their discoveries, including their de- scriptions of plants and animals of economic value or imputed medicinal virtues. Zoological results of the four transatlantic voy- ages commanded by Christopher Columbus— the first (1492-1493) and second (1493-1496) to the Antilles, the third (1498) to the Antilles and Ven- ezuela, and the last (1 502-1 504) to Middle Amer- ica—included reports of a variety of mammals. The kinds seen were identified with such familiar Old World forms as lion, tiger, bear, fox, dog, ferret, rabbit, deer, boar, goat, sheep, rodent, mon- key, and ape. Characterizations given were less descriptions of external morphology than of gen- eral mien, gross habitat, behavior in response to human confrontations or predation on human property, gastronomic qualities, and use, if any, in medical treatment, ceremonial rites or magic, or as household pets. Those who followed Columbus in the discovery and exploration of the mainland returned with additional bits of information on mammals noted by the attendant Spanish chroniclers. Among the more important of these voyagers of discovery were Pinzon, who followed Columbus to the Ven- ezuelan coast in 1 500, and Amerigo Vespucci, who sailed first with Ojeda to Brazil in 1499 and in- dependently again in 1 502 and 1 503 in the service of Portugal. Pedro Cabral, however, had already claimed Brazil for Portugal in 1 500 on his way to India. In 1516 Juan Diaz de Solis discovered the estuary of the Rio de la Plata, and Sebastian Cabot, in the service of Spain, sailed in 1526 up the Rio Parana. Vasco Nunez de Balboa accompanied En- ciso to Panama in 1510, and in 1513, with Fran- cisco Pizarro, crossed the isthmus to behold the vast Pacific Ocean. Pizarro visited Panama again in 1531, recrossed the isthmus, and sailed south along the west coast of South America to the dis- covery and conquest of Peru. Cabeza Alvarez Nu- nez de Vaca arrived in Buenos Aires in 1541 and continued overland into Paraguay. Pedro de Val- divia visited Venezuela in 1 530, Peru in 1 532, and Chile in 1540, 1541, and 1552. The explorations of Colombia by Gonzalez Jimenez de Quesada from 1536 to 1539 and again in 1569 to 1571 signaled the end of the period of discovery and conquest. III. Spanish Chroniclers of New World Discoveries The recorders or chroniclers of New World dis- coveries, conquests, happenings, and natural phe- nomena were the clerics and scribes who accom- panied the explorers or awaited their return to Spain for recording the news. Most of the accounts or records remained unpublished, but some of the manuscripts are reportedly preserved in the ar- chives of Spain or the Vatican. The chroniclers whose published narratives contain interesting in- formation on mammals include the following. Peter Martyr of Anghiera (1455-1526), Italian by birth, and the first and most prestigious chron- icler of the Discovery, was a member of the Royal Spanish Council of the Indies, Prothonotary of the Catholic Church, correspondent of Popes, confi- dant of Christopher Columbus, and friend of sea captains, clergymen, and other contemporary voy- agers to the New World. News he received from his informants constitutes the first records of New World discoveries. His chronicles, known as the Decades and addressed to the Pope, began to ar- rive at the Vatican in 1 494. The first Decade de Orbe Novo, with first notices of American mam- mals, was published in 1516, but pirated Italian 14 HELDIANA: ZOOLOGY Fig. 1. Map, South America of the Colonial period from the Stevens (1726) translation of Herrera y Tordesillas. editions appeared in 1504 and 1507. The fourth Decade was published in 1521, and the complete set of eight of the projected 10 appeared posthu- mously in 1587. Gonzalo Fernandez de Oviedo y Valdes (1478- 1557) was appointed royal chronicler of news sent directly to him by provincial governors and other New World officials. Included were Oviedo's own observations and results of investigations during his residence as representative of the Spanish Crown in the Provinces of Darien, Panama, Gua- temala, Cuba, and Santo Domingo. He published HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY IS the first part of his Historia de las Indias in 1 526, other parts in 1535 and 1 547. The entire work was printed between 1851 and 1855 in Madrid. The Spanish Jesuit Jose de Acosta ( 1 539-1 600) wrote his Historia Natural y Moral de las Indias during a residence in Peru from 1571 to 1 587 and saw it published in 1 590. Acosta's philosophical inquiries extended to all asjjects of nature in the New World and greatly influenced the thinking of his contemporaries. Antonio Vazquez de Espinosa (b. between 1560 and 1575, d. 1630), a Carmelite missionary, lived many years in Spanish America, most of them in Peru and Mexico. His natural history notes are compiled from many sources, including his per- sonal observations and testimony of people he met in travels connected with his clerical duties. The forgotten manuscript of his Compendium was dis- covered in the Vatican library by Charles Upson Clark in the early part of the 20th century. Clark's English translation of the work was published in 1 942 by the Smithsonian Institution, and his tran- scription of the original Spanish in 1 948 by the same institution. Antonio de Herrera y Tordesillas (1559-1625), historiographer to the King of Spain, compiled the General History of the Vast Continent and Is- lands of America from archived reports by the New World discoverers and conquistadores, gov- ernors, clergy, colonists, and travelers. He also borrowed heavily from published accounts, in- cluding those of other chroniclers. There is no in- dication that his notices on mammals were based on personal observations. The first edition of Herrera's History was published in 1601, another in 1 60 1-1 6 1 5. These and a 1 728 Spanish edition in the Library of Congress are cited in the bibliography. I have not seen these works. The Stevens translation, published 1725-1726, was used here. Whatever the quality of the translation, I find no fault with the descriptions of mammals, and the stories about them are in line with similar accounts in other sources. IV. First Mammals: Anecdotal Period Island Mammals of the Discoverers The first Columbian voyage, in 1492, resulted in the discovery of the Antillean islands of Cuba, Hispaniola, and part of the Bahaman Archipelago. According to Peter Martyr, who reported results of the voyage in his first Decade (1504, 1516), quadrupeds were not seen, but three kinds of "rabbits" were said to occur in Hispaniola (Haiti and Dominican Republic). The same animals, ac- tually caviomorph rodents, were described later by Oviedo during his residence in Santo Domingo. The following accounts are freely translated or paraphrased from the Spanish of the Paraguayan (1944-1945) edition of Oviedo's work. Hutia, the first "rabbit" (1944, libro XII, cap. I), is smaller than the ordinary rabbit, its ears smaller and tail ratlike. The hutia is said to be dark grayish in color and very good eating. It was hunted and killed by the barkless dogs of the na- tives, but is no longer found, except rarely. Gerrit S. Miller (1929, p. 12), studied the re- mains of mammals in kitchen middens of the Sa- mana Bay region, Dominican Republic, and con- cluded that the original description of the hutia "would apply as well to the species of Plagiodon- tia, and presumably also to the Isolobodons [sic] that there seems to be no reason to doubt that these were the animals Oviedo had in mind." The quemi, second of the "rabbits" (1944, libro XII, cap. II), is said to be blackish like the hutia and similar in form, but larger like an ordinary hound. Natives of the island who saw and ate the animal found it savory. Oviedo believed them ex- tinct. All attributions to the quemi, according to Mil- ler ( 1 929, p. 13), agree with those of a "large rodent whose remains I found in the caves near St. Mi- chel, Haiti, in 1925. Consequently, I proposed for it the generic name Quemisia. The presence of the same creature in the Boca del Infiemo kitchen midden appears to confirm my guess." The mohuy "rabbit" (1945, libro XII, cap. Ill), is smaller than the hutia, a paler brown or grayish in color, its flesh highly esteemed by the island's caciques and noblemen. The pelage, unlike that of the hutia, is stiff", sharply pointed, and erect. Ovie- do saw no mohuy, but knew persons on the island who did and reportedly regarded its flesh as better than that of the other "rabbits." "There be little if any doubt," says Miller ( 1 929, p. 13), "that the animal Oviedo thus described was Brotomys voratus ... its remains have been found in every kitchen midden that has been ex- amined in the Dominican Republic. . . . The ac- count of stiff", pointed, erect-standing hairs of the back seems especially applicable to a relative of the South American spiny-rats." The cori, a fourth "rabbit," described by Oviedo 16 HELDIANA: ZOOLOGY (1945, libro XII, cap. IV), is almost certainly the domestic guinea pig. Miller (1929, p. 14) ques- tioned whether the guinea pig was pre-Columbian or a Spanish introduction. He inclined to the sec- ond alternative "chiefly because remains of the animal have been found in only one midden." It appears, however, that one Simone Verde, who accompanied Columbus on his first voyage, men- tioned in a letter dated 20 March 1494 (cf. Martyr in Gaffarel, 1907 trans, p. 12, footnote 2; p. 14, footnotes 1 , 2) the existence on the island of a black and white dormouse-like animal without tail. The guinea pig or cui, domesticated in Peru, was carried by pre-Columbian Indians for food and barter and introduced into islands and many parts of mainland South America where cavies do not naturally occur. Many of them, such as completely isolated colonies I saw in Colombia near Bogota, had become feral, their coloration having reverted to the wild or agouti pattern. Other Hispaniolan mammals mentioned by Oviedo are the barkless domestic dogs and house rats, the latter certainly brought by the Spaniards. Apart from the extinct insectivore Nesophontes, Miller found no remains of mammals the size of mice or rats in kitchen middens or owl pellets. Two additional native West Indian mammals observed by Oviedo in 1 523 or 1 524 in Cuba differ from those of Hispaniola. My paraphrased trans- lation of Oviedo's Spanish descriptions follows. The guabiniquinax is somewhat larger than a rabbit, its feet similar, the tail long and ratlike, the pelage smoother than that of a badger, the skin white, the flesh savory. It lives and breeds in the mangroves along the coast. To capture it, the In- dians position their canoes beneath the mangroves where it nests, then shake the tree to cause the animal to fall into the water where it is seized. The animal as described above is certainly a form of Capromys, but Oviedo continues as fol- lows: "The animal is the size of a hare, looks like a fox, its color is dark brown mixed with reddish, the tail hairy and the head shaped like that of a ferret. It abounds along the Cuban coast." The characterization and habitat are obviously out of place and probably were meant to be included with the description of the ayre, the second of the Cu- ban mammals reported by Oviedo. Herewith my paraphrased translation of his description of that animal. The ayre is reddish brown, the size of a rabbit with pointed muzzle, its flesh exceedingly tough. Notwithstanding, the natives cook or roast as many of the animals as they can capture, for they are abundant. But no matter how long the meat may be cooked or roasted, it is no less tough to chew. This characterization seems to fit the insectivore Solenodon. On the other hand, the flesh of Cap- romys, as of most if not all caviomorphs, is tender and, as a rule, delectable. From his correspondents Oviedo received no- tice of still another mammal, the guacabitinax, an inhabitant of the islands near those of Las Perlas in the Golfo de San Miguel and the Isla de las Culebras or Gorgona, off" the southwest coast of Colombia. The name, not to be confused with the preceding, and the description and details of the animal's habits, are unmistakably those of the paca {Agouti paca Linnaeus). Manatees sighted at sea at various times by Co- lumbus and his men were believed to be mer- maids, albeit ugly ones. Martyr's narrative of a captive manatee as given in the available French translation of his third Decade (Gaffarel, 1907) is composite. The account by Herrera of the same manatee (in Stevens's translation, 1725, vol. 1, p. 278) appears to hew closer to the original source of information: The Spaniards at this Time found a new Sort of Fish, which was a considerable ad- vantage to them; tho' in those Parts there is much Variety. It is call'd Manati, in shape like a Skin they use to carry Wine in, having only two Feet at the Shoulders, with which it swims, and it is found both in the Sea and in Rivers. From the Middle it sharpens off" to the Tail, the Head of it is like that of an Ox, but shorter, and more fleshy at the Snout; the Eyes small, the Colour of it grey, the Skin very hard, and some scattering Hairs on it. Some of them are twenty Foot long, and ten in Thickness. The Feet are round, and have four Claws on each of them. The Females bring forth like the Cows, and have two Dugs to give suck. The Taste of it is beyond Fish; when fresh it is like Veal, and salted like Tunny-Fish, but better, and will keep longer; the Fat of it is sweet, and does not grow rusty. Leather for Shoes is dress'd with it. The Stones it has in the Head are good against the Pleurisy and the Stone. Sometimes they are taken ashore, grazing near the Sea, or Rivers, and when young they are taken with Nets. Thus the Cazique Caramestex took one, and fed it twenty-six Years in a Pond, and it grew sensible and tame, and would come when call'd by the Name of Mato, HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 17 which signifies Noble. It would eat what- soever was given it by Hand, and went out of the Water to feed in the House, would play with the Boys, let them get upon him, was pleas'd with Musick, carry'd Men over the Pool, and took up ten at a Time, without any Difficulty. Martyr's third Decade mentions many "rab- bits" and deer encountered in 1 5 1 6 by Andre Mo- rales on the forested Isla Rica (now San Jose) of the Archipielago de las Perlas in the Golfo de Pa- nama. The deer, most likely Mazama gouazoubira permira Kellogg, 1 946, and rabbits (Sylvilagus sp.) were said to be so abundant that Spaniards could shoot them with arrows from horseback. The re- tiring tapeti, Sylvilagus brasiliensis, the only rabbit known from mainland Panama and the Pearl Is- lands, would have been an unlikely target for the equestrian Spaniards. Mainland Mammals of the Discoverers First knowledge of mainland American mam- mals was contained in reports of the Paria Pen- insula, Venezuela, by Columbus on his third trans- atlantic voyage in 1498 and Vicente Yaiiez Pinzon, who followed in the tracks of Columbus. Martyr's first Decade carried the news of their encounters with the common opossum (Didelphis marsupi- alis), sloths, armadillos, anteaters, deer (Odocoi- leus, Mazama), peccaries, tapir, kinkajou (Potos flavus), barkless dog {Canis familiaris), jaguar, puma and their color varieties, vampire(?) bats, and red howler monkey (Alouatta seniculus). On his fourth and last voyage (1 502-1 504), Co- lumbus explored the Atlantic coast of Middle America from the Golfo de Uruba to Guatemala. Spanish emissaries charged with establishing set- tlements followed quickly. Mammals reported by them and noted in Martyr's second and third Decades, and by Oviedo, include the common opossum, bats, monkeys, three-toed sloths, ant- eaters, armadillos {Dasypus novemcinctus), white- tail deer, red brocket, collared and white-lipped peccaries, squirrels, a composite of carnivorous species identified as raposas (including Didel- phis?), zorros (foxlike Camivora), lobos (Dusicyon or Lutra), rabbits (Sylvilagus brasiliensis), "hares" (Dasyprocta sp.. Agouti paca, and perhaps the newly introduced European hare or rabbit). The domes- tic dog, like that first seen in the Antilles, was barkless. An encounter with vampire bats by Pinzon's men is reported by Martyr in the first Decade. Vampires are also mentioned in the second De- cade in connection with Enciso's disastrous ex- periences in the Darien and in the third Decade in accounts of the animals of the Golfo de Uruba. The following characterization of a vampire bat by Herrera (in Stevens's translation, 1726, vol. 2, p. 7 1 ) is a translation from the original sources in Spanish. "This venomous Creature has one quality that tho' it bites one man among an hundred one Night, the next Time it only bites in the very same Place, tho' the Person bit be among two hundred; which it does either on the Toes, the Fingers, or the Head, and much Blood runs from it." That the same vampire bat should visit the same person sleeping in the same place on successive nights may not be unusual. An experience of mine in 1935 on the Rio Napo in Ecuador is of interest in this regard. Two Indian families and I, alto- gether 10 persons including a five-year-old girl, traveled three days upstream in a large dugout canoe. The river was low and we could bivouac on sandbars at the end of each day's travel. On each of the three nights, a vampire bat visited the little girl, scraped the skin of her nose, and fed on the trickling blood. No other member of the party was attacked. It seems improbable that the same bat should have found the same victim at each of the three different bivouacs. Perhaps the child slept more soundly than the others of the party, or her blood was more attractive to the vampires which abounded in the region. The last of Martyr's eight Decades includes de- scriptions of Spanish settlements in the Golfo de Paria, Venezuela. In addition to those mammals previously mentioned by Martyr (above) are the lesser anleater (Tamandua tetradactyla), capuchin monkey (Cebus apella or C. nigrivittatus), peccary, deer (Odocoileus virginianus), jaguar {Felis onca), spotted cats {Felis pardalis and/or F. wiedii), wea- sels (Mustela frenata), skunk (Conepatus chinga), porcupine (Coendou prehensilis), and manatee {Trichechus manatus). Oviedo described the same animals of the region in greater detail, but with no additions of sp>ecies. Vazquez de Espinosa, who in 1628 presumably visited the northern Venezuelan coast and the town of Santo Tomas above the mouth of the Rio Orinoco, reported the same mammals as well as squirrels (Sciurus aestuans) 18 FIELDIANA: ZOOLOGY and many kinds of monkeys. He claimed that Isla Margarita, off the Venezuelan coast, was overrun with rabbits {Sylvilagus floridanus). Many of the larger mammals of Colombia in the territories of the Muso and Colima Indians north of Bogota were already known by 1544. With bats and other small mammals omitted, more kinds were reported by Herrera than could be re- corded today from the same region on the basis of extant specimens preserved in museums. Her- rera, in the English version by Stevens (1726, vol. 6, p. 191), states: There are a great number of grey Swine [Tayassu pecan] that have the Navel on the Back, and a smaller sort of several Colours [Tayassu tajacu] much like wild boars. Ti- gers (Felis onca) not numerous but very fierce; Lions (Felis concolor) that do no harm, except only among the Cattle and two other sorts of Tigers that were inoffensive besides another sort that are always in the water, like Greyhounds, and all their four feet are like those of a Goose [Lutra annectens]. The black wild cats [Felis yagouaroundi] seize the Hens, carry them away under one of their front legs and run away on the other three. The black Bears [Tremarctos ornatus] like those in Spain, do no hurt but only to the small Cattle. The Ant-Bears [Myrme- cophaga tridactyla] when they go, lay their Tail, which is long, on their Heads, winding them about their Necks, and so walk from Ant-hill to Ant-hill, stretch out their Tongues near half a Yard which are soon cover'd with Pismires, then they draw them back and eat them. There are Dantas [Tapirus pinchaque or T. terrestris]. Deer [Odocoileus virgini- anus] like ours in Europe, and others red like wild Goats [Mazama rufina or Mazama americana], and the Bezoar stones found in them are best. The Guadatinajas [Agouti paca] are like Hares; and the Zorillas [Di- delphis marsupialis] or little foxes, that have a purse under their Belly, in which they carry their Cubs, the ever so many, are very mis- chievious to the Henroosts. The little Crea- tures call'd Umazia [Marmosa] have a dug growing out for every one of their young, and they stick to it till bred up. The Ar- madillo [Dasypus novemcinctus] which has been spoken of having five claws in each Forefoot, with which it throws up the Earth, is tame and eaten. The Perico Ligero [Bra- dypus variegatus] is three hours climbing a Tree, goes about in the Night, gives a cry every time it lifts a Foot, and is half an Hour, between every Step, is as big as a Barbary monkey, and fierce, yet does no harm. There are cats (?) that sleep all the Day, and all the Night catch Birds and Mice. The Pizma [Na- sua nasua] about as big as a large Lap Dog, has a bad countenance, a long Snout, its voice like a Bird, defends itself against Dogs, and the Spaniards call them Badgers. The Hedgehogs [spiny echimyids] are like those in Spain, the largest like Porcupines [Coen- dou sp.] darting out their Prickles. There are many sort of Apes, squirrels. Elsewhere, in the Province of Bogota, Herrera (in Stevens's translation, 1726, vol. 6, p. 77) notes "innumerable apes, monkeys, ferrets [marsupi- als?], squirrels, weasels [Mustela frenata], deer [Odocoileus virginianus], roebuck [Mazama rufi- na] and Rabbits [Sylvilagus brasiliensis] . . . but not Hares." Manatees were reported from the Rio Magdalena. From coastal Colombia, at Zaragoza, 30 leagues from Caceres in the lower Rio Cauca Valley, Vaz- quez de Espinosa records jaguar, puma, danta (Tapirus terrestris), oso (Myrmecophaga or Ta- mandua), cuchumbi (Nasua), armadillo (Dasy- pus), raposa (Dusicyon thous), chucha (Didelphis marsupialis), "three" species of sahinos or pec- cary, perico ligero (Bradypus variegatus), nutria (Lutra or Chironectes), and guadatinaja (Agouti paca). Acosta's long residence in Peru made him fa- miliar with some of the mammals in the vicinities of Cuzco and Lima and others about which he may have learned from travelers or records. He described sahinos (peccaries), dantas (tapir), ar- madillos, perico ligero (three-toed sloth), osos (anteaters), otoronco (bear), chinchilla, vizcacha, cui (guinea pig). The "liebres verdaderas" or true hares are certainly the introduced European hare. He affirmed that conejos or rabbits (Sylvilagus brasiliensis) occur in the Reino de Quito (Ecua- dor). Acosta declared there were monkeys of all kinds throughout America, but those he described were Middle American. At Capira near Nombre de Dios, Panama, he saw monkeys (presumably spider monkeys) swing by their tails from a tree on one side of a stream to another tree on the opposite HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 19 side. Where the river was too wide for this ma- neuver, the monkeys of the troop, he related, crossed by forming a hanging chain holding on to each other's tail, then swinging so that the endmost could grab the branch of a tree on the other side of the river and let all the others clamber up. The anecdote is less a fabrication than an ex- aggeration. Individual howlers and spider mon- keys, usually the alpha male or an old suckling female, may bridge a narrow gap in the canopy pathway by holding with its prehensile tail the branch on one side of the gap and with it swinging the body to catch, with outstretched arms, the nearest branch of the far side. Monkeys too small or weak to hurdle the gap run or scramble over the bridging back of their elder. I have seen strong young adults take advantage of the same conve- nience. Acosta also narrated the tale of a monkey that resided in the palace of a provincial governor. As related to him, the simian was trained to fetch wine from the town tavern. The animal would set off with the empty wine pitcher in one hand, the wine money gripped in the other. Not before the pitcher was filled to the brim did the sage monkey release his coins to the tavern keeper. There were times on these errands when taunting street ur- chins chased and hurled stones at the monkey. Annoyed by this sport, the simian halted, set down the pitcher, and returned the stones with sufficient force and accuracy to rout his tormentors. Retriev- ing his pitcher, he moved on serenely to deliver the wine at the palace. Peruvian "sheep" or camelids were greatly ad- mired by the Spaniards when first seen. Acosta's interesting account of them was suitably appre- ciated by Herrera, and the English translation by Stevens (1726, vol. 4, p. 36) is quoted herewith. There are no such Vicunas and Sheep in New Spain [Mexico] As those of Peru, and those Sheep are Tame, and very serviceable; but the Vicunas are wild, and have no Horns, the like of them not to be seen in the whole World, but only in Peru and Chile, bigger than Goats, but smaller than Calves, their Colour almost Murrey, breeding on the highest Mountains, in cold and desert Places, which they call Punas. They go in flocks, run swiftly, and when they see any Men, fly and drive their Young before them. Of their Wooll are made very valuable Mantles, which never lose their Colour, because it is natural; they are said to be good for Inflam- mations in the Kidneys, as are Quilts made of the Wooll, because they moderate the Heat, and the same in the Gout; and in them the Bezoar Stones are found. The abundance and ubiquity of llamas may have inspired some Spaniards to attempt to raise Old World camels in Peru. According to Acosta, some brought from the Canary Islands were bred for a while. Sebastian Cabot's journal of conquest and ex- ploration of the Province of Rio de La Plata, then consisting of modem northern Argentina, cisan- dean Bolivia, and southeastern Brazil, included data on natural history. As recorded by Herrera, the mammals seen were the hairy armadillo {Chaetophractus sp.) and several other kinds, ca- vies {Cavia), swamp deer (Blastocerus dichotomus), pampa deer {Blastoceros bezoarticus), brockets (Mazama sp.), tapirs {Tapirus terrestris), peccaries (any or all of the known species), howler monkeys (Alouatta), canids (Dusicyon), lesser anteater (7a- mandua tetradactyla), jaguar (Felis onca), and puma (Felis concolor). Southern Brazilian mam- mals in particular included deer, peccary, tapir, "rabbits" with small, round ears {Dolichotisl), paca {Agouti paca), armadillo, sloth (Bradypus torqua- tus), opossum {Didelphis albiventris), monkeys, and coastal seals, most likely Arctocephalus australis. Vazquez de Espinosa adds capybaras, armadil- los (tatu and quirquincho specified), and guanacos. In the vicinity of Chuquisaca (La Plata), Bolivia, the missionary notes brockets {Mazama), vicufia, guanaco, dark gray wildcats known as oscollos, jaguar called "otorongo," puma locally called poma, a large beast called lilisto with a horselike head that lures cattle and humans, a ferret called siqui {Mustela frenatal), skunks or anatiria {Co- nepatus), bear {Tremarctos ornatus), antbears (probably Tamandua), vizcacha {Lagidium), and cuis {Cavia porcellus). The occurrence of sea lions {Otaria flavescens) and fur seals {Arctocephalus) on both southern continental coasts was mentioned by Vazquez de Espinosa. The sea lions along the coast of Are- quipa, Peru, he reported come out of the water onto the rocks and make low sounds at night. The animals were hunted by the Indians for their hides. In northern Chile, the natives of Arua and Ata- cama converted the hides into balloon-like floats for support of their seagoing fishing rafts. The conquest of Chile by Pedro de Valdi via in 1 54 1 provided the chroniclers with additional in- formation on mammals. Vazquez de Espinosa re- 20 HELDIANA: ZOOLOGY ported huemul {Hippocamelus bisulciis), "fallow deer" (spotted fawns of huemul), guanaco, and vicuna in the vicinity of Osomo. According to the same authority, the Rio Guasco valley (29°S) har- bored "squirrels" (chinchillas) with very fine fur. V. Brazil: Mammalogy Through 18th Century Andre Thevet (1503-1592) The French missionary Andre Thevet arrived in 1555 in Rio de Janeiro, the principal port of a French colony in the ephemeral France Antarc- tique. Thevet returned to France via the Antilles a year later, and the accounts of his travels were published in 1557 or 1558. Father Thevet's cu- riosity about all he saw in the New World knew no bounds, and he became an avid collector of Indian artifacts, local birds, and insects. Not all objects and events described in his book con- formed to popular European prejudices or gen- erally accepted misconceptions. The work stirred up considerable debate and was rejected by many not prepared to accept the realities that opossums had pouches or that the bodies of American In- dians were not densely furred. The Brazilian mammals described or men- tioned by Thevet include the locally common opossum (Didelphis albiventris), tapeti (Sylvilagus brasiliensis), agouti {Dasyprocta leporina, declared good eating), peccaries, deer (probably Mazama), coati {Nasua nasua), tapir (Tapirus terrestris), ca- puchin monkey {Cebus apella), golden tamarin {Leontopithecus rosalia), armadillos, jaguar (Felis onca), and deer-hunting canids (Speothos"?), but no lions or wolves. The three-toed sloth was abun- dant, but never observed eating or drinking. The- vet adds, however, that there are those who believe the beast sustains itself solely by the small, slender leaves of a very high tree called amahut. Georg Marcgraf (or Marggrav or Marggraf] (1610-1644) Most illustrious of the pre-Linnaean naturalist- explorers of Brazil was Georg Marcgraf Bom in Liebstad, Saxony, educated in Holland with em- phasis on astronomy and botany, he sailed for Brazil in 1638 on a scientific expedition led by Johann Moritz, Count of Nassau-Siezen. The par- ty, which included the young physician Piso (1611- 1678), landed in Pemambuco. Explorations were restricted to northeastern Brazil in the present states of Pemambuco, Paraiba, and Rio Grande do Norte. Among MarcgraPs accomplishments were the construction of an astronomical observatory, the first of its class in the New World, and a mono- graphic study of the plants and animals of the region. After turning over his notes and illustra- tions to Moritz, for preparation and publication, the naturalist sailed for Africa, where he died shortly after arrival. MarcgraPs monumental His- toriae Rerum Naturalia Brasiliae, a part of Willen Piso's Historia Naturalis Brasilia, was published in 1648 in Amsterdam. Of the mammals of the northeastern region of Brazil described by Marcgraf, 32 were native species, the others introduced. Their detailed de- scriptions and life history notes, together with crude but useful woodcuts (fig. 2), were among the pri- mary references on which Linnaeus based bino- mials in the 10th (1758) and 12th (1766) editions of his Systema Naturce. The mammals are listed in Table 1 by the in- digenous names used by Marcgraf and their cur- rent scientific names. Provenance of the forms which served as types for binomialists, mainly Linnaeus, was restricted for taxonomic purposes to Pemambuco by Thomas (1911). Alexandre Rodrigues Ferreira (1756-1815) The first Brazilian naturalist of European ex- traction, Alexandre Rodrigues Ferreira, was bom in Salvador, Bahia. He pursued higher studies in Portugal, received his doctorate in 1779 from the University of Coimbra, and was then appointed Naturalist of the Museu Real d'Ajuda in Lisbon. He retumed to Brazil in 1783 commissioned by the museum to collect samples of plants, animals, and minerals and to record all matters of scientific and political interest within his scope. The expe- dition, or "Viagem Filosofica," explored the prov- inces of Grao Para, Rio Negro, Mato Grosso, and Cuiaba from 1 783 to 1 792 (fig. 3). Rodrigues Fer- reira retumed to Lisbon the following year. The scientific materials collected in Brazil, with notes and illustrations, were deposited in the Mu- seu d'Ajuda. Included were 4 1 7 species of animals represented by 592 specimens. Of these, 76 spec- imens represented 65 species of mammals. The whole collection was confiscated by the invading armies of Napoleon and taken to Paris for study HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 21 •c 1 2-* 03 so 'a$5M /a/acM Linnaeus, 1758 2 Hydrochaeris hydrochaeris Linnaeus, 1 766 2 Sciurus aestuans Linnaeus, 1 766 Dasypus septemcinctus Linnaeus, 1758 Dasypus novemcinctus Linnaeus, \1 5% 2 Tolypeutes trici net us Linnaeus, 1758 2 Felis tigrina Schreber, 1775 Coendou prehensilis prehensilis. Linnaeus, 1758 2 Pteronura brasiliensis Gmelin, 1 788 2 Blastoceros bezoarticus Linnaeus, 1 758 Blastoceros bezoarticus Linnaeus, 1758 F(e//5 onca Linnaeus, 1758 2 Felis onca (melanistic) 2 Felis concolor Linnaeus, 1771 * Editors' Note: Here and elsewhere in this paper. Article 51(c) of the International Code of 2kx)logical Nomen- clature, governing the use of parentheses in scientific names, is not followed. by Etienne Geoffroy St.-Hilaire of the Museum National de Histoire Naturelle in Paris. Monkeys constituted a sizeable part of the loot, and the following were described as new by Etienne Geoffroy St.-Hilaire in 1812 and by others as not- ed in brackets; the current form of each name is used: Callithrix jacchus penicillatus, Callithrix jacchus geoffroyi [Humboldt], Callithrix jacchus aurita, Callithrix humeralifer, Callithrix argentata melanura, Saguinus labiatus, Saimiri ustus [I. Geoffroy], Callicebus amictus, Callicebus person- al us, Pithecia monachus, Alouatta fusca, Cebus apella cirrifer. Cebus flavus, and Lagothrix la- gothricha canus. Mounted specimens of previ- ously named forms also brought to Paris from the Lisbon museum included Callithrix jacchus Lin- naeus, Leontopithecus rosalia Linnaeus, Chiro- potes satanas Hoffmannsegg, Brachyteles arach- noides E. Geoffroy, Inia geoffrensis Blainville, and probably others lost or discarded. Except for the descriptions by the French zo- ologist, the specimens and manuscripts of Rod- rigues Ferreira were largely neglected during the naturalist's lifetime. The several portions of the memoirs published posthumously were heavily edited. In 1972, however, the entire Viagem Fi- losofica, in two text volumes and two of colored plates, was published by the Conselho Federal de Cultura of the Brazilian Ministry of Education and Culture. Treatment of mammals in the zoological mem- oir was a model of its kind for the times. Each species was described, with bibliographic refer- ences for the ones better known, external char- acters and what was learned of habitat, habits, reproduction, utilization by man, and gastronomic rating. With respect to the last, Rodriguez Ferreira grouped the Brazilian mammals according to those used most widely for food (peccary, deer, tapir, paca, agouti), those eaten only by Indians and some HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 23 ' W' KOIFIRO |)A AIAC.KM l'llA POR AlfXANDRK RODRU.I Is l-F.RRHRA. \l MA hISI ANCIA APROXIMADA l>E Vf J7'.' kM (17M \- Fio. 3. Map of Brazil showing routes (bold lines) of Alexandre Rodrigues Ferreira, during the " Viagem Filosofica, 1783-1792; from Rodrigues Ferreira (1972). 24 FIELDIANA: ZOOLOGY Table 2. Mammals illustrated in the Viagem Filosoftca by Rodrigues Ferreira (1971), Plate no. Brazilian name Current scientific name F gure 118 Gamba Didelphis marsupialis Linnaeus 119 Macaco-da noite Aotus sp. 120 Zogue-zogue; uapuca Callicebus moloch Hoffmannsegg 121 Parauacu Pithecia monachus E. Geoffroy 122 Cuxiu Chiropotes satanas chiropotes Humboldt 123 Cuxiu-preto Chiropotes satanas satanas Hoffmannsegg 124 Guariba-vermelho Ahuatta seniculns Linnaeus 125 Guariba-da-mao-ruiva Alouatta belzebul Linnaeus 126 Mico-de-cheiro Saimiri ustus \. Geoffroy 127 Quata-de-cara-vermelha Ateles paniscus Linnaeus 128 Barrigudo-cinzento Lagothrix lagothricha Humboldt 129 Sauitinga Callithrix argentata argentata Linnaeus 130 Saui dourado Callithrix humeralifer chrysoleuca Wagner 131 Saui Callithrix jacchus penicillata E. Geoffroy 132 Saui-de-mao-ruiva Saguinus midas midas Linnaeus 133 Tamarin Saguinus midas tamarin Link 134 Saui-de-bigode-branco Saguinus labiatus labiatus E. Geoffroy 135 Tamandua-mirim Tamandua tetradactyla Linnaeus 136 Tamanduai Cyclopes didactylus Linnaeus 137 Tamanduai Cyclopes didactylus Linnaeus 138 Preguifa-de-tres-dedos Bradypus variegatus Schinz 139 Tatu-galinha Dasypus novemcinctus Linnaeus 140 Tatu peba Euphractus sexcinctus Linnaeus 141 Guaraxaim Procyon cancrivorus F. Cuvier 142 Janauira Speothos venaticus Lund 143 Guara Chrysocyon brachyurus Illiger 144 Quati Nasua nasua Linnaeus 145 Jupara Potosflavus Schreber 146 Furao Galictis vittata Schreber 147 Irara Eira barbara Linnaeus 148 Ariranha Pteronura brasiliensis Gmelin 149 Maracaja Felis geoffroyi d'Orbigny and Gervais 150 Jaguartirica Felis pardalis Linnaeus 151 Su9uarana Felis concolor Linnaeus 152 Jaguar Felis onca Linnaeus 153 On9a preta Felis onca Linnaeus 154 Peixe-boi, male & female Trichechus inunguis Natterer 155 Caitetu Tayassu tajacu Linnaeus 156 Veado vermelho Mazama americana Erxleben 157 Cariacu Odocoileus virginianus cariacou Boddaert 158 Quatipuru- vermelho Sciurus igniventris Wagner 159 Quatipuru-preto Sciurus spadiceus Olfers 160 Quatipuru-louro Sciurus igniventris Wagner 161 Rato-d'agua Nectomys squamipes Brants 162 Prea Cavia aperea Erxleben 163 Cutia-vermelha Dasyprocta leporina Linnaeus 164 Cutia-preta Dasyprocta fuliginosa Wagler 165 Acutiuaia Myoprocta exilis Wagler 166 Paca Agouti paca Linnaeus 167 Cuandu Coendou prehensilis Linnaeus 168 Uiara Inia geoffrensis Blainville 169 Tucuxi Sotalia fluviatilis Gervais and Deville white residents (anteaters, armadillos, sloths, por- cupines, monkeys, jaguar), and animals not eaten by humans (marsupials, melanistic felids, squir- rels, capybara). Bezoar stones and certain parts of the animal, usually tegumentary, were also cited for their medicinal merits, particularly as anti- venins for headaches and female sterility, or as aphrodisiacs. A memoir on the peixe boi or river manatee (Tricheciis inunguis Natterer) provides detailed in- formation on such topics as hunting, harpooning, reproduction, size, weight, blubber, butchery, HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 25 Fio. 4. Four monkeys of the "Viagem Filosofica" collections: upper left, parauaco (Pithecia monachus E. Geoffroy), possibly the holotype; upper right, saui-de-bigode-branco {Saguinus labiatus labiatus E. GeofTroy), possibly the holotype; lower left, mico-de-cheiro (Saimih ustus I. Geoffroy), possibly the holotype; lower right, saui (Callithrix jacchus penicillata E. Geoffroy), possibly the holotype; from Rodrigues Ferreira (1972). 26 HELDIANA: ZOOLOGY preservation, and market value of the flesh. The author decried the slaughter of the young and not- ed the disappearance of manatees in certain lakes. Of all Brazilian mammals described or merely listed in the Viagem Filosofica, those depicted in color in the 50 plates (each 1 9 x 29 cm) are rep- resentative. They are listed in Table 2 by plate number with their Brazilian and current scientific names. The animals were postured as prepared by taxidermists (fig. 4). Many of the monkeys are those later described by E. Geoffroy. VI. Brazil: Mammalogy to Middle of 19th Century Introduction Growth of science in South America during the first third of the 19th century shifted from the Spanish colonies, with their wars for independence and internal political turmoil, to the relatively sta- ble Portuguese colony of Brazil. Following the in- vasion of Portugal by the Napoleonic armies, the royal family fled to Brazil and made Rio de Janeiro its capital and center of cultural activities. During previous years Brazil had been closed to foreigners to prevent the mines of precious metals and min- erals from passing out of control of the ruling Por- tuguese. Dom Joao VI, however, opened the ports and changed the environment to one befitting an enlightened monarch in residence. Cultural insti- tutions, including museums, libraries, and uni- versities, were built, and scientific investigations were promoted. Betrothal of the Archduchess Leo- poldina, daughter of the Emperor of Austria, with Dom Pedro, Crown Prince of Portugal and Brazil, became the most important single factor in the advancement of science in the New World during the first half of the 1 9th century. The entourage of the bride on her voyage to Brazil included some of the best and most adventurous of the younger scientists of Austria and Bavaria. The Viennese naturalists of the party included the field collector Johann Natterer, and from the court of Munich, the zoologist Spix and the bot- anist Martins. Two years earlier, in 1815, the most accomplished of the naturalist-travelers, Maxi- milian Prinz Wied zu Neuwied of Prussia, arrived on the scene. Modem Brazilian mammalogy begins with the scientific accounts of the collections and travels of these naturalists. Johann Baptist Ritter von Spix (1781-1826) and Carl Friedrich von Martins (1794-1866) The German naturalist Johann Baptist Ritter von Spix first studied for the priesthood, but after two years his attention turned to medicine and natural history. His doctorate was earned in 1 806. That same year he was appointed assistant in the Museum of the Munich Academy of Science, with responsibility for the organization of the zoolog- ical collections. In 1816 he was ordered by the King of Bavaria to undertake a two-year scientific expedition to Brazil, together with the museum's assistant in botany, Carl Friedrich von Martins. The two departed on 10 April 1817 through the port of Trieste, and after considerable delay, they arrived in Rio de Janeiro on 15 July 1817. The exuberance and variety of the native plant life in eastern Brazil at first awed and bewildered the two young naturalists. Everything they saw was new to them, and all they could possibly collect and preserve was easily reached along the trails they traveled from Rio de Janeiro to Minas Gerais and beyond. Real or fantasized dangers lurking in what they imagined as dark, brooding, impene- trable forests restrained their urges for stepping ofl" the beaten path. The strange and wonderful wild- life encountered on the roads was enough to gratify their utmost expectations and inspired them to record their impressions in ecstatic prose. On the trip from Ipanema, Sao Paulo, to Vila Rica, Minas Gerais, they described, as translated into English by Lloyd in equally romanticized and tortured prose, the numerous flocks of little monkeys [that] run whistling and hissing to the recesses of the forest; the cavies, running about on the tops of the mountains, hastily secrete themselves under loose stones; the American ostriches (Emas), which herd in families, gallop at the slightest noise, like horses through the bush- es, and over hills and valleys, accompanied by their young; the dicholopus {Seriemas), which pursues serpents, flies, sometimes sinking into the grass, sometimes rising into the trees, or rapidly climbing the summits of the hills, where it sends forth its loud deceitful cry, resembling that of the bustard; the terrified armadillo {Tatu Canastra, Peba, Bola) runs fearfully about to look for a hid- ing place, or, when the danger presses, sinks into its armour; the ant-eater {Tamandud, Bandeira mirim) runs heavily through the HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 27 ^^ ^'Salvador (Bahia) Januaria liPbrto de Salgado) aOiamantina \>^ ^-■^buro Preto (Villa ricQ) Rio de Janeiro Sao Poulo Kane von Brasilien mit dem eingezeichneten Reiseweg von Johann Baptist von Spix und Carl Friedrich Philipp von Martius anlaBlich ihrer Expedition in den Jahren 1817-1820. Fig. 5. Map of Brazil showing routes of the Spix and Martius expedition (1817-1820); only principal stations plotted; from Tiefenbacher(1983). plain, and, in case of need, lying on its back, threatens its pursuers with its sharp claws. Far from all noise, the slender deer, the black tapir or a pecari, feed on the skirts of the forest. Elevated above all this, the red-head- ed vulture (urubii) soars in the higher re- gions; the dangerous rattle-snake {Casca- vel), hidden in the grasses, excites terror by its rattle; the gigantic snake sports suspended from the tree with its head upon the ground; and the crocodile resembling the trunk of a tree, basks in the sun on the banks of the pools. After all this has passed during the day before the eyes of the traveler, the ap- proach of night, with the chirping of the grasshoppers, the monotonous cry of the goat-sucker {Jodo corta pdo), the barking of the prowling wolf, and of the shy fox, or the roaring of the ounces, complete the singular picture of the animal kingdom in these peaceful plains. For the next three years, the zoologist and bot- anist explored the eastern states of Brazil from Sao Paulo and Minas Gerais north to Para. Most of July and August of 1 8 1 9 was spent in Belem (Para). On 2 1 August they shipped up the Rio Amazonas, making stopovers at the mouth of the Rio Tocan- tins, the Rio Xingu ( 1 0 September), Santarem on the Rio Tapajos ( 1 8 September), Obidos (23 Sep- tember), Parintins, and Vila Nova da Rainha (1 October). The mouth of the Rio Madeira was passed 1 5 October, and on 22 October they landed at Barra do Rio Negro (Manaus). Travel upstream 28 HELDIANA: ZOOLOGY continued in November with a stop at Tefe (for- merly Ega) on 26 November. Spix then traveled alone up the Solimoes to Tabatinga at the Peru- vian border, arriving 9 January 1820. Martius, for his part, ascended the Rio Japura to Araracuara in eastern Colombia. Spix returned to Manaus on 3 February 1820. On 1 1 February he ascended the Rio Negro to Barcelos and was back again in Manaus 28 Feb- ruary to continue his travels downstream to Be- lem, where he arrived on 16 April. He embarked on 14 June 1820 for Europe from Rio de Janeiro (fig. 5). In the Reise. Spix and Martius (1828, p. 541) made up an impressive list of the mammals of the sertao (scrub country) of Campos Gerais de Sao Felipe in the angle between the Rio Sao Francisco and its eastern tributary, the Rio Verde Grande, northern Minas Gerais. The data were evidently compiled uncritically from a number of sources, including local informers, personal observations, and publications based on the Wied-Neuwied (1826) collections. Their use and misuse of names are too involved to unravel here. Except for the missing bats (given elsewhere by Spix, 1823) and some small rodents, it is unlikely that a similar or larger number of mammalian species of the area, based on actual specimens, could be made today. The sertao mammals of the Spix and Martius ex- pedition are listed in Table 3 by current scientific names of the species only, with the Spix and Mar- tius equivalents omitted. In his journey up the Amazon, Spix noted habits of the inia {Inia geoffrensis) (Spix &, Martius, 1831, p. 1 1 1 9) and of the manatees (Trichechus inunguis) (Spix & Martius, 1831, p. 1122). The results of the expedition are recorded in several publications, including the Simiarum et Vespertilionum Brasiliensium by Spix (1823). The account of the nearly three-year journey or Reise in Brazil by Spix and Martius (1823-1831) is re- plete with observations on the biology, geography, geology, paleontology, mineralogy, meteorology, and the various human cultures and industries of the parts of the country they traveled. Many kinds of mammals are mentioned, but except for bats and monkeys, few of them were collected. The zoological material actually collected con- sisted of thousands of invertebrates and 498 species of vertebrates, of which 34 were monkeys and 15 bats. Altogether, according to Avila Pires (1974, p. 139), 85 species of mammals were collected. Spix (1823) reported only on the monkeys and bats and illustrated in color the types of all species. Table 3. Mammals of the sertao of Campos Gerais de Sao Felipe, Minas Gerais, recorded by Spix and Mar- tius (1828, p. 541, footnote 3). Current scientific names to species only are used. The Spix and Martius usage of local, German, and scientific names is too confused for tabulation. The arrangement is phylogenetic. Marsupialia Caluromys philander Linnaeus Didelphis marsupialis Linnaeus Primates Callithrix jacchus Linnaeus Cebus apella Linnaeus Alouatta fusca E. GeofTroy Alouatta caraya Humboldt Edentata Tamandua tetradactyla Linnaeus Myrmecophaga tridactyla Linnaeus Bradypus torquatus Desmarest Bradypus variegatus Schinz Dasypus novemcinctus Linnaeus Tolypeutes tricinctus Linnaeus Priodontes maximus Kerr Euphractus sexcinctus Linnaeus Carnivora Dusicyon thous Linnaeus Chrysocyon brachyurus Illiger Nasua nasua Linnaeus Procyon cancrivorus G. Cuvier Conepatus chinga Molina Eira barbara Linnaeus Pteronura brasiliensis Gmelin Felis wiedii Schinz Felis tigrina Schreber Felis pardalis Linnaeus Felis concolor Linnaeus Felis onca Linnaeus Felis yagouaroundi E. Geoffroy Perissodactvla Tapirus terrestris Linnaeus Artiodactyla Mazama gouazoubira Fischer Mazama americana Erxleben Blastoceros bezoarticus Linnaeus Lagomorpha Sylvilagus brasiliensis Linnaeus RODENTIA Sciurus aestuans Linnaeus Wiedomys pyrrhorhinos Wied-Neuwied Echimys and/or Proechimys species? Myocastor coypus Molina Kerodon rupestris Wied-Neuwied Cavia aperea Linnaeus Dasyprocta leporina Linnaeus Agouti paca Linnaeus Coendou insidiosus Kuhl Chaetomys subspinosus Olfers HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 29 d. c 5? 11 c 5 1^ 2 i" •S 30 HELDIANA: ZOOLOGY Table 4. Monkeys (Primates) of the Spix and Martius Expedition described by Spix (1823); the arrangement is phyiogenetic. Current name Spix and Martius synonym Figure Cebuella pygmaea Spix, 1 823 Callithrix jacchus jacchus Linnaeus, 1758 Callithrix jacchus penicillatus E. Geoffroy, 1812 Saguinus bicolor bicolor Spix, 1823 Saguinus fuscicollis fuscicollis Spix, 1823 Saguinus mystax mystax Spix, 1823 Saguinus nigricollis nigricollis Spix, 1 823 Saguinus oedipus geoffroyi Pucheran, 1845 Callicebus cupreus Spix, 1823 Callicebus personatus personatus E. Geoffroy, 1812 Callicebus personatus nigrifrons Spix, 1 823 Callicebus personatus melanochir Kuhl, 1820 Callicebus torquatus torquatus Hoffmannsegg, 1 807 Callicebus cinerascens Spix, 1823 Aotus vociferans Spix, 1 823 Actus azarae infulatus Kuhl, 1820 Pithecia monachus monachus E. Geoffroy, 1812 Pithecia pithecia pithecia Linnaeus, 1 766 Chiropotes satanas chiropotes Humboldt, 1812 Cacajao melanocephalus ouakary Spix, 1823 Alouatta caraya Humboldt, 1812 Alouatta belzebul discolor Spix, 1823 Alouatta fusca Spix, 1823 Alouatta seniculus stramineus Humboldt, 1812 Cebus albifrons unicolor Spix, 1 823 Cebus apella libidinosus Spix, 1823 Cebus apella macrocephalus Spix, 1823 Cebus apella xanthosternos Wied-Neuwied, 1 820 Lagothrix lagothricha lagothricha Humboldt, 1812 Lagothrix lagothricha carta E. Geoffroy, 1812 Brachyteles arachnoides E. Geoffroy, 1 806 Jacchus albicollis Spix, 1 823 Midas oedipus (varietas), Spix, 1823 Callithrix gigot Spix, 1823 Callithrix amicta E. Geoffroy, 1812 Nyctipithecus felinus Spix, 1823 Pithecia hirsuta Spix, 1823; Pithecia inusta Spix, 1823 Pithecia capillamentosa Spix, 1823 Brachyurus israelita Spix, 1 823 Mycetes barbatus Spix, 1823 Cebus gracilis Spix, 1823 Cebus cucullatus Spix, 1823; Cebus xanthocephalus Spix, 1823 Gastrimargus infumatus Spix, 1823 Gastrimargus olivaceus Spix, 1 823 Brachyteles macrotarsus Spix, 1 823 most life-size. Separate reports on all groups of animals collected by Spix have been brought to- gether in a Festschrift in his honor edited by Tie- fenbacher (1983). The mammals are treated by Kraft (1983). The 31 presently recognized species and sub- species of monkeys ( 1 5 new) and the 1 4 recognized species of bats (six new) are listed in Tables 4 and 5 by current names with synonyms in parentheses. Maximilian Prinz von Wied-Neuwied (1782-1867) Maximilian Prinz von Wied-Neuwied was bom in Prussia and studied biological sciences at the University of Gottingen under the famous natu- ralist-anthropologist Blumenbach. His ambition to travel and study nature in South America was realized when he sailed for Rio de Janeiro from England the first week of May 1815, and arrived on 17 July. After a few excursions in the surroundings of Rio de Janeiro, Wied-Neuwied left for Cabo Frio on 15 August 1815, stopping at many fazendas and villages along the way. He left Cabo Frio on 8 September for Sao Salvador dos Campos dos Goitacazes (now simply Campos) on the Rio Pa- raiba, and arrived on 25 September. After more excursions and more collections in the state of Rio de Janeiro, he crossed the Rio Itabapoana on 26 November into the state of Espirito Santo. A con- siderable amount of time was devoted there to explorations of the Rio Doce region. February 1816 saw Wied-Neuwied in Bahia, where he occupied himself until May 1817. The coastal town of Bel- monte, where he arrived in August 1816, was the base for explorations of Botocudo Indian territory. In December 1816 Wied-Neuwied established II- heus as center for travel westward to Sao Pedro de Alcantara, now Itabuna, and the border of Mi- nas Gerais. On 1 0 May Wied-Neuwied embarked at Salvador for Lisbon, then transshipped to Ger- many through an English port. Wied-Neuwied's itinerary is difficult to track be- HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 31 Table 5. Bats (Chiroptera) of the Spix and Martius Expedition described by Spix (1823); the arrangement is phylogenetic. Current name Spix and Martins synonym Rhynchonycteris naso Wied-Neuwied, 1820 Noctilio albiventris albiventris Desmarest, 1818 Noctilio leporinus leporinus Linnaeus, 1 758 Tonatia bidens Spix, 1823 Trachops cirrhosus Spix. 1823 Glossophaga sohcina Pallas, 1 766 Carollia perspiciUata Linnaeus, 1758 Artibeus planirostris Spix, 1 823 DiphyUa Spix, 1823 Diphylla ecaudata Spix, 1 823 Thyroptera Spix. 1 823 Thyroptera tricolor Spix, 1823 Eptesicm brasiliensis Desmarest, 1823 Promops nasutus Spix Molossus ater E. Geoffroy, 1 805 Proboscidea rivalis Spix, 1 823; Proboscidea saxatilis Spix, 1 823 Noctilio albiventer Spix, 1823 Noctilio rufiis Spix, 1823 Glossophaga amplexicaudata Spix, 1823 Vampynts soricinus Spix, 1823 Molossus fumarius Spix, 1823 Molossus ursinids Spix, 1823 cause of his many roundabout journeys and short excursions with too few dates for fixing comings and goings. To add to the difficulty, the names of many localities he visited no longer exist or were never plotted on any official map; a few names have changed. Bokermann's (1957) gazetteer of nearly all localities of the Reise, with page refer- ences to their mention in Wied-Neuwied's works, is indispensable for study of the naturalist's op- erations in Brazil. Wied-Neuwied was interested in all aspects of nature, but the fauna and Indians engaged most of his attention. His species accounts are models of precision, his descriptions detailed, and com- parisons where needed are made with published descriptions by Humboldt, Azara, Buffon, and others. The bibliographic references to the species are complete. Observations of habitats and repro- duction are carefully recorded, and geographic range is usually given with circumspection. Wied- Neuwied's account of Geoffi"oy's tufted-ear mar- moset (his Hapale leucocephalus) is an example (my translation): I found it in the state of Espirito Santo. I am unable to determine if it extends north of the Rio Doce or beyond as I could not hunt often in the dark forests of this river because of the Botocudo Indians. I can therefore state that the habitat of this species lies between 20° and 21" south latitude. The animal is common in the forests of the Rio Espirito Santo, especially in the outlying bush and the mangue bush {Conocarpus and Av- icennis) bordering the river, as well as in the low palm {Allagoptera pumila and others)- covered sandy coastal districts not far from the mouth of the Espirito Santo. . . . The following excerpt of Wied-Neuwied's (1826, p. 161) observations on the golden lion tamarin (Leontopithecus rosalia rosalia Linnaeus) brings together his observations on distribution, habits, habitat, food, and reproduction: The sahuim vermelho is nowhere abundant; we saw only single individuals or family groups, particularly in the Serra da Inua, the forests of Sao Joao, and in the hilly forest surrounding Ponta Negra and Gurupina. The animal lives just as well on bushy sandy plains as in the high mountain forests. It feeds on fruits and insects and hides from strangers by disappearing into the leafy tree- tops. One or two young are produced at a birth. The female carries the offspring on her back or at her breasts [when suckling] until they are strong enough to follow her on their own. . . . Any excitement causes them to erect the long hair surrounding their faces. In gen- eral, however, their habits are similar to those of other sahuis. Wied-Neuwied also accurately delimited the distribution of the subspecies Leontopithecus ro- salia chrysomelas and added information on hab- its and reproduction. Wied-Neuwied notes (1826, p. 1 59) that "sahuis bom in Europe are carried by the father but I have never seen this here." Although generally careful in interpreting his 32 HELDIANA: ZOOLOGY ". ^.... Fig. 7. Some animals of the Wied-Neuwied Brazilian expedition: upper left, Hapale chrysomelas Wied-Neuwied (= Leontopithecus rosalia chrysomelas), possibly the holotype; upper right, Mus pyrrhorhinos Wied-Neuwied (= fViedomys pyrrhorhinos), possibly the holotype; lower left, Desmodus rufus Wied-Neuwied (= Desmodus rotundus E. Geoffroy); lower right, Felis macroura Wied-Neuwied (= Felis wiedii Schinz), possibly the holotype; from Wied- Neuwied (1822-1831). data, Wied-Neuwied could arrive at unwarranted conclusions. Among the bats collected, the leaf- nosed Phyllostomus hastatus was largest and for this reason was regarded as a blood-sucking vam- pire, although Wied-Neuwied found only insects and no blood in the stomach of this or any other bat he had examined. After confessing he had nev- er seen a bat feed on blood, he correctly blamed the large bats seen fluttering around the pack mules at night for causing them to appear next morning covered with blood. Convinced in his judgment, he described the wartlike excrescences around the mouth of innocent phyllostomine bats as adap- tations for blood-sucking. Ironically, Wied-Neu- wied (1824, 1826) later described and figured the external and dental characters of a bat he named Desmodus rufus, unaware it was a real blood-suck- ing vampire. Wied-Neuwied noted, however, that he had no opportunity to observe the live animal, because it had been captured and prepared as a specimen by assistants during his absence. The food and habits of this bat, he believed, were no different from those of other bats. The mammals of Wied-Neuwied's Brazilian ex- HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 33 pedition are described or recorded in several pub- lications. Those under Wied-Neuwied's own name are found in Isis ( 1 820, 1821), the Reise nach Bra- silien in two volumes ( 1 820, 1821), the Abbildun- gen zur Naturgeschichte Brasi liens ( 1 822-1 83 1 , see fig. 7 for some samples), and the four-volume Bei- trdge zur Naturgeschichte von Brasilien. The first volume of the last title is on reptiles, the second on mammals ( 1 826), the third and fourth on birds. Some diagnoses and binomials that Wied-Neu- wied proposed for new forms received duly ac- knowledged advance publication by Kuhl (1820) and Schinz (1821). Authorship of such newly named forms continues to be attributed to Wied- Neuwied, according to Articles 1 1 and 50 of the International Code of Zoological Nomenclature. In the few cases where Kuhl or Schinz proposed names other than those used by Wied-Neuwied, authorship is determined by priority. The actual number of mammals collected by Wied-Neuwied is unknown. According to him, they represented 82 species, but the number recognized today as valid is 7 1 . The specimens were preserved in his private museum, but duplicates were dis- tributed to the natural history museums of Berlin, Frankfurt, Leiden, and Paris. After Wied-Neu- wied's death, the remainder of the collection was sold, and the American Museum of Natural His- tory in New York acquired a part in 1869. Avila Pires ( 1 965, p. 3) affirms that fewer than 600 spec- imens of the original collection are registered in the catalogue of mammals of the New York in- stitution. Of these, only 38 skins and 16 skulls are of South American origin. Included are holotypes (or syntypes) of Didelphis aurita Wied-Neuwied, Didelphis cinerea Temminck, Molossus plecotus Wied-Neuwied, Phyllostoma brevicaudum Wied- Neuwied, Vespertilio leucogaster Wied-Neuwied, Vespertilio naso Wied-Neuwied, Hypudeus dasy- trichos Wied-Neuwied, and Mus pyrrhorhinos Wied-Neuwied. Table 6 lists all mammalian species recorded by Wied-Neuwied. Current names are used; syn- onyms used by Wied-Neuwied are included. Johann Natterer (1787-1843) Johann Natterer, bom near Vienna, was well schooled in the sciences, especially biology, and in modem languages and illustration. Natterer's father, the imperial falconer and collector of birds and insects, taught him to hunt and preserve an- imals as museum specimens. In 1 8 1 6 he was em- ployed as assistant in the Imperial Natural History Museum of Vienna and in 1817 was appointed member of an expedition to investigate the Bra- zilian biota. He arrived in Rio de Janeiro on 5 November accompanied by Mikan and Pohl, both naturalists, and Schott, a botanist. Within a year Mikan, Sochor, a hunter, and two artists who were to accompany Natterer, retumed to Europe. Pohl and Schott retumed in 1821. Natterer was primarily a bird collector, but his interest in collecting extended to mammals, other vertebrates, insects, and parasitic helminths. He traveled light and, as a rule, worked alone or with few native helpers (Ihering, 1 902). He collected in most of the eastem coastal states and in Mato Grosso and the Amazonian region between the Rios Tapajos and Madeira and in the Rio Negro basin north of the Rio Amazonas (fig. 8). His main base for the first five years was Ipanema, Sao Pau- lo. His itinerary— with goings and comings, side trips, short stopovers in some sites, long delays in others— was arranged chronologically by Pelzeln (1871,1883) into "Reisen" (or journeys), with dates given for all points visited, and is summarized below. Only general areas or terminal points and inclusive dates are given. Johann Natterer's Brazilian Reisen, 1817-1835. I. Rio de Janeiro, 5 November 1817 to 1 No- vember 1818. II. Eastem Sao Paulo, 2 November 1818 to March 1820. III. Southern Sao Paulo to boundary between Rio Grande do Sul and Rio de Janeiro, July 1820 to 1 February 1821. IV. Rio de Janeiro, Sao Paulo, 1 February to September 1822. V. Northern Sao Paulo, Goias, eastem Mato Grosso, Minas Gerais, October 1822 to 31 December 1824. VI. Mato Grosso, January 1825 to July 1829. VII. Mato Grosso, Rio Madeira, and upper trib- utaries to Borba in Amazonas (Capitania Rio Negro), 15 July 1829 to June 1830. VIII. Borba to Rio Negro, Rio Casiquiare, Ven- ezuelan border, retum to Barcelos and Bor- ba, June 1830 to 31 August 1830. IX. Rio Negro from Barcelos to Rio Branco, 5 September 1831 to 2 July 1832; Barra do Rio Negro, 29 August 1832 to 7 July 1834; Rio Tapajos, August 1834. X. Para, Maranhao, Rio Grande, Paraiba, Per- 34 FIELDIANA: ZOOLOGY ■ ■ ■ /yj/__/"/»«-»r ornt .\tm^ni6er A*!// <»«.»<' ' 6fx Ar^rintr fS'2/ /«y.//r/«- /«;'/< 17'f/l'ri. ./mm /^ifA/jt . 1ntft,*l /xV 1/:^/- /S'y? .-r -IZ Fig. 8. Map of Brazil showing routes of Johann Natterer (bold line); from [brother of Johann] Natterer (1833, Oken's Isis, heft VI, pi. 14). nambuco, Bahia, Rio de Janeiro, September Natterer's enormous collections were sent to the 1834 to September 1835 (no mammal col- Vienna museum and, except for the birds and lections). mammals, were never fully reported. His friend Sailed for Europe 15 September 1835. Andreas Wagner (1797-1861) described most of HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 35 Table 6. Brazilian mammals recorded by Wied-Neuwied (1826) with some figured in the Abbildungen (1822- 1831); the arrangement is phylogenetic. Current name Wied-Nenwied synonym Figure Marsupiaua Afarmosa murina Linnaeus, 17S8 Marmosa cinerea Temminck, 1 824 Philander opossum frenata Olfers, 1818 Didelphis marsupialis aurita Wied-Neuwied, 1826 Chiroptera Rhynchonyaeris naso Wied-Neuwied. 1820 (Reise) Centronyaeris maximiliani Fischer, 1 829 Peropteryx macrotis Wagner. 1843 Diclidurus albus Wied-Neuwied, 1819 Noctilio leporinus Linnaeus, 1 758 Xfacrophyllum macrophyllum Wied-Neuwied, in Schinz, 1821 Phyllostomus hastatus Pallas, 1867 PhvUostomus obscunis Wied-Neuwied, in Schinz, i821 Glossophaga soricina Pallas, 1 766 Anoura caudifera E. Geoffrey, 1818 CaroUia brevicauda Wied-Neuwied, 1821 Carollia perspicillata Linnaeus, 1 758 Artibeus liturcaus 0\fcT%, 1818 Desmodus rotundus E. Geoffiroy, 1810 Myotis albescens E. Geoflfroy, 1806 Myotis nigricans Wied-Neuwied, in Schinz, 1821 Eumops perotis Wied-Neuwied, in Schinz, 1 82 1 Primates Callithrix jacchus penicillatus E. Geoffix)y, 1812 Callithrix jacchus geoffroyi Humboldt, 1812 Leontopithecus rosalia chrysomelas Kuhl, 1 820 Leontopithecus rosalia rosalia Linnaeus, 1758 Callicebm personatus persoruUus E. GeoflSroy, 1812 Callicebus personatus melanochir Wied-Neuwied, 1820 (Reise) Alouatta caraya Humboldt, 1812 Alouatta fusca E. Geoffroy, 1812 Cebus apella nigritus Goldfiiss, 1 809 Cebus apella robustus Kuhl, 1 820 Cebus apella xanthostemos Wied-Neuwied, 1820 (Reise) Brachyteles arachnoides E. Geoflfroy, 1806 Edentata Tamarulua tetradactyla Linnaeus, 1758 Myrmecophaga tridactyla Linnaeus, 1758 Edentata Bradypus torquatus Desmarest, 1816 Didelphys myosuros Temminck. 1 825 Didelphis marsupialis. Wied-Neuwied, 1826, not Linnaeus Vespertilio caJcaratus Wied-Neuwied, in Schinz, 1821, not Rafinesque, 1818 Vespertilio caninus Wied-Neuwied, in Schinz, 1821, not Blumenbach, 1797 Diclidurus freyreissii Wied-Neuwied 1822, Abbild. Noctilio dorsatus Desmarest, 1818; Noctilio unicolor Desmarest, 1818 Artibeus planirostris Spix, 1823 Glossophaga amplexicaudata E. Geoflftoy, 1818 Phyllostoma bernicaudum {sic) Wied-Neuwied, in Schinz, 1821 Phyllostoma brachyotos (sic) Wied-Neuwied, in Schinz, 1821 Phyllostoma superciliatum Wied-Neuwied, in Schinz, 1821 Rhinolophus ecaudatus Wied-Neuwied, in Schinz, 1821; ZJesmorfus rw^ Wied-Neuwied, 1824 Vespertilio leucogaster Wied-Neuwied, in Schinz, 1821 Hapale penicillatus kuhlii Wied-Neuwied, 1826 p. 142)* Hapale leucocephalus Kuhlii (sic), Wied-Neuwied, 1826t Mycetes niger Kuhl, 1820 Mycetes ursinus Humboldt, 1812, not Humboldt, 1805 Cebus cirrifer E. Geoffroy, 1812, not Cebus fatuel- lus Linnaeus ? Cebus flavus E. Geoffroy, 1812 Ateles hypothanthus Kuhl, 1820 Myrmecophaga jubata Linnaeus, 1766 Bradypus tridactylus Wied-Neuwied, 1826, not Linnaeus, 1758 36 HELDIANA: ZOOLOGY Table 6. Continued. Current name Wied-Neuwied synonym Figure Cabassous unicinctus Linnaeus, 1758 Euphractus sexcinctus Linnaeus, 1758 Dasypus novemcinctus Linnaeus, 1 758 Priodontes maximus Kerr, 1 792 Carnivora Dusicyon thous brasiliensis Wied-Neuwied, in Schinz, 1821 Chrysocyon brachyurus Illiger, 1815 Nasua nasua solitaria Wied-Neuwied, in Schinz, 1821 Procyon cancrivorus G. Cuvier, 1 798 Potosflavus nocturnus Wied-Neuwied, 1826 Eira barbara Linnaeus, 1758 Pteronura brasiliensis Gmelin, 1 788 Felis wiedii Schinz, 1821 Felis pardalis mitis F. Cuvier, 1820 Felis yagouaroundi eyra Fischer, 1814 Felis concolor Linnaeus, 1 77 1 Felis onca Linnaeus, 1758 SiRENIA Trichechus manatus lAnnditns, 1758 Perissodactyla Tapirus terrestris lArmaitxis, 1758 Artiodactyla Tayassu tajacu Linnaeus, 1758 Tayassu pecari Link, 1795 Mazama gouazoubira Fischer, 1814 Mazama americana Er\\ehcn, Mil Blastoceros bezoarticus Linnaeus, 1758 Blastocerus dichotomus Illiger, 1815 Lagomorpha Sylvilagus brasiliensis Linnaeus, 1758 RODENTIA Sciurus aestuans Linnaeus, 1 766 Wiedomys pyrrhorhinos Wied-Neuwied, 1821 (Reise) Oxymycterus rufus dasytrichos Wied-Neuwied, in Schinz, 1821 Proechimys myosuros Lichtenstein, 1818 Cavia aperea Erxleben, 1 777 Kerodon rupestris Wied-Neuwied, 1820 (Isis) Hydrochaeris hydrochaeris Linnaeus, 1 766 Dasyprocta leporina aguti Linnaeus, 1 766 Agouti paca Linnaeus, 1 766 Coendou insidiosus Olfers, 1818 Chaetomys subspinosus Olfers, 1818 Dasypus setosus Wied-Neuwied, 1 826; Dasypus gilvipes Illiger, 1815 Dasypus longicaudus V^icd-Neuwied, 1826 Dasypus gigas Cuvier, 1822 Canis azarae Wied-Neuwied, 1 823 Canis campestris Wied-Neuwied, 1826 A^asMfl 50c/a//5 Wied-Neuwied, 1826 Mustela gulina Wied-Ncuwied, 1821 Felix macroura Wied-Neuwied, 1 823 Felis pardalis. Wied-Neuwied, 1826 Felis yaguarundi, Wied-Neuwied, 1 826 Felis brasiliensis Wied-Neuwied, 1 82 1 Manatus americanus Link, 1795 Tapirus americanus Gmelin, 1788 Dicotyles torquatus Cuvier, 1817 Cervus simplicicornis Illiger, 1815 Cervus rufus Cuvier, 1817 Cervus campestris 'Wied-Nexxwied, 1826, not Cuvier, 1817 Cervus paludosus Desmarest, 1 822 Hypudeus dasytrichos Wied-Neuwied, 1826 * The name is a correctly formed trinomial but this form was not in use at the time, and Wied-Neuwied used no trinomials elsewhere in his publications on Brazilian mammals. t The name appears to be a trinomial although the patronymic, properly in the genitive, is not italicized. Most likely Wied-Neuwied meant to cite Kuhl for this and the preceding taxon as authority for his use of the names in question. It was common practice at the time to cite the author who replaced an earlier generic name with a different one. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 37 the new mammalian species in a series of reports published in the Archivfur Naturgeschichte ( 1 842, 1843), in the Abhandlungen der Akademie .Vfiin- chen ( 1 847-1 849), and his supplementary volumes of Schreber's Sdugethiere (1840-1855). Finally, Pelzeln (1883) brought together most, if not all, available taxonomic, descriptive, and geographic data in a single report. Natterer intended to work up the entire collection himself, but died within a few years of his return to Europe. His journal, with notes on habits, reproduction, and anatomy of the Brazilian animals collected, was lost. Natterer collected 78 1 specimens of mammals, representing more than half (58%) of the currently known Brazilian genera and nearly as many (44%) of the species (table 3). Most poorly represented are bats, mice, and mouse opossums. Had Natterer been equipped with suitable traps and trammel nets known at the time but not used in fieldwork, he might have collected nearly all the mammalian genera and species now known to occur in Brazil. Still, his collection represented more species and included t\pes of more new species than had been collected in Brazil by anyone else in the century, or p)ossibIy at any time. The numbers of genera and species of mammals collected by Natterer, as identified by Pelzeln (1883), are listed in Table 7. The totals are com- pared with the numbers currently recognized, some genera having been increased and some species eliminated by synonymy. The revised numbers of genera and species are shown, in turn, as percent- ages of the estimated total numbers of currently known Brazilian genera and species of mammals. VII. Guianas: Mammalogy to End of 18th Century Pierre Barrere (1690-1755) The physician, botanist, and correspondent of the French Royal Academy of Sciences, Pierre Barrere, resided three years (1752-1755) in Cay- enne, with instructions to prepare a detailed report on the natural history of French Guiana. The work he finally published in 1 74 1 , however, is no more than an abbreviated glossary- of Guianan minerals, plants. moUusks, fishes, reptiles, birds, and mammals. The list of mammals was uncrit- ically compiled from Marcgraf and others. Species previously recorded by early chroniclers from the lower Rio Orinoco region which occur throughout the Guianas but were not mentioned by Barrere are the golden handed tamarin {Saguinus midas), red brocket (Mazama americana), red acuchi (A/y- oprocta exilis), tayra (Eira barbard), white-lipped peccary (Tayassu pecari), and silky anteater (Cy- clopes didactylus). Jose Gumilla (d. 1750) A natural history and geography of the Rio Ori- noco region in Spanish, published by Father Jose Gumilla, provides interesting, but largely erratic, descriptions of the countryside and human inhab- itants, but nothing of interest regarding native mammals. Gumilla's explorations of the interior led him to deny the reported existence of a con- nection between waters of the Orinoco and Negro rivers. Jacques Nicolas Bellin (1703-1772) The description of the Guianan possessions of France, Spain, Holland, and Portugal, from the Orinoco River to the Amazonas River, by Jacques Nicolas Bellin, published in 1763, contains infor- mation on natural history, but adds nothing note- worthy to the then-known mammalian fauna. Edward Bancroft (1744-1821) The English physician Edward Bancroft lived three years in Dutch Guiana, now Suriname, prac- ticing medicine and gathering notes for his Essay on the Natural History of Guiana. The work, pub- lished in 1 769, deals broadly with plants and an- imals, but the author's knowledge of mammals was mostly limited to hearsay, although he also made some observations on animals brought to him by natives or seen in captivity or during short walks into the countryside. Persistent reports of the existence of apes or ape-men in South America were recounted by Bancroft (p. 1 30) in these terms: The Orang-Outang of Guiana is much larger than either the African or the Oriental, if the accounts of the natives may be relied on; for I do not find that any of them have been seen by the White inhabitants of this coast, who never penetrate far into the woods. These animals, in all the different languages of the Natives, are called by names signi- 38 HELDIANA: ZOOLOGY Table 7. Numbers of mammalian genera and species collected by Johann Natterer in Brazil, 1817-1835, based on Pelzeln (1883), and compared with currently known totals. Table 7. Continued. Total Taxon Number reported by Pelzeln (1883) Current equiva- lent number cur- rently known for Brazil (esti- mated, 1984) Percent- age of current total col- lected by Natterer Marsupialia Genera Species 2 18 6 15 8 30 75% 50% Chiroptera Genera Species 10 48 28 40 60 125 47% 32% Primates Genera Species 12 45 14 28 16 50 87% 56% Edeimtata Genera Species 10 16 10 12 12 15 83% 75% Carnivora Genera Species 11 17 10 14 14 25 71% 56% PiNNIPEDIA Genera Species 0 0 0 0 2 2 0% 0% SlRENlA* Genera Species 1 1 1 1 1 2 100% 50% Perissodactyla Genera Species 1 1 1 1 1 1 100% 100% Artiodactyla Genera Species 4 7 4 6 5 7 80% 86% Lagomorpha Genera Species 1 1 1 1 1 1 100% 100% RODENTIA Sciuropmorpha Genera 1 Species 5 1 3 3 6 33% 50% Myomorpha (Murinae Genera 3 Species 1 7 excluded) 5 17 20 45 25% 24% Caviomorpha Genera Species 11 24 15 22 23 47 65% 47% Cetacea* Genera Species 2 2 2 2 2 2 100% 100% Taxon Total cur- rently known Percent- Number for age of reported Current Brazil current by equiva- (esti- total col- Pelzeln lent mated, lected by (1883) number 1984) Natterer Totals Genera 69 99 170 58% Species 202 156 358 44% * Fresh water only. fying a Wild Man. They are represented by the Indians as being near five feel in height, maintaining an erect position, and having a human form, thinly covered with short black hair; but I suspect that their height has been augmented by the fears of the Indians, who greatly dread them, and instantly flee as soon as one is discovered, so that none of them have ever been taken alive, much less at- tempts made for taming them. The Indians relate many fabulous stories of these ani- mals; and, like the inhabitants oi Africa and the East, assert, that they will attack the males, and ravish the females of the human species. Philippe Fermin (1720-1790) Philippe Fermin, the author of an account pub- lished in 1769 of the history, geography, and nat- ural objects of colonial Suriname, was one of those European men who "never penetrate far into the woods." Indeed, Fermin believed that all Euro- [)eans and Creoles were physically incapable of coping with the difficulties of surveying the natural fauna of the countryside, let alone the wilderness, or resisting the diseases generated by the "foul" air of forests and swamps. Notwithstanding this, Fermin compiled a fair list of the mammals. The didelphids included Didelphis marsupialis, Phi- lander opossum, and Marmosa spp. All three kinds of anteaters and the two- and three-toed sloths are mentioned. The two native squirrels, Sciurus aes- tuans and Sciurillus pusillus, are distinguished. Other rodents are the capybara, paca, a porcupine, cavy, spiny rats or echimyids (most likely of the genera Proechimys and Echimys), and a water rat HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 39 (probably Nectomys squamipes). The carnivores include tayra, otter, jaguar, puma, margay, ocelot, and two kinds of bush dogs (Dusicyon thous and possibly Speothos venaticus). Monkeys are Sa- guinus midas, Saimiri sciureus, Pithecia pithecia, Chiropotes sat anas, Cebus apella, Alouatta seni- culus, Ateles paniscus. and variants of some of them regarded as distinct species. African simians introduced with the slave trade and mentioned by Marcgraf are included. The ungulates are tapir, brocket {Mazama americana), and the collared and white-lipped peccaries. Regarding white-lipped peccaries, Fermin af- firms they form herds of as many as 300 individ- uals. Hunters, he states, tremble when they hear the sound of their clicking tusks. When attacked, only two avenues of escape are open: The first is a tree, if it can be climbed; the second and surest is standing ground and urinating, the odor of the urine, he affirms, being a powerful peccary repel- lant. Monsieur Bajon (1763?) The French physician, surgeon, and anatomist Bajon, with 1 2 years' residence in French Guiana, investigated climate, agriculture, natural history, and human diseases. The knowledge he gained was acquired firsthand, much of it new or supplemen- tary to what was already contained in the ency- clopedic volumes on natural history by Buffon and Daubenton. In the second of his two-volume work, Bajon (1778, p. 178) declared that, contrary to popular belief, the jaguar feared man and did not attack without provocation. His accounts of habits and detailed descriptions of intestinal morphology and female genitalia of peccaries supplement Dauben- ton's (in Buffon) gross anatomy of a male collared peccary. Bajon clarified the differences between the agouti (Dasyprocta leporina) and acouchi {My- oprocta exilis). He described the male agouti penis, with its peculiar complement of spines, erectile spears, and sharp blades. Descriptions with life history notes are given for the chien sauvage {Du- sicyon thous), eira {Eira barbara), and chien cra- bier (Procyon cancrivorus). Marsupials fascinated him, particularly the role of the pouch in females of the pean {Didelphis marsupialis), quatre-ouel (Philander opossum), and also the pouchless rat de bois {Marmosa sp.). The commonly held belief that each didelphid young is bom and develops at the end of a teat was rejected by Bajon, but despite numerous observations and dissections, he failed to solve the mystery of marsupial birth. Bajon's monographic account of the tapir {Tap- irus terrestris) includes detailed, but not always accurate, descriptions of anatomy, reproduction, development, behavior, food, vocalization, hunt, and human utilization. John Gabriel Stedman (1744-1797) A soldier of the Scots Brigade of the Nether- lands, John Gabriel Stedman arrived in Suriname in 1773 to help subdue the uprising of the African slaves. Most of the fighting was already over when he landed, so Stedman devoted much of his time to recording his observations of life in the country and wilderness. His Narrative, published in 1 796 in two volumes, contains much on the natural history of Suriname, with illustrations by his own hand (fig. 9). The mammals, some only listed, oth- ers described, often with anecdotes, are the fol- lowing. Stedman used local names, current bino- mials are in parentheses. Volume I, p. 14. Narwhal (Monodon monoceros). Sighted from shipboard at Devil's Island off Cayenne. ". . . appeared but six or eight feet in length, and its horn about four. . . . The narwhal ... is more frequently found in cold than warm climates. The female is said to be unprovided with that protuberance so re- markable in the male. It appears that some authors have confounded this animal with the sword-fish, to which however it does not prove to have the very smallest resemblance." The locality record for the circumpolar narwhal is unexpected, and no doubt erroneous. Nevertheless, Stedman's description is accu- rate albeit the dimensions given seem small. At the same time, Stedman provided a de- tailed description and good figure of a sword- fish or sawfish to prove it was not a sawfish he saw! Volume I, p. 153, pi. 16. Sicapo (Bradypus tri- dactylus). Volume I, p. 153, pi. 16. Dago luyaree (Choloepus didactylus). Volume I, p. 153. Ourang-outang. "I have never seen, nor heard described, while I was in this country. . . ." Volume I, p. 166, pi. 18. Micoo or mecoo {Cebus apella) (fig. 9). 40 HELDIANA: ZOOLOGY •^ jT^^-..^ _^ p* \^^^ )& 'J] J^ u* JC^ # 1 H^'^^!^ %sj ■ .^.:.A^ ■ '^^^ •c s u E t^ < c j= o s (> o it . ^L O J \ (MAY 1800) ^^ ( ■"^ s,y//f ^ ; Cerro Dulda \ / \ '^ --*^f\ ^^^^ — . Vv" jL Q La Esneralda^^ / jg ,• Ms) ^' H**"^" PORTAGE f~^ .i^-r O'^*""*'*"'^-?^ \ ( ^d— — 3" ^< ^^ y^ - 3° »^>^ ^ ^\ y"^ y^ \y< \. / —r' ^^/\J * "'^ / y^^"^ Sa "^vK ^}^^^ ^ Boca MaWcarJ v* /^ V— ^ / /<^ ^X- , fr^ / 1 /^\ *"# ~ — M L ( til 31 ■ *^ \ y^ / ^^ \ 1 f Al *N / ( • 1 _ Solano ^' ^^x^^ / ^^-^^ 2° / \ . „Jl ""^ "^o Negro/ ^ n^s^^ 2" ^\ ^\ *. Vi WJ r-^^ \ \ \ ^ ^^\} r\ A ^>^ V \. J \ 7 • ^ I ^Vr'^ \ ( / \" \ A V-P^ • tf 3V ""^^""^'^ \ \ ( \ ^ X * \0 ^ >. l^.'°\.Cucuy 3\ \ ( ^^y -^'j 1° ^^^"^^^ '^ > A^E^ Catarat. de^^ Cerro de la Nebllna_ _^« \ / ^ "^ 1° ^ :>a* '*"* X ^ ^ \i 1 r^ 1 ) 1 \ 67° 66° 6S- Fig. 1 2. Route of Humboldt and Bonpland in Amazonas, Venezuela, from the Rio Orinoco-Atabapo to the Rio Guainia-Negro via p>ortage between the Rio Temi and Rio Pimichin and the Rio Casiquiare connecting the Negro and Orinoco. expedition up the Rio Orinoco for verification of its reputed connection with the Amazonian Rio Negro. The exploration began on 27 March 1800 with a three-day inspection of a western tributary, the Rio Apure. The journey then continued up the mainstream to the Spanish mission of San Fer- nando de Atabapo near the confluence of the Rios Atabapo and Guaviare with the Orinoco. At this point, the travelers left the Orinoco and continued up the Atabapo to the tributary Temi, which they followed to the tiny mission of Yavita, arriving on 1 May. On 10 May, after portage to the Rio Pimichin, a tributary of the Guainia, they attained San Carlos de Rio Negro at the mouth of the Rio Casiquiare. The next day they headed up the Ca- siquiare and, after 10 days' travel by water, reen- tered the Orinoco on 21 May (figs. 1 1-12). Having confirmed the connection between the HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 53 waters of the Orinoco and Amazon rivers, the ex- plorers shipped 750 miles downstream to arrive at Angostura (Ciudad Bolivar) in mid-June 1800. After more work on the coast, Humboldt and Bon- pland departed Venezuela on 24 November 1800 for Havana, Cuba. They remained there until 1 7 March 1 80 1 , then sailed for Colombia with land- ings along the Rio Sinu on 25 March and Carta- gena on 30 March. The journey thereafter was devoted mainly to explorations of the Cordilleras of Colombia and Ecuador, then through moun- tains, deserts, and the upper Amazonia of Peru south to Lima. The few mammals observed or described during this part of the journey are men- tioned in Humboldt's (1805-181 1) Recueil. From Lima, Humboldt and Bonpland em- barked on 24 December 1 802 for Guayaquil and left 15 February 1803 for Mexico. Humboldt's lively Personal Narrative evokes vi- sions of Venezuelan life and landscapes from coastal plains to the headwaters of the Rio Ori- noco. The narrative is replete with descriptions of geography, ecology, astronomical orientations, widths, depths, and volumes of rivers, histories, languages and customs of Indians, Catholic mis- sions, missionaries, and the human interest trials and tribulations of the travelers. Information on mammals, however, is comparatively meager, but some interesting bits can be quoted or paraphrased from the Ross translation of the original French (Humboldt, 1884). Humboldt and Bonpland found manatees abun- dant in the Rio Orinoco and tributaries Meta and Apure, but absent above the cataracts of Mai- pures. Some of the animals they caught were 1 0 to 12 feet long and weighed 500 to 800 pounds. Humboldt's dissection of one (fig. 1 3) revealed "no vestige of nails on the external surfaces of the fins which were quite smooth, but little rudiments of nails appear at the third phalanx when the skin of the fins is taken off." The lungs, they observed, consisted of "large cells resembling immense swimming bladders; they [the lungs] are 3 feet long. Filled with air they have a bulk of more than a thousand cubic inches [Humboldt, Ross transla- tion, 1884, vol. II, p. 169]." Its distinction from T. manatus was not appreciated, however, until 1 883 when described by Natterer (in Pelzeln, 1 883). There is also considerable doubt that a clawless manatee does occur in the Rio Orinoco basin or anywhere outside the Amazonian watershed. Dolphins (Sotalia) were seen above and below the great cataracts of the Orinoco and often swam alongside the canoe. In the inundated forest of the divide between the waters of the Orinoco and Ne- gro, the travelers "were astonished by an extraor- dinary noise. On beating the bushes a shoal of toninas (fresh-water dolphins) four feet long sur- rounded our boat. They fled across the forest, throwing out those spouts of compressed air and water. . . ." Other Venezuelan mammals mentioned in the narrative include the expected jaguar, otter, deer, peccaries, capybara, and vampire bats. Monkeys, however, absorbed more of Hum- boldt's attention than other animals. He carried with him a number of live simians captured in the upper Rio Orinoco region for shipment to the Jar- din des Plantes in Paris via the Antillean island of Guadeloupe. The newly discovered bearded saki {Chiropotes satanas chiropotes Humboldt; fig. 14) died before transshipment, but its skin was saved and arrived in Paris. The type specimen of red howler, Simla urslna Humboldt (= Alouatta se- nlculus arctoides Cabrera) survived the journey, whereas the first-known douroucouli or night monkey {Aotus trivlrgatus Humboldt; fig. 14) suc- cumbed in Guadeloupe. Humboldt often mentioned the ubiquitous, highly visible howler or araguato {Alouatta seni- culus). At one time he saw from the road below troops of 30 to 40 individuals crossing through the trees. In a carefully deployed experiment in Ara- gua, he calculated the distance the howler's vo- calization could be heard as 800 toises (6 ft 4.73 inches x 800 = 5,1 15 ft) or nearly 1 mile (5,280 ft). Humboldt (Ross translation, 1884, vol. II, p. 453) recounts the Indian tale of bearded sakis (Chiropotes) and uacaries (Cacajao) of the Orinoco "placing themselves in a circle and, by striking the shell [of the Brazil nut pericarp] with a stone, suc- ceed in opening it so as to take out the triangular nuts." Although Humboldt dismissed the story as fabulous, he did believe that the monkeys cracked the shell of the Bertholletia nut with their teeth to obtain the meat which they devoured with gusto. Belief in the existence of a hairy man of the woods was practically universal. The missionary Father Gili gravely related to Humboldt the tale of a woman "in the town of San Carlos in the Llanos of Venezuela who much praised the gentle character and attentions of the man of the woods. She is stated to have lived several years with one in great domestic harmony, and only requested some hunters to take her back because she and the children (a little hairy also) were weary of living so far from the church and the sacraments." Hum- 54 FIELDIANA: ZOOLOGY IkbU. ^ % Fig. 1 3. The Orinoco clawless manatee, supposedly Trichechus inunguis Natterer left, lateral (1) and ventral (2) views; right, head from above (1), mouth, upper inner view (2), mouth, lower inner view (3), mouth, side view (4), and trunk, sagittal section (5); original illustrations by Humboldt; from Humboldt (1838). boldt resented that he and Bonpland "were every- where blamed, in the most cultivated class of so- ciety, for being the only persons to doubt the reality of the great anthropomorphic monkey of Ameri- ca." Humboldt's Recueil d 'Observations de Zoologie et d 'Anatomic Comparee, a collection of memoires published as a volume in 181 1-1812, deals with many species of invertebrates and vertebrates, but a large share of the text is about monkeys. One memoir with excellent illustrations by Humboldt is on the comparative anatomy of the hyoid bone and larynx of the cotton-top tamarin (Saguinus oedipus oedipus Linnaeus; fig. 1 4), and that of the red howler {Alouatta seniculus seniculus Lin- naeus), the Colombian squirrel {Sciurus granaten- sis granatensis Humboldt; fig. 14), birds, and croc- odiles, all from the Rio Magdalena region. Another memoir on the carnivores includes descriptions of Gulo quitensis (= Conepatus chinga quitensis Humboldt) from Quito, Ecuador, Mustela sinuen- sis (= Eira barbara sinuensis Humboldt), from the Rio Sinu, Colombia, and a discourse on other mustelids and the kinkajou {Potos Jlavus Schre- ber). The memoir on monkeys of the upper Rio Orinoco and connecting Rios Casiquiare and Ne- gro includes the original descriptions o^ Aotus tri- virgatus, Chiropotes satanas chiropotes, Cacajao melanocephalus, Callicebus torquatus lugens, La- gothrix lagothricha, and Cebus albifrons. A chap- ter on the monkeys of Colombia and the upper Amazonian region includes the description of a representative each of Cebus capucinus Linnaeus from the Rio Sinu, A teles belzebuth marginatus E. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 55 V M. PSIT TAl- 1 S AR Al' RAN A. N" ^ N° Ml. SI n HIS (.RANATF.NSIS. _;?•«• V. vm. SIMIA OKDIHl S. -. •»« ^ Fig. 1 4. Monkeys and anatomical dissections from Humboldt (1811): upper left, Simia melanocephala Humboldt (= Cacajao melanocephalus), holotype; lower left, two views of Simia trivirgata Humboldt (= Aotus trivirgaius), holotype; upper right, Simia satanas Hoffmannsegg (= Chiropotes satanas satanas), lectotype; lower right, throat cartilages of Psiltacus araurana Linnaeus (= Ara araurana), Sciurus granatensis Humboldt, and Simia oedipus Linnaeus (= Saguinus oedipus oedipus). 56 HELDIANA: ZOOLOGY Table 8. New World monkeys (Platyrrhini) recorded by Humboldt (1812); the arrangement is phylogenetic. Current name Humboldt synonym Figure Callitrichidae Callithrix jacchus jacchus Linnaeus, 1758 Callithrix jacchus penicillata E. Geoffroy, 1812 Callithrix jacchus geoffroyi Humboldt, 1812 Callithrix jacchus aurita E. Geoffroy, 1812 Callithrix humeralifer humeralifer E. Geoffroy, 1812 Callithrix argentata melanura E. Geoffroy, 1812 Callithrix argentata argentata Linnaeus, 1771 Saguinus fuscicollis fuscus Lesson, 1840 Saguinus labiatus labiatus E. Geoffroy, 1812 Saguinus midas niger E. Geoffroy, 1 803 Saguinus midas midas Linnaeus, 1758 Saguinus oedipus oedipus Linnaeus, 1758 Leontopithecus rosalia rosalia Linnaeus, 1 766 Cebidae Saimiri sciureus cassiquiarensis Lesson, 1 840 Callicebus moloch moloch Hoffmannsegg, 1 808 Callicebus torquatus lugens Humboldt, 1811 Callicebus torquatus torquatus Hoffmannsegg, 1 808 Callicebus personatus personatus E. Geoffroy, 1812 Actus trivirgatus Humboldt, 1811 Actus azarae azarae Humboldt, 1811 Pithecia mcnachus monachus E. Geoffroy, 1812 Pithecia pithecia pithecia Linnaeus, 1 766 Chiropotes satanas satanas Hoffmannsegg, 1 808 Cacajac melanocephalus Humboldt, 1811 Alouatta caraya Humboldt, 1812 Alouatta seniculus arctoidea Cabrera, 1 940 Alouatta seniculus straminea Humboldt, 1812 Cebus capucinus capucinus Linnaeus, 1758 Cebus nigrivittatus nigrivittatus Wagner, 1 848 Cebus apella apella Linnaeus, 1 758 Cebus apella xanthosternos Wied-Neuwied, 1 820 Cebus apella nigritus Goldfuss, 1810 Lagcthrix lagcthricha lagothricha Humboldt, 1812 Lagothrix lagcthricha cana E. Geoffroy, 1812 Lagcthrix flavicauda Humboldt, 1811 Atetes paniscus chamek Humboldt, 1812 Ateles paniscus paniscus Linnaeus, 1 766 Ateles belzebuth belzebuth E. Geoffroy, 1 806 Ateles belzebuth marginatus E. Geoffroy, 1809 Brachyteles arachnoides E. Geoffroy, 1 806 Jacchus leucccephalus Geoffroy, 1812 Simla leonina Humboldt, 1805, not Shaw, 1800 Simla Ursula Hoffmannsegg, 1808 Not Simla sciurea Linnaeus Simla amicta Humboldt, 1811 Pithecia rufiventer E. Geoffroy, 1812; Simla leuco- cephala E. Geoffroy, 1812 Simla ursina Humboldt, 1805, not Bechstein, 1800 Simla hypoleuca Humboldt, 1811 Simla capucina Humboldt, 1812, not Linnaeus, 1758 Cebus barbatus Humboldt, 1812, attributed to E. Geoffroy Simla variegata Humboldt, 1812, not Kerr, 1 792 Simla cirrifera Humboldt, 1812; Cebus niger E. Geoffroy, 1812 13 14 13 13 Simla chuva Humboldt, 181 1, p. 340; 1812, p. 362, footnote 2 Geoffroy from lower Amazonia, Alouatta senicu- lus Linnaeus from the Rio Magdalena, and La- gothrix flavicauda Humboldt from northern Peru. In an addendum, Humboldt listed all platyrrhine monkeys known to 1812. They are arranged in Table 8 by current scientific names with Hum- boldt's synonyms. X. PARAGUAY The Paraguayan province, claimed by Spain, was first visited in 1526 by Sebastian Cabot and then explored by Cabeza Alvarez Nunez de Vaca in 1541. For the next two centuries, waves of mis- HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 57 Fig. 1 5. Map of Azara's Paraguay and adjacent parts of Brazil and Argentina; from Azara (1809). 58 FIELDIANA: ZOOLOGY sionaries and colonists penetrated to the remotest comers of the province in quest of climates or environments that resembled or could be trans- formed into the familiar ones of Spain. The monumental Histoire du Paraguay by the Jesuit missionary Pierre Francois Xavier de Char- levoix (1682-1761), published in 1757, describes the land that extended from the Atlantic to the eastern base of the Andes between latitudes 1 5° and 35° in the drainage basin of the Rio Parana- Paraguay. It relates the history of the province from the time of the conquest, describes native customs, conversions to Christianity, and estab- lishment of missions. The little of natural history in the text adds nothing about wild mammals not already recorded by others. Two decades later Fe- lix de Azara wrote the most complete natural his- tory account of the mammalian fauna of Paraguay for its time and ever since. Felix de Azara (1746-1811) The Spaniard Don Felix de Azara (1 746-1 811), an army engineer, was commissioned in 1781 to assist in defining the boundaries between Spanish and Portuguese territories. Unmapped territories between Brazil and Paraguay were assigned to Azara, but the Portuguese showed no interest in their share of the work. With time on his hands and a disposition toward the natural sciences, Azara devoted nearly the full 20 years, from 1781 to 1800, of his American residence to the study of geography, Guarani Indians, and the birds and mammals of Paraguay and northeastern Argen- tina between 24° and 36°S and about 54°3r to 56°W (or 60°W of Greenwich) (fig. 15). With no schooling in the natural sciences and no books for reference or guidance, Azara de- pended on his own resources. They proved ade- quate. Azara recorded his observations with care, precision, meticulous attention to detail, and rig- orous exclusion of speculation and fantasy. His anatomical descriptions, measurements, and ac- counts of behavior were based on animals ob- served in the wild or in captivity, usually in his own home or garden. Useful information received from others was credited to the informants. Pop- ular beliefs and hearsay were labeled as such. Without other sources of information, Azara used the Guarani names for most of the amimals he described and Spanish epithets for the remainder. The manuscript of the mammals or quadrupe- dos of Paraguay contained accounts of 66 species. Shortly after its completion, the author received a shipment of several volumes of a Spanish trans- lation of Buffon's Histoire Naturelle. Not surpris- ingly, Azara found in them much with which to disagree, but some of his adverse criticism was unfair. Azara knew Paraguayan mammals better than anyone else, but only a minority of the species were the same as the Neotropical species described in the Histoire Naturelle, and those that were the same did not always behave in the same way at different times or in different places. Azara sent a copy of the manuscript of the quad- rupedos to his brother, Jose Nicolas, then Spanish ambassador to Paris, who arranged for publication in that city after translation into French by M.-L.-E. Moreau de Saint-Mery. A year after his return to Spain in 1801, Azara secured publication in Ma- drid of the original Spanish manuscript with emendations and addition of 1 1 species, for a total of 77. Azara may not have been aware that as many as 62 of the 77 species he described were still un- known to science. His clear and precise charac- terization of each of the species or subspecies, however, provided contemporary cataloguers and systematists with the bases for the descriptions of 50 new species, many with their vernacular ap- pellations in the binomial. Actual specimens served as types for the remaining 1 2 species. The mammals described by Azara are listed be- low, with the scientific name of each given first followed by its local name(s). The page references are to Azara's works in French (Essais, 1801), Spanish (Apuntamientos, 1802), and the Voyage (1809). The last is a French translation in four volumes of Azara's travels in Paraguay with sep- arate atlas, but only the first volume and atlas contain information on mammals. Tapirus terrestris Linnaeus, 1758 Mborebi, Essais I, p. 1; Mborebi, Apunt., I, p. 1; Mborebi ou tapir. Voyage, p. 246. Tayassu G. Fischer, 1814 Coure ou Tayazou, Essais, I, p. 18; Cures o Tayaziis, Apunt., I, p. 14; Cure ou tayazii. Voyage, p. 248. Tayassu pecari albirostris Illiger, 1815 Tagnicati, Essais, I, pp. 2 1 , 25; Taiiicati, Apunt., p. 19; Tanicati, Voyage, p. 249. Bibliographic type of the subspecies. Tayassu tajacu Linnaeus, 1758 Taytetou, Essais, I, pp. 21,31; Taytetvi, Apunt., I, p. 23; Taytetu, Voyage, p. 249. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 59 CERVIDAE Gazou, Essais. I, p. 43; Venados, Apunt., I, p. 29; Guazu, Voyage, p. 250. Blastocenis dichotomus Illiger, 1815 Gouazoupoucou, Essais, I, p. 70; Guazu-pucu, Apunt., I, p. 33; guazu-pucu. Voyage, p. 250. Bibliographic type of the species. Blastoceros bezoarticus leucogaster Goldfuss, 1817 Gouazouti, Essais, I, p. 77; Giiazu-ti, Apunt., I, p. 41; Guazu-ti, Voyage, p. 251. Bibliographic type of the subspecies. Mazama americana gouazoupita Fischer, 1814 Gouazoupita, Essais, I, p. 82; GUazu-pita, Apunt., I, p. 5 1 ; Guazu-pita, Voyage, p. 252. Bibliographic type of the subspecies. Mazama gouazoubira gouazoubira Fischer, 1814 Gouazoubira, Essais, I, p. 86; Giiazu-bira, Apunt., I, p. 57; Guazu-bira, Voyage, p. 252. Bibliographic type of the species. DIDELPHIDAE Micoures, Essais, I, p. 240; Fecundos, Apunt., I, p. 204; Feconds, Voyage, p. 281. Didelphis albiventris Lund, 1 840 Micoure premier, ou micoure propement dit, Essais, I, p. 244; Micure, Apunt., I, p. 209; Micure, Voyage, p. 283. Caluromys lanatus Olfers, 1818 Micoure second, ou Micoure laineux, Essais, I, p. 175; Lanoso, Apunt., I, p. 221; Lanoso, Voyage, p. 287. Holotype in alcohol, no. 528, Museo de Cien- cias Naturales, Madrid, captured 22 July 1789, by Felix d'Azara (Cabrera, 1916, Bol. Real Soc. espanola Hist. Nat., 16, p. 1). Lutreolina crassicaudata Desmarest, 1 804 Micoure troisieme, ou micoure a queue grosse, Essais, I, p. 284; Coligrueso, Apunt., I, p. 229; Coligrueso, Voyage, p. 290. Bibliographic type of the species. Marmosa pusilla Desmarest, 1 804 Micoure quatrieme, ou micoure a queue longue, Essais, I, p. 290; Colilargo, Apunt., I, p. 251; Colilargo, Voyage, p. 291. Bibliographic type of Marmosa macrura Ol- fers, 1818 (= M. pusilla Desmarest). Micoure sixieme, ou micoure nain, Essais, I, p. 304; Enano, Apunt., I, p. 262; Enano, Voy- age, p. 284. Bibliographic type of Marmosa pusilla Des- marest, 1804. Monodelphis brevicaudis Olfers, 1818 Micoure cinquieme, ou micoure a queue courte, Essais, I, p. 295; Colicorto, Apunt., I, p. 258; Colicorto, Voyage, p. 293. Bibliographic type of the species. MYRMECOPHAGIDAE Hormigueros, Apunt., I, p. 61. Myrmecophaga tridactyla Linnaeus, 1758 Gnouroumi, ou Yoquoui, Essais, \, p. 89; Nu- rumi o Yoqui, Apunt., I, p. 66; Nurumi ou tamandua. Voyage, pp. 253, 255. Tamandua tetradactyla Linnaeus, 1758 (fig. 16) Cagouare, Essais, \, p. 103; Cagiiare, Apunt., \, p. 74; Cagiiare, Voyage, pp. 253, 256; Atlas, pi. VII (tamandua noir), pi. VIII (Cag- uouare). FELIDAE Gatos, Apunt., I, p. 85. Felis onca Linnaeus, 1758 Yagouarete, £'55a/5, 1, p. 1 14; Yaguarete,^;?Mm., I, p. 91; Yaguarete, Voyage, p. 258; Atlas, pi. IX. Yagiiarete negro, Apunt., I, p. 114; Yaguarete noir. Voyage, p. 267. Felis concolor Linnaeus, 1771 Gouazouara, Essais, I, p. 133; Giiazuara, Apunt., I, p. 120; Guazuara, Voyage, p. 268. Felis geoffroyi D'Orbigny and Gervais, 1 844 Mbaracaya, ^/7Mn/., I, p. 147; Baracaya, Voyage, p. 271. Note: Said not to exist in Paraguay. Felis species? Negro, Apunt., I, p. 154; Chat noir. Voyage, p. 273. Felis pardalis Linnaeus, 1758 Chibigouazou, Essais, I, p. 152; Chibi-giiazii, Apunt., I, p. 132; Chibi-guazu, Voyage, p. 269. Herpailurus yagouaroundi eyra Fischer, 1814 (fig. 16) Yagouaroundi, Essais, I, p. 171; Yaguarundi, Apunt., I, p. 156; Yaguarundi, Voyage, p. 273, Atlas, pi. X (Yagouarondi, black phase); Eyra, Essais, I, p. 177; Eyra, Apunt., I, p. 159; Eyra, Voyage, p. 274 (red phase). Bibliographic type of the subspecies. Felis colocolo pajeros Desmarest, 1816 Chat pampa, Essais, I, p. 1 79; Pajero, Apunt., I, p. 160; Pajero, Voyage, p. 274. Bibliographic type of the species. Note: Said not to exist in Paraguay. 60 FIELDIANA: ZOOLOGY .(■ ').i.'..M.ire of the species. Myotis albescens E. Geoffroy, 1 806 Chauve-souris douzieme, ou chauve-souris brun-obscur, Essais, II, p. 294; Pardo ob- scuro, Apunt., II, p. 309. Bibliographic type of the species. Johann Rudolph Rengger (1795-1832) Azara was followed by Johann Rudolph Reng- ger, a Swiss pharmacist and naturalist, who arrived in Paraguay in 1819 and devoted himself to the study of its mammals. His six-year study culmi- nated in the Naturgeschichte der Saeugethiere von Paraguay, published 1830. A total of 59 species was described, including four as new of which only Calomys callosus and Proechimys longicaudatus survived revisions. Azara distinguished 77 species, or 1 8 more, but several are not strictly Paraguayan. Among the Paraguayan forms missed by Rengger but recognized by Azara are the murine opossum (Marmosa), hairy armadillo (Chaetophractus), three-banded armadillo (Tolypeutes), skunk (Co- nepatus), tucotuco (Ctenomys), four cricetine ro- dents, and two bats. Well over 100 species are presently known from Paraguay. No doubt Azara set standards for the high qual- ity and accuracy of Rengger's descriptions and be- havioral accounts. The wealth of information in the Naturgeschichte has hardly been tapped by modem mammalogists. XI. Chile Giovanni Ignazio Molina (1737-1829) Knowledge of Chilean land mammals as a re- gional fauna begins with publication of the Saggio in 1782 by the Jesuit priest Don Giovanni (Juan) Ignazio Molina, who lived in Chile the first 30 years of his life. Expulsion of the Jesuits from the country obliged Molina to emigrate in 1768 and settle in his ancestral Italy. What Molina knew about Chilean mammals he learned before 1768; much of what he wrote about them thereafter suf- fered from a decayed memory. Molina was a naturalist in the broadest sense 64 HELDIANA: ZOOLOGY and was familiar with the Systemce of Linnaeus. He was not, however, particularly dedicated to any one branch of science, and his descriptions of the Chilean mammals are, for the most part, vague, inaccurate, and sometimes composite. A few of his subjects were fanciful, and none of the re- mainder were closely examined. Nevertheless, by dint of elimination and stretches of the imagina- tion, modem mammalogists have come to agree- ment on the application of most of the Linnaean names proposed by Molina for the likeliest species he may have had in mind. Thirty kinds of mammals were described in the Saggio. According to Osgood (1943, p. 15), five of them are unidentifiable, four (armadillos) are extraterritorial, two are but one and the same, and one is duplicated. The 14 still valid, with names dating from Molina, 1782, are Lutra felina, My- ocastor coypus, Conepatus chinga, Galictis cuja, Dusicyon culpaeus, Felis guigna, Felis colocolo, Felis concolor puma, Spalacopus cyanus, Octodon degus, Vizcacia vizcacia, Pudu puda, Vicugna vi- cugna, and Hippocamelus bisulcus. Remaining species, notably the larger mammals, recorded by Molina were well known to early voyagers, chron- iclers, and naturalists and had already received Linnaean names. first volume (1847) of eight on zoology contains virtually all Chilean mammals known at the time. Fifty-four species are described, with accounts of habits, habitat, and geographic distribution of each. For the most part. Gay worked from actual spec- imens brought to him by natives or observed by him on his travels throughout the country. On his return to France, Gay included in his studies the Chilean material preserved in the Paris Natural History Museum. The species recorded by Gay include Marsupi- alia, 2 (4% of the total); Chiroptera, 7 (13%); Gar- ni vora, 12 (22%); Pinnipedia, 6(11%); Rodentia, 23 (43%; myomorphs, 24%, caviomorphs, 18%); Artiodactyla, 3 (5%). Among the 30 species re- corded by Molina, only 3 or 10% are rodents. Of the 20 Chilean species collected by Darwin, 12 or 60% are rodents. In this volume Patterson and Feigl recognize 93 living Chilean sF)ecies, of which 53 or 57% are rodents (33% myomorphs, 24% caviomorphs), and 1 0 or 1 1 % are bats. XII. Peru Johann Jacob von Tschudi (1818-1889) Eduard Friedrich Poeppig (1798-1868) The German naturalist Eduard Poeppig is known for his Reise in Chile, Peru, and on the Rio Ama- zonas during the years 1827-1832. The account of his travels, in two volumes, was published 1835-1836. The Chilean mammals recorded in- clude seals, sea lions, and elephant seals, the degu, Spalacopus cyanus Molina (Psammomys nocti- vagus Poeppig, a synonym), the coypu, and a small canid, probably Dusicyon griseus Gray. In Antuco, Province of Bio Bio, he encountered the pudu, huemul, and two species of bats, one described as Nyticyus varius (= Lasiurus borealis bonariensis Lesson & Gamot, 1827), the other as Nycticyus macrotus (currently Histiotus macrotis Poeppig, 1835). Claudio Gay (1800-1873) Between the years 1 844 and 1871, Claudio Gay, French naturalist and longtime resident of Chile, produced 25 volumes, including two of plates, on the history, geography, and biota of Chile. The The Swiss biologist Johann von Tschudi was bom in the town of Glarus and studied the sciences at Swiss, French, and German universities. In- spired by the accounts of the travels of Humboldt and Darwin in South America, Tschudi sailed on 27 February 1838 from Le Havre for Peru. The first landing on the continent was made 5 June 1 838 on the Chilean island of Chiloe. After a delay of about three weeks and many observations of the natural history of the island, von Tschudi reembarked for Callao, Pern, with short stopovers in Valdivia and Juan Femandez. From August 1838 through most of 1843, von Tschudi traveled over much of Peru. Of particular interest to him were the higher vertebrates and the physical factors controlling their geographic dis- tribution. He distinguished faunal zones based on mling ecological features. The major zones were Pacific coast, Andean altitudinal zones of westem and eastem slopes, and the tropical Amazonian selva. Apparently, no one had preceded von Tschudi in the recognition of definable biogeo- graphic areas in the New World. The narrative of von Tschudi's travels in Pern was published in 1846 in German, followed in 1847 by Thomasina Ross's English translation. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 65 It li I ^1 si ll « 5 3 C U S Si, If J it ^6^ . 3 X) O w 66 HELDIANA: ZOOLOGY The scientific accounts of the mammals are found in a preliminary report (1844a) and first part of the Untersuchungen liber die Fauna Peruana, pub- lished later the same year ( 1 844b). Although von Tschudi attempted to provide the fullest account possible of Peruvian mammals, it appears he had little or no contact with the ma- jority of them. Most of his characterizations and life history accounts are taken from Humboldt, Spix, Wied-Neuwied, other European travelers and natives. Camelids, the dominant animals of the Peruvian landscape fascinated von Tschudi, and he wrote more about them than of other animals. His description of a vicuria hunt is quoted below from the Ross translation (Tschudi, 1 847, pp. 2 1 9- 220). The Indians seldom employ fire-arms in hunting the vicunas. They catch them by what they term the chacu. In this curious hunt, one man at least belonging to each family in the Puna villages takes a part, and women accompany the train, to officiate as cooks to the hunters. The whole company, frequently amounting to seventy or eighty individuals, proceeds to the Altos (the most secluded parts of the Puna), which are the haunts of the vicuiias. They take with them stakes, and a great quantity of rope and cord. A spacious open plain is selected, and the stakes are driven into the ground in a circle, at intervals of from twelve to fifteen feet apart, and are connected together by ropes fastened to them at the height of two or two and a half feet from the ground. The circular space within the stakes is about half a league in circumference, and an opening of about two hundred paces in width is left for en- trance. On the ropes by which the stakes are fastened together the women hang pieces of colored rags, which flutter about in the wind. The chacu being fully prepared, the men, some of whom are mounted on horseback, range about within a circuit of several miles, driving before them all the herds of vicurias they meet with, and forcing them into the chacu. When a sufficient number of vicunas is collected, the entrance is closed. The timid animals do not attempt to leap over the ropes, being frightened by the fluttering rags sus- pended from them, and, when thus secured, the Indians easily kill them by the tolas. These bolas consist of three balls, composed either of lead or stone; two of them heavy, and the third rather lighter. They are fas- tened to long, elastic strings, made of twisted sinews of the vicuria, and the opposite ends of the strings are all tied together. The Indian holds the lightest of the three balls in his hand, and swings the two others in a wide circle above his head; then taking his aim at the distance of about fifteen or twenty paces, he lets go the hand-ball, upon which all the three balls whirl in a circle, and twine round the object aimed at. The aim is usually taken at the hind legs of the animals, and the cords twisting round them they become firmly bound. It requires great skill and long prac- tice to throw the bolas dexterously, espe- cially when on horseback: a novice in the art incurs the risk of dangerously hurting either himself or his horse, by not giving the balls the proper swing, or by letting go the hand-ball too soon. The vicuiias, after being secured by the bolas, are killed, and the flesh is distributed in equal portions among the hunters. The skins belong to the Church. The price of a vicuna skin is four reals. When all the ani- mals are killed, the stakes, ropes, &c., are packed up carefully, and conveyed to another spot, some miles distant, where the chacu is again fixed up. The hunting is continued in this manner for the space of a week. The number of animals killed during that inter- val varies according to circumstances, being sometimes fifty or sixty, and at other times several hundred. During five days I took part in a chacu hunt in the Altos of Huayhuay, and in that space of time 122 vicurias were caught. With the money obtained by the sale of the skins a new altar was erected in the church of the district. The flesh of the vicuiia is more tender and better flavored than that of the llama. Fine cloth and hats are made of the wool. When taken young, the vicurias are easily tamed, and become very docile; but when old, they are intractable and ma- licious. At Tarma I possessed a large and very fine vicuria. It used to follow me like a dog whenever I went out, whether on foot or on horseback. The frequent hunting seems not to have the effect of diminishing the numbers of these animals. If in the vicinity of the villages where chacus are frequently established, they are less numerous than in other parts, it is because, to elude the pursuit of the hunters, HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 67 Table 9. Peruvian mammals according to Tschudi (1844a,b); current scientific names are used followed by Tschudi's synonym or misidentification, local names, and figure in this text; extralimital species are bracketed; arrangement of taxa follows Tschudi. Current wune Tschudi synonym or misidentification Local name Figure Aieles paniscus chamek Hum- boldt LagothrLx lagothricha poeppigi Schinz LagolhrLx flavicauda Humboldt Alotmtta seniculus Linnaeus [Mycetes rufimanus Kuhl] Cebus apella Linnaeus Cebus albifrons Humboldt [Cebus capucinus Linnaeus] Saimiri boliviensis peruviensis Hershkovitz Callicebus torquatus Hofimann- segg (subsp.?) [Callicebus personatus E. Geof- froyl Aotus nigriceps E)ollman [Chiropotes] Saguinus mystax mystax Spix Saguinus nigricollis Spix Saguinus fuscicollis Spix [Saguinus midas midas Lin- naeus] [Leontopithecus rosalia chryso- melas Kuhl] Chiroptera Phyllostomus elongatus E. Geof- fh)y Phyllostomus hastatus Pallas Phvlloslomus discolor Wagner, i843 Artibeus cinereus Gervais Stumira erythromos Tschudi Sturnira oporophilum Tschudi Glossophaga soricina Pallas Anoura geoffroyi peruana Tschudi Eptesicus innoxius Gervais Histiotus macrotus Poeppig Noctilio leporinus Linnaeus Noctilio albiventris Desmarest Tadarida brasiliensis I. Geoffroy Xfolossus molossus Pallas? Eumops auripendulus Shaw Molossus ater E. Geofiroy [Promops nasutus Spix] Carnivora Tremarctos omatus F. Cuvicr Nasua nasua montana Tschudi Potosflavus Schreber Eira barbara Linnaeus Ateles marginatus; Aieles ater; Ateles pentadactylus Lagothrix humboldti; Lagothrix canus Mycetes flavicaudatus (sic) Mycetes stramineus Alouatta belzebul Cebus robustus Chrysothrix sciureus Callithrix amictus Callithrix personatus Nyctipithecus trivirgatus IPithecia* [Midas labiatus] [Midas labiatus] [Midas labiatus] [Midas rufimanus] [Midas chrysomelas] Chuva; maquisapa; chamek; mahmonda; machucusillo; supaya Mono oki; choko Coro [= coto?] Macaquito Tocon Hatmnmasu Phyllostomus innominatum Tschudi Phyllostomus (Artibeus) pusil- lum Phyllostomus (Sturnira) oporo- philum Tschudi Glossophaga amplexicauda Glossophaga (Choeronycteris) peruana Tschudi Vespertilio innoxius Vespertilio ( Vesperugo) velatus Noctilio unicolor Noctilio affinis Molossus (Dysopes) naso Molossus (Dysopes) velox Molossus (Dysopes) ferox; Dy- sopes longimanus Molossus (Dysopes) myosuros Tschudi; Molossus anonymus Tschudi Dysopes fitmarius Ursus fivgilegus Tschudi Nasua socially, Nasua solitaria, Nasua leucorhynchos Tschudi Cercoleptes caudivolvidus Galictis barbara Hucamari Achuna, mishash Cushumbi Omeyro 17 68 HELDIANA: ZOOLCXJY Table 9. Continued. Current name Tschudi synonym or misidentification L4>cal name Figure Carnivora {continued) Mustela frenata agilis Tschudi Conepatus chinga Molina Lutra felina Molina Lutra montana Tschudif Dusicyon thous Linnaeus Felis concolor Linnaeus Felis onca Linnaeus Felis pardalis Linnaeus Feiis wiedii Schinz Felis yagouaroundi E. Geoffroy PlNNIPEDIA Otaria flavescens Shaw Marsupialia Didelphis marsupialis Linnaeus Metachirus nudicaudatus E. Geoffroy Philander opossum Linnaeus Marmosa noctivaga Tschudi Marmosa impavida Tschudi Marmosa murina Linnaeus Caluromys lanatus ornatus Tschudi RODENTIA Sciurus aestuans Linnaeus Sciurus pyrrhinus Thomas Sciurus stramineus Eydoux and Souleyet Sciurus spadiceus tricolor Tschudi Proechimys sp.? Chinchilla brevicaudata Water- house Lagidium peruanum Meyen Lagidium viscacia Molina^ [Octodon degus Molina] [Myocastor coypus Molina] Coendou bicolor Tschudi Dasyprocta leporina Linnaeus Dasyprocta variegata Tschudi Akodon boliviensis Meyen Phyllotis darwini Waterhouse Oryzomys longicaudatus de- structor Tschudi Oryzomys melanostoma Tschu- di Rhipidomys leucodactylus Tschudi Agouti paca Linnaeus Hydrochaeris hydrochaeris Lin- naeus Cavia porcellus Linnaeus Molina (Thiosmus) mapurita; Mephitis furcata; Mephitis amazonica Lutra chilensis Canis azarae Felis onza Felis macrura (sic = Felis ma- croura); Felis celidogaster Felis yaguaruruii Otaria jubata; Otaria ulloae Tschudi; Otaria aurita Hum- boldt (in Tschudi) Didelphys azarae Didelphys myosuros Poma, leon Choque china, yana cheque, tigre Uturunco Mucamuca, jarachupa 17 \Sciurus variabilis] [Echinomys leptosoma] Eriomys chinchilla Lagidium peruvianum {sic) Lagidium pallipes [Octodon cummingii] [Myopotomus coypus] Sphingurus {sic) bicolor Dasyprocta aguti Linnaeus Acodon boliviense Hesperomys darwini Hesperomys destructor Hesperomys melanostoma Hesperomys {Rhipidomys) leu- codactylus Coelogenys fulvus Hydrochoerus capybara Cavia cutleri Cutspi or cushpi 17 Cuy del monte HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 69 Table 9. Continued. Current name Tschudi synonym or misidentification Local name Figure Lagomorpha Sylvilagus brasiliensis Linnaeus Edentata Bradypus vahegatus Schinz [Bradypus torquatus lUiger] Dasypus novemcinctus Linnaeus Cabassous unicinctus Linnaeus Tamandua tetradactyla Lin- naeus Cyclopes didactylus Linnaeus Perissodactyla Tapirus terrestris Linnaeus Tapirus pinchaque Roulin Artiodactyla Tayassu tajacu Linnaeus Tayassu pecari Link Lama glama Linnaeus Lama pacos Linnaeus Lama guanicoe Miiller Vicugna vicugna Molina Mazama americana Erxleben Mazama gouazoubira peruana Tschudi Hippocamelus antisensis d'Or- bigny Lepus brasiliensis Bradypus infuscatus Dasypus 9-cinctus {sic) Dasypus tatuay {sic = tatouay) Myrmecophaga tamandua Myrmecophaga didactyla Quirquincho Tapirus americanus Tapirus villosus Dicotyles torquatus Dicotyles labiatus Auchenia lama Llama Auchenia paco Alpaca Auchenia huanaco Auchenia vicuna Vicuna Cervus rufus Cervus nemorivagus var. per- Liucho, venado uana Cervus antisiensis Tarush, taruga 17 * Sakis {Pithecia) evidently not seen by von Tschudi. His descriptions are of bearded sakis {Chiropotes) after Humboldt (1811), which do not occur in Peru, t May not be an otter, according to Thomas (1908, p. 393). % The species was known to occur in parts of formerly southwestern Peru now in Chile. they seek refuge in the Altos, where they are found in vast numbers. Several modem travelers have lamented the diminution of the vicunas, but without reason. In fonner times those animals were hunted more ac- tively than at present. Von Tschudi's journeys in the puna inspired him to poetic descriptions of the habits, particu- larly the visual propensities, of its denizens. Herds of vicuiias approached me with cu- rious gaze, and then on a sudden fled with the swiftness of the wind. In the distance I observed stately groups of huanacos turning cautiously to look at me, and then passing on. The Puna stag (tarush) slowly advanced from his lair in the mountain recesses, and fixed on me his large, black, wondering eyes, whilst the nimble rock rabbits (viscachas) playfully disported and nibbled the scanty herbage growing in the mountain crevices. (Tschudi, Ross translation, 1 847, p. 249) On descending the eastern slope of the Cordi- lleras to the subtropical zone inhabited by a greater variety of different kinds of mammals, von Tschu- di (Ross translation, 1847, p. 275) romanticized that: . . . the swift-footed roe [Mazama sp.] of the Cordillera roams here and dwells in the thickets, avoiding the warm forest. The dark brown coati {Nasua montana, Tsch.) howls and digs at the root of trees in search of food, the shy opossum crawls fearfully under the foliage; the lazy armadillo creeps into his hole, but the ounce [Felis onca] and the lion [Felis concolor] seldom stray hither to con- test with the black bear {Ursus frugilegus Tsch.) the possession of his territory. The 70 FIELDIANA: ZOOLOGY little hairy tapir (Tapirus villosus, Wagn.) ventures only at twilight out of his close am- bush to forage in the long grass. The systematic arrangement in the Untersu- chungen is said to include all mammals known at the time to occur in Peru. By von Tschudi's count, the fauna consists of 1 1 9 species in 48 genera. These totals include domestic animals, the intro- duced house mouse, some duplicated names of native species, and a number of others not known to occur in Peru. In terms of currently recognized species found in Peru, von Tschudi's combined lists (1844a, pp. 244-255; 1844b, pp. 6-20; 21- 264) consist of 87 species in 58 genera. The species are listed in Table 9 with von Tschudi's synonyms or misidentifications. Author attributions of the synonyms are omitted unless they are to von Tschudi himself Vernacular names, if given, are included. Extralimital species are shown in brack- ets. In the case of unrevised groups or where two or more subspecies occur in Peru without possi- bility of determining which were described by von Tschudi, only the specific names are given. XIII. Patagonia Alcide Charles Victor d'Orbigny (1802-1857) The French-bom Alcide d'Orbigny was educat- ed by his country's leading naturalists. His apti- tudes were recognized by authorities of the Mu- seum National d'Histoire Naturelle, and with that institution's financial and material assistance, he sailed for South America charged with making a scientific survey of the southern half of the con- tinent. Circumstances restricted his studies and collections of mammals almost entirely to Argen- tina and Bolivia. D'Orbigny left France 3 1 July 1 826 and arrived in Rio de Janeiro 24 September 1826 on his way to Montevideo where he landed on 29 September. The natural history of the region between Mal- donado east of Montevideo and Buenos Aires en- gaged his attention for several months. On 14 February 1827, d'Orbigny ascended the Rio Parana and arrived 1 5 March at the important fluvial port of Corrientes, capital of the province of the same name. With the town as base, d'Or- bigny explored the province throughout much of one year. On his return to Buenos Aires in April 1828, he made stops in Entre Rios and Santa Fe. Beginning June 1828 and continuing through 1829, his at- tentions were devoted to faunal studies in the provinces of Buenos Aires and Rio Negro. The chronology of the early part of 1829, as given by d'Orbigny (1835-1847) in the Voyage, confuses time spent in the two provinces with that spent in Corrientes. In any event, d'Orbigny was clearly in Buenos Aires and Rio Negro during the last half of 1 829. He returned to Montevideo in December 1 829 and on 29 December sailed on to Patagonia and Chile. Cape Horn was rounded on 19 January 1830 and Valparaiso, Chile, was reached 16 February. Because of the political unrest in the country, d'Or- bigny sailed to the then Bolivian port of Cobija, where he landed on 8 April; 20 April found him in Arica and Tacna, both ports then in Peru's pos- session. After some investigation of the coast, d'Orbigny left Tacna on 19 May for La Paz, the mountain capital of Bolivia, arriving there 28 or 29 May. For the next three years, d'Orbigny explored, mapped, and sampled the natural resources of the country. He crisscrossed Bolivia from La Paz east to the Paraguayan border and from Potosi in the south to the lower Rio Mamore in the north. D'Or- bigny's actual itinerary is almost impossible to track because of the inaccuracies of the then avail- able maps. Modem maps aided Pilleri and Arvy (1977) in their reconstmction of the itinerary in chronological sequence (fig. 1 8). A complete account of d'Orbigny's South Amer- ican joumey with observations on and descrip- tions of the geology, paleontology, living plants, animals, and Indians is contained in seven huge volumes published serially from 1 835 through 1847 in Paris under the title Voyage dans I'Amehque Meridionale. A full report on the mammals was reserved for the last, or perhaps a separate pub- lication, but a turn in d'Orbigny's fortunes inter- rupted the work. A number of colored plates of mammals believed new to science and a few short articles on others had already been published. So that all would not be lost, a synoptic systematic report on the mammals collected was published in 1 847 jointly with the distinguished mammal- ogist Paul Gervais, as number 2 of volume 4 of the Voyage. Brief notes on distribution and be- havior accompany the abbreviated descriptions of each species. The species are listed in Table 10 with abstracted locality data. Scientific names used are current with synonyms and misidentifications added. The specimens are deposited in the Mu- seum National d'Histoire Naturelle in Paris. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 71 5 : ' '■ i =. I ? I •? :^ ^ < Is I § I 1 1 5 , 1 , ; = £ II ^ I I 1 1 s I §• z X i -i £ 3. ^ ^ > f- }. i. i^ EccSccccc'.cc :^ ;? j: c ?. ■?. ■?. f. ■?. a i ■>. I •e o CO 3 P I 4f 72 FIELDIANA: ZOOLOGY CAI.ITKIN M».pkn" '*• r M.I.ITIIKIN J..— i^M.. I. . ■■! .'■■■!■ Fig. 1 9. Animals of the d'Orbigny Bolivian Expedition: upper left, Callithrix entomophagus d'Orbigny (= Saimiri boliviensis boliviensis I. Geoffroy and Blainville); upper right, Callithrix donacophilus d'Orbigny (= Callicebus do- nacophilus donacophilus); lower left, Felis geoffroyi d'Orbigny and Gervais (= Felis colocolo geoffroyi); lower right. Mephitis humboldtii (= Conepatus chinga suffocans Illiger); from d'Orbigny and Gervais (1847). HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 73 Table 10. Mammals of the southern half of South America, mostly Bolivia and Argentina, recorded by d'Orbigny and Gervais (1847); the arrangement is phylogenetic. Current name d'Orbigny and Gervais synonym Locality Figure Chiroptera Noctilio albiventris Desmarest, 1818 Noctilio leporinus rufipes d'Or- bigny, 1835 Tonatia sylvicola d'Orbigny, 1835 Artibeus planirostris Spix, 1 823 Desmodus rotundus E. Geoffroy, 1810 Myotis nigricans Wied-Neu- wied, 1821 Eptesiciis furinalis d'Orbigny and Gervais, 1 847 Myotis albescens E. Geoffroy, 1806 Myotis ruber E. Geoffroy, 1 806 Histiotus velatus I. Geoffroy, 1824 Tadarida brasiliensis I. Geof- froy, 1824 Molossus crassicaudatus E. Geof- froy, 1805 Primates Saimiri boliviensis boliviensis I. Geoffroy and Blainville, 1 834 Callicebus donacophilus donaco- philus d^Orhigny, 1835 Alouatta seniculus sara Elliot, 1910 Cebi4s apella paraguayanus Fischer, 1829 Carnivora Dusicyon gymnocercus Fischer, 1814 Chrysocyon brachyurus Illiger, 1815 Tremarctos ornatus F. Cuvier, 1825 Procyon cancrivorus nigripes Mivart, 1886 Nasua nasua solitaria Wied- Neuwied, 1821 PotosflavusSchxc\xT, 111 A Lyncodon patagonicus Blain- ville, 1842 Galictis cujafurax Thomas, 1907 Conepatus chinga suffocans Illi- ger, 1815 Lutra platensis Waterhouse, 1838 Noctilio affinis d'Orbigny, 1835 BOLIVIA: Moxos Province BOLIVIA: Chiquitos and Moxos provinces BOLIVIA: Yuracare territory, base of eastern Cordillera BOLIVIA: Chiquitos Province Not Vespertilio perspicillatus Linnaeus, 1758 Desmodus rufus Wied-Neuwied, 1824; Edostoma cinerea d'Or- bigny, 1835 Vespertilio hypothrix d'Orbigny and Gervais, 1847 Vespertilio isidori d'Orbigny and Gervais, 1847 Molossus rugosus d'Orbigny, 1835, not Molossus nasutus Spix, 1823 Molossus moxensis d'Orbigny, 1835; Molossus velox Tem- minck, 1827 Calithrix (sic) entomophagus d'Orbigny, 1835 Not Stentor stramineus E. Geof- froy Cebus fulvus var. Not Canis cancrivorus Desma- rest, 1820 Canis jubatids Desmarest, 1820 Not Procyon cancrivorus Cuvier, 1798 Nasua fusca Desmarest, part Cercoleptes caudivolvulus Schre- ber, 1774 Not Mustela brasiliensis Gme- lin, 1788 Mephitis castaneus d'Orbigny and Gervais, 1 847, not Me- phitis humboldtii Gray, 1837 BOLIVIA: Chiquitos BOLIVIA: Moxos ARGENTINA: Corrientes ARGENTINA: Corrientes ARGENTINA: Corrientes BOLIVIA: Chuquisaca ARGENTINA: Corrientes BOLIVIA: Moxos and Chiqui- tos provinces BOLIVIA: Chiquitos; Moxos; Santa Cruz BOLIVIA: Moxos Province BOLIVIA: Santa Cruz; Chiqui- tos; Moxos BOLIVIA: near Santa Cruz de la Sierra BOLIVIA: Chiquitos Tropical South America to 41°S BOLIVIA: Cochabamba; Chu- quisaca BOLIVIA: Chiquitos; ARGEN- TINA: Corrientes BOLIVIA: tropics to 30'^ BOLIVIA: foot of eastern Cordi- llera ARGENTINA: Rio Negro 19 19 19 ARGENTINA: Rio Parana in Provinces Buenos Aires and Corrientes 74 FIELDIANA: ZOOLOGY Table 10. Continued. Current name d'Orbigny and Gervais synonym Locality Figure Carnivora Felis colocolo pajeros Desma- rest, 1816 Felis geoffroyi d'Orbigny and Gervais, 1847 Felis concolor Linnaeus, 1771 Felis onca Linnaeus, 1758 PiNNIPEDIA OtariaflavescensShs^N, 1800 Arctocephalus australis Zimmer- mann, 1782 Mirounga leonina Linnaeus, 1758 Artiodactyla Mazama gouazoubira Fischer, 1814 Blastoceros bezoarticus Lin- naeus, 1758 Hippocamelus antisensis d'Or- bigny, 1834 Blastocerus dichotomus Illiger, 1815 RODENTIA Sciurus spadiceus Olfers, 1818 Eligmodontia typus F. Cuvier, 1837 Octodon degus Molina, 1782 Octodontomys gliroides, Gervais and d'Orbigny, 1 844 Ctenomys boliviensis Water- house, 1848 Ctenomys magellanicus Bennett, 1835 Microcavia australis Gervais and d'Orbigny, 1833 Galea flavidens Brandt, 1835 Dolichotis patagonum Zimmer- man, 1780 Dasyprocta azarae Lichtenstein, 1827 Cetacea Inia boliviensis d'Orbigny, 1834 [Pontoporia blainvillei Gervais and d'Orbigny, 1 844; not part of d'Orbigny collection] Lagenorhynchus cruciger Quoy and Gaimard, 1 824 Lissodelphis peroni Lacepede, 1804 Otaria jubata Schreher, 1776 Otaria porcina Molina, 1782 Phoca proboscidea Peron, 1817 Cervus simplicicornis Illiger, 1815 Not Cervus campestris F. Cu- vier, 1817 Cervus paludosus Desmarest, 1822 Not Sciurus igniventris Wagner, 1842 Not Ctenomys brasiliensis Blainville, 1826 Dasyprocta patachonica Desma- rest, 1820 Not Dasyprocta nigricans Wag- ner, 1842 ARGENTINA: from 35°-45'S ARGENTINA: Pampas to 44'S 19 BOLIVIA; ARGENTINA: to Straits of Magellan Tropical South America not be- yond 40^; ARGENTINA: Pampas; Serrania de Tandil ARGENTINA: S mouth Rio Negro ARGENTINA: coast; PERU: coast ARGENTINA: Rio Negro, near mouth Tropical South America to 28°S Lowland savannas to northern Patagonia BOLIVIA: La Paz; Cochabam- 20 ba; Chuquisaca; rarely below 3500 m ARGENTINA: Corrientes; BO- LIVIA: Chiquitos BOLIVIA: Chiquitos ARGENTINA: Corrientes CHILE: Santiago de Chile BOLIVIA: La Paz ARGENTINA: Corrientes; BO- LIVIA: Santa Cruz de la Sie- rra ARGENTINA: northern Pata- gonia ARGENTINA: Rio Negro BOLIVIA: Cochabamba; Chu- quisaca; La Paz ARGENTINA: northern Pata- gonia; Corrientes Tropical South America BOLIVIA: rivers of Moxos and 20 Chiquitos URUGUAY: Montevideo Atlantic Ocean (57»-76'«, E and S of Cape Horn) Atlantic Ocean (48°-64'«); At- lantic-Pacific Oceans around Cape Horn HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 75 ■' S, : i ^ 'b 1 1 \Vi «J M S V ' i ,^ \ I ^ ja ■•»■ ^ •c f- is ■* \ \ ■_ 1 tj J "3 11 ^ u \ i •1 ■■i I « 'm ■/. 1U "b t^ 1 1 ■Si <« ;§« ^ t "^ e SO 11 X S c^. •2 12 S 1 pa CO .!>•§ O > e s ected, soon found it. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 77 lOV Coplapo'U a Coqulabo pLl- Coocepcirfn J9 \ l Buenos Aires ^-xJUjtontevldeo Y (SIERRA DF, ^fo Plata CHARLES DARUIN TOTAGE OF B.M.S. BEACLE SOUTH AMERICAN LOCALITIES (1832-1835) CBOHOS*- AXCRIPELACO^^ ,' -■w T. Good Success (■snia dc Good Uickf 80* Hp-y^^ ^/San Julian FALKLAND ISLANDS tllUA DEL ruioo Canai fagla J 1 L. 60* 50* 30* Fig. 2 ! . Map showing principal South American stations visited by Charles Darwin ( 1 832-1 835) on world cruise of H.M.S. Beagle (1832-1836). 78 HELDIANA: ZOOLOGY u HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 79 Table 1 1 . Mammals collected or observed by Oanvin in the Maldonado Region, Uruguay and parts of Argentina, and those recorded by Waterhouse (1838-1839); the arrangement is phylogenetic. Current name Waterhouse synonym or misidentification Locality Figure Marsupialia Didelphis albiventris Lund Lutreolina crassicaudata Des- marest Monodelphis dimidiata Wagner Chiroptera Tadarida brasiliensis I. Geoffroy Edentata Dasypus hybridus Desmarest Zaedyus pichiy Desmarest Chaetophractus villosus Desma- rest Tolypeutes matacus Desmarest Carnivora Dusicyon gymnocercus Fischer Felis colocolo pajeros Desmarest Galictus cujafurax Thomas Lutra platensis Waterhouse Conepatus chinga gibsoni Thomas Felis concolor acrocodia Gold- man Felis onca palustris Ameghino Artiodactyia Blastoceros bezoartiats Lin- naeus Lama guanicoe Muller RODENTIA Myomorpiia Oryzomys flavescens Waterhouse CaJomys laucha Olfers Eligmodontia typus Cuvier Holochilus brasiliensis darwini Thomas Reithrodon physodes typicus Waterhouse Akodon azarae Fischer Akodon colibreve Brants Scapteromys tumidus Water- house Oxymycierus rufus nasutus Waterhouse Caviomorpiia Cavia porcellus Linnaeus Hydrochaeris hydrochaeris Lin- naeus Didelphis azarae AucL Didelphis brachyura Auct. Not Dysopes nasutus Spix Dasypus minutus AucL Not Canis azarae Wied-Neu- wied Not Galictis vittata Schreber Not Cervus campestris Cuvier Mus bimaculatus Waterhouse; Mus gracilipes Waterhouse Mus elegans Waterhouse Mus arenicola Waterhouse Mus obscurus Waterhouse Cavia cobaia Auct. Hydrochoerus capybara Auct. URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado ARGENTINA: Banda Oriental, Entre Rios ARGENTINA: Bahia Blanca (observed) ARGENTINA: Bahia Blanca (observed) ARGENTINA: La Plata (ob- served) ARGENTINA: Bahia Blanca URUGUAY: Maldonado URUGUAY: Maldonado ARGENTINA: Bahia Blanca (observed) ARGENTINA: the pampas (ob- served) ARGENTINA: in the Rio Pa- rana (observed) URUGUAY: Maldonado; AR- GENTINA: Bahia Blanca; Rio Negro ARGENTINA: Rio Negro (ob- served) URUGUAY: Maldonado ARGENTINA: Bahia Blanca ARGENTINA: Bahia Blanca ARGENTINA: Bahia Blanca URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado URUGUAY: Maldonado 22 80 HELDIANA: ZOOLOGY Table 1 1 . Continued. Current name Waterhouse synonym or misidentification Locality Figure Caviomorpha (continued) Dolichotis patagonum Zimmer- man Vizcacia maximus Desmarest Ctenomys brasiliensis Blainville Lagostomus trichodactylus Brookes ARGENTINA: Rio Negro (ob- served) URUGUAY: Maldonado URUGUAY: Maldonado Darwin could not explain the viscacha's behavior. Pampas deer (Blastoceros bezoarticus) were abundant throughout the La Plata region. Darwin (1839, p. 55) saw very many small herds, containing from five to seven animals each, near the Sierra Ven- tana, and among the hills north of Maldo- nado. If a person crawling close along the ground, slowly advances toward a herd, the deer frequently, out of curiosity, approach to reconnoitre him. I have by this means killed, from one spot, three out of the same herd. Although so tame and inquisitive, yet when approached on horseback, they are ex- ceedingly wary. In this country nobody goes on foot, and the deer knows man as its en- emy only when he is mounted and armed with the bolas. The jaguar by some accounts is a man-killer, by others, fears man. Darwin (1839, p. 159) records several instances reported to him of man-killing jaguars of the Rio Parana region. The Beagle left the Rio Plata on December 1 833 for Puerto Deseado, or Port Desire, on the Pata- gonian coast. The mammals collected by Darwin and reported by Waterhouse (1838-1839), with descriptions and supplementary notes by Darwin, are listed in Table 1 1, with the Waterhouse syn- onyms (misidentifications included). Added are the few species Darwin mentioned in his Journal but did not collect. Unless otherwise indicated, all species are from the neighborhood of Maldonado, Uruguay. The geology and natural history of Patagonia investigated by Darwin included those of the Straits of Magellan and Tierra del Fuego (December 1 832- January 1833; May-June 1834), Puerto Deseado (Port Desire) (December 1833-January 1834), Santa Cruz (April-May 1834), and the Falkland Islands (March 1834). The Beagle itself (fig. 22) sailed up the Rio Santa Cruz to a point 1 40 miles from its mouth in the Atlantic Ocean to about 60 miles from the nearest arm of the Pacific Ocean on the opposite side of the cordillera. Darwin was greatly impressed by the number, variety, and great size of fossil mammals, mostly Pleistocene, exposed on the Patagonian plains. These, he (1839, p. 209) believed, were confir- mation of the "law" that existing animals in an area have a close relation in form with extinct species in the same area. The natural causes for extinction, however, eluded Darwin. After pro- posing and rejecting a number of explanations, the nonevolutionist Darwin (1839, p. 212) concluded that the whole series of animals, which have been created with f>eculiar kinds of organization, are confined to certain areas; and we can hardly suppose these structures are only ad- aptations to peculiarities of climate or coun- try; for otherwise, animals belonging to a distinct type, and introduced by man, would not succeed so admirably even to the exter- mination of the aborigines. On such grounds it does not seem a necessary conclusion that the extinction of species, more than their creation, should exclusively depend on the nature (altered by physical change) of their country. All that at present can be said with certainty, is that, as with the individual, so with the species, the hour of life has run its course, and is spent. The small number of extant large mammals and great number and variety of small mammals, also impressed Darwin (1839, p. 215). Patagonia, poor as she is in some resj)ects, can, however, boast of a greater stock of small rodents than p)erhaps, any other coun- try in the world. Several sjjecies of mice are externally characterized by large thin ears and a very fine fur. These little animals HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 81 swarm amongst the thickets in the valleys, where they cannot for months together taste a drop of water. They all seem to be can- nibals, for no sooner was a mouse caught in one of my traps than it was devoured by others. A small and delicately-shaped fox which is likewise very abundant, probably derives its entire support from these small animals. The guanaco was regarded as the characteristic quadruped of the Patagonian plains (Darwin, 1 839, p. 215). Herds of fifty or a hundred were common, and, as I have said, we saw one which must have contained at least five hundred. The puma with the condor in its train, follows and preys upon these animals. The footsteps of the former were to be seen almost every- where on the banks of the river [Santa Cruz]; and the remains of several guanaco, with their necks dislocated, and bones broken, showed how they had met their death. In March 1834, Darwin visited the Falkland Islands. On a tour he encountered large numbers of horses, cattle, swine, and rabbits {Oryctolagus cuniculus Linnaeus [= Lepus magellanicus Lesson and Gamot] in domestic and feral states. The an- imals had been brought by French colonists in 1 764. Darwin wrote (p. 249), The only quadruped native to the island, is a large wolf-like fox [Dusicyon (Iculpaeus) australis Kerr] which is common to both East and West Falkland. I have no doubt it is a peculiar species and confined to this archijjelago. . . . These wolves are well known . . . [for] their lameness and curiosity. ... To this day their manners remain the same. . . . As far as I am aware, there is no other in- stance in any part of the world, of so small a mass of broken land, distant from a con- tinent, possessing so large a quadruped pe- culiar to itself. Their numbers have rapidly decreased; they are already banished from that half of the island which lies to the east- ward of the neck of land between St. Sal- vador Bay and Berkeley Sound. Within a very few years after these islands shall have become regularly settled, in all probability this fox will be classed with the dodo, as an animal which has perished from the face of the earth. The mammals collected by Darwin in the Ar- gentine Patagonia (including Falkland Islands) and bordering parts of the Chilean Straits of Magellan are listed in Table 12. The Chilean leg of the cruise began in May 1 834 with the passage of the Beagle into the eastern mouth of the Straits of Magellan and ended July 1835 with departure from Copiapo in northern Chile. While the Beagle sailed up and down the Chilean coast, Darwin explored the coast, islands, archipelagos, and cordillera. He crossed the Andes on 21 March 1835 through the Portillo Pass south of Santiago, and proceeded to the town of Men- doza in Argentina. Differences observed between the biota of eastern and western versants of the cordillera impressed Darwin. The mountains, he (1839, p. 399) reasoned, have existed as a great barrier, since a period so remote that whole races of animals must subsequently have perished from the face of the earth. Therefore, unless we suppose the same species to have been created in two different countries, we ought not to expect any closer similarity between the organic beings on opposite sides of the Andes, than on shores separated by a broad strait of the sea. The correlation between geographic isolation and faunal peculiarity was noted in other circum- stances. Darwin (1839, p. 439) observed that next to lizards, mice appear to be able to support existence on the smallest and driest portions of the earth— even on islets in the midst of great oceans. I believe it will be found, that several islands, which possess no other warm-blooded quadruped, have small rodents peculiar to themselves. Ratadas or rat plagues in Chile also caught Dar- win's attention. One of the earliest recorded for Oryzomys longicaudatus longicaudatus, viewed through the eyes of Darwin (in Waterhouse, 1 838, p. 40), "overran the wooded country south of Con- cepcion, in swarms of infinite numbers." The mammals of Tierra del Fuego tallied by Darwin (1839, p. 300) included, besides cetaceans and phocids, one bat [not named but likely Histiotus montanus magellanicus Philippi], a mouse with grooved front teeth {Reithrodon of Waterhouse) and two other species, the tu- 82 HELDIANA: ZOOLOGY Table 12. Patagonian mammals collected by Darwin and recorded by Waterhouse (1838-1839); arrangement is phylogenetic. Current name Waterhouse synonym or misidentification Locality Figure Carnivora Dusicyon australis Kerr Dusicyon griseus Gray Felis colocolo pajeros Desmarest Artiodactyla Lama guanicoe Miiller RODENTIA Oryzomys longicaudatus magel- lanicus Bennett Akodon xanthorhinus Water- house Akodon canescens Waterhouse Auliscomys micropus Water- house Graomys griseojlavus Water- house Phyllotis xanthopygus Water- house Reithrodon physodes cunicu- loides Waterhouse Euneomys chinchilloides Water- house Myocastor coypus Molina Microcavia australis I. Geoffroy and d'Orbigny Dolichotis patagonum Zimmer- man Cetacea Lagenorhynchus cruciger Quoy and Gaimard Canis antarcticus Shaw Not Canis azarae Wied-Neu- wied Auchenia llama Desmarest Cavia patachonica Shaw Delphinus fitzroyi Waterhouse Falkland Islands ARGENTINA: PaUgonia ARGENTINA: Santa Cruz ARGENTINA: Patagonia ARGENTINA: Puerto de Hambre (Port Famine); CHILE: Straits of Magellan CHILE: Peninsula de Hardy ARGENTINA: Puerto Deseado (Port Desire); Santa Cruz ARGENTINA: Rio Santa Cruz, Santa Cruz ARGENTINA: Rio Negro ARGENTINA: Puerto Deseado (Port Desire); Santa Cruz ARGENTINA: Puerto Deseado (Port Desire); San Julian; Rio Santa Cruz, Santa Cruz ARGENTINA: Eastern entrance to Straits of Magellan ARGENTINA: Rio Chubut ARGENTINA: 41°S to Straits of Magellan ARGENTINA: Patagonia ARGENTINA: Golfo San Jose, 42°30'S, Chubut 22 cotuco (the greater number of these rodents are confined to the eastern and dry part), a fox, sea-otter, guanaco, and one deer [un- named but likely Hippocamelus bisulcus]. The latter animal is rare, and is not, I be- lieve, to be found south of the Straits of Magellan, as happens with the others. With respect to geographic distribution, Darwin (1839, p. 300), observing the general correspondence of the cliffs of soft sandstone, mud, and shingle, on the opposite side of the Strait, together with those on some intervening islands [was] strongly tempted to believe that the land was once joined and thus allowed animals so del- icate and helpless as the tucotuco, and Reithrodon to pass over. The tucotuco in question is Ctenomys magellan- icusfueginus Philippi (Osgood, 1943, p. 1 19). Dar- win (1839, p. 327) also mentioned the occurrence of the puma {Felis concolor) in Tierra del Fuego, and related something of its habits in other parts of Chile and Argentina. The type specimen of Darwin's zorro {Dusicyon fulvipes Martin), peculiar to the island of Chiloe, was discovered by Darwin (p. 34 1 ) on 6 December 1834 sitting on the rocks and so intently absorbed in watching the maneuvers of two ship's officers engaged in surveying, that I was able, by quietly walking up be- hind, to knock him on the head with my geological hammer. This fox, more curious or more scientific, but less wise, than the generality of his brethren, is now mounted in the museum of the Zoological Society. Sea otters {Lutrafelina Molina) were described by Darwin (in Waterhouse, 1838, p. 24) as ex- HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 83 ceedingly common amongst the innumerable channels and bays which form the Chonos Ar- chipelago. . . . they may generally be seen quietly swim- ming with their heads just out of water amidst the great entangled beds of kelp, which abounds on this coast. They burrow in the ground, within the forest, just above the rocky shore, and I was told, that they some- times roam about through the woods. This otter does not, by any means, live exclu- sively on fish. One was shot whilst running to its hole with a large volute-shell in its mouth; another (I believe the same species) was seen in Tierra del Fuego devouring a cuttle fish. But in the Chonos Archipelago perhaps the chief food of this animal, as well as that of the immense herds of great seals, and flocks of terns and cormorants, is a red- coloured crab (belonging to the family Mar- rouri) of the size of a prawn, which swims near the surface in such dense bodies, that the water appears of a red colour. This spec- imen weighed nine pounds and a half The vampire bat which Darwin (1839, p. 25) recognized as a species of d'Orbigny's genus Edos- toma (= Des modus) was singled out as often the cause of much trouble, by biting the horses on their withers. The injury is generally not so much owing to the loss of blood, as to the inflammation which the pressure of the saddle afterwards produces. The whole circumstance has lately been doubted in England; I was therefore fortun- ate in being present when one was actually caught on a horse's back. We were bivouack- ing late one evening near Coquimbo, in Chile, when my servant, noticing that one of the horses was very restive, went to see what was the matter, and fancying he could dis- tinguish something, suddenly put his hand on the beast's withers, and secured the vam- pire. In the morning, the spot, where the bite had been inflicted, was easily distinguished from being slightly swollen and bloody. The third day afterwards we rode the horse, with- out any ill effects. The Chilean mammals collected and others, only observed by Darwin, are listed in Table 13. Departing Chile on 12 July 1835, the Beagle sailed north along the Peruvian coast before turn- ing west to the Galapagos Islands. Darwin's ac- counts of the stopovers in Iquique, Callao. and Lima make no mention of indigenous mammals. The Galapagos Archipelago, it seemed to Dar- win ( 1 839. p. 454), was "a little world within itself; the greater number of its inhabitants both vege- table and animal being found nowhere else." Dar- win (p. 464) "endeavoured to make as nearly a p)erfect collection in every branch as time permit- ted" but the only land mammals he found were the Chatham Island rice rats described by Water- house as Oryzomys galapagoensis (fig. 23), and the introduced Rattus on James Island. Darwin's investigations of the Galapagos fauna, Ijarticularly the birds, lizards, tortoises, and cer- tain plants stirred old beliefs and generated new and conflicting thoughts. However, at the time he wrote his journal in October 1835, Darwin (1839, p. 474) made no attempt to come to any definite conclusions, as the sp)ecies have not [as yet] been accu- rately examined; but we may infer, that, with the exception of a few wanderers, the organic beings found on this archipelago are peculiar to it; and yet their general form strongly par- takes of an American character. . . . This similarity in type between distant islands and continents, while the sp)ecies are distinct, has scarcely been noticed. The circumstances would be explained according to the views of some authors, by saying that the creative power had acted according to the same law over a wide area. Writers on Darwin, quoting from his revised (1845) edition of the Journal, attribute to Darwin more foresight on the origin of species than is ap- parent in the first (1839) edition quoted here. At the time of its publication, two years delayed, Dar- win, still a creationist and believer in the immut- ability of species, had yet to know the identity or specific aflinities of the vast majority of the plants and animals he had collected. This knowledge served him later for definition and elaboration of thoughts expressed in the second and other revised editions of the Journal, but not in the first. The following impressions of the biota of the Galapagos Islands in the second edition (p. 372 of an 1 899 "authorized edition") and oft quoted in whole or in part by various authors, are absent in the first. The natural history of the islands is emi- nently curious and well deserved attention. Most of the organic productions are aborig- inal creations, found nowhere else; there is 84 FIELDIANA: ZOOLOGY Table 13. Chilean mammals collected or only observed by Darwin, and those identified by Waterhouse (1838- 1839); the arrangement is phylogenetic. Current name Waterhouse synonym or misidentification L4>cality Figure Marsupialia Marmosa elegans Waterhouse Chiroptera Histiotus montanus magellani- cus Philippi Myotis chiloensis Waterhouse Tadarida brasiliensis I. Geoffroy Desmodiis rotundus dorbignyi Waterhouse Carnivora Dusicyon culpaeus magellanicus Gray Dusicyon fulvipes Martin Dusicyon griseus Gray Lutra felina Molina Felis concolor Linnaeus Artiodactyla Hippocamelus bisulcus Molina RODENTIA Oryzomys longicaudatus longi- caudatus Bennett Oryzomys longicaudatus magel- lanicus Bennett Akodon olivaceus olivaceus Waterhouse Akodon olivaceus brachiotus Waterhouse Akodon xanthorhinus xantho- rhinus Waterhouse Abrothrix longipilis longipilis Waterhouse Phyllotis darwini darwini Water- house Reithrodon chinchilloides Waterhouse Abrocoma bennetti Waterhouse Spalacopus cyanus Molina Myocastor coypus Molina Octodon degus Molina Ctenomys magellanicus fueginus Philippi Not Dysopes nasutus Spix Not Canis azarae Wied- Neuwied Lutra chilensis Bennett Mus renggeri Waterhouse Abrocoma cuvieri Waterhouse Poephagomys ater Cuvier Octodon cummingii Bennett Valparaiso Tierra del Fuego (observed) Chiloe Valparaiso Coquimbo Copiapo; Straits of Magellan Chiloe 22 Copiapo; Straits of Magellan Chonos Archipelago Tierra del Fuego and central Chile to 10,000 ft elevation (observed) Tierra del Fuego (observed) Concepcion Puerto de Hambre, Straits of Magellan Valparaiso; Coquimbo Chonos; Chiloe Hardy Peninsula, Tierra del Fuego Coquimbo Coquimbo 22 Straits of Magellan Valparaiso; Aconcagua Valparaiso Chonos Archipelago Valparaiso Tierra del Fuego (observed) even a difference between the inhabitants of the different islands; yet all show a marked relationship with those of America, though separated from that continent by an open space of ocean, between 500 and 600 miles in width. The archijjelago is a little world within itself, or rather a satellite attached to America, whence it has derived a few stray colonists, and has received the general char- acter of its indigenous productions. Consid- ering the small size of these islands, we feel the more astonished at the number of their aboriginal beings, and at their confined range. Seeing every height crowned with its crater, and the boundaries of most of the lava- streams still distinct, we are led to believe that within a period, geologically recent, the unbroken ocean was here spread out. Hence, both in space and time, we seem to be brought somewhat near to that great fact— that mystery of mysteries— the first appear- ance of new beings on this earth. HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 85 "-(SI 86 HELDIANA: ZOOLOGY XIV. Georges Louis Leclerc de Buffon (1707-1788) Georges Louis Leclerc de Buffon was bom into wealth and devoted his life to scientific labors; he won recognition as the leading naturalist of his time. In 1739 he was appointed keeper of the Jar- din du Roi in Paris (now the Jardin des Plantes), which he turned into one of the most important centers of biological research during the 18th cen- tury. Buffon's lifetime work was a general natural history in 36 volumes. The first volume dealt with science in general, the second with man, the next 1 3 with nonhuman mammals ( 1 750-1 767). These were followed by nine volumes on birds, seven volumes (1789) supplementary to the preceding, and the last five on minerals, including fossils. Treatment of most species in the Histoire Na- turelle is usually monographic. Gross descriptions, including measurements and weights, are based on individuals received in the Jardin du Roi. Geo- graphic distribution of the species is included with the description. Habits observed in captivity and mentioned in the literature are recorded. Anatom- ical descriptions by Daubenton, Buffon's collab- orator, are of the skeleton, with soft parts and tegumentary structures of particular interest. Complete bibliographic references and synony- mies, including those to the 10th edition of the Linnaean Systema Natura, accompany each spe- cies account. Buffon drew together much if not everything known of a species, often an indiscriminate com- posite of species. Most of the information was compiled, some of it original. Many life history notes were received from correspondents, partic- ularly M. de la Borde, the royal physician resident in Cayenne, French Guiana. Another correspon- dent, M. Saint Lurrent of Trinidad, believed he had solved the mystery of marsupial birth (cf p. 40). At a certain stage of development, he in- formed Buffon, the embryonic op>ossum crawled from the uterus through a tube at the end of which it found a long teat to which it remained attached until fully developed. An easily verifiable discov- ery by Daubenton (in Buffon) was that tapirs have simple stomachs, not the complex ruminant type claimed by Bajon (above). Buffon reported that domestic cats kill but do not eat shrew- or short- tail opossums of the genus Monodelphis. House cats do indeed kill these animals and usually de- posit them whole in the middle of the path leading from the house to the garden. Most of the illustrations of mammals and all anatomical drawings of the Histoire Naturelle are original. A small sampling is reproduced here (figs. 24-25). Buffon was the first naturalist to recognize re- gional faunas as such and to discriminate between Old World species and different but similar ap- pearing or like-named species of the New World. He perceived the platyrrhine-catarrhine dichoto- my of primates, and the phylogenetic distance be- tween the groups. He further distinguished pre- hensile-tailed monkeys from non-prehensile-tailed species, and cebids from callitrichids by their un- gues and teeth. Buffon's sense of rivalry with the contemporary Linnaeus led him to find fault with and cast scorn on the binomial system used in the Systema Na- turae. Buffon argued for retention of vernacular names for species as well as a makeshift vernacular terminology for generic or supergeneric groups. Lack of a scientific system of nomenclature in Buffon's work, and the almost universal adoption of the Linnaean binomial system by contemporary and later authors caused the Histoire Naturelle to be regarded as no better than a layman's encyclo- pedia of science. It has been republished with many revisions in many editions and languages. It is unfortunate that Buffon's important contributions to life histories, morphology, and evolutionary bi- ology were largely ignored by Darwin and are little appreciated today. It seems that the greater luster credited to Darwin owes much to the dimming of Buffon's because of his lack of organization and consistency in his writings. XV. Faunal Origins and Distribution Early attempts to explain observed similarities and differences between Old and New World mammals all supposedly descended from occu- pants of Noah's ark, began with the 1 6th century philosopher and chronicler Acosla and in some quarters continues to this day. Jose de Acosta (1539-1600) Jose de Acosta argued that the animals of the New World had not been carried there by man. His evidence indicated that New World man brought nothing but himself over a land route. The possibility that animals migrating from the ark might have crossed the Atlantic Ocean by swim- HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 87 IIJI.IXI /W ,.♦ j^^ ,„ J"/ X-W /-.ff //y I.B 5.VOOI IV \ ri.(;.\|HK.\IKNl AvPBi.r. Msnt, i>K :«i»iT. /■//. 11// /..y i.A URAXDK riiAivr-soiRis mn- im- i.ajht h" h iu'vaxnb IK c- \i;; \i Fig. 24. Mammals figured in the Histoire Naturelle of Buffon: upper left, le saki (= Pithecia pithecia Linnaeus, male; from Buffon, 1767); upper right, le sagouin singe de nuit (= Pithecia pithecia Linnaeus, female; from Buffon, 1 789); lower left, la grande chauve-souris fer-de-lance de la Guyanne (= Phyllostomus hastatus Pallas; from Buffon, 1 789); lower right, le cabiai (= Hydrochaeris hydrochaeris Linnaeus; from Buffon, 1 764). 88 FIELDIANA: ZOOLOGY rtijii j-^ lit I'fFK l"|-K 111' fllll.l l>K. MIR IK llf, l.A <.l'^.^NNK /■/ Win ..^ f- Xm >U 11 'CXXl/.fja td> 1 n \ 1 1,/ ^ r 1 m 1i- B^^Bl 1 ^i»' Via '■ 1 , \*«.ll m )<': fwfl 1 ■^/ i '^fjS 1 '' ■'u t 1 r "*■■ ■ r 1 -,-:5V ^>^^%P vi.: ' 'utA m ^ ■^3^5^ »«yr'-^->— ^ :^; i^?5??S»«— , -ttt; IK I \1l\RIN NK«i«l L II* ossr.rsE i>K 1.A noxvi! iir. i.'Ar.ot'Arrr. Fig. 25. Mammals figured in the Histoire Naturelle of BufTon: upper left, la mouffette du Chili (= Conepatus chinga Molina); upper right, la grande marte de la Guyanne (= Eira barbara Linnaeus); lower left, le tamarin n^e (= Saguinus midas Linnaeus); lower right, hyoid apparatus and thyroid cartilage of the throat ofAlouatta seniculus Linnaeus); from Buffon (1789). HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 89 ming or island hopping was also dismissed be- cause, as Acosta pointed out, none of the animals was known to occur on oceanic islands. The leg- endary island of Atlantis, which might have been part of a former transatlantic archipelago, was treated as fable. Other conjectures discarded, Acosta resolved that New and Old World northern continents were or had been connected, or very nearly approximated, at their polar extremities. Differences between New and Old World species, he affirmed, could be explained by the disappear- ance of connecting Old World populations, mu- tations among the New World sijecies engendered by their isolation in different environments, or by degeneration. No accounting or explanations were needed for "imperfect" organisms such as rats, frogs, insects, or vermin in general. These, it was commonly held, arose spontaneously from decay- ing matter. Antonio Vazquez de Espinosa (1560/1575-1630) Antonio Vazquez de Espinosa agreed with the explanation of a northern migratory or connecting route but added (1948, chap. 36), in the awkward phrasing of Clark's translation, that near the Straits of Magellan in what is called Tierra del Fuego, which is still not well known or explored, and there are numerous other quarters where the mainland of the New World could have communicated with that of the Old, or at least have lain so close as to afford passage not merely for the peo- ples who settled the New World, but the various kinds of animals which live in them— many of species well known in Eu- rope and elsewhere, and others peculiar and unique in the New World, like the Peruvian sheep [llamas], the guanacos [regarded as the wild form of llamas], vicuiias and tarugas [Hippocamelus bisulcus]. Carolus Linnaeus (1707-1778) Linnaeus played no direct role in the develop- ment of Neotropical mammalogy apart from pro- viding scientific names for some species discov- ered or described by others. His impact on the scientific world, however, was enormous. Like the philosophers before him, he believed all species were created as they are now in one place from which they spread in search of the habitats for which they were specially created. This idea of a staging area as the center of origin and dispersal, still dominant today in the minds of some students of Neotropical mammalogy, was critically reex- amined by Buffon. Georges Louis Leclerc de Buffon (1707-1788) Buffon, on comparing faunas of New and Old Worlds, was impressed by similarities between some of the species and differences between others. His explanation for similar-apjjearing sp>ecies, like that of the chroniclers, was that a land connection permitted passage of animals from Old to New World. Differences between New and Old World species, he suggested, in agreement with the chron- icler Jose Acosta, could have resulted from de- generation, environmental pressures, or isolation of the New World derivatives. On the other hand, Buffon argued, species peculiar to the New World or without Old World analogs must have arisen in situ, an opinion already intimated by Vazquez de Espinosa. Buffon was the first naturalist to envision the mammals of the region as a community or fauna that might well have originated indeF>endently of other regional faunas. Noah's ark had no place in his concept of faunal origins, and he rejected as too short the scripturally based 6,000-year esti- mate of the earth's age. Buffon's ideas of organic evolution and multiple centers of origin were nov- el and prepared the minds of his and succeeding generations for the acceptance of Darwinian evo- lution. Linnaeus, the arch exponent of the fixity of species and their origin and dispersal from a single center, conceived the elements of his binomial sys- tem as symbols for nailing down his credo. The system was so good it proved to be the best yet devised for the expression of genetic relationships between species and the surest base for the con- struction of evolutionary sequences in nominate terms. On the other hand, Buffon, independent of the religious constraints of his time and evolu- tionist in thought if not always in words, never attained the stature of his contemporary for lack of a competing system, key, code, or standard that would bring cohesion to his rambling philoso- phies. 90 FIELDIANA: ZOOLOGY Johann Andreas Wagner (1797-1861) Johann Andreas Wagner, the foremost masto- zoologist of his generation and author of a mono- graph on the geographic distribution of mammals, summarized (1844, p. 13) the three current but disparate opinions on mammalian origin and dis- persal. First, all species were created in one and the same region and spread from there to all cor- ners of the earth. Second, the species could have been created in separate localities in the same or different regions. Finally, each species could have arisen spontaneously anywhere and developed ac- cording to its peculiar constraints. Zoogeographers of the early half of the 19th cen- tury divided the world into major faunal regions correlated primarily with climate. Wagner (1844) separated the earth into four provinces: the Nord- liche north of 30°N, the Mittlere between BCN and 30°S, excluding the Australische roughly be- tween 0°S-55°S and 1 30°W-200''W, and the South American Magellanische, south of 30°S. The South American portion of the pantropical Mittlere Province extended from Mexico southward to southern Brazil and central Chile. Wagner's de- scriptions of the provincial faunas included tab- ulations of their resfjective genera and included species. Maximilian Prinz von Wied-Neuwied (1782-1867) Scriptural constraints were not evident in the thinking of the field naturalists. Maximilian Prinz von Wied-Neuwied recognized the limitations of geographic range as a property of a species. Johann Jacob von Tschudi (1818-1889) Tschudi attempted to follow Wied-Neuwied in defining specific ranges, but nearly all were based on the presumption that the geographic range of the species coincided with ecological life zones. The ecological life zones of Peru described by von Tschudi on the basis of fauna, flora, and climate, are the first of their kind for any Neotropical re- gion. Charles Robert I>arwin (1809-1882) The young Darwin also recognized geographic limitations of distribution in the light of physical barriers such as mountains and large bodies of water. XVI. Inventories to Middle of 19th Century Systema Naturm of Linnaeus, 1758, 1766 The 1 0th edition of the Systema Naturce pub- lished in 1758 by the Swedish naturalist Carolus Linnaeus ( 1 707-1 778), marks the beginning of the consistent application of his binomial system of zoological nomenclature. According to the uni- versally accepted International Code of Zoological Nomenclature, names for animals published be- fore 1 758 are not available, no matter how clearly defined the species. Likewise, zoological names for species published after 1757 that are not binomial or do not satisfy all provisions of the Code are not available. The effect of the Code in practice is that species without Linnaean names are treated as un- known to science. The 1 0th edition of the Systema Naturce lists a total of 1 72 species of mammals, exclusive of ma- rine cetaceans, each with its binomen consisting of a defined generic and defined specific name. Subsequent revisions of the bases for the names revealed that some represented more than a single species, others were duplicates or synonyms, and a few were equivocal or belonged to unidentifiable animals. The revisions, however, made no signif- icant change in the total number of real mam- malian species known to Linnaeus in 1758. The 1 2th and last revised edition of the Systema Naturae by Linnaeus himself, published in 1766, lists a world total of 208 mammalian species. Ta- ble 14 compares the relative numbers of world. Neotropical, and Nearctic genera and species in the Linnaean 10th and 12th editions of the Sys- tema Naturce with the totals in Buffbn's Histoire Naturelle. Cetaceans are omitted because they are oceanic species known before the discovery of America. Primary sources for the definition and naming of the Linnaean New World species were speci- mens preserved in the Swedish museums, partic- ularly the Adolphi Friderici Regis Museum, and primary bibliographic references. Such references for the Neotropical mammals were the works of Marcgraf (1648), Anson (1748), Browne (1756), and Seba (1 734-1 765). For both Neotropical and HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 91 Table 14. Number of world and New World mammals known to Linnaeus (1758, 1766) and to Buffon (1750- 1 789) and their percentage of world total. Author Date World total Neotropica Nearctica Linnaeus 1758 35 genera 18(51%) 15(43%) 1 72 species 45 (26%) 23(13%) Linnaeus 1766 36 genera 20 (55%) 15(42%) 208 species 55 (26%) 28(13%) BulTon 1750-1789 251 species 78(31%) 47(19%) Nearctic species, Linnaeus cited Hernandez (1651) and Edwards ( 1 743- 1751), and for Nearctic species alone, Catesby (1731) and Kalm (1753). Histoire Naturelle of Buffon, 1750-1789 In volume 9 of his Histoire Naturelle, published in 1 76 1 , Buffon estimated a world total of ap- proximately 200 known mammalian species of which 1 30 were Old World, 70 New World. When the sp)ecies of all 1 3 volumes on mammals and the supplementary volumes are counted, the total is 251, of which 78 are Neotropical, 47 Nearctic. The greater number of species recognized by Buffon as compared with those of Linnaeus re- flects his better knowledge of mammals and wider use of the available literature. Neither authority searched the works of the chroniclers for descrip- tions or figures of New World mammals. Synopsis Mammalium of Schinz, 1844 The Schinz catalogue of Recent mammals of the world, published 1844, brings the inventory of mammalian species to near the cutoff date of this part of the history of Neotropical mammalogy. The totals indicate that about 50% of all New World mammalian species now known had been de- scribed. The vast majority of the remaining 50% described since the middle of the 19th century are as small as or smaller than common tree squirrels. With respect to Neotropical mammals, by mid- 1 9th century about 90% of known sp)ecies larger than common tree squirrels had been described. In contrast, no more than about 1 0% of the smaller forms, mainly marsupials, bats, and rodents, had been named. The Schinz catalogue is summarized in Table 1 5. A first glance at the figures of the first order, the Marsupialia, may suggest the list is skewed. Only one marsupial species, Didelphis virginiana, is known to occur in Nearctica. Schinz recorded three because the species had been ovemamed at the time. It is believed, however, that "under- named" or composite species, as well as over- named identical species, are more or less evenly distributed in all three columns. As a result, the bottom line totals, particularly the percentages, are fair estimates of the real number of species known to science at the time of Schinz's compilation. The percentages have not changed significantly since. XVII. Summary Knowledge of Neotropical mammals from 1492 to the mid- 17th century was mainly an ag- gregation of anecdotes often riddled with myth, folklore, and untested generalizations. Eurojjeans identified New World species with similar-ap- pearing or similar-behaving Old World species and used the same names for them. Descriptions of mammals were usually comparisons with familiar European animals; measurements, rarely given, were rough estimates. Habitat when mentioned was usually on the order of "forest," "plain," or "river." Descriptions of animals often included use as food or pets, medicinal merits, or value of rawhide or bones in the manufacture of artifacts. Habits were usually described in terms of reactions to man when hunted or in captivity, or as harmful or benign to his person or property. Mammals— the term had not yet been concocted— were the hairy beasts of the earth. Whales and manatees were fish and could be eaten on Fridays. Bats were something else, mostly vampires; mice and other small mammals were vermin, in a class with frogs and cockroaches. Mammal collecting during this period was gen- erally limited to capture of live individuals for domestication, as pets, or for exhibition in zoos, circuses, or fairs. Dead animals were sometimes skinned and stuffed or bottled in brandy. The crudely prepared or pickled specimens, if not live 92 HELDIANA: ZOOLOGY animals, often served as models for the woodcut drawings of early treatises on natural history. Some specimens were purchased for museums or cabi- nets of collectors, including those of Linnaeus, King Frederick Adolph of Sweden, Reaumur of Paris, the King of France (Jardin du Roi), or the shelves of the Dutch pharmacist, Albert Seba. Most of the Neotropical specimens probably originated in the South American possessions of Holland and France. The few crude attempts at classification of mam- mals during the 16th and 17th centuries were hard- ly more than random arrangements equivalent to shopping lists. Species, being individually created kinds, were unrelated to other created kinds, or simply arose spontaneously from putrefying mat- ter. The scientific study of mammals, or mammal- ogy, of the Neotropical Region began with the ex- plorations of northeastern Brazil by Georg Marc- graf and culminated with the publication in 1648 of his Historic Rerum Naturalium Brasilia. His accounts of the included 32 sp)ecies of mammals reveal the glimmer of an attempt at natural group- ings of kinds or the beginnings of a classification of Neotropical mammals. Insofar as is known, none of MarcgraPs animals were preserved. Lin- naean names for the species of the Historice were based on bibliographic references to their descrip- tions and figures (cf. fig. 2, table 1). The first expedition to the Neotropical Region actually committed to the collection and perma- nent preservation of mammals (and other objects) for scientific study was the Brazilian Viagem Fi- losofica, 1783 to 1792, conceived by the Portu- guese government and conducted by the Brazilian- bom naturalist Alexandre Rodrigues Ferreira. The large number of specimens gathered by the ex- pedition was deposited in Lisbon's Museu d'A- juda. The specimens of monkeys that had been carried away to the Paris Natural History Museum were studied by the French scientist Etienne Geof- froy St.-Hilaire. His descriptions were published without reference to source of material. Alexander von Humboldt followed on the heels of the Viagem Filosofica with his explorations of northwestern South America from 1799 through 1802. His expedition was highly successful and in scope has rarely been equaled by other "one-man" surveys of a large portion of the South American continent. The personal narrative of his travels inspired successive naturalist-travelers, most no- tably the explorers of Brazil, Spix and Martius, Maximilian Wied-Neuwied, and Johann Natterer. Table 1 5. Number of world, Neolropic, and Nearc- tic species (subspecies) of mammals known to Schinz ( 1 844); species common to both regions are included in both. Order World Neotropica Nearctica Marsupialia 138 31(22%) 4(3%) Insectivora 114 2(2%) 21 (18%) Chiroptera 326 110(34%) 21 (6%) Primates 281 73 (26%) 0(0%) Edentata 31 24 (77%) 0(0%) Camivora 303 58(19%) 41 (13%) Pinnipedia 39 1 1 (28%) 2(5%) Sirenia 5 2 (40%) 1 (20%) Perissodactyla 23 2(9%) 0(0%) Artiodactyla 186 1 1 (6%) 12(6%) Lagomorpha 52 4(8%) 14(27%) Rodentia 563 152(27%) 104(18%) Cetacea 4 2 (50%) 0(0%) Proboscidea 2 0(0%) 0(0%) Hyracoidea 5 0(0%) 0(0%) Totals 2,072* 482 (23%) 220(11%) * The estimated number of species in 1972 (Hersh- kovitz, p. 332, table 3) is Neotropica 810, Nearctica 442 or an approximate doubling since 1 844 in both regions, with a slightly greater increase (less than 2%) in Nearctica relative to Neotropica. Increase since then has been al- most exclusively Neotropical. Later there was von Tschudi, who traveled in Peru; d'Orbigny, who journeyed in Patagonia but did his finest and most lasting work in Bolivia on a scale almost equal to that of Humboldt's; and Darwin, who voyaged around the southern half of South America and the Galapagos Islands in H.M.S. Beagle. The brothers Schomburgk, motivated by Hum- boldt's trip up the Rio Orinoco to its connection via the Casiquiare Canal with the Rio Negro trib- utary of the Rio Amazonas, completed the trajec- tory by plotting the course of the upper Rio Negro to its connection with the Casiquiare. Their ex- plorations of the British Guianas and bordering parts of Brazil and Venezuela yielded the first large collection of Guianan mammals, all depos- ited in the Berlin Natural History Museum. Chilean mammals became fairly well known through the reports of Molina (1782), Poeppig (1836), and Gay (1847). The mammals of Para- guay, their distribution, habits, or biology in gen- eral, became better known through the efforts of Felix de Azara than those of other Neotropical countries. The 200-year period from MarcgraPs ( 1 648) re- port to the last of those of Schomburgk (1 848) was one of survey and inventory of South American mammals. The published reports and personal HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 93 narratives of travel provided much reliable data on geographic distribution, habitat, life histories, ecological backgrounds, itineraries and maps of the expeditionary routes, and stopping and col- lecting localities. Descriptions of the collected mammals, most of them by the naturalist-trav- elers themselves, were often based on skeletal, dental, and soft parts in addition to purely tegu- mentary characters. Their classifications were pu- tatively natural groupings on the ordinal, family, and, as a rule, the generic levels. The prevailing belief in the biblical version of creation and fixity of species, not confessed in writing, did not blind systematists of the period to evident relationships between species and their clusterings into supra- specific groups. Descriptions of species were, nevertheless, typological. Subspecies or geograph- ic races were, at best, vaguely conceived but de- scribed as species. The infrequent or rare use of trinomials was accidental or equivocal and not certainly intended for a clearly defined geographic race. The term usually used for deviates firom "types" was "variety." Controversies regarding origin of species or fau- nas centered on where, not how. Philosopher- chroniclers of the first era accepted Noah's ark literally as the one place of origin and disjiersal of the Recent fauna. Acosta may have been the first to suggest the former existence of intercontinental connections for passage of Old World animals into the New World. More and better knowledge of the world's fauna during the second era revealed the weaknesses or fallacy of the ark dogma. Staunch creationists such as Linnaeus pointed instead to a vaguely located region as the place from which all species dispersed to occupy predestined habitats for which they had been created. Other authorities like Buffon argued for multiple centers of origin, with sp)ecies origi- nating in the habitats for which they were adapted. Darwin also believed in multiple places of origin, or faunal regions separated by geographic barriers but with some trepidation. The belief in multiple creations was heretical. Inconsistencies between religious dogma and realities did not prevent Wied-Neuwied from rec- ognizing the geographic range of a species (or sub- species) as a property of that species. Another advance beyond scriptures was the concept of eco- logical life zones contributed by von Tschudi, who plotted them for Peru on the basis of plants, an- imals, and climate. The total of named Neotropical sjjecies of mam- mals counted ft-om 1758, the year of publication of the 10th edition of the Linnaean Systema Na- tura and starting date of zoological nomenclature, to mid- 1 9th century, exceeded by far that of the Nearctic region and any other equivalent area of the world. Neotropical mammals were also better known than those of other continents except west- em Europe. By mid- 1 9th century, about 90% of currently known Neotropical mammalian species larger than common tree squirrels had already been described, but no more than about 10% of the smaller forms. The great number and variety of Neotropical mammals (and animals generally) known to sci- ence by mid- 19th century and the accumulated knowledge gained from study of living and pre- served specimens in field and laboratory, much of it contributed by Charles Darwin, helped pscvt the way to the Darwinian revolution of the next half century. XYIII. Acknowledgments I am indebted to Benjamin W. Williams, As- sociate Librarian and Librarian of Rare Books, Field Museum of Natural History, for p)ermission to consult at pleasure in the Museum's Mary W. Runnells Rare Book Room the books needed for writing this article; and to Bruce D. Patterson, Robert M. Timm, Ronald H. Pine, Debra Mos- kovitz, and J. A. Gagliano for reviewing the manu- script. Map>s shown in Figures 11, 12, and 2 1 were prepared by the author with assistance of Mary Anne Rogers from accounts of the travelers cited and other sources. Photographic reproductions of the figures are by Field Museum of Natural His- tory Staff Photographer Ron Testa. Technical Assistants Barbara Brown and Mary Anne Rogers typed the manuscript and contributed in other ways toward its completion. XIX. Literature Cited AcosTA, JosE DE- 1590. Historia natural y moral de las Indias en que se tratan las cosas notables del cielo, elementos, metales, plantas y animales dellas y los ritos y ceremonias, leyes y gobiemo y guerras de los Indios. Juan de Leon, Seville, 535 + (36) pp. [not seen]. . 1894. Historia natural y moral de las Indias. Escrita por el P. Joseph de Acosta de la Compaiiia de Jesus. Publicada en Sevilla en 1 590 y ahora fielmente 94 nELDL\NA: ZOOLOGY reimpresa de la primera edicion. Ramon Angles, Ma- drid, 2 vols. Anghiera, see Martyr. Anson, George. 1 748. A Voyage Round the World in the Years MDCCXL, I, II, III, IV by George Anson . . . Esq. Compiled and published by Richard Walter. J. & P. Knapton, London. AviLA Pikes, Fernando Dias de. 1 965. The type spec- imens of Brazilian mammals collected by Prince Maximilian zu Wied. American Museum Novitates, no. 2209: 1-21. . 1 974. CaracterizafSo zoogeografica da Provin- cia Amazonica. I. Expedi^oes cientificas na Amazonia Brasiliera. Anales Academia Brasiliera de Ciencias, 46(1): 133-158. AzARA, Felix de. 1801. Essais sur I'histoire naturelle des quadrupedes de la Province du Paraguay. Trans- lated from the original Spanish by M. L. E. Moreau- Saint-Mery. Charles Pougens, Paris, vol. 1, xxvii + 366 pp.; vol. 2, 499 pp. . 1802. Apuntamientos para la historia natural de los quadrupedos del Paraguay y Rio de la Plata. La Viuda de Ibarra, Madrid, vol. 1, xiii + 318 pp.; vol. 2, (3) + 328 pp. . 1809. Voyages dans I'Amerique Meridionale. Contenant la description geographique politique et civile du Paraguay et de la riviere de La Plata; I'histoire de la decouverte et de la conquete de ces contrees; des details nombreaux sur leur histoire naturelle, et sur les peuples sauvages qui les habitent, le recit des moyens employes par les Jesuits pour assujetir et civiliser les indigenes, etc. Publics d'apres les manuscrits de Tau- teur, avec une notice sur sa vie et ses ecrits. Translated from the original Spanish and annotated by M. Son- nini, with notes by G. Cuvier; edited by C. A. Walck- enaer. Dentu, Paris, 4 vols, with atlas. Bajon, M. 1777-1778. Memoires pour servir a I'his- toire de Cayenne et de la Guiane Frangaise. Grange, Paris, vol. 1, 1777; vol. 2, 1778. Bancroft, Edward. 1769. An Essay on the Natural History of Guiana in South America. T. Beckett and P. A. DeHondt, London, (8) + 402 + (2) pp. Barrere, Pierre. 1741. Essai sur I'histoire naturelle de la France Equinoxiale ou de'nombrement des plantes, des animaux, y des mineraux, que se trouvent dans I'lsle de Cayenne, les Isles de Remire, sur les cotes de la mer, & dans le continent de la Guyane. Piget, Paris. Bellin, Jaques Nicolas. 1763. Description geogra- phique de la Guiane. Didot, Paris. Bokermann. Werner C. A. 1957. Atualiza^o do iti- nerario de viajem do Principe de Wied do Brasil (1815- 1817). Arquivos de 2^oologia (Rio de Janeiro), 10(3): 209-251. Browne, Patrick. 1756. The Civil and Natural His- tory of Jamaica. Privately printed, London. BuFFON, George Louis Leclerc, Comte de. 1 753-1 789. Histoire naturelle, generale et particuliere, avec la de- scription du cabinet du roy. L'Imprimerie Royale, Paris, 44 vols., 1 750-1804; vols. 4-15, Quadrupedes par M. de Buffon et M. Daubenton, 1753-1767; vol. 9, 1761; supplement, vol. 107 par M. de Buffon, 1774-1789. Catesby, Mark. 1731-1743. The Natural History of Carolina, Florida and the Bahama Islands. Privately printed, London, vol. 1,(2) + vii + (1) + xliv + 100 + 2 pp., 100 pis., fold, map; vol. 2, 2 + 120 + (6) + 2 pp., 120 pis. Charlevoix, Pierre Francois Xavier DE. 1757. His- toire du Paraguay. Oesaint, David, Durand, Paris, 6 vols. Darwin, Chari,es Robert. 1839. Journal of Re- searches into the Geology and Natural History of the Various Countries Visited by H.M.S. Beagle. Under the Command of Captain Fitzroy, R.N., from 1832 to 1836. Henry Colbum, London, xiv + 615 pp., 16 pis. Edwards, George. 1743-1751. A Natural History of Uncommon Birds, and of Some Other Rare and Un- described Animals. Printed for the author at the Col- lege of Physicians, London, 4 vols. Fermin, Philippe. 1769. Description generale, histo- rique, geographique et physique de la colonic de Su- rinam. E. van Harrevelt, Amsterdam, 2 vols. Ferreira Rodriguez, Alexandre, see Rodriguez Fer- reira, Alexandre. Gaffarel, Paul. 1907. See Martyr, Anghiera Pierre. Gay, Claudio. 1 847. Historia fisica y politica de Chile. Vol. 1, Zoologia. Museo de Historia Natural de San- tiago [de Chile], Malde y Renow, Paris. GuMiLLA, Jose. 1741. El Orinoco ilustrado, historia natural, civil y geograhca, de este gran rio, y de sus caudalosas vertientes; gobiemo, usos y costumbres de los Indios, sus habitadores con nuevas, y utiles noticias de animales, arboles, frutos, aceytes, resinas, yervas, y singulares a nuestra santa fe, y casos de mucha edi- licacion. M. Fernandez, Madrid. . 1758. Histoire naturelle, civile et geographique de rOrenoque. Translated from the Spanish. 2nd edi- tion by M. Eidous. Veuve de F. Firard, Avignon, 3 vols. Hernandez, Francisco. 1651. Nova plantarum, ani- malium et mineralium mexicanorum historia. B. Deuersini & Z. Masotti, typis V. Masacardi, Rome, (16) + 950 + 90 pp. . 1959. Historia natural de Nueva Espafia, vol. 2. Translated by Jose Rojo Navarro. Universidad Na- cional de Mexico, D.F. Herrera y Tordesillas, Antonio de. 1 60 1 . Historia general de las Indias occidentalis: o, de los hechos de los castellanos en las islas y tierra firme de mar oceano, escrita por Antonio de Herrera, cronista mayor de Su Magestad de las Indias y de Castilla. En la Emplen- terea, 4 vols, [work not seen, cited from reviewer com- ments]. . 1601-1615. Historia general de las Indias oc- cidentalis: o, de los hechos de los castellanos en las islas y tierra firme del mar oceano, escrita por Antonio de Herrera, cronista mayor de Su Magestad de las Indias y de Castilla. Madrid, 9 parts in 4 vols, [work not seen, cited from reviewer comments]. 1725-1726. The General History of the Vast Continent and Islands of America, Commonly Called, the West Indies, from the First Discovery Thereof: With the Best Accounts the People Could Give of the Antiquities. Collected from the original relations sent HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 95 to the Kings of Spain. Translated into English by Capt. John Stevens. Batley, London, vol. 1, 1725; vols. 2- 6, 1726. . 1728. Historia general de las Indias occiden- talis: o, de los hechos de los castellanos en las islas y tierra firme del mar oceano, escrita por Antonio de Herrera, cronista mayor de Su Magestad de las Indias y de Castilla. En ocho decadas. Sigue a la ultima de- cada. la Description de las Indias, por el mismo autor. J. B. Verdussen, Amberes [work not seen, cited from Library of Congress catalogue]. Hershkovitz, Philip. 1972. The Recent mammals of the Neotropical Region: A zoogeographic and ecolog- ical review, pp. 31 1-431. In Keast, A., F. C. Erk, and B. Glass, eds.. Evolution, Mammals and Southern Continents. State University of New York Press, Al- bany. Humboldt. Alexander von. 1811-1812. Voyage de Humboldt et Bonpland. Observations de zoologie et d'anatomie comparee. Pt. 2, vol. 1 , Recueil d'obser- vations de zoologie et d'anatomie comparee, faites dans I'ocean Atlantique dans Tinterieur du nouveau continent et dans la mer du sud pendant les annees 1799, 1800, 1801, 1802 et 1803. Levrault Schoell, Paris, viii + 368 pp., 40 pis. . 1838. Ueber den Manati des Orinoko. Trans- lated from the original manuscript in French by the editor. Wiegmann's Archiv fiir Naturgeschichte, 4(1): 1-10. . 1884. Personal Narrative of Travels to the Equinoctial Regions of America During the Years 1799-1804 by Alexander von Humboldt and Aime Bonpland. Translated from the original French and edited by Thomasina Ross. George Bell and Sons, London, 3 vols. Ihering, Herman von. 1 902. Natterer e Landsdorff. Exploradores antigos do estado de Sao Paulo. Revista do Museu Paulista (Sao Paulo), 5: 13-34. Kalm, Pehr. 1753-1761. En resa til Norra America. Pa Kong. Svenska Vetenskapsacademiens befallning, och publici Kostnad. Stockholm, 3 vols. Kraft, Richard. 1983. Die von J. B. von Spix bes- chriebenen neotropischen Primaten und Chiropteren. Verzeichnis der in der Zoologischen Staatssammlung Miinchen aufbewahrten Typusexemplare. Spixiana Suppl., 9: 429-441. KuHL, Heinrich. 1 820. Beitrage zur Zoologie und ver- gleichenden Anatomic, Abt. 1. Frankfurt am Main, 1 52 + unnumbered pp. Langguth, A. 1978. Revived application in the case of the names for South American rodents published by Brants ( 1 827). Z.N.(S.) 1 775. Bulletin of Zoological Nomenclature, 35(2): 115-120. Linnaeus, Carolus. 1758. Systema Naturae. Regnum animate, ed. 10, vol. 1. Holmiae, 4 + 824 + 3 pp. . 1 766. Systema Natura per regna tria naturae, secundum classis, genera, species, cum characteribus, differentius, synonymis locis . . . , ed. 12 reformata, vol. 1. L. Salvii, Holmiae, 1 + 327 + 36 pp. Marcgraf [or Marcgrav, or Marggraf] de Lieb- STAD, Georg. 1648. Historiae rerum naturalium Brasiliae, libro octo cum appendice de Tapuyis, et Chi- lensibus. loannes de Laet, Antvverpianus, In ordinem digessit & annotationes addidit multas, &. varia ab auctore omissa supplevit & illustravit. In Piso, W., Historia Naturalis Brasilliae, asspicio et beneficio II- lustras. I. Maritii Com. Nassau. Issius provincije et Maris summi praefecti adomata in qua non tantum plantx et animalia sed et indiginarum morbi, ingenia et mores describuntur et iconibus scipta quingentas illustrantus. Fransiscum Hackium, et Amstelodami, Lud, Elizevirium, Leiden. Martyr, Anghiera Pierre. 1907. De orbe novo. Les huit decades. Translated from Latin with notes and comments by Paul Gaffarel. Recueil de voyages et de documents pour servir a I'histoire de la geographic depuis le XIII' jusqu'a la fin du XVI"^ siecle. No. XXI. Imprimerie Nouvelle Noel Texier, La Rochelle, France. . 1912. De orbe novo. The Eight Decades of Peter Martyr d'Anghiera. Translated from the Latin with notes and introduction by Francis Augustus MacNutt. G. P. Putnam's Sons, New York & London, 2 vols. Miller, Gerrit S., Jr. 1929. Mammals eaten by In- dians, owls, and Spaniards in the coast region of the Dominican Republic. Smithsonian Miscellaneous Collections, 82(5): 1-16. Molina, Giovanni Ignazio. 1 782. Saggio sulla storia naturale del Chili. S. Tommaso d'Aquino, Bologna, 367 + 1 pp. Orbigny, Alcide Dessalines d'. 1835-1847. Voyage dans I'Amerique Meridionale (le Bresil, la Republique orientale de Uruguay, la Republique Argentine, la Pa- tagonie, la Republique du Chile, la Republique de Bo- livia, la Republique du Perou) execute pendant las annees 1826, 1827, 1828, 1829, 1830, 1831, 1832, et 1833. Pi tois- Levrault et cie, Paris, Strasbourg, 9 vols. Orbigny, Alcide d', and Paul Gervais. 1847. Mam- miferes, vol. 4, pt. 2. In Orbigny, A. d'. Voyage dans I'Amerique Meridionale (le Bresil, la Republique ori- entale de I'Uruguay, la Republique Argentine, la Pa- tagonie, la Republique du Chili, la Republique de Bo- livia, la Republique du Perou) execute pendant les annees 1826, 1827, 1828, 1829, 1830, 1831, 1832, 1833. Pitois-Levrault et cie, Paris, Strasbourg. Osgood, Wilfred H. 1943. The mammals of Chile. Publications of Field Museum of Natural History, Zoological Series, 30: 1-268. OviEDO Y Valdes, Gonzalo Fernandez de. 1 944- 1 945 . Historia general y natural de las Indias, Islas y Tierra- Firme del Mar Oceano. Prologo de J. Natalicio Gon- zalez. Notas de Jose Amador de los Rios. Editorial Guarania, Asuncion (Paraguay), 1 4 vols, [text and il- lustrations said to be reproduced from the original manuscript]. Pelzeln, August von. 1871. Ueber Omithologie Brasiliens. Resultate von Johann Natterer's Reisen in den Jahren 1817 bis 1835. A. Pichler's Witwe & Sohn, Vienna. . 1883. Brasilische Saugethiere. Resultate von Johann Natterer's Reisen in den Jahren 1 8 1 7 bis 1 835. Verhandlungen der kaiserlich-koniglichen zoologisch- botanischen Gesellschaft in Wien, 33: 1-140. PiLLERi, G., and Lucie Arvy. 1977. La expedicion de Alcide d'Orbigny en Bolivia y el descubrimiento del delphin boliviano de agua dulce Inia boliviensis (d'Or- 96 FIELDIANA: ZOOLOGY bigny, 1834). Himanatomisches Institut, Bern, Swit- zerland, 15 pp., 4 figs., 16 pis., fold. map. PoEPPiG, Eduard. 1835-1836. Reise in Chile, Peru und auf dem Amazonenstrome wahrend der Jahre 1827-1832. Friedrich Fleischer, Leipzig, vol. 1, xii + 466 pp.; vol. 2, V + 464 pp. Rengger, J. R. 1830. Naturgeschichte der Saeuge- thiere von Paraguay. Schweighauserschen Buchhand- lung, Basel, Switzerland, 16 + 394 pp. RoDRiGUES Ferreira, ALEXANDRE. 1971. Viagem fi- losofica pelas capitanias do Grao-Para, Rio Negro, Mato Grosso e Cuiaba. Iconografia. Vol. 2, Zoologia. Conselho Federal de Cultura, Rio de Janeiro. Editora Monumento S. A., Sao Paulo. Artes Graficas Gomes de Souza S. A., Rio de Janeiro, 168 color plates. . 1972. Viagem filosofica pelas capitanias do mecophagusjubata). Proceedings of the Zoological So- ciety. Annals of Natural History, ser. 1 , 4(23): 202. 1840. Information respecting botanical trav- Grao-Para, Rio Negro, Mato Grosso e Cuiaba. Me- morias. Zoologia, Botanica [nomenclatorial notes by Luiz Carlos Souto; introduction by Jose Candido de Melo Carvalho]. Conselho Federal de Cultura. De- partamento de Imprenta Nacional, Rio de Janeiro, 246 pp. Roth, Walter E. 1931. See Schomburgk, Richard, and Schomburgk, Robert Hermann. ScHiNZ, Heinrich R. 1821. Das Thierreich eingetheilt nach dem Bau der Thiere als Grundlage ihrer Natur- geschichte und der vergleichenden Anatomic von dem Herm Ritter von Cuvier. Vol. 1 , Saugethiere und Vo- gel. J. G. Cotta'schen Buchhandlung, Stuttgart and Tiibingen. . 1844-1845. Systematisches Verzeichniss aller bis jetzt bekannten Saugethiere oder Synopsis Mam- malium nach dem Cuvier'schen System. J. Gassmann, Sohn, Solothum, Switzerland, 2 vols. Schomburgk., Richard. 1847-1848. Reisen in Brit- ish-Guiana in den Jahren 1840-1844. Im auftrag Sr. Majestat des Konigs von Preussen. Nebst einer Fauna und Flora Guiana's nach Vorlagen von Johannes Miil- ler, Ehrenberg, Erichson, KJotzsch, Troschel, Cabanis und Andem. Mit Abbildungen und einer Karte von British-Guiana afgenommen von Sir Robert Schom- burgk. Versuch einer Fauna und Flora von British- Guiana ( 1 848). J. J. Weber, Leipzig, vol. 1 , 1 847; vols. 2-3, 1848. . 1922-1923. Travels in British Guiana During the Years 1840-1844 Carried out Under the Com- mission of His Majesty the King of Prussia. Together with a Fauna and Flora of Guiana According to the Works of Johannes Miiller, Ehrenberg, Erichson, KJotzsch, Troschel, Cabanis and Others. Including il- lustrations and a map of British Guiana drawn by Sir Robert Schomburgk. Translated and edited, with geo- graphical and general indices and route maps, by Wal- ter E. Roth. "Daily Chronicle," Georgetown, British Guiana, vol. 1 (1922), 38 + 402 pp., 8 pis., 4 maps, 2 figs.; vol. 2 (1923), (12) + 443 pp., 8 pis., 7 maps. Schomburgk, Robert Herman. 1 840. A Description of British Guiana, Geographical and Statistical: Ex- hibiting Its Resources and Capabilities, Together with the Present and Future Condition and Prospects of the Colony. Simpkin, Marshall, and Co., London. . 1 840. Remarks on the greater ant-bear (Myr- ellers. Recent expedition in Guiana. Annals of Natural History, ser. 1, 4(24): 262-266; (25): 322-328; (26): 429-434; 5(28): 29-35; (31): 282-288; (32): 343-348; (33): 399-403. . 1841. Reisen in Guiana und am Orinoko wah- rend der Jahre 1835-1839. Nach seinen Berichten und Mitteilungen an die geographische Gesellschaft in London. Edited by O. A. Schomburgk, with a foreword by Alexander von Humboldt. Georg Wigand, Leipzig. . 1931. Travels in Guiana and on the Orinoco During the Years 1835-1839. According to his Re- ports and Communications to the Geographical So- ciety of London. Edited by O. A. Schomburgk, with a preface by Alexander von Humboldt. English trans- lation by Walter E. Roth. "The Argosy" Company, Limited, Georgetown, British Guiana. Seba, Albertius. 1734-1765. Locuplestissimi rerum naturralium thesauri accurata descripti et iconibus ar- tificiosissimisexpressio, per universam historiam. . . . Jansson-Waesberg, Amsterdam, 4 vols., 449 color pis. Spix, Jean de. 1 823. Simiarum et vespertilionum bra- siliensium species novae: ou histoire naturelle des es- pecies nouvelles de singes et de chauve-souris obser- vees et recueillies pendant le voyage dans I'interieur du Bresil. Typis Francisci Serephici Hiibschmanni, Monaco, 16 + 72 pp., 38 color pis. Spix, Johann Baptist von, and Carl Friedrich von Martius. 1823-1831. Reise in Brasilien auf Befehl Sr. Majestat Maximilian Joseph 1., Konigs von Baiem, in den Jahren 1817 bis 1820. M. Lindauer, Munich, vol. 1, 1823; vol. 2, 1828; vol. 3, 1831. . 1824. Travels in Brazil in the Years 1817- 1820, Undertaken by Command of His Majesty the King of Bavaria. Translated from the original German by H. E. Lloyd. Longman, Hurst, Rees, Orme, Brown, and Green, London, vol. 1, 24 + 327 pp., 4 pis.; vol. 2, 8 + 298 pp., 5 pis. Stedman, John Gabriel. 1 796. Narrative of a Five Years' Expedition Against the Revolted Negroes of Surinam in Guiana on the Wild Coast of South Amer- ica; from the Year 1772 to 1777: Elucidating the His- tory of that Country, and Describing Its Productions, viz. Quadruf>eds, Birds, Fishes, Reptiles, Trees, Shrubs, Fruits, & Roots; with an Account of the Indians of Guiana, & Negroes of Guinea. J. Johnson and J. Ed- wards, London, vol. 1,18 + 423 pp., 18 pis., 3 maps; vol. 2, 9 + 419 + (5) pp., 38 pis., 2 maps. Stevens, John. 1726. See Herrera y Tordisillas, An- tonio de. Tamayo H., Manuel, and C. Daniel Frassinetti. 1 980. Catalogo de los mamiferos fosiles y vivientes de Chile. Boletin Museo Nacional de Historia Natural de Chile, 37: 323-399. Thomas, Oldfield. 1908. On certain African and S.- American otters. Annals and Magazine of Natural History, ser. 8, 1: 387-395. . 1911. The mammals of the tenth edition of Linnaeus; an attempt to fix the type of the genera and the exact bases and localities of the species. Proceed- HERSHKOVITZ: HISTORY OF NEOTROPICAL MAMMALOGY 97 ings of the Zoological Society of London, 1911: 158. 120- TiEFENBACHER, L., ED. 1983. Fcstschrift zu Ehren von Dr. Johann Baptist Ritter von Spix. Spixiana, Suppl. 9. TscHUDi, Johann Jacob von. 1844a. Mammalium conspectus quae in Republica Peruana reperiuntur et pleraque observata vel collecta sunt in itinerere. Ar- chiv fiir Naturgeschichte, 10(1): 244-255. . 1844b. Untersuchungen iiber die Fauna Pe- ruana. Physiognomik von Peru (1844); Therologie (1844b); Omithologie [with J. Cabanis] (1845-1846); Herpetologie (1845 [1846]). Scheitlin und ZoUikofer, St. Gallen, Switzerland. . 1 846. Peru. Reiseskizzen aus den Jahren 1 838- 1842. Scheitlin und Zollikofer, St. Gallen, Switzer- land. 1847. Travels in Peru During the Years 1838- 1 842. On the Coast, in the Sierra, Across the Cordi- lleras and the Andes into the Primeval Forests. English translation from the original German by Thomasina Ross. Wiley and Putnam, New York. Vazquez DE EspiNOSA, Antonio. 1942. Compendium and description of the West Indies. Translated by Charles Upson Clark. Smithsonian Miscellaneous Collections, 102: i-xii; 1-862. . 1948. Compendio y descripcion de las Indias Occidentals. Transcrito del manuscrito original por Charles Upson Clark. Smithsonian Miscellaneous Collections, 108: i-xii; 1-801. Wagner, Johann Andreas. 1840-1855. Die Sauge- thiere in Abbildungen nach der Natur mit Beschrei- bungen von Dr. Johann Christian Daniel von Schre- ber, Supplementenbanden 1-5. Erlangen, Germany. . 1 842. Diagnosen neuer Arten brasilischer Sau- gethiere. Archiv fur Naturgeschichte, 8(1): 356-362. . 1843. Diagnosen neuer Arten brasilischer . 1844-1846. Die geographische Verbreitung der Saugethiere. Abhandlungen der Mathem.-physika- lischen Classe der Koniglich Bayerischen Akademie der Wissenschaften, 4(1): 1-116 (1844); (2): 39-108 (1845); (3): 1-114 (1845); (3): 241-244, explanation of maps, plates I-IX (1846). 1850. Beitrage zur Kenntniss der Saugethiere America's. Abhandlungen der Mathem.-physikali- schen Classe der Koniglich Bayerischen Akademie der Wissenschaften (Miinchen), 5: Erste Abt. [Marsupi- alia, Chiroptera], 1 19-208 (1847); Zweiten Abt. [Ro- dentia], 269-332 (1847); Dritte Abt. Vierte Ordnung Affen, 405-480(1848). Waterhouse, George R. 1838-1839. Mammalia. /« Darwin, C, The Zoology of the Voyage of H.M.S. Beagle, Under the Command of Captain Fitzroy, Dur- ing the Years 1832-1836. With Notes by Charles Dar- win, part II. Smith, Elder and Co., London, no. 1, 1838; no. 2, 1838; no. 3, 1838; no. 4, 1839. Wied-Neuwied, Maximilian Alexander Philipp, Prinz von. 1820. Ueber ein noch unbeschriebenes Sau- gethier aus der Familie der Nager. Isis (Oken's), 1: 43. . 1820-1821. Reise nach Brasilien in den Jahren 1815 bis 1817. Heinrich Ludwig Bronner, Frankfurt A.M., vol. 1, 1820; vol. 2, 1821. . 1821. Ueber die Brasilianischen Hirschart. Isis (Oken's), 2: 649-651. . 1822-1831. Abbildungen zur Naturgeschichte Brasiliens. (Recueil de Planches Coloriees d'Animaux du Bresil.) Weimar, Germany, 90 color pis. with de- scriptive text issued in 1 5 parts with 6 pis. each (30 pis. of mammals in parts 1-5, 1822-1824). . 1826. Beitrage zur Naturgeschichte von Brasilien, vol. 2. Gr. H. S. priv. Landes-Industrie- Comtpoirs, Weimar, Germany. Handflugler. Archiv fiir Naturgeschichte, 9(1): 365- 368. 98 HELDIANA: ZOOLOGY A New Superfamily in the Extensive Radiation of South American Paleogene Marsupials Rosendo Pascual and Alfredo A. Carlini ABSTRACTS Significant new mammals have been recovered from the Colhuehuapian mammal-bearing beds (latest Oligocene) exposed in the Gaiman region of Patagonia (Chubut Province, Argentina). Some fragmentary mandibles and isolated teeth belong to a new genus and species, Patagonia peregrina. The mandibular and dental specializations of this form are so distinctly convergent on those of some fossorial rodents that it is regarded as a distinct clade of South American marsupials. It represents the type of a new family, Patagoniidae, which is assigned to a new superfamily, Patagonioidea, which represents a natural evolutionary group in the same sense as other superfamilies of marsupials recognized by Simpson. Its systematic position within the superorder Marsupialia awaits comprehensive analysis of those enigmatic marsupials (Groe- berioidea and Argyrolagoidea) apparently most closely related to it. Varios nuevos y significativos mamiferos han sido recogidos de capas mamaliferas del Col- huehuapense (Oligoceno tardio) expuestas en la region de Gaiman, Patagonia (Chubut, Argen- tina). Algunos fragmentos mandibulares y dientes aislados pertenecen a un nuevo genero y especie, Patagonia peregrina. Esta forma presenta especializaciones mandibulares y dentarias tan distintamente convergentes hacia las de albunos roedores cavadores que es considerada como un distinto clado de marsupiales sudamericanos. Representa el tipo de una nueva Familia, Patagoniidae, que es asignada a la nueva Superfamilia Patagonioidea, porque representa un grupo evolutivo natural como los de otras Superfamilias de marsupiales reconocidas por Simp- son. Su posicion sistematica dentro del Superorden Marsupialia depende del analisis integrado de aquellos marsupiales enigmaticos (Groeberioidea y Argyrolagoidea) aparentemente mas estrechamente relacionados a el. Novos mamiferos foram recuperados dos leitos de Colhuehuapian (do alto Oligoceno), ex- postos na regiao de Gaiman, Patagonia (Provincia de Chubut, Argentina). Fragmentos man- dibulares e dentes isolados pertencem a um novo genero e especie, Patagonia peregrina. As especializa9oes mandibulares e dentais encontradas sao tao claramente convergentes as de alguns roedores fossorios, que esta forma e considerada uma classe distinta de marsupiais sulameri- canos. A especie representa o tipo de uma nova familia, Patagoniidae, a qual e designada a uma nova superfamilia, Patagonioidea, por formar um grupo evolutivo bem definido, como o formam as outras familias de marsupiais, reconhecidas por Simpson. A posi9ao sistematica dos Patagonioidea, dentro da superordem Marsupialia, aguarda uma analise mais compreensiva dos marsupiais ainda enigmaticos (como Groeberioidea e Argyrolagoidea) aparentemente e seus relativos mais proximos. From the Division Paleontologia Vertebrados, Museo de La Plata, Paseo del Bosque, 1900 La Plata, Argentina; and CONICET, Argentina. PASCUAL & CARLINI: NEW SUPERFAMILY OF PALEOGENE MARSUPIALS 99 Introduction The taxon described in this paper is yet another example of the great adaptive radiation and dis- persal of marsupials in South America. It repre- sents a second line of marsupials that is convergent on the rodent adaptive zone (cf Groeberioidea— Patterson, 1952; Simpson, 1970c; Clemens & Marshall, 1976). However, it is distinct from pre- viously named forms, not only phylogenetically but also ecologically. The new form does not suggest that marsupials attained the breadth and diversity of rodent ad- aptations, but it does show that marsupials oc- cupied the rodent adaptive zone in previously un- imagined ways. This new marsupial indicates that marsupial radiations in South America were al- most as broad and reached as great extremes as those in Australia. The find is consistent with the view that "A complete record of South American marsupials would certainly include a large number of taxa, probably some of high categorical rank, now unknown" (Simpson, 1970a, p. 59). This and other forms recently found in northwestern Ar- gentina (Pascual, 1980a, b, 1981, 1983) validate Simpson's prophetic suggestion that ". . . major parts of marsupial evolution were occurring in areas and facies inadequately sampled, if at all, by the known fossil deposits and the collections so far made" (Simpson, 1970a, p. 58). These deposits indicate the value of applying new sample-col- lecting techniques at mammal-bearing localities that are supposedly well known; it is only neces- sary to find new, appropriate facies. The new ecological type from the Paleogene pro- vides evidence to support Gould's (1983) view of "early experimentation, later standardization," with a consequent reduction in diversity. As in therians (Pascual et al., 1985) the diversification of South American marsupials took place princi- pally in the Paleogene. Measurements reported in Table 1 are depicted in Figure 3 and are given in millimeters. The ab- breviation MACN CH is used for the Museo Ar- gentino de Ciencias Naturales "Bernardino Ri- vadavia" (Buenos Aires), Coleccion Chubut. Classification Superfamily PATAGONIOIDEA nov. The only known family of this taxon is the Pat- agoniidae. The superfamily is sufficiently charac- terized by the diagnoses of that family and its only known species. Justification for its superfamilial rank is given in a later section on affinities. Family Patagoniidae nov. Type— Patagonia gen. nov. The only known ge- nus. Known Distribution— Late Oligocene. Col- huehuapian from Central Patagonia (Chubut Province, Argentina). Diagnosis.— Small marsupials with the same reduced number of lower teeth as the Groeberi- idae, but with a different dental formula: 1.1.0.3. Open-rooted and rodent-like lower incisor, oval in cross section, strongly curved, although not as much as in the Groeberiidae, and with the intra- alveolar portion differently arranged. The incisor extends lingually along the ventral border of the mandible to the root of the inflected crest beneath the last molar, where it forms a prominence sim- ilar to that of hystricognathous rodents, but ven- trally. Lower canine smaller, procumbent, appar- ently incisor-like and closed-rooted, separated from the cheekteeth by a short, crested diastema at al- veolar level. Lower cheekteeth rectangular in cross section, decreasing in size posteriorly, hypselo- dont, rootless, wholly surrounded by enamel, and slightly curved, with the concavity forward. Hor- izontal ramus of the mandible short and deep, with the highest part posterior, beneath the masseteric fossa, where the body of the mandible becomes strongly convex and inflected; deep pterygoid fos- sa, limited ventrally by a flange like that found in Argyrolagidae and in some Australian marsupials (e.g., Macropodidae); strong, salient coronoid pro- cess; masseteric fossa relatively deep but reduced, dorsally situated with a prominent masseteric crest; subvertical symphysis unfused, with nearly smooth symphyseal surfaces. PATAGONIA gen. nov. Etymology- From Patagonia, its geographical record. Type— Patagonia peregrina sp. nov. Known Range and Diagnosis— Same as that of the family. Patagonia peregrina sp. nov. Figures 1-3 Etymology— From Latin peregr inns, strange or rare. 100 FIELDIANA: ZOOLOGY «f *^ « Fig. 1 . Patagonia peregrina gen. et sp. nov. A-B, Stereopairs of MACN CH-865, a fragment of a right mandibular ramus with m,.,: A, occlusal view; B, posterior view; C-D, X-ray of fragments of two right mandibular rami with i,, alveolus of c,, and m,., complete (C, holotype; MACN CH-869) and with alveoli of i,, and c,, and m,., complete (D, MACN CH-865). Graphic scale = 2 mm. HoLOTYPE-MACN CH-869 (fig. 2A-B). Frag- ment of right mandibular ramus with three cheek- teeth, intra-alveolar portion of the incisor, and alveolus of the canine. Hypodigm — Holotype and the following: MACN CH-864, part of right mandibular ramus with first and second cheekteeth, part of alveolus of the third, and part of alveoli of incisor and canine; MACN CH-865, part of right ramus with three cheekteeth and alveoli of the incisor and PASCUAL & CARLINI: NEW SUPERFAMILY OF PALEOGENE MARSUPIALS 101 B 'C, alveolus alveolus m 1 m 2 m alveolus Fig. 2. Patagonia peregrina gen. et sp. nov. A-B, Stereopairs of MACN CH-869 (holotype), a fragment of a right mandibular ramus, with i,, alveolus of c,, and m,.,: A, labial view; B, lingual view. Graphic scale = 2 mm. C, Stereopairs of MACN CH-867, a fragment of a left mandibular ramus, with alveoh of i,, c, and mj, and m,_2; occlusal view. canine; MACN CH-866, part of left ramus with the second and third cheekteeth, and part of al- veoli of the first cheektooth and the incisor, MACN CH-867, part of the left ramus with the first and second cheekteeth, and part of alveoli of the third cheektooth, incisor, and canine; MACN CH-868, part of the right ramus with three cheekteeth and alveolus of the incisor; MACN CH-870, part of left ramus with first and second cheekteeth and part of the alveolus of the third; MACN CH-874, part of right ramus with the second cheektooth, alveoli of the first and third cheekteeth, and part of the alveolus of the incisor; MACN CH-875, part of right ramus with the second and third 102 HELDIANA: ZOOLOGY cheekteeth and part of the alveolus of the incisor; and MACN CH-876, three isolated upper(?) cheekteeth. Horizon and Locality— Both the holotype and the hypodigm come from the Trelew Member of the Sarmiento Formation (see Mendia & Bayar- sky, 1981) and are Colhuehuapian (Late Oligo- cene) in age. Apparently they were found in the upper unit, exposed on the south side of the Chu- but River valley, Chubut Province, Argentina (Central Patagonia; see Fleagle & Bown, 1983, pp. 242-244). Quite probably this corresponds to Simpson's "stratum F of Fig. 1," which is part of his "Trelew beds" (= "Trelewense"). The material was recovered by O. E. Donadio, M. Soria, J. G. Fleagle, and T. M. Bown (see Fleagle & Bown, 1 983) through dry-screening local deflation lag de- posits. Diagnosis— The only known species of the fam- ily. Description— Dentition— See Figures 1 A,C-D; 2-3. Each side of the lower jaw has one fully ro- dent-like gnawing incisor, only incompletely pre- served in the holotype; it is posteriorly bordered by a relatively shallow and conical alveolus (the tooth being absent in all specimens at hand) sep- arated from the medial one by bone and set at a relatively oblique angle (figs. IC-D; 2C). Homol- ogies of these teeth are uncertain, but the rodent- like medial tooth is surely an incisor, designated for description as i,. The shape and disposition of the second alveolus agrees with the procumbent canine of Polydolopidae (Epidolopinae; cf Paula Couto, 1952, 1961; Pascual & Bond, 1981) and Prepidolopidae(Pascual, 1980b, fig. 2D-E); it thus appears that this tooth is c,. This alveolus is fol- lowed by a short diastema at alveolar level, then three cheekteeth, all rectangular in cross section (with some differences among them) and in close approximation, forming a molariform series. They are surrounded by enamel on all sides and are not strictly lobate, nor are the trigonid, talonid, or original cusps clearly indicated, as occlusion with the uppers was mediated through practically flat areas. The dentine forms a shallow basin sur- rounded by the highest enamel layer, which is slightly higher on the lingual side. There is a slight- ly deeper anteroposterior wear groove, extending from the anterolabial comer to the posterolingual one (fig. lA). Grinding involved a longer propal- inal movement and a shorter ectental stroke. The homologies of these teeth with the more numerous ancestral series cannot be determined. Plausibly B /r- ^[ii^dEJE).--]^ Fig. 3. Patagonia peregrina gen. et sp. nov. Outline of a right mandibular ramus fragment, with alveoli of i, and c,, and m,., complete (MACN CH-865), showing the measurements of Table 1 . A, Labial view; B, occlusal view; C, cheekteeth series (m ,.3); D, lingual view. Graph- ic scale = 2 mm. they are homologous with those typically desig- nated m,_3 in marsupials and are so designated here, yielding the lower dental formula 1.1.0.3. which is provisionally homologized as i,, c,, m,.,. However, many specialized marsupials from the South American fossil record show tendencies (1) PASCUAL &. CARLINI: NEW SUPERFAMILY OF PALEOGENE MARSUPIALS 103 Table 1 . Dimensions of specimens of Patagonia peregrina gen. et sp. nov. (see fig. 3 for measurement references). Dimensions Specimen MACN CH-864 MACN CH-865 MACN CH-866 MACN CH-867 MACN CH-868 MACN CH-869 MACN CH-870 MACN CH-874 MACN CH-875 3.68 4.08 1.40 1.12 4.08 3.60 1.00 1.00 0.96 1.40 1.20 1.08 1.08 1.12 0.96 5.40 1.00 1.08 1.00 0.92 1.20 1.00 0.96 0.96 1.00 1.00 1.04 1.00 0.92 1.36 1.08 1.00 0.96 0.96 0.92 5.12 1.40 1.00 1.08 1.00 1.00 0.96 4.80 1.28 1.16 1.24 1.08 6.76 2.80 3.00 4.60 2.20 4.40 6.20 1.80 1.60 1.80 2.40 to elongate or modify either the pj (Polydolopidae and Parabderitini caenolestids; see Marshall, 1980) or the m, (Abderitinae and Palaeothentinae caen- olestids), and (2) to reduce (e.g., Caenolestidae) or lose (Polydolopidae) the m4. All teeth except the c, are completely hypselodont and rootless. The cheekteeth are slightly curved, with the concavity forward (fig. IC-D). Incisor ("/J— Incompletely preserved in the ho- lotype (fig. 2A-B). The anterior end is broken, but the posterior end is unaltered, showing an open pulp cavity with no sign of root formation. It is as elongate and curved as in the most specialized caviomorphs (e.g., Ctenomyidae), although not upcurved posteriorly. The extra-alveolar part would have been nearly vertical, although not so much as in the Groeberiidae (cf. fig. 4B,D). The tooth extends along the ventral border lingually, first beneath the m, and then lingually to far below the mj, terminating where the ventral border be- comes an inflected flange (MACN CH-865, fig. 1 B) reminiscent of kangaroos; the base of the alveolus shapes a superficial prominence similar to that of hystricognathous rodents, although enveloped by the inflected ventral border. It is approximately oval in cross section, with the long axis oriented dorsoventrally and with a flatter medial surface. Apparently enamel covers the entire tooth, but a noticeably heavier layer extends as a ventral band. Canine fie J— None of the specimens in the hy- podigm include this tooth, but its alveolus (figs. 1 C-D; 2C) is oblique, tapered, and relatively shal- low, indicating a closed-rooted tooth; its oblique orientation suggests that the occlusal apex was ap- pressed against the occlusal tip of i,. First Molar (m,)—The first molar is separated from the c, by a crested diastema (fig. 1 A) as long as m,. It is the largest cheektooth, almost rect- angular in cross section, with the longer lateral faces slightly concave; the lingual face sometimes bears a very shallow groove along the intra-alveo- lar portion. The anterior face is convex, occasion- ally somewhat pointed; the posterior face is almost flat, forming angles with the lateral faces slightly greater than 90°. Second Molar (^wj— The second molar is irreg- ularly quadrate, with the lateral faces slightly con- vex and the anterior and posterior ones flatter. The posterolingual angle is less than 90°, whereas the others are almost 90°. Its width is similar to that of m, (fig. lA). Third Molar (m^J—The third molar is the small- est cheektooth, being subtriangular in cross section rather than square. The anterior face is slightly convex, lingually flatter, and labially more strongly curved; the labial face converges posterolingually with the lingual face, forming a rounded pillar rather than a well-defined posterior face (fig. lA). Mandible— No nearly complete mandible is known, but parts of the horizontal ramus and base of the coronoid process are known. These parts indicate the mandible is extremely short and deep, like that of Groeberia minoprioi, although very different in other respects (cf. fig. 4B,D). The sym- physis is subvertical and unfused, with a nearly smooth symphyseal surface (i.e., normal in struc- ture instead of fused and forming the odd medial posterior projection peculiar to Groeberiidae). The depth of the mandible increases abruptly toward the mj. The deep masseteric fossa appears to be F)eculiarly confined to a dorsal position, as the masseteric crest is situated at a level between the alveolar rim and the lowest level of the rounded and inflected ventral border (figs. 1 B; 2 A). A sim- ilar condition is found in some Abderitinae caen- olestids (e.g., Parabderites bicrispatus; Marshall, 1976, fig. 8a), although in P. bicrispatus the man- dibular body is not as deep and the alveolar border 104 FIELDIANA: ZOOLOGY not as extensively inflected. The coronoid process has its root beneath the m,, forming a strong, sa- lient lamina (known only by its root), so that a conspicuous diagonal valley is formed between the coronoid and the alveolar border behind m, (fig. lA); a similar structure is present in Groeberia minoprioi (see Patterson, 1952, p. 41); the valley is open labially and lingually limited by a prom- inence similar to that present in Australasian Po- toroinae. In many respects this strong, salient, ascending ramus and correlated features are reminiscent of highly fossorial caviomorphs, such as burrow-in- habiting Ctenomyidae. Although the mandibular angle is not preserved in any of the specimens, it probably was inflected, as suggested by the inflec- tion of the ventral border, beginning at the level of mj, which defines a lingual flanged crest (figs. IB; 2B) similar to that producing the extremely inflected angle in the Macropodidae. This lingual ventral flanged crest seems to be the lingual border of an expanded and relatively deep pterygoid fos- sa, resembling that of argyrolagids (see Simpson, 1 970a). There is a relatively large alveolar foramen within the pterygoid fossa, level with the alveolar border and within a pit (fig. 3D), and a mental foramen beneath the anterior face of m, at the level of the alveolus of i, (fig. 2 A). Affinities As in the case of Groeberia (see Simpson, 1 970c), the conclusion that Patagonia is a marsupial rests on a combination of definite, negative, and indi- rect evidence. The most definite evidence for its being a marsupial is the inflected ventral border of the mandible and probably the related inflected angle. This evidence alone is inconclusive, as a few marsupials lack an inflected angle and a few placentals have one. However, no known placental has such an extended and upturned flange-shaped inflection, and even in marsupials it is rarely so well developed (e.g., Groeberiidae [Patterson, 1952]. Argyrolagidae [Simpson, 1970a,b], and the Australasian Macropodidae). Unlike Groeberia, Patagonia has other characters supporting its mar- supial affinities, namely the lower procumbent in- cisor-like canine. In the Epidolopinae (Pascual &. Bond, 1981) there are three procumbent lower teeth, the third being unquestionably the canine. Within the more advanced Polydolopidae (Poly- dolopinae), there are one or three procumbent lower teeth; in the latter case, evidence suggests they consist of two incisors and a canine, the me- dial incisor being quite reduced and the canine well developed, single, and closed-rooted. As in Groeberia the negative evidence is that Patagonia has no features precluding its reference to the Marsupialia. It does exhibit characters mak- ing reference to any Eutheria highly improbable. Its habitus is rodent-like, but its two differentially procumbent lower teeth rule out reference to the Rodentia. While the incisor is rodent-like in shape, it is oriented differently than that in rodents, ex- tending along the ventral border of the horizontal ramus, first below the m,, then lingually to other molars, without curving upward. It apparently shapes the ventral border of the mandible. In ad- dition, the short diastema extends at the level of alveoli. Among known rodents, only Paramyidae and Ischyromyidae developed diastemas at the al- veolar level, but even in these groups, the incisor extends as in other rodents, not as in Patagonia. A more-or-less rodent-like habitus was also char- acteristic of some notoungulates, especially among Typotheria and Hegetotheria, but insofar as known not so extreme in development as in Patagonia. Neither the enlargement of the incisor nor the re- duction of the cheekteeth is known in any prim- itive Paleocene notoungulates or in other South American "ungulates." Even later rodent-like no- toungulates were much less specialized than the Oligocene Patagonia. South American marsupials diverged very early into unique evolutionary lin- eages (see Simpson, 1970a-c, 1971, 1 980; Pascual, 1980a,b, 1981; Paula Couto, 1979; Reig, 1981). Patagonia peregrina is unquestionably a mar- supial because its unique and diagnostic combi- nation of characters are unknown in any eutherian. Nevertheless, it could be regarded as another of the extinct South American mammals considered by some as incertae sedis and by others as a tertium quid with regard to the eutherian-marsupial di- chotomy (McKenna, 1980; Reig, 1981). However, the marsupial affinities of other peculiar fossil mammals from South America remain unques- tioned, despite weaker support than that offered here for Patagonia. For example, the basis for con- sidering the Polydolopidae as marsupials is the combination of an inflected mandibular ramus, palatal vacuities, and a cheektooth formula of 1-3 1-4 P— -- and ^-r~z- These characters were formerly used to exclude the polydolopids from the Allothe- ria. But, as these characters are present in prim- itive therians outside South America, their diag- nosis of marsupials can be considered an "act of PASCUAL & CARLINI: NEW SUPERFAMILY OF PALEOGENE MARSUPIALS 105 Fig. 4. Labial (1) and occlusal (2) outlines of mandibles, showing the diflFerent development of incisor. A, Ar- gyrolagus parodii Rusconi; B, Groeberia minoprioi Patterson; C, Proargyrolagus bolivianus Wolff, D, Patagonia peregrina gen. et sp. nov. Graphic scale = 2 mm. faith based on . . . geography and stratigraphic po- sition rather than on . . . biology" (McKenna, 1 980, pp. 58-59). We beheve that assignment of poly- dolopids to marsupials represents the most par- simonious conclusion. Like the newly described Proargyrolagus boli- vianus (Wolff, 1984), Patagonia peregrina is another peculiar marsupial that appears in the fos- sil record without known ancestors (see Simpson, 1970c, p. 16) only to vanish again soon afterward: Groeberiidae (Divisaderan Age, Late Eocene); Pat- agoniidae (Colhuehuapian Age, Late Oligocene); Necrolesiidae (Santacrucian Age, Early Miocene); Argyrolagidae (Huayquerian to Uquian Ages, Late Miocene to Early Pleistocene). We believe there are cedent reasons to think of Proargyrolagus bo- 106 nELDL\NA: ZOOLOGY livianus Wolff, 1984, described as a Deseadan ar- gyrolagid, as possibly representing a distinct fam- ily of Argyrolagoidea. This raises the question of the position of Pat- agonia among the varied ranks of South American marsupials. The previous descriptions and illus- trations demonstrate that Patagonia peregrina has many peculiarities that are rare, differently devel- oped, or completely absent in other marsupials (cf fig. 4). The most striking of these are: 1 . Mandible extremely short and deep, with un- fused subvertical symphysis, dorsally posi- tioned masseteric fossa, ventral border in- flected at level of the mj, enveloping there the alveolus of the incisor. 2. Presence in each ramus of mandible of one rodent-like rootless incisor that extends lin- gually along ventral border of mandible to below the m,. 3. Presence in each ramus of one procumbent canine, single- and closed-rooted, scarcely separated from the incisor and with the oc- clusal apex probably appressed to the inci- sive apex. 4. Three rectangular and continuously growing cheekteeth arranged in close sequence. These and other less striking characters under- score the unique specializations of Patagonia per- egrina, leading to its assignment to a new family, Patagoniidae. But the distinctive combination of characters in the Patagoniidae identify it as a dis- tinct evolutionary group, that is, a different clado- genetic unit. Simpson (1945, 1970a, 1980) des- ignated natural evolutionary groups of marsupials as suF>erfamilies. Following this line of reasoning, Patagoniidae should be allocated to a new super- family, the Patagonioidea. What are the affinities of this new superfamily to other superfamilies within the superorder Mar- supialia? Any discussion of its affinities depends on the systematics of other taxa, many of which are problematic. The systematics of fossil and ex- tant South American marsupials, including the merits of recognizing Marsupialia as a superorder, are discussed by Simpson (1970a, 1971) and Pas- cual (1980b). The majority of South American marsupials represent the order Polyprotodonta; this is roughly equivalent to Ride's ( 1 964) Marsupicamivora, but also includes Ameghino's Paucituberculata (see Pascual, 1980b; contra Kirsch, 1977a,b; Reig, 1981). There is as yet no compelling argument to include any South American families within the Australasian order Diprotodonta (Reig, 1981), de- spite some suggestions to the contrary (e.g., Pas- cual & Herrera, 1 973, 1 975). While the allocations of these groups seems unambiguous, the positions of most remaining groups (e.g., Argyrolagidae, Necrolestidae, and Groeberiidae) remain uncer- tain. With some reservation, Kirsch (1977b) in- cluded the Necrolestidae in the polyprotodont Borhyaenoidea (as did Patterson, 1958), and the Groeberioidea and Argyrolagoidea within the Paucituberculata. Independently, Clemens and Marshall ( 1 976) also treated these animals as mar- supials, recognizing each as superfamilies: Argy- rolagoidea, Necrolestoidea, and Groeberioidea. Like Simpson, they made these assignments with disclaimers that the interrelationships of these groups were far from clear. Reig (1981, p. 60) not only questioned whether the Argyrolagidae (his Microtragulidae) were mar- supials, as none of its known characters are ty- pologically diagnostic, but conjectured probable affinities to the Anagalida. Further, without rig- orous analysis, he suggested that the Argyrolagidae could be treated as an independent order, pro- posing the name Argyrolagida. He concluded that only more intensive study or additional records could substantiate allocation of this order to the Metatheria or the Eutheria. Remains of Patagoniidae exhibit a unique mo- saic of characters, some of which are absent or differently developed in Groeberiidae and Argy- rolagidae. Despite their similarities, each of these taxa apF)ears prima facie to represent indep)endent evolutionary trends. To assess their interrelation- ships, common and distinctive characters of each must be carefully weighed. Remains of Argyro- lagoidea obtained in the same horizon and locality as the hypodigm of Patagonia peregrina should be particularly useful in this regard and are now under study. Ordinal and subordinal allocation of the Patagonioidea await this more comprehensive analysis. Known representatives of this taxon are so highly derived, as is the case with other peculiar marsupials, that their relationships to other mar- supial groups are obscure and can only be clarified by an expanded record of earlier forms. Ecology and Historical Biogeography Biological inferences of Patagonia are necessar- ily limited to the mandibular fragments thus far PASCUAL & CARLINL NEW SUPERFAMILY OF PALEOGENE MARSUPIALS 107 known. These demonstrate unique characters among marsupials, Hving or extinct, which are ob- viously related to a particular mode of life. No known eutherian possesses such mandibular fea- tures. Superficially it is similar to Groeberia, both being rodent-like marsupials: each has a short and deep mandible with a single enlarged, open-rooted incisor, deeply extended along the mandible, with the extra-alveolar part apparently nearly vertical. These represent functional not phylogenetic sim- ilarities, as similar states were attained by different routes: in Groeberia this tooth extended within an odd medial posterior projection of the symphysis, whereas in Patagonia the intra-alveolar portion is truly rodent-like, in being extended along the hor- izontal ramus (cf fig. 4B,D). No doubt both were powerfial gnawers as the lower incisor worked al- most vertically, much more so than in most ro- dents. The unknown face and snout of Patagonia was probably short and deep; whether it had two pairs of lagomorph-like upper incisors like Groe- beria remains unknown. Related to this gnawing specialization, both Groeberia and Patagonia show a short diastema near the alveolar level and a re- duced number of postincisive teeth, four in both; however, Patagonia has three cheekteeth, whereas Groeberia has four. The rodent-like habitus of Pat- agonia is especially advanced, because the three cheekteeth are truly hypselodont, rectangular- shaped in cross section, with at most only shallow lateral grooves representing the remnants of an- cestral bilobate cheekteeth. This combination of features suggests food was obtained by gnawing and prepared for swallowing by grinding. It represents extraordinary conver- gence on some desert-adapted and fossorial forms, such as the Octodontidae. The evolution of cheek- teeth toward a rectangular shape and numerically reduced sequence has been recognized as occurring within the Octodontoidea (from the Octodontidae to the Ctenomyidae; Pascual et al., 1965). The dental features of Patagonia are also convergent on those of the desert-dwelling African Bathyer- gidae, particularly to the sand rat Heterocephalus glaber, and to the North American Geomyidae. These convergent anatomical features suggest that Patagoniidae were probably fossorial marsupials. Anatomical convergence of Patagonia on des- ert-dwelling fossorial rodents is curious, because prevailing conditions in central Patagonia during the Colhuehuapian Age were not highly favorable to desert dwellers. The first record of platyrrhine monkeys in Patagonia occurs at the same locality and level (Fleagle & Sown, 1983) as Patagonia. Many other vertebrate remains recovered at this site (see Bordas, 1939; Donadio, 1983) suggest an environment of well-watered tropical woodlands. Conversely, however, both Argyrolagoidea and very advanced Cephalomyidae rodents from this site (currently under study) show dental features reminiscent of desert or at least drier environ- ments. Generally, the Colhuehuapian vertebrate fauna from central Patagonia (see Pascual, 1970; Pascual & Odreman Rivas, 1971; Marshall et al., 1983) is composed of both forest and open-coun- try types, presumably brought together in a sub- tropical savanna. Thus, the Patagoniidae, Ceph- alomyidae, and Argyrolagoidea occurred in apparently inappropriate environments, probably restricted to xeric patches in the subtropical sa- vanna mosaic. Because the Colhuehuapian Pata- gonioidea were already highly specialized for xeric habitats, they probably evolved earlier in the Pa- leogene. It therefore seems likely that ancestral forms existed in the Deseadan (Early Oligocene). Another highly specialized group of marsupials, the Argyrolagoidea, suggests this hypothesis. For- merly believed present in the record fi-om the Huayquerian (Late Miocene) to the Uquian (Early Pleistocene; see Marshall et al., 1983), argyrola- goids have now been reported from the Deseadan of Bolivia (Wolff, 1984), and here from the Col- huehuapian beds of central Patagonia. The pre-Deseadan record contains no potential ancestor for either Argyrolagoidea or Patagonioi- dea. Simpson (1970c, p. 17) proposed that "these groups (including Groeberioidea) evolved in what are now (and quite likely were then) the tropics and are picked up in our record only when they spread rather briefly to what was for them a mar- ginal area." It seems quite probable that the en- vironments responsible for their initial divergence were poorly or not represented in the known fossil record. Global diastrophic movements in the Late Eocene, and apparently related climatic and en- vironmental changes, are thought to be responsi- ble for the cosmopolitan turnover in Early Oh- gocene mammal communities (Kurten, 1971). This turnover also occurred in South America (Pascual, 1984). Mammal communities in the Deseadan (Early Oligocene) are substantially different from Eocene communities in composition (see Pascual et al., 1 985), apparently reflecting Stehlin's '^^grande coupure."" The apparently sudden occurrence of the Argyrolagoidea. and probably the Patagonioi- dea, in the Deseadan Age is probably another ex- ample of this global turnover. 108 nELDL\NA: ZOOLOGY It is remarkable that, to the numerous succes- sive parallel trends ("successive trends" or "iter- ation"; Simpson, 1953, pp. 248-259; 1961, p. 127) in the evolutionary history of South American mammals, especially from the Deseadan on, can be added the convergence of Oligocene patagoniid marsupials and Pliocene to Recent ctenomyid ro- dents on a common morphology. These conver- gences are products of similar responses to re- peated environmental conditions. The anatomical and functional similarities of Patagonia peregrina with the extant Ctenomys are so striking that we are tempted to call the former the "marsupial tuco- tuco." Acknowledgments All of the material studied here was discovered by 1983 and 1984 expeditions of the Museo Ar- gentino de Ciencias Naturales "Bernardino Ri- vadavia" (MACN), in which Lie. Oscar E. Don- adio and Lie. Miguel Soria (both of MACN) and the American paleontologists John G. Fleagle and Thomas M. Bown participated. Dr. Jose F. Bo- naparte, Chief of the Seccion Paleontologia Ver- tebrados, MACN, and responsible for these ex- peditions, generously put this and other marsupial material at our disposal. The x-ray plates were made by Dr. Roberto Guevara, Profesor de Odon- tologia, Universidad Nacional de La Plata, by the authority of the Dean, Dr. Oscar Barletta. We thank all of them very much. Literature Cited BoRDAS, A. F. 1939. Diagnosis sobre algunos mami- feros de las capas con Colpodon del valle del Rio Chu- but. Physis, 14:413-433. Clemens, W. A., and L. G. Marshall. 1976. Amer- ican and European Marsupialia. Pars 123. Marsupi- alia. In Westphal, F., ed., Fossilium Catalogus. I: An- imalia. W. Junk, The Hague, 1 14 pp. DoNADio, O. E. 1983. Los lacertilios del Colhuehua- pense de la provincia del Chubut, Argentina. Circular Informativa de la Asociacion Paleontologica Argen- tina, 11: 5-6. Fleagle, J. G., and T. M. Bown. 1983. New Primate fossil from Late Oligocene (Colhuehuapian) localities of Chubut Province, Argentina. Folia Primatologica, 41: 240-266. Gould, S. J. 1 983. Nature's Great Era of experiments. Natural History, 7: 12-21. KiRSCH, J. A. W. 1977a. The classification of marsu- pials, pp. 1-48. In Hunsaker II, D., The Biology of Marsupials. Academic Press, New York, San Francis- co, London, 537 pp. . 1977b. The comparative serology of Marsu- pialia, and a classification of marsupials. Australian Journal of Zoology, Supplementary Series, 52: 1-152. KuRTfeN, B. 1971. The Age of Mammals. Columbia University Press, New York, 250 pp. Marshall, L. G. 1976. Revision of the South Amer- ican fossil marsupial subfamily Abderitinae (Mam- malia, Caenolestidae). Publicaciones Museo Munici- pal de Ciencias Naturales de Mar del Plata "Lorenzo Scaglia," 2(3): 57-90. . 1980. Systematicsofthe South American mar- supial family Caenolestidae. Fieldiana: Geology, n.s., 5: i-vii, 1-145. Marshall, L. G., R. Hoffstetter, and R. Pascual. 1983. Mammals and stratigraphy: Geochronology of the continental mammal-bearing Tertiary of South America. Palaeovertebrata, Memoire Extraordinaire, Montpellier, 1983: 1-93. McKenna, M. C. 1980. Early history and biogeogra- phy of South America's extinct land mammals. In Ciochon, R. L., and A. B. Chiarelli, eds.. Evolutionary Biology of the New World Monkeys and Continental Drift. Plenum Publishing Corp., New York, xvi + 528 pp. Mendia, J. E., AND A. Bayarsky. 1981. Estratigrafia del Terciario en el valle inferior del rio Chubut. VII Congreso Geologico Argentino, San Luis (20-26 Sep- tiembre 1981), Actas 3: 593-606. Pascual, R. 1970. Evolucion de comunidades, cam- bios faunisticos e integraciones biocenoticas de los vertebrados cenozoicos de Argentina. Actas IV Con- gresso Latinoamericano de Zoologia (Caracas, 10-16 de Noviembre de 1968), 2 (Paleontologia Sudameri- cana): 991-1088. . 1980a. Nuevos singulares tipos ecologicos de marsupiales extinguidos de America del Sur (Paieo- ceno tardio o Eocene temprano) del Noroeste Argen- tino. Actas II Congreso Argentino de Paleontologia y Bioestratigrafia y I Congreso Latinoamericano de Pa- leontologia (Buenos Aires, 2-6 de Abril 1 978), 2: 151- 173. 1 980b. Prepidolopidae, nueva familia de Mar- supialia Didelphoidea del Eoceno sudamericano. Ameghiniana, 17(3): 216-242. . 1981. Adiciones al conocimiento de Bona- partherium hinakusijum (Marsupialia, Bonapartheri- idae) del Eoceno temprano del Noroeste Argentino. Anais II Congresso Latino-Americano de Paleonto- logia (26 a 30 de Abril, 1981, Porto Alegre-Brasil), 2: 507-520. . 1983. Novedosos marsupiales paleogenos de la Formacion Pozuelos (Grupo Pastes Grandes) de la Puna, Salta, Argentina. Ameghiniana, 20(3-4): 265- 280. . 1984. La sucesion de las Edades-mamifero, de los climas y del diastrofismo sudamericanos durante el Cenozoico: fenomenos concurrentes. Anales de la Academia Nacional de Ciencias Exactas, Fisicas y Naturales, 36: 1 5-37. PASCUAL 8c CARLINI: NEW SUPERFAMILY OF PALEOGENE MARSUPIALS 109 Pascual, R., and M. Bond. 1981. Epidolopinae subfam. nov. de los Polydolopidae (Marsupialia, Po- lydolopoidea). Anais II Congresso Latino- Americano de Paleontologia (26 a 30 de Abril, 1981, Porto Alegre- Brasil), 2: 479-488. Pascual, R., AND H. E. Herrera. 1973. Adiciones al conocimiento de Pliolestes tripotamicus Reig, 1955 (Mammalia, Marsupialia, Caenolestidae) del Plioceno superior de la Argentina. Ameghiniana, 10(1): 36-50. . 1975. Stilotherium Ameghino, 1887, el mas primitivo Caenolestidae conocido. Consideraciones sobre la transicion Didelphidae-Caenolestinae (Mar- supialia). Actas I Congreso Argentine de Paleontologia y Bioestratigrafia (12-16 Agosto, 1 974, Tucuman, Ar- gentina), 2: 417-430. Pascual, R., and O. E. Odreman Rivas. 1971. Ev- olucion de las comunidades de los vertebrados del Terciario argentine. Los aspectos paleozoogeograficos y paleoclimaticos relacionados. Ameghiniana, 8(3-4): 372-412. Pascual, R., J. PiSANO, and E. J. Ortega. 1965. Un nuevo Octodontidae (Rodentia, Caviomorpha) de la Formacion Epecuen (Plioceno medio) de Hidalgo (Provincia de La Pampa). Ameghiniana, 4(1): 19-30. Pascual, R., M. G. Vucetich, G. J. Scillato-Yane, AND M.Bond. 1985. Main pathways of mammalian diversification in South America, pp. 219-247. In Stehli, F., and S. D. Webb, eds.. The Great American Biotic Interchange. Plenum Publishing Corp., New York. Patterson, B. 1952. Un nuevo y extraordinario mar- supial deseadiano. Revista del Museo Municipal de Ciencias Naturales, Mar del Plata, 1: 39-44. . 1958. Affinities of the Patagonian fossil mam- mal Necrolestes. Breviora (Museum of Comparative Zoology), 49: 1-14. Paula Couto, C. de. 1952. Fossil mammals from the beginning of the Cenozoic in Brasil. Marsupialia: Poly- dolopidae and Borhyaenidae. American Museum Novitates, 1559: 1-27. . 1961. Marsupiais fosseis do Paleoceno do Bra- sil. Anais Academia Brasileira Ciencias, Rio de Ja- neiro, 33(3/4): 321-333. . 1979. Tratado de Paleomastozoologia. Aca- demia Brasileira de Ciencias, Rio de Janeiro, 590 pp. Reig, O. A. 1981. Teoria del origen y desarrollo de la fauna de mamiferos de America del Sur. Monogra- phiae Naturae (Publicadas por el Museo Municipal de Ciencias Naturales "Lorenzo Scaglia"), Mar del Plata, 1: 1-162. Ride, W. D. L. 1964. A review of Australian fossil marsupials. Journal of the Royal Society of Western Australia, 47(4): 91-131. Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bulletin of the Amer- ican Museum of Natural History, 85: i-xxvi, 1-350. . 1953. The Major Features of Evolution. Co- lumbia University Press, New York, 248 pp. 1 96 1 . Life of the Past. Yale Paper Bound, Yale University Press, New Haven, Conn. 1970a. The Argyrolagidae, extinct South American marsupials. Bulletin of the Museum of Comparative Zoology, 139: 1-86. 1970b. Additions to knowledge of the Argy- rolagidae (Mammalia, Marsupialia) from the Late Ce- nozoic of Argentina. Breviora (Museum of Compar- ative Zoology), 361: 1-9. 1970c. Addition to knowledge of Groeberia (Mammalia, Marsupialia) from the Mid-Cenozoic of Argentina. Breviora (Museum of Comparative Zool- ogy), 362: 1-17. 1971. The evolution of marsupials in South America. Anais Academia Brasileira de Ciencias, Su- plemento, 43: 103-118. 1 980. Splendid Isolation. The Curious History of South American Mammals. Yale University Press, New Haven and London, IX + 266 pp. Wolff, R. G. 1984. A new Early Oligocene Argyro- lagid (Mammalia: Marsupialia) from Salla, Bolivia. Journal of Vertebrate Paleontology, 4(1): 108-1 13. 110 FIELDIANA: ZOOLOGY An Additional 14-Chromosome Karyotype and Sex-Chromosome Mosaicism in South American Marsupials Milton H. Gallardo and Bruce D. Patterson ABSTRACTS The karyotype of Rhyncholestes Osgood is described for the first time. The karyotype has 2n = 14 and is similar in most respects to karyotypes of similar number found in other American and Australasian genera in several families. The karyotype of somatic (bone marrow) tissues from male Dromiciops Thomas is presented for the first time; surprisingly, it differs from the 2n = 14 complement previously reported from female bone marrow and male gonads. The 2n = 1 3 karyotype found in bone marrow of male Dromiciops lacks a minute element thought to be the Y chromosome. This instance of somatic chromosome elimination represents the first case reported for American marsupials and presents an interesting parallel to sex-chromosome mosaicism among Australasian Peramelidae and Petauridae. El cariotipo de Rhyncholestes Osgood es descrito por primera vez. El cariotipo consta de 2n = 14 y es muy similar a cariotipos de igual mimero encontrados en otros generos americanos y australoasiaticos de varias familias. El cariotipo de tejidos somaticos (medula osea) de un Dromiciops Thomas macho es presentado por primera vez; sorprendentemente, difiere del complemento 2n = 14 reportado previamente de medula osea femenina y gonadas masculinas. El cariotipo 2n = 1 3 encontrado en medula osea del Dromiciops macho carece de un diminuto elemento que supuestamente corresponde al cromosoma Y. Este ejemplo de eliminacion so- matica de cromosomas representa el primer caso reportado en marsupiales americanos y pre- senta un interesante paralelo con el extenso mosaicismo de los cromosomas sexuales descrito entre las formas australoasiaticas. Descreve-se pela primeira vez, o cariotipo de Rhyncholestes Osgood. O cariotipo e de 2n = 1 4, e, na maioria de seus aspectos, assemelha-se aos cariotipos de numeros similares encontrados em outros generos americanos e austral^sios. O cariotipo de tecidos somdticos (da medula ossea) de Dromiciops Thomas machos e descrito pela primeira vez. Supreendentemente, este cari- otipo difere do complemento de 2n = 1 4, previamente descrito para a medula ossea das femeas e para as gonadas dos machos. No cariotipo de 2n = 13, encontrado na medula ossea de Dromiciops machos, falta um elemento miudo, possivelmente o cromossomo Y. Este e o primeiro exemplo documentado da elimina9ao somatica de um cromossomo em marsupiais americanos, e apresenta um paralelo interessante ao mosaico frequentemente encontrado nos cromossomos sexuais de outras formas austral^sias. From the Institute de Ecologia y Evolucion, Univer- sidad Austral de Chile, Casilla 567, Valdivia, Chile; and Division of Mammals, Field Museum of Natural His- tory, Chicago, IL 60605-2496. GALLARDO & PATTERSON: KARYOTYPES OF MARSUPIALS 1 1 1 Introduction Several unusual cytological features, including low diploid number (Hayman & Martin, 1969; Reig et al., 1977), paternally derived X inactiva- tion (Lyon, 1974a,b), multiple sex-chromosome systems (Hayman & Martin, 1969; Schneider, 1977), somatic elimination of sex-chromosomes (Schneider, 1977; Close, 1984), and sperm con- jugation (Biggers & Creed, 1962; Biggers & De- Lamater, 1965), have made marsupials interesting subjects of cytological research. These studies have clarified fundamental cytological mechanisms. Additionally, results of the research have shed Ught upon directions of chromosomal evolution and upon interrelationships of lineages (Hayman & Martin, 1969; Reig et al., 1977; Sharman, 1982). A 14-chromosome karyotype occurs in several distinct lineages in all living American families: Didelphidae (Reig et al., 1 977), Microbiotheriidae (Spotomo & Fernandez, 1971; Reig et al., 1972), and Caenolestidae (Hayman et al., 1971). This karyotype also occurs in several Australian mar- supial lineages (Hayman & Martin, 1969) and is therefore considered the primitive chromosome number for Metatheria (Reig et al., 1977). Direc- tion of chromosome evolution in Metatheria has proceeded via centromeric dissociations— with pericentric inversions superimposed on the basic Robertsonian mechanism— to give rise to the re- maining 2n = 1 8 and 2n = 22 karyotypes known for American forms (Hayman & Martin, 1969; Reig et al., 1977). Extremes of karyotypic varia- tion in Australasian marsupials are 2n = 10 to 32 (Schneider, 1977). Two autochthonous and endemic South Amer- ican genera, Rhyncholestes and Dromiciops, are especially interesting from an evolutionary view- point. Both are represented by a single species and occur only in the temperate Valdivian rainforests of southern Chile and Argentina. Rhyncholestes. one of three extant genera of Caenolestidae, is widely isolated from its relatives in the northern Andes and presents some striking morphological specializations. Dromiciops. thought by some to have special affinities with Australasian lineages (Sharman, 1982; Szalay, 1982), is the only hving genus of the otherwise extinct Microbiotheriidae (Marshall, 1982). Its affinities with other marsu- pial genera are currently uncertain. In this note we present the first somatic karyotypes of male Rhyn- cholestes raphanurus and Dromiciops australis. Additionally, we document the first instance of somatic sex-chromosome mosaicism in South American marsupials. Materials and Methods Seven specimens of D. australis (five males and two females) from Valdivia (39°32'S, 72°52'W), Osorno (4r06'S, 72°30'W), and Concepcion (37»26'S, 73°19'W) provinces, Chile, were ana- lyzed by the in vivo colchicine-hypotonic citrate technique using bone marrow as a source of mi- toses (Patton, 1967). Modifications of the same procedure were used for the one R. raphanurus collected at La Picada, Volcan Osorno (41*^6'S, 72''30'W); incubation with colchicine lasted 2.5 hours and a slightly more hypotonic solution of sodium citrate was used. A total of 4 1 9 mitotic plates was examined: 29 1 from male and 1 20 from female D. australis and 10 from R. raphanurus. Museimi specimens were deposited in the Collec- tion of Mammals, Instituto de Ecologia y Evolu- cion, Universidad Austral de Chile, and Field Mu- seum of Natural History. Results and Discussion Rhyncholestes raphanurus presents a 2n = 14 complement, consisting of three pairs of large metacentric, one pair of medium-sized metacen- tric, and two pairs of small metacentric autosomes. The sex-chromosomes are an acrocentric X and a minute Y (fig. 1). This karyotype differs morpho- logically from the didelphid 2n = 1 4 in not show- ing a clear break between chromosome groups A and B. It also differs in arm ratios (table 1) from the other living caenolestids, Lestoros and Caeno- lestes (see Hayman et al., 1971). Moreover, the interstitial region of the short arm of pair two shows an achromatic area, resembling a secondary con- striction, not described in other caenolestids (but see discussion in Sharman, 1982). Nevertheless, a 2n = 14 karyotype characterizes all three genera of Caenolestidae, which supports previous claims that this karyotype is primitive for Metatheria (Hayman & Martin, 1969; Hayman et al., 1971; Reig et al., 1 977) and reinforces the pattern of low karyotypic variation within marsupial families. Secondary constrictions can serve as chromo- some markers and are thus useful, in the absence 112 FIELDIANA: ZOOLOGY of banding data, for phylogenetic reconstruction. However, the secondary constriction evident in the karyotype of Rhyncholestes is unreported in other South American marsupials, although sec- ondary constrictions are widespread among Aus- tralasian marsupials (Hayman &. Martin, 1969). Considering commonality and in-group and out- group comparisons, we regard the secondary con- striction of Rhyncholestes as apomorphic. Thus, the similar structures of Australasian marsupials were apparently independently derived and can- not be traced back to some marsupicamivorous or other common ancestor. Chromosome counts from all four male D. aus- tralis consistently indicated 2n = 13 chromo- somes. The diploid number for females was 2n = 14 as was previously reported (Spotomo & Fer- nandez, 1971; Reig et al., 1972). No differences among our specimens from geographically isolated localities were detected, nor were secondary con- strictions evident. Electron microscope studies of sex-chromo- somes in spermatocytes of D. australis and the didelphid Marmosa elegans demonstrate striking similarities (Fernandez et al., 1979). These simi- larities suggest that a 2n = 14 karyotype should be present in D. australis, its Y chromosome should resemble that of M elegans, and both genera should exhibit an XX/XY sex-chromosome system. We have consistently found 2n = 13 chromo- somes in somatic tissues of male Dromiciops and 2n = 14 in female somatic tissue. The missing chromosome in males is dotlike and probably the Y chromosome (fig. 1). Translocation of the Y to an autosome is an unlikely mechanism for the dif- ferences between sexes because males have 2n = 14 in germinal cells and because the sex vesicle appears normal (Fernandez et al., 1979). While it is possible that such a small chromosome might be overlooked in one or a few chromosomal spreads, its universal absence in all counted plates makes this alternative highly unlikely. Available data favor a somatic elimination of the Y chro- mosome. Previous studies have shown that both consti- tutive and facultative heterochromatin can be de- leted from marsupial cells in vivo without appar- ent deleterious effects on cell replication and survival (Hayman & Martin, 1969). Most exam- ples of somatic elimination of sex-chromosomes in marsupials involve the X chromosome in dos- age compensation (e.g., perameUds and petaurids; Close, 1984). Mitotic figures from the testes of B (( a (» A-1 a B-1 II C-1 B-1 A-2 A-3 II •« C-2 XY D& Si! \l A-1 A-2 A-3 U lOjLL »'' A« ^^ c-1 c-2 XX iiii !3 W A-1 A-2 A-3 B-1 Ift 6A c-1 c-2 X Fig. 1 . Karyotypes from bone marrow cells of A, Rhyncholestes raphanurus, male; B, Dromiciops aus- tralis (2n = 1 4), female; C, Dromiciops australis (2n = 13), male. GALLARDO & PATTERSON: KARYOTYPES OF MARSUPIALS 113 Table 1 . Arm ratios Gong arm/short arm) of Rhyn- cholestes autosomes (ratios are based on 1 0 counted plates; sex chromosomes are acrocentric). Pair 1: 1.43 Pair 2: 1.53 Pair 3: 1.39 Pair 4: 1.44 Pair 5: 1.42 Pair 6: 1.46 Dromiciops do show the XY constitution. There- fore, male zygotes begin development as XY, and the Y is retained in the germinal cell line, but is lost in at least some somatic tissues. More studies will be needed to determine the extent of this mo- saicism in other tissues. We believe this instance of sex-chromosome mosaicism probably represents a parallel, inde- pendently derived case from that in Australasian forms. However, it could be used to support Sza- lay's (1982) assertion that Dromiciops is more closely related to Australasian lineages than any other American form, belonging in the Australa- sian cohort Australidelphia. In this regard it is noteworthy that Sharman's (1982) analysis of gross chromosomal morphology suggested that the 2n = 1 4 karyotype of Dromiciops (virtually identical to those of some burramyids, peramelids, and Vom- batus ursinns) might be highly similar to that of the common ancestor of Australian marsupials. This instance of sex-chromosome mosaicism also bears on Archer's (1976) contention that pera- melids, which also exhibit sex-chromosome mo- saicism, appear to be derivatives of didelphids in basicranial anatomy. Banding studies of chro- mosomal morphology in these groups are needed to help resolve these various suggestions. A "ratchet" model for the evolution of the Y chromosome and dosage compensation has been suggested (Charles worth, 1978). Initially an active chromosome, the Y is homologous to the X, but chiasmata formation (and thus recombination be- tween the two) is suppressed (e.g., Ohno, 1967). A gradual accumulation of deleterious mutations could account for its erosion over time, leading to minute size. In the didelphid Monodelphis dimid- iata. the synaptonemal complex is absent in the X-Y pairing region. Structural elements of the complex are present, but their assembly seems in- hibited by the shortness of the Y chromosome. It could be argued that, in Monodelphis and other metatherians with dotiike Y chromosomes, Y function is apparently reduced to sex determina- tion, unnecessary in at least some somatic tissues. We favor the evolutionary erosion of the Y chro- mosome and its lack of function in the bone mar- row tissue of Dromiciops australis as ultimate causes for this sex-chromosome mosaicism. Late replication of highly heterochromatinized DNA, a proximate mechanism for sex -chromosome mo- saicism suggested in dosage compensation (Hay- man &. Martin, 1974), may account for the acci- dental loss of the minute Y chromosome during mitotic divisions of somatic cells of Dromiciops. Acknowledgments We thank Brian K. Lang and Peter L. Meserve for assistance in obtaining specimens at La Picada. We received financial support from the Direccion de Investigacion, Universidad Austral de Chile (S- 83-03), Field Museum of Natural History, Amer- ican Philosophical Society (Johnson Fund # 1 646), and National Geographic Society (#2582-82). Dr. R. Fernandez kindly facilitated study of Dromi- ciops material. The constructive criticisms of J. A. W. Kirsch, P. Myers, and J. L. Patton on an earlier draft are thankfiiUy acknowledged. Literature Cited Archer, M. 1976. The basicranial region of marsupi- camivores (Marsupialia), inter-relationships of car- nivorous marsupials, and affinities of the insectivo- rous marsupial peramelids. Zoological Journal of the Linnean Society, 59: 2 1 7-322. BiGGERS, J. D., AND R. F. S. Creed. 1962. Conjugate spermatozoa of the North American opossum. Nature (London), 196: 1112-1113. BiGGERS, J. D., AND E. D. DeLamater. 1965. Mar- supial spermatozoa pairing in the epididymis of American forms. Nature (London). 208: 402—404. Charlesworth, B. 1978. Model for evolution of Y chromosomes and dosage compensation. Proceedings of the National Academy of Science, USA, 75: 5618- 5622. Close, R. L. 1984. Rates of sex chromosome loss dur- ing development in different tissues of the bandicoots Perameles nasuta and Isoodon macrourus (Marsupi- alia: Peramelidae). Australian Journal of Biological Science, 37: 53-61. Fernandez-D., R., S. Berrios, and J. Pincheira. 1979. Position of the nucleolus within the nuclei of pachy- tene spermatocytes of Dromiciops australis and Mar- mosa elegans (EHdelphoidea-Marsupialia). Experien- tia (Basel), 35: 1021-1023. Havman, D. L., and p. G. Martin. 1969. Cytogenetics 114 FIELDIANA: ZOOLOGY of marsupials, pp. 191-217. In Benirschke, K., ed.. Comparative Mammalian Cytogenetics. Springer, New York. . 1974. Mammalia. I: Monotremata and Mar- supialia, pp. 1-110. In John, B., ed.. Animal Cyto- genetics. Vol. 4: Chordata 4. Gebriider Bomtraeger, Berlin. Hayman, D. L., J. A. W. KiRscH, P. G. Martin, and P.F.Waller. 1971. Chromosomal and serological studies of the Caenolestidae and their implications for marsupial evolution. Nature (London), 231: 194-195. Lyon, M. F. 1974a. Evolution of X-chromosome in- activation in mammals. Nature (London), 250: 651- 653. . 1974b. Mechanisms and evolutionary origins of variable X-chromosome activity in mammals. Pro- ceedings of the Royal Society of London, Series B, 187: 243-268. Marshall, L. G. 1982. Systematics of the South American marsupial family Microbiotheriidae. Field- iana: Geology, n.s., 10: 1-75. Ohno, S. 1967. Sex Chromosomes and Sex-Linked Cells. Springer- Verlag, Berlin, 192 pp. Patton, J. L. 1967. Chromosome studies of certain pocket mice, genus Perognathus (Rodentia: Hetero- myidae). Journal of Mammalogy, 48: ll-'il. Reig, O. a., R. Fernandez, and O. A. Spotorno. 1 972. Further occurrence of a karyotype of 2n = 14 chro- mosomes in two species of Chilean didelphoid mar- supials. Zeitschrift fur Saugetierkunde, 37: 37—42. Reig, O. A., A. L. Gardner, N. O. Bianchi, and J. L. Patton. 1977. The chromosomes of the Didelphi- dae (Marsupialia) and their evolutionary significance. Biological Journal of the Linnean Society, 9: 191-216. Schneider, L. K. 1977. Marsupial chromosomes, cell cycles, and cytogenetics, pp. 51-93. In Hunsaker II, D. D., ed.. The Biology of Marsupials. Academic Press, New York. Sharman, G. B. 1 982. Karyotypic similarities between Dromiciops australis (Microbiotheriidae, Marsupi- alia) and some Australian marsupials, pp. 711-714. In Archer, M., ed.. Carnivorous Marsupials, Vol. IL Royal 2toological Society of New South Wales, Sydney, 804 pp. Spotorno, O. A., and D. R. Fernandez. 1971. The chromosomes of the "monito del monte" Dromiciops australis Philippi. Mammalian Chromosomes News- letter 12(2): 40-41. SzALAY, F. S. 1982. A new appraisal of marsupial phy- logeny and classification, pp. 621-640. In Archer, M., ed.. Carnivorous Marsupials, Vol. II. Royal Zoological Society of New South Wales, Sydney, 804 pp. GALLARDO & PATTERSON: KARYOTYPES OF MARSUPIALS 115 Notes on the Black-Shouldered Opossum, Caluromysiops irrupta Robert J. Izor and Ronald H. Pine ABSTRACTS Caluromysiops is distinct from the three species of Caluromys in external, cranial, dental, skeletal, and phallic characters, although the two genera are certainly more closely related to each other than to any other extant genus. Much uncertainty remains regarding the ecology and distribution of this rare opossum. Caluromysiops es distinto de las tres especies de Caluromys en caracteres extemos, craneales, dentales, esqueletales y falicos, aunque los dos generos son por cierto mas cercanamente rela- cionados entre si que lo es ningun otro genero existente. Todavia hay mucha incertidumbre en relacion a la ecologia y distribucion de este rara raposa. Caluromysiops difere das tres especies de Caluromys en carateres extemos, craniais, dentais, esqueletais e falicos, embora sejam os dois generos claramente mais proximos entre si do que entre qualquer outro genero atualmente existente. A ecologia e a distribuifao desta rara especie continuam muito pouco conhecidas. Introduction The black-shouldered opossum, Caluromysiops irrupta Sanborn, is the rarest of the larger didel- phids. Its history as a subject of scientific study is peculiar, beginning with a very late discovery (195 1); also, many more specimens have been dis- played in zoos (15) than have been collected for museums directly from the wild (2). Despite the paucity of associated data and other shortcomings, zoo animals have been the source of some valuable information during this study. Materials and Methods All zoos known or suspected to have kept Cal- uromysiops were contacted for information on the From the Division of Mammals, Field Museum of Natural History, Chicago, Illinois 60605-2496. Dr. Pine's present address is Illinois Mathematics and Science Academy, Aurora, Illinois 60506-1039. acquisition, history in captivity, and eventual dis- position of animals. All known specimens pre- served in collections were examined by one or both of us, and all tag data were recorded. Results Small sample sizes have hampered previous work on this species, and have affected this study to some extent. Most published information is based on single specimens. Some of the characters described by Sanborn (1951) as diagnostic are in- dividually or ontogenetically variable and not re- liable for identification in all cases. For example, the extent of the hair on the dor- sum of the tail is distinctive in Caluromysiops, although not as extreme as originally described. On immature animals such as the holotype, the furred area reaches nearly to the end of the tail. Adults, however, lack fur on the distal 1 5-20 mm. IZOR & PINE: CALUROMYSIOPS IRRUPTA 117 A^ ■ -i; Fig. 1. Ventral and lateral views of the cranium and lateral view of the mandible of adult male Caluromysiops irrupta. fmnh 60698. Certain incisors have fallen out and have been lost. 118 FIELDIANA: ZOOLOGY Fig. 2. Adult female Caluromysiops irrupta with two young (one is specimen no. cvg M-17 BE 173). Photo courtesy of Edward T. Maruska and the Cincinnati Zoo. The portion of the tail covered dorsally with fur is still much more extensive than even in Calu- romys lanatus, in which only the proximal 50%- 70% is covered. Except perhaps for some Glironia, Caluromysiops is unique among didelphids in that the fur extends onto a distal unpigmented portion of the tail. The distal one-quarter to one-third of the tail fur is also white. In other genera of didel- phids, individuals with some distal portion of the tail skin unpigmented have fur of the tail confined to the proximal pigmented area of the tail skin. The most striking external feature of Caluro- mysiops is probably the pair of dark lateral and dorsal stripes. These typically arise on the back of the hand and run up the inner side of the forelimb onto the shoulder, where they reach their greatest width of 15-30 mm. They approach each other middorsally but usually do not merge, and run in narrowing parallel bands to the rump. In one old individual, cvg M-30 BE 95, which had been dis- played for six years and eight months at the Cin- cinnati Zoo, the pattern is obscured by a general grizzling. A common variant of the pattern has the back of the hand white, with the dark stripe beginning as a sharply delineated black band around the wrist. This feature may occur on one or both forefeet. As Sanborn and others have noted, some in- dividuals of the woolly opossums Caluromys der- bianus and C. lanatus have coloration suggesting the characteristic dorsal markings of Caluromy- siops. In the species of Caluromys, there is typi- cally a darker brown or reddish dorsal area which grades into the paler, grayer sides of the body. In some individuals, this darker region is bisected on the back of the head, neck, and shoulders by a middorsal gray streak. The supposed similarity to Caluromysiops, however, is not at all close. The darker dorsal areas in Caluromys are most sepa- rated in the place where in Caluromysiops they are closest to merging. Moreover, the individuals of Caluromys having the gray middorsal stripe IZOR & PINE: CALUROMYSIOPS IRRUPTA 119 Table 1 . Caluromysiops irrupta formerly exhibited in zoos. Sex Date arrived Date died Disposition of remains Acquisition data Bronx 2too (New York 2^ological Society) F 10 Sept. 1962 28 July 1969 Incinerated? F 20 Nov. 1963 26 Dec. 1964 amnh 208101 National Zoo (Smithsonian Institution, Washington, D.C.) 'Iquitos, Ecuador" (presumably Peru) M 4 Nov. 1969 M 31 Mar. 1971 12 Apr. 1971 (Sent to Lincoln Park Zoo, 11 Oct. 1972) usNM 396160 C. Chase, Miami From Oklahoma City Zoo Oklahoma City Zoo F 23 Nov. 1965 M 28 Jan. 1967 F 19 Dec. 1965 20 Oct. 1970 5 Aug. 1967 Discarded? (Sent to National Zoo, 31 Mar. 1971) Discarded? C. Chase, Miami Lincoln Park Zoo, Chicago F 18 Aug. 1972 M 18 Aug. 1972 (pouch young) F 18 Aug. 1972 (pouch young) M 11 Oct. 1972 7 4 Apr. 1973 25 Mar. 1973 16 Sept. 1974 Discarded? FMNH 121522 FMNH 60154 FMNH 60398 ... Brookfield Zoo (Chicago Zoological Society) M 21 Apr. 1970 19 Aug. 1971 FMNH 60698 R. Baudy, Center Hill, Fla. Tarpon Springs Zoo M (1 Aug. 1972 "rec'd in lab") USNM 397626 ... Cincinnati Zoo F 25 July 1965 F 25 July 1965 (young w/F) F 25 July 1965 (young w/F) M 1 July 1967 8 Mar. 1967 6 Nov. 1965 11 Dec. 1967 27 Feb. 1974 Discarded? CVGM-17BE 173 Discarded? cvG M-30 BE 95 Peru, via Animated Shippers, Miami Peru, via Animated Shippers, Miami Peru, via Animated Shippers, Miami Cuxio (= Cuzco?), Peru, via C. Chase, Miami Data from E. Maruska, M. Jones, J. Eisenberg, A. Dittmar, A. Hamer, C. Chase (all in litt.), and Collins (1973). AMNH = American Museum of Natural History, New York; usnm = National Museum of Natural History, Washing- ton, D.C; FMNH = Field Museum of Natural History, Chicago; cvg = personal collection of E. Maruska, Director, Cincinnati Txio. Table 2. Measurements of Caluromysiops irrupta. Sex No. Total Hind length Tail length foot Greatest Medial skull Basal palatal length Condylo- length length (incl. incisive (incl. (incl. Ear incisors) length incisors) incisors) 6 CVG M-30 BE 95 617 333.5 29 63.7 61.0 57.3 30.0 S USNM 397626 630+ 330 + 52 34 64.5 63.4 59.1 32.3 6 USNM 396 1 60 590 340 51 32 63.6 62.5 57.3 30.5 S FMNH 60698 63.4 60.6 56.5 30.5 9 AMNH 208101 570 310 47 37 62.6 60.7 56.4 29.7 120 FIELDIANA: ZOOLOGY also strongly tend to have the palest extremities, whereas Caluromysiops has extremities with broad blackish bands (on the inner side of the forelimbs and outer side of the hind limbs). Other differences in pelage include the Mar- mosa-like eye rings and the median facial stripe of all Caluromys, which are completely lacking in most Caluromysiops and only faintly suggested in a few. There is no feature of the color pattern indicating that Caluromys and Caluromysiops represent simple variants of a single evolutionary trend. Cranially, the extant didelphids present a rather restricted array of morphologies. All have the same dental formula. The skulls differ primarily in size, in the presence and arrangement of palatal vacu- ities, and in details of the masticatory apparatus such as sagittal crests, shape of the zygomata, and the postorbital processes. To our knowledge, a key to the skulls of the genera has never been con- structed. It is not surprising, therefore, that it is difficult to find trenchant cranial characters sup- porting the distinctiveness of Caluromysiops as a genus. In the context of the family's relative uni- formity, this does not necessarily argue against generic distinction. Pine, however, indicated in Honacki et al. (1982) that he prefers to regard Caluromysiops as a subgenus o^ Caluromys, most- ly because of similarity in skull shape. The dentition of Caluromysiops irrupta was de- scribed by Sanborn (1951) as having larger M'-- and m,., than Caluromys. He noted the absence in the holotype of M^, M^*, and m4 and attributed the lack of an M^ to its probable loss in the cleaning of the skull, but did not discuss the absence of the other molars. The holotype is a juvenile and the developing alveolus of the m4 is quite evident, so the tooth is probably unerupted. The larger size of the molars is generally a valid character distin- guishing Caluromysiops from Caluromys. Some individuals oi Caluromysiops may never have had the minute P', which is frequently lost in adults, but otherwise the dental formula conforms to that of the other didelphids. The single root of the usu- ally spicule-like P' differs from the condition in Caluromys, in which the tooth is double rooted, or at least very broad with an incipient division. There is a strong tendency in Caluromys for the small cusps on the labial stylar shelf to be subdi- vided into as many as nine small, low cusps. Cal- uromysiops typically has five such cusps, each being higher and more distinct than in Caluromys. Caluromys and Caluromysiops are united by the apparently derived character (Archer, 1 982) of clo- sure of the maxillary palatal fenestrae. This feature alone is sufficient to distinguish them from all oth- er living New World marsupials, with the possible exception of some Marmosa. Archer apparently erred in attributing such closure to Glironia. Cal- uromysiops is slightly farther along in the process than Caluromys, with only small, round, paired foramina remaining at the maxillopalatal suture. Species of Caluromys have more or less elongate foramina. Several cranial features of Caluromysiops sug- gest adaptations for strong biting forces. The sag- ittal crest in adults is very pronounced, and the zygomatic arches are robust and widely bowed outward. Rostral length is relatively shorter than in Caluromys, and the mandible is deeper, with the ascending ramus broader and more upright. This seems incongruous in view of the description by Janson et al. (1981) of nectarivorous behavior. Zoo animals, however, have readily accepted a varied diet including animal products (Collins, 1973), and the species probably only exploits nec- tar and pollen opportunistically. Cranial asymmetry is prevalent in our sample. About half of the skulls examined had some sort of deviation of the rostral axis relative to that of the braincase, or deflection of the sagittal crest from the midline. Table 2. Continued. Post- Breadth zygo- Depth Inter- post- Post- matic brain- Maxil- Man- orbital orbital orbital Zygo- brain- Length case Length Length lary dibular con- pro- con- matic case longer (incl. of mand. tooth- M'- tooth- striction cesses striction breadth width nasal bullae) mandible ramus row M' row 12.7 22.1 8.2 38.2 23.4 25.1 22.2 45.8 47.2 22.8 9.5 28.3 13.9 21.1 36.9 23.2 24.4 20.6 46.8 48.6 23.2 8.7 28.8 12.4 21.4 7.8 39.2 22.9 25.1 22.1 47.5 23.1 29.4 14.0 20.1 9.5 37.0 23.6 23.9 9.3 11.3 18.3 9.2 38.0 23.3 25.8 22.3 9.1 IZOR & PINE: CALUROMYSIOPS IRRUPTA 121 Postcranial anatomy of the black-shouldered opossum displays some interesting but as yet inexplicable differences from that of woolly opos- sums. The hind limbs of Caluromysiops are rel- atively much shorter than the forelimbs. The fore- arm is especially long. In addition, all of the skeletal elements are more heavily built than in Caluro- mys, with larger articular surfaces. Both genera exhibit a slightly offset articulation of the second metacarpal, which allows the animals to spread the second and third digits and grasp small branch- es between them. This schizodactylous grip, also found in phalangeroids, is useful for slow, delib- erate climbers which may back up along a branch rather than turn around to proceed headfirst. The tail has 30-31 vertebrae, compared to 36-38 in Caluromys, and has well-developed chevron bones throughout its length. Rosenthal ( 1 972, 1 975) noted that a female Cal- uromysiops was received at the Lincoln Park Zoo with pouch young, which 40 days later still lacked markings and body hair. Details of pouch anatomy were not provided. All of the didelphids examined to date have a more or less cleft glans penis. Biggers ( 1 966) noted that Caluromys derbianus differed from other species he examined in the greater extent of the cleft (half the length of the penis), in the contin- uation of medial urethral grooves to the apices, and in the rounded, slightly bulbous ends of the glans. The single available dissected-out specimen of a Caluromysiops penis (fmnh 60698) suffered some postmortem deterioration and may not be completely representative, but still shows clearly a very deeply split glans (ca. 4 cm) with distinctly enlarged, rounded tips. The urethral grooves also seem to extend nearly to the ends. These characters of the genitalia would seem to ally Caluromys and Caluromysiops. However, Caenolestes also has a deeply cleft glans penis (Os- good, 1921), and many Australian marsupials ex- hibit some version of the same phenomenon, so it may represent a shared primitive character. Moreover, a large majority of didelphid species have not been evaluated in this regard, and the significance cannot be properly assessed. Genitalia of mammals lacking bacula generally have been less studied, even though soft tissue structure can be equally informative (Woolley, 1982), and our cursory survey of preserved material indicates considerable undocumented variety. A remarkable feature, poorly preserved on fmnh 60698, but manifest on the protruding penial apex of FMNH 60398, is a dense covering of small (ca. 1 mm), comified, recurved spines. These are dis- tributed primarily on the rather rugose tip and medial sides of the glans, along the urethral groove. | Osgood (1921) described the glans of Caenolestes as rugose proximally and covered distally by small i circular papillae, but Biggers ( 1 966) noted no such structures on Caluromys or other didelphids ex- j amined. The taxonomic affinities of Caluromysiops ir- rupta have been controversial at both the generic and suprageneric levels. Cabrera (1958), Hersh- kovitz (in Marshall, 1982), and Pine (in Honacki et al., 1982) have suggested that its evident rela- tionship to Caluromys might be better expressed by including it in the latter genus. The present authors are divided on the question of whether this change would improve the current arrange- ment. Reig's (1955) assertion that this species belongs in the Microbiotheriidae has received adequate refutation (Segall, 1969; Szalay, 1982). Kirsch's (1977) attempt to subdivide the Didelphidae is undermined by the fact that his subfamily names Caluromyinae and Dactylopsilinae, as proposed, are nomina nuda. Given his uncertainty about the contents of the supposed subfamilies of didel- phids, this fact could spare future workers consid- erable confusion, although the names may have since become available inadvertantly in subse- quent publications. As most zoo animals have changed hands sev- eral times before reaching their final destinations, there is little likelihood of accurate field data ac- companying them. Among dubious origins re- ported for zoo-held Caluromysiops are Sao Paulo, Brazil, and Iquitos, Ecuador (sic). According to J. A. Davis, Jr. (in litt.), the latter animal "was said by the dealer to have been captured in a backyard on the outskirts of Iquitos, Peru"; see also Bridges (1968) and Davis (1965). Another purported lo- cality, Cuxio, Peru, has not been located and may represent a transcription error for Cuzco. There are only three unquestioned locality rec- ords, all from southern Amazonian Peru, as fol- lows: Peru: Depto. Cuzco; Prov. Quispicanchis, Quince Mil (13°16'S, 70°38'W), 680 m, fmnh 68336 (the holotype). Peru: Depto. Madre de Dios; Itahuania (12°47'S, 7 1°1 3'W), skull is fmnh 84426, skin is in the Museo Nacional de Historia Natural "Javier Prado", Lima. Peru: Depto. Madre de Dios; Manu National 122 FIELDIANA: ZOOLOGY Park, Cocha Cashu Biological Station ( 1 1°55'S, 7 1°1 8'W) (Janson et al., 1981; Terborgh et al., 1984; Emmons, 1984). These three localities are within 1 50 km of each other, along the western margin of the Amazon basin, between 400-700 m elevation. The only sympatric species of Caluromys recorded is C la- natus. Simonetta's (1979) report of a Caluromysiops near Leticia, Colombia, is a problem. Although we are unable to locate the original account, it is our opinion that this record is best discounted. The photograph appears to have been staged with a captive specimen, since the species is nocturnal (Collins, 1973; Janson et al., 1981; Terborgh et al., 1984). Leticia is at least 900 km from the three well-documented localities, and one of the mu- seum specimens we examined (usnm 397626) is known to have passed through Leticia from an unknown source en route to a zoo in Florida. Le- ticia is the location of a major animal dealership, and the point of exportation of many Amazonian species to the U.S. The dusky brown color on the crown of the head, which Simonetta suggests may differentiate his Colombian specimen subspecifi- cally, is variable in the material we examined, and is probably of no taxonomic importance. Conclusions Caluromysiops irrupta is a species which has often been erroneously or incompletely character- ized in the scientific literature. There are now enough specimens in collections to allow reason- ably complete treatments of its morphology, al- though it remains an almost complete ecological and behavioral enigma. Acknowledgments The authors thank those individuals and insti- tutions listed in Table 1 for their invaluable as- sistance in compiling these data, and for loans of specimens in their care. Anita McQuaig, Linda E. Pine, Nobuko Etoh Pine, Joyce Shaw, and Mary Reed helped with the manuscript. Joseph A. Davis and the editors and reviewers made many helpful suggestions. Literature Cited Archer, M. 1982. A review of Miocene thylacinids (Thylacinidae, Marsupialia), the phylogenetic position of the Thylacinidae and the problem of apriorisms in character analysis, pp. 445-476. In Archer, M., ed., Carnivorous Marsupials. Royal Society of New South Wales. BiGGERS, J. D. 1 966. Reproduction in male marsupials, pp. 251-280. In Rowlands, I. W.. ed., Symposia of the Zoological Society of London, 15: 1-559. Bridges, W. 1968. The Bronx Zoo Book of Wild An- imals. New York Zoological Society and Golden Press, New York, 8 unnumbered pp. + 304 pp. Cabrera, A. 1958. Catalogo de los mamiferos de America del Sur. I. (Metatheria-Unguiculata-Camiv- ora). Revista del Museo Argentino de Ciencias Na- turales "Bernardino Rivadavia": 2k>ologia (1957), 4: 1-307. Collins, L. R. 1973. Monotremes and Marsupials. Smithsonian Publication 4888, Smithsonian Institu- tion, Washington, D.C., 323 pp. Davis, J. A., Jr. 1965. Agreat year for rarities. Animal Kingdom, 68(5): 130-133. Emmons, L. H. 1 984. Geographic variation in densities and diversities of non-flying mammals in Amazonia. Biotropica, 16: 210-222. HoNACKJ, J. H., K. E. Kjnman, and J. W. Koeppl, eds. 1 982. Mammal Species of the World. Allen Press and Association of Systematics Collections, Lawrence, Kansas, 694 pp. Janson, C, J. Terborgh, and L. H. Emmons. 1981. Non-flying mammals as pollinating agents in the Am- azonian forest. Biotropica, 12(Suppl.): 1-6. KiRSCH, J. A. W. 1977. The comparative serology of Marsupialia, and a classification of marsupials. Aus- tralian Journal of Zoology, supp. sen, 52: 1-152. Marshall, L. G. 1982. Evolution of South American Marsupialia, pp. 251-272. In Mares, M. A., and H. H. Genoways, eds.. Mammalian Biology in South America. Special Publication Series, F*ymatuning Lab- oratory of Ecology, University of Pittsburgh, 6: 1-539. Osgood, W. H. 1921. A monographic study of the American marsupial, Caenolestes. Field Museum of Natural History, Zoological Series, 14: 1-156. Reig, O. A. 1955. Noticiapreliminarsobrelapresencia de microbiotherinos vivientes en la fauna sudameri- cana. Investigaciones Zool6gicas Chilenas, 2: 121-130. Rosenthal, M. A. 1972. Observations on the water opossum or yapok (Chironectes minimus). Proceed- ings 48th Annual Conference American Association of Zoological Parks and Aquariums held in Portland, Oregon, Oct. 1-5, 1972: 95-98. . 1975. Observations on the water opossum or yapok Chironectes minimus in captivity. International Zoo Yearbook, 15: 4-6. Sanborn, C. C 1951. Two new mammals from south- em Peru. Fieldiana: Zoology. 31: 473-477. Segall, W. 1 969. The middle ear region of Dromi- ciops. Acta Anatomica, 72: 489-50 1 . IZOR & PINE: CALUROMYSIOPS IRRUPTA 123 SiMONETTA, A. M. 1979. First record of Ca/Mrow>'5;op5 cies of Cocha Cashu Biological Station, Manu Na- from Colombia. Mammalia, 43: 247-248. tional Park, Peru. Fieldiana: Zoology, n.s., 21: 1-29. SzALAY, P. S. 1982. A new appraisal of marsupial phy- Woolley, P. A. 1982. Phallic morphology of the Aus- logeny and classification, pp. 621-640. //J Archer, M., tralian species of Anlechinus (Dasyuridae: Marsupi- ed., Carnivorous Marsupials. Royal Society of New alia): A new taxonomic tool?, pp. 767-781. /« Archer, South Wales. M., ed.. Carnivorous Marsupials. Royal Society of New Terborgh, J. W., J. W. FiTZPATRiCK, AND L. Emmons. South Walcs. 1984. Annotated checklist of bird and mammal spe- 124 FIELDIANA: ZOOLOGY Feeding Habits of the Opossum {Didelphis marsupialis) in Northern Venezuela Gerardo A. Cordero R. and Ruben A. Nicolas B. ABSTRACTS The food items in the annual diet of the opossum {Didelphis marsupialis) in northern Ven- ezuela are reported by season, sex, and dental age. One hundred eight opossums were sampled in 21 different sites on a monthly basis from March 1983 to March 1984. The number of food items recorded varies seasonally. By volume, animal foods (63.5%) are more important than plant foods (22.9%) throughout the year. Birds (2 1 .5%), mammals (1 5.3%), insects (14.8%), and fruits (12.8%) are the most prominent foods, by volume. Feeding habits of males and females do not differ significantly. However, diets of young and old animals are different. Se seiialan los componentes de la dieta anual del rabopelado {Didelphis marsupialis) en el norte de Venezuela por epoca del ano, sexo y edad. El muestreo se hizo mensualmente colec- tandose 108 animales desde Marzo 1983 a Marzo 1984 de 21 localidades diferentes. El numero de componentes de la dieta varia estacionalmente. En terminos de volumen, los alimentos de origen animal (63.5%) son mas importantes que los de origen vegetal (22.9%) a traves del aiio. Las aves (21.5%), los mamiferos (15.3%), los insectos (14.8%) y las frutas (12.8%) son las alimentos mas sobresalientes, en terminos de volumen. Los habitos alimentarios de los machos y las hembras no difieren significativamente. Sin embargo, las dietas de los animales jovenes y viejos son diferentes. Relata-se os componentes da dieta anual do gamba {Didelphis marsupialis) no norte da Venezuela, por epoca, sexo e idade. Amostras foram coletadas mensalmente de marfo de 1983 a marfo de 1 984, e de 2 1 locais diferentes, para um total de 108 animais examinados. O nvimero dos componentes da dieta varia sasonalmente. Em termos de volume, os alimentos de origem animal (63,5%) sao mais importantes do que os de origem vegetal (22,9%) atraves do ano. Aves (21,5%), mamiferos (15,3%), insetos (14,8%), e frutos (12,8%) foram os alimentos mais abun- dantes por volume. Apesar dos habitos alimentares nao diferirem entre machos e femeas, a dieta dos animais jovens difere da dieta dos adultos. Introduction of at least seven of 70 species are known (Fleming, 1972; Hunsaker, 1977; Atramentowicz, 1982; Feeding habits of neotropical didelphid mar- Streilein, 1982; Charles-Dominique, 1983; cf. supials are poorly known, in spite of their high Kirsch & Calaby, 1977). However, the informa- diversity and broad geographical distribution. Diets tion reported for most species is based on quali- tative data. This paper reports the food items in- gested by opossums {Didelphis marsupialis) in From the Facultad Ciencias Institute de Zoologia northern Venezuela throughout the year by sea- Tropical, Apartado 47058, Caracas 1041 -A, Venezuela. son, sex, and dental age. CORDERO & NICOLAS: FEEDING HABITS OF OPOSSUMS 125 Study Area Fieldwork was conducted mainly in the Barlo- vento region of the State of Miranda and within the city of Caracas and its surroundings in north- em Venezuela (10°00'-l(y30'N, 66°00'-67°00'E). The climate is highly seasonal, with a humid pe- riod of nine months (May-January) and a dry pe- riod of three months (February-April) in Barlov- ento and seven months of rainfall (May- November) and five months of drought (Decem- ber-April) in Caracas. Annual mean temperature of Barlovento is 26° C versus 20.6° C in Caracas and its surroxmdings. Rainfall is 2,053 mm at Bar- lovento and 1,011 mm at Caracas. Elevations sampled range from 40 m to more than 1 ,000 m above sea level. According to the Holdridge Life Zones (Ewel et al., 1976), the vegetation of Bar- lovento is primarily a humid tropical forest, whereas that of Caracas is mostly in premontane humid forest. Materials and Methods The sample of 108 opossums was assembled fi"om March 1 983 to March 1 984, either from road- kills or hunting. Fifty-two animals were taken from nine localities at Barlovento, whereas 56 speci- mens were taken from 1 2 localities in or near Ca- racas. Body measurements, sex, and dental age of each animal were recorded. Age determination was based on tooth eruption and wear (Petrides, 1 949; Tyndale-Biscoe & MacKenzie, 1976), permitting their grouping into seven age classes (Cordero, un- publ. data, see Appendix 1). Stomach contents were analyzed according to Korschgen's (1980) rec- ommendations. Each stomach and its contents were placed in a fine sieve ( 1 -mm diameter mesh screen) and thoroughly washed under running water in order to separate fine from coarse material. After measuring the entire volume of the contents, each item was separated under a dissecting microscope and its volume recorded. A reference collection was used for the identification of insects. Results Opossum Foods and Seasonal Variation Six (5.6%) of the 108 stomachs we examined were empty. Numbers of stomachs with items were: dry season, 1 6(15.7%) and wet season, 86 (84.3%). Data for these 102 stomachs appear in Table 1 and Figure 1 . Percentage of volume and frequency of occurrence are shown for each class of items in Table 1. Considerable seasonal variation exists in the number of food items recorded. During the dry season, the most important food items are mam- mals, birds, and insects. In the wet season, birds are more important by volume than mammals or insects, and fruits seem to be of greater impor- tance. Gastropods are ingested in a higher pro- portion during dry season than wet season. Snakes, toads, and earthworms are consumed only in the latter period. Food of animal origin is more important (63.5% by volume) than plant food (22.9%) in the diet of opossums throughout the year. By volume, birds (21.5%), mammals (15.3%), insects (14.8%), and fi-uits (12.8%) are the principal foods ingested by opossums. In terms of frequency, insects (49.1%), fiiiits( 18.6%), birds (12.7%), and mammals (8.8%) contribute to the annual diet. Domestic cats {Felis catiis) and rats (Rattus rat- tus) were considered as prey items of opossums because no dipteran carrion larvae were observed in stomach contents. However, unidentified mam- malian remains are more important than those of cat and rat by both volume and frequency. Birds ingested by opossums were either chickens (Gallus sp.) or young birds which were more numerous in the wet season; during the dry season, chickens were recorded as carrion. Avian material account- ed for 12.7% of the stomach contents and 21.4% by volunie. Snakes and toads are consumed at low levels in relation to their abundance in study sites, suggesting that these food items are of little im- portance for opossums in northern Venezuela. Insects of at least nine families occurred in 49.0% of the stomachs, with an annual volume of 14.8%. Beetles and grasshoppers accounted for the ma- jority of insects consumed. Slugs (Veronicellidae) were recorded in the rainy season (1.7% by volume), whereas Vulimulidae are important in the dry season (6.3% by volume). Centipedes and earthworms were poorly repre- sented in the stomachs. Fruits such as Psidium guajava and Giiazuma ulmifolia are very important in the diet of opos- sums. By both volume and frequency of occur- rence, ftiiits are more important in the rainy sea- son. Miscellaneous foods such as garbage (paper, plastic bags, felt, thread filaments), particulate ma- 126 FIELDIANA: ZOOLOGY Table 1. Percentages of volume (V) and frequency (F) of food items of opossums in northern Venezuela in 1983 and 1984, by season and for the year. Wet season Dry season Annual Food items % V %F %V %F %V %F Animals 65.95 90.65 51.22 114.40 63.48 97.03 Mammalia 15.42 5.81 14.75 20.19 15.31 8.82 Felis catus 4.14 2.32 3.44 1.96 Rattus rattus 0.41 7.69 0.07 0.98 Mammal remains 11.28 3.49 14.34 12.50 11.80 5.88 Aves 23.52 13.95 11.27 7.69 21.46 12.74 Gallus sp. 6.42 2.32 5.34 1.96 Young birds 10.04 3.49 8.3S 4.90 Bird remains 7.06 8.14 11.27 7.69 7.77 5.88 Reptilia 0.41 1.16 0.34 0.98 Snake remains 0.41 1.16 0.34 0.98 Amphibia 1.97 1.16 1.64 0.98 Bufo sp. 1.97 1.16 1.64 0.98 Insecta 14.82 47.65 14.70 63.45 14.81 49.01 Coleoptera 7.36 23.24 3.69 23.07 6.76 21.56 Passalidae 0.21 1.16 1.20 7.69 0.28 1.96 Scarabaeidae 5.07 11.63 1.54 7.69 4.48 9.80 Coccinelidae 1.54 7.69 0.26 0.98 Curculionidae 0.31 2.32 0.26 1.96 Meloidae 0.31 1.16 0.26 0.98 Carabidae 0.12 1.16 0.10 0.98 Remains 1.34 5.81 1.12 4.90 Orthoptera 5.12 16.28 0.87 7.69 4.41 14.71 Acrididae 5.12 16.28 0.87 7.69 4.41 14.71 Cursores 0.64 2.32 8.09 25.00 1.89 5.88 Blattaria 0.43 1.16 8.09 25.00 1.72 4.90 Phasmida 0.21 1.16 0.17 0.98 Lepidoptera 1.66 4.65 2.05 7.69 1.72 5.88 Larvae 1.66 4.65 2.05 7.69 1.72 5.88 Homoptera 0.04 1.16 0.03 0.98 Cicadidae 0.04 1.16 0.03 0.98 Mollusca 1.71 8.14 6.25 7.69 2.47 5.88 Veronicellidae 1.71 8.14 1.42 4.90 Vulimulidae 6.25 7.69 1.05 0.98 Chilopoda 1.73 5.81 0.15 7.69 1.46 4.90 ■ Annelida 1.20 5.81 1.00 4.90 Lumbricidae 1.20 5.81 1.00 4.90 Carrion 5.17 1.16 4.10 7.69 4.99 3.92 Dendrophidion parcarinatum 5.17 1.16 4.30 0.98 Gallus sp. 4.10 7.69 0.69 2.94 Plants 22.05 43.01 27.26 32.69 22.92 39.22 Fruits 14.19 20.92 6.15 7.69 12.84 18.63 Psidium gtmjava 5.70 13.95 6.15 7.69 5.77 13.73 Guazuma ulmifolia 6.21 2.32 5.17 1.96 Passiflora sp. 1.91 3.49 1.59 1.96 Mangifera sp. 0.37 1.16 0.31 0.98 Grass remains 0.83 2.32 0.69 1.96 Plant remains 7.03 19.77 21.11 25.00 9.39 18.63 Miscellaneous 2.38 19.76 8.71 12.40 1.98 18.62 Paper trace 11.63 trace 9.80 Plastic bags trace 4.65 trace 6.20 trace 4.90 Felt trace 1.16 trace 0.98 Thread filaments trace 2.32 trace 6.20 trace 2.94 Particulate Material 9.62 12.79 21.52 18.75 11.62 11.76 CORDERO &. NICOLAS: FEEDING HABITS OF OPOSSUMS 127 GRASS 0.7 % SNAKES 0.3% CENTIPEDES 1.5% EARTHWORMS 1.0% Fig. 1. Proportionate annual volumes of major groups of items from 102 stomach contents of opposums from northern Venezuela between March 1983 and March 1984. terial, and plant remains comprised 2.0%, 1 1.6%, and 9.4% by volume, respectively. Garbage items were only recorded for those animals collected in or near Caracas. Variation of Food Items by Sex Feeding habits of male and female opossums are compared in Table 2. By volume, males con- sume mainly fruits (22.8%), birds (17.1%), plant remains (15.4%), and insects (14.5%), whereas fe- males consume mammals (31.4%), birds (14.5%), insects (11.6%), and fruits (8.8%). However, by frequency of occurrence, males consume primarily insects (30%), fruits (19.2%), and plant remains (15.6%); females consume insects (28.6%), plant remains (12.2%), mammals (10.2%), and fruits (10.2%). Both comparisons by means of a Mann- Whitney U test indicate no significant differences between the sexes. Table 2. Food items, by sex, in terms of volume (V) and frequency (F). Males Females (N = = 53) (N = 31) Food items % V %F % V %F Mammalia 9.6 3.6 31.4 10.2 Aves 17.1 6.0 14.5 8.1 Reptilia 0.7 1.2 Amphibia 3.3 1.2 Insecta 14.5 30.0 11.6 28.6 Mollusca 2.3 6.0 4.0 6.1 Chilopoda 1.2 3.6 0.08 2.0 Annelida 0.6 2.4 1.8 4.1 Carrion 2.0 2.0 Fruits 22.8 19.2 8.8 10.2 Plant remains 15.4 15.6 7.6 12.2 Miscellaneous 0.7 3.6 5.2 6.1 Particulate material 12.4 7.2 13.0 10.2 N = Sample size. 128 FIELDIANA: ZOOLOGY A Mann-Whitney U test was also used to com- pare volumes of principal food groups (mammals, birds, amphibians, fruits, insects, and plant re- mains) in the diets of male and female opossums; no significant differences were detected. Despite this, the composition of the diet suggests that males are more arboreal than females. However, a / test comparing the capture frequencies of both sexes on the ground in a 26-hectare grid indicates no significant differences {P > 0.05; Cordero, unpubl. data). The grid contained 18x18 National live- traps, with a distance of 30 m between stations and rows, and was run from December 1981 to May 1984. Variation of Food Items by Age Food of opossums by age classes appears in Ta- ble 3. Note that the number of food items increases as animals become older. By volume and fre- quency of occurrence, animals of younger ages (I, II, III, and IV) consume mainly invertebrates, fruits, and plant remains, while older animals (ages V, VI, and VII) take those items plus mammals and birds, which become more important as the animal ages. Diets of opossums were compared by successive ages, that is, II with III, III with IV, and so on, by Mann-Whitney U tests. No significant differences were detected. However, when diets of young and old animals were compared, significant differences were demonstrated. Nothing has been published on age-related diet variation for D. virginiana, D. albiventris, or any other marsupial. Discussion These results provide a preliminary view of the annual diet of Didelphis marsupialis in north- western Venezuela. This study shows that opos- sums, while omnivorous, are more carnivorous and insectivorous than herbivorous or frugivo- rous. However, we accept these patterns guardedly because they may represent methodological arti- facts: (1) most of our specimens (84.3%) were tak- en in the wet season, so that trophic habits during the dry season are imprecisely known; and (2) the rinsing step in processing stomach contents may have inadvertently washed away traces of fruit pulp that might have been studied using other methods. Our results indicate that insects, fruits, birds, and mammals figure prominently in the annual diet. These figures contrast with those reported by oo ■ • ■* • • 1/^ CTn O 00 00 ■ • O • • 00 r^ 00 r~ (^ . • r~ • . so — »o 00 00 Ov 00 00 vO PO rn »0 (N rn • ■^ ^_ ^ r-' — — • vd r~-' o O 00 d — ■ . . . . vo VO ■O ! : : : i~^' 00 iri 1^ ■<)...«-) — Tf f^ ^ r) o\ O m . . . . O ; ; ; ; O : : : : 8 c .2 ••£ „ •c 2 S « Ji = 3 ^ t» o ? o • - cu S cu CORDERO & NICOLAS: FEEDING HABITS OF OPOSSUMS 129 Molins de la Sema and Lorenzo (1 982) in a study of stomach contents of 47 Didelphis marsupialis sampled from February 1981 to May 1982 in the lowlands of Sierra de Perija in the State of Zulia, northwestern Venezuela. In their study, the order of importance of food items, by frequency, is as follows: plant leaves (68.3%), fruits (56.2%), rep- tiles (42.6%), insects (29.2%), amphibians (28.8%), birds (14.3%), mammals (15.1%), mollusks ( 1 2.2%), and seeds ( 1 1 .4%). The effects of seasonal and habitat differences in the two studies may ex- plain these differences, since the main vegetation types of the lowlands of Perija are dry and humid tropical forests, with eight months of rainfall (April-November) and four months of drought (December-March). Other studies have also shown that opossums feed on vertebrates. The volume we recorded for mammalian prey (15.3%) is low in comparison with diets determined for the Virginia opossum {Didelphis virginiana), except for Lay's ( 1 942) 7% value. Hopkins and Forbes (1980) also recorded cats and rats in low frequencies and volumes in the diets of opossums in Oregon. Similarly, do- mestic chickens figured prominently in the diet of our specimens and have been reported as prey or carrion of D. virginiana in New York (Hamilton, 1951, 1958), Missouri (Reynolds, 1945), Iowa (Wiseman & Hendrickson, 1950), Michigan (Taube, 1947), and Kansas (Sandidge, 1953). In contrast, snakes and toads were taken infrequent- ly, paralleling the results of Blumenthal and Kirk- land (1976), who reported traces of amphibians in the diets of Pennsylvania Didelphis, and of Wise- man and Hendrickson (1950), who showed rep- tiles have a frequency of 1% in the diet of Iowa opossums. The importance of insects in the diet of our animals is somewhat lower than that pre- viously reported for opossums in Michigan (30.4%; Gardner, 1982, citing Dearborn, 1932), Missouri (34.2%; Reynolds, 1 945), and Kansas (42.2%; San- didge, 1953). However, the volumes we report are higher than those in literature records for New York (Hamilton, 1951, 1958), Oregon (Hopkins & Forbes, 1980), and Pennsylvania (Blumenthal & Kirkland, 1976). Records for other inverte- brates are also similar to those in existing literature reports (e.g., Taube, 1947; Hamilton, 1951, 1958; Reynolds, 1945; Sandidge, 1953). Our data and literature records indicate that Di- delphis species have similar diets, embracing a wide range of food items. More detailed studies, espe- cially of food-use in relation to availability, will be needed to establish the degree of euryphagy. Acknowledgments This study was partly granted by CONICET Project SI- 1158. We thank J. Ojasti for sugges- tions and review of the manuscript. We greatly appreciate the editorial assistance of B. Patterson. The staff members of the Estacion Experimental Rio Negro, Universidad Simon Rodriguez pro- vided logistical support during fieldwork. L. Du- que and R. Martinez helped us in the identification of snakes and slugs, and E. Pannier provided some stomach contents. To all of them, our thanks. Literature Cited Atramentowicz, M. 1982. Influence du milieu sur I'activite locomotrice et la reproduction de Caluromys philander (L). Revue d'Ecologie Appliquee (Terre Vie), 36: 373-395. Blumenthal, E. M., and G. L. Kirkland. 1976. The biology of the opossum, Didelphis virginiana in south- central Pennsylvania. Proceedings of the Pennsylvania Academy of Science, 50: 81-85. Charles-Dominique, P. 1983. Ecology and social ad- aptations in didelphid marsupials: Comparison with eutherians of similar ecology, pp. 395-422. In Eisen- berg, J. F., and D. G. Kleiman, eds., Advances in the Study of Mammalian Behavior. Special Publication of the American Society of Mammalogy, no. 7. EwEL, J. J., A. Madriz, and J. A. Tosi. 1976. Zonas de vida de Venezuela. Fondo Nacional de Investiga- ciones Agropecuarias, Caracas, 265 pp. Fleming, T. H. 1972. Aspects of the population dy- namics of three species of opossums in the Panama Canal Zone. Journal of Mammalogy, 53: 619-623. Gardner, A. L. 1982. Virginia opossum {Didelphis virginiana), pp. 3-36. In Chapman, J. A., and G. A. Feldhamer, eds., Wild Mammals of North America. Johns Hopkins University Press, Baltimore. Hamilton, W. J. 1951. The food of the op>ossum in New York State. Journal of Wildlife Management, 15: 258-264. . 1958. Life history and economic relations of the opossum {Didelphis marsupialis virginiana) in New York Slate. Cornell University Agricultural Station Memoir, 354: 1-48. Hopkins, D. D., AND R. B. Forbes. 1980. Dietary pat- terns of the Virginia oix)ssum in an urban environ- ment. The Murrelet, 61: 20-30. HuNSAKER, D. 1977. Ecology of New World marsu- pials, pp. 95-156. In Hunsaker II, D., ed.. Academic Press, New York. KiRSCH, J. A. W., andJ. H. Calaby. 1977. The species of living marsupials: An annotated list, pp. 9-26. In Stonehouse, B., and D. Gilmore, eds.. The Biology of Marsupials. The Macmillan Press Ltd., London and Basingstoke. 130 FIELDIANA: ZOOLOGY KoRSCHGEN, L. J. 1980. Procedures for food-habits analyses, pp. 1 13-127. /n Schemnitz, S. D., ed.. Wild- life Management Techniques. The Wildlife Society, Washington, D.C. Lay, D. W. 1 942. Ecology of the opossum in eastern Texas. Journal of Mammalogy, 23: 147-159. MOLINS DE LA SeRNA, M., AND J. LORENZO PrIETO. 1982. Alimentacion del rabipelado {Didelphis marsupialis) de la Sierra de Perija. Acta Cientifica Venezolana, 33: 410. Petrides, G. a. 1949. Sex and age determination in the opossum. Journal of Mammalogy, 30: 364-378. Reynolds, H. C. 1 945. Some aspects of the life history and ecology of the opossum in central Missouri. Jour- nal of Mammalogy, 26: 361-379. Sandidge, L. L. 1953. Food and dens of the opossum {Didelphis virginiana) in northeastern Kansas. Kansas Academy of Science, 56: 97-106. Streilein, K. E. 1982. Ecology of small mammals in the semiarid Brazilian Caatinga. I. Climate and faunal composition. Annals of Carnegie Museum, 51: 79- 107. Taube, C. M. 1947. Food habits of Michigan opos- sums. Journal of Wildlife Management, 11: 97-103. Tyndale-Biscoe, C. H., and R. B. Mackenzie. 1976. Reproduction in Didelphis marsupialis and D. albi- ventris in Colombia. Journal of Mammalogy, 57: 249- 265. Wiseman, G. L., and G. D. Hendrickson. 1 950. Notes on the life history and ecology of the oi>ossum in south- east Iowa. Journal of Mammalogy, 31: 331-337. Appendix 1. supialis. E>ental age classes for Didelphis mar- Tooth Age Age eruption Wear class (months) dP' M' 0 I 3.0-3.5 dP' M^ 0 II 4.5-5.0 dP' M' 0 III 6.2-6.7 P' M' 0 IV 7.9-8.7 P' M* 0 V 10.9-11.7 P' M- P^ M'-2 VI 12.8-14.1 P' M* P' M^ VII > 16.1 Source: G. A. Cordero (unpublished data). CORDERO & NICOLAS: FEEDING HABITS OF OPOSSUMS 131 Notes on Distribution of Some Bats from Southwestern Colombia Michael S. Alberico ABSTRACTS Noteworthy range extensions are presented for Noctilio albiventris, Rhinophylla alethina, Sturnira aratathomasi, and Lonchophylla handleyi, including the second Colombian report for the last. A previous report of Molossops brachymeles is clarified as representing M. abrasus. Se presentan algunas notables extensiones del rango de distribucion para las especies Noctilio albiventris, Rhinophylla alethina, Sturnira aratathomasi y Lonchophylla handleyi, este ultimo siendo el segundo reporte para Colombia. Un reporte anterior de Molossops brachymeles se clarifica como representativo de M. abrasus. Apresentam-se notaveis exten^oes mas distribui96es das especies Noctilio albiventris, Rhin- ophylla alethina, Sturnira aratathomasi, e Lonchophylla handleyi, esta ultima sendo apenas o segundo registro para a Colombia. Clarifica-se o registro anterior de Molossops brachymeles como representativo de M. abrasus. Introduction Despite considerable interest in Neotropical mammals, southwestern Colombia remains poor- ly understood in this respect. This is mainly a result of a lack of adequate collections caused by the inaccessible nature of much of the zone. Early collecting expeditions to which we owe much of our knowledge were undertaken around the turn of the century by personnel of the American Mu- seum of Natural History and summarized by Allen (1916). Bats were typically underrepresented in these early collections because of inadequate col- lecting techniques in use at the time. Now, with the aid of Japanese mist nets, we are able to obtain more complete samples of bat communities. In this report I present results of a continuing col- lecting effort during the past five years in this poor- ly known region, extending the known distribution From the Departamento de Biologia, Universidad del Valle, Call, Colombia. of Noctilio albiventris, Lonchophylla handleyi, Rhinophylla alethina, Sturnira aratathomasi, and Molossops abrasus. All specimens mentioned were collected in mist nets, prepared as standard study skins with skulls, and deposited in the mammal collection of the Departamento de Biologia, Universidad del Valle, Cali, Colombia (UV). Distribution Noctilio albiventris The lesser bulldog bat was recently reviewed by Davis (1976) and by Hood and Pitocchelli (1983). Both mapped the distribution as including eastern Colombia across the Llanos and Amazonas and the northern Caribbean coast. Davis (1976) re- ported the altitudinal range of the species as ex- tending up to 1 , 1 00 m. We have found this species to be common in the upper Cauca valley, between ALBERICO: DISTRIBUTION OF COLOMBIAN BATS 133 the Cordillera Central and the Cordillera Occi- dental of the Andes, where the elevation reaches this approximate limit. Fifteen specimens from the Departamento (= state) del Valle del Cauca and adjacent Departamento del Cauca were com- pared with the descriptions and measurements of all subspecies recognized by Davis (1976). This population is indistinguishable from N. a. minor in all characters examined and undoubtedly fol- lows the Rio Cauca south from the Caribbean low- lands. A similar southern extension is most prob- able in the valley of the Rio Magdalena to the Departamento de Huila, but has yet to be con- firmed by collections. Specimens Examined— Cauca: Rio Palo, 18 km S, 5 km E Puerto Tejada, 3°04'N, 76°22'W, 1,050 m (3 92, UV3 1 3, 324, 325); Valle del Cauca: 2 km S, 4 km W Candelaria, 3°23'N, 76°23'W, 1,000 m (1 (5, UV676); Universidad del Valle (Melendez Campus), 8 km S Cali, 3°22'N, 76°32'W, 1,000 m (5 S6, UV2602, 2603, 2604, 2608, 2609; 2 99, UV2605, 2607); 13 km S, 1 km E Cali, 3°22'N, 76°32'W, 1,000 m (2 66, UV2620, 2611; 1 9, UV2612). Lonchophylla handleyi This species was described on the basis of spec- imens from Peru and southern Ecuador by Hill (1980), who suggested that some individuals in existing collections might be misidentified as L. robusta. Lonchophylla handleyi was first reported for Colombia by Alberico and Orejuela (1982), who collected a single individual from near the Ecuadorian border at 850 m. A specimen recently collected from the Departamento del Valle del Cauca at 480 m provides the second record for Colombia. Both specimens are larger (greatest length of skull, 28.4 and 28.6 mm, respectively) than the largest L. robusta reported by Hill (1980) for Peru and Ecuador and are larger than any L. robusta in our collections from western Colombia. Both Colombian specimens of L. handleyi are from the lower slope Andean forests, probably one of the last habitats to be intensively sampled for mammals in this country. The presence of this species in a relatively narrow elevational band within this habitat type attests to the importance of continued collecting in the Pacific slope of the Andes in southwestern Colombia. Specimens Examined— Nariilo: 5 km E Junin, l''20'N, 78°08'W, 850 m (1 6, UV3007); Valle del Cauca: Rio Cajambre, approx. 60 km S Buena- ventura, 3°20'N, 77°00'W, 480 m (1 9, UV3694). Rhinophylla alethina This species was described based on specimens from western Colombia in the Departamento del Valle del Cauca (Handley, 1 966) and until recently was known only from the type locality. Albenco and Orejuela (1982) reported it from Narifio near the Ecuadorian border and suggested that it might have a broader geographic range than previously thought, which was confirmed by Baud (1982) who reported the species for Ecuador. Our collections show R. alethina to be relatively common in the Pacific lowlands and the adjacent lower slopes of the western Andes up to 850 m. That this species was only recently described and remains poorly known is undoubtedly due to insufficient collect- ing in the forests of this zone. Specimens Examined— Nariflo: 5 km E Junin, 1°20'N, 78°08'W, 850 m (3 66, UV3029, 3033, 3036; 5 99, UV3030, 3031, 3032, 3034, 3035). Valle del Cauca: Alto Anchicaya, 35 km S, 20 km E Buenaventura, 3°34'N, 76°54'W, 400 m (2 66, UV3166, 3167); Rio Azul, 5 km N, 25 km W Darien, 3°59'N, 76°44'W, 560 m (1 9, UV3391); Rio Cajambre, approx. 60 km S Buenaventura, 3°20'N, 77°00'W, 480-520 m (l 6, UV3702; 1 9, UV3703); Rio Cahma, 13 km N, 14 km E Bue- naventura, 4°00'N, 76°59'W, 40 m (1 9, UV2809). Stumira aratathomasi In their description of this species, Peterson and Tamsitt (1968) reported three specimens, the ho- lotype from the Departamento del Valle del Cauca in western Colombia and two from an unknown locality in Ecuador. They stated that it might be restricted to the Pacific side of the Andes. Thomas and McMurray ( 1 974) provided measurements for the holotype and six individuals collected near the type locality and suggested that this species may be common at high elevations in the western An- des of Colombia. Our recent collections extend the known range some 150 km to the north in the Cordillera Occidental and, more importantly, re- cord the presence of S. aratathomasi in the Cor- dillera Central, where it was previously unknown. This species appears to inhabit medium to high elevation forests which are relatively continuous 134 FIELDIANA: ZOOLOGY in Colombia, and its occurrence both farther to the north and in the Cordillera Oriental is likely. Specimens Examined— Valle del Cauca: Cor- dillera Central: Hacienda "Los Alpes," 6 km S, 1 1 km E Florida, 3°16'N, 76°09'W, 2,400 m (1 9, UV3482); Cordillera Occidental: Betania, 10 km N, 15 km W Bolivar, 4°26'N, 76''19'W, 1,800 m (1 9, UV3876); Parque Nacional "Los Farallones de Cali," 1 0 km S, 1 6 km W Cali, 3°22'N, 76°4 1 'W, 2,600 m (1 9, UV3373); Paso de Galapagos, 8 km N, 4 km E El Cairo, 4°50'N, 76°12'W, 1,800 m {2 66, UV4131,4133). Molossops abrasus This species was reported for Colombia by Al- berico and Naranjo-H. (1982) as M. brachymeles, based on specimens from the Cauca valley in northern Valle del Cauca. Although often referred to by this latter specific epithet (see Cabrera, 1958; Freeman, 1981), the holotype of Dysopes abrasus from Brazil has been shown to represent this species (Husson, 1962; Carter & Dolan, 1978). The Co- lombian record extends the known distribution of M. abrasus in western South America from An- dean Peru some 1 ,600 km to the north. Specimens Examined— Valle del Cauca: 1 1 km S, 2 km W Cartago, 4''39'N, 75°56'W, 930 m (2 66, UV2451, 2452; 1 9, 2453). Acknowledgments This report is the result of the combined efforts of many friends and students, too numerous to mention by name, who have collaborated either by accompanying the author in the field, by sharing specimens collected during other activities, or both. However, a few individuals have contributed more than could be expected in the normal turn of events, and their support in the field and out has been especially important in the present study: Eduardo Velasco, Gloria Giral, Alonso Gonzalez, Guiller- mo Cantillo, and Luz Marina Alberico. To these, the author is most appreciative. Literature Cited Alberico, M., and L. G. Naranjo-H. 1982. Primer registro de Molossops brachymeles (Chiroptera: Mo- lossidae) para Colombia. Cespedesia, II: 141-143. Alberico, M., AND J. E. Orejuela. 1982. Diversidad especifica de dos comunidades de murcielagos en Na- rino, Colombia. Cespedesia, Suplemento no. 3(4 1-42): 31-40. Allen, J. A. 1916. List of mammals collected in Co- lombia by the American Museum of Natural History expeditions, 1910-1915. Bulletin of the American Museum of Natural History, 35: 191-238. Baud, F. J. 1982. Presence de Rhinophylla alethina (Mammalia, Chiroptera) en Equateur et repartition actuelle du genre en Amerique du Sud. Revue Suisse de Zoologie, 89: 8 1 5-82 1 . Cabrera, A. 1958. Catalogo de los mamiferos de America del Sur. Revista del Museo Argentino de Ciencias Naturales "Bernardino Rivadavia", Ciencias Zoologicas, 4: 1-307. Carter, D. C, and P. G. Dolan. 1978. Catalogue of type specimens of neotropical bats in selected Euro- pean museums. Special Publications, The Museum, Texas Tech University, 15: 1-136. Davis, W. B. 1976. Geographic variation in the lesser noctilio, Noctilio albiventris (Chiroptera). Journal of Mammalogy, 57: 681-107. Freeman, P. W. 1981. A multivariate study of the family Molossidae (Mammalia, Chiroptera): Mor- phology, ecology, evolution. Fieldiana: Zoology, n.s., 7: 1-173. Handley, C. O., Jr. 1966. Descriptions of new bats {Choeroniscus and Rhinophylla) from Colombia. Pro- ceedings of the Biological Society of Washington, 79: 83-88. Hill, J. E. 1 980. A note on Lonchophylla (Chiroptera: Phyllostomatidae) from Ecuador and Peru, with the description of a new species. Bulletin of the British Museum (Natural History), Zoology Series, 38: 233- 236. Hood, C. S., and J. Pitocchelll 1983. Noctilio al- biventris. Mammalian Species, 197: 1-5. HussoN, A. M. 1962. The bats of Suriname. Rijks- museum van Natuurlijke Historic, Leiden, 58: 1-282. Peterson, R. L., and J. R. Tamsitt. 1968. A new species of bat of the genus Sturnira (family Phyllosto- matidae) from northwestern South America. Life Sci- ences Occasional Papers, Royal Ontario Museum, 12: 1-8. Thomas, M. E., and D. N. McMurray. 1974. Ob- servations on Sturnira aratathomasi from Colombia. Journal of Mammalogy, 55: 834-836. ALBERICO: DISTRIBUTION OF COLOMBIAN BATS 135 Distributional Records of Bats from the Caribbean Lowlands of Belize and Adjacent Guatemala and Mexico Timothy J. McCarthy ABSTRACTS Thirty new species records are presented for the bat fauna of Belize, along with secondary records for eight bats that had been recorded previously from that country. Contiguous lowland localities in Guatemala provided new department records: nine for El Peten, five for Izabal, and two for Alta Verapaz. The El Peten records include the first confirmation of Vampyrum spectrum in Guatemala. One state record for Quintana Roo, Mexico, is reported. These species represent the genera Saccopteryx, Balantiopteryx, Diclidurus, Noctilio, Pteronotus, Mormoops, Micronycteris, Lonchorhina, Macrophyllum, Tonatia, Mimon, Phyllostomus, Phylloderma, Trachops, Chrotopterus, Vampyrum, Glossophaga, Uroderma, Vampyrops, Vampyrodes, Vam- pyressa, Chiroderma, Artibeus, Centurio, Diphylla, Natalus, Myotis, Eptesicus, Lasiurus, Bau- erus, Eumops, and Molossus. Range extensions are acknowledged for Saccopteryx leptura, Diclidurus virgo, Noctilio leporinus, Micronycteris nicefori, Macrophyllum macrophyllum, Phyl- lostomus discolor, Vampyrum spectrum, Glossophaga commissarisi, Uroderma bilobatum, Vam- pyrodes caraccioli, Artibeus toltecus, and Bauerus dubiaquercus. A checklist of the bat fauna of Belize, which stands at 66 species, is presented. Se registran 30 especies que no habian sido citadas antes para la fauna de murcielago de Belice, con registros secundarios para ocho murcielagos ya conocidos de ese pais. En ciertas localidades contiguas de las tierras bajas de Guatemala, se obtuvieron nuevos registros depar- tamentales: nueve de El Peten, cinco de Izabal, y dos de Alta Verapaz. Los registros de El Peten incluyen la primera confirmacion de Vampyrum spectrum en Guatemala. Ademas, se presenta un nuevo registro estatal para Quintana Roo, Mexico. Las especies obtenidas estan segregadas en los generos Saccopteryx, Balantiopteryx, Diclidurus, Noctilio, Pteronotus, Mormoops, Mi- cronycteris, Lonchorhina, Macrophyllum, Tonatia, Mimon, Phyllostomus, Phylloderma, Trach- ops, Chrotopterus, Vampyrum, Glossophaga, Uroderma, Vampyrops, Vampyrodes, Vampyressa, Chiroderma, Artibeus, Centurio, Diphylla, Natalus, Myotis, Eptesicus, Lasiurus, Bauerus, Eu- mops, y Molossus. Para cada una de las siguientes especies de murcielagos se anota el alcance geografico de su distribucion conocida: Saccopteryx leptura, Diclidurus virgo, Noctilio leporinus, Micronycteris nicefori, Macrophyllum macrophyllum, Phyllostomus discolor, Vampyrum spec- trum, Glossophaga commissarisi, Uroderma bilobatum, Vampyrodes caraccioli, Artibeus tolte- cus, y Bauerus dubiaquercus. Se incluye una lista de 66 especies que representan la fauna de murcielagos de Belice. Apresenta-se records de 30 novas especies de morcegos para Belice, e de oito especies pouco conhecidas no pais. Areas adjacentes, na Guatemala, providenciaram novos records para: El From the Department of Mammalogy, American Mu- seum of Natural History, Central Park West at 79th Street, New York, NY 10024. MCCARTHY: DISTRIBUTION OF BATS 137 Peten (nove especies), Izabal (cinco especies), e Alta Verapaz (dois especies). Os records de El Peten incluem as primeiras confirma^oes de Vampyrum spectrum na Guatemala. Um novo record para Quintana Roo, Mexico, e incluido. Estas especies representam os generos Saccop- teryx, Balantiopteryx, Diclidurus, Noctilio, Pteronotus, Mormoops, Micronycteris, Lonchorhina, Macrophyllum, Tonatia, Mimon, Phyllostomus, Phylloderma, Trachops, Chrotopterus. Vam- pyrum, Glossophaga, Uroderma, Vampyrops, Vampyrodes, Vampyressa, Chiroderma, Artibeus, Centurio. Diphylla, Natalus, Myotis, Eptesici4s, Lasiurus. Bauerus, Eumops, e Molossus. Re- conhece-se extensoes nas areas onde sao encontrados Saccopteryx leptura, Diclidurus virgo, Noctilio leporinus, Micronycteris nicefori, Macrophyllum macrophyllum, Phyllostomus discolor, Vampyrum spectrum, Glossophaga commissarisi, Uroderma bilobatum, Vampyrodes caraccioli, Artibeus toltecus. e Bauerus dubiaquercus. Apresenta-se uma lista da fauna de morcegos em Belice, que agora conta com 66 especies. Introduction Inventories of bat communities in Mexico and Central America have increased significantly dur- ing the last twenty-five years (Jones et al., 1977). Although the resulting data have enhanced our knowledge of the distributions and the zoogeo- graphical relationships of species, incomplete sur- veys exist for certain regions. The northern low- lands along the Caribbean coast of Honduras, Guatemala, Belize, and Quintana Roo, in Mexico, is one such region. Travel within this coastal ver- sant has improved with agricultural and settle- ment expansion. The isolation of Belize from its neighbors has been reduced with the construction of roads in Guatemala's frontier of El Peten and Mexico's former territory of Quintana Roo. A paved road from Izabal now connects El Peten and Belize with the Pan-American Highway in western Guatemala. Road development continues within Belize for all-weather travel. Belize is situated within the Caribbean lowland drainage of northern Central America. Contiguous with Belize on this eastern slope is the eastern portion of the department of El Peten to the west and, to the south, the department of Izabal, both of Guatemala. Southern Quintana Roo of penin- sular Mexico borders to the north (see fig. 1 and Gazetteer). The topography of these Caribbean lowlands extends from the lower ranges (600 m and below) of the eastern Sierra de Chama, the Sierras de las Minas, the Sierra de Santa Cruz, the Sierra del Meredon, and the Montarias del Mico in Alta Verapaz and Izabal, and the Maya Moun- tains of southern Belize and southeastern El Peten to the low undulating relief of southern Quintana Roo. The Maya Mountains represent a heavily eroded Paleozoic formation that now ranges at the top from 671 to 853 m in elevation, with the high- est peak at 1113m (Wright et al., 1959). Annual rainfall in portions of Izabal averages from 3,000 to nearly 5,000 mm (Portig, 1976). Over 4,500 mm of rainfall was reported from the most south- em coastal area of Belize. North and northwest- ward of the Maya Mountains, rainfall decreases appreciably to less than 1 ,500 mm in north-central El Peten and northern Belize, where less than 1 ,400 mm was recorded near the Quintana Roo border (Walker, 1973). The severity of this northward reduction of rainfall is intensified by the increased lack of surface drainage into the Yucatan Penin- sula of Mexico. Because the limestone shelf of northern Belize has geological affinities with the Yucatan Peninsula (Wright et al., 1 959), the south- em limit of this peninsula can be considered the fault line extending from north of the Maya Moun- tains westward through the northem shore of Lake Peten-Itza, El Peten (Wadell, 1938; West, 1964). Effectively, the northem plain of Belize and north- em El Peten are portions of the Yucatan Peninsula. The northward shift from alluvial soils to shallow calcareous soils, along with the mentioned cli- matic changes, create edaphic conditions that af- fect the composition and the structure of the vege- tation that can be supported (Lundell, 1 934, 1 937; Standley & Record, 1936; Wright el al., 1959; Pen- nington & Samkhan, 1968). The potential effect of this transitional physiography on the distribu- tion and relative abundance of bats in this Carib- bean lowland region will require further inventory studies. This paper documents 30 new records for Belize. A checklist of the known bat fauna for this country is annotated in the Appendix. Sixty-six species are cited. Included here are also records from nearby localities for El Peten, Izabal, and Alta Verapaz, Guatemala, and Quintana Roo, Mexico. Nine species records from EI Peten, five records from 138 HELDIANA: ZOOLOGY Izabal, and two records from Alta Verapaz in- crease the number of reported species for these departments to 35, 31, and 40, respectively (see Jones, 1966; Carter etal., 1966; Rick, 1968; Smith, 1972; LaVal, 1973a; Martinez R., 1980; Mc- Carthy, 1982). Jones et al. (1973) and Bimey et al. (1974) summarized the records for 31 species from Quintana Roo, and this paper provides one additional record. Materials and Methods The bats that I collected during the years 1974- 1984 in Belize and El Peten (Parque Nacional Ti- kal), Guatemala, were obtained principally with mist nets set at ground level; aerial netting and the use of a bat trap were limited. Unless otherwise stated, mist netting was carried out during the first half of the night. A limited number of specimens were obtained with hand nets or plastic funnel traps at roost sites. Specimens were prepared as standard museum skins with skulls and/or skele- tons, or as fluid-preserved specimens. These vouchers are housed in Field Museum of Natural History, Chicago (FMNH); The Museum, Mich- igan State University, East Lansing (MSU); Car- negie Museum of Natural History, Pittsburgh (CM); and American Museum of Natural History, New York (AMNH). A survey of 45 museum collections in the United States, Canada, Mexico, Guatemala, and England resulted in additional specimens from Belize, El Peten, Izabal, Alta Verapaz, and Quintana Roo. Pertinent specimens (147) have been included in this report from the following institutions [collec- tors in brackets]: American Museum of Natural History, New York [N. Sullivan]; British Museum (Natural History), London, England (BM) [R. H. L. Disney; P. Williams; A. M. Hutson; R. E. Steb- bings]; Carnegie Museum of Natural History [N. A. Bitarj; Field Museum of Natural History [L. de la Torre]; Florida State Museum, University of Florida, Gainesville (FSM) [F. J. Bonaccorso]; Museum of Zoology, Louisiana State University, Baton Rouge (LSUMZ) [D. M. Uy]; Royal On- tario Museum, Toronto, Canada (ROM) [R. L. Peterson; J. Kamstra; J. Fragoso]; Texas Coop- erative Wildlife Collections, Texas A«&,M Uni- versity, College Station (TCWC) [D. C. Carter; M. D. Engstrom]; Texas Tech University, Lubbock (TTU) [P. Diamond]; and United States National Museum of Natural History, Washington, D.C. (USNM) [E. L. Tyson]. Systematic arrangement of species accounts and nomenclature, unless otherwise indicated, follow Jones et al. (1977) and Handley (1980). Disney (1968) did not provide data for the first records of Pteronotus davyi, Tonatia minuta, and Eptesi- cusfurinalis from Belize. Those data are presented in the respective accounts of this report, with ad- ditional records. Further secondary records from Belize of Mimon crenulatum, Trachops cirrhosus, Glossophaga commissarisi, Vampyressa pusilla, and Eumops auripendulus are also included. All of the species accounts are discussed in the context of their range and elevational distributions in Mexico and Central America. Hall (1981) was the primary reference for this unless cited otherwise. Forest types in Belize follow Wright et al. ( 1 959), whose classification was partially based on the sea- sonal formation series (Beard, 1 944), which refers to structural appearance. The correct political alignments between the states of the Yucatan Pen- insula are inconsistent among a number of pub- lished maps. The state boundary between Quin- tana Roo and Campeche on the map in Figure 1 (see also Gazetteer) is based on a number of Gov- ernment of Mexico (Secretaria de Programacion y Presupuesto) maps, including "Carta Topografica, Merida" (1:1 ,000,000; 1 979 and 1 983) and "Mapa Geografica" (1:5,000,000; 1980). Species Accounts Family EMBALLONURIDAE Subfamily EMBALLONURINAE Saccopteryx leptura (Schreber, 1 774) Specimen Examined— BELIZE. Toledo: 2. 1 km NNE Salamanca Camp, Columbia Forest, 1 9 (cm). The known distribution of this small sac-winged bat north of Panama extends through Costa Rica and Nicaragua to Chiapas along the Pacific ver- sant. The presence of predominantly lowland Sac- copteryx leptura in southern Belize represents a country record and an extension of its distribution along the Caribbean side from southeastern Nic- aragua. Small bats were observed foraging up to heights of 13-13.5 m during the twilight period of the evening. Flight appeared to be concentrated within MCCARTHY: DISTRIBUTION OF BATS 139 a small, open area below the lower canopy of ev- ergreen seasonal forest. A short mist net was hand- hoisted to capture (24 March) this adult specimen. Saccopteryx bilineata was collected shortly after the capture of S. leptura. Balantiopteryx io Thomas, 1 904 Specimens Examined— GUATEMALA. EI Pe- ten: Poptun, Finca Ixobel, \S 66, 1 8 99 (cm). The restricted distribution of Balantiopteryx io ranges from the Gulf lowlands of Veracruz, Oa- xaca, and Tabasco to the lowlands of Belize and eastern Guatemala. Kirkpatrick et al. ( 1 975), Cart- wright and Kirkpatrick (1977), and Sanborn (1936) represent the previous records for Belize and Iza- bal. The Poptun locality represents the first record for El Peten. The specimens reported here were collected ( 1 2 June) by N. A. Bitar as they exited from a cave surrounded by secondary forest. The distribution of this colonial species may be restricted in part by the availability of adequate cave habitats as roosting sites. Subfamily DICLIDURINAE Diclidurus virgo Thomas, 1903 Specimen Examined— BELIZE. Cayo: 1.5 km SSW Roaring Creek, 1 6 (fmnh). The white bat is represented by relatively few localities in Middle America, which extend from western (Nayarit) and eastern (southern Veracruz) Mexico through Central America. Specimens from southwestern El Peten were reported by Jones (1966). The single specimen from Belize repre- sents a northward range extension in the Carib- bean lowlands from northwest Honduras (Carter & Dolan, 1978) and a record for the country. The single bat apparently was roosting on the trunk of a fig tree (Ficus insipida) overhanging a pool along the Roaring Creek River. It was cap- tured (May) by C. Tzul after being observed on a number of occasions roosting near, but not among, a group of Rhynchonycteris naso. Jones ( 1 966), Starrett and Casebeer ( 1 968), and Handley ( 1 976) commented on the high foraging habits of Dicli- durus. Similar to the molossid bats, Diclidurus probably concentrates its foraging efforts at levels well above the tree canopy and beyond the reach of conventional collecting techniques, except fire- arms. This may explain why there are few speci- mens available in collections. Goodwin ( 1 969) considered Diclidurus virgo at best not more than subspecifically different from D. albus. Both species were recognized by Ojasti and Linares (1971), who questioned Goodwin ( 1 969) because they believed that his South Amer- ican comparative material represented D. virgo and not D. albus. Family NOCTILIONIDAE Noctilio leporinus mastivus (Vahl, 1797) Specimens Examined— BELIZE. Cayo: Banana Bank, 1 <5, 1 9 (fmnh); Barton Creek, at Western Hwy., 1 9 (fmnh). Stann Creek: Melinda, Stann Creek River, 1 6 (fmnh). Toledo: 1.2 km E Agua- cate, Aguacate River, 2 66 (cm), 1 9 (bm); Big Fall, vicinity Rio Grande Bridge, 1 9 (fmnh); Salamanca Camp, 1 9 (bm). The fishing bat occurs along the riparian habi- tats of river systems, inland lakes, and coastlines in primarily lowland regions from northwestern (northern Sinaloa), eastern (southern Veracruz), and peninsular (Yucatan) Mexico throughout Cen- tral America (Davis, 1973;Hellebuycketal., 1985). Dickerman et al. (1981) reported a locality for Noctilio leporinus from Alta Verapaz as in the Ca- ribbean drainage when it was clearly in the Rio Usamacinta drainage of the Gulf lowlands. The Belizean localities extend northward the recorded occurrence of A^. leporinus from Izabal and north- western Honduras (Carter et al., 1966). All of the specimens were obtained (March, April, May, July, August) over rivers and a pond except for one individual, which was mist netted (28 August) low over a pasture adjacent to a flood- ed river. This bat was foraging primarily for insects since its feces contained the chitinous remains of these prey. Additional fishing bats from the lo- calities in Cayo and Stann Creek districts were captured, banded, and released. This bat was com- mon along the South Stann Creek drainage. Cocks- comb Basin. A specimen belonging to M. Craig, Belize Audubon Society, was collected at Indian Church (Lamanai), New River Lagoon, Orange Walk District. An old specimen of M leporinus in the collections of British Museum (Natural His- tory) was registered in 1 909 without pertinent field data. The two peninsular records from Campeche (Jones et al., 1973) and Yucatan (Bimey et al., 1974) were obtained in coastline habitat along the 140 FIELDIANA: ZOOLOGY Gulf of Mexico. Although subsurface drainage predominates north of Belize into Quintana Roo, shallow inland "lagunas" are fairly common and probably support Noctilio populations. Family MORMOOPIDAE Pteronotus davyi fulvus (Thomas, 1 892) Specimens Examined— BELIZE. Cayo: Central Farm, 1 <5 (cm), 1 $ (fmnh); Ontario, 5.5 km W Teakettle, 1 S (fmnh); Unitedville, 9 km WSW Teakettle, 1 <5 (fmnh). Orange Walk: Tower Hill, B.S.I, compound, 3 99 (fmnh). Toledo: Aguacate, 1 (3 (cm); 1.2 km E Aguacate, 1 $ (bm), 1 $ (cm); Rice Station, 2 6$ (fmnh); 0.4 km W Rice Station, 1 $ (fmnh); San Antonio, 1 $ (fmnh). GUATE- MALA. El Peten: Parque Nacional Tikal, 1 $ (msu). Smith (1972) summarized the majority of the capture localities for this subspecies of naked- backed bat, which ranges from northwestern (So- nora), northeastern (Tamaulipas), and peninsular (Yucatan) Mexico southeastward to Honduras and El Salvador, but omitted the only record for Belize (Disney, 1968). Parque Nacional Tikal is the first record for El Peten, and the Belizean specimens provide additional records for Belize. Disney ( 1 968) did not present data for his single Pteronotus davyi specimen. This male was ob- tained ( 1 November) in Cayo District, at Listowel along the Belize River, and is housed in British Museum (Natural History). The subsequent spec- imens reported here were collected (October-De- cember, May, July, August) in open areas, bor- dering on vegetation and buildings, and over water. The specimen from El Peten was captured (25 March) along a trail in upland deciduous forest. An additional P. davyi from Tikal was captured, banded, and released. Pteronotus personatus psilotis (Dobson, 1878) Specimens Examined— BELIZE. Toledo: 1 .2 km E Aguacate, Aguacate River, 1 9 (bm), 4 33, 1 9 (CM); Big Fall, 1.5 km WSW Rio Grande Bridge, 1 $ (fmnh); 0.8 km NW Blue Creek, 1 9 (fsm); Crique Jute, 1 $ (fmnh); Crique Lagarto, 1 km NW San Antonio, 1 9 (fmnh); Jacinto Creek, at Punta Gorda Road, 1 3 (msu); 0.4 km W Rice Station, 1 9 (fmnh); Salamanca Camp, 1 9 (usnm); San Antonio, 1 9 (fmnh); 0.9 km WNW San Pedro Columbia, 1 9 (fmnh). The distribution of Pteronotus personatus psi- lotis extends from western (southern Sinaloa) and eastern (Tamaulipas) Mexico southeastward to Honduras and El Salvador (Smith, 1972), with Caribbean lowland localities in Campeche (Jones et al., 1973), El Peten (Jones, 1966), and Alta Ver- apaz (Jones, 1966). Elevations range from 123 to 984 m. These localities from southern Belize are the first records for the country. Fifty-three percent of the small moustache bats were collected (March, May, July) over open water; the remainder were foraging (January, April, Au- gust, December) in open areas adjacent to build- ings or corralled cattle. Mormoops megalophylla megalophylla Peters, 1864 Specimens Examined— BELIZE. Belize: 6.6 km N Churchyard, 1 9 (cm). Cayo: 1.6 km NW Au- gustine, Rio Frio, 1 3 (ttu). Stann Creek: Melinda, 1 3 (fmnh). Toledo: Forest Home, 1 3 (msu); Pueb- lo Viejo, 1 3 (fmnh). GUATEMALA. Izabal: 25 km SSW Puerto Barrios, 1 3 (tcwc). The leaf-chinned bat has been reported through- out Mexico, Guatemala, El Salvador, and Hon- duras (Smith, 1972). Davis and Carter (1962), Jones (1966), and Taibel (1977) provided lowland records for El Peten and Alta Verapaz. Elevations range from near sea level to 2270 m. These lo- calities are the first records for Belize and Izabal. Except for one Belizean specimen, which was captured (9 June) in a cave, these leaf-chinned bats were associated (March, April, December) with open areas bordering on forest or orchard edges, including pine savanna. One Mormoops specimen, which was registered into the British Museum (Natural History) collections in 1892, may have been obtained in the vicinity of Belize City. The Guatemalan specimen was collected by D. C. Car- ter. Family PHYLLOSTOMIDAE Subfamily PHYLLOSTOMINAE Micronycteris brachyotis (Dobson, 1878) Specimens Examined— BELIZE. Cayo: 1 km NW Augustine, 2 $S (fmnh). Toledo: Crique Ne- gro, Columbia Forest, 1 3 (bm). The first Middle American specimen of Micro- McCARTHY: DISTRIBUTION OF BATS 141 nycteris brachyotis was initially reported from Nic- aragua as M. syhestris by Goodwin (1946), but was correctly identified by Sanborn (1949). Sub- sequent records are from the Gulf-Caribbean low- lands of southern Veracruz (Medellin L. et al., 1983), Oaxaca (Schaldach, 1964), Chiapas (Davis et al., 1964), and El Peten (Jones, 1966; Rick, 1 968; McCarthy, 1 982); and the Pacific-Caribbean versants of Costa Rica (Howell & Burch, 1974; Starrett, 1976; LaVal & Fitch, 1977) and Panama (Handley, 1966; Fleming etal., 1972; Bonaccorso, 1979). Reported elevations range from 40 to 594 m. The present specimens are the first records of the yellow-throated bat for Belize. Two specimens were captured (29 July) as they exited from a cave into a low deciduous seasonal forest, and a third bat was taken (28 May) along a path in an evergreen seasonal forest. The two specimens from Cayo, which I obtained while as- sisting a histoplasmosis survey, were listed as "M. bardyoiis" in a preliminary report (Quinones et al., 1978, p. 559) and no specific locality data were provided. Micronycteris megalotis mexicana Miller, 1898 Specimens Examined— BELIZE. Corozal: San Antonio, 2 km NW Corozal, 1 6 (fmnh). Orange Walk: San Antonio, Rio Hondo, 2 55, 1 2 (fmnh). Toledo: Aguacate, 1 6 (cm); Big Fall, 2 km E Rio Grande Bridge, 1 2 (bm); Cuevas Creek Bridge, 10 km NW Punta Gorda, 1 5, 1 2 (bm), 1 S (amnh), 1 2 (msu); Nimli Punit, 1 2 (cm); Rocky Run Ranch, 4.8 km NW Punta Gorda, 1 3, 1 2 (bm); Union Camp, 2 22 (bm); Vista Hermosa Ranch, 3.7 km WNW Punta Gorda, 1 2 (cm). GUATEMALA. El Peten: Parque Nacional Tikal, 1 <5 (fmnh). The distribution of this subspecies of big-eared bat extends from western (Jalisco), eastern (south- em Tamaulipas), and peninsular (Yucatan) Mex- ico, along the Pacific coastal and highland regions, to Costa Rica. Gardner et al. ( 1 970) suggested that the southern extent of Micronycteris megalotis mexicana is in the Cordillera Talamanca of Costa Rica. This species has been recorded most often at lowland-moderate elevations, up to 2870 m. Specimens from Isla Cozumel, Quintana Roo, rep- resent the only record for Quintana Roo (Jones et al., 1973). The records of A/, m. mexicana which are reported here are the first for Belize and El Peten. Belizean specimens were obtained (May, July, August, November) in diurnal roost sites (shallow caves and limestone chambers, bridge approach- ments, abandoned rum factory boiler) and col- lected in forest habitats (riparian marsh, evergreen and semi -evergreen, deciduous semi -evergreen, and deciduous seasonal). The Tikal specimen was captured (6 June) roosting in a passageway of an excavation tunnel within a ruin complex. A second juvenile male was captured, banded, and released (29 July) in escobal palm (Cryosophila argentea) forest, 1.9 km SE Tikal Reservoir. Micronycteris nicefori Sanborn, 1 949 Specimen Examined— BELIZE. Toledo: 0.4 km NE Aguacate, 1 2 (fmnh). Handley ( 1 966) documented the first specimens of Micronycteris nicefori north of South America, from Panama. Subsequently, it has been reported from southeastern Nicaragua (Baker & Jones, 1975) and both the dry Pacific (Starrett, 1976) and wet Caribbean (LaVal, 1977) lowlands of Costa Rica. These Central American localities range from near sea level to over 100 m. This first record from Belize also represents a significant Central Amer- ican range extension along the Caribbean versant. The M. nicefori specimen reported here was mist netted on 1 5 December along a track in hilltop, evergreen seasonal forest. Micronycteris schmidtorum Sanborn, 1935 Specimens Examined— BELIZE. Corozal: Pat- chakan, 2 22 (fmnh). Orange Walk: 1 .3 km W San Antonio, Rio Hondo, 1 6 (fmnh). Toledo: Big Fall, 1 km E Rio Grande Bridge, 1 6 (cm). Micronycteris schmidtorum was described (San- bom, 1935) from specimens collected in the Ca- ribbean lowlands of Izabal. An additional Gua- temalan specimen was recorded in the Pacific piedmont (Dickerman et al., 1981). The remaining Central American records represent both the Pa- cific and Caribbean lowland slopes from Honduras (Sanbom, 1941), Nicaragua (Davis et al., 1964; Baker & Jones, 1 975), Costa Rica (Starrett & Case- beer, 1968; Fleming et al., 1972; Howell & Burch, 1974; LaVal & Fitch, 1977), and Panama (Han- dley, 1 966). Specimens from Yucatan assigned to M. schmidtorum by Villa-R. (1966) were reiden- tified as M. megalotis by Jones et al. (1973). An identification of M. schmidtorum (Jones et al., 1 973) for a specimen from Isla Cozumel, Quintana Roo, was questioned by Hall (1981) because this 142 FIELDIANA: ZOOLOGY specimen previously was identified as M. mega- lot is (Jones & Lawlor, 1 965). I examined this spec- imen (University of Kansas 9 1 539) and agree that it is M. schmidtorum. The northern distribution of this big-eared bat extends to the Caribbean coast of the Yucatan Peninsula. The specimens reported here are the first records for Belize. At Parque Nacional Tikal, one juvenile and two adult females, which were captured (30 July) in a hollow tree (Bursera semirouba) of an upland de- ciduous seasonal forest, were photographed, band- ed, and released. This site was revisited during the following March, but no Micronycteris were found. These individuals of M. schmidtorum were the first seen in El Peten. Similarly, Sanborn (1935) and Starrett and Casebeer (1968) reported indi- viduals from tree hollows. The Belizean specimens were captured (February, September, November) in the orchard vegetation of a village, along a sec- ondary forest edge, and in riparian secondary vegetation. Lonchorhina aurita aurita Tomes, 1863 Macrophyllum macrophyllum (Shinz, 1821) Specimens Examined— BELIZE. Cayo: Sibun River at Indian Creek, 1 $ (fmnh). Toledo: Big Fall, 1.7 km NE Rio Grande Bridge, 1 $ (cm). Tabasco, Mexico, represents the northernmost occurrence for the long-legged bat, which is known from both the Caribbean and Pacific regions of Central America. Primarily a lowland inhabitant, Macrophyllum macrophyllum ranges from 40 to almost 600 m. These specimens represent a Ca- ribbean lowlands range extension from north- western Honduras (Valdez «& LaVal, 1971) and the first records for Belize. Harrison and Pendleton (1974), Gardner (1977), and Dickerman et al. (1981) indicated that long- legged bats may be closely associated with aquatic habitats. Similarly, the Belizean specimens were obtained ( 1 7 March, 1 April) from along the Sibun River, although not directly above water, and over the surface of the Rio Grande. The first bat was taken at approximately 0340 in a stand of shade trees, dominated by cohune palms (Orbignya co- hune), at the edge of an open pasture. Specimens Examined— BELIZE. Stann Creek: 5.3 km WNW Quam Bank, Cockscomb Basin, 1 9 (CM). Toledo: 0.8 km NW Blue Creek, 1 3, 1 2 (amnh); Crique Jute Village, 1 9 (cm); Crique Ne- gro, Columbia Forest, 1 S (bm), 1 $ (usnm); 2. 1 km NNE Salamanca Camp, Columbia Forest, 3 66 (CM). GUATEMALA. El Peten: Poptun, Finca Ixobel, 2 66 (cm). Lonchorhina aurita was first recorded in Middle America from Panama (Miller, 1912). Subsequent collecting has found this cave-dwelling bat north- ward through Central America to southeastern (southern Veracruz, Oaxaca, Tabasco) and pen- insular (Quintana Roo) Mexico. Predominately lowland, this distinctive leaf-nosed bat extends up to more than 1500 m in representative habitats. Jones et al. (1973) reported the only record from Quintana Roo, while specimens from Izabal (San- bom, 1936) are apparently the next Caribbean ver- sant record north of eastern Costa Rica (Nelson, 1965); records from Nicaragua and Honduras are lacking. The specimens examined for this account are the first records from Belize and El Peten. All specimens from Belize were captured (March, April, May, August) in deciduous seasonal and evergreen seasonal forests. The Guatemalan bats were captured by N. A. Bitar as they exited from the cave discussed in the Balantiopteryx io ac- count. Tonatia bidens bidens (Spix, 1823) Specimens Examined— BELIZE. Cayo: Rio Frio, 1 .6 km W Augustine, 1 9 (cm). Toledo: Nimli Punit, 1 6 (cm); Orange Creek, 1.5 km SW Punta Gorda, 1 6 (msu); 2. 1 km NNE Salamanca Camp, Columbia Forest, 1 6 (cm); 2.2 km NNE Sala- manca Camp, Columbia Forest, 1 9 (cm). Goodwin (1946) first recorded Tonatia bidens in Central America from the Pacific lowlands of Costa Rica. Other humid lowland records include both the Caribbean and Pacific versants of Pan- ama, continuing along the Caribbean corridor of Nicaragua, Honduras, and Guatemala. The north- ernmost record is from eastern Chiapas (Medellin L., 1983). The Guatemalan records are from the Caribbean lowlands of El Peten (McCarthy, 1982) and Izabal (Carter et al., 1966). Elevations range from near sea level to around 660 m. The present specimens constitute the first records from Belize. Four adult males were taken (March, April) over a creek in a low transitional forest, in a high ev- ergreen seasonal forest, and in a deciduous sea- sonal forest. A subadult male was captured (24 September) in the courtyard of a Mayan archae- ological site located in a high deciduous seasonal forest. MCCARTHY: DISTRIBUTION OF BATS 143 Tonatia evotis Davis and Carter, 1978 Specimen Examined- GUATEMALA. El Pe- ten: Parque Nacional Tikal, 1 6 (fmnh). Davis and Carter ( 1 978) described Tonatia evo- tis on the basis of its smaller size in comparison to T. sylvicola; a female from Izabal was designated as the holotype. El Peten is part of a Gulf-Carib- bean distribution which extends from southern Veracruz, Tabasco, Chiapas, and Campeche to Be- lize, and continues along northern Honduras (Da- vis & Carter, 1978). Martinez R. (1980) recorded an additional eastern Guatemalan locality in Aha Verapaz. All recorded elevations are less than 100 m. The T. evotis from Tikal represents the first record for El Peten. Two adult males and one pregnant female were mist netted (20 February, 29 and 25 March) in Tikal along the Uaxactun Road, at a permanent water pool in escobal palm forest, and in an upland deciduous seasonal forest. One male and the fe- male were banded and released. Tonatia minuta Goodwin, 1 942 Specimens Examined— BELIZE. Cayo: 1.1 km W Augustine, 1 2 (fmnh); Central Farm, at Belize River, 1 9 (fmnh); 1.2 km E Macaw Bank, 1 2 (fmnh). Toledo: Big Fall, 1.7 km NE Rio Grande Bridge, 1 2 (msu); San Lucas, 1 2 (msu). This small Tonatia was originally described from the Caribbean coast of Nicaragua as T. nicaraguae (Goodwin, 1942a). Its Middle American distri- bution is lowland ( 1 5 to 6 1 0 m) along Caribbean and Pacific versants, from southern Veracruz (Lackey, 1 970) to El Peten, Guatemala (McCarthy, 1982) and Belize (Disney, 1968), continuing through Honduras (LaVal, 1969; Valdez & LaVal, 1971; Greenbaum & Jones, 1978), Nicaragua (Jones et al., 1971; Greenbaum & Jones, 1978), and Costa Rica (Gardner et al., 1 970; LaVal, 1 977), to Panama (Davis et al., 1964; Handley, 1966). This account represents additional records for the small round-eared bat in Belize. Disney (1968) reported no data for the first To- natia minuta specimen from Belize, which was a female collected (25 November) in Cayo District, at Listowel, along the Belize River. This specimen was deposited in British Museum (Natural His- tory). The additional specimens reported here were captured (November, January, February, April, May) over rivers or in a deciduous seasonal forest. The name minuta is applied in accordance with the discussion by McCarthy ( 1 982). Gardner ( 1 976) referred to a personal communication with C. O. Handley, Jr., who suggested that all small Tonatia (including minuta) represent a single species, T. brasiliense. Because the taxonomy is poorly under- stood, a systematic review of this group would be useful. Mimon cozumelae Goldman, 1914 Specimens Examined— BELIZE. Belize: Churchyard, Sibun River, 1 2 (fmnh). Cayo: "Mountain Pine Ridge", 2 33, 1 2 (bm); 0.8 km W Augustine, 1 6 (cm); 1 km NW Augustine, 2 $6 (fmnh); Barton Creek, at Western Hwy., 2 $S, 3 22 (fmnh). Toledo: vicinity Aguacate, 2 $S, 2 22 (cm), 1 (5 (fmnh); Crique Negro, Columbia Forest, 1 2 (bm); Pueblo Viejo, 1 3, 1 2 (fmnh); 2. 1 km NNE Salamanca Camp, Columbia Forest, 2 6S (cm); 2.2 km NNE Salamanca Camp, Columbia Forest, 1 <5 (cm); vicinity Union Camp, 2 5(5, 1 2 (bm), 2 22 (cm). This spear-nosed bat ranges from southeastern (northern Oaxaca, southern Veracruz) and pen- insular (Yucatan, Quintana Roo) Mexico south- eastward along the humid Caribbean side of Cen- tral America. Specimens from Isla Cozumel, Quintana Roo, provided the original description for this species (Goldman, 1914). Recorded ele- vations extend to 495 m. The Belizean localities reported here are the first records for the country. Mimon cozumelae were collected (January, March, May, July, August, September, December) along the edge of deciduous and semi-evergreen seasonal forests bordered with pasture, on riparian flood plains, over rivers, along paths in high de- ciduous, semi-evergreen seasonal forests, and in caves. Schaldach (1964), Villa-R. (1966), and Hall (1981) considered cozumelae a subspecies of ben- nettii. I tentatively accept cozumelae at the specific level. Minion crenulatum keenani Handley, 1 960 Specimens Examined— BELIZE. Cayo: Listow- el, Baking Pot, 1 S (fmnh). Toledo: Crique Negro, Columbia Forest, 1 6 (usnm). There are few records for Mimon crenulatum keenani from Middle America. The distribution of this distinctive spear-nosed bat extends along the Caribbean versant, from Panama (Handley, 144 FIELDIANA: ZOOLOGY 1966; Bonaccorso, 1979), Costa Rica (Gardner et al., 1970; LaVal, 1977), Nicaragua (Greenbaum & Jones, 1978), Belize (Ruiz, 1983), El Peten (McCarthy, 1982), and Campeche (Jones, 1964) to the Gulf lowlands of eastern Chiapas (Medellin L., 1 983). All recorded elevations range below 265 m. These specimens are the second and third rec- ords from Belize. The first record (Ruiz, 1 983) was obtained near Blue Hole, 14 km SE Belmopan, Cayo District. One Mimon crenulatum was captured (8 Oc- tober) in a house after it flew through an open window. The house was situated along the Belize River in an agricultural area. The second specimen was netted (29 March) along a path in evergreen seasonal forest. E. L. Tyson collected the specimen from Toledo District. Phyllostomus discolor verrucosus Elliot, 1905 Specimens Examined— BELIZE. Toledo: Cri- que Lagarto, 1 km NW San Antonio, 1 S (fmnh); 1 km NNE Salamanca Camp, Columbia Forest, 1 (3 (cm). GUATEMALA. Alta Verapaz: Lanquin, Lanquin Cave, approx. 1 49 km WSW Puerto Bar- rios, 1 (5, 1 5 (fmnh). Records of Phyllostomus discolor extend from southern (Oaxaca, Veracruz) Mexico along both the Pacific and Caribbean corridors of Central America. Records are more common at lower el- evations, less than 600 m. The new records from southern Belize provide a limited range extension northward from eastern Izabal (Sanborn, 1936). An adult from Crique Lagarto was captured ( 1 January) along the edge of low secondary forest bordering this settlement. The head of the bat was covered with yellow pollen. The second specimen was netted (21 March) in secondary vegetation, which resulted from slash-bum agriculture. Whit- ish pollen dusted the face, chest, and ventral wing surfaces. A male subadult Phyllostomus discolor that was taken ( 1 3 July) along a fenceline of sec- ondary vegetation between two pastures, 1 .9 km ENE Rio Grande Bridge, Big Fall, Toledo District, was photographed, banded, and released. L. de la Torre apparently captured (3 June) the two Phyl- lostomus from Alta Verapaz inside the entrance of Lanquin Cave. I tentatively follow Jones et al. (1977) in as- signing the specimens of Phyllostomus discolor from the Caribbean lowlands to the subspecies verrucosus. Sanborn (1936, p. 98) recognized ver- rucosus subspecifically, stating the "available mea- surements of rf/5co/or would place them much clos- er to verrucosus."'' He suggested the Panamanian P. d. discolor are assignable to verrucosus based on larger size. Felten (1956) and ^urt and Stirton (1961) concurred with his statement by referring a large series from El Salvador to verrucosus; with the availability of greater series of specimens, Da- vis and Carter (1962) indicated they could not recognize two subspecies of P. discolor in Central America and northern South America, acknowl- edging only P. d. verrucosus. Handley ( 1 966) ap- parently disagreed as he recognized the subspecies discolor in Panama. Multivariate analysis of mor- phological data (Power & Tamsitt, 1 973) suggested this species might be monotypic. Phylloderma stenops septentrionalis Goodwin, 1940 Specimens Examined— BELIZE. Toledo: Cri- que Negro, Columbia Forest, 1 2 (usnm); 2. 1 km NNE Salamanca Camp, Columbia Forest, 2 $S (CM). This rarely encountered species has been re- corded north of Panama from the Caribbean coast of Costa Rica (LaVal, 1977), the highlands of Hon- duras (Goodwin, 1940), the Caribbean lowlands of Guatemala (McCarthy, 1982), and the Gulf lowlands of Chiapas (Carter et al., 1966). Limited elevational data are from lowland to approxi- mately 1320 m. The specimens oi Phylloderma stenops from Belize represent the eighth, ninth, and tenth specimens north of Panama and the first records from Belize. All specimens were mist netted (March, Decem- ber) in similar evergreen seasonal forest habitats. E. L. Tyson collected the specimen from Crique Negro. Handley ( 1 966) regarded the Panamanian spec- imens to be Phylloderma stenops stenops, and those from northward into Middle America were thought to be subspecifically different from the nominal species. LaVal (1977) did not designate a subspe- cies for his Costa Rican specimen. Trachops cirrhosus coflini Goldman, 1925 Specimens Examined— BELIZE. Orange Walk: Richmond Hill (Goat Hill), 8.9 km SSW Orange Walk Town, 1 3, 1 $ (cm). Toledo: 2.2 km NNE Salamanca Camp, Columbia Forest, 1 $ (cm). MCCARTHY: DISTRIBUTION OF BATS 145 GUATEMALA. Izabal: 25 km SSW Puerto Bar- rios, 1 $ (tcwc). This lowland subspecies of the fringe-lipped bat is recognized from eastern (southern Veracruz) and southeastern (eastern Oaxaca) Mexico southeast- ward to Nicaragua. Recorded elevations are from near sea level to approximately 330 m. Jones (1966), Rick (1968), and McCarthy (1982) pro- vided records for El Peten. The description of this subspecies was based on specimens from eastern El Peten (Goldman, 1925). The first Belizean rec- ords were reported from Belize District in the vi- cinity of Belize City (Sanborn, 1941) and Rock- stone Pond (Pendergast, 1979). The specimen from Izabal is the first record for that Guatemalan de- partment. D. C. Carter obtained the single specimen from Izabal on 19 February. The additional Belizean specimens were mist netted (March, April) in de- ciduous marsh and evergreen forests. Chrotopterus auritus (Peters, 1856) Specimens Examined— BELIZE. Toledo: vicin- ity Crique Negro, Columbia Forest, 1 9 (fmnh); 1.6 km NNE Salamanca Camp, Columbia Forest, 1 9 (fmnh). Chrotopterus was first reported in Central Amer- ica from El Salvador (Burt & Stirton, 1961). Sub- sequently, this carnivorous bat has been recorded from southern (southern Veracruz, northern Oa- xaca, Chiapas) and peninsular (Yucatan, Quintana Roo) Mexico southeastward throughout Central America at lowland and upland elevations (40 to over 1 880 m). Chrotopterus auritus has been re- ported from Quintana Roo (Jones et al., 1 973) and El Peten (Rick, 1968; McCarthy, 1982). These specimens from southern Belize provide the first records for the country. The Belizean specimens were netted (10 April, 28 July) in an evergreen seasonal forest at ground level along a path and at a height of about 13.7m over an intermittent stream bed. Both were active during the morning hours, 0418 and 0330, re- spectively. The subspecific name Chrotopterus auritus au- ritus has been applied to Middle American pop- ulations (Jones et al., 1971). Carter and Dolan (1978) stated the type specimen for Vampyrus au- ritus Peters, 1856, actually was collected in Santa Catarina, Brazil, not in Mexico. The discussion by Carter and Dolan (1978, p. 37) suggested that Pe- ters based his description on one or more speci- mens from Brazil and compared these with a spec- imen from an unrecorded locality in Mexico as the "verwandten Art aus Mexico." Handley ( 1 966) doubted that subspecies were recognizable. Vampyrum spectrum (Linnaeus, 1758) Specimen Examined— BELIZE. Toledo: Santa Elena, 1 S (fmnh). Two localities in southern Veracruz, Mexico (Goldman, 1917; Navarro L., 1 979) are the north- westernmost records of the false vampire bat's Middle American distribution, which continues in Nicaragua (Dobson, 1 878; Allen, 1910), Costa Rica (Casebeer et al., 1963; Armstrong, 1969; Gardner et al., 1970; Howell &. Burch, 1974; Vehrencamp et al., 1977; LaVal & Fitch, 1977), and Panama (Handley, 1966; Peterson & Kirmse, 1969; Bo- naccorso, 1979). Although primarily lowland in distribution, its highest recorded elevation was about 1815m. The occurrence of Vampyrum spec- trum in the Caribbean lowlands of Belize is doc- umented by this specimen. There appears to be no definite record of this carnivorous bat from Guatemala (Jones, 1966). Dobson (1878, p. 471) recorded "Guatemala" as part of the Central American range for Vampyrum, but did not list any examined specimens. Alston (1879-1882, p. 39) stated Dobson (pers. comm.) saw specimens from Guatemala, although Alston realized the collector, O. Salvin, had not obtained specimens of Vampyrum; hence, the identification of this species is doubtful. Five false vampire bats were mist netted on three separate dates in Parque Nacional Tikal, El Peten. Two females were cap- tured during the dry season (22 and 24 March) in an upland deciduous seasonal forest, in the vicin- ity of Central Plaza of the archaeological site, and at a permanent water pool in escobal palm forest, 2.6 km SE Central Plaza. Two females and one male were netted during the wet season (22 July) at a location along an archaeological transect in escobal palm forest, 1 km SE Tikal Reservoir. All of these bats were released after being observed, measured, and/or photographed. These individ- uals provide the first record for Guatemala and, along with the specimen from Belize, bridge an intermittent distribution that now extends north- ward toward peninsular Mexico. The Vampyrum spectrum from Belize was cap- tured (8 April) during the early morning (0300) in 146 FIELDIANA: ZOOLOGY an open field. We were "trapping" Desmodus ro- tundus during a vampire bat control effort in the village. This large bat was captured after it made a number of low passes over horses and mules, which were encircled by mist nets. The bat died while enroute to captivity via an assistant. The Central American population of Vampy- rum was described as a distinct subspecies, V. s. nelsoni (Goldman, 1914), but Handley (1966) ar- gued that the species was monotypic. Subfamily GLOSSOPHAGINAE Glossophaga commissarisi commissarisi Gardner, 1962 Specimens Examined— BELIZE. Belize: Rock- stone Pond, 2 SS, 3 99 (rom). Toledo: Aguacate, 1 9 (fmnh), 1 9 (cm); Big Fall, 1 km SE Rio Grande Bridge, 2 $S (cm); Forest Home, 1 9 (fmnh); 2.8 km NNW Punta Gorda, 1 9 (fmnh). GUATE- MALA. Izabal: 25 km SSW Puerto Barrios, 7 SS, 6 99 (tcwc). Webster and Jones (1982) summarized the dis- tribution for this subspecies of nectivorous bat, which was documented from eastern (Veracruz) and southern (Oaxaca, Chiapas) Mexico and southern Belize southeastward throughout Central America. Hellebuyck et al. (1985) recently re- ported records from El Salvador. The specimens from Izabal are the first records from this Gua- temalan department. The specimens from Belize District extend northward the distribution of Glos- sophaga commissarisi along the Caribbean low- lands. According to D. C. Carter's field notes, the ma- jority of the Guatemalan Glossophaga commis- sarisi were mist netted (February, March) over a stream and in the adjacent undisturbed forest. Many of these nectivorous bats were captured in association with night-blooming "bat flowers" bordering on a stream. The Belizean specimens reported (Webster & Jones, 1982) from Lubaan- tun, Toledo District, were collected ( 1 8 April) in a disturbed semi-evergreen seasonal forest. Ad- ditional specimens were secured (January, July, September, December) in secondary and orchard vegetation of villages, in riparian secondary vege- tation, and from the hollow of a mamey tree (Pou- teria mammosa). Subfamily STENODERMATINAE Uroderma bilobatum molaris Davis, 1968 Specimen Examined— MEXICO. Quintana Roo: 2 km N, 8 km W Bacalar, 1 $ (tcwc). Davis (1968) recognized this subspecies of the tent-making bat from the Gulf-Caribbean versant of southern Veracruz, Tabasco, northeast Oaxaca, northern Chiapas, Belize, Honduras, Nicaragua, Costa Rica, and northwest Panama. Disney (1968) and Pendergast (1979) also reported the occur- rence of Uroderma bilobatum from Belize. The specimen reported here represents the first record for Quintana Roo and a marginal range extension into the Mexican peninsula of Yucatan. The above specimen was taken in a net on 6 August by M. D. Engstrom along a path leading to an inland lagoon. Vampyrops helleri helleri Peters, 1866 Specimens Examined— BELIZE. Cayo: Banana Bank, 5 99 (fmnh); 0.8 km W Macaw Bank, 1 6 (fmnh). Toledo: Big Fall, 1.9 km ENE Rio Grande Bridge, 1 9 (amnh), 1 9 (cm), 1 $ (msu); Crique Negro, Columbia Forest, 1 $ (bm); Forest Home, 1 (5 (fmnh), 1 (5 (msu); Salamanca Camp, 1 S (bm), 1 (5 (fmnh), 1 9 (usnm); 1.8 km NNE Salamanca Camp, Columbia Forest, 1 9 (fmnh); vicinity Union Camp, 1 9 (bm), 2 99 (cm). The Middle American records of this fruit bat indicate a distribution from sea level to elevations of over 1 300 m and a range from southeastern Mexico (southern Veracruz, Oaxaca, Tabasco) throughout Central America. Lowland records have been reported from El Peten (Rick, 1968) and Izabal (Carter et al., 1966). This account con- stitutes the first records from Belize. Eighty-seven percent of the Vampyrops helleri specimens were captured along or in proximity to waterways. Eleven additional individuals were re- leased at Banana Bank, where a concentration of stenodermatines (Sturnira, Uroderma, Vampyres- sa, Chiroderma, Artibeus, and Vampyrops) was observed. The remaining localities were in upland evergreen seasonal forest and in disturbed village vegetation. A specimen in the collection of St. John's College, Belize City, was collected by E. L. Tyson in Columbia Forest. I follow Dickerman et al. (1981) for the taxo- nomic assignment of the subspecific epithet. MCCARTHY: DISTRIBUTION OF BATS 147 Vampyrodes caraccioli major G. M. Allen, 1908 Specimens Examined— BELIZE. Toledo: Agua- cate, 1 (5 (CM); Big Fall, 1 .9 km ENE Rio Grande Bridge, 1 S (cm), 1 6 (fmnh); Big Fall, 2.1 km E Rio Grande Bridge, 1 S (bm); Crique Negro, Co- lumbia Forest, 1 S (bm), 1 5, 1 9 (msu); Salamanca Camp, 1 S (usnm); 1.6 km N Salamanca Camp, Columbia Forest, 1 S (fmnh); 2. 1 km NNE Sala- manca Camp, Columbia Forest, 4 66, I 9 (cm); San Antonio, 1 9 (fmnh). The published distribution of Vampyrodes car- accioli major northwestward of Costa Rica and Panama is confined to the Gulf-Caribbean low- lands as far as southern Mexico (Oaxaca, southern Veracruz, Chiapas); elevational data are less than 300 m. The records from Belize extend the range of this stenodermatine north of Izabal (Sanborn, 1936). The Belizean localities represent habitats of ri- parian lowland and upland evergreen seasonal for- ests and village secondary vegetation. The capture dates cover both the dry and wet seasons (March, April, May, July-September, December). I follow Carter and Dolan (1978) for the correct spelling of Vampyrodes caraccioli. Vampyressa pusilla thyone Thomas, 1 909 Specimens Examined— BELIZE. Cayo: 1.6 km NW Augustine, 3 66, I 9 (cm); Banana Bank, 1 9 (fmnh); Blancaneaux, 8.3 km NNE Augustine, 1 9 (fsm). Toledo: vicinity Aguacate, 1 9 (bm), 3 99 (cm); 1.2 km E Aguacate, 1 3, 1 9 (cm); Big Fall, 1 km E Rio Grande Bridge, 1 9 (cm); Big Fall, 2. 1 km E Rio Grande Bridge, 1 6 (cm); Big Fall, 1 .9 km ENE Rio Grande Bridge, 1 5, 1 9 (cm), 1 9 (fmnh); Crique Negro, Columbia Forest, 1 6 (msu), 1 6 (usnm); Forest Home, 1 6 (msu); Pueblo Viejo, 1 9 (fmnh); 1 .6 km NNE Salamanca Camp, Co- lumbia Forest, 1 3, 2 99 (fmnh). The general distribution of the little yellow-eared bat extends from southern (Oaxaca, southern Ve- racruz, Chiapas) and peninsular (Campeche) Mex- ico and continues southeastward along the Carib- bean slope to both the Pacific and Caribbean corridors of southern Nicaragua, Costa Rica, and Panama, into South America. Elevational data are primarily lowland, from sea level up to a recorded 2200 m. Peterson (1966) reported the only record of Vampyressa pusilla in Belize, from Rockstone Pond, Belize District. There are also previous rec- ords from El Peten (Rick, 1 968) and southeastern Campeche (Jones et al., 1973). This account pro- vides additional records of this species. These specimens of Vampyressa pusilla were collected (February-May, July-September, De- cember) in moist habitats, the majority of which were associated directly with riparian vegetation or in village and pasture-edge vegetation situated near rivers. Evergreen seasonal forest provided an upland habitat. Chiroderma villosum jesupi J. A. Allen, 1900 Specimens Examined— BELIZE. Cayo: Banana Bank, 1 3, 5 99 (fmnh). Corozal: Chan Chen, 1 6 (fmnh). Toledo: Big Fall, vicinity Rio Grande Bridge, 1 6 (fmnh); Big Fall, 1 .7 km NE Rio Grande Bridge, 1 9 (msu); Big Fall, 1 .9 km ENE Rio Grande Bridge, 1 3, 1 9 (cm); San Antonio, 1 6 (fmnh); 1 km WNW San Pedro Columbia, 1 9 (fmnh). GUA- TEMALA. EI Peten: Parque Nacional Tikal, 1 6 (fmnh). The Middle American occurrence of Chiroder- ma villosum has been documented in southern (Oaxaca, southern Veracruz, Chiapas) and pen- insular (Campeche, Quintana Roo) Mexico, Gua- temala, Nicaragua, Costa Rica, and Panama. Hel- lebuyck et al. (1985) recently reported this fruit bat from El Salvador. Locality records reach from the coastal lowlands to upland habitats at 1 300 m. Southeastern Campeche (Jones et al., 1973) and northern Quintana Roo (Bimey et al., 1974) are previous Caribbean lowland localities, in addition to these first records from Belize and El Peten. All but one of the Belizean Chiroderma were associated either directly with or in the vicinity of riparian evergreen or semi-evergreen seasonal for- ests (April, May, August, September, December). One individual was captured (15 November) in village orchard vegetation. Five additional indi- viduals were released at Banana Bank. The Tikal specimen was captured (24 March) along the per- manent water pool mentioned in the Tonatia ev- otis account. Artibeus toltecus toltecus (Saussure, 1 860) Specimens Examined— BELIZE. Cayo: vicinity Augustine, 2 66, 4 99 (fsm); 1 .6 km NW Augustine, Rio Frio, 1 3, 1 9 (fmnh), 5 66 (ttu), 4 66 (cm); "Rio On," ? km N Augustine, 1 9 (ttu); 1.1 km S Baldy Beacon, Bald Hills, 3 99 (cm); vicinity San Luis, 7.1 km SSW Augustine, 1 9 (ttu). Toledo: 148 FIELDIANA: ZOOLOGY Orange Point, 1 2 (fmnh); Pueblo Viejo, 3 9$ (fmnh); Union Camp, 5 S6, 4 92 (cm). In his revision of the small Artibeus of Middle America, Davis (1969) recognized the range of Artibeus toltecus toltecus from southern Tamau- lipas, Mexico, southeastward along the mountain- ous region of the Gulf versant, upland of southern Mexico, Guatemala, Honduras, Nicaragua, and Costa Rica. He did not examine Panamanian spec- imens. Handley (1966) summarized the Pana- manian localities for /I. toltecus. This bat primarily occurs at elevations between 328 and 1640 m, although elevations near sea level were recorded (Davis, 1969). Consequently, the occurrence of ^. toltecus in the Maya Mountain range of southern Belize and southeastern El Peten was not unex- pected. These Belizean localities represent the first northern Caribbean lowland records. The Belizean localities range in elevation from near sea level to approximately 720 m. Artibeus toltecus is more common at the higher elevations. These dark-colored Artibeus were captured (De- cember-February, April, June, September) in hab- itats of deciduous seasonal forest, semi-evergreen seasonal forest, transitional forest, and pine forest- savanna. The subspecies toltecus is applied, based on the proximity of Belize to its distribution as defined by Davis (1969). Centurio senex senex Gray, 1 842 Specimens Examined— BELIZE. Belize: 1 .4 km S San Pedro, Ambergris Caye, 1 3, 1 9 (fmnh). Cayo: 1.6 km NW Augustine, Rio Frio, 1 <5 (ttu); vicinity Augustine, Rio On, 1 9 (ttu); Blanca- neaux, 8.3 km NNE Augustine, 1 9 (fsm); Central Farm, 1 3, 1 9 (fmnh); Teakettle, Young Gal Road at Belize River, 1 3, 1 9 (fmnh); Xunantunich, 1 $ (fmnh). Corozal: 1.2 km E, 1.6 km N Corozal, 1 (5 (LSUMZ). Orange Walk: 1.6 km NW San An- tonio, Rio Hondo, 1 9 (fmnh). Toledo: Big Fall, 1 .9 km ENE Rio Grande Bridge, 1 S (cm); Crique Negro, Columbia Forest, 1 5, 1 9 (usnm); Forest Home, 1 9 (amnh); vicinity Union Camp, 2 99 (BM), 1 9 (CM). GUATEMALA. Alta Verapaz: Lan- quin, vicinity Lanquin Cave, approx. 149 km WSW Puerto Barrios, 1 $ (amnh). Izabal: 25 km SSW Puerto Barrios, 1 3, 5 99 (tcwc). The recorded distribution of the wrinkle-faced bat extends from western (southern Sinaloa), northeastern (southern Tamaulipas), and penin- sular (Campeche and Quintana Roo) Mexico and continues southeastward through Central America at principally lower to upland elevations (sea level to 1882 m). The records given here are the first for Belize, Alta Verapaz, and Izabal. The distribution of this unusual bat in Belize reflects apparent ecological flexibility. Centurio se- nex has been captured in low littoral forest and mangrove swamp edge on the coastal sand strip of Ambergris Caye, to about 720 m in evergreen and semi-evergreen seasonal forest on the south- ern slope of the Maya Mountains. Evergreen sea- sonal and transitional forests, secondary forest, and agriculturally disturbed areas provide additional habitats. This bat was captured throughout the year. Two males and one female were mist netted and released at Orange Point, Toledo District. Brother N. Sullivan collected (15-17 January) the specimen from Alta Verapaz, but I assume the bat was captured outside of Lanquin Cave. The spec- imens from Izabal were obtained (February, March) by D. C. Carter and field party. Field data are limited, but four Centurio were captured over a stream. Diphylla ecaudata Spix, 1823 Specimens Examined— BELIZE. Cayo: vicinity Augustine, 1 S (rom); San Antonio, 1 6 (fmnh). Toledo: Crique Jute, 1 $ (amnh); San Antonio, 1 9 (fmnh); Santa Elena, 1 9 (fmnh). The distribution of Diphylla ecaudata appears primarily restricted along the Gulf side and in the Yucatan Peninsula of Mexico southeastward throughout Central America, where this bat occurs from the coastal lowlands up into the mountainous highlands (1880 m). The hairy-legged vampire bat has been recorded from El Peten (McCarthy, 1 982) and Quintana Roo (Jones et al., 1973). The spec- imens reported here are the first records from Be- lize. Four of the localities represent village environ- ments where Diphylla was captured (April, July, August, December) along with Desmodus rotundus during vampire bat control activities. Mist netting was carried out in direct immediacy to domestic livestock and homes. The feeding activities of £>/- phylla in these villages were not documented, al- though one blood meal was obtained for analysis. P. Boreham, Imperial College Field Station, En- gland, reported (in litt.) a weak precipitin reaction for a mammal host from the blood meal sample without a response for bird or reptile. It is not known if this blood meal was obtained in the vil- McCARTHY: DISTRIBUTION OF BATS 149 lage (Santa Elena). Gardner (1977) summarized the sanguivorous preference ofDiphylla as for pri- marily avian hosts. The hairy-legged vampire from Augustine was apparently taken (22 February) in a deciduous seasonal forest. Family NATALIDAE Natalus stramineus saturatus Dalquestand Hall, 1949 Specimens Examined— BELIZE. Cayo: 1.6 km NW Augustine, Rio Frio, 2S6,2 9i (fsm); 0.8 km W Augustine, 2 66, 3 22 (cm); 1.5 km N Augustine, 5 22 (cm); Sibun Camp, Hummingbird Hwy. at Silver Creek, 1 2 (fmnh). Orange Walk: Richmond Hill (Goat Hill), 8.9 km SSW Orange Walk Town, 1 2 (cm). Stann Creek: Kendal, 1 6 (fmnh). Toledo: vicinity Aguacate, 1 3, 3 22 (cm); 1.2 km E Agua- cate, 1 2 (cm); Vista Hermosa, 3.7 km WNW Punta Gorda, 8 66, 6 22 (fmnh). The northern range of Natalus stramineus sa- turatus extends from both northwestern (Sinaloa) and northeastern (Nuevo Leon) Mexico, including the Yucatan Peninsula, southeastward through Central America where the number of records for this species is noticeably reduced beyond Guate- mala to Panama. Although predominately a low- land species, elevations were recorded as high as 2400 m. The presence of the funnel-eared bat in Belize was anticipated, as it appears to be well reported throughout the Gulf-Caribbean versant. Those specimens obtained (April, August, Sep- tember) at roost sites in Belize were from caves. Other capture localities include low riparian forest and open areas bordering on forest, in orchard habitats, and alongside a building. Family VESPERTILIONIDAE Subfamily VESPERTILIONINAE Myotis elegans Hall, 1962 Specimens Examined— BELIZE. Belize: Belize City, Landivar, 1 2 (amnh), 1 3, 1 2 (fmnh), 1 6 (msu); Mussel Creek, 7.5 km W Burrell Boom, 1 6, 1 2 (fmnh). LaVal (1973a) summarized the lowland distri- bution of Myotis elegans. ranging from the Gulf (eastern San Luis Potosi, Veracruz), Pacific coastal (Chiapas), and peninsular (southeastern Cam- peche) regions of Mexico to Honduras, Nicaragua, and northeastern Costa Rica. Subsequent records were reported from the Pacific side of Costa Rica and the Caribbean lowlands of El Peten (LaVal, 1977; McCarthy, 1982). The majority of eleva- tions are less than 1 20 m, ranging to 750 m. These additional Caribbean lowland localities are the first records from Belize. Two elegant Myotis were netted (1 July) along a tractor track, in low riparian vegetation domi- nated by bamboo and thistle palms. Four indi- viduals were obtained (January, February, May, December) at a coastal locality in low vegetation bordering on disturbed mangrove (Rhizophora mangle, Avicennia germinans) habitat. Eptesicus furinalis gaumeri (J. A. Allen, 1897) Specimens Examined— BELIZE. Belize: Belize City, Landivar, 1 2 (cm). Cayo: Central Farm, 2 22 (cm), 5 66, 16 22 (fmnh), 2 66,19 (TTu); Little Vaquero Creek, 9.3 km NNW Augustine, 1 5, 1 2 (fsm); Ontario, 5.5 km W Teakettle, 1 2 (fmnh); Teakettle, 1 6 (fmnh). Corozal: Estero Lagoon, 4 km W Patchakan, 1 3, 1 2 (fmnh); Santa Clara, 1 2 (fmnh). Orange Walk: Honey Camp Lagoon, 1 6, 2 22 (fmnh); Tower Hill, B.S.I, compound, 3 66 (CM), 1 5, 4 22 (fmnh); 2 km SSW Tower Hill Bridge, 1 2 (cm). Stann Creek: Melinda, 3 22 (fmnh); Dangriga (Stann Creek), 1 <5 (usnm). Toledo: Or- ange Creek, 1.5 km S Punta Gorda, 1 2 (msu); Punta Gorda, 1 2 (msu). The Mexican distribution of Eptesicus furinalis gaumeri ranges from the western (Jalisco) and the eastern (San Luis Potosi) versants southeastward to South America. Davis (1965), Disney (1968), and Starrett and Casebeer (1968) reported records from all of the Central American countries except El Salvador. Lowland elevations range from near sea level to 1800 m, the majority being below 500 m. This tropical brown bat has been reported from El Peten (Rick, 1968; McCarthy, 1982) and Quin- tana Roo (Jones et al., 1973). The localities here are additional records for Belize. Disney ( 1 968) did not present locality data for his two specimens of Eptesicus furinalis. Both were males, captured (16 November, 29 December) in Cayo District, near Central Farm and Esperanza (4.5 km W Central Farm). These are located in British Museum (Natural History). An additional 1 96 individuals were captured from three of the localities reported here; the majority of these were ISO FIELDIANA: ZOOLOGY banded and released during a behavioral study. The majority was found in direct association with buildings, utilizing the infrastructure of the walls or floors and the space behind window shutters as roost sites. Individuals have been taken over water (creeks and a swimming pool) at three localities and in riparian vegetation along two lagoons. Lasiurus borealis (Muller, 1776) Speomens ExAMnsfED— BELIZE. Orange Walk: Tower Hill, B.S.I, compound, 1 9 (fmnh). Stann Creek: 5.3 km NNW Quam Bank, Cockscomb Basin, 1 9 (cm). GUATEMALA. El Peten: Parque Nacional Tikal, 1 9 (fmnh). The subspecies teliotis ranges southward from both the western and eastern regions of Mexico to Oaxaca and the northern Yucatan Peninsula. Specimens of Lasiurus borealis from the Guate- malan central highlands were assigned by Jones (1966) to the Central American subspeciesyra«/z/7, based on Handley (1960). Carter et al. (1966) as- signed specimens from both lowland and highland localities in Chiapas Xofrantzii, suggesting that the region of the Isthmus of Tehuantepec represents the break heXwecn front zii and teliotis. Hall (1981) concurred with this arrangement. Similarly, Jones et al. (1973) suggested that southern Mexico, in- cluding the Yucatan Peninsula, may represent a zone of intergradation between frantzii and teli- otis. Few specimens of L. borealis are available from El Salvador (Burt & Stirton, 1961), Honduras (Goodwin, 1942b), Nicaragua (Davis & Carter, 1962— as L. b. teliotis), Costa Rica (Goodwin, 1 946; Gardner et al., 1 970) and Panama (Handley, 1 966). Recorded elevations (near sea level to about 2540 m) are primarily low or moderate (< 1 155 m). Koopman (1959) reported the only record from Quintana Roo. This account represents the first records for Belize and eastern Guatemala from El Peten. The red bats captured in Belize (April, May) were netted over a stream and a swimming pool. The Tikal specimen was taken (30 July) while it was flying in an open area near a large man-made reservoir. I hesitate to assign a subspecific designation be- cause I see no practical purpose in doing so until adequate series of specimens from throughout the range of Lasiurus borealis become available. Han- dley (1960) had fewer specimens of L. borealis at hand for a proper evaluation of subspecific vari- ation. Consequently, the limits of the distributions for the recognized subspecies remain unresolved. Lasiurus ega (Gervais, 1855) Specimens Examined— BELIZE. Belize: Trop- ical Park, Mi. 14.5 Western Hwy., 1 S (fmnh). Orange Walk: Tower Hill, B.S.I, compound, 2 6i (FMNH), 1 (3 (cm). Stann Creek: 5.3 km WNW Quam Bank, Cockscomb Basin, 2 66, I 9 (cm). Toledo: Big Fall, 1.7 km NE Rio Grande Bridge, 1 9 (cm); Orange Creek, 1 .5 km SW Punta Gorda, 1 6 (msu). Similar to Lasiurus borealis, the distribution for the two recognized subspecies of the yellowish bat is not well understood. While L. e. panamensis was recognized along the Pacific versant of Chia- pas (Baker &. Patton, 1 967) and Guatemala (Dolan & Carter, 1979;Dickermanetal., 1981), Goodwin ( 1 969) identified panamensis from the moderate elevations of the Gulf drainage in northern Oaxaca and suspected L. e. xanthinus may occur in the drier Pacific portion of that state. Baker et al. ( 1 97 1 ) determined the variation in karyotypes and pelage color of L. ega from near Brownsville, Texas, re- sembled those from eastern coastal and southern Mexico and referred the Texas specimens to L. e. panamensis. Meanwhile, L. e. xanthinus was rec- ognized in the Yucatan Peninsula (Jones et al., 1973; Bimey et al., 1 974). The yellow bat is poorly represented from the remainder of Central Amer- ica, which includes Honduras (Goodwin, 1942b; LaVal, 1969; Greenbaum &. Jones, 1978), Costa Rica (Goodwin, 1946; Starrett & Casebeer, 1968; Gardner et al., 1970; LaVal & Fitch, 1977), and Panama (Handley, 1966). Where designated, the subspecies panamensis has been applied to these preceding Central American localities, although Hall (1981) did not acknowledge panamensis north of Costa Rica. Elevational data are similar to those for L. borealis. Ingles (1958) reported two L. ega from Quintana Roo. Jones et al. (1973, p. 23) translated Ingles's locality from Spanish as "Puer- to Morelos" when it was actually a collection site only 16 km east of the state border with Yucatan, along the highway from Valladolid (Yucatan) to Puerto Morelos (Quintana Roo). Alvarez and Ra- mirez-P. (1972) cited an additional Caribbean lowland record from southeastern Campeche. This account provides the first L. ega records from Be- lize. Eight yellowish bats were captured (April, May) over streams, a river, and a swimming pool. Another was netted ( 1 8 August) at about 5 m above MCCARTHY: DISTRIBUTION OF BATS 151 the ground while circhng a building located in grass- sedge savanna. Lasiurus intermedius intermedius (H. Allen, 1862) Specimen Examined— BELIZE. Toledo: Crique Jute, 1 9 (cm). The range of this subspecies of the large yellow bat extends southeastward from Mexico to Hon- duras (Handley, 1960; Carter et al., 1966), El Sal- vador (Hellebuyck et al., 1985), and Guatemala (Carter et al., 1 966). Lasiurus intermedius has been recorded in Mexico from the northern Yucatan Peninsula and Chiapas northwestward to Texas along the eastern coast and to Sinaloa on the Pa- cific side. Recorded elevations range from lowland to highland (1620 m) habitats. A single specimen of L. intermedius from northern Quintana Roo (Bimey et al., 1974) provided the only record for that Mexican state. This Belizean specimen rep- resents the first record for the country. The above specimen was obtained on 30 March over the stream Crique Jute surrounded by sec- ondary vegetation. Bauerus dubiaquercus (Van Gelder, 1959) Specimens Examined— BELIZE. Cayo: 1.6 km NW Augustine, Rio Frio, 1 2 (rom). Toledo: 2. 1 km NNE Salamanca Camp, Columbia Forest, 1 6 (CM). The published localities of the rarely encoun- tered Bauerus dubiaquercus are scattered from the Islas Tres Marias (Nayarit), Jalisco, and southern Veracruz in Mexico to eastern Honduras and Cos- ta Rica (Engstrom & Wilson, 1981; Dinerstein, 1985). Mainland elevations range from approxi- mately 460 to 1450 m and appear to represent mid-elevation and montane forest habitats (Pine, 1 966; Pine et al., 1 97 1 ; Engstrom & Wilson, 1981; Dinerstein, 1985). These first occurrences of Bau- erus in Belize extend northward a scattered dis- tribution along the northern Caribbean lowlands in Central America. J. Kamslra and J. Fragoso collected (8 July) one specimen inside the main Rio Frio cave, located in a deciduous seasonal forest at approximately 410m. The second Bauerus was netted (26 March) along an open forestry track in an evergreen forest at about 180 m. Engstrom and Wilson (1981) and Martin and Schmidly (1982) evaluated the taxonomic status of Antrozous (Bauerus) dubiaquercus and con- cluded the chromosomal, cranial, postcranial, and phallic differences between this bat and Antrozous (Antrozous) pallidus were sufficient to recognize Bauerus as a distinct genus. I follow their conclu- sions and agree that the species is monotypic since the mainland sample size that previously was as- signed to A. d. meyeri Pine, 1971, was limited to a total of five specimens representing both sexes. Family MOLOSSIDAE Eumops auripendulus auripendulus (Shaw, 1 800) Specimens Examined— BELIZE. Orange Walk: Orange Walk Town, 1 ? (cm); Tower Hill, B.S.I, compound, 1 9 (fmnh). The recorded distribution of Eumops auripen- dulus auripendulus includes both moist uplands and drier lowland coastal and plateau areas, rang- ing from eastern Oaxaca, Tabasco, Quintana Roo, and Belize, through Guatemala, western Hondu- ras, El Salvador, western Nicaragua, Costa Rica, and Panama, into South America (Eger, 1974; Greenbaum & Jones, 1978). Villa-R. (1956) and Eger (1974), respectively, reported this free-tailed bat from Quintana Roo and Belize (Belize District: Rockstone Pond). This account provides the sec- ond and third records for Belize. The Orange Walk specimen consists of a man- dible and partial skull, which were recovered from an owl {Tyto alba) roost in a church tower. The second specimen was discovered (July) alive by L. G. Hoevers, after it apparently was attacked by a bird. Eumops bonariensis nanus (Miller, 1900) Specimens Examined— BELIZE. Orange Walk: Orange Walk Town, 2 ?? (cm). Eger (1977) summarized the few available Mid- dle American localities for this small mastiff bat, which are limited to southeastern Mexico (south- em Veracruz, Tabasco, Yucatan), eastern Hon- duras, and Panama. These and additional locali- ties in Panama (Dolan & Carter, 1979) and Nicaragua (Hall, 1981) are restricted to coastal lowland environments. This is the first recording of Eumops bonariensis for Belize. 152 HELDIANA: ZOOLOGY Entire specimens oi Eumops bonariensis as yet are unavailable from Belize. Documentation is based on two fragmented sets of maxillary tooth- rows, which were sifted from regurgitated rubble beneath an owl {Tyto alba) roost in a church tower. Moiossus ater nigricans (Miller, 1 902) Specimens Examined— GUATEMALA. Izabal: 25 km SSW Puerto Barrios, 2 33, 4 99 (tcwc). The black mastiff bat is a common inhabitant of roof spaces throughout its lowland Middle American range, from western (Sinaloa) and east- em (Tamaulipas) Mexico southeastward into South America. This species has been reported from the Caribbean lowlands of Quintana Roo (Jones et al., 1973) and Belize (Murie, 1935; Pendergast, 1979). These Moiossus ater from Izabal are the first rec- ord for that department. Apparently, all of these specimens were collected on 1 5 February over a stream by D. C. Carter and his field party. Addendum While this paper was in press, other papers and additional information concerning bats in Belize came to my attention. Two recent papers provide new country records of the glossophagines Li- chonycteris obscura (Hill, 1985) and Hylonycteris underwoodi (McCarthy & Blake, 1987), which increase the known bat fauna to 68 species. Both records are from Toledo District. McCarthy and Blake (1987) also reported the occurrence of the following bats: Rhynchonycteris naso, Saccopteryx bilineata, Balantiopteryx io, Noctilio leporinus, Pteronotus parnellii, Micronyctehs megalotis, M. nicefori, M. schmidtorum, Tonatia evotis, Mimon cozumelae, Phyllostomus discolor, Trachops cir- rhosus, Chrotopterus auritus, Vampyrum spec- trum, Glossophaga soricina, Carollia brevicauda, C. perspicillata, Sturnira lilium, Uroderma bilo- batum, Vampyressa pusilla, Artibeus jamaicensis, A. lituratus, A. phaeotis, A. watsoni, Centurio se- nex, Natalus stramineus, Rhogeessa tumida, Bau- erus dubiaquercus, and Eumops underwoodi. Hill, J. E. 1 985. The status of Lichonycteris degener Miller, 1931 (Chiroptera: Phyllostomidae). Mam- malia, 49(4): 579-582. McCarthy, T. J., and M. Blake. 1 987. Noteworthy bat records from the Maya Mountains Forest Re- serve, Belize. Mammalia, 51(1): 161-164. A bat specimen from Stann Creek was reported (Miller & Allen, 1928, p. 180) as Myotis nigricans nigricans. I examined this specimen at USNM and found it to be an immature Eptesicus furinalis. It is included in this paper under the species account for the latter species. Miller, G. S., Jr., and G. M. Allen. 1928. The American bats of the genera Myotis and Pizonyx. Bulletin of the United States National Museum, 144: 1-218. Silva-Taboada and Koopman (1964, p. 3) re- ported specimens of Tadarida laticaudata (= Nyc- tinomops laticaudatus) from Corozal District. Most of the bat species discussed in an unpublished dis- sertation by A. M. Cartwright were also reported by Kirkpatrick et al. (1975) and Cartwright and Kirkpatrick (1977). The remaining identifications (Cartwright, 1977, pp. 240, 242-246, 250, 251), which were from Belize District, included Rhyn- chonycteris naso, Saccopteryx bilineata, Carollia brevicauda, C. perspicillata, Sturnira lilium, Arti- beus lituratus, A. phaeotis, Des modus rotundus, Eptesicus furinalis, Rhogeessa tumida, and Mo- iossus moiossus. Baker et al. (1985, p. 236) re- ported cytogenetic data from specimens of Rho- geessa tumida that I collected in Belize District. Baker, R. J., J. W. Bickham, and M. L. Arnold. 1985. Chromosomal evolution in Rhogeessa (Chi- roptera: Vespertilionidae): Possible speciation by centric fusions. Evolution, 39(2): 233-243. Cartwright, A. M. F. 1977. Patterns of Neotrop- ical chiropteran reproduction including histological and ecological aspects of bats collected in Belize. Ed.D. diss.. Ball State University, Muncie, Ind., 278 pp. Silva-Taboada, G., and K. F. Koopman. 1964. Notes on the occurrence and ecology of Tadarida laticaudata yucatanica in eastern Cuba. American Museum Novitates, 2175: 1-6. A mammal checklist was included in a resource paper on Belize (Hartshorn et al., 1984). The list of bats supposedly was a compilation of known and expected species. The result is inaccurate and undocumented. The reader is referred to the checklist in the Appendix as correct. Hartshorn, G., et al. 1984. Belize. Country En- vironmental Profile. A Field Study. United States MCCARTHY: DISTRIBUTION OF BATS 153 Agency for International I>evelopment, San Jose, Costa Rica. 151 pp. Additional specimens of Micronycteris mega- lotis (2: Belize District, Cayo District), Artibeiis toltecus (11: Cayo District), Centurio senex (2: Cayo District), and Diphylla ecaudata (3: Cayo District) were found in the mammal collection of Royal Ontario Museum. D. J. Tallman collected speci- mens of M. megalotis ( 1 ) and Mimon cozumelae (2) from Orange Walk District, which were de- posited in Bell Museum of Natural History, Uni- versity of Minnesota. I secured further voucher specimens (AMNH) of Mimon cozumelae (1), Vampyrodes caraccioli ( 1 ), Vampyressa pusilla ( 1 ), and Bauerus dubiaquercus (1) from Toledo Dis- trict. Certain specimens (Sturnira lilium and Rho- geessa tumidd) that were catalogued by Dobson (1878, pp. 540, 246) were listed as collected in "Honduras." These were obtained by D. Dyson and H. Cuming between November 1844 and late 1845. During that time, "Honduras" correspond- ed to the present region that extends from southern Quintana Roo, Mexico, southeastward to northern Honduras. Many early collectors did not differ- entiate between the area of Belize ("British settle- ment in Honduras") and that of the Republic of Honduras ("Spanish Honduras"), but recorded only "Honduras" or "Bay of Honduras" without further locality data. The above specimens did not originate from present day Honduras, but were collected in Belize. Additional specimens of Mi- cronycteris megalotis (Dobson, 1 878, p. 479) from the "Bay of Honduras" and Rhynchonycteris naso (Dobson, 1878, p. 368) from "Honduras" remain orphaned records of the historical literature. Uroderma bilobatum was reported (Sanchez-H. et al., 1986) from southern Quintana Roo while this volume was delayed. Ten specimens were col- lected at Ruinas de Kohunlich (18°23'N; 88''42'W), about 16 km W Estevez on the Belizean border. Four other species {Pteronotus davyi, Mormoops megalophylla, Tonatia evotis, T. minuta) were documented for the first time from Quintana Roo, from localities within 35 km of the northern bor- der of Belize. The known bat fauna of Quintana Roo is now represented by 36 species. Sanchez-H., O., G. Tell£z-G., R. A. Medelun, and G. Urbano-V. 1986. New records of mammals from Quintana Roo. Mexico. Mammalia, 50(2): 275- 278. Acknowledgments Grateful acknowledgments are due H. Flowers, O. Rosado, R. Belisle, and E. O. Bradley, E>e- partment of Forestry, Belize, and H. Topsey, W. Branche, E. A. Graham, and J. Palacio, Depart- ment of Archaeology, Belize, for the necessary per- mission to conduct fieldwork. The Department of Agriculture, Belize, provided opportunities and support while I headed the Vampire Bat Education and Control Program (1975-1978). Division of Mammals, Field Museum of Natural History, pro- vided partial equipment needs (1974) and partial funding ( 1979). Mellon North American Mammal Research Institute, O'Neil Fund, and Section of Mammals, Carnegie Museum of Natural History, supported further fieldwork (1982, 1984). Nu- merous colleagues and friends have been of direct assistance to my needs in Belize, and the following are but a few: E. Awe, P. Boreham, H. Bums, W. D. Burton, J. Cab, G. Cal, G. C. Canto, J. Chun, M. Craig, L. de la Torre, L. Dieckman, R. Foster, P. W. Freeman, P. Gamble, E. Garcia, J. P. and M. Garcia, H. H. Genoways, E. Gomez, L. Hen- derson, R. W. Henderson, L. G. and E. Hoevers, A. M. Hutson, R. J. Izor, T. Juring, E. King, K. Leacock, D. Owen-Lewis, N. MacKenzie, G. T. McCarthy, H. Pastor, R. H. Pine, A. Rabinowitz, M. L. Reed, C. J. Rushin, K. and T. Salisbury, R. Schmidt, B. M. Silva, M. F. Valentine, L. Waight, L. Wilkins, and A. C. S. Wright. Various regiments and the Royal Air Force of Her Majesty's British Forces provided certain logistical support over the years. My fieldwork in Parque Nacional Tikal, El Pe- ten, Guatemala, was made possible while assisting D. J. Howell and J. G. Cant during their respective visits to the park. M. Dary-R., Universidad de San Carlos, and F. Polo-Sifontes, Instituto de An- tropologia y Historia, graciously assisted and per- mitted my work in 1979. J. R. Martinez R. kindly provided a copy of his thesis. The courtesies rendered to me by the personnel of the following institutions while providing either information, the loan of specimens, or assistance during visits are greatly appreciated: American Museum of Natural History; British Museum (Natural History); Carnegie Museum of Natural History; Field Museum of Natural History; Flor- ida State Museum, University of Florida; Museum of Natural History, University of Kansas; Mu- seum of Zoology, Louisiana State University; The Museum, Michigan State University; Royal On- tario Museum; Texas Cooperative Wildlife Col- 154 HELDIANA: ZOOLOGY lection, Texas A «& M University; The Museum, Texas Tech University; United States National Museum of Natural History. I am grateful to N. A. Bitar, F. J. Bonaccorso, D. C. Carter, M. D. Engstrom, A. M. Hutson, J. Kamstra, R. L. Pe- terson, and E. L. Tyson for sharing information concerning their field efforts in either Belize, El Peten, Izabal, or Quintana Roo. Besides J. R. Choate, J. K. Jones, Jr., K. F. Koopman, and R. M. Timm, two anonymous re- viewers enhanced earlier manuscripts with their comments and editorial skills. E. Mendez kindly reviewed my Spanish abstract. Verification of the state boundaries in the Yucatan Peninsula was provided by T. Dachtera, National Geographic Society, and G. de la Torre M., Secretaria de Pro- gramacion y Presupuesto, Mexico. M. A. Schmidt, R. A. Rollin, V. Risoli, and S. Cramer typed var- ious versions of the manuscript. S. M. Velez guid- ed the manuscript during my absence. This report is a contribution of the Mammals of Belize Pro- gram. Literature Cited Allen, J. A. 1910. Additional mammals from Nica- ragua. Bulletin of the American Museum of Natural History, 28:87-115. Alston, E. R. 1879-1882. Biologia Centrali-Ameri- cana. Mammalia. Taylor and Francis, London, xx + 220 pp., 22 pis. Alvarez, T., and J. Ramirez-P. 1972. Notas acerca de murcielagos mexicanos. Anales de la Escuela de Ciencias Biologicas, Mexico, 19: 167-178. Andersen, K. 1908. A monograph of the chiropteran genera Uroderma, Enchisthenes, and Artibeus. Pro- ceedings of the Zoological Society of London, 1908: 204-319. Armstrong, D. M. 1969. Noteworthy records of bats from Costa Rica. Journal of Mammalogy, 50(4): 808- 810. Baker, R. J., and J. K. Jones, Jr. 1975. Additional records of bats from Nicaragua, with a revised check- list of Chiroptera. Occasional Papers of The Museum, Texas Tech University, 32: 1-13. Baker, R. J., T. Mollhagen, and G. Lopez. 1971. Notes on Lasiunis ega. Journal of Mammalogy, 52(4): 849-852. Baker, R. J., and J. L. Patton. 1 967. Karyotypes and karyotypic variation of North American vespertilion- id bats. Journal of Mammalogy, 48(2): 270-286. Beard, J. S. 1 944. Climax vegetation in tropical Amer- ica. Ecology, 25(2): 127-158. Birney, E. C, J. B. Bowles, R. M. Timm, and S. L. Williams. 1974. Mammalian distributional records in Yucatan and Quintana Roo, with comments on reproduction, structure, and status of peninsular pop- ulations. Occasional Papers, Bell Museum of Natural History, University of Minnesota, 13: 1-25. Bonaccorso, F. J. 1979. Foraging and reproductive ecology in a Panamanian bat community. Bulletin of the Florida State Museum, Biological Sciences, 24: 359-408. Burt, W. H., andR. A. Stirton. 1961. The mammals of El Salvador. Miscellaneous Publications of the Mu- seum of Zoology, University of Michigan, 117: 1-69. Carter, D. C, and P. G. Dolan. 1978. Catalogue of type specimens of Neotropical bats in selected Euro- pean museums. Special Publications of The Museum, Texas Tech University, 15: 1-136. Carter, D. C, R. H. Pine, and W. B. Davis. 1966. Notes on Middle American bats. Southwestern Nat- uralist. 11(4): 488-499. Cartwright, a. M., and R. D. Kjrkpatrick. 1977. A range extension of Peropteryx kappleri (family Em- ballonuridae) in Central America. Proceedings of the Indiana Academy of Sciences, 86: 466. Casebeer, R. S., R. B. Linskv, and C. E. Nelson. 1 963. The phyllostomid bats, Ectophylla alba and Vampy- rum spectrum, in Costa Rica. Journal of Mammalogy, 44(2): 186-189. Davis, W.B. 1965. Re\iev/ ofxheEptesicusbrasiliensis complex in Middle America with the description of a new subspecies from Costa Rica. Journal of Mam- malogy, 46(2): 229-240. . 1 968. Review of the genus Uroderma (Chirop- tera). Journal of Mammalogy, 49(4): 676-698. 1969. A review of the small fruit bats (genus Artibeus) of Middle America. Southwestern Natural- ist, 14(1): 15-29. 1970a. A review of the small fruit bats (genus Artibeus) of Middle America. Part II. Southwestern Naturalist, 14(4): 389-402. 1970b. The large fruit bats (genus Artibeus) of Middle America, with a review of the Artibeus Ja- maicensis complex. Journal of Mammalogy, 51(1): 105-122. 1973. Geographic variation in the fishing bat. Noctilio leporinus. Journal of Mammalogy, 54(4): 862- 874. 1984. Review of the large fruit-eating bats of the Artibeus "" lituratus"" complex (Chiroptera: Phyl- lostomidae) in Middle America. Occasional Papers of The Museum, Texas Tech University, 93: 1-16. Davis, W. B., and D. C. Carter. 1962. Notes on Central American bats with description of a new sub- sf)ecies of Mormoops. Southwestern Naturalist, 7(1): 64-74. . 1978. A review of the round-eared bats of the Tonatia silvicola complex, with descriptions of three new taxa. Occasional Papers of The Museum, Texas Tech University, 53: 1-12. Davis, W. B., D. C. Carter, and R. H. Pine. 1964. Noteworthy records of Mexican and Central American bats. Journal of Mammalogy, 45(3): 375-387. Dickerman, R. W., K. F. Koopman, and C. Seymour. 1981. Notes on bats from the Pacific lowlands of Guatemala. Journal of Mammalogy, 62(2): 406-41 1. MCCARTHY: DISTRIBUTION OF BATS 155 DiNERSTEiN. E. 1985. First records of Lasiurus casta- neus and Antrozous dubiaquercus from Costa Rica. Journal of Mammalogy, 66(2): 41 1—412. Disney. R. H. L. 1 968. Observations on a zoonosis: Leishmaniasis in British Honduras. Journal of Ap- plied Ecology . 5( 1 ): 1-59. DoBSON. G. E. 1878. Catalogue of the Chiroptera in the Collection of the British Museum. British Mu- seum, London, xlii -^ 567 pp.. 30 pis. OoL\N, P. G., AND D. C. Carter. 1979. Distributional notes and records for Middle American Chiroptera. Journal of Mammalogy, 60(3): 644-649. Eger. J. L. 1 974. A new subspecies of the bat Eumops auhpendulus (Chiroptera: Molossidae). from Argen- tina and eastern Brazil. Life Sciences Occasional Pa- pers, Royal Ontario Museum, 25: 1-8. . 1977. Systematics of the genus fumops (Chi- roptera: Molossidae). Life Sciences Contributions, Royal Ontario Museum, 110: 1-69. Engstrom, M. D., and D. E. Wilson. 1981. System- atics oi Antrozous dubiaquercus (Chiroptera: Vesper- tilionidae), with comments on the status of Bauerus Vjui Gelder. Annals of Carnegie Museum, 50( 1 5): 37 1- 383. Felten, H. 1956. Fledermause (Mammalia, Chiroi>- tera) aus El Salvador. Senckenbergiana Biologica, 37(3- 4): 179-212. Fleming, T. H., E. T. Hooper, and D. E. Wilson. 1 972. Three Central .\merican bat communities: Structure, reproductive cvcles, and movement patterns. Ecology-. 53(4): 555-569. G.ardner, a. L 1 976. The distributional status of some Peruvian mammals. Occasional Papers of the Mu- seum of Zoolog>-, Louisiana State Universitv, 48: 1- 18. . 1977. Feeding habits, pp. 293-350. In Baker. R. J., J. K. Jones, Jr., and D. C. Carter, eds.. Biology of bats of the New World family Phyllostomatidae. Part IL Special Publications of The Museum, Texas Tech University, 13: 1-364. Gardner, A. L.. R. K. LaVal, and D. E. Wilson. 1 970. The distributional status of some Costa Rican bats. Journal of Mammalog>. 51(4): 712-729. Gaumer, G. F. 1917. Monografia de los mamiferos de Yucatan. Departamento de Talleres Graficos de la Secretaria de Fomento, Mexico, xli -^ 33 1 pp.. 1 map. Goldman, E. A. 1914. A new bat of the genus Mimon from Mexico. Proceedings of the Biological Society of Washington, 27: 75-76. . 1917. New mammals from North and Middle .\merica. Proceedings of the Biological Society of Washington, 30: 107-1 16. 1925. A new bat of the genus Trachops from . 1942b. Mammalsof Honduras. Bulletin of the American Museum of Natural History, 79: 107-195. . 1946. Mammalsof Costa Rica. Bulletin of the Guatemala. FVoceedings of the Biological Society of Washington, 38: 23-24. Goodwin, G. G. 1 940. Three new bats from Honduras and the first record of Enchisthenes harti (Thomas) for North America. American Museum Novitates, 1075: 1-3. . 1942a. A summary of recognizable species of American Museum of Natural History, 87: 275—473. . 1969. Mammals from the state of Oaxaca, Mexico, in the American Museum of Natural History. Bulletin of the American Museum of Natural History, 141: 1-269. Greenbaum, I. F., AND J. K. Jones, Jr. 1978. Note- worthy records of bats from El Salvador. Honduras, and Nicaragua. Occasional Papers of The Museum, Texas Tech University. 55: 1-7. Haxx, E. R. 1981. The Mammals of North America, vol. 1 . John Wiley and Sons. New York, xvii + 600 + 90 pp. Handley. C. O.. Jr. 1960. Descriptions of new bats from Panama. Proceedings of the United States Na- tional Museum. 112(3442): 459-479. . 1 966. Checklist of the mammals of Panama. pp. 753-795. In Wenzel. R. L., and V. J. Tipton, eds.. Ectoparasites of Panama. Field Museum of Natural Histor>', Chicago. 861 pp. . 1976. Mammals of the Smithsonian Venezue- lan Project. Brigham Young University Science Bul- letin. Biological Series, 20(5): 1-91. 1980. Inconsistencies in formation of family- Tonatia. with descriptions of two new species. Journal of Mammalog>, 23(2): 204-209. group and subfamily-group names in Chiroptera, pp. 9-13. In Wilson. D. E.. and A. L. Gardner, eds.. Pro- ceedings Fifth International Bat Research Conference. Texas Tech Press, Lubbock, 434 pp. H.ARRISON, D. L., AND N. I*endleton. 1974. A second record of Wied's long-legged bat (Macrophyllum ma- crophyllum Schinze, 1821, Chiroptera: Phyllostomat- idae) in El Salvador, with notes on the palate, repro- duction, and diet of the species. Mammalia, 38(4): 689-693. Hellebuyck, v., J. R. Tamsitt, and J. G. Hartman. 1985. Records of bats new to El Salvador. Journal of Mammalog>, 66(4): 783-788. HERSHKOvrrz, P. 1951. Mammals from British Hon- duras. Mexico. Jamaica, and Haiti. Fieldiana: Zool- ogy. 31(47): 547-569. HowTLL, D. J. 1977. Time sharing and body parti- tioning in bat-plant pollination svstems. Nature, 270: 509-510. Howell, D. J.. AND D. BuRCH. 1974. Food habits of some Costa Rican bats. Revista de Biologia Tropical. 21(2): 281-294. Ingles. L. G. 1958. Notas acerca de los mamiferos mexicanos. Anales del Instituto de Biologia, Mexico, 29: 379-408. Jones, J. K., Jr. 1 964. Bats from western and southern Mexico. Transactions of the Kansas Academy of Sci- ence. 67(3): 509-516. . 1966. Bats from Guatemala. University of Kansas Publications. Museum of Natural History, 16(5): 439-472. Jones. J. K.. Jr.. and T. E. Lawlor. 1965. Mammals from Isla Cozumel. Mexico, with description of a new species of harvest mouse. University of Kansas Pub- lications. Museum of Natural History. 16(3): 409- 419. 156 FIELDIANA: ZOOLOGY Jones, J. K., Jr., J. D. Smith, and H. H. Genoways. 1973. Annotated checklist of mammals of the Yu- catan Peninsula, Mexico. I. Chiroptera. Occasional Papers of The Museum, Texas Tech University, 13: 1-31. Jones, J. K., Jr., J. D. Smith, and R. W. Turner. 1971. Noteworthy records of bats from Nicaragua, with a checklist of the chiropteran fauna of the country. Oc- casional Pap)ers of the Museum of Natural History, University of Kansas, 2: 1-35. Jones, J. K., Jr., P. Swanepoel, and D. C. Carter. 1977. Annotated checklist of the bats of Mexico and Central America. Occasional Papers of The Museum, Texas Tech University. 47: 1-35. Kjrkpatrick, R. D., a. M. Cartwright, J. C. Brier, AND E. J. Spicka. 1 975. Additional mammal records for Belize. Mammalia, 39(2): 330-331. KooPMAN, K. F. 1959. The zoogeographical limits of the West Indies. Journal of Mammalogy, 40(2): 236- 240. Lackey, J. A. 1970. Distributional records of bats from Veracruz. Journal of Mammalogy, 51(2): 384-385. LaVal, R. K. 1969. Records of bats from Honduras and El Salvador. Journal of Mammalogy, 50(4): 819- 822. . 1973a. A revision of the Neotropical bats of the genus Myotis. Natural History Museum of Los Angeles County, Science Bulletin, 15: 1-54. 1973b. Systematics of the genus Rhogeessa (Chiroptera: Vespertilionidae). Occasional Papers of the Museum of Natural History, University of Kan- sas, 19: 1-47. . 1977. Notes on some Costa Rican bats. Bre- nesia, 10/11:77-83. LaVal, R. K., and H. S. Fitch. 1 977. Structure, move- ments and reproduction in three Costa Rican bat com- munities. Occasional Papers of the Museum of Nat- ural History, University of Kansas, 69: 1-27. Lundell, C. L. 1934. Preliminary sketch of the phy- togeography of the Yucatan Peninsula. Carnegie In- stitute of Washington, 436: 257-321. . 1937. The vegetation of Peten. Carnegie In- stitute of Washington, 478: 1-207. Martin, C. O., AND D. J. ScHMiDLY. 1982. Taxonomic review of the pallid bat, Antrozous pallidus {Le Conte). Special Publications of The Museum, Texas Tech University, 18: 1-48. Martinez R., J. R. 1980. Deteccion de virus rabico en cerebros de murcielagos de la cuenca del Rio Po- lochic. Aha Verapaz e Izabal. Tesis de grado, Facultad de Medicina Veterinaria y Zootecnia, Universidad de San Carlos, Guatemala, 54 pp. McCarthy, T. J. 1982. Bat records from the Carib- bean lowlands of EI Peten, Guatemala. Journal of Mammalogy. 63(4): 683-685. MedellIn L., R. a. 1983. Tonatia bidens and Mimon crenulatum in Chiapas, Mexico. Journal of Mam- malogy, 64( 1 ): 150. Medellin L., R. A.. D. Navarro L.. W. B. Davis, and V. J. Romero. 1983. Notes on the biology of Mi- cronycteris brachyotis (Dobson) (Chiroptera) in south- em Veracruz, Mexico. Brenesia, 21: 7-11. Miller, G. S., Jr. 1912. A small collection of bats from Panama. Proceedings of the United States Na- tional Museum, 42(1882): 21-26. Murie, a. 1935. Mammals from Guatemala and Brit- ish Honduras. Miscellaneous Publications of the Mu- seum of Zoology, University of Michigan, 26: 1-30. Navarro L., D. 1979. Vampyrum spectrum (Chxrop- tera. Phyllostomatidae) in Mexico. Journal of Mam- malogy, 60(2): 435. Nelson, C. E. 1 965. Lonchorhina aurita and other bats from Costa Rica. Texas Journal of Science, 17: 303- 306. Ojasti, J., and O. J. Linares. 1971. Adiciones a la fauna de murcielagos de Venezuela con notas sobre los especies del genero Diclidurus (Chiroptera). Acta Biologica Venezolana, 7(4): 421-441. Pendergast, D. M. 1971. Excavations at Eduardo Quiroz Cave, British Honduras (Belize). Occasional Paper. Art and Archaeology, Royal Ontario Museum, 21: 1-123. . 1979. Excavations at Altun Ha, Belize, 1964- 1970, vol. 1. Royal Ontario Museum, Toronto. 226 pp. Pennington, T. D., and J. Sarukhan. 1968. Manual para la identificacion de campo de los principales ar- boles tropicales de Mexico. Instituto Nacional de In- vestigaciones Forestales y Organizacion de las NN.UU. para la Agricultura y la Alimentacion, Mexico, D.F., 413 pp. Peterson, R. L. 1 966. The genus Vampyressa recorded from British Honduras. Journal of Mammalogy, 46(4): 676. Peterson, R. L., and P. Kjrmse. 1 969. Notes on Vam- pyrum spectrum, the false vampire bat, in Panama. Canadian Journal of Zoology, 47(1): 140-142. Pine, R. H. 1966. Baeodon meyeri (Chiroptera: Ves- pertilionidae) a new species of bat from Veracruz. Southwestern Naturalist, 11(2): 308-310. . 1972. The bats of the genus Cara///a. Technical Monograph of the Texas Agricultural Experimental Station, Texas A & M University, 8: 1-125. Pine, R. H., D. C. Carter, and R. K. LaVal. 1971. Status of Bauerus Van Gelder and its relationships to other nyctophiline bats. Journal of Mammalogy, 52(4): 663-669. Portig, W. H. 1 976. The climate of Central America, pp. 405-478. In Schwerdtfeger, W.. ed.. Climates of Central and South America. Elsevier Scientific Publ. Co., New York. 532 pp. Power, D. M., and J. R. Tamsitt. 1973. Variation in Phyllostomus discolor (Chiroptera: Phyllostomatidae). Canadian Journal of Zoology. 51(4): 461-468. QuiNONES, E., J. P. Koplan, L. Pike, F. Staine, and L. Ajello. 1978. Histoplasmosis in Belize, Central America. American Journal of Tropical Medicine and Hygiene. 27(3): 558-561. Rick, A. M. 1968. Notes on bats from Tikal, Guate- mala. Journal of Mammalogy, 49(3): 516-520. Ruiz, R. G. 1983. Mimon crenulatum keenani (Chi- roptera: Phyllostomidae) from Belize. Southwestern Naturalist, 28(3): 374. MCCARTHY: DISTRIBUTION OF BATS 157 Salvin, O. 1 864. A fortnight amongst the sea birds of British Honduras. Ibis. 1864: 372-387. Sanborn. C.C. 1935. New mammals from Guatemala and Honduras. Field Museum of Natural History, Zo- ology Series. 20( 1 1): 81-85. . 1936. Records and measurements of Neotrop- ical bats. Field Museum of Natural History, Zoology Series, 20(13): 93-106. 1937. American bats of the subfamily Embal- lonurinae. Field Museum of Natural History. Zoology Series. 20(24): 321-354. 1 94 1 . Descriptions and records of Neotropical bats, pp. 371-387. In Papers on mammalogy. Pub- lished in honor of Wilfred Hudson Osgood. Field Mu- seum of Natural History, Zoological Series, 27: 1-395. 1 949. Bats of the genus Micronycteris and its subgenera. Fieldiana: Zoology, 31(27): 215-233. Sanderson, I. T. 1941. Living Treasure. The Viking Press. New York, 290 pp. ScHALDACH, W. J., Jr. 1964. Notas breves sobre al- gunos mamiferos del sur de Mexico. Anales del In- stituto de Biologia. Mexico, 35: 129-137. Smith, J. D. 1972. Systematics of the chiropteran fam- ily Mormoopidae. Miscellaneous Publications of the Museum of Natural History, University of Kansas, 56: 1-132. Standley, p. C, and S. J. Record. 1936. The forests and flora of British Honduras. Field Museum of Nat- ural History. Botanical Series, 12: 1-432. Starrett, a. 1976. Comments on bats newly recorded from Costa Rica. Natural History Museum of Los Angeles, Contributions in Science, 277: 1-5. Starrett, A., AND R. S. Casebeer. 1968. Records of bats from Costa Rica. Los Angeles County Museum, Contributions in Science, 148: 1-21. Taibel, a. M. 1977. Mammiferi del Guatemala, con speciale riguardo alia regione del Petdn, raccolti dal Maggio al Settembre 1932. Atti Socieia Ilaliana di Scienze Naturali, Museo Civico di Storia Nalurale di Milano, 118(3-4): 379-401. Valdez, R., AND R. K. LaVal. 1971. Records of bats from Honduras and Nicaragua. Journal of Mammal- ogy, 52(1): 247-250. Vehrencamp, S. L., F. G. Stiles, and J. W. Bradbury. 1977. Observations on the foraging behavior and avi- an prey of the Neotropical carnivorous bat, Vampy- rum spectrum. Journal of Mammalogy, 58(4): 469- 478. ViLLA-R., B. 1956. Otros murcielagos nuevos para la fauna de Mexico. Anales del Institute de Biologia, Mexico, 26: 543-545. . 1966. Los murcielagos de Mexico. Universi- dad Nacional Autonoma de Mexico, Institute de Bio- logia, Mexico, D.F., xvi + 491 pp. Wadell, H. a. 1938. Physical-geological features of Peten, Guatemala. Carnegie Institute of Washington, 437: 336-348. Walker, S. H. 1973. Summary of climatic records for Belize. Land Resources Division. Surrey. England. Supplementary Report. 3: 1-145. Webster, W. D., and J. K. Jones, Jr. 1982. A new subspecies of Glossophaga commissarisi (Chiroptera: Phyllostomidae) from western Mexico. Occasional Pa- pers of The Museum, Texas Tech University, 76: 1-6. West, R. C. 1 964. Surface configuration and associated geology of Middle America, pp. 33-83. In West, R. C, ed.. Handbook of Middle American Indians, vol. 1. University of Texas Press, Austin, 570 pp. Wright, A. C. S., D. H. Romney, R. H. Arbuckle, and V. E. Vial. 1959. Land in British Honduras. Report of the British Honduras Land Use Survey Team. Co- lonial Research Publication, London, 24: 1-327. Gazetteer The numbers for localities are plotted in Figure 1 . MEXICO QUINTANA ROO GUATEMALA El PETfeN 1. Bacalar 18°43'N; 88°22'W 2. Parque Nacional Tikal 1 7°20'N; 89°39'W 3. Poptun 16°2rN; 89''26'W 158 FIELDIANA: ZOOLOGY Fig. 1. Distribution of localities in the districts of Belize; Quintana Roo, Mexico; and El Peten and Izabal, Guatemala. The numbers refer to those listed in the Gazetteer. This map does not display the Caribbean lowlands of Guatemala and Mexico in their entirety. MCCARTHY: DISTRIBUTION OF BATS 159 IZABAL 4. Puerto Barrios 15°43'N; 88°36'W BELIZE CoROZAL District 5. Chan Chen 18°26'N; 88»27'W 6. Corozal 18°24'N; 88°24'W 7. Patchakan 18°24'N; 88''29'W 8. Santa Clara 18°18'N; 88°30'W Orange Walk District 9. Honey Camp Lagoon 18°03'N; 88°27'W 10. Orange Walk Town 18°05'N; 88°34'W 11. San Antonio, Rio Hondo 18°1 1'N; 88°39'W 12. Tower Hill, Belize Sugar Industries (B.S.I.) 18°02'N; 88''34'W Belize District 13. Belize City 17°30'N; 88°12'W 14. Churchyard 17°18'N; 88°33'W 15. Mussel Creek 17°39'N; 88°24'W 16. Rockstone Pond (Altun Ha) 17°46'N; 88°22'W 17. San Pedro, Ambergris Caye 17°55'N; 87°58'W 18. Tropical Park 17°28'N; 88''23'W 16°47'N: 30. Teakettle 17°13'N; 88°5rw 31. Xunantunich 17«t)5'N; 89°08'W Stann Creek District 32. Kendal 16°49'N; 88°22'W 33. Melinda 17°00'N; 88°18'W 34. Quam Bank, Cockscomb Basin 88°28'W Toledo District 35. Aguacate 16°10'N; 89°06'W 36. Big Fall 16''15'N; 88°53'W 37. Blue Creek 16°12'N; 89°03'W 38. Crique Jute and Salamanca Camp (Forestry Camp) 16°16'N; 89°0rw 39. Crique Negro, Columbia Forest 16°17'N; 89°02'W 40. Cuevas Creek and Jacinto Creek Bridges, at Punta Gorda Road 16°09'N; 88°53'W 41. Forest Home 16'^8'N; 88°50'W 42. Nimli Punit 16°20'N; 88''48'W 43. Orange Point 16°04'N; 88°49'W 44. Pueblo Viejo 16°13'N; 89°09'W 45. Punta Gorda 16°07'N; 88°48'W 46. Rice Station (Agricultural Station) 16°08'N; 88°51'W 47. San Antonio 16°15'N; 88°02'W 48. San Lucas (deserted) 16°05'N; 89°06'W 49. San Pedro Columbia 16°17'N; 88°58'W 50. Santa Elena 16°14'N; 89'X)6'W 51. Union Camp 16°24'N; 89°08'W Cayo District 19. Augustine 16°58'N; 88°59'W 20. Baldy Beacon, Bald Hills 17°0rN; 88°47'W 21. Banana Bank 17°15'N; 88''48'W 22. Barton Creek at Western Hwy. 17°13'N; 89''57'W 23. Central Farm and Listowel 17°irN; 89°00'W 24. C.I.T.A.,Sibun River at Indian Creek 17''16'N; 88''34'W 25. Macaw Bank 17°05'N; 89'^4'W 26. Roaring Creek 17°15'N; 88''47'W 27. San Antonio 17°05'N; 89°0rW 28. San Luis 16°54'N; 89°00'W 29. SibunCamp 17°05'N; 88°39'W Appendix This district checklist of the bat fauna of Belize is based on published accounts. The citations refer to the initial taxonomic treatments of specimens. Districts are arranged from north (left) to south (right). Abbreviations are as follows: Cz = Coro- zal; OW = Orange Walk; Bz = Belize; Cy = Cayo; SC = Stann Creek; Td = Toledo. In order to give an accurate list of bats, it is necessary to present certain discrepancies that have appeared in the literature. Dobson (1878) referred to certain early sp>ecimens that may have origi- nated from Belize. One reference (Dobson, 1878, 160 FIELDIANA: ZOOLOGY p. 520) to "Half-Moon Key, Honduras" for a spec- imen of Artibeus perspicillatus (= jamaicensis) be- longs to Belize, since this specimen was collected by O. Salvin (see Salvin, 1864). Sanderson (1941, p. 228) recalled "Anoura sp." in his descriptive narrative of a visit to then British Honduras. This species was not identified (Hershkovitz, 1951) in the Sanderson bat collection. Diaemus youngi was cited from Belize (Villa-R., 1966, p. 340), but R. L. Peterson (pers. comm.) stated that the specimen in question was actually from Guyana. Specimens of 17 species of bats were listed in Disney (1968) without locality data. These specimens, which are housed in British Museum (Natural History) and Royal Ontario Museum, are all from Cayo Dis- trict. J. L. Eger (pers. comm.) identified the ques- tionable specimen of Molossus bondae in Disney ( 1 968, p. 7) as M. molossus. Quinones et al. ( 1 978, p. 559) reported six species, which I collected and identified, without the exact locality information other than "the Maya Mountains region." This locality is 1 km NW Augustine, Cayo District. Sixty-six bat species are recognized in Belize. Species Cz OW Bz Cy sc Td References Rhynchonycteris naso X Murie, 1935, pp. 17-18; Disney, 1968, p. 7 Saccopteryx bilineata X X X X Sanborn, 1937, p. 331; Hershkovitz, 1951, p. 553; Disney, 1968, p. 7; Pendergast, 1979, p. 10 Saccopteryx leptura X This paper Peropteryx kappleri X Cartwright & Kirkpatrick, 1977, p. 466 Peropteryx macrotis X Hershkovitz, 1951, p. 553 Centronycteris maximilliani X Sanborn, 1941, p. 372 Balantiopteryx io X X Kirkpatrick et al., 1975, p. 330; Cartwright & Kirkpatrick, 1977, p. 466 Diclidurus virgo X This paper Noctilio leporinus X X X X This paper Pteronotus davyi X X X Disney, 1968, p. 7; this paper Pteronotus parnellii X X X Hershkovitz, 1951, pp. 553-554; Smith, 1972, p. 74; Cartwright & Kirkpatrick, 1977, p. 466; Quinones et al., 1978, p. 559; Pendergast, 1979, p. 10 Pteronotus personatus X This paper Mormoops megalophylla X X X X This paper Micronycteris brachyotis X X This paper Micronycteris megalotis X X X This paper Micronycteris nicefori X This paper Micronycteris schmidtorum X X X This paper Lonchorhina aurita X X This paper Macrophyllum macrophyllum X X This paper Tonatia bidens ... X X This paper Tonatia evotis X X Sanborn, 1941, pp. 372-373; Disney, 1968, p. 7 Disney, 1968, p. 7; this paper Tonatia minuta X X Mimon cozumelae X X X This paper Mimon crenulatum X X Ruiz, 1983, p. 374; this paper Phyllostomus discolor X This paper Phylloderma stenops X This paper Trachops cirrhosus X X X Sanborn, 1941, p. 374; Pendergast, 1979, p. 10; this paper Chrotopterus auritns X This paper Vampyrum spectrum X This paper Glossophaga commissarisi X X Webster & Jones, 1982, p. 5; this paper Glossophaga soricina X X Villa-R., 1966, p. 231; Disney, 1968, p. 7; Cartwright & Kirkpatrick, 1977, p. 466; Howell, 1977, p. 510; Quinones el al., 1978, p. 559; Pendergast, 1979, p. 10 Continued on next page MCCARTHY: DISTRIBUTION OF BATS 161 Species Cz OW Bz Cy SC Td References CaroUia brevicauda Carollia perspicillata Sturnira lilium Uroderma bilobatum Vampyrops helleri Vampyrodes caraccioli Vampyressa pusilla Chiroderma villosum Artibeus intermedius* Artibeus jamaicensis* Artibeus lituratus* Artibeus phaeotis X X X X Artibeus toltecus Artibeus watsoni X X X X Centurio senex Desmodus rotundus X X X X X X X X Diphylla ecaudata Natalus stramineus Thyroptera tricolor Myotis elegans Myotis keaysi X X X X X X X X X X Eptesicus furinalis Lasiurus horealis Lasiurus ega Lasiurus intermedius Rhogeessa tumida X X X X X X X X X X X X X X X Bauerus dubiaquercus Nyctinomops laticaudatus Eumops auripendulus Eumops bonariensis Eumops glaucinus Eumops underwoodi Molossus ater X X X X X X X X X X Molossus molossus Molossus sinaloae X X X X Hershkovitz, 1951, pp. 555-556; Villa-R., 1966, p. 270; Disney, 1968, p. 7; Pine, 1972, p. 42; Quinones et al., 1978, p. 559; Pendergast, 1979, p. 10 Hershkovitz, 1951, p. 555; Disney, 1968, p. 7; Pine, 1972, p. 42; Quinones et al., 1978, p. 559; Pendergast, 1979, p. 10 Disney, 1968, p. 7; Pendergast, 1979, p. 10 Disney, 1968, p. 7; Davis, 1968, p. 697; Pendergast, 1979, p. 10 This paper This paper Peterson, 1966, p. 676; this paper This paper Davis, 1984, p. 14 Dobson, 1878, p. 520; Andersen, 1908, p. 263; Gaumer. 1917, p. 298; Disney, 1968, p. 7; Davis, 1970b, p. 118; Pen- dergast, 1971, p. 102 Murie, 1935, p. 19; Hershkovitz, 1951, pp. 556-557; Disney, 1968, p. 7; Pendergast, 1979, p. 10 Disney 1968, p. 7; Davis, 1970a, p. 398; Pendergast, 1979, p. 10 This paper Davis, 1970a, p. 393; Pendergast, 1979, p. 10 This paper Hershkovitz, 1951, p. 557; Disney, 1968, p. 7; Cartwright & Kirkpatrick, 1977, p. 466; Quinones et al., 1978, p. 559; Pen- dergast, 1979, p. 10 This paper This paj)er Sanborn, 1941, p. 382 This paper Hershkovitz, 1951, pp. 557-558; LaVal, 1973a, p. 25; Quinones et al., 1978, p. 559 Disney, 1968, p. 7; this paper This paper This paper This paper LaVal, 1973b, p. 29; Kirkpatrick et al., 1975, p. 331 This paper Murie, 1935, p. 19 Eger, 1974, p. 5; this paper This paper Eger, 1977, p. 42 Eger, 1977, p. 55 Murie, 1935, p. 19; Pendergast, 1979, p. 10 Murie, 1935, p. 19 Hershkovitz, 1951, p. 559; Disney. 1968, p. 7 * Davis (1984) examined the Artibeus "lituratus" complex in Middle America and restored Artibeus intermedius J. A. Allen to specific status. Specimens cited in the publications listed for A. jamaicensis and A. lituratus should be reevaluated. 162 HELDIANA: ZOOLOGY New Species of Mammals from Northern South America: Fruit-Eating Bats, Genus Artibeus Leach Charles O. Handley, Jr. ABSTRACTS The larger species of Artibeus of the Amazon Basin are defined, and a new giant species is named and described from Venezuela and Colombia. Artibeus fallax, A. Hercules, and A. pla- nirostris are regarded as subspecies of Artibeus jamaicensis, by far the most variable of the larger Artibeus of the region. The smaller Artibeus are keyed and arranged in six species groups. A new dwarf species is described from Brazil, Ecuador, Guyana, Peru, and Venezuela. Distribution and diversity of the smaller species are discussed. Artibeus cinereus, once thought to range throughout Central America and much of South America and to include all of the smaller taxa except A. concolor and A. hartii, is restricted to include only the nominate form and A. quadrivittatus of the lower Amazon Basin and adjacent coastal areas. With these additions and changes in status, at least nine species of Artibeus now are known to occur in northeastern South America. Las especies de gran tamaiio de Artibeus de la Cuenca del Rio Amazonas son definidas y una nueva especie gigante de Venezuela y Colombia es nombrada y descrita. Artibeus fallax, A. hercules, y A. planirostris son consideradas como subespecies de Artibeus jamaicensis, que es el mas variable de los grandes Artibeus de la region. Una clave es preparada para las especies de Artibeus menores, y las especies son arregladas en seis grupos. Una nueva especie enana de Brasil, Ecuador, Guyana, Peru, y Venezuela es descrita. La distribucion y la diversidad de las especies menores son discutidas. Artibeus cinereus, que antes se penso estaba distribuida en Centro America y una gran parte de Sudamerica, y que incluyera todas las taxa mas pequeiias (a excepcion de A. concolor y A. hartii), es ahora restringuida para incluir solamente la especie nominal y A. quadrivittatus a la Cuenca baja del Rio Amazonas y a las areas costeras adyacentes. Con estas adiciones y cambios de "status," por lo menos nueve especies de Artibeus ya son conocidas y se encuentran en el nordeste de Sudamerica. Sao definidas as especies maiores de Artibeus que ocorrem na Bacia Amazonica, e uma especie nova, gigante, e descrita. Artibeus fallax, A. hercules, e A. planirostris sao consideradas subes- pecies de Artibeus jamaicensis, certamente a especie mais variavel dos Artibeus maiores da regiao. Uma chave para os Artibeus menores, os quais foram designados a seis grupos de especies, e fomecida. Uma especie nova aiia e descrita do Brasil, Equador, Guiana, Peru, e Venezuela. A diversidade, e as distribui^oes geograficas destas especies, sao discutidas. Artibeus cinereus, o qual acreditava-se abranger toda America Central e grande parte da America do Sul, alem de incluir todos taxa menores com excessao de A. concolor e A. hartii, e reduzido a um unico taxon, restrito ao sul da Bacia Amazonica e as suas areas adjacentes. Incluindo as adi9oes e mudan^as de status propostas neste trabalho, sao reconhecidas, atual- mente, ao menos nove especies de Artibeus na regiao nordeste da America do Sul. From the National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. HANDLEY: NEW SPECIES OF ARTIBEUS 163 Introduction Mammals and their ectoparasites were collected in Venezuela between 1965 and 1968 by the Smithsonian Venezuelan Project (SVP), supported in part by a contract (DA-49-MD-2788) of the Medical Research and Development Command, Office of the Surgeon General, U.S. Army. Nu- merous papers have described the ectoparasites and mammals of the Project. Throughout these papers undescribed species of mammals have been referred to by alphabetical designations. Some of these have been named subsequently by Handley and Ferris (1972), Handley and Gordon (1980), and Handley (1984). This paper deals with fruit- eating bats of the genus Artibeus Leach. The cranial measurements reported here were taken as outlined by Handley (1959, p. 98). Hind foot, tibia, calcar, and forearm were measured on dry museum si)ecimens or on specimens preserved in alcohol; all other external dimensions were measured on fresh specimens in the field. All mea- surements are in millimeters. Coloration was de- termined under Examolites (Macbeth Corp., New- burg, NY 12533) with natural light excluded. A New Giant Artibeus It is now generally agreed that in and around the Amazon Basin there are three large species of Artibeus. Handley (1976) recognized them as: (1) /1./m//^;>jo51« Gray— smaller, molars 3/3, rostrum arched, postorbital process poorly developed, fur long, coloration blackish, facial stripes faint or ab- sent, interfemoral membrane (IM) naked; (2) A. jamaicensis Leach — larger, molars 3/3, rostrum arched, postorbital process poorly developed, fur short, coloration gray-brown, facial stripes present but not sharply defined, IM naked; and (3) A. li- turatus Olfers— larger, molars 2/3, rostrum flat- tish, postorbital process well developed, fur short, coloration chocolate brown, facial stripes promi- nent and well defined, IM hairy. However, as shown by Koopman (1978) and Honacki et al. (1982), there is no consensus on the delimitation of these species. The difficulty in defining the species arises pri- marily from the fact that Artibeus jamaicensis is unusually variable geographically in morphology. The other species show very little variation in this region. Artibeus jamaicensis is large in the Ama- zon Basin, so large in fact that the subspecies there, A. j. fallax Peters and A. j. Hercules Rehn, until recently have been aligned by most authors with the universally large A. lituratus, although they differ from it in many characteristics other than size. To the southeast of the Amazon Basin {A. j. planirostris Spix) and to the north of it (A. j. trin- itatis Andersen), A. jamaicensis is dramatically smaller, in fact similar in size to A. fuliginosus. Everywhere east of the Andes A. jamaicensis has 3/3 molars; west of the Andes and in Central America it has 2/3 molars. Specimens in the SVP collection show that the large Artibeus jamaicensis fallax and small A. j. trinitatis apparently intergrade in the Llanos of Venezuela where the habitat is marginal for A. jamaicensis and where it is an uncommon bat. Furthermore, intergradation between the small, 1 2-molar A. j. trinitatis and the slightly larger, 10- molar A. j. aequatorialis Andersen of the north- west coast of South America can be seen in spec- imens from northern Colombia. These two zones of intergradation are of crucial importance in the nomenclature oi Artibeus, for they serve to link "/I. jamaicensis^'' of the West Indies and Central America and "A. planirostris" of eastern South America. They are especially im- portant in the present context because of the dis- covery of a fourth large Artibeus, superficially sim- ilar to but larger than A. j. fallax, occurring together with it in southern Venezuela and with the small A. j. trinitatis in western Venezuela and northern Colombia. It can be recognized as follows: Artibeus amplus new SF>ecies HoLOTYPE— USNM 440932, adult female with suckling young, skin and skull, collected 1 5 April 1968 by Norman E. Peterson, F. P. Brown, Jr., and J. O. Matson at Kasmera, 2 1 km SW Machi- ques, Estado Zulia, Venezuela, 270 m, in a damp cave in a cliff across the Rio Yasa from the Kas- mera Biological Station, eastern foothills of the Sierra de Perija. Original number, svp 22086. Etymology— Latin amplus, large, referring to the large size of this bat, one of the largest Artibeus. Distribution— Northern foothills of the Cen- tral Andes in Colombia; lower eastern slopes of the Sierra de Perija and the Venezuelan Andes in western Venezuela; and the vicinity of Cerro Dui- da and the low mountains of southeastern Bolivar in southern Venezuela. It probably occurs in ad- 164 HELDIANA: ZOOLOGY jacent parts of Guyana and Brazil as well. The SVP collectors found A. arnplus near streams and in other moist areas (98%); in evergreen forest (90%) and in forest openings such as yards and orchards (10%). Most specimens were mist netted (86%), but some ( 1 4%) were found roosting in the twilight zone of caves. Elevational range, 24-1200 m. Holdridge life zones (Ewel & Madriz, 1 968): Trop- ical humid forest (10%), Tropical very humid for- est (22%), Premontane humid forest (12%), Pre- montane very humid forest (2%), Premontane rain forest (4%), Lower montane very humid forest (10%), and Lower montane rain forest (40%). Ridge slopes and valley floor in the area where the ho- lotype was collected were clothed with second growth evergreen forest, while lawns, shrubbery, banana and papaya plants, and scattered grapefruit trees characterized the grounds of the biological station. Description— Size large (forearm 70.0, greatest length of skull 31.3, maxillary toothrow 11.2 — averages of Venezuelan specimens). Coloration of fur as in sympatric Artibeus jamaicensis (dorsum blackish brown to brown; facial stripes present but obscure; underparts blackish brown, usually frost- ed with white; underarms with abundant long, usually whitish hairs); ears dark fuscous to black, paler basally; lips and noseleaf blackish; mem- branes blackish; wing tips undifferentiated or gray- ish, never white. Horseshoe of noseleaf bound down mediobasally; legs and interfemoral mem- brane slightly hairy, the latter particularly medio- ventrally, where hairs extend as a short fringe be- yond edge of membrane; forearm long. Skull superficially like that of Artibeus jamai- censis, but relatively longer and narrower; rostrum long and flattish; supraorbital ledges subparallel and together with postorbital processes often poor- ly develojDed or even ill-defined; zygomata not very flared from skull, usually subparallel to one another, and in side view, thin and fragile; posterolateral angle of skull not particularly flared; palate rela- tively narrow and toothrows ovoid in outline; postpalatal extension usually long, narrow, and parallel sided; dentition as in A. jamaicensis, ex- cept I' only weakly bilobed; dental formula 2/2- 1/1-2/2-3/3 X 2 = 32. This bat is the only known host of Strebla paramirabilis Wenzel and Tri- chobius assimilis Wenzel (Diptera: Streblidae), so it can be distinguished from other Artibeus by its parasites as well as its morphology. Measurements of the holotype, an adult female: total length 101, tail vertebrae 0, hind foot (dry) 17, ear from notch 25, forearm 69.2, tibia 24.1, calcar 6.2, weight 70.4 g. Greatest length of skull 31.9, zygomatic breadth 18.3, postorbital breadth 7.7, breadth of braincase 13.3, depth of braincase 1 1.6, length of maxillary toothrow 1 1.2, postpal- atal length 9.8, palatal breadth outside of M^ 12.9, rostral breadth at base of canines 8.3. See Table 1 for additional measurements. Comparisons— Four large species of Artibeus occur in Venezuela, all of them together in the southern part of the country. Among these, Arti- beus amplus and A. jamaicensis are most alike; but despite the superficial resemblance, the two can be distinguished by many characters, both ex- ternal and cranial. All A. amplus examined have the lower edge of the noseleaf horseshoe bound down, while about 95% of A. jamaicensis from the same localities have it free; all A. amplus have the interfemoral membrane slightly hairy and fringed medially, hwX A. jamaicensis never does; and while A. jamaicensis often has the wings white-tipped, A. amplus never does. Cranially, A. amplus differs from A. jamaicensis in having a longer, narrower skull; longer, somewhat more ffattened rostrum (most easily seen in dimensions of rostral shield); less arched nasals; margins of supraorbital nearly parallel, rather than converging posteriorly, and usually not as well developed; zygomata thinner and more fragile and usually subparallel rather than diverging markedly posteriorly; posterolater- al angle of skull not so flaring; palate narrower and toothrows usually less nearly circular in outline; and postpalatal extension usually longer and nar- rower, parallel sided (not flaring posteriorward). The two species are hosts of different species of parasitic streblid ffies. Specimens Examined— Total 55. COLOMBIA. Antioquia: La Tirana, 33 km SW Zaragoza, 520 m (2 usnm). VENEZUELA. Apure: Nulita, Selvas de San Camilo, 29 km SSW Santo Domingo, 24 m (2 usnm). Bolivar: 21 to 33 km NE Icabani, 775-851 m (6 usnm); Km 125, 85 km SSE El Dorado, 826-1165 m (5 usnm). T.F. Amazonas: Belen, Rio Cunucunuma, 56 km NNW Esmeralda, 1 50 m (9 usnm); Cabecera del Caiio Culebra, Cerro Duida, 40 km NNW Esmeralda, 1 1 40-1 200 m (2 1 usnm); Caiio Culebra, Cerro Duida, 50 km NNW Esmeralda, 800 m (2 usnm); Tamatama, Rio Ori- noco, 2 km above Boca del Casiquiare, 135 m (2 usnm). Zulia: Kasmera, 21 km SW Machiques, 270 m (3 USNM, 1 ucv); 15 km W Machiques (1 amnh); Novito, 19 km WSW Machiques, 1 135 m (1 usnm). Remarks— In previous publications of SVP, Ar- tibeus amplus has been known as ""Artibeus sp. D". HANDLEY: NEW SPECIES OF ARTIBEUS 165 Table 1 . Measurements of adult Artibeus ampins and A. jamaicensis. For each measurement, line 1 includes the mean plus or minus two standard errors, line 2 the extremes, and line 3, in parentheses, the number of specimens measured. All specimens are from Venezuela unless otherwise stated. Total length 100.4 ± 3.88 93-104 (5) 89.9 ± 2.22 80-100 (21) 86.4 ± 2.90 77-95 (19) 83.8 ± 4.14 73-91 (10) 88.1 ± 1.62 80-93 (17) 86.3 ± 1.60 82-90 (8) Hind foot (dr>) Ear Forearm Greatest length Zygomatic breadth 15.9 ± 0.40 14-18 (19) 15.4 ± 0.32 15-16 (10) 17.8 ± 0.32 16-19 (17) 18.3 ± 0.32 18-19 (8) 16.9 ± 0.30 16.0-17.7 (12) 17.0 ± 0.28 16.6-17.6 (6) 19.1 ± 0.14 18.7-19.6 (13) 19.3 ± 0.32 18.6-20.2 (8) Postorbital breadth Artibeus amplus, males and females, Zulia and Colombia 18.4 ± 0.60 23.7 ± 1.28 70.8 ±1.78 31.4 ± 0.26 18.6 ± 0.20 17-19 22-26 68.6-75.3 31.0-31.9 18.1-18.8 (7) (7) (7) (7) (7) A. amplus, males and females, T.F. Amazonas and Bolivar 18.3 ± 0.30 23.0 ± 0.98 69.1 ± 0.90 31.2 ± 0.24 18.4 ± 0.14 17-20 18-26 65.0-73.2 30.3-32.8 17.4-19.1 (22) (21) (22) (29) (30) A. jamaicensis, females, Zulia 22.6 ± 0.96 61.1 ± 0.64 27.7 ± 0.24 17-25 58.9-64.2 26.7-28.5 (19) (19) (19) A. jamaicensis, males, Zulia 22.2 ± 0.98 59.3 ± 0.92 27.4 ± 0.14 20-25 56.2-61.4 27.1-27.7 (10) (10) (10) A. jamaicensis, females, T.F. Amazonas 24.6 ± 0.46 66.8 ±1.12 30.7 ± 0.28 23-26 62.1-70.1 29.4-31.3 (17) (17) (14) A. jamaicensis. males, T.F. Amazonas 24.6 ± 0.52 65.4 ± 1.42 30.7 ± 0.26 24-26 62.4-68.6 30.2-31.4 (8) (8) (8) 7.9 ± 0.20 7.6-8.3 (7) 7.8 ± O.IO 7.3-8.4 (31) 6.8 ± 0.08 6.5-7.1 (19) 6.8 ± 0.14 6.3-7.0 (10) 7.5 ± 0.10 7.1-7.7 (14) 7.6 ± 0.16 7.2-7.9 (8) A New Dwarf Artibeus The taxonomy of the smaller Artibeus is in a state of flux. As recently as 35 years ago all of the smaller species except A. concolor Peters and A. hart a Thomas were believed to be variants of .-i. cinerens Gervais. Since then, first one and then another of the supposed subspecies of .4. cinereus has been shown to be independent species. Today only A. bogotensis Andersen, A. glaucus Thomas, A. pumilio Thomas. A. quadrivittatus Peters. A. rosenbergi Thomas, and A. watsoni Thomas re- main associated with A. cinereus (Honacki et al., 1982). However, except for^. quadrivittatus, these do not properly belong with A. cinereus either. Artibeus glaucus and A. bogotensis intergrade in Ecuador and form an Andean-northern South American species sympatric with A. cinereus in southern Venezuela. Artibeus glaucus thus has two subspecies, the nominate form and A. g. bogoten- sis. Artibeus watsoni Thomas of northwestern South Ameinca and Central America is closely related, but intergradation with A. g. glaucus or A. g. bo- gotensis has not been obsei^ed. Artibeus pumilio is an enigmatic taxon. Many museum specimens bear the name A. pumilio, but perhaps the only specimen properly associated with the name is the holotype. This specimen may be only an odd variant of one of the other species, but not of the species described here. For the pres- ent, A. pumilio must be regarded as unplaceable. The same can be said for A. rosenbergi, charac- teiized by a curiously long, narrow skull such as can be found occasionally in large samples of most species o{ Artibeus. Because of their equivocal sta- tus, neither A. pumilio nor A. rosenbergi is includ- ed in the appended list of species and key. The characteii sties and status of these taxa will be the subject of another paper. Thus, .A. cinereus now has been shorn of all of its supposed subspecies except A. c. quadrivittatus. Its supposed range has been reduced from encom- passing most of Central Ameiica and tropical South America to occupying only the Amazon Basin (possibly only the lower basin) and adjacent coast- al areas. Sympatric with A. cinereus in much of its range is a distinctive dwarf species which can be known as: 166 HELDIANA: ZOOLOGY Table 1. Continued. Braincase breadth Braincase depth Maxillary Postpalatal toothrow length Width at molars Width at canines Tibia 13.5 ± 0.20 13.2-14.0 (7) Artibeus amplus. males and females, 1 1.0 ± 0.26 1 1.2 ± 0.16 9.7 ± 0.20 10.6-11.6 11.1-11.5 9.3-10.0 (7) (7) (7) Zulia and Colombia 13.2 ± 0.26 8.4 ±0.16 12.7-13.5 8.2-8.8 (7) (7) 25.9 ± 1.28 24.1-28.2 (6) 13.3 ± 0.12 12.9-14.0 (30) A. amplus, males and females, T.F. Amazonas and Bolivar 11.2 ±0.06 11.2 ±0.10 9.8 ±0.14 13.3 ± 0.10 8.6 ± 0.08 10.7-11.5 10.7-11.8 9.1-10.6 12.8-13.9 8.3-8.9 (30) (31) (28) (30) (30) 24.8 ± 0.38 23.1-26.2 (22) 12.3 ± 0.10 12.0-12.8 (19) 10.2 ± 0.16 9.5-10.9 (19) A. jamaicensis. females 10.0 ±0.12 8.6 ±0.14 9.4-10.5 8.2-9.3 (17) (18) , Zulia 12.1 ± 0.16 11.2-12.7 (19) 7.6 ± 0.10 7.2-8.0 (19) 22.4 ± 0.50 20.1-23.8 (19) 12.1 ± 0.14 11.8-12.4 (10) 10.3 ± 0.18 10.0-10.9 (10) A. jamaicensis. males, 10.1 ± 0.16 8.4 ± 0.12 9.7-10.4 8.2-8.8 (8) (10) Zulia 12.2 ± 0.18 11.8-12.6 (10) 7.7 ± 0.12 7.4-7.9 (9) 21.7 ± 0.52 20.1-23.0 (10) 13.2 ± 0.12 12.9-13.5 (14) 10.8 ± 0.18 10.2-11.5 (14) A. jamaicensis. females, T.F 11.4 ±0.20 9.4 ±0.18 11.0-12.0 8.9-10.0 (13) (14) . Amazonas 13.7 ± 0.22 13.2-14.4 (13) 8.6 ± 0.12 8.3-8.9 (13) 24.1 ± 0.46 22.3-25.4 (17) 13.4 ± 0.22 12.8-13.7 (8) 11.0 ± 0.18 10.7-11.4 (8) A. jamaicensis, males, T.F. 11.4 ±0.18 9.3 ± 0.24 11.1-11.8 8.7-9.8 (8) (8) Amazonas 13.8 ± 0.30 13.0-14.3 (8) 8.8 ± 0.12 8.6-9.1 (8) 23.3 ± 0.52 22.0-24.4 (8) Artibeus gnomus new species HoLOTYPE— USNM 387534, adult female, skin and skull, collected 14 June 1966 by A. L. and M. D. Tuttle at El Manaco (= Km 74), 59 km SE El Dorado, Bolivar, Venezuela, 150 m, in a mist net in an orchard. Original number, svp 9298. Etymology— Latin gnomus, diminutive fabled being, dwarf, alluding to the small size of this species, one of the smallest Artibeus. Distribution— The Amazon Basin and bor- dering regions; from northern Amazonas Territory (14 km SSE Pto. Ayacucho) and northern Bolivar State (28 km SE El Manteco) in Venezuela and northern Guyana, to Para (Belem) and Mato Gros- so (Serra do Roncador), Brazil, and Loreto (Santa Rosa), Peru. SVP collectors netted A. gnomus mostly in moist sites (92%) in evergreen forest (52%) or openings such as savannas (25%) and yards and orchards (23%). Elevations range 1 19- 161 m in Venezuela, sea level to 530 m in Brazil. Holdridge life zones: Tropical dry forest (22%), Tropical humid forest (67%), Tropical very humid forest (2%), and Premontane humid forest (9%). Description— Body size small (forearm aver- ages 36-38, greatest length of skull 18.5-18.7, and maxillary toothrow 5.7-6.0). Dorsal coloration gray-brown to brown; underparts paler; facial stripes very white and sharply defined. Soft parts coloration in life (usnm 36 1 742, male, Belem, Bra- zil): ear narrowly edged with yellow, brightest to- ward base; antitragus entirely yellow; tragus yel- low, brightest distally and on posterior basal lobe; noseleaf and horseshoe gray-brown medially, cream color laterally; lips and chin gray-brown; iris brown; forearm and fingers brownish flesh col- or; wings blackish, except membrane between fin- gers II and III transparent, grayish; interfemoral membrane sooty brown; legs and feet dark brown; claws horn color. Face short; shape and propor- tions of ears, noseleaf, horseshoe, lips, chin, and interfemoral membrane as in Artibeus cinereus; noseleaf minutely hirsute; lower edge of horse- shoe free; basal part of forearm hairy; hind ex- tremities (except for short hairs on feet) appear naked. Skull small, short, and broad; zygomata sub- parallel; rostrum narrow, very short, moderately HANDLEY: NEW SPECIES OF ARTIBEUS 167 Table 2. Measurements of adult male and female (combined) Artibeus gnomus and A. glaucus bogotensis. For each measurement, line 1 includes the mean plus or minus two standard errors, line 2 the extremes, and line 3, in parentheses, the number of specimens measured. All specimens are from Venezuela. Total length 47.5 ± 1.40 44-54 (13) 52.2 ± 0.86 49-56 (19) Hind foot (dry) Ear Forearm Greatest length Zygomatic breadth A. glaucus bogotensis. Km 125, 85 km SSE El Dorado 10.6 ± 0.22 17.4 ±0.38 39.6 ± 0.60 20.3 ± 0.10 11.6 ±0.10 10-11 16-19 36.8-41.9 19.4-21.2 10.8-12.1 (19) (19) (19) (50) (44) Postorbital breadth Artibeus gnomus, Rio Supamo, Los Patos, and EI Manaco 9.5 ± 0.28 16.9 ± 0.62 36.7 ± 0.54 18.5 ± 0.18 11.0 ±0.18 9-10 14-19 34.0-38.3 17.9-19.1 10.4-11.2 (13) (13) (13) (14) (8) 4.9 ± 0.10 4.5-5.2 (14) 5.0 ± 0.04 4.6-5.3 (51) deep and arched, and much swollen posterolater- ally (part on rostral shield, part within orbit, above eye); excavation for orbital nerve large and deep; braincase short and deep, with swelling at pos- terodorsal apex interrupting junction of sagittal and lambdoidal crests; postpalatal extension rel- atively short; internal edge of pterygoid fossa strongly ridged, narrowing mesopterygoid fossa and cupping pterygoid fossa which opens straight back; vomerine ridge visible in mesopterygoid fossa; va- cuities in roof of posterior nares much anterior to mesopterygoid fossa and not easily seen; outline of maxillarv toothrows nearly circular; upper ca- nine small (especially in basal diameter); M' with accessory internal ridge on lateral cusps, and with relatively wide talon; m, present (75 of 79 speci- mens examined). Measurements of the holotype, an adult female: total length 47, tail vertebrae 0, hind foot (dry) 9, ear from notch 18, forearm 36.5, tibia 12.6, calcar 4.9, weight 10.5 g. Greatest length of skull 18.2, zygomatic breadth 10.8, postorbital breadth 4.8, breadth of braincase 8.5, depth of braincase 7.2, length of maxillary toolhrow 5.5, postpalatal length 6.5, palatal breadth outside of M' 7.1, rostral breadth at base of canines 4.6. See Table 2 for additional measurements. Comparisons— ^r//Z)e«5 gnomus differs from A. concolor and A. hartii in many ways, but most significantly in lack of M\ From all other small Anibeus {A. anderseni, A. cinereus, and A. glaucus bogotensis) that occur within its range, A. gnomus can be distinguished by its possession of mj. Among the specimens examined, m, is consistently absent in these other taxa while it is consistently present in A. gnomus (except in southern Venezuela, where it is absent from both mandibles in four of 53 specimens and from one mandible only in two others). In addition, A. gnomus differs from all of the sympatric taxa in its more prominent white facial stripes; more colorful ears, noseleaf, and lips; average browner, less grayish coloration of pelage; shorter face and rostrum (except when compared with A. concolor); more swollen supraorbital re- gion; average larger and deeper orbital nerve ex- cavation (sometimes equally large and deep in A. g. bogotensis); and more cupped pterygoid fossa, with internal ridge so enlarged as to significantly narrow the mesopterygoid fossa. Artibeus gnomus differs from the sympatric taxa individually in several other ways. It is much smaller than A. concolor (forearm averages 36-38 vs. 46-48). In contrast to A. hartii it has notched inner upper incisors, brownish rather than dark chocolate coloration, and a wide, unfringed inter- femoral membrane. Compared with A. aruierseni (including the holotype, fmnh 21331), ^. gnomus is similar in size (slightly larger than Rio Madeira A. anderseni); has rostrum much deeper, more arched, narrower, and shorter; face not dished; orbit larger; zygomata more nearly parallel; and vacuities in roof of p)OSterior nares far forward of mesopterygoid fossa, rather than opening in it or close to it. At Belem, Brazil, both Artibeus gnomus and A. cinereus were numerous and were often taken in the same nets. There, fresh specimens of the two species were compared. Artibeus gnomus is smaller in size, and has a smaller head and shorter face; facial stripes much brighter, more sharply defined, and more prominent; ears, noseleaf, and lips more brownish, less grayish; ear edgings, antitragus, and tragus bright yellow, rather than cream; and nose- leaf edged with cream, rather than plain gray- brown. Furthermore, it has zygomata more nearly parallel; rostrum deeper and shorter; supraorbital area much swollen and its edges nearly parallel; and smaller teeth. 168 HELDIANA: ZOOLCX}Y Table 2. Continued. Braincase breadth Braincase depth Maxillary toothrow Postpalatal length Width at molars Width at canines Tibia 8.5 ± 0.14 8.1-9.0 (14) 9.0 ± 0.06 8.5-9.5 (48) Artibeus gnomus, Rio Supamo, Los Patos, and El Manaco 7.4 ± 0.14 5.7 ± 0.06 6.3 ± 0.10 7.5 ± 0.12 4.9 ± 0.08 13.2 ± 0.46 7.1-8.0 5.5-5.9 6.0-6.7 7.1-7.9 (14) (14) (14) (14) A. glaucus bogotensis, Km 125, 85 km SSE El Dorado 7.9 ± 0.06 6.5 ± 0.04 7.0 ± 0.08 8.0 ± 0.06 7.2-8.3 6.0-6.8 6.5-7.5 7.5-8.7 (48) (51) (47) (50) 4.6-5.1 (14) 5.1 ± 0.06 4.9-5.6 (50) 11.2-14.4 (13) 13.6 ± 0.36 12.3-15.8 (19) In southern Venezuela Artibeus gnomus is sym- patric with A. glaucus bogotensis. Compared with Venezuelan specimens and with the holotype (bm 99.11.4.35) of this taxon, A. gnomus is much smaller and shorter faced; has a deeper, shorter rostrum; disproportionately wider zygomatic spread; and smaller teeth. In addition to comparisons of /I. gnomus with sympatric species, two other small Artibeus need to be considered: 1. Artibeus g. glaucus —This species occurs nearby in the Andes. It (including the holotype, BM 94.8.6.13) possesses m„ and its skull has the basic shape of /i. gnomus. However, it is much larger and darker in color, has the hind extremities much hairier, the supraorbital region usually less swollen, and the pterygoid fossa much less cupped and opening to the mesopterygoid fossa. 2. Artibeus fvatsoni— West of the Andes and ex- tending into Central America is another small species, A. watsoni Thomas, which like A. gnomus possesses m,. It (including its holotype, bm 0.7. 1 1 . 1 9) is larger than A. gnomus; has larger teeth; longer rostrum, with reduced supraorbital swell- ing; shallower and less well-defined orbital nerve excavation; and like A. glaucus has the pterygoid fossa not cupped and opening into the mesopter- ygoid fossa (which consequently is not narrowed by the inner pterygoid ridge). Discussion— The ten small species of Artibeus recognized here can be associated in six species groups: 1. Artibeus concolor Group— Amazon and up- per Orinoco basins and Guianas. Includes only Artibeus concolor. 2. Artibeus hartii Group— Mexico and Central America, across northern South America to Trin- idad, and south to Peru east of the Andes and to Ecuador west of the Andes. Includes only Artibeus hartii. 3. Artibeus glaucus Group— Mexico, Central America, and South America to Mato Grosso and Peru. Includes Artibeus glaucus (with two subspe- cies, A. g. bogotensis and A. g. glaucus), A. gnomus, and A. watsoni. 4. Artibeus toltecus Group— Mexico and Cen- tral America. Includes Artibeus aztecus Andersen and Artibeus toltecus Saussure, each with several subspecies. 5. Artibeus cinereus Group— Guiana region, coastal Brazil, and lower Amazon Basin (dubious- ly also upper Amazon Basin). Includes only Ar- tibeus cinereus, with A. c. quadrivittatus as a sub- species. 6. Artibeus phaeotis Group — Mexico, Central America, and South America to upper Amazon Basin and western Ecuador. Includes Artibeus an- derseni Osgood and Artibeus phaeotis Miller, ' with several subspecies. Diversity in the small Artibeus is greatest in east- ern South America, where representatives of five of the six groups occur and where three of the groups are endemic. Altogether six species occur in and around the Amazon Basin, while only one is known with certainty in the central portion of the Basin; there are three in the lower Amazon ' Until recently (Koopman, p. 152, in Honacki et al., 1982) it has not been generally recognized that Artibeus phaeotis and A. ravus are conspecific. They inlergrade in eastern Panama and western Colombia. Both names date from Miller (1902). Although A. ravus was named first, on an earlier page, A. phaeotis became embedded in the literature as the name of this species. HANDLEY: NEW SPECIES OF ARTIBEUS 169 and on the southern fringes in Brazil and Bohvia, four or five in southern Venezuela, and five in east- em Peru, Ecuador, and Colombia. In contrast, only three of the species groups occur in Central Amer- ica, and only one of them is endemic there. Several distributional patterns are represented in the complex of Amazonian species. Artibeus concolor is found throughout the Basin but scarce- ly beyond it; A. cinereus occurs in the lower Am- azon and along the coast for some distance north and south of the river; A. anderseni is known from the upper Amazon and an isolated area in northern Colombia; A. glaucus is higher up, in the Andes, and eastward around the northern edge of the Ba- sin in Venezuela; the range of A. hartii resembles that of A. glaucus, but extends on into Central America; and the dwarf /I. gnomus has a p)eculiar circular range, completely ringing the Amazon Ba- sin but apparently not extending into its interior. Key to the Smaller Species of Artibeus 1. Molars 3/3 (mj large) 2 r. Molars 2/3 (mj minute) or 2/2 3 2. V notched; facial stripes absent; coloration pale brown; interfemoral membrane broad and naked; forearm 43-52 mm Artibeus concolor 2'. I' not notched; facial strip>es present; color- ation dark chocolate brown; interfemoral membrane narrow and fringed; forearm 36- 42 mm Artibeus hartii 3. Supraorbital region much swollen; molars 2/3 (2/2 m A. g. bogotensis and occasionally in the others) . . . Artibeus glaucus Group, 4 3'. Supraorbital region little, or not at all, swol- len; molars 2/2 7 4. Rostrum short and moderately arched; pter- ygoid fossa cupijed and opening back, causing mesopterygoid fossa to be narrowed; forearm 34-38 mm Artibeus gnomus 4'. Rostrum long and much or only moderately arched; pterygoid fossa not cupped, opening into and not narrowing mesopterygoid fossa 5 5. Rostrum much arched; orbital nerve exca- vation shallow and often ill-defined; dorsum pale brownish; ears pale; forearm 35-41 mm Artibeus watsoni 5'. Rostrum moderately arched; orbital nerve excavation deep and well defined 6 6. Molars usually 2/3; dorsum dark grayish or blackish; ears dark; forearm 38-42 mm . . . Artibeus glaucus glaucus 6'. Molars 2/2; dorsum pale brownish or grayish; ears pale; forearm 37-41 mm Artibeus glaucus bogotensis 7. Interfemoral membrane narrow and fringed; coloration blackish Artibeus toltecus Group, 8 7'. Interfemoral membrane broad, naked; col- oration brownish 9 8. Larger, forearm 42-48 mm Artibeus aztecus 8'. Smaller, forearm 37-41 mm Artibeus toltecus 9. Rostrum deep and arched; palate long and moderately wide Artibeus cinereus 9'. Rostrum shallow and flattened; palate short and very wide Artibeus phaeotis Group, 10 10. Maxillary toothrow 5.2-6.2 mm; rostrum often tilted up anteriorly Artibeus anderseni 10'. Maxillary toothrow 6.7-7.1 mm; rostrum usually not tilted up anteriorly^ Artibeus phaeotis Specimens Examined— /irf/Aeiii anderseni — BRAZIL. Amazonas: Borba, Rio Madeira (1 amnh). Rondonia: Porto Velho (2 amnh, 2 fmnh, including holotype of ^. anderseni); Sto. Antonio do Hauayara (4 amnh). COLOMBIA. Bolivar: Ca- tival. Upper Rio San Jorge, 120 m (16 fmnh). Antioquia: Aljibos, 26 km S and 22 km W Zara- goza, 630 m (2 usnm); nr. La Tirana, 24 km S and 22 km W Zaragoza, 520 m (2 usnm). ECUAIX)R. Napo: Rio Suno (Abajo) (4 amnh). Pastaza: Mon- talvo, Rio Bobonaza ( 1 fmnh); Rio Pindo Yacu (1 fmnh); Rio Yana Rumi (1 fmnh). PERU. Hua- nuco: Monte Alegre (1 amnh). Loreto: Boca Rio Curaray (1 amnh); Boca Rio Peruate, Rio Ama- zonas, 90 m (1 FMNH); Lagarto, Alto Ucayali (1 amnh); Mazan (1 amnh); 59 km W Pucallpa (1 usnm); Puerto Indiana, Rio Amazonas (2 amnh); Rio Morona (Quebr. Pushaga), Alto Amazonas, 220 m (2 FMNH); Rio Yavari Mirim (Quebr. Es- peranza), 200 m (2 fmnh); Santa Cecilia, Rio Man- iti, Iquitos, 110 m (3 fmnh); Santa Luisa, Rio Nanay, Iquitos, 160 m (1 fmnh); Sarayacu, Rio Ucayali (1 amnh). Pasco: San Juan, Oxapampa, ^ Couplet 10 will separate Artibeus anderseni and A. phaeotis in South America and in southern Central America, but it will not distinguish A. anderseni from Mexican A. phaeotis nanus. In such a comparison, A. anderseni can be recognized by its relatively broader skull. 170 FIELDIANA: ZOOLOGY j 274 m (3 usnm). Departamento (?): Yuhucumayo, 1200 ft [= Puno: Yahuaramayo, 366 m?] (1 mcz). Artibeus cinereus c/iiere«5— BRAZIL. Amazo- nas: Sta. Clara, Vila Bela Imperatriz [nr. Parintins] (1 amnh). Para: Fordlandia, Rio Tapajos (2 amnh); Maracano, Rio Jamunda [= Nhamunda?], Faro (5 amnh); Rio Yumunda, Faro ( 1 bm). Artibeus cinereus quadrivittatus — BRAZIL,. Maranhao: Juryassu [= Turia9u?] (1 bm). Para: Belem (10 usnm); Benevides (1 bm); Para [= Be- lem] (1 bm); Ilha do Taiuna, Rio Tocantins (3 amnh). Pemambuco: Pemambuco [= Recife] (2 bm). Rio Grande do Norte: Natal (1 usnm). SUR- IN AME. Surinam ( 1 bm). VENEZUELA. Bolivar: Hato San Felipe, Serrania de Nuria (1 ucv); Hato San Jose, 20 km W La Paragua, 300-324 m (2 usnm). Artibeus glaucus bogotensis — COLOMBIA. Cundinamarca: Bogota (2 bm): nr. Bogota ( 1 bm); Curiche, nr. Bogota (2 bm, including holotype of A. bogotensis); Fomeque (1 amnh); Fusagasuga (2 mcz); Rio Negro, nr. Bogota (2 bm). GUYANA. Kanuku Mts. (3 bm). VENEZUELA. Bolivar: El Manaco, 59 km SE El Dorado, 150 m (3 usnm); Hato San Jose, 20 km W La Paragua, 300-324 m (3 USNM); 23 to 45 km NE Icabaru, 824-851 m (3 USNM); Km 125, 85 km SSE El Dorado, 826-1 165 m ( 1 20 usnm); Rio Supamo, 50 km SE El Manteco, 1 50 m (2 usnm). T.F. Amazonas: Belen, Rio Cu- nucunuma, 56 km NNW Esmeralda, 150 m (1 usnm); Caiio Culebra, Cerro Duida, 50 km NNW Esmeralda, 800 m (3 usnm). Artibeus glaucus glaucus— BOIAW A. Santa Cruz: Buenavista, 400 m (1 fmnh). ECUADOR. Napo: Baeza (1 bm). PERU. Cuzco: Collpa de San Lorenzo, Quincemil, 700 m (1 1 fmnh); Hda. Ca- dena, Quincemil, 1000 m (9 fmnh). Junin: Chan- chamayo, 1000 m (2 bm, including holotype oi A. glaucus); Huacapistana ( 1 f>4Nh). Puno: Rio Inam- bari, 670 m (3 amnh); Santo Domingo (1 amnh); Yahuaramayo, 366 m (1 bm, 1 usnm). Artibeus gnomus-ToXdA 104. BRAZIL. Mate Grosso: Serra do Roncador, 264 km N (by road) Xavantina, 533 m (17 usnm). Para: Belem, Sta. A, IPEAN (7 usnm); Belem, Utinga (5 usnm); Be- lem, Benevides (2 usnm). ECUADOR. Pastaza: Canelos, upper Rio Bobonaza (1 amnh). GUY- ANA. E. Berbice District: Wikki River (3 usnm). Mazaruni-Potaro District: Issano Road, 1 2 mi W of Bartica-Potaro Road (1 usnm). PERU. Loreto: 59 km SW Pucallpa (1 usnm); Santa Rosa, Alto Ucayali (10°42'S, 73''50'W) (2 amnh). VENE- ZUELA. Bolivar: El Manaco, 59 km SE El Do- rado, 150 m (12 usnm); Km 38, SE El Dorado, 100 m (1 ucv); Los Patos, 28 km SE El Manteco, 1 50 m (4 usnm); Rio Supamo, 50 km SE El Man- teco, 150 m (1 usnm); Salto Chalimaha, Rio Pa- ramichi, Rio Paragua (1 ucv); Salto Ichun, Rio Paragua (2 ucv). T.F. Amazonas: Belen, Rio Cu- nucunuma, 56 km NNW Esmeralda, 150 m (2 usnm); Boca Mavaca, 84 km SSE Esmeralda, 1 38 m (1 usnm); Caiio Leon, Cerro Duida, 325 m (1 amnh); Capibara, Brazo Casiquiare, 106 km SW Esmeralda ( 1 usnm); Esmeralda, Cerro Duida, 325 m (3 amnh); 14 to 65 km S, SSE, and SSW Pto. Ayacucho, 119-161 m (16 usnm); Rio Mavaca, 1 08 km SSE Esmeralda, 1 40 m (7 usnm); San Juan, Rio Manapiare, 163 km ESE Pto. Ayacucho, 155 m (6 usnm); Tamatama, Rio Orinoco, 135 m (7 usnm). Artibeus phaeot is— HoXoXyTpcs o{A. phaeotis and A. ravus, plus hundreds of other specimens from Mexico, Central America, and NW South Amer- ica. Artibeus pumilio— PERU. Loreto: Masisea, Tushemo, Rio Ucayali, 328 m (1 bm, holotype of A. pumilio). Artibeus watsoni— PANAMA. Chiriqui: Boga- va, 250 m (5 bm, including holotype of A. watsoni); Progreso (34 usnm); Puerto Armuelles (2 usnm). Remarks— In previous publications of SVP, Artibeus gnomus has been known as '''' Artibeus sp. A". Acknowledgments Among the persons who helped me put together this paper I am especially grateful to Sally DeMott, who measured the SVP skulls; Linda Gordon, who compiled the tables and worked with me on the comparisons; and Jane Ailes Small, who read and criticized the manuscript and did the word pro- cessing. Curators of several collections kindly per- mitted me to study specimens under their care in the preparation of these descriptions: American Museum of Natural History (AMNH), British Museum (Natural History) (BM), Field Museum of Natural History (FMNH), Museum of Com- parative Zoology, Harvard University (MCZ), and Universidad Central de Venezuela (UCV). The SVP collection is in United States National Mu- seum of Natural History (USNM); a portion of its specimens have been returned to Venezuela. HANDLEY: NEW SPECIES OF ARTIBEUS 171 Literature Cited EwEL, J. J., AND A. Madriz. 1968. Zonas de Vida de Venezuela. Ministerio de Agricultura y Cria, Caracas, 265 pp., map. Handley, C. O., Jr. 1959. A revision of American bats of the genera Euderma and Plecotus. Proceedings of the United States National Museum, 110: 95-246. . 1976. Mammals of the Smithsonian Venezue- lan Project. Brigham Young University Science Bul- letin, Biological Series, 20(5): 1-91. . 1984. Newspeciesof mammals from northern South America: A long-tongued bat, genus Anoura Gray. Proceedings of the Biological Society of Wash- ington. 97: 513-521. Handley, C. O., Jr., and K. C. Ferris. 1972. De- scriptions of new bats of the genus Vampyrops. Pro- ceedings of the Biological Society of Washington, 84: 519-523. Handley, C. O., Jr., and L. K. Gordon. 1980. New species of mammals from northern South America. Mouse possums, genus Marmosa Gray, pp. 65-72. In Eisenberg, J. F., ed.. Vertebrate Ecology in the North- em Neotropics. Smithsonian Institution Press, Wash- ington, D.C. HoNACKJ, J. H., K. E. Kjnman, and J. W. Koeppl, eds. 1982. Mammal Species of the World. Allen Press, Inc., and the Association of Systematic Collections, Lawrence, Kansas, 694 pp. Koopman, K. F. 1978. Zoogeography of Peruvian bats with special emphasis on the role of the Andes. Amer- ican Museum Novitates, 2651: 1-33. Miller, G. S., Jr. 1902. Twenty new American bats. Proceedings of the Academy of Natural Sciences of Philadelphia, 54: 389-412. 172 HELDIANA: ZOOLOGY Seasonality of Reproduction in Peruvian Bats Gary L. Graham ABSTRACTS The reproductive conditions of 3,489 specimens were used to determine seasonal patterns of pregnancy and parturition in Peruvian bats. More species that are trophic generalists yield birth records for the dry season than do trophic specialists. Relatively more highland than lowland species have births recorded for both seasons (dry and wet). Presumably, trophic generalists and highland species experience less seasonal variation in food supplies compared to the other groups. A larger percentage of nectarivorous species than frugivores have birth records for the dry season; the reverse is true for the wet season. These patterns are associated with greater floral resource abundance during the dry season and a greater abundance of fruit resources during the wet season. Las condiciones reproductivas de 3489 especimenes fueron usadas para describir patrones estacionales de embarazo y alumbramiento en murcielagos peruanos. Mas especies de gener- alistas alimenticios tienen registros de nacimiento durante los dos estaciones (seca y mojada) que tienen especialistas. Relativamente mas especies desde tierras altas que desde tierras bajas tienen registros de nacimiento durante la estacion seca. Possiblamente, generalistas alimenticios y especies de tierras altas sufrir menos variacion estacional de provisiones alimentos que los otros grupos. Un mayor porcentaje de especies nectivoras que frugivoras tienen registros de nacimiento durante la estacion seca pero, el opuesto exista para la estacion mojada. Estos patrones son asociados con un mayor abundancia de recursos de flores durante la estacion seca y con un mayor abundancia de recursos de frutas durante la estacion mojada. As condi96es reprodutivas de 3489 especimes de morcegos p)eruanos foram usadas para descrever seus padroes de parturi^ao. Os morcegos de habitos alimentares generalizados pos- suem uma propor9ao maior de especies que possuem registros de partos durante a epoca da seca, do que os morcegos com habitos alimentares especializados. Proporcionalmente mais especis de morcegos de areas elevadas, do que das planicies, possuem registros de partos durante as duas epocas (seca e chuvosa). Morcegos generalizados e os de areas elevadas provavelmente experimentam menos varia^ao epocal em a quantidade da comida, do que os morcegos das planicies ou morcegos com habitos alimentares especializados. Uma porcentagem maior das especies nectarivoras, do que das especies frugivoras, possuem registros de partos durante a epoca da seca, sao a regra ao inves da exce9ao entre os morcegos que consumen fruta. Estes padroes podem ser interpretados como adpata96es de individuos a recursos alimentares que sao regularmente, ou ocasionalmente, obteniveis durante a epoca da seca. From the Department of Biology, University of New Mexico, Albuquerque, New Mexico 87131. GRAHAM: PERUVIAN BATS 173 Introduction Although the timing of reproductive events is an important Hfe history adaptation, there are few studies of the reproductive phenology of South American bats (Racey, 1982). Most of these stud- ies are of individual species and cover only part of the year. Gestation, parturition, lactation, and weaning should be timed to correspond with vari- ations in the abundance and diversity of food sup- plies (Fleming et al., 1972; Bradbury & Vehren- camp, 1977; Bonaccorso, 1979; Wilson, 1979; August & Baker, 1 982). These variations are known to be seasonal in most of the Neotropics (Janzen, 1967, 1973; Foster, 1982; Wolda, 1982; Smythe, 1982; Terborgh, 1983). Which reproductive event is actually synchronized with peak food supplies is likely to be determined by the relative cost of each event. Lactation is the most costly period for most females, but weaning and dispersal pose the greatest survival problem for young bats (Wilson, 1979; Racey, 1 982). These authors, and Tuttle and Stevenson (1982), agree that the weaning of young is the most critical period for individuals of most bat populations. Most species of Neotropical bats that have been studied are polyestrous (Fleming et al., 1972 Thomas, 1972; Bradbury & Vehrencamp, 1977 Myers, 1977; Wilson, 1979; Bonaccorso, 1979 Humphrey & Bonaccorso, 1979). Each year, fe- male bats confront the problem of timing two (rarely three in a few vespertilionids) periods of lactation and weaning with variations in food availability. Individuals of most species handle this problem by producing their young so that the first is weaned at the beginning of the wet season as fruit supplies are reaching their peak, and the second later in the wet season when fruit avail- ability reaches a second peak or remains relatively high (Wilson, 1979; Tuttle «& Stevenson, 1982). Exceptions to this general pattern of seasonal poly- estry have been observed in most studies (Fleming et al., 1972; Thomas, 1972; Myers, 1977; Myers & Wetzel, 1983; Bonaccorso, 1979; see Mares & Wilson, 1 97 1 ; Bradbury & Vehrencamp, 1 977; and August & Baker, 1982 for good discussions of dry season birth periods). The purpose of this study is to examine some of the details of dry season reproduction in Pe- ruvian bats. I demonstrate that the different feed- ing assemblages and the faunas of different zoo- geographical regions differ in the proportion of species with birth records during the dry season. I also offer suggestions as to how dry season births may be adaptive. Methods The bat specimens from Peru housed in the col- lections of the American Museum of Natural His- tory, Field Museum of Natural History, Louisiana State University Museum of Zoology, National Museum of Natural History, and the Texas Co- operative Wildlife Collection were examined in late 1977 and early 1978 for information on re- production. Data were taken directly from Tuttle (1970) and Bowles et al. (1979) for specimens not in the above collections but included in their re- ports. Information on reproductive condition of bats was also obtained from the notes of A. L. Gardner and J. L. Patton. Fieldwork at several localities in Peru, conducted from June through August 1977 and from June through November 1978, enabled me to collect and record the repro- ductive data for many of the bats now in the Lou- isiana State University collection. Diet informa- tion was taken from the literature (Heithaus et al., 1975; Paradiso, 1975; Gardner, 1977) and is given in Graham ( 1 983). Nomenclature follows Graham (1983) except for Artibeus glaucus, which I now consider a distinct species. Recorded data consisted of species identifica- tion, locality, elevation, sex, age, and macroscopic reproductive condition. Individuals with incom- pletely ossified phalangeal epiphyses were classi- fied as juveniles. Females were designated as preg- nant or lactating if this information was included on specimen tags or indicated by dissection of fluid preserved females. The crown-rump lengths of embryos (including extra-embryonic membranes) were either measured (in mm) or taken from spec- imen tags. Length of testes was also noted when it was recorded on sp>ecimen tags, but was not used to establish reproductive patterns because of un- certainties in the relationship between testes size and sexual activity (Taddei, 1976; Thomas, 1972). Monthly pregnancy frequencies are based only on those female SF>ecimens for which I am confident that the presence (or absence) of embryos had been properly recorded. This limitation excludes most of the specimens collected prior to 1960. I classified each species with adequate data (those with evidence of at least one birth period) into those that can give birth in the dry (May-Septem- ber) and/or wet (October-April) seasons. I as- 174 FIELDIANA: ZOOLOGY sumed that a birth between 1 May and 3 1 August also indicates weaning of the young during the dry season. Species with births recorded only for the month of September (usually a dry month) were not included in either breeding season, since these young could be weaned either as the dry season ended or as the wet season began. The above clas- sifications were accomplished by determining the distribution of births across all months, as indi- cated by the distribution of juveniles, lactating fe- males, and embryo sizes. If a single birth was in- dicated for a given season, then that species was identified as having the potential to produce young during that season. The presence of juveniles or lactating females was assumed to indicate partu- rition during the month prior to capture. If a fe- male was carrying a large, near-term embryo, par- turition was assumed to take place in that month. On the other hand, if the embryo was very small relative to the adult body length, birth was as- sumed to take place four months later for most species, three months later for small insectivorous species (Findley & Wilson, 1 974), and five months later for emballonurids (Bradbury & Vehrencamp, 1977) and Desmodus rotundus (Wilson, 1979). Classification of a species as one that can give birth during the dry season does not necessarily mean that it does so each year. I was unable with this method to determine the actual frequency of births during either season. A dry season classi- fication simply means that individuals can at least occasionally produce young during the dry season. The lack of birth records for a given species for either season can be the result of an actual absence of birthing during that season, or it may be due to an inadequate sample from that season. If I as- sume that the lack of a breeding record for a given season is not due to inadequate sampling, then I can use the G test (Sokal & Rohlf, 1981) to assess the significance of the differences between the groups of species compared. This is a safe as- sumption, since most of the specimens were col- lected during the Peruvian dry season (the season with which this study is principally concerned) and because the groups of bats predicted to have the greatest proportion of dry season births are the groups most poorly sampled (see Appendix and below). Regions and Climate I follow Koopman's ( 1978) division of Peru into three zoogeographical regions. He lists SF)ecies as lowland if they were collected below 1000 m east of the Andes and highland if collected above this elevation. Pacific coastal species are those col- lected either along the arid coast or adjacent An- dean foothills (including the mesic areas of north- western Peru). The Pacific coastal plain is characterized by low (< 50 mm) annual rainfall, most of which falls between December and April. The mesic areas on the northwestern Andean slopes experience great- er rainfall over an extended period. Rainfall east of the Andes (fig. 1 ) is seasonal and abundant, with the greatest amounts deposited between the ele- vations of 1 000 and 3000 m (Bowman, 1916). The wet season begins in October and continues through April and the dry season begins in May and ex- tends through September. Most investigators agree that food resources are primarily modulated by rainfall seasonality (Jan- zen, 1967, 1973; Frankie et al., 1974; Ricklefs, 1975; Buskirk & Buskirk, 1976; Wolda, 1978a,b, 1982; Foster, 1982; Smythe, 1982; Terborgh, 1983). Many of the following generalizations on seasonal changes in food supplies for bats of mid- dle and higher elevations are based on the as- sumption that Peruvian plants and insects respond to environmental conditions in a manner similar to closely related organisms in Costa Rica, Pan- ama, or Puerto Rico (as described by the above authors). Terborgh's (1983) study of the changes in fruit, flower, and insect supplies in Manu Na- tional Park indicates that in the lowlands of south- eastern Peru, fruit abundance and diversity usually increase in October with the onset of the wet sea- son, peak in November and December and again in March, and decrease to the lowest level in May and June. Not only is the first peak greater but it also includes a greater proportion of plants in the family Moraceae (Terborgh, pers. comm.), which have fruit favored by bats (Gardner, 1 977). Flower abundance peaks in the dry season, but some flow- ers are present throughout the year (Terborgh, 1983). At middle to higher elevations, the phe- nology of fruit and flower species is probably sim- ilar to lowland species, but less seasonal (Ter- borgh, 1977; Nevling, 1971). Changes in insect resources are more difficult to generalize because the species in different size classes (Smythe, 1982; Wolda, 1982) and in different habitats (Bradbury & Vehrencamp, 1976; Terborgh, 1983; Janzen, 1983) have different periods of peak abundance and diversity. If the supply across size classes and habitats is considered, then it is probably rather high throughout the year. I will not discuss the GRAHAM: PERUVIAN BATS 175 600. 500. 400. 5 300. I 200_ 100. Month Fig. 1. Average rainfall amounts at Iquitos (lowland) and Yurac (middle el- evation), Department of Loreto, Peru. Data are taken from a map published in 1971 by the Servicio Nacional de Me- teorologia e Hidrologia in Peru and are averaged from 1 0 years of records. seasonal changes in food levels west of the Andes in Peru because less is known of this region. Results More than 3,400 female specimens divided among 109 species were included in this study. Twins were recorded for only one female, a Car- ollia perspicillata, collected in November bearing two well-developed embryos (24 and 29 mm). Specimens were collected almost exclusively be- tween May and December, with June through Au- gust having the largest samples (see sample sizes in Appendix). The reproductive records are also unevenly distributed among the species (Appen- dix). These sampling problems made it difficult to identify the reproductive patterns for most species. If, however, the percentages of all the females that were pregnant are determined for each species for each month, many species have data that fit the pattern of seasonal polyestry. This pattern is well illustrated by the pregnancy curve ofCarollia per- spicillata (fig. 2). Pregnancy levels peak in Septem- ber and December and are followed by periods of parturition, as suggested by the October drop in the frequency of pregnant females and by the ju- veniles collected in January (Appendix). The be- ginning of the second reproductive cycle of the season is also indicated by the females of Carollia castanea (l), Vampyrops dorsalis (7), and Artibeus planirostris (1) that were simultaneously lactating and pregnant in November and December (Ap- pendix). On the other hand, the pregnancy curve of Myotis (fig. 2) and the records of juveniles and lactating females of Carollia (fig. 2) and Artibeus (Appendix) for May-September indicate that births can occur outside of the wet season. Individuals of some species (i.e., Desmodus rotundus, Myotis nigricans, and perhaps Glossophaga soricina and Artibeus planirostris; see Appendix) may be able to produce young during any month of the year. Table 1 lists the number of species in each fam- ily or subfamily with births recorded during the seasons. I was able to identify birth periods for 79 (72.5%) of the 109 species listed in the Appendix. Most species (63 of the 79. 79.7%) have birth rec- ords for the wet season but a surprisingly large proportion of all of the species (46 out of 79, 58.2%) have records of dry season parturition. The ab- sence of birth records during the wet season 176 HELDIANA: ZOOLOGY Carollia perspiclllata 60-1 0 0 0 3 3 24 38 100 18 15 23 15 Fig. 2. Changes in pregnancy levels of Carollia perspiclllata and Myotis nigricans. The numbers across the top refer to the sample sizes for each month. The ordinate represents the percentage of all females for each month that were recorded as pregnant. (Mormoopidae) and the dry season (Furipteridae and Thyropteridae) may represent real periods of no births but may also be the result of inadequate sampling. Are there groups of bats that have a greater ten- dency than other groups toward parturition during the dry season? If seasonal fluctuations in food supplies are less in the highlands than in the low- lands, then proportionately more highland than lowland species could be expected to produce and wean young during the dry season and, if most species are polyestrous, during both seasons. Thir- ty-one of the 58 lowland species (53.4%) and 17 of the 27 highland species (63.0%) have records of dry season births (table 2). Although the 9.6% difference is in the predicted direction, it is not significant {P > 0.25). The highland region has a greater proportion of species with birth records from both seasons (51.9% vs. 34.5%, P < 0.05). The highland Sturnira bogotensis that were lac- tating and pregnant in June and August (Appen- dix) provide evidence that some species with dry season birth records are polyestrous, becoming pregnant again during the dry season. Bats of the Pacific coastal and slope region also show a strong tendency toward dry season births. If trophic generalists (those that consume more than one type of food, such as fruit and insects) can switch to another resource when one type be- comes scarce, they should be less vulnerable to seasonal fluctuations in their food supply than are trophic specialists (those species that use only one major type of food). This reduced seasonality of food resources should be reflected by a tendency for births to occur during the dry season and, if polyestry is common, for both seasons. Nine of the 1 1 generalist species (81.8%) and 23 of the 68 specialists (33.8%, table 2) have records of indi- viduals that have given birth during the dry sea- son, and relatively more generalists (63.6%) than GRAHAM: PERUVIAN BATS 177 Table 1. Number of species of Peruvian bats with births recorded for each season. Family Dry Season* Wet Both Total spe- ciest Emballonuridae 1 3 4 Noctilionidae 2 1 1 2 Mormoopidae Phyllostomidae Phyllostominae Glossophaginae Carolliinae 2 6 6 2 6 4 5 3 3 2 2 9 7 5 Stumirinae 4 6 4 6 Stenodermatinae 10 22 10 22 Desmodontinae 1 2 1 2 Furipteridae Thyropteridae Vespertilionidae Molossidae 7 5 1 1 7 5 4 2 1 1 10 8 Totals 46 63 30 79 * The number of species with records of birth in either the dry or the wet season or in both. A species is listed in all three columns if it has birth records for both sea- sons. t The total number of species that have data indicating birth periods. The sum of the dry and wet season records does not equal the total, because some species have rec- ords for both seasons. specialists (29.7%) have records for births during both seasons (table 2). These large proportional differences are significant (P < 0.01 and P < 0.05, respectively). Since floral resources are greatest in the dry sea- son when fruit resources are lowest, nectarivorous species should show a greater tendency for indi- viduals to produce and wean young during the dry season compared to frugivores. Six of the seven (85.8%) bats that are at least partially nectarivo- rous have birth periods in the dry season com- pared with less than half (1 7 of 40, 42.5%) of those that are partially frugivorous (table 2). This large relative difference is significant (P < 0.05). Dry season production and weaning of young is prob- ably the common pattern for most bats that eat nectar. This is indicated by the large number of juveniles (13) and lactating females (35) recorded for this group from May through August. Since the resource abundances are reversed during the wet season, the pattern of frugivorous sp>ecies hav- ing relatively more birth records (39 of the 40 species that are partially frugivorous, 97.5%) than nectarivores (3 of the 7 species that are partially nectarivorous, 42.9%) was expected {P < 0.001). Percentages of the species in these trophic groups that have birth records for both seasons are similar (40.0% and 42.9% for frugivores and nectarivores, resp)ectively, P > 0.50). The number of insectivorous species that have birth records for each season is almost equal (17 for the dry, 1 8 for the wet). This is the expected pattern if insect abundance and diversity remain high throughout the year. If a species is seasonally polyestrous, the prob- ability of recording a birth for the wet season for that species is greater, since both young are pro- duced during that season. If most species are sea- sonally polyestrous, then the majority of species should have wet season birth records. All of the ecological groups mentioned above (except the nectarivores) show considerable evidence of par- turition during the wet season (table 2). Discussion Seasonal polyestry is the dominant pattern in other studies of Neotropical bat reproduction (Hemingetal., 1972; Thomas, 1972;Taddei, 1976; LaVal & Fitch, 1977; Myers, 1977; Bonaccorso, 1979; Wilson, 1979; August & Baker, 1982). The majority of bats in Peru, especially frugivores, give birth during the wet season. Although this pattern is expected if most species are seasonally polyestrous most of the time, this conclusion is compromised by the possibility that some species may be monestrous, producing and weaning their young during the wet season. I was unable, for most species, to distinguish between these two re- productive patterns. However, the females col- lected from November and December that were simultaneously lactating and pregnant provide evidence for seasonal polyestry. Females that fol- low this pattern wean their first young concur- rently with the onset of the heavy rains, as fruit supplies are reaching their peak, and wean their second young near the middle of the wet season when food resources remain plentiful. In Peru, as in many of the other Neotropical areas discussed in the above studies, young bats are at least occasionally bom and weaned in the dry season. These bats are members of zoogeo- graphical and ecological groups whose tendencies toward dry season births can be predicted. If food availability for a particular group of bats does not fluctuate greatly throughout the year relative to a second group, then the first group should include a larger percentage of species with birth records for the dry season and, if the species are polyes- 178 HELDIANA: ZOOLOGY Table 2. Number of species of Peruvian bats with births recorded for each season, by geographical region and ecological group. Season* Region Dry Wet Both Total and group % (N) % (N) % (N) species* Geographical Regions Lowlands 53.4(31) 79.3 (46) 34.5(19) 58 Highlands 63.0(17) 88.8 (24) 51.9(14) 27 Pacific coastal 80.0 (8) 60.0 (6) 30.0 (3) 10 Trophic Groups Insectivores 58.6(17) 62.1 (18) 22.2 (6) 29 Frugivores 38.7(12) 100 (33) 38.7(12) 33 Nectarivores 66.7 (2) 33.3(1) ... 3 Piscivores 100 (1) ... 1 Vampires 50.0(1) 100 (2) 50.0(1) 2 Frugivore/insectivores 60.0 (3) 80.0 (4) 40.0 (2) 5 Nectarivore/insectivores 100 (3) 66.7 (2) 66.7 (2) 3 Piscivores/insectivores 100 (1) 100 (1) 100 (1) I Omnivorest 100 (2) 100 (2) 100 (2) 2 Trophic Specialists:}: 33.8 (23) 79.4 (54) 29.7(19) 68 Trophic Generalists§ 81.8(9) 90.9(10) 63.6 (7) 11 * See notes to Table 1 . N = Number of species; % = % of total species. t Includes species that eat more than two major types of items, t Includes insectivores through vampires (see text, p. 1 77). § Includes frugivores/insectivores through omnivores. trous, for both seasons. Two groups of Peruvian bats, highland species and trophic generalists, pos- sibly experience less seasonal fluctuation in food supplies relative to lowland species and trophic specialists. However, the reasons for the damp- ened fluctuations differ with each group. At high elevations in Peru, even during the dry season, clouds cover the forests most of the time (Terborgh, 1977; pers. obs.), providing moisture that is largely unavailable to lowland plants. This yearlong availability of moisture presumably keeps plant productivity and ultimately food availability from fluctuating as greatly as in the lowlands. This reduced fluctuation in food supplies is reflected by a weak trend toward relatively more highland birth records for the dry season and by a stronger high- land trend for birth records for both seasons. Species with generalized feeding habits are thought to be able to switch resources when one becomes reduced. This dampens the effect of sea- sonal changes in food availability and explains their larger percentages of birth records for the dry season and for both seasons. Fleming et al. ( 1 972), Heithaus et al. ( 1 975), and Bonaccorso ( 1 979) pro- vide data from Costa Rica and Panama that il- lustrate seasonal switches made by several of the same species included in my study as feeding gen- eralists (i.e., Phyllostomus discolor and Glossoph- aga soricina). Relatively more nectarivorous than frugivorous species have individuals with dry season birth and weaning records, but the records for both seasons are not different. These were the expected patterns, since both groups presumably experience seasonal variations in their food supplies; floral resources are greatest and fruit resources lowest during the dry season (Terborgh, 1983). Hence, dry season births may actually be the rule for bats that con- sume nectar but the exception for frugivores. The reverse is true for the wet season. These seasonality differences explain the almost equal proportions of species with birth records for both seasons. Thus, for trophic generalists, highland, and nec- tarivorous bat species, many females produce their first young in the dry season and become pregnant again, producing their second young during the wet season. Obviously, many more data are need- ed to clarify these patterns and to address the pos- sibility of differential survival of young between the two seasons. In Peru, the onset and termination of the rainy season can vary annually (Terborgh, 1983). Food supplies track these variations (Terborgh, 1983; Foster, 1982), and so may bat reproduction. I do not believe that climatic variability is the principal factor responsible for the births recorded in Peru between May and August because the ecological groups that I compared (frugivores and nectari- GRAHAM: PERUVIAN BATS 179 vores, trophic specialists and generalists) showed different seasonal birth patterns. If unusual dry season rains were the principal factor, then equal proportions of the different ecological groups should have responded by producing young during the normally dry season. My data support the hypothesis that dry season births are adaptations for producing and perhaps weaning young when food resources are available predictably or occasionally during this time. These patterns need to be confirmed by long term studies of actual bat communities, and of the resources used by individuals at both highland and lowland localities and during both seasons. Acknowledgments I am grateful to the many people who assisted me in Peru and in the United States. Antonio Brack E., Eric Cardich Briceno, Richard Bustamante M., and Susana Moller-H. of the Direccion General Forestal y de Fauna, Ministerio de Agricultura, Lima, provided the necessary collecting permits. Hernando de Macedo R. and Ramon Ferreyra of the Museo de Historia Natural "Javier Prado" were also helpful. I thank Arturo Koenig R., Man- uel A. Plenge, and Gustavo del Solar for their hospitality and help in making this study possible. I appreciate the assistance provided by Faucet and Aero Peru. Without the reliable assistance of my Peruvian associates, Manuel Sanchez, Klaus Wehr, and Reyes Rivera, the fieldwork would have been much more difficult. Karl F. Koopman (American Museum of Nat- ural History), Don E. Wilson (National Museum of Natural History), David J. Schmidly and Wil- liam B. Davis (Texas Cooperative Wildlife Col- lection), and Patricia W. Freeman (formerly of Field Museum of Natural History) were very help- ful during trips to their museums. Koopman and James L. Patton (Museum of Vertebrate Zoology) loaned fluid-preserved bats. James B. Cope (Jo- seph Moore Museum), Alfred L. Gardner (Na- tional Museum of Natural History), and Patton were kind enough to send me unpublished manu- scripts and field notes. For valuable assistance in the field I thank Linda J. Barkley, J. William Eley, Gary R. Graves, John P. O'Neill, Theodore A. Parker, III, Thomas S. Schulenberg, and Morris D. Williams. The excellent suggestions on the manuscript by Philip Myers were appreciated. I thank Susan T. Graham for her support during the several revisions of this work. 1 gratefully acknowledge the financial support of the LSUMZ Peruvian fieldwork by John S. Mcllhenny, Irving and Laura R. Schweppe, E. W. Mudge, and the late Babette M. Odom. Travel to other museums was made possible in part by an LSU Foundation-Graduate Student Travel Award. Literature Cited August, P. V., and R. J. Baker. 1982. Observations on the reproductive ecology of some Neotropical bats. Mammalia, 46: 177-181. BoNACcoRso, F. J. 1979. Foraging and reproductive ecology in a Panamanian bat community. Bulletin of the Florida State Museum, Biological Sciences, 24: 359^08. Bowles, J. B., J. B. Cope, AND E. A. Cope. 1979. Bi- ological studies of selected Peruvian bats of Tingo Maria, Departamento de Huanuco. Transactions of the Kansas Academy of Sciences, 82: 1-10. Bowman, I. 1916. The Andes of Southern Peru: Geo- graphical Reconnaissance Along the Seventy-Third Meridian. H. Holt and Co., New York, 336 pp. Bradbury, J. W.. and S. L. Vehrencamp. 1 976. Social organization and foraging in emballonurid bats. I. Field studies. Behavioral Ecology and Sociobiology, 1:337- 381. . 1977. Social organization and foraging in em- ballonurid bats. IV. Parental investment patterns. Be- havioral Ecology and Sociobiology, 2: 1 9-29. BusKiRK, R. E., and W. H. Buskjrk. 1976. Changes in arthropod abundance in a highland Costa Rican forest. American Midland Naturalist, 95: 288-298. FiNDLEY, J. S., AND D. E. WiLSON. 1974. Observations on the Neotropical disk-winged bat, Thyroptera tri- color Spi\. Journal of Mammalogy. 55: 562-571. Fleming, T. H., E. T. Hooper, and D. E. Wilson. 1972. Three Central American bat communities: Structure, reproductive cycles, and movement patterns. Ecology, 53: 555-569. Foster, R. B. 1982. The seasonal rhythm of fruitfall of Barro Colorado Island, pp. 151-172. In Leigh. E. G., Jr., A. S. Rand, and D. M. Windsor, eds.. The Ecology of a Tropical Forest. Smithsonian Institution Press, Washington, D.C.. 468 pp. Frankie, G. W., H. G. Baker, and P. A. Opler. 1974. Comparative phenological studies of trees in tropical wet and dry forests in the lowlands of Costa Rica. Journal of Ecology, 62: 881-919. Gardner, A. L. 1977. Feeding habits, pp. 293-350. In Baker, R. J., J. K. Jones, Jr., and D. C Carter, eds.. Biology of bats of the New World family Phyllosto- matidae. Part II. Special Publications, The Museum, Texas Tech University, 13: 1-364. 180 HELDIANA: ZOOLOGY Graham, G. L. 1983. Changes in bat species diversity along an elevational gradient up the Peruvian Andes. Journal of Mammalogy, 64: 559-571. Heithaus, E. R., T. H. Fleming, and P. A. Opler. 1975. Foraging patterns and resource utilization in seven species of bats in a seasonal tropical forest. Ecology, 56:841-854. Humphrey, S. R., and F. J. Bonaccorso. 1979. Pop- ulation and community ecology, pp. 409-441. In Ba- ker, R. J., J. K. Jones, Jr., and D. C. Carter, eds., Biology of bats of the New World family Phyllosto- matidae. Part III. Special Publications, The Museum, Texas Tech University, 16: 1-441. Janzen, D. H. 1967. Synchronization of sexual repro- duction of trees within the dry season in Central Amer- ica. Evolution, 21: 620-637. . 1973. Sweep samples oftropical foliage insects: Effects of seasons, vegetation types, elevation, time of day, and insularity. Ecology, 54: 687-708. . 1983. Insects: Introduction, pp. 619-645. In Janzen, D. H., ed., Costa Rican Natural History. The University of Chicago Press, Chicago, 816 pp. Koopman, K. F. 1978. Zoogeography of Peruvian bats with special emphasis on the role of the Andes. Amer- ican Museum Novitates, 2651: 1-33. LaVal, R.K., AND H.S. Fitch. 1977. Structure, move- ments and reproduction in three Costa Rican bat com- munities. Occasional Papers, Museum Natural His- tory, University of Kansas, 69: 1-27. Mares, M. A., and D. E. Wilson. 1971. Bat repro- duction during the Costa Rican dry season. Bio- Science, 21: 471^77. Myers, P. 1977. Patterns ofreproduction of four species of vespertilionid bats in Paraguay. University of Cal- ifornia Publications in Zoology, 107: 1-41. Myers, P., and R. M. Wetzel. 1983. Systematics and zoogeography of the bats of the Chaco Boreal. Mis- cellaneous Publications, Museum of Zoology, Uni- versity of Michigan, 165: 1-59. Nevling, L. I., Jr. 1971. The ecology of an elfin forest in Puerto Rico, 16. The flowering cycle and an inter- pretation of its seasonality. Journal of the Arnold Ar- boretum, 52: 586-613. Paradiso, J. L., ED. 1975. Walker's Mammals of the World, 3rd ed., vols. 1 and 2. Johns Hopkins Uni- versity Press, Baltimore, 1,500 pp. Racey, p. a. 1 982. Ecology of bat reproduction, pp. 57-104. In Kunz, T. H., ed.. Ecology of Bats. Plenum Press, New York, 425 pp. RiCKLEFS, R. E. 1975. Seasonal occurrence of night- flying insects on Barro Colorado Island, Panama Canal Zone. Journal of the New York Entomological Society, 83: 19-32. Smythe, N. 1982. The seasonal abundance of night- flying insects in a Neotropical forest, pp. 309-318. In Leigh, E. G., Jr., A. S. Rand, and D. M. Windsor, eds.. The Ecology of a Tropical Forest. Smithsonian Insti- tution Press, Washington, D.C., 468 pp. SoKAL, R. R., AND F. J. Rohlf. 1981. Biometry, 2nd ed. W. H. Freeman, San Francisco, Calif, 859 pp. Taddei, V. A. 1976. The reproduction of some Phyl- lostomidae (Chiroptera) from the northwestern region of the State of Sao Paulo. Boletim de Zoologia, Uni- versidade de Sao Paulo, 1: 313-330. Terborgh, J. 1977. Bird species diversity on an An- dean elevational gradient. Ecology, 58: 1007-1019. . 1983. Five New World Primates. Princeton University Press, Princeton, N.J., 260 pp. Thomas, M. E. 1972. Preliminary study of the annual breeding patterns and population fluctuations of bats in three ecologically distinct habitats in southwestern Colombia. Ph.D. diss., Tulane University, New Or- leans, La., 161 pp. TuTTLE, M. D. 1970. Distribution and zoogeography of Peruvian bats, with comments on natural history. University of Kansas Science Bulletin, 49: 45-86. TuTTLE, M. D., AND D. Stevenson. 1 982. Growth and survival of bats, pp. 105-150. In Kunz, T. H., ed., Ecology of Bats. Plenum Press, New York, 425 pp. Wilson, D. E. 1979. Reproductive patterns, pp. 3 1 7- 378. In Baker, R. J., J. K. Jones, Jr., and D. C. Carter, eds., Biology of bats of the New World family Phyl- lostomatidae. Part III. Special Publications, The Mu- seum, Texas Tech University, 16: 1-441. WoLDA, H. 1978a. Fluctuations in rainfall, food and abundance oftropical insects. Journal of Animal Ecol- ogy, 47: 369-381. . 1978b. Fluctuations in abundance of tropical insects. American Naturalist, 112: 1017-1045. . 1982. Seasonality of Homoptera on Barro Col- orado Island, pp. 319-330. In Leigh, E. G., Jr., A. S. Rand, and D. M. Windsor, eds., The Ecology of a Tropical Forest. Smithsonian Institution Press, Wash- ington, D.C., 468 pp. GRAHAM: PERUVIAN BATS 181 s I a o - & g i I i I I 1 Ci — — § "" ^^ ^ ^ _ -^ ^^ ,«i^ ^^ .^-^ ir> ■* o VO 7^ ^m^ M f^ ir> (N ^^ o f^ ■* o OS O 00 f<^ vo iri r> Tf 00 .^ik^ ^-^ ^-^ ,^i^ ■* 00 (N 00 . . o ^^ f«^ O "5 ^ c ■«: B '^ c: ft, vj H K ^ ^ ^ C .s ft. ft.^ S^ S^ ft. ^ U 5 3 .J U Qu U o: :©:: :o: • • .; • ■ • -ooo- • o u X hJ hJ U u ^ ^ ^ U u cu u X u .J -j u u u u .'^ .'^ c -2 -2 111 Z ft. ft^ 3 3 S3 I o o o o - c s: c c "OOP o 1 1? ? ? 3 If t?.s 5 S 5 3'^§ a ."> .'^ .'o .'^ .'^ .'^ 2 C C C C C C - i» ii» w ^ ^ %j ^ o 5; a: a: (S; ^ 6^ O ft. s: k: p o c c c 2 P P P ■? 5 C a a 5 :* .Vj .U 3 .5 o to ^ 3 s a bo ft, ts c 2 •^-2 2 S -2 S c c ^ — -2 -2 -2 -2 -2 -2 ;: 13 := o ti^^^^^ 3 3a pot • 0 e-. Q^QQ^c^. QQQ ^ : o : : o — O Tf c^ v y—s^-s ^-s On — . Tt — Tf vo . r<^ m . . . . O . fn r-~ . . : o o o m o 1/^ r^ n >n /^ tT) o o 00 O r> o t^ • >n ^ o o ^^ --M ^-M ^* ?^ ^^ r<^ r- ^^ CTn ,^1^ ^^1^ (N On ^.^ Tt ^M. ^^^ VO >n ^^ ^M ^^ O •-^ VO t— 1 (N r«% ^H <— 4 VO (N Tf 00 »-H OS — M O m o rs| PO On O O O^ ro " ■* fN o m so f«^ n ^^ ^^ ^». ^^ OS OS ^.^Tt v?~. OS ^^^.v^»,^^ >N«r> c<^ fN — -H o (N — OO CA >/~i rri o o : o ^ : o ■ o UXQu,Bu.JUXhJ.-1hJ3:.4X.-IXnJ0UnJ^ : ^ _) ^ X ^ £ >J iJ ^u^XXX^XX <3 <3 2 S '5 — 5 5 ^ t ■t^ > ?• So so ^ £! ^ « .2 .2 .2 sill P r'3 r'^ r^ U - >>. a a C^, ft, Sk. Ss £ £ S S a ,a ^ -5 5j a c^ c c S 5 5:5 I 2 5 5 5 S .«o .^ »» »» Ik g 333333333 3 3 3 Co (>5 l>5 ^ GRAHAM: PERUVIAN BATS 183 a < i ^ ec CT ^^ ^ ^ ^ ^^^^ ^ Q .. .. ^ ^^Q^c^ ^^^^^Q^c^ QQQQc^ Q^ ..a. ^^ 50(2) 20(15)' 69(13)' 100(1) 0(4)' o : 0(1) 73(30)' 50(2) 1^ i i : o : S" o 00 m — " . O 00 O r\ ir> o : — f;i t^ CM oo" o o i- ^vO • • • o r'^ : '. '. tn 00 50(2) 50(2) 33(3)' . . (N 00 . : (N 00 : ■^ 00 oo : — . O VO op : : : : g OO vO O ^^o ^-^^^3" f«-, _ _ 00 — O O O O rn 5 . ^irTC"-, o o : oo : oo O O ON 1^ o o r^ r^ fo O ^ ^*^ O CM r- m (^ >/^ ■y-^ O m PO m vO (N 1^ >/% m vp f»^ _ o o o O O O o Tf o m o >o — ' r) — * o l/^ ,^^ rr fN ^^ m ^M o o o o o O O «-> 00 r~ m r~ o so o . •c ^ 2 'c ^ •c o c 4> a ^ r .!>> 2 >> i! 1 H > .Jo .50 g :§ ::§ :g S S -id a a a ^ I'll >- o c ^ V, i a ^ 3 3 >i .^ .~ -2 c: c .lo .to •i C c ^ ^ 5 <3 . , a -o •=: ^ S 3 3 ■^ yi to .tn .*o .>) .^ ,^ .'o .^o .tn .l^ .^ .^ .y I 5^^ 3 3 S o o 5> 5 5 5 9

3 ^ (^ GRAHAM: PERUVIAN BATS 185 I' ^ e-. ^ e-. C-. e-- Q ^ Q ^ Q Q c^. Q . m — • g2 O O s a M 9 ^J^^ f^:^:^^ s s s I ^ ^ 1^ ? S S -S .'^ c ^ ^ 'J "2 ^ <3 53 CiO Ci. g to to » c^ i 5 5 s a a; uj t*j uj ttj a •c 3 •o Ka a, -c CL, m c II « o S ^ T3 00 c s < £ x^ o E w « 00 S o .2 u ^ u tu Q. a Q.-= JQ 2 •o U (A - Si ifi c a* ti C = E C/5 186 FIELDIANA: ZOOLOGY Tent Construction by Bats of the Genera Artibeus and Uroderma Robert M . Timm ABSTRACTS Herein, I describe new styles of tents cut and utilized by Artibeus anderseni, A. glaucus, A. gnomus, A. phaeotis, A. toltecus, A. watsoni, Uroderma bilobatum, and U. magnirostrum; review and summarize the literature on tent use by Artibeus and Uroderma; and discuss the effectiveness of tents as diurnal roosts. Artibeus anderseni alters the shape of Heliconia leaves by cutting the lateral nerves and interconnected tissue extending out from the midrib. Artibeus glaucus cuts the basal lateral nerves in Xanthosoma, causing the two sides of the leaf to collapse downward around the midrib. Artibeus phaeotis cuts the lateral nerves and interconnected tissues in both banana and Heliconia imbricata; the basal cuts veer out from the midrib such that a distinctive V-shaped enclosure is formed by the hanging leaf Artibeus toltecus cuts the basal nerves on Anthurium, causing the sides of the leaf to fold down around the midrib to form a pyramid- shaped tent. Artibeus watsoni was found to make four distinctive styles of tents, including simple V-shaped cuts on bifurcated palms, cuts of a few side veins on aroids to produce a rounded pyramid, elongate J-shaped cuts on banana and Heliconia, and polygonal cuts on Carludovica pal mat a. Artibeus watsoni has the greatest repertoire in tent styles, and uses the most diverse array of plant species and leaf shapes. Two styles of tents constructed by Uroderma bilobatum are reported for the first time, one on the large pinnately leafed palm Scheelea rostrata and the second on banana. The common denominator between the Uroderma bilobatum tents reported herein and those previously described is that all are on large, broad leaves and all have a distinctive V-shaped pattern cut by the bats. Uroderma magnirostrum also creates an inverted elongate V-shaped tent on pinnately leafed palms. All New World tent-makers described to date are tropical members of the phyllostomid subfamily Stenoderminae. Each species of tent-making bat has one or more distinctive style of tent. Bats select leaves of specific shapes, sizes, and angles for tent construction. Most species appear to be obligate tent-roosters. Tents provide bats with a cryptic diurnal roost site, in addition to providing shelter from both the sun and rain and an early warning to the approach of predators. Aqui yo describo nuevos estilos de carpas cortadas y utilizadas por Artibeus anderseni, A. glaucus, A. gnomus, A. phaeotis, A. toltecus, A. watsoni, Uroderma bilobatum, y U. magnirostrum reviso y hago un sumario de la literatura acerca del uso de carpas por filostomidos; y discuto la efectividad de las carpas como perchas diumas. Artibeus anderseni altera la forma de las hojas de Heliconia cortando las nervaduras centrales y tejido interconectado que se extiende From the Division of Mammals, Field Museum of Natural History, Chicago, Illinois 60605-2496. The au- thor's present address is Museum of Natural History and Department of Systematics and Ecology, University of Kansas, Lawrence, Kansas 66045. TIMM: ARTIBEUS AND URODERMA 187 exteriormente desde la nervadura central. Artibeus glaucus corta la base de las nervaduras laterales en Xanthosoma, causando asi que los dos lados de la hoja colapsen hacia abajo al rededor de la nervadura central. Artibeus phaeotis corta las nervaduras laterales y tejidos in- terconectados en banana y Heliconia imbricata; los cortes basales viran hacia afuera desde la nervadura central de tal modo que una distintiva cavidad en forma de V es formada por la hoja colgante. Artibeus toltecus corta las nervaduras basales de Anthurium, adi que los lados de la hoja doblen en redor de la nervadura central de tal modo que una cavidad es formada en forma de una piramide. Artibeus watsoni fue encontrada haciendo cuatro distintos estilos de carpas, incluyendo simples cortes en forma de V en palmas bifurcadas, cortes en unas pocas nervaduras laterales en araceas para producir una piramide redondeada, cortes alargados en forma de J en banana y Heliconia. y cortes poligonales en Carludovica pal mat a. Artibeus watsoni tiene el mas grande repertorio en estilos de carpas y usa la mas diversa serie de plantas y formas de hojas. Dos estilos de carpas construidas por Uroderma bilobatum son reportadas por primera vez; una en la larga y pinnada hoja de palma Scheelea rostrata y la segunda en banana. El comun denominador entre carpas de Uroderma bilobatum repnirtadas aqui y aquellas previa- mente descritas es que todas usan hojas grandes y anchas y todas tienen un distintivo patron en forma de V cortado por los murcielagos. Uroderma magnirostrum tambien corta una carpa en forma de V-invertido en hojas pinnadas de palmas. Todos los filostomidos cortadores y utilizadores de carpas descritos del Nuevo Mundo son membros tropicales de la subfamilia Stenoderminae. Cada especie tiene uno o mas estilos distintos de carpas. Los murcielagos escogen hojas de formas especificas, y constroen sus carpas en angulos peculiares. La mayor parte de las especies parecen utilizar las carpas obligatoriamente. Las carpas oferecen una percha oculta durante el dia, asi como un abrigo del sol, de la lluvia, y de predadores. Neste trabalho, (1) descrevo novos estilos de tendas cortadas e utilizadas por Artibeus an- derseni, A. glaucus, A. gnomus, A. phaeotis, A. toltecus, A. watsoni, Uroderma bilobatum, e U. magnirostrum; (2) reviso e resumo a literatura sobre o uso de tendas pelos morcegos da familia Phyllostomidae, e (3) discuto a eficiencia de tendas como alojamentos diumos. Artibeus an- derseni altera a forma das folhas de Heliconia, cortando as veias laterais e os tecidos interligados que estendem da veia central. Artibeus glaucus corta as veias basilares laterais em Xanthosoma, causando com que os dois lados da folha caiam contra o centro. Artibeus phaeotis corta as veias laterais e os tecidos interligados nas folhas de bananas e de Heliconia imbricata. Os cortes basilares partem da veia central, formando um abrigo distinto em forma de "V". Artibeus toltecus corta as veias basilares de Anthurium, causando com que os lados da folha dobrem em volta da veia central, criando um abrigo em forma de piramide. Artibeus watsoni constroi quatro tipos diferentes de tendas, incluindo simples cortes em forma de "V" em folhas de palmeiras bifurcadas; cortes em algumas das veias laterais em folhas de trepadeiras (resultando em piramides redondas); cortes alongados, em forma de "J", em folhas de bananas e de Heliconia; e cortes poligonos em fdlhas de Carludovica palmata. Artibeus watsoni possue o maior repertorio de estilos de tendas, e usa o conjunto mais di verso de especies de plantas e de configura96es de folhas. Dois estilos de tendas construidas por Uroderma bilobatum sao descritos pela primeira vez; um nas fdlhas grandes da palmeira Scheelea rostrata, e outro nas folhas de bananas. Fatores comuns entre as tendas construidas por Uroderma bilobatum aqui descritas, e as descritas previamente, sao a forma distinta em "V" cortada pelos morcegos, e o uso de folhas grandes e largas para a construfao das tendas. Uroderma magnirostrum tambem constroe tendas em forma de "V" invertido nas folhas de palmeiras. Todos morcegos construidores de tendas no Novo Mundo p)ertencem a subfamilia Steno- derminae (familia Phyllostomidae), e cada especie exibe um ou mais estilos caracteristicos de constru9ao. A maioria destas especies de morcegos parecem alojar-se obrigatoriamente em tendas, as quais oferecem nao so um abrigo diumo camuflado, mas tambem prote9ao contra sol, chuva, e predadores. 188 HELDIANA: ZOOLOGY Introduction The use of cut leaves for diurnal roosting sites by bats was first described by Thomas Barbour (1932), who reported on diurnal roosts of Uro- derma bilobatum near the Panama Canal. He found these bats roosting under the leaves of two culti- vated palms identified as Livistona chinensis and Prichardia pacifica. Not only were these bats found roosting under palm fronds, but they had also al- tered the leaf to produce a diurnal roosting struc- ture. Barbour (1932, p. 307) stated that "by nip- ping the ridges of the plications on the under side the leaf is weakened and as the bitten spots are skillfully and serially distributed the leaf finally is sufficiently weakened so that the distal portion droops sharply downward." Chapman (1932, p. 555) discovered Artibeus watsoni roosting under the cut veins of a bifurcated palm, Geonoma cu- neata (reported as G. decurrens), on Barro Colo- rado Island, Panama, and first called these mod- ified leaves "tents." Ingles (1953) also reported A. watsoni constructing tents in the palm Geonoma oxycarpa (reported as G. binervia) on Barro Col- orado Island. Goodwin and Greenhall (1961, p. 262) found Artibeus cinereus roosting "under cut leaves of palm trees and on the under side of ba- nana leaves" and Uroderma bilobatum roosting under cut leaves of the carat palm, Sabal glau- cescens, and coconut palm, Cocos nucifera, on Trinidad. Ectophylla alba was reported by Timm and Mortimer (1976) to alter the leaves of five species of Heliconia in Costa Rica; the bats se- lected specific leaves for both size and angle to the ground. Artibeus jamaicensis was found by Foster and Timm (1976) roosting under the cut leaflets on a pinnately leafed palm, Scheelea rostrata, in a tropical dry forest in Costa Rica. Recently, Timm ( 1 984) reported tent construction by another phyl- lostomid, Vampyressa pusilla, in Costa Rica, and Koepcke (1984) found Mesophylla macconnelli utilizing similar tents in Peru. Only one Old World bat, Cynopterus sphinx (Pteropodidae), has been reported to alter the shape of palm leaves to pro- duce a diurnal roosting structure (Goodwin, 1 979). Reviews of roosting site selection by bats were provided by Tuttle (1976) and Kunz (1982). To date, seven species of phyllostomids {Arti- beus cinereus, A. jamaicensis, A. watsoni, Ecto- phylla alba, Mesophylla macconnelli, Uroderma bilobatum, and Vampyressa pusilla) have been re- ported to modify leaves of plants to produce diur- nal roosting sites herein referred to as tents. All are tropical members of the phyllostomid subfam- ily Stenodermatinae. The phyllostomid genus Artibeus, which in- cludes some 1 5 species, is widespread in the Noo- tropics from northern Mexico southward to Ar- gentina and Chile. These bats range in size from 10 g (A. anderseni and A. watsoni) to 70 g (A. lituratus). Uroderma, a closely related genus of medium-sized stenodermatines that includes only two species, is found from southern Mexico through the Amazon Basin of South America. The better known of the two species, U bilobatum, weighs from 13 to 21 g. Herein I describe tent construction and utili- zation by Artibeus anderseni, A. glaucus, A. gno- mus, A. phaeotis, A. toltecus, A. watsoni, Uroderma bilobatum, and U. magnirostrum; describe several new styles of tents; review and summarize the lit- erature on tent use by Artibeus and Uroderma; discuss the effectiveness of tents as diurnal roosts; and suggest directions for future research. Methods Descriptions of Study Areas Costa Rica— Bosque Brrancia, also known lo- cally as Bosque Blanco, is located 0.8 km west of Cuarto Cruces on the south side of the Pan Amer- ican Highway (Route 1) in Guanacaste Province, in the Pacific lowlands of western Costa Rica. Bos- que Brrancia is a restricted area of undisturbed lowland forest classified as Tropical Moist Forest; the dominant vegetation includes A nacardium ex- celsum and Scheelea rostrata. This stand of forest probably has not been logged previously, and rep- resents a close approximation to the original (pre- Columbian) forests of this part of Guanacaste. Further descriptions of this forest can be found in Janzen (1971) and Wilson and Janzen (1972). Parque Nacional de Corcovado is located on the Osa Peninsula of southwestern Costa Rica, Puntarenas Province (between 08°27'N and 08°39'N, and 83°25'W and 83°45'W); the eleva- tion ranges from sea level to 400 m. Corcovado lies within the Tropical Wet Forest Life 21one (Holdridge, 1967), with lowland evergreen forest being the dominant forest type. Mean annual rain- fall is 3,800 + mm and the wettest months are from August through November; mean monthly tem- peratures range from 25.0°C to 26.5°C. Vegetation and habitat types at Corcovado have been de- TIMM: ARTIBEUS AND URODERMA 189 scribed by Herwitz (1981) and Hartshorn (1983). Areas surveyed included both primary forest and secondary scrub along the coast. The La Selva Biological Station is the field sta- tion of the Organization for Tropical Studies lo- cated 1 km SW of Puerto Viejo de Sarapiqui, Heredia Province in the Caribbean lowlands of northeastern Costa Rica (10°27'N,84°00'W); ele- vation ranges from 29 to 100 m. Mean annual rainfall is 3,990 mm, with the wettest months being November, December, and February; mean monthly temperatures range from 24.5°C (Decem- ber) to 26.1°C (April). La Selva lies within the Tropical Wet Forest Life Zone, with lowland ev- ergreen forest being the dominant forest type. Vegetation and habitat types of La Selva have been described by Slud (1960), Holdridge et al. (1971), Sawyer and Lindsey (1971), and Hartshorn ( 1 983). One unusual feature of the subcanopy of the La Selva forest is the diversity and abundance of dwarf palms (Hartshorn, 1983), especially the broad- leaved species, Geonoma congesta and Asterogyne martiana. These species are regularly utilized by Artibeus watsoni for tent construction. The tran- sect surveys, which included all Artibeus tents ob- served, were restricted to primary forest. The Uro- derma tents described from this site were restricted to an open banana patch. Palo Verde (Refugio Nacional de Fauna Sil- vestre Dr. Rafael Lucas Rodriguez Caballero) is a wildlife refuge located 2 km S and 12 km E of Bolson, in the Pacific lowlands of Guanacaste Province, northwestern Costa Rica (10''30'N, 85°20'W); the elevational range is from 3 to 183 m. Palo Verde lies within the Tropical Dry Forest Life Zone, with lowland deciduous forest and riv- erine swamp forest being the dominant forest types. The vegetation and habitat types of Palo Verde have been described by Slud ( 1 980) and Hartshorn (1983). Mean annual rainfall is 1,700+ mm, with the wettest months being April, May, September, and October; mean monthly temperatures range from 26.0°C (November) to 29.7°C (April). The immediate vicinity of the survey at this site was in a mosaic of primary forest and secondary scrub that included considerable bananas scattered throughout. The newly expanded Parque Nacional Braulio Carrillo is located in northeastern Costa Rica (be- tween 10°05'N and 10°25'N, and 83°54'W and 84'^5'W); the elevation of the park ranges from 100 to 2600 m. Braulio Carrillo is located on the eastern, Caribbean slope of Volcan Barba in He- redia Province. The elevational range at which Ar- tibeus toltecus and associated tents were observed was from 700 to 1400 m, within the Premontane Rain Forest Zone with midelevational evergreen forest and tall palms being the dominant forest types. Mean annual rainfall at this elevational range is perhaps as much as 5,000 mm, although no exact measurements are yet available. The vege- tation and habitat types at the midelevational ranges also have yet to be described. On 14-15 April 1986, 3 km of trail ranging from 700 to 1 100 m were surveyed for bat tents. Additionally, an intensive netting effort with Richard K. LaVal and Don E. Wilson was conducted in this area over a 7-day period to determine what species of bats were present and their relative abundance. At Finca Las Cruces (2 km S of San Vito, Pun- tarenas Province, 08°45'N, 82°58'W, 1200 m) in the Premontane Wet Forest-Rain Forest transition area, approximately 2 km of trail leading down to the Rio Jaba was surveyed for tents on 1 3 August 1982. Ecuador— Cascada San Rafael lies 17 km (by road) west of the village of Reventador, Napo Province, in northeastern Ecuador (00°5.8'S, 77°34.4'W). Rainfall averages 1,500 to 2,000 mm; average temperatures range from 18°C to 22°C. The elevation is 1200 m. Reventador lies within the Humid Subtropical Life Zone. Lagarto Cocha and Zancudo Cocha are military encampments along the Rio Aguarico of Ama- zonian Ecuador named for prominent lagoons. Both areas are undisturbed primary lowland rain forests classified as Moist Forest, with an annual rainfall averaging from 2,000 to 3,000 mm; the elevation is approximately 200 m. Peru— Hacienda Amazonia lies just north of the confluence of the Alto Rio Madre de Dios and the Rio Pinipini in the department of Madre de Dios, southeastern Peru (12°58'S, 71'>09'W), just north of Atalaya. The Hacienda is located just east of Parque Nacional del Manu in the Upper Trop- ical Zone on the eastern foothills of the Andes. On 25 July 1985 Barbara L. Clauson searched the ridge above the Hacienda for bat tents at an ele- vation of 825 m in primary rain forest that had received some selective timber harvest; she re- turned to this site again on 3 November 1985. Cerro de PantiacoUa lies above the Rio Palotoa, 10-15 km NNW of Shintuya, in the department of Madre de Dios, southeastern Peru (12°35'N, 71''18'W). On 15 November 1985 Clauson searched a steep sloping ridge at 600 m. The sur- 190 HELDIANA: ZOOLOGY Table 1 . Individual measurements (in cm or degrees) collected from four tents constructed by Artibeus anderseni on Heliconia in Ecuador. Blade Petiole Basal height Length Width Angle Length Angle Remarks 176 158 154 140 34 38 36 27 40 15 50 80 75 88 60 80 55 80 70 200 220 260 140 1 adult male A. anderseni 3 male A. anderseni rounding forest was primary rain forest in the Up- per Tropical Zone on the eastern Andean foothills. Reference specimens of the bats have been de- posited at Field Museum of Natural History, Chi- cago; Escuela Politecnica Nacional, Quito, Ecua- dor; and Universidad Nacional de Costa Rica and Servicio de Parques Nacionales, San Jose, Costa Rica. Voucher specimens of plants have been de- posited in the herbaria at Field Museum, Duke University, Missouri Botanical Garden, and/or University of Wisconsin. Accounts of Species Artibeus Artibeus anderseni Osgood, 1916 Artibeus anderseni occupies an extensive range in western Amazonia; however, little is known of its biology. This species has long been considered a junior synonym of A. cinereus. In resurrecting A. anderseni as a distinct species, Koopman (1978, p. 14) stated, "Besides its shorter face and more abrupt forehead, A. anderseni apparently always lacks the last lower molars, which A. cinereus in western Amazonia almost always has." I concur with Koopman in recognizing A. anderseni as a distinct species. In late October and early November of 1983, the trails and forest surrounding the military en- campments were searched at Lagarto Cocha and Zancudo Cocha in eastern Ecuador. Artibeus an- derseni was found to alter the shape of leaves of several small, forest Heliconia species to produce diurnal roosting structures. To create a tent from a Heliconia leaf, the bat severs the lateral nerves and interconnecting veins that extend along both sides of the midrib. The cuts ran along the central 90% of the leaf from near (0 to 14 cm) the base to near (10 to 20 cm) the tip. Nerves and inter- connected tissues were severed, but not complete- ly, so that they did afford some support for the sides of the leaf. Cuts ran parallel to the midrib for most of its length, but did flare outward slightly toward the base. The lateral nerves were cut from 3 to 8 mm from the midrib; the midrib was not cut. Claw marks where the bats roosted started 50 cm from the base in one tent and ran for 1 6 cm distally; in another they started at 70 cm and ran for 10 cm distally. Measurements of the blade length, blade width, blade angle, petiole length, petiole angle, and basal height of four tents are provided in Table 1 . At Zancudo Cocha one Heliconia tent was un- occupied for two days in succession, then on the third day was occupied by an adult male with enlarged testes, an adult lactating female, and a juvenile male Artibeus anderseni (fig. 1). Another Heliconia tent was unoccupied. At Lagarto Cocha 13 tents were found in Hel- iconia. One was occupied by three subadult males not in breeding condition; a second tei t contained a single adult male with enlarged testes. Artibeus cinereus (Gervais, 1856) Artibeus cinereus, Gervais's fruit-eating bat, is found on the islands of Grenada, Trinidad, and Tobago, and throughout the Amazon Basin and adjacent coastal areas. Surprisingly little has been published on roosting behavior or ecology of this widely distributed species. On Trinidad, Goodwin and Greenhall (1961, p. 262) stated, "It roosts in small colonies of a few individuals under the cut leaves of palm trees and on the under side of ba- nana leaves." On Tobago, Husson (1954, p. 64) reported a single male Artibeus cinereus "hanging in a banana tree in cultivated country near the shore." TIMM: ARTIBEUS AND URODERMA 191 Fig. 1. Photograph of an adult male, adult female, and juvenile Artibeus anderseni roosting in a Heliconia leaf tent. Details of the cut side nerves can be seen along the midrib of the leaf Artibeus glaucus Thomas, 1 893 Artibeus glaucus is found at midelevalions along the eastern slopes of the Andes from Venezuela to Peru. The status of the name glaucus has long been in a state of flux; it often has been considered a subspecies of A. cinereus. I concur with Handley (1987) in regarding .4. glaucus as a distinct species. A single old adult male Artibeus glaucus (fmnh 124844) was observed roosting under a cut Xan- thosoma leaf at Cascada San Rafael, Ecuador, on 21 September 1983. Four cut leaves on separate plants were observed in close proximity to each other; a fifth cut leaf was observed approximately 10 to 1 5 m to the south. Only the one cut leaf was occupied by the single bat. An adult female A. glaucus was netted in the vicinity that evening. All tent leaves were cut down at the time and four were measured. On 26 November, two more cut leaves were found in this Xanthosoma population; one contained three A. glaucus. The Xanthosoma in which the bats were roosting were part of a population of Xanthosoma that occupied approx- imately 1 ha on a steep east-facing hillside. The basal lateral nerves from 2 to 5 (usually 3) of the Xanthosoma leaves were severed near the midrib, causing the two sides of the leaf to collapse downward around the midrib. The midrib was not cut. Four of the five cut leaves were measured in September; one found in November was measured (table 2). The leaves that were selected by Artibeus glaucus for tents all had the midrib running ap- proximately parallel to the ground, whereas the majority of unaltered leaves in the p>opulation stood at more vertical angles. Artibeus gnomus (Handley, 1987) Artibeus gnomus, the dwarf fruit-eating bat, is found in a peculiar circular range ringing the west- em edge of the Amazon Basin. Although this species has a wide distribution from Venezuela and Guyana to Peru, it was only recently recog- 192 HELDIANA: ZOOLOGY Table 2. Individual measurements (in cm or degrees) collected from five tents constructed by Artibeus glaucus on Xanthosoma in Ecuador. Blade Length Width Angle Petiole length Basal height Remarks 57 68 60 140 230 3 basal cut veins on each side, 2-3 cm from midrib 53 • • • 40 • • • 230 3 basal cut veins on each side, 1.5-3.0 cm from midrib 55 62 40 • •■ 230 2 veins cut left, 4 cut right side, 2-3 cm from midrib 64 36 45 137 300 3 basal veins cut on each side, 1.5-3.0 cm from midrib; adult male A. glaucus hanging 16 cm from base 61 66 20 145 275 5 basal veins cut on each side, 3.5-4.0 cm from midrib; 3 ^. glaucus nized as a distinct species, and little biological in- formation is available (Handley, 1987). On 15 November 1985 Barbara L. Clauson found a single adult male Artibeus gnomus roost- ing under a cut Monstera lechleriana leaf. The Monstera was growing as an epiphyte approxi- mately 10 m off the ground on a tree on a sloping hillside at 600 m elevation at Cerro de Pantiacolla, southeastern Peru. The single cut leaf was green and healthy and hung horizontally. No other cut leaves were observed in the immediate vicinity. The altered leaf was 70 cm long and 38.5 cm at its widest point. All lateral nerves along the basal nearly two-thirds of the leaf were severed imme- diately adjacent to the thick midrib; this included the basal 12 nerves on one side and 14 on the other. The midrib was severed at 44.5 cm from the base which caused the apical third of the leaf to droop downward perpendicular to the midrib. All nerves proximal to the midrib cut were sev- ered. The lateral nerves along the apical, drooping 25.5 cm were unaltered. The tent resulting from these cuts was quite en- closed, being formed by the sides of the leaf col- lapsing downward around the midrib and the dis- tal third of the leaf folding down, perpendicularly to the midrib. The lone Artibeus gnomus was roosting 9 cm toward the base from the severed midrib of the leaf. Artibeus jamaicensis Leach, 1821 Artibeus jamaicensis, the Jamaican fruit-eating bat, is found throughout much of tropical Central America, the northern half of South America, and the Greater and Lesser Antilles. At many localities this is one of the most common species of bats encountered and consequently has received more study than any other phyllostomid. Artibeus jamaicensis was reported roosting un- der the cut leaflets of Scheelea rostrata in Costa Rica by Foster and Timm (1976). Scheelea ros- trata is a large, pinnately leafed palm with the leaflets extending out at right angles from the hor- izontal rachis. Leaflets within the middle 1.3 m region of the frond were cut at varying distances that increased going up to the center of the cut area, then decreased. "As a result . . . the distal parts of the leaflets folded perpendicularly, hung vertically below the frond, and formed a broadly lanceolate tent" (Foster & Timm, 1976, p. 266). Although several Artibeus jamaicensis occupied the roost, only two males were captured, one an adult with enlarged testes, and the second a smaller male not in breeding condition. Artibeus jamai- censis apparently has a harem mating system, in which a single breeding male defends a roost used by several females and their offspring; nonbreed- ing males may be found either singly or in small groups (Morrison, 1979; Kunz et al., 1983). Ar- tibeus jamaicensis has been found roosting in a wide variety of situations, including caves, hollow trees, buildings, and under unaltered leaves (Tut- tle, 1976), and thus is certainly not an obligate tent-roosting species, as apparently are the smaller species oi Artibeus. Artibeus phaeotis (Miller, 1 902) Artibeus phaeotis, the pygmy fruit-eating bat, is found from central Mexico to northern South America (Timm, 1985). Most accounts of habitat for pygmy fruit-eating bats mention their being netted in close proximity to stands of bananas, Musa X para^/5/aca (Ramirez-Pulidoetal., 1977; Watkins et al., 1972). Davis (1970) suggested that they might roost under the leaves of bananas. During the summer of 1982, Artibeus phaeotis TIMM: ARTIBEUS AND URODERMA 193 Fig. 2. Top, Dorsal view of a banana leaf {Midsa x paradisiacd) showing the cut nerves running aloQg the midrib and flaring out toward the base; bottom, tent of Art ibeus phaeotis made from a banana leaf was observed roosting under the leaves of banana and Heliconia imbricata at Palo Verde and La Selva, respectively, in Costa Rica. In all cases the leaves had been altered to produce tents. Art ibeus phaeotis constructs roosts in both ba- nanas and Heliconia imbricata by biting the lateral nerves and interconnecting tissue that extend at right angles from the midrib, causing the blade to fold over in a V-shaped enclosure. The two sides of the leaf collapse downward, hanging beneath the midrib (fig. 2). Nerves and interconnected tis- sues are not completely severed, thus the sides of the leaf provide some support for the entire length of the cut. The cuts ran from the base of the leaf to near the tip (table 3). Near the base, the cuts flared out from the midrib to the sides to form an elongate J-shaped pattern. The uncut tip and basal portion of the leaf provide additional strength. The undersides of roost leaves are obscured from view from almost all angles except from directly be- neath the tent. To characterize Musa clumps, each of which presumably represented an individual plant, the number of stalks (ramets) per clump was counted; the height of each ramet was estimated to the near- est half meter. This was done for all clumps of Musa in the patches that could be located, both those with tents and without bat tents. When a bat tent was located, the following measurements were taken in centimeters: the angle of the petiole (mid- way from stalk to blade), angle of blade (midway on blade), petiole length, blade length, blade width, height of base of blade, height of tip of blade, height of roost, isolation distance (distance from nearest solid object on the same vertical level), length of uncut basal portion of blade, and length of uncut distal portion. I also noted whether the roosts were in direct sunlight or in shade (table 3). Also for all tents the number of stalks on that column was counted, and the age of the leaf rel- ative to other expanded leaves was recorded. The oldest (lowest leaf) on a plant was assigned the number 1 , then the rest counted up from there. I scored 232 individual stalks of Musa, which 194 HELDIANA: ZOOLOGY Table 3. Individual measurements (in cm or degrees) collected from 19 tents constructed by Artibeus phaeotis on leaves of banana, Musa x paradisiaca, in Costa Rica. Petiole Length Blade Angle Width Baiwl . height Height Tip Roost Isolation Uncut Length Angle Base Tip 43 40 110 -27 20 196 156 182 81 0 15 S3 38 136 -14 19 230 146 225 120 10 45 43 42 146 -40 22 213 170 228 125 1.5 48 49 12 120 -55 23 250 180 240 53 5 8 43 S2 100 -16 18 205 200 70 0 30 SO 55 120 -3 21 200 165 210 25 3 8 S4 50 142 -36 20 235 160 230 115 1 44 40 75 80 + 15 19 190 195 206 80 0,25 20 36 68 175 -12 26 320 260 320 >300 0 1.5 60 70 195 +25 30 360 400 400 > 300 0,35 0 44 66 130 +23 25 250 260 290 120 0 40 41 50 94 +5 20 140 155 170 40 0,40 25 34 45 97 -22 20 205 175 200 100 0 0 SO 68 130 0 23 180 220 230 60 0 2 S6 60 147 -37 27 260 180 240 35 0 28 33 55 85 -12 21 205 215 220 18 0 5 16 52 60 +28 13 130 140 35 0 12 46 50 129 + 15 21 250 220 260 120 0 65 40 54 88 + 15 14 170 180 220 0 3 were in 4 1 clumps scattered in 1 00 m of second growth forest. The mean number of ramets per clump was 5.4; the range from 1 to 23. I located 26 leaves that were cut by bats, a ratio of roughly one tent per 8. 1 ramets. Fourteen clumps had cut leaves; if a clump had cut leaves it had a mean of 1.9 cut leaves (range 1 to 2). Of the 26 altered leaves, complete data were taken on 19. The re- maining tents were decomposing; in a few cases the bats had not completed the tents. On the par- tially completed tents, only one side of the midrib was cut along one-quarter to one-third the distance of the blade, but did not cause the side to collapse. Only 2 of the 1 9 complete tents were located di- rectly in the sun. Bat tents were found in clump sizes ranging from 1 to 23 (mean = 6.2). The av- erage height of all plants (N = 232) was 2.3 m. The average height of plants with tents was 2.0 m (N = 16). One adult male with enlarged testes and six pregnant females were watched over a 3 -day p)e- riod in one of the roosts. Additionally, two solitary nonpregnant female Artibeus phaeotis roosted un- der separate leaves (fig. 3). Artibeus phaeotis appears to select banana leaves with specific characteristics. Usually these are the oldest fully expanded leaves, just over 2 m above the ground, with the center of the midrib nearly horizontal to the ground and positioned far enough from nearby stems and branches to limit access by predators. Roost sites generally are located in the shade of surrounding forest trees where ap- parently they are protected by the forest overstory from wind, blowing rain, and sunlight. Although bananas are not native to the New World, they are now common throughout the range of Artibeus phaeotis and probably provide roost sites in other localities. Heliconia, Calathea, and broad-leafed palms are uncommon at Palo Verde, hence are not readily available to A. phaeotis for tent sites there. Artibeus phaeotis used only banana leaves for tent making at Palo Verde, but con- structed similar tents in Heliconia imbricata at La Selva. Heliconia imbricata is the largest species of Heliconia in Costa Rica, and its leaves are similar in size and shape to banana. The tents in Heliconia were similar in all respects to those in banana leaves. An adult male and adult female were found roosting together under a single Heliconia leaf in late June. Villa-R. ( 1 967) found a single specimen roosting near the mouth of a small cave in Mexico. Artibeus toltecus (Saussure, 1 860) Artibeus toltecus, the lowland fruit-eating bat, is found along the coasts of eastern and western Mex- ico from Nuevo Leon and Sinaloa south through Central America and perhaps to extreme north- western Colombia. This species appears to be re- stricted to midelevational slopes, and in Costa Rica TIMM: ARTIBEUS AND URODERMA 195 Fig. 3. Photograph of an adult female Artibeus phaeotis roosting in a banana leaf tent. Details of the cut side nerves can be seen along the midrib of the leaf. I have found it only from 650 to 1 500 m in ele- vation. In April 1986 I observed a single Artibeus tol- tecus roosting under a cut leaf of Anthurium ca- peratum in Braulio Carrillo National Park, north- eastern Costa Rica. The Anthurium was growing as an epiphyte on a small tree at 800 m (IVi km S, 11 km E of San Miguel, 10°17'N, 84'>05'W). One leaf on the plant was altered; it was IVi m off the ground and the midrib hung parallel to the ground. Four or five lateral nerves were cut basally on each side, causing the sides of the leaf to fold down around the midrib. A break of the midrib at its midpoint caused the distal half of the leaf to droop down (fig. 4). Seven additional tents of this style were observed on Anthurium in this area, ranging in elevation from 700 to 1400 m. It is assumed that they were made by A. toltecus, the only small species of Artibeus we netted there, al- though these tents were not occupied. Six tents were observed in a 3-km transect ranging from 700 to 1 100 m in elevation. Davis (1944) reported that Artibeus toltecus roosts imder banana leaves, although he did not indicate that the bats were modifying the leaves. Davis (1944, p. 378) stated: . . . they had regularly established roosts tin- der the large, drooping leaves of the banana trees, each one easily recognized by the man- ner in which the vane of the leaf hung limply suspended from the midrib. The closely ap- pressed vanes of the leaf, plus the natural darkness within the depths of the grove, af- forded good concealment. These bats, too, were wary and that feature coupled with the nature of their retreat caused considerable difficulty in procuring specimens. In light of Davis's description of the roost sites of A. toltecus in banana leaves and my own obser- vations on A. toltecus, I suspect that this species was creating tents similar to those I observed for A. phaeotis in Costa Rica. The tents formed by A. phaeotis in Musa (see fig. 2) are similar in ap- 196 HELDIANA: ZOOLOGY Fig. 4. Top, Ventral view of an Anthurium caperatum leaf showing the cut nerves along the base of the leaf and the broken midrib; bottom, tent of Art ibeus toltecus. TIMM: ARTIBEUS AND URO DERMA 197 pearance to those described by Davis (1944) for A. toll ecus. Artibeus toltecus has also been reported in caves (Davis et al., 1964; Jones, 1966; Jones & Alvarez, 1 964), and Goodwin ( 1 934, p. 1 2) reported a single specimen collected "in one of the buildings (church?) at San Lucas . . . [the] rest of the con- gregation seemed to be Glossophaga." Artibeus watsoni Thomas, 1 90 1 Artibeus watsoni, Thomas's fruit-eating bat, is one of the smaller members of the genus Artibeus and found from southern Veracruz south through Central America to northern South America. It appears to be restricted to lowland and midele- vation humid forests. During the summers of 1974, 1982, 1984, and 1 986, numerous individuals of Artibeus watsoni were seen roosting under 19 different sj)ecies of broadleafed plants at several localities in Costa Rica. At Parque Nacional de Corcovado, Costa Rica, trails were surveyed on three separate oc- casions for the presence of tents made by Artibeus watsoni, in June and August 1982 and again in August 1984. In mid-June 1982, the following groupings of Artibeus watsoni were observed: two (both adults, a male and pregnant female), two (pregnant female and one not captured), two (not captured), and six hanging singly (of which two were captured and found to be adult males). Ad- ditionally, several tents on banana and Heliconia were marked for relocation later in the summer. A tent marked on Heliconia imbricata was relo- cated 60 days later. The tent was still intact, al- though it was beginning to break down; a single A. watsoni v/zs using it. All other marked tents had decomposed. On 10 August 1982 I found 90 tents constructed by Artibeus watsoni along the trail through Cor- covado's "Monkey Woods." These tents were made from the following species of plants: Musa x par- adisiaca (49, 54%), Anthurium ravenii (13, 14%), unidentified aroid (11, 12%), Heliconia imbricata (9, 10%), Heliconia latispatha (1, 1%), Heliconia sp. (3, 3%), and Calathea insignis (4, 4%). Tents located on Anthurium ravenii were most often found clumped, with an average of 2.6 tents per plant, whereas in the other species of plants it was uncommon to find more than one tent per indi- vidual plant. Bats were found singly (five) or in three groupings of four, three, and two individuals. Four of the single bats were all adult males. A trail running up to a ridge top was surveyed from 9 through 1 1 August 1982, with the following results: 25 tents found of which 16 were on Hel- iconia imbricata (64%), 8 on Calathea insignis (32%), and 1 on Carludovica palmata (4%). Three tents were occupied by two (sexes unknown), one male, and one female. Near the mouth of the Rio Llorona on 8 and 9 August I counted the following groups of bats: eight (three adult females, three young, and two not caught), two (adult female with volant young), two (sexes unknown), and three singles (one a nonreproductive adult female). Ad- ditional tents were observed in banana, coconut palm (Cocos nucifera), Calathea insignis, and Car- ludovica cf. drudei. In August 1 984, 1 found 63 tents constructed by Artibeus watsoni. These were distributed on the following plants: Anthurium ravenii (36, 57%), Heliconia sp. (14, 22%), Musa x paradisiaca (7, 1 1%), Calathea insignis (3, 5%), Carludovica pal- mata (\, 1.6%), IVelfia georgii (1 , 1.6%),and(/eo/i- oma sp. (1, 1.6%). Only 3 of the 63 tents were occupied; one had two bats and two each had sin- gle bats. As I noted in 1982, tents on Anthurium ravenii were often clumped on the same plant with an average of 2.5 tents per plant. At La Selva in July 1982, 43 Artibeus watsoni tents were located over a 5 -day period in the fol- lowing species of plants: Asterogyne martiana (33, 77%), Geonoma congest a (6, 14%), Geonoma cu- neata (2, 5%), and an unidentified species of Cy- clanthaceae (2, 5%). One adult male A. watsoni was found under an Asterogyne martiana tent on the first day. On the fifth day an adult female with young was found under another A. martiana tent that had been unoccupied for the previous fovu" days, as was a third adult (not captured) under another A. martiana tent. All other tents were un- occupied. In 1974 1 surveyed approximately 10 km of trails at La Selva and found 29 tents on the following species of palms: Asterogyne martiana (19, 66%), Bactris wendlandiana ( 1 , 3%), Geonoma congesta (2, 7%) and Geonoma cuneata (7, 24%); all were unoccupied. Foster and Timm (1976) reported tents in these palms, although they were not able to associate bats with the tents. My recent studies at La Selva have confirmed that these tents were made by A. watsoni. At Finca Las Cruces in mid-August 1982, 13 tents constructed by Artibeus watsoni were located; 198 HELDIANA: ZOOLOGY 2 '^ e o •a 00 Si eo| C * (/J c* . k» a T •2 o I a c g . u C f 3 ■-> P « « to x: CO ♦^ o S li -3 e o V TIMM: ARTIBEUS AND URODERMA 199 Fig. 6. Top, Dorsal view of the cyclanth, Carludovica palmata, showing the polygonal cuts; bottom, tent of Artibeus waisoni on C. palmata. 200 HELDIANA: ZOOLOGY Fig. 7. Photograph of adult female Artibeus watsoni and her subadult offspring roosting in a Carludovica leaf tent. Details of the polygonal cuts and folds can be seen in the background. one was occupied by an adult male. The tents were distributed on the following species of cyclanths: Asplundia euryspatha (6, 46%), Carludovica drudei (4, 31%), and Cyclanthus bipartitus (3, 23%). Artibeus watsoni uses a variety of species of plants and a wide array of leaf shapes for diurnal roosts. I have found four distinct styles of tents at a single locality (Corcovado). These styles include the sim- ple V-shaped cuts on bifurcated palms (fig. 5), cut- ting a few side veins on aroids to produce a round- ed pyramid, the elongated J-shaped cuts on banana and Heliconia leaves, and the polygonal cuts on Carludovica (figs. 6-7). For each distinct leaf shape, the cuts create a well-concealed diurnal roost. Ar- tibeus watsoni probably is an obligate tent-rooster, as it has only been found roosting under cut leaves. On several instances a bat occupied the same tent, or tents in close proximity, for two to three days in succession. Those tents might then remain unoccupied for several days in succession. Dis- turbed bats generally flew directly to another tent from 20 to 50 m away, or attempted to return to the tent where originally found. Tents generally are found clumped, both on a single plant if leaf morphology and age are appro- priate, and in restricted areas. Up to five tents have been found on a single Anthurium ravenii, and when present the mean number of tents was 2.5. At Parque Nacional de Corcovado, Choe and Timm (1985) found that Artibeus watsoni showed strong preference for Anthurium ravenii leaves that were medium sized, low within the plant, and grew closer to the ground than average A. ravenii leaves. Also at this site, Boinski and Timm (1985) doc- umented that squirrel monkeys {Saimiri oerstedi) were major predators on A. watsoni, with the adult male monkeys being the most successful at cap- turing bats. Additionally, double-toothed kites {Harpagus bidentatus) followed troops of foraging squirrel monkeys, using them as "beaters." When tent-making bats were flushed by the monkeys and escap)ed, they were routinely captured and con- sumed by the attending double-toothed kites. Artibeus watsoni has long been known to cut palm tents for diurnal roosts, although prior to this study little had been published on roosting TIMM: ARTIBEUS AND URODERMA 201 Fig. 8. Dorsal view of Scheelea rostrata showing the leaflets cut by Uroderma bilobatum to form a tent 202 HELDIANA: ZOOLOGY behavior of this species. The elongate J-shaped cuts made on bifurcated palms were first described and illustrated by Chapman (1932, p. 555). He stated that "both vanes of the leaf whence the bat flew were cut diagonally to the midrib of the leaf, so that their terminal portions drooped downward to form a tentlike shelter." Chapman appropri- ately termed these three-sided diurnal roosts "tents," and I have expanded the use of the word tents to include all modified leaves by bats. Barbour's (1932) original description of Uro- derma bilobatum cutting palm leaves for roosts also provides a secondhand report (p. 308) by H. C. Clark stating that "Clark has just found for the first time a youngish coconut palm, a single leaf of which was being cut by bats of the genus Uro- derma in a very similar way." The common use of young coconut palms {Cocos nucifera) by Ar- tibeus watsoni in Costa Rica, coupled with the total lack of evidence that Uroderma bilobatum uses juveniles of this palm, leaves that small, or roosts that close to the ground, suggests that the tents seen by Clark were in fact made by A. watsoni. Artibeus watsoni also is abundant on Barro Col- orado Island, Panama. Apparently no voucher specimens of the bats were preserved at the time. Allen (1939, p. 69) reported "a specimen of A. watsoni that was hanging by day from the under- side of a banana leaf" Perhaps the natural-looking folds caused by the cuts running parallel to the midrib were not noticed at the time. Ingles (1953) reported on tents of A. watsoni in two species of Geonoma on Barro Colorado Island; one tent was occupied by three individuals. Thomas's fruit-eat- ing bat has been found roosting in an artificial tent, an inverted hanging box. Wilson (1970) reported that several females raised young in the comer of a suspended box on Barro Colorado Island. Uroderma Uroderma bilobatum Peters, 1865 Uroderma bilobatum has been given the dis- tinctive "common" name of Peters's tent-making bat. Tents constructed by Uroderma bilobatum were seen at three separate localities in Costa Rica during the summers of 1982, 1984, and again in 1986, the first at Bosque Brrancia near Cuarto Cruces in the Pacific lowlands of northwestern Costa Rica, the second at Corcovado on the Osa Peninsula, and the third at La Selva in the Carib- bean lowlands. On 25 August 1982 a colony of Uroderma bi- lobatum was roosting under a modified frond of the palm Scheelea rostrata at Bosque Brrancia. The colony included an adult male with enlarged testes and four adult females. The Scheelea rostrata frond in which the colony of Uroderma bilobatum roosted was a mature leaf, approximately 6.5 m in length. The bats were hanging approximately 4.5 m off"the ground; most were clustered together, although a few were spread out over 50 cm of the frond. The cut leaflets started at about 3.5 m off" the ground and proceeded up the frond for the next 2.5 m (fig. 8). The general pattern of the cut leaflets was a tapering effect, with the cuts on the lowermost leaflets being farthest from the midrib. Leaflets along the proximal 2 m and the distal 50 cm were unmodified. Only the midrib of the leaflets was cut. Each leaflet had a distinctive V-shaped fold at its base where it was attached to the midrib. The bats were hanging from the leaflets rather than the midrib. From the dorsal aspect of the leaf, the proximal portion of the tent (cut leaflets) extended 50 cm further down on the right side than on the left to include 10 basal leaflets whose opposites on the left were unaltered. The basalmost cut leaflet was cut 34 cm from the midrib. Proceeding distally, the length of the unmodified basal portion of each leaflet decreases. The basalmost cut leaflet on the left was cut 19 cm from the midrib. The overall appearance of the tent was a sharp, convergent taper for the next meter. Following this section, there was a 7 5 -cm section in which the cuts were close to the midrib (within 3 cm). On the distal- most 30 cm of the tent, the leaflets were cut closer to the midrib on the left side than on the right. Similar tents, each housing a colony of Uroderma bilobatum, were found in a large stand of Scheelea rostrata at Corcovado in 1984, and William A. Haber (pers. comm.) informed me that he has seen similarly cut leaves in the same species of palm at Cahuita (09°44'N, 82°49'W) in the Caribbean low- lands of southeastern Costa Rica in 1 984. In June of 1982 and again in March of 1986, I found numerous banana leaves cut by Uroderma bilobatum just to the north of the field station at La Selva. The midrib on vertical leaves was cut to the extent that the distal portion of the leaf collapsed downward to form a two-sided tent (fig. 9). Severing the midrib on vertical leaves had the effect of folding the leaf back upon itself creating a tight, dark crevice at the fold where the bats roosted (fig. 1 0). In addition to severing the mid- rib, the bat cut a large V-shaped pattern running TIMM: ARTIBEUS AND URODERMA 203 Fig. 9. Left, Dorsal view of a banana leaf showing the cut midrib and the large V-shaped cuts through the side nerves; right, tent of Uroderma bilobatum made from a banana leaf. from the midrib to the base of the leaf. The side veins and interconnected tissues were partly sev- ered. However, because the leaf stood nearly ver- tical, these V-shaped cuts did not cause further folding of the leaf The only cut creating the tent was that of the midrib. In 1982 five tents in widely separated banana leaves were located (table 4). One was occupied by eight Uroderma bilobatum, which included one adult male with enlarged testes and seven females. In 1986 eight tents were observed in the same banana patch. On this occasion eight U. bilobatum 204 FIELDIANA: ZOOLOGY TIMM: ARTIBEUS AND URODERMA 205 tents were clustered in three clumps of bananas. Only one tent was occupied; it contained 1 3 bats. This folded broad-leaved style of tent is un- doubtedly the tent style illustrated by Walker ( 1 960, p. 30) in his photograph of roosting Uroderma bilobatum, although he did not describe it nor mention where it was observed. Interestingly, I have searched several dozen banana and larger Heliconia groves throughout Costa Rica and Ec- uador specifically looking for this style of tent, and none were observed. Uroderma bilobatum is an abundant and widespread species in the lowlands, but employment of this particular style of tent appears spotty as I have not observed it elsewhere. In Panama, Barbour (1932) found that colony size under a single cut leaf of Prichardia pacifica varied from a few bats to 56. Prior to Barbour's discovery that Uroderma altered leaves, Goldman (1920, p. 199) stated of these bats in Panama. In the forest near Gatun Uroderma biloba- tum was located several times, a few in a place, clinging during the day in clusters to the midribs on the under sides of large palm leaves. They usually choose darkened spots where the leaf was folded over, or over- hanging pinnae shut out much of the light. Burt and Stirton (1961) reported U. bilobatum in El Salvador roosting beneath the leaves of bananas and coconut palms. I strongly suspect that the bats had altered these leaves and that these authors had failed to notice it or failed to associate the hanging bats with the damage to the leaves. Bloedel (1955, p. 234) stated of Uroderma bilobatum in Panama that: I observed these bats only in their palm-leaf tents. ... In the latter part of March most females have nursing young, and are roost- ing in clusters of 20 to 40, while the males are separated from them, usually solitary or in small groups of from 2-5. In Trinidad, Goodwin and Greenhall (1961, p. 254) found them on the under side of the fan-shaped leaves of certain palm trees, especially the carat palm (Sabal glaucescens). . . . The bat makes a series of cuts across the pleated surface of a leaf, causing half of the leaf to bend at an angle to form a protected retreat. Table 4. Individual measurements (in cm or de- grees) collected from four tents constructed by Uroderma bilobatum on banana, Musa x paradisiaca, in Costa Rica. Blade Petiole Cut midrib Length Width angle from base Remarks 210 95 40 117 Not occupied 220 90 70 90 8 U. bilobatum 80 1 U. bilobatum 80 Not occupied 210 95 Not occupied They were found roosting in colonies of 1 0 or more individuals. In Suriname, Husson (1962, p. 161; 1978, p. 143) collected three pregnant female U. bilobatum "in a plantation where they were found hanging on the imder side of a leaf of [a] so-called 'paloeloe,' Ravenala guyariensis." In Nicaragua, Jones ( 1 964, p. 507) collected four female U. bi- lobatum that "hung together about 10 feet above the ground in the 'tent' formed by a cut palm frond. Each was pregnant with a single embryo." In Gua- temala, Dickerman et al. (1981, p. 409) reported, "Palm leaf tents were frequently found occupied by one to seven individuals, but nursing females were usually found alone or with juveniles." Leaves selected by Uroderma bilobatum for tents are all large, and of a variety of shapes. The large V-shaped pattern cut into the leaves is a charac- teristic of Uroderma tents. Artibeus watsoni also uses a variety of leaf shapes for tents; however, the nature of the cuts and style of tent created vary with leaf shape. Uroderma bilobatum, on the other hand, makes patterned cuts that appear to be an innate response to large leaves, regardless of the shape. As noted in the banana tents, a single cut across the midrib creates the tent, and the V-shaped pattern had no effect upon tent shape; this perhaps represents wasted effort by the bats. There are a few records of Uroderma bilobatum being found roosting in a hollow tree and one "un- der the eave of house" (Davis, 1968, p. 695). In all cases these have been of single individuals, and I suspect represent either recently dispersed young that have yet to join a breeding colony or bachelor males. Uroderma bilobatum roosting under a cut Prichardia leaf was illustrated by Kunz (1982). A colony of U. bilobatum roosting under a banana leaf tent was illustrated by Keller ( 1 986). Mac- donald's ( 1 984, p. 806) photograph of two tent- making bats under a Heliconia tent is erroneously labeled Uroderma bilobatum. These bats are ac- 206 HELDIANA: ZOOLOGY Fig. 11. Tent of Vroderma magnirostrum on the pinnately leafed palm Astrocaryum muru- muru. tually a small species of Artibeus, probably A. phaeotis; the tent style and size and coloration of the bats are typical oi A. phaeotis. Uroderma magnirostrum Davis, 1968 Although Vroderma magnirostrum is a widely distributed bat found from Mexico to Bolivia, it was not recognized as a species distinct from U. bilobatum until 1968 (Davis, 1968), and few spec- imens are represented in collections. When W. B. Davis described this new species he commented that although it was widely distributed only seven specimens had been collected prior to the wide- spread use of mist nets in the 1960s, and that all specimens available to him had come from local- ities less than 1000 feet in elevation. He stated that "These facts strongly suggest basic differences in the habits of the two species and that those bats with a deep rostrum are not 'tent-makers' as are members of the species Uroderma bilobatum" (Davis, 1968, p. 678). There have been no reports on the behavior or ecology of U. magnirostrum. On 25 July 1985 Barbara L. Clauson discovered a colony of two male and three female Uroderma TIMM: ARTIBEUS AND URODERMA 207 magnirostrum roosting under the cut leaflets of the pinnately leafed palm, Astrocaryum murumuru. The single occupied tent found was on the ridge above Hacienda Amazonia at 825 m in south- eastern Peru. The entire colony was collected by John W. Fitzpatrick. On 3 November Clauson re- turned to the site to measure ,»ie leaf and noted an additional cut leaf in the same plant. When first observed, four bats were hanging to- gether and one was hanging several centimeters away. When observed an hour later, all five were hanging together in a tight cluster approximately 7.5 m off the ground. The bats were hanging from the leaflets rather than the midrib, approximately 200 cm from the tip of the leaf The colony in- cluded one adult male with enlarged testes, two adult females, and two subadults, one female and one male. The roosting structure of Uroderma magniros- trum was in a pinnately leafed palm (fig. 1 1 ). The bats severed the leaflets along the upper two-thirds of the leaf; those along the lower third were un- altered, as were the leaflets at the very tip. As the leaflets proceeded up the tent they were severed closer to the midrib forming an elongate, conver- gently tapering tent (fig. 1 1). The general appear- ance of the U. magnirostrum tent is similar to that described herein for U. bilobatum on the pinnately leafed palm Scheelea rostrata. The Astrocaryum frond in which the colony of Uroderma magnirostrum roosted was a mature leaf, approximately 6.1 m in length and 1.9 m in width at the widest point, with the petiole 1.1m long. The leaf left the trunk (d.b.h. .4 m) at 3 m from the ground and hung at an angle of approximately 50°. The bats were hanging approximately 7.5 m off" the ground. The cut leaflets started at 1.5 m from the lowest leaflet and proceeded up the frond for the next 2.9 m to nearly the tip (fig. 1 1). The cuts on the lowermost leaflets were furthest from the midrib. The lowest severed leaflets were cut up to 34 cm from the midrib, whereas the distal leaflets were severed only 2 cm from the midrib of the leaf The midribs of the leaflets were cut causing the distal portion of the leaflets to fold downward. Leaflets along the proximal 1.5 m and the distal .5 m were unmodified. The trunk, pet- iole, and midrib of this palm were covered with sharp, penetrating spines several centimeters in length. After 14'/2 weeks this tent was still alive and green, most of the leaflets appearing as fresh in November as they did in July. This Astrocaryum contained a second cut leaf that was unoccupied. This roost was also in a ma- ture leaf which was an older leaf than the occupied tent, with many broken, yellowed, and brown leaf- lets. The leaf was approximately 6 m in length and 2 m in width at the widest point, with a petiole .8 m long. The leaf left the trunk at 2.9 m from the ground and hung at an angle of approximately 60°. Cuts were distributed asymmetrically along the length of the leaf The cut leaflets on the left side of the leaf started 3.05 m from the lowest leaflet and proceeded up the frond for the next 1 .3 m, to .88 m from the tip. The cut leaflets on the right side of the leaf started 3.25 m from the lowest leaflet and proceeded up the frond for the next 1.49 m, to .49 m from the tip. The cuts on the lower leaflets were farthest from the midrib. The lowest severed leaflets were cut up to 35 cm from the midrib, whereas the distal leaflets were severed as close as 1.5 cm from the midrib. Leaflets along the proximal 3.05 m and 3.25 m and the distal .88 m and .49 m were unmodified. I propose the common name of Davis's tent- making bat for this species. Conclusions A review of the literature on tent-making bats contains some 32 primary references covering the 55-year period from 1932 through 1986. Surpris- ingly, we actually know very little about the bi- ology of these bats. As late as 1975 Eisentraut was yet doubting th^t bats were cutting leaves to make tents, stating: . . . observers maintain that the bats form these tentlike structures themselves, by making a series of holes running across the middle of a large palm leaf The bats then supposedly bend the outer half of the leaf around, so they can then rest inside this 'tent'. . . . On the basis of personal obser- vations in tropical regions in Africa, I tend to believe instead that these holes were made by insect larvae while the leaves were still rolled up. A storm can then easily break the leaf along the line of holes and form the tent roof which is so convenient for the bats (Ei- sentraut, 1975, p. 142). Eisentraut, by his own admission, had never seen a bat tent. I believe that if he had, he would have come to the same conclusion Thomas Barbour did 208 HELDIANA: ZOOLOGY nearly a half century earlier, that the bats and not insects were making the cuts. Although we have yet to actually observe bats cutting leaves to form the roosting structures de- scribed herein, I hope the volume of data pre- sented here and in my other works establishes for a fact that many species of small and medium sized stenodermines are indeed tent-makers. The ob- servations presented represent data collected from several hundred tents located over a 1 5 -year pe- riod. Several facts consistently emerge between my observations and those independently corrobo- rated by others. Bats of the genera Artibeus and Uroderma (as well as Ectophylla, Mesophylla, and Vampyressa) roost under cut leaves. These leaves may be on a wide variety of species of plants, but generally the shape of the leaves is similar. The shape of the cuts is very characteristic for each species of bat and the patterns and styles of tents created by the bat species are consistent. The behavioral repertoire associated with tent- making in bats certainly evolved more than once, as evidenced by the patchy distribution of tent- making species with the chiropteran suborders Megachiroptera and Microchiroptera. Within the Megachiroptera, a single species of tent-maker is known, Cynopterus sphinx. Within the Microchi- roptera, tent-makers are known only from one subfamily of the Phyllostomidae, the Stenoder- minae. The Stenoderminae constitute an ex- tremely speciose and diverse group of bats, with more than 30 species currently recognized. Tent construction within stenodermines may be a trait that evolved once, twice, or as many as three times. The Artibeus- Uroderma group are sister genera and form one clade of the tent- making repertoire. Sec- ondly, the Mesophylla- Vampyressa group are sis- ter genera (and perhaps should be considered con- generic) and would constitute the second clade. Finally, Ectophylla would constitute a third lin- eage. The relationship between these three lineages is uncertain and warrants further investigation. Knowing that bats modify the leaves of several species of plants to produce diurnal roosting struc- tures led to the following questions: (1) Are bats selecting specific species of plants for tents? (2) What styles of tents are cut by bats and do these differ between species? (3) Do bats select for a particular angle, size, or shape of leaf for diurnal tents? (4) Are leaves selected preferably in larger clumps or smaller clumps? (5) Are older or youn- ger leaves selected? (6) Are leaves of a particular height class selected? (7) Are leaves that are not adjacent to solid objects selected? (8) What do typical tents look like? (9) How and why did tent construction evolve? On occasion I have found "cheaters," species of bats roosting in a tent made by another species. Is cheating an evolved strategy of roost site selec- tion of some bats? Bats of the genera Artibeus and Uroderma ac- tively modify leaves to produce diurnal roosting structures, but by biting the tissue between veins along the midrib and leaving the midrib and most veins intact, do not kill the leaves. The resulting tent is available for use as a roost for an extended period of time; one was observed in use for more than 60 days. Bats select for specific sizes and shapes of leaves. Tents provide concealment from predators and protection from the rain, wind, and sun. This type of roost offers the additional ad- vantage that the bats are warned about the ap- proach of a potential predator, because even slight movements of the leaf stem or the leaf itself are transmitted as magnified vibrations to the roosting bat. Tents may provide bats with suitable roosting sites that would not otherwise be available in close proximity to prime food resources. One of the most productive areas for future re- search will be exploring aspects of the biology of these bats from an evolutionary perspective. Fu- ture subjects I will be addressing include the role of tent roosting in controlling ectoparasites and the correlation between complexity of tents and social systems in these bats. I believe that tent- making originated as an antipredation strategy and has since, secondarily, evolved to play a major role in controlling ectoparasites and in social be- havior. Many factors influence the choice of roost site selection by bats. Included among these are vul- nerability to predation, physical stability of the site, proximity to food sources, and general ap- propriateness of the nest microenvironment for the rearing of young. It seems likely that tent con- struction requires considerable time and energy expenditure by bats, attesting to intense selection pressures involved. Acknowledgments I thank Eduardo Lopez Pizarro and El Depar- tamento de Vida Silvestre and Fernando Cortes and Servicio de Parques Nacionales of Costa Rica for making this study possible. The Organization for Tropical Studies (OTS), Rebecca Butterfield, TIMM: ARTIBEUS AND URODERMA 209 William A. Haber, Gary Hartshorn, Charles E. Schnell, and Joe M. Wunderle are gratefully ac- knowledged for assistance with logistics. Robert J. Izor assisted with field logistics in Costa Rica and Peru, and John W. Fitzpatrick assisted in Peru. In Ecuador I thank the Comandancia General del Ejercito Ecuatoriano, the Corporacion Estatal Pe- trolera Ecuatoriana, and the Ministerio de Agri- cultura y Ganaderia for making our studies there possible. Luis Albuja, Ramiro Barriga, Angelitos Garrett, Myriam Ibarra, Gustavo Orces, and Don- ald J. Stewart provided logistic assistance in Ec- uador. Kerry A. Barringer, William C. Burger, Thomas B. Croat, Robin B. Foster, Barry Ham- mel, and Timothy Plowman provided identifica- tions or confirmed identifications of the plants. Barbara L. Clauson, Alfred L. Gardner, Lawrence R. Heaney, Karl F. Koopman, Thomas H. Kunz, Bruce D. Patterson, and Timothy Plowman pro- vided valuable suggestions on earlier drafts of the manuscript. My wife, Barbara, provided superb assistance with all aspects of this project, including providing several of the photographs used and all data on Uroderma magnirostrum. Rosanne Mie- zio prepared the illustrations. Nina Cummings, Ron Testa, and Diane White expeditiously and cheerfully executed my photography requests. This project was funded in part by grants from the Rice Foundation of Chicago, the National Science Foundation [INT-8303 194], National Geographic Society, and Field Museum of Natural History. I especially thank Mr. and Mrs. Arthur A. Nolan, Jr. for their continuing support of my research. This paper is dedicated to Philip Hershkovitz in recognition of his contributions to Neotropical mammalogy, and most especially for the friend- ship he has shown me. Literature Cited Allen, G. M. 1939. Bats. Harvard University Press, Cambridge, Massachusetts, 368 pp. Barbour, T. 1932. A peculiar roosting habit of bats. Quarterly Review of Biology, 7: 307-3 1 2. Bloedel, p. 1955. Observations on the life histories of Panama bats. Journal of Mammalogy, 36: 232-235. BoiNSKJ, S., AND R. M. TiMM. 1985. Predation by squirrel monkeys and double-toothed kites on tent- making bats. American Journal of Prima tology, 9: 121- 127. Burt, W. H.. AND R. A. Stirton. 1961. The mammals of El Salvador. Miscellaneous Publications, Museum of Zoology, University of Michigan, 117: 1-69. Chapman, F. M. 1932. A home-making bat. Natural History, 32: 555-556. Choe, J. C, and R. M. Timm. 1985. Roosting site selection by Art ibeus watsoni {ChiToplera: Phyllostom- idae) on Anthurium ravenii (Araceae) in Costa Rica. Journal of Tropical Ecology, 1: 241-247. Davis, W. B. 1 944. Notes on Mexican mammals. Jour- nal of Mammalogy, 25: 370-403. . 1 968. Review of the genus Uroderma (Chirop- tera). Journal of Mammalogy, 49: 676-698. -. 1970. A review of the small fruit bats (genus Artibeus) of Middle America. Part II. Southwestern Naturalist, 14: 389-402. Davis, W. B.. D. C. Carter, and R. H. Pine. 1964. Noteworthy records of Mexican and Central Ameri- can bats. Journal of Mammalogy, 45: 375-387. Dickerman, R. W., K. F. Koopman, and C. Seymour. 1981. Notes on bats from the Pacific lowlands of Guatemala. Journal of Mammalogy, 62: 406—4 1 1 . Eisentraut, M. 1975. The insectivorous bats, pp. 1 1 1- 1 48. /n Grzimek, H. C. B., ed., Grzimek's Animal Life Encyclopedia. Vol. 1 1 , Mammals 11. Van Nostrand Reinhold Company, New York, 635 pp. Foster, M. S., and R. M. Timm. 1976. Tent-making by Artibeus jamaicensis (Chiroptera: Phyllostomati- dae) with comments on plants used by bats for tents. Biotropica, 8: 265-269. Goldman, E. A. 1920. Mammals of Panama. Smith- sonian Miscellaneous Collections, 69(5): 1-309. Goodwin, G. G. 1934. Mammals collected by A. W. Anthony in Guatemala, 1924-1928. Bulletin of the American Museum of Natural History, 68: 1-60. Goodwin, G. G.. and A. M. Greenhall. 1961. A review of the bats of Trinidad and Tobago: Descrip- tions, rabies infection, and ecology. Bulletin of the American Museum of Natural History, 122: 187-302. Goodwin, R. E. 1979. The bats of Timor Systematics and ecology. Bulletin of the American Museum of Natural History, 163: 73-122. Handlev, C. O., Jr. 1987. New species of mammals from northern South America: Fruit-eating bats, genus Artibeus Leach, pp. 163-172. In Patterson, B. D., and R. M. Timm, eds.. Studies in Neotropical mammal- ogy: Essays in honor of Philip Hershkovitz. Fieldiana: Zoology, U.S., 39: 1-506. Hartshorn, G. S. 1983. Plants, pp. 1 1 8-1 57. /« Jan- zen, D. H., ed., Costa Rican Natural History. Uni- versity of Chicago Press, Chicago, xi + 816 pp. Herwttz, S. R. 1981. Regeneration of selected tropical tree species in Corcovado National Park, Costa Rica. University of California Publications in Geography, 24: 1-109. HoLDRiDGE, L. R. 1967. Life Zone Ecology. Tropical Science Center, San Jose, Costa Rica, 206 pp. HoLDRiDGE, L. R.. W. C Grenke, W. H. Hatheway, T. Liang, and J. A. Tosi, Jr. 1971. Forest Environ- ments in Tropical Life Zones: A Pilot Study. Pergamon Press, New York, 747 pp. Husson, a. M. 1954. On Vampyrodes caracciolae (Thomas) and some other bats from the island of To- bago (British West Indies). Zoologische Mededelingen, 33: 63-67. 210 HELDIANA: ZOOLOGY . 1962. Thebatsof Suriname. Zoologische Ver- handelingen, No. 58: 1-282. 1 978. The mammals of Suriname. Zoologische monographieen van het Rijksmuseum van Natuur- lijke Historie, No. 2: 1-569. Ingles, L. G. 1953. Observations on Barro Colorado Island mammals. Journal of Mammalogy, 34: 266- 268. Janzen, D. H. 1971. The fate of Scheelea rostrata fruits beneath the parent tree: Predispersal attack by bru- chids. Principes, 15: 89-101. Jones, J. K., Jr. 1964. Bats new to the fauna of Nic- aragua. Transactions of the Kansas Academy of Sci- ences, 67: 506-508. . 1966. Bats from Guatemala. University of Kansas Publications, Museum of Natural History, 16: 439-472. Jones, J. K., Jr., AND T. Alvarez. 1964. Additional records of mammals from the Mexican state of San Luis Potosi. Journal of Mammalogy, 45: 302-303. Keller, W. E. 1 986. Phenomena, comment and notes. Smithsonian, 17: 28, 30-31. KoEPCKE, J. 1984. "Blattzelte" als Schlafplatze der Fledermaus Ectophylla macconnelli (Thomas, 1901) (Phyllostomidae) im tropischen Regenwald von Peru. Saugetierkundliche Mitteilungen, 31: 123-126. KoopMAN, K. F. 1978. Zoogeography of Peruvian bats with special emphasis on the role of the Andes. Amer- ican Museum Novitates, 2651: 1-33. KuNZ, T. H. 1982. Roosting ecology of bats, pp. 1-55. In Kunz, T. H., ed., Ecology of Bats. Plenum Press, New York, 425 pp. Kunz, T. H., P. V. August, and C. D. Burnett. 1983. Harem social organization in cave roosting Artibens jamaicensis (Chiroptera: Phyllostomidae). Biotropica, 15: 133-138. Macdonald, D. 1984. The Encyclopedia of Mam- mals. Facts on File Publications, New York, 895 pp. Morrison, D. W. 1 979. Apparent male defense of tree hollows in the fruit bat, Artibeus jamaicensis. Journal of Mammalogy, 60: 11-15. Ramirez-Pulido, J., A. Martinez, and G. Urbano. 1977. Mamiferos de la Costa Grande de Guerrero, Mexico. Anales del Instituto de Biologia, Universidad Nacional Autonoma de Mexico, Serie Zoologia, 48: 243-292. Sawyer, J. O., and A. A. Lindsey. 1971. Vegetation of the life zones in Costa Rica. Indiana Academy of Science Monograph, 2: 1-214. Slud, p. 1960. The birds of Finca "La Selva," Costa Rica: A tropical wet forest locality. Bulletin of the American Museum of Natural History, 121: 49-148. . 1980. The birds of Hacienda Palo Verde, Guanacaste, Costa Rica. Smithsonian Contributions to Zoology, 292: 1-92. TiMM, R. M. 1984. Tent construction by Vampyressa in Costa Rica. Journal of Mammalogy, 65: 166-167. . 1985. .4m"/?^M.s pAaeo/w. Mammalian Species, 235: 1-6. TiMM, R. M., AND J. Mortimer. 1976. Selection of roost sites by Honduran white bats, Ectophylla alba (Chiroptera: Phyllostomatidae). Ecology, 57: 385-389. TuTTLE, M. D. 1976. Collecting techniques, pp. 71- 88. In Baker, R. J., J. K. Jones, Jr., and D. C. Carter, eds.. Biology of Bats of the New World Family Phyl- lostomatidae. Part I. Special Publications, The Mu- seum, Texas Tech University, 10: 1-218. ViLLA-R., B. 1 967. Los murcielagos de Mexico. Anales del Instituto de Biologia, Universidad Nacional Au- tonoma de Mexico, xvi + 49 1 pp. Walker, E. P. 1960. Studying our Fellow Mammals. The Animal Welfare Institute, New York, 1 74 pp. Watkins, L. C, J. K. Jones, Jr., and H. H. Genoways. 1972. Bats of Jalisco, Mexico. Sjjecial Publications, The Museum, Texas Tech University, 1: 1-44. Wilson, D. E. 1970. An unusual roost o{ Artibeus ci- nereus watsoni. Journal of Mammalogy, 51: 204-205. Wilson, D. E., AND D. H. Janzen. 1972. Predation on Scheelea palm seeds by bruchid beetles: Seed density and distance from the parent palm. Ecology, 53: 954- 959. Appendix List of scientific names of plants mentioned in the text and used by Neotropical bats for tent con- struction. Anacardiaceae Anacardium excelsum (Bertero & Balbis) Skeels Araceae (cont'd.) Monstera lechleriana Schott Xanthosoma sp. Araceae Cyclanthaceae Anthurium caperatum Croat &. Baker Anthurium ravenii Croat & Baker Asplundia euryspatha Hari. Carludovica dmdei Masters TIMM: ARTIBEUS AND URODERMA 211 Cyclanthaceae (cont'd.) Carludovica palmata R. & P. Cyclanthus bipartitus Poit. Marantaceae Calathea insignis Petersen Musaceae Heliconia imbricata (Kuntze) Baker Heliconia latispatha Benth. Musa X paradisiaca L. Plenakospermum guyanense Endl. (syn. Ravenala guyanensis Petersen) Palmae Asterogyne martiana (H. Wendl.) H. Wendl. ex Hemsley Astrocaryum murumuru Mart. Bactris wendlandiana Burret Cocos nucifera L. Geonoma congesta H. Wendl. ex Spruce Geonoma cuneata H. Wendl. ex Spruce (syn. G. decurrens H. Wendl.) Geonoma oxycarpa Martius (syn. G. binervia Oerst.) Livistona chinensis (Jacq.) R. Br. ex. Mart. Prichardia pacifica Seem. & H. Wendl. Sabal mauritiiformis (Karsten) Griseb. &. H. Wendl. ex Griseb. (syn. S. glaucescens Lodd. ex H. E. Moore) Scheelea rostrata (Oersted) Burret Welfia georgii H. Wendl. ex Burret 212 FIELDIANA: ZOOLCX5Y Comparative Ultrastructure and Evolutionary Patterns of Acinar Secretory Product of Parotid Salivary Glands in Neotropical Bats Carleton J. Phillips, Toshikazu Nagato, and Bernard Tandler ABSTRACTS Secretory products produced by acinar cells of the parotid salivary glands of 1 5 species of Neotropical bats {Pteronotus parnellii, Phyllostomus elongatus, P. latifolius, Tonatia bidens, T. sylvicola, Trachops cirrhosus, Glossophaga soricina, Leptonycteris sanborni, Sturnira lilium, Artibeus jamaicensis, Ariteus flavescens, Eptesicus lynni, E. brasiliensis, Tadarida brasiliensis, and Molossus molossus) were compared by transmission electron microscopy. Extensive inter- and intrageneric differences were found in the ultrastructure of the mature acinar secretory granules. This variation in secretory cell product exceeded any previously reported intraordinal phenotypic variation at the cellular level, but was in keeping with previously reported bio- chemical data on salivary protein polymorphism in primates and rodents. Data from molecular biology and systematics lend support to the hypothesis that the microscopic variations are directly representative of genie differences among species. It also is postulated that intrageneric microscopic differences at least partly are due to neutral (nonfunctional) differences in molecular structure or charge (or both) rather than evolutionary selection. Among the phyllostomids, a general trend in parotid acinar cell product was found in Artibeus and Ariteus, in which a decrease in enzymatic content of the product could be correlated with ultrastructural differences. The secretory product in Artibeus and Ariteus also differed significantly from that of the genus Sturnira. and it is proposed that the phenotypic differences between Sturnira and the other two stenodermatines represent a major genetic difference of systematic importance. The ultrastruc- tural appearance and substructure of the parotid acinar secretory granules could not be con- sistently correlated with diet alone, although insectivorous-animalivorous SF)ecies have enzyme- rich, mostly electron-dense granules, whereas two fruit bats, Artibeus and Ariteus, have pale, enzyme-poor parotid granules. Productos de secrecion producidos por celulas acinares de las glandulas salivales de 1 5 esF>ecies de murcielagos neotTopicales (Pteronotus parnellii, Phyllostomus elongatus, P. latifolius, Tonatia bidens, T. sylvicola, Trachops cirrhosus, Glossophaga soricina, Leptonycteris sanborni, Sturnira lilium, Artibeus jamaicensis, Ariteus flavescens, Eptesicus lynni, E. brasiliensis, Tadarida bra- siliensis y Molossus molossus) fueron comparados mediante microscopio electronico de trans- mision. Extensivas diferencias inter- e intragenericas fueron encontradas en la estructura de granulos glandulares acinares maduros. Esta variacion en productos de celulas secretoras excedio From the Department of Biology, Hofstra University, Hempstead, NY 11550 (Phillips); and Department of Oral Biology, School of Dentistry, Case Western Reserve University, Cleveland, OH 44106 (Nagato and Tandler). Dr. Nagato's present address is Department of Oral and Maxillofacial Surgery, Ehime University School of Med- icine, Shizukawa, Shigenobu, Onsen-gun, Eshime 79 1 -02, Japan. PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 213 cualquier variacion fenotipica intraordinal previamente reportada a nivel celular, pero estuvo en armonia con datos bioquimicos previamente reporlados de polimorfismo en proteinas sa- livales de primates y roedores. Datos de biologia molecular y sistematica proveen apoyo a la hipotesis de que las variaciones microscopicas son direclamente representativas de diferencias geneticas entre especies. Es lambien postulado que diferencias microscopicas intragenericas, al menos parcialmente, son debidas a diferencias neulrales (no funcionales) en estructura molecular o carga (o ambas) mas bien que a seleccion evolutiva. Entre los filostomidos, una tendencia general en el producto parotido de las celulas acinares fue encontrado en Artibeus y Ariteus, en los cuales una disminucion en contenido enzimatico del producto p>odria estar relacionado con diferencias ultraestructurales. El producto secretado en Artibeus y Ariteus tambien se diferencio significativamente de aquel del genero Sturnira y es propuesto que las diferencias fenotipicas entre Sturnira y los otros dos stenodermatinos representan una mayor diferencia genetica de importancia sistematica. La apariencia ultraestructural y subestructura de los granulos secre- torios parotidos acinares podria no estar consistentemente correlacionada solo con la dieta, aunque especies insectivoras-animalivoras tienen granulos ricos en enzimas mayormente densos en electrones, mientras que los murcielagos frugivoros, Artibeus y Ariteus, tienen granulos parotidos palidos y pobres en enzimas. Produtos secretorios produzidos por acinos das glandulas parotidas de 15 especies de mor-' cegos neotropicos (Pteronotus parnellii, Phyllostomus elongatus, P. latifolius, Tonatia bidens, T. sylvicola, Trachops cirrhosus, Glossophaga soricina, Leptonycteris sanborni, Sturnira lilium. Artibeus jamaicensis, Ariteus flavescens, Eptesicus lynni, E. brasiliensis, Tadarida brasiliensis, e Molossus molossus) foram comparados atraves da microscopia de transmissao eletronica. Vastas diferengas intra- e intergenericas foram encontradas na ultra-estrutura dos granulos maduros dos acinos secretorios. Esta variafao no produto das celulas secretorias supera qualquer varia^ao fenotipica intraordinal previamente relatada para o nivel celular, porem concorda com as rela96es publicadas sobre dados bioquimicos do polimorfismo de proteinas salivares em primatas e em roedores. Dados sistematicos, e de biologia molecular, apoiam a hipotese que variances microscopicas sao diretamente representativas das diferen^as geneticas entre especies. Propoese tambem, que estas diferen^as microscopicas intragenericas sao ao menos parcialmente causadas por diferen9as neutras (i.e., naofuncionais) nas estruturas moleculares ou nas suas cargas eletricas (ou ambas), ao inves de serem consequencias da sele^ao evolutiva. Entre os morcegos da familia Phillostomidae, foi encontrado em Artibeus e em Ariteus um padrao geral nos produtos dos acinos parotideos, onde uma redu9ao do conteudo enzimatico' e correlacionado a diferencas nas ultraestruturas dos granulos produzidos. Os produtos secre- torios em Artibeus e em Ariteus sao significantemente diferentes dos produtos do genero Sturnira. e prop6e-se que as diferen9as fenotipicas entre Sturnira e os outros dois stenodermatinos representam uma grande diferen9a genetica, de importancia sistematica. A aparencia ultraes- trutural, e a subestrutura dos granulos secretorios dos acinos parotideos, nao se correlacionam consistentemente com a dieta por si, apesar de que as especies insetivoras-animalivoras possuem granulos ricos em enzimas e densos em eletrons, equanto que dois morcegos frugivoros, Artibeus e Ariteus, (wssuem granulos parotideos palidos e com poucas enzimas. Introduction Neotropical bats are extremely diversified; ex- isting species represent perhaps the most outstand- ing mammalian example of ecomorphological ad- aptation. Dentitions, jaw morphology, brains, kidneys, tongues, and digestive tracts are only a few examples among the anatomical features that have been investigated in recent years (e.g., Phil- lips, 1971; Forman, 1972; Phillips et al., 1977, 1984; Freeman, 1979, 1981; Eisenberg «t Wilson, 1978; Studier et al., 1983). The major salivary glands are yet another system that has attracted attention, primarily because histological, ultra- structural, and histochemical investigations have consistently revealed striking interspecific differ- ences and unusual histological and secretory fea- tures (Wimsatt, 1956; DiSanto, 1960; Junqueira & Fava de Moraes, 1965; Junqueira et al., 1967, 1973; Phillips et al., 1977; Mineda, 1977, 1978; 214 HELDIANA: ZOOLOGY Pinkstaff et al., 1982; Tandler & Cohan, 1984; Nagatoetal., 1 984; Tandler & Phillips, 1985; Phil- lips & Tandler, 1985, 1987; Tandler et al., in press). Mammalian salivary glands are highly complex organs that not only secrete digestive enzymes but also can secrete hormones, antibodies, and toxins, to name but a few known products (e.g., Tandler, 1972; Hand, 1980b). Data are available on the biochemistry of saliva in humans and common laboratory species; but very little is known about the specific biochemistry of the parotid acinar se- cretory granules except for laboratory rats in which some of the proteins have been characterized (Ro- binovitch & Sreebny, 1969; Ball, 1974; Wallach et al., 1975). The complex structure and function of salivary glands is underscored by data from studies of Neotropical bats, which recently have been analyzed by both transmission electron mi- croscopy and histochemistry. For example, the ac- cessory submandibular gland of Trachops cirrho- sus has been shown to differ histologically from any known mammalian salivary gland, with the exception of the same gland in Megaderma lyra and M. spasma. Megaderma lyra is an Old World ecological equivalent of Trachops (Phillips &. Tandler, 1985, 1987; Phillips et al., 1987). Both of these unrelated species feed on frogs (Lekagul & McNeely, 1977; Tuttle & Ryan, 1981), which possibly has been a factor in the convergent evo- lution of their submandibular glands. A previously unknown cellular organelle has been described in another Neotropical bat, Tonatia sylvicola (Na- gato et al., 1984). In this species, the presence of xmique crystalloid smooth endoplasmic reticulum in seromucous acinar cells is sex-linked (being found only in submandibular acinar cells in males); a steroid product produced by this organelle pos- sibly serves as a species-isolating mechanism or as part of a chemo-behavioral system, or both. Lastly, a comparative investigation of the secre- tory product in seromucous acinar cells in sub- mandibular glands of five species ofArtibeus has revealed that the ultrastructural characteristics of secretory products can have systematic implica- tions (Tandler et al., 1983, 1986). This study was of particular interest because the salivary gland data matched genie data independently derived by Koop and Baker (1983). For the present investigation we surveyed par- otid acinar cell secretory products in a selected group of 1 5 species of Neotropical bats. Our group comprised four families (Mormoopidae, Phyllo- stomidae, Vespertillionidae, and Molossidae) and included a group of species in which dietary habits ranged from insectivory and animalivory to om- nivory and frugivory. This investigation is the first comprehensive interspecific survey of secretory product ultrastructure, and addresses the follow- ing questions: ( 1 ) what is the range of variation in secretory products; (2) what are the evolutionary patterns in parotid secretory product; and (3) what systematic conclusions can be reached by com- parative ultrastructural analysis? Materials and Methods Numbers and sex of specimens used in the pres- ent study are given in the Appendix. All of these bats were collected during fieldwork in Mexico, Jamaica, and Suriname. Voucher specimens for all species and collecting localities are deposited in the mammal collections of either the Carnegie Museum of Natural History or The Museum, Tex- as Tech University. Bats typically were collected at night with mist nets, and were kept overnight without food until they were killed between 0900 and 1 200 the following morning. The animals were anesthetized with 0.25 ml of sodium pentobarbital (50 mg/ml, intraperitoneally) and the salivary glands removed, placed on dental wax, flooded with freshly mixed fixative, and diced into pieces measuring approximately 1 mm'. Two fixation protocols were used at different times during the project. Specimens collected in Mexico and Jamaica were fixed in 2% glutaral- dehyde in 0. 1 M phosphate buffer and then stored unrefrigerated in fresh fixative. The specimens col- lected in Suriname were fixed in a modified tri- aldehyde-dimethylsulfoxide (dmso) mixture, first described by Kalt and Tandler (1971) and later modified slightly for fieldwork (Phillips, 1 985). The trialdehyde fixative, consisting of 3% glutaralde- hyde, 1% paraformaldehyde, 0.5% acrolein, 2.5% DMSO, and 1 mM CaClj in a 0.05 M cacodylate buffer and sucrose at pH 7.2, proved superior to the simple glutaraldehyde fixative in that (1) a higher percentage of tissues proved to be ade- quately fixed for study, and (2) the mitochondria tended to remain intact instead of being disrupted. Although the composition of the fixatives was dif- ferent, we have not found any evidence that these differences introduced undesirable artifacts that would influence our analysis. Additional details about the techniques can be found in both Phillips (1985) and Nagato et al. (1984). To remove unbound aldehydes, the tissue blocks were subjected to prolonged washing in phos- PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 215 acinar lumen. acinar cell secretory graauLej striated Lntercalated acinus duct duct Fig. I . Schematic diagram of a secretory unit in a typical parotid gland. A cluster of secretory cells, arranged around a lumen to form an acinus, drains into an intercalated duct that in turn empties into a striated duct. In reality, more than one acinus may be associated with an intercalated duct, which itself may be branched. The acinar cells elaborate the primary saliva, which has plasma-like concentrations of electrolytes, and add their organic secrelory products to it. The precise role of the intercalated duct cells has not been established, but they probably play a role in secretion and transport. The principal function of the striated ducts is resorption of electrolytes, mainly sodium, from the primary saliva, rendering it hypotonic. phate-buffered sucrose. The blocks were postfixed for two hours in phosphate-buffered 2% OSO4, rinsed in distilled water, soaked overnight in cold aqueous 0.25 uranyl acetate, rinsed again in dis- tilled water, dehydrated in ascending concentra- tions of ethanol, passed through propylene oxide, and embedded in Epon-Maraglas (Tandler & Wal- ter, 1977). Thin sections were doubly stained with methanolic uranyl acetate (Stempak & Ward, 1 964) and lead citrate (Venable & Coggeshall, 1965). All sections were examined in a Siemens Elmiskop la transmission electron microscope (tem). Semithin sections ( 1 nm) used for orientation were stained with toluidine blue (Bjorkman, 1962) and exam- ined in a Zeiss Ultraphot. Tissue samples used for comparisons were se- lected carefully from among available tissue blocks. We generally avoided edges of tissues where me- chanical trauma often affects not only the cellular ultrastructure but also the microscopic appearance of the secretory granules. In selecting representa- tive "mature" granules we took into account the full array of inter- and intracellular variability as well as the often complex substructural geometry of the granules. Selection of the most representa- tive granules admittedly was subjective, but based on our experience, four principal criteria were applied: (1) cells selected for analysis showed no signs of shrinkage or swelling, and sensitive or- ganelles, such as mitochondria, were not distorted; (2) the appearance of the granule was not altered in any meaningful way by variations in fixation; (3) the appearance of the secretory granules had to be unrelated to their location within the spec- imen block; and (4) the development of the gran- ules could be traced (from Golgi complex to apical cytoplasm) without any major breaks in devel- opmental sequence. Results ' The basic histology of the parotid salivary gland was similar in all 15 sp)ecies examined; in each species the acinus was formed by a cluster of cells connected to the striated duct by intercalated ducts of varying lengths (fig. 1). At the tem level, acinar cell secretory products were found to differ in all species examined (figs. 2-5). These differences could not be related to granule ontogeny by com- paring "immature" Golgi-GERL (Golgi-endoplas- mic reticulum-lysosomes)-associated granules with "mature" granules in the apical cytoplasm. In our samples that consisted of both males and females, we found no evidence of sexual dimorphism in secretory granule substructure, and in our largest samples (5 to 1 0 specimens) we found no examples of individual or geographic variation that could not be attributed to typical inter- or intracellular variation or to granule geometry. Although vari- 216 FIELDIANA: ZOOLOGY PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 217 53 F •8 c u <3 55 tn "2 g § t3 4J 5s es ? c O s* ! J3 a S 2 i/i o o ■ 1 u d V u ^ U c •i 00 w c o u ■^ o CA CO r<^ _3 s r-' **^ u "3 c X to C 4J C CQ i/i (^ & O U ^ 1 k Q O ■S. o 1 S., 5 is £ T3 ^ 3 C o C8 CO o s s s X 3 ^ to c c §5 1 1 tw 8 a o c •->, V U 1 > ■•5 t 2 s ^ 3 c o Y-. '^ o \j ss ^ 5 ■"O vo' ■jj u eripheral light material had a layered structure (fig. 4, top). Eptesicus brasiliensis— As in E. lynni. the lim- iting membrane was subtended by layered mate- rial, usually in several plies, that extended into the dense matrix. These extensions were less tortuous than those in E. lynni and appeared to subdivide the granule interior (fig. 4, bottom). Tadarida brasiliensis— The secretory granules in this species exhibited a spectrum of patterns. The most common was relatively simple, with either short, flat, dense prisms or dense dots being suspended in a slightly less dense matrix. A few granules contained dense hollow spheres, whereas others had a mazelike configuration based on light and dark laminations (fig. 5, top). PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 219 Fig. 4. Higher magnification of mature secretory granules in parotid acinar cells in Eptesicus lynni (top) and E. hrasiliensis (bottom). Note the subtle but consistent differences in the granule substructure. Eptesicus lynni, x 63,000; E. brasiliensis, x 76,800. 220 HELDIANA: ZOOLOGY Fig. 5. Parotid acinar cell granules in Tadarida brasiliensis (top) are compared to those found in Molossus molossus (bottom). Note the range of variation in granule substructure in T. brasiliensis. Tadarida brasiliensis, x 26,000; Molossus molossus, x 24,000. PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 221 MoLOSSUs MOLOSSUS— The variable morphol- ogy of the granules in this species appeared to depend on their stage of maturation. Early gran- ules were small, with a moderately dense matrix in which were some prominent dense particles. As the granules matured (based on size, density, and spatial relationship to the Golgi complex), they enlarged and the particles decreased somewhat in density. The mature granules had a farinaceous matrix probably resulting from comminution of the dense particles (fig. 5, bottom). Discussion The idea of using comparative ultrastructural analysis of homologous, morphologically differ- entiated cells to study evolutionary pathways and to explore systematic relationships is new to the study of mammalian orders, but has been used in broader studies of vertebrates and invertebrates (Eakin, 1968; Rieger & Tyler, 1979; PhiUips & Tandler, 1987). Nevertheless, the potential value of ultrastructural comparisons has been demon- strated recently by studies of gastric mucosa, ret- ina, and submandibular salivary glands (Phillips et al., 1984; Feldman & Phillips, 1984; Tandler et al., 1983, 1986; Phillips & Tandler, 1987). The discovery of patterns of intersF>ecific differences in cellular architecture and cellular secretory prod- ucts is in keeping with previously successful use of histology in evolutionary and systematic mam- malogy (e.g.. Quay, 1954; Forman, 1972; Phillips & Oxberry, 1972; Sands et al., 1977; Naumova, 1981; Hood & Smith, 1982, 1983). Secretory cells, such as the parotid acinar cells used in the present study, seem to hold special promise for comparative investigation. Firstly, the entire secretory process— from nuclear dna to syn- thesis of proteins and complex carbohydrates and packaging of materials into secretory granules and their subsequent discharge— has been studied in- tensively over the past several decades (e.g., Jamieson & Palade, 1971; Castle et al., 1975; Pa- lade, 1975). Secondly, the secretory process seems to be relatively conservative— in the sense that basic pathways are the same in virtually all secre- tory cells— and, therefore, interspecific compari- sons are facilitated and interpretation is somewhat simplified (Phillips & Tandler, 1987). In the present study we limited our descriptive comparisons to "mature" secretory granules. These are the secretory granules that accumulate within the apical cytoplasm of the acinar cell; this product either is discharged from the cell into the acinar lumen (thus becoming part of the formative saliva) or, after an unknown storage interval, is broken down through autophagy and recycled within the cell. The glycoprotein components of the acinar secretory product are elaborated by the rough en- doplasmic reticulum acting in concert with the Golgi complex and are perhaps the most interest- ing feature from a comparative point of view. Syn- thesis of these glycoproteins begins with transcrip- tion of a very small segment of the genome into mRNA. The exportable proteins of the secretory granules could be regarded as providing a "win- dow" on the genome because, although all somatic cells have the same genome (Briggs & King, 1 952, 1957), only a small portion actually is operational in any given fully differentiated cell. A protein synthesized for export thus directly reflects a por- tion of the operational genome. Any polysaccha- ride components of the secretory granules prob- ably are at least one more step removed from the genome because synthesis of complex carbohy- drates is enzymatically determined and usually takes place within saccules in the Golgi complex (Tandler, 1978). Even so, the polysaccharides also have considerable potential for comparative anal- ysis. Production of secretory granules can be consid- ered in terms of ontogeny. The first "granules" typically are seen in direct proximity to the Golgi complex, where carbohydrates are linked to pro- teins and the membrane that will encase the gran- ule is synthesized in a process involving both the GERL and the Golgi complex itself (Tandler, 1978; Hand, 1 980a; Hand & Oliver, 1 984). Newly formed ("immature") granules differ greatly from "ma- ture" granules (see Castle et al., 1975) and were not used by us in describing the product for each of our species. Nevertheless, in view of species differences found by us, it is interesting to ask whether or not immature granules of one species might resemble mature granules in another species. Such similarities might be expected if hetero- chronic differences accounted for differences in mature product. However, because no such cross- species similarities were found by us between immature and mature secretory granules, hetero- chrony does not seem to account for sp>ecies dif- ferences, at least among the 1 5 bats examined here. No two Neotropical sfiecies examined here ex- hibited the same mature secretory granules in their parotid gland acinar cells, although in a few cases there was a degree of resemblance. This extreme variability is easily the greatest ever reported for 222 HELDIANA: ZOOLOGY homologous cells within an order or, as in the case of the Phyllostomidae, within a family of mam- mals (or any other vertebrates). To what can we attribute this striking finding? One consideration is fixation, which certainly affects the appearance of any cellular feature as viewed with transmission electron microscopy. In- deed, it can be said that the appearance of cells and their products essentially is the consequence of their intrinsic chemistry combined with the chemistry of the fixative at the moment in time when fixation occurred. Different fixatives and tis- sue processing can have profound effects on the microscopic appearance of secretory granules in salivary glands (Simson et al., 1978). However, we used consistent processing techniques and two similar fixatives. Our specimens of Ariteus and Leptonycteris both were fixed with 2% glutaral- dehyde, whereas all of our other specimens were fixed with a trialdehyde-DMSO fixative (Kalt & Tandler, 1971; Phillips, 1985). We were able to eliminate fixation as a source of variation because we also have examined specimens of Artibeus that had been fixed in both solutions, coincidentally with Leptonycteris, Ariteus, and all of the other species examined (see A. phaeotis parotid in Phil- lips et al., 1977). Generally speaking, given consistent prepara- tion techniques, microscopic differences in secre- tory granule substructure can be ascribed to bio- chemical differences among the granules. Microscopically detectable sequestration of indi- vidual types of macromolecules within secretory granules has been demonstrated only rarely (Ra- vazzola & Orci, 1980; Kousvelari et al., 1982) but clearly is the best available explanation of intra- granule substructure. Based on fundamental prin- ciples of biochemistry, it thus can be concluded that macromolecules packaged in the secretory granules most likely sort themselves out according to charge and steric effects, as well as chemical interactions, to yield a characteristic pattern for each species. However, it also should be noted that a homogeneous appearance of intragranular sub- stance does not in itself preclude sequestration of different enzymes within the granule. Separate lo- calization of different enzymes (a-amylase and chymotrypsinogen A or a-amylase and trypsino- gen) within pancreatic cell zymogen granules has been demonstrated with a combination of hrp (horseradish-p)eroxidase)-labeled and ferritin-la- beled antibodies (Ono et al., 1980). At present it is impossible to correlate exactly secretory granule substructure with particular chemical components such as certain enzymes or mucosubstances, so from microscopic images alone we cannot say precisely how the granules in our species differ chemically from one another. How- ever, some conclusions can be inferred from the literature. For example, an electron-dense image (see, for example, fig. 2), labeled classically as "se- rous," can be associated with granules rich in en- zymes. Such an image is typical in species such as laboratory rodents and primates, for which some data are available on the biochemistry of parotid saliva (e.g., Jacobsen «Sc Hensten-Pettersen, 1974). On the other hand, the presence of electron-dense "serous" granules does not preclude the presence of mucosubstances within the secretory granules. Pinkstaff'et al. (1982) reported that, although the parotid product in the little brown bat, Myotis lucifugus, was "serous" with standard histological techniques, both neutral and acidic mucosub- stances could be demonstrated histochemically. The parotid granules in Artibeus and Ariteus are interesting in this regard because electron-dense material is scarce (especially in Artibeus, fig. 3) and enzyme production is extremely low, at least in Artibeus (Junqueira et al., 1973). By way of con- trast, Sturnira lilium has largely electron-dense granules (fig. 3) and thus differs considerably from the other two stenodermatines; Sturnira produces saliva rich in enzymes (Junqueira et al., 1973). Although the parotid granules in Artibeus (and, by extension, Ariteus) are low in enzyme content, to what can we attribute their tem image? This is an intriguing question because Wimsatt (1956) re- ported that the gland was negative for mucosub- stances, whereas Radtke (1972) reported the pres- ence of sialomucins in parotid acinar cells. In part, this apparent disagreement is the result of differ- ences in techniques that cannot be resolved by transmission electron microscopy. What can be determined about the parotid sal- ivary glands in Neotropical bats that relates to their diets, evolutionary history, or systematic re- lationships? Clearly the ultrastructure of parotid secretory granules has systematic significance, be- cause no two genera (or species either) are exactly alike. This finding is in keeping with our studies of the submandibular gland in five species of Ar- tibeus (Phillips et al., 1977; Tandler et al., 1983, 1986). The seromucous cells in this gland were found to contain granules that allowed for three groupings of Artibeus species {A. cinereus-A. phaeotis, A. jamaicensis-A. lituratus, and A. con- color) that matched the independently derived genie data (from isozyme analysis) reported by PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 223 Koop and Baker ( 1 983). The differences in parotid acinar product in Eptesicus lynni and E. brasi- liensis (fig. 4) also are interesting from this per- spective because independent genie data from 1 9 presumptive loci show that E. lynni probably orig- inated from the E. fuscus species complex, sepa- rate from the origin of E. brasiliensis (Arnold et al., 1980). Eptesicus lynni was found to share only 62% of analyzed alleles with E. brasiliensis. It might seem surprising that the microscopic images of salivary gland secretory products are generically (and often specifically) distinctive, but variation of such a fine resolution actually is in keeping with numerous genetic studies of mam- malian saliva. Several genetic markers have been found in human saliva (Azen & Opp)enheim, 1 973; Ashton & Balakrishinan, 1974; Tan & Ashton, 1976), and salivary proteins in particular tend to be polymorphic (Azen, 1972, 1973). Additionally, both sex and strain differences in salivary proteins have been reported in laboratory mice (Ikemoto & Matsushima, 1984). Although none of these dif- ferences has been demonstrated microscopically (all are based on biochemical analysis alone), their occurrence nevertheless is significant to compar- isons among mammalian species. Although the sometimes subtle but consistent microscopic differences between species within the genera Phyllostomus, Tonatia, and Eptesicus (figs. 2, 4) are in keeping with our hypothesis of the sensitivity of secretory product to genie differ- ences, they cannot readily be related to any known ecological differences between species within each genus. It is altogether possible that such species differences do not represent direct evolutionary selection. Instead, the differing images might rep- resent species-specific protein polymorphisms of a type that would not significantly affect "perfor- mance" of the saliva even though differences in primary molecular structure or surface charges (or both) could be indirectly detected by transmission electron microscopy (Phillips & Tandler, 1987). The existence of such "nonfunctional" (and pre- sumably nonselected, i.e., "neutral") interspecific variation in a protein molecule has been demon- strated previously in the otherwise conservative hemoglobin molecule (Perutz, 1983). In this ex- ample the tertiary and quaternary structures ap- parently are conserved regardless of large numbers of functionally neutral amino acid substitutions in the primary structure. Although these substitu- tions in themselves most often have no effect on the functional capacity of the molecule, they nevertheless are known sometimes to change mo- lecular surface charges (Perutz, 1983). For the present we postulate that similar nonselected, non- functional variations could account for the micro- scopically detectable intrageneric differences in salivary proteins in Tonatia bidens and T. sylvi- cola, Phyllostomus latifolius and P. elongatus, and Eptesicus lynni and E. brasiliensis. This explana- tion seems more parsimonious than the alterna- tive, which would be to assume that the acinar cell component of the parotid saliva is "functionally" different in closely related, ecologically similar species. The extent or degree of microscopic dif- ferences in the mature acinar cell product ulti- mately might tell us more about relative times of divergence than about ecological differences among congeneric species of bats. Microscopic comparisons of salivary gland se- cretory products could be valuable to cladistic studies of chiropteran families, at least to judge from our data. The phyllostomines generally are regarded as the most primitive (least derived) of the Phyllostomidae (Smith, 1976), and their par- otid acinar products thus would qualify as plesio- morphic (following Henning, 1 966). The products in all five species examined (Phyllostomus elon- gatus, P. latifolius, Tonatia bidens, T. sylvicola, and Trachops cirrhosus) contained large amounts of electron-dense material (fig. 2), as did the se- cretory product in Pteronotus parnellii (a closely related mormoopid), Eptesicus lynni and E. bra- siliensis (Vespertilionidae), and Tadarida brasi- liensis and Molossus molossus (Molossidae), all of which serve as "outgroups." If the electron-dense, enzyme-rich secretory granules are regarded as plesiomorphic in micro- chiropteran bats, then it would be reasonable to regard a granule with less electron-dense material as apomorphic (derived). The glossophagine gen- era examined (Glossophaga and Leptonycteris) are representative of a phyllostomid evolutionary trend in which dentition, tongues, and associated mus- culature were modified for feeding on fruit, pollen, and nectar (Park & Hall, 1951; Phillips, 1971; Greenbaum & Phillips, 1974; Griffiths, 1982, 1983; Smith 8l Hood, 1 984). In these genera the electron- dense component has been reduced in comparison to the phyllostomines. Carollia perspicillata, which is omnivorous, fits into this category in that the parotid secretory granules contain relatively little electron-dense material. In this bat the secretory granules are very distinctive because the electron- dense inclusions are often in the form of cagelike geodesic structures (Phillips & Tandler, 1987; Tandler et al., in press). 224 FIELDIANA: ZOOLOGY Perhaps the most interesting systematic finding in our data lies within the nominal subfamily Stenodermatinae. These bats represent an evolu- tionary trend toward frugivory that includes ex- treme gastric adaptation at the gross, histological, histochemical, and cellular levels (Forman, 1972; Forman et al., 1979; Phillips & Studholme, 1982; Phillips et al., 1984). The pale parotid secretory granules in Artibeus and Ariteus are synapomor- phous, whereas the parotid granules in Sturnira are electron-dense, enzyme-rich, and more nearly like the plesiomorphic granules of the phyllosto- mines, in particular Trachops cirrhosus (figs. 2-3). This example is interesting because Sturnira al- ways has been something of an enigma. Although de la Torre (1961) allied this genus with Vampy- rops-\\\it stenodermatines and Smith (1976) in- cluded it with the "long-faced" stenodermatines, others (e.g., Walton & Walton, 1968) previously had placed the genus in a separate subfamily (Stur- nirinae) based on a variety of morphological fea- tures that seemed inconsistent with the other, more traditional, stenodermatines such as Artibeus. In- deed, Slaughter (1970) pointed out that Sturnira has some distinctive dental features that possibly link the genus to the glossophagines. The differ- ences in the parotid secretory granules thus are in keeping with a variety of other phenotypic differ- ences. While it is reasonable to suggest that secre- tory granule differences of this magnitude repre- sent a major interspecific difference in the operational segment of the genomes of homolo- gous parotid acinar cells, the eventual systematic value of such data will await availability of data about still other phyllostomid genera. Although our comparative data clearly docu- ment great microscopically detectable plasticity in the secretory product of parotid acinar cells, many questions about their evolution remain unan- swered. For example, what has been the role of diet? At first glance, our data suggest that insec- tivorous and animalivorous species have enzyme- rich, electron-dense granules, whereas frugivores have enzyme-poor, pale secretory granules. Yet, what about Sturnira lilium, which certainly in- cludes large amounts of fruit in its diet (Gardner, 1977)? At least three factors interfere with any effort to correlate parotid acinar cell product with diet. First, acinar cells are but one cell type among several that influence the biochemistry of parotid saliva. Second, and possibly more important, the par- otid is but one salivary gland among a host of glands (the submandibular and sublingual and mi- nor glands) that are located throughout the oral region and that contribute substances to the saliva. Some data suggest that different glands in different species might have been more responsive, in an evolutionary sense, to changes in diet. For ex- ample, it is the accessory submandibular glands of Trachops and Megaderma that are unique in histology and might correlate with feeding on frogs (Phillips &. Tandler, 1985; Phillips et al., 1987); the parotid in Trachops is similar to that of other phyllostomines. A third factor, for which we presently have no data, is the possibility that parotid acinar cells can respond in some way to diet at the individual level. Our sample sizes are large enough to convince us of the near uniformity of secretory granule mor- phology within a population. However, Schick et al. (1984) have recently published the first report of a secretory cell responding to dietary intake by a shift in enzyme production. This is the first such detailed molecular data known to us and, although pancreatic acinar cells in laboratory rats were the source of the data, the implications for compar- ative studies of salivary glands are worth noting. A final factor for consideration is the complex role that salivary glands play in the lives of mam- mals. Salivary glands in Artibeus, Chiroderma, and Ametrida might contribute to gastric cytoprotec- tion (Studier et al., 1983; Phillips et al., 1984). In Tonatia sylvicola an unusual organelle found in submandibular seromucous cells of males might be related to species isolation or sex recognition, or both (Nagato et al., 1984). If these examples are typical for bats, then digestion is but one of several major functions of salivary glands. In conclusion, this first systematic microscopic survey of a secretory cell product has demonstrat- ed a previously unknown, extraordinary degree of variation within a group of related sjjecies. Based on our findings, one might conclude that Neo- tropical microchiropteran bats will serve as a sig- nificant model for study of how secretory cells have evolved in mammals; meanwhile, data from comparative investigations will contribute to our knowledge of genie relationships among these an- imals. Acknowledgments Financial support for field and laboratory re- search that led to the data reported here came from a variety of sources, which we are pleased to ac- knowledge: Research Corporation Grant C-1251 PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 225 (to Phillips); NIH Grant DE 03455-02 (to Phil- lips); NSF Grant CDP-80 1-8653 (to Phillips); Hofstra University HCLAS Executive Committee grants (to Phillips); NIH Grant AM-08305 (to Tandler); NIH Grant RO DE 07648-01 Al (to Tandler & Phillips); the Alcoa Foundation (through Dr. Hugh H. Genoways, Carnegie Museum of Nat- ural History); and the Graham Netting Research Fund via a grant from the Cordelia S. May Char- itable Trust (through Genoways). In addition to recognizing our financial support, we also wish to thank several colleagues whose field companion- ship in such places as Suriname and Jamaica, hard work, and willingness to share time and ideas in their laboratories played a major part in our study. These are: Hugh H. Genoways, Carnegie Museum of Natural History; Gary W. Grimes and Dorothy E. Pumo, Hofstra University; Robert J. Baker, Texas Tech University; and Henry A. Reichart, formerly of STINASU, in Suriname. Field assis- tance was provided by K. M. Studholme, S. L. Williams, N. M. Sposito, J. Groen, R. L. Honey- cutt, B. Koop, M. Arnold, B. A. Oxberry, P. Bil- leter, J. Bickham, and J. Patton. We also acknowl- edge technical assistance of Carol Ayala, Case Western Reserve University, and typing by Linda Cossen, Special Secretarial Services, Hofstra Uni- versity. Artwork and lettering were done by Helen Tandler. All specimens were collected and im- ported under proper permits issued by the relevant authorities in Mexico, Jamaica, Suriname, and the United States. Literature Cited Arnold, M. L., R. J. Baker, and H. H. Genoways. 1 980. Evolutionary origin of Eptesicus lynni. Journal of Mammalogy, 61: 319-322. AsHTON, G. C, AND C. R. Balakrishinan. 1 974. Poly- morphisms of human salivary proteins. American Journal of Human Genetics, 26: 145-153. Azen, E. A. 1972. Genetic polymorphism of basic pro- teins from parotid saliva. Science, 176: 673-674. . 1973. Properties of salivary basic proteins showing polymorphism. Biochemical Genetics, 9: 69- 85. Azen, E. A., and F. G. Oppenheim. 1973. Genetic polymorphism of proline-rich human salivary pro- teins. Science, 180: 1067-1069. Ball, W. D. 1974. Isolation of salivary ribonuclease, deoxyribonucleaseand amylase from the parotid gland of the rat, Rattus norvegicus albinus. Biochimica et Biophysica Acta, 341: 305-317. Bjorkman, N. 1962. Lx)w magnification electron mi- croscopy in histological work. Acta Morphologica Neerlando-Scandinavica, 4: 344-348. Briggs, R., AND T. J. King. 1952. Transplantation of living nuclei from blastula cells into enucleated frogs' eggs. Proceedings of the National Academy of Sci- ences, USA, 38: 455-463. . 1957. Changes in the nuclei of differentiating endoderm cells as revealed by nuclear transplantation. j Journal of Morphology, 100: 269-3 1 2. \ Castle, J. D., J. D. Jamieson, and G. E. Palade. 1975. Secretion granules in the rabbit parotid gland. Isola- tion, subfractionation, and characterization of the membrane and content subfractions. Journal of Cell Biology, 64: 182-210. DE LA Torre, L. 1 96 1 . The evolution, variation, and systematics of the Neotropical bats of the genus Stur- nira. Ph.D. diss., University of Illinois, Urbana. DiSanto, P. E. 1960. Anatomy and histochemistry of the salivary glands of the vampire bat, Desmodus ro- tundus murinus. Journal of Morphology, 106: 301- 335. Eakjn, R. M. 1968. Evolution of photoreceptors, pp. 194-242. In Dobzhansky, T., M. K. Hecht, and W. C. Steere, eds.. Evolutionary Biology, vol. 2. Appleton- Century-Crofts, New York, 452 pp. EiSENBERG, J. F., AND D. E. WiLSON. 1978. Relative brain size and feeding strategies in the Chiroptera. Evolution, 32: 740-751. Feldman, J. L., AND C.J. Phillips. 1984. Comparative i retinal pigment epithelium and photoreceptor ultra- ' structure in nocturnal and fossorial rodents: The east- em woodrat, Neotomafloridana, and the plains pocket gopher, Geomys bursarius. Journal of Mammalogy, 65:231-245. FoRMAN, G. L. 1 972. Comparative morphological and histochemical studies of stomachs of selected Amer- * ican bats. University of Kansas Science Bulletin, 59: 591-729. FoRMAN, G. L., C. J. Philups, and C. S. Rouk. 1979. Alimentary tract, pp. 205-227. In Baker, R. J., J. K. Jones, Jr., and D. C. Carter, eds.. Biology of bats of the New World family Phyllostomatidae. Part III. Spe- ] cial Publications, The Museum, Texas Tech Univer- 1 sity, 16: 1-441. ' ' Freeman, P. W. 1979. Specialized insectivory: Beetle- eating and moth-eating molossid bats. Journal of Mammalogy, 60: 467-479. . 1981. Correspondence of food habits and mor- phology in insectivorous bats. Journal of Mammal- ogy, 62: 166-173. Gardner, A. L. 1977. Feeding habits, pp. 293-350. In Baker, R. J., J. K. Jones, Jr., and D. C. Carter, eds.. Biology of bats of the New World family Phyllosto- matidae. Part II. Special Publications, The Museum, Texas Tech University, 13: 1-364. Greenbaum, I. F, and C. J. Phillips. 1974. Compar- ative anatomy and general histology of tongues of long- nosed bats {Leptonycteris sanborni and L. nivalis) with reference to infestation of oral mites. Journal of Mam- malogy, 55: 489-504. Griftiths, T. A. 1982. Systematics of the New World nectar-feeding bats (Mammalia, Phyllostomidae), based on the morphology of the hyoid and lingual regions. American Museum Novitates, 2742: 1-45. 226 HELDIANA: ZOOLOGY . 1983. On the phylogeny of the Glossophaginae and the proper use of outgroup analysis. Systematic Zoology, 32: 283-285. Hand, A. R. 1980a. Cytochemical differentiation of the Golgi apparatus from GERL. Journal of Histo- chemistry and Cytochemistry, 28: 82-86. . 1980b. Salivary glands, pp. 336-370. /« Bhas- kar, S. N., ed., Orban's Oral Histology and Embryol- ogy. C. V. Mosby Co., St. Louis, Mo., 482 pp. Hand, A. R., and C. Oliver. 1 984. Effects of secretory stimulation on the Golgi apparatus and GERL of rat parotid acinar cells. Journal of Histochemistry and Cytochemistry, 32: 403-412. Henning, W. 1966. Phylogenetic Systematics. Trans- lated by D. D. Davis and R. 2^nger. University of Illinois Press, Urbana, 263 pp. Hood, C. S., and J. D. Smith. 1982. Cladistical anal- ysis of female reproductive histomorphology in phyl- lostomatoid bats. Systematic Zoology, 31: 241-251. . 1983. Histomorphology of the female repro- ductive tract in phyllostomoid bats. Occasional Pa- pers, The Museum, Texas Tech University, 86: 1-38. Ikemoto, S., and Y. Matsushima. 1 984. Sex and strain differences of mouse variant salivary proteins. Differ- entiation, 26: 55-58. Jacobsen, N., AND A. Hensten-Pettersen. 1974. Some biochemical characteristics of the salivary gland prod- ucts of Macaca irus and Cercopithecus aethiops. Com- parative Biochemistry and Physiology, 49B: 605-614. Jamieson, J. D., AND G. E. Palade. 1971. Synthesis, intracellular transport, and discharge of secretory pro- teins in stimulated pancreatic exocrine cells. Journal of Cell Biology, 50: 135-158. Junqueira, L. C. U., and F. Fava de MorAes. 1965. Comparative aspects of the vertebrate major salivary glands biology, pp. 36-48. In Wohlfarth-Bottermann, K. E., ed., Funktionelle und Morphologische Orga- nisation der Zelle. Sekretion und Exkretion. Springer- Verlag, Berlin. Junqueira, L. C. U., F. Fava de MorAes, and A. M. Toledo. 1967. Sialic acids in vertebrate salivary glands, saliva and pancreas. Archives of Oral Biology, 12: 151-157. Junqueira, L. C. U., A. M. S. Toledo, and A. I. Doine. 1 973. Digestive enzymes in the parotid and subman- dibular glands of mammals. Anais da Academia Bra- sileira de Ciencias, 45: 629-643. Kalt, M. R., AND B. Tandler. 1971. A study of fix- ation of early amphibian embryos for electron mi- croscopy. Journal of Ultrastructure Research, 36: 633- 645. Koop, B. P., AND R. J. Baker. 1983. Electrophoretic studies of relationships of six species ofArtibeus (Chi- roptera: Phyllostomidae). Occasional Papers, The Mu- seum, Texas Tech University, 83: 1-12. Kousvelari, E. E., F. G. Oppenheim, and L. S. Outler. 1982. Ultrastructural localization of salivary acidic proline-rich proteins from Macaca fascicularis. Jour- nal of Histochemistry and Cytochemistry, 30: 274- 278. Lekagul, B., and J. A. McNeely. 1977. Mammals of Thailand. Kurusapha Ladprao Press, Bangkok, 758 pp. MiNEDA, T. 1977. An electron microscopic investiga- tion on the submandibular gland of the Japanese horseshoe-bat {Rhinolopus (sic) ferrumequinum nip- pori). Histological and histochemical studies. Aichi- Gakuin Journal of Dental Science, 15: 35-47. . 1978. An electron microscopic study of the parotid gland of the Japanese horseshoe-bat {Rhino- lopus [sic] ferrumequinum nippon) in the functional and hibernating period. Aichi-Gakuin Journal of Dental Science, 16: 139-145. Nagato, T., B. Tandler, and C. J. Phillips. 1984. Unusual smooth endoplasmic reticulum in subman- dibular acinar cells of the male round-eared bat, To- natia sylvicola. Journal of Ultrastructure Research, 87: 275-284. Naumova, E. L 1981. Functional morphology of the digestive system of rodents and lagomorphs. USSR Academy of Sciences, Moscow, 261 pp. (In Russian.) Ono, T., Y. Sakai, N. Yamamoto, and K. Yasuda. 1 980. Double staining for immunohistochemistry: Si- multaneous demonstration of two enzymes in an iden- tical zymogen granule of an ultra-thin frozen section. Acta Histochemica et Cytochemica, 13: 436-448. Palade, G. 1975. Intracellular aspects of the process of protein synthesis. Science, 189: 347-358. Park, H., and E. R. Hall. 1951. The gross anatomy of the tongues and stomachs of eight New World bats. Transactions of the Kansas Academy of Sciences, 54: 64-72. Perutz, M. F. 1983. Species adaptation in a protein molecule. Molecular Biology and Evolution, 1: 1-28. Phillips, C. J. 1971. The dentition of glossophagine bats: Development, morphological characteristics, variation, pathology, and evolution. Miscellaneous Publications, University of Kansas, Museum of Nat- ural History, 54: 1-138. . 1985. Field fixation and storage of museum tissue collections suitable for electron microscopy. Acta Zoologica Fennica, 170: 87-90. Phillips, C. J., G. W. Grimes, and G. L. Forman. 1 977. Oral biology, pp. 1 2 1-246. In Baker, R. J., J. K. Jones, Jr., and D. C. Carter, eds., Biology of bats of the New World family Phyllostomatidae. Part II. Special Pub- lications, The Museum, Texas Tech University, 13: 1-364. Phillips, C. J., and B. Oxberrv. 1972. Comparative histology of molar dentitions of Microtus and Cleth- rionomys, with comments on dental evolution in mi- crotine rodents. Journal of Mammalogy, 53: 1-20. Phillips, C. J., and K. Studholme. 1982. Compara- tive ultrastructure of gastric chief cells in carnivorous and frugivorous bats. IRCS Medical Science, Cell and Membrane Biology, 10: 573-574. Phillips, C. J., K. M. Studholme, and G. L. Forman. 1 984. Results of the Alcoa Foundation Suriname ex- peditions. VIII. Comparative ultrastructure of gastric mucosae in four genera of bats (Mammalia: Chirop- tera), with comments on gastric evolution. Annals of Carnegie Museum, 53: 71-1 17. Phillips,C. J., AND B. Tandler. 1985. Unique salivary gland structure in two genera of tropical bats. Ana- tomical Record, 211: 152A. PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 227 . 1987. Mammalian evolution at the cellular level, pp. 1-66. In Genoways, H. H., ed.. Current Mammalogy. Plenum Press, New York, 519 pp. Phiixips, C. J., B. Tandler, and C. A. Pinkstaff. 1987. Unique salivary glands in two genera of tropical mi- crochiropteran bats: An example of evolutionary con- vergence in histology and histochemistry. Journal of Mammalogy, 68: 235-242. Pinkstaff, C. A., B. Tandler, and R. P. Cohan. 1982. Histology and histochemistry of the parotid and the principal and accessory submandibular glands of the little brown bat. Journal of Morphology, 172: 271- 285. Quay, W. B. 1954. The meibomian glands of voles and lemmings (Microtinae). Miscellaneous Publica- tions, Museum of Zoology, University of Michigan, 82: 1-24. Radtke, B. M. 1972. Observacoes estruturais, histo- quimicas e bioquimicas das glandulas salivares, par- otida e submandibular, de morcegos (Chiroptera, Mammalia). Ph.D. diss.. Institute of Biomedical Sci- ence, University of Sao Paulo, Brazil. Ravazzola, M., AND L. Orci. 1980. Glucagon and gli- centin immunoreactivity are topologically segregated in the a granule of the human pancreatic A cell. Na- ture, 284: 66-67. RiEGER, R., AND S. Tyler. 1979. The homology theo- rem in ultrastructural research. American Zoologist, 19: 655-664. RoBiNovrrcH, M. R., AND L. M. Sreebny. 1969. Sep- aration and identification of some of the protein com- ponents of rat parotid saliva. Archives of Oral Biology, 14: 935-949. Sands, M. W., R. P. Coppinger, and C. J. Phillips. 1 977. Comparisons of thermal sweating and histology of sweat glands of selected canids. Journal of Mam- malogy, 58: 74-78. Schick, J., R. Verspohl, H. Kern, and G. Scheele. 1 984. Two distinct adaptive responses in the synthe- sis of exocrine pancreatic enzymes to inverse changes in protein and carbohydrate in the diet. American Journal of Physiology, 247 (Gastrointestine and Liver Physiology, 10): G61 1-G616. SiMSON, J. A., R. M. DoM, P. L. Sannes, and S. S. Spicer. 1978. Morphology and cytochemistry of acinar se- cretory granules in normal and isoproterenol-treated rat submandibular glands. Journal of Microscopy, 113: 185-203. Slaughter, B. H. 1970. Evolutionary trends of chi- ropteran dentitions, pp. 51-83. In Slaughter, B. H., and D. W. Walton, eds.. About Bats. Southern Meth- odist University Press, E>allas, Texas, 339 pp. Smith, J. D. 1976. Chiropteran evolution, pp. 49-69. In Baker, R. J., J. K. Jones, Jr., and D. C. Carter, eds.. Biology of bats of the New World family Phyllosto- matidae. Part I. Special Publications, The Museum, Texas Tech University, 10: 1-218. Smith, J. D., and C. S. Hood. 1984. Genealogy of the New World nectar-feeding bats reexamined: A reply to Griffiths. Systematic Zoology, 33: 435-460. Stempak, J. G., AND R. T. Ward. 1964. An improved staining method for electron microscopy. Journal of Cell Biology, 22:697-701. Studier, E. H., B. C. Boyd, A. T. Feldman, R. W. Dapson, and D. E. Wilson. 1983. Renal function in the Neotropical bat, Artibeus jamaicensis. Com- parative Biochemistry and Physiology, 74A: 199-209. Tan, S. G., and G. C. Ashton. 1 976. An autosomal glucose-6-phosphate dehydrogenase (hexose-6-phos- phate dehydrogenase) polymorphism in human saliva. Human Heredity, 26: 1 13-123. Tandler, B. 1972. Microstructure of salivary glands, pp. 8-21. In Rowe, N. H., ed., A Symposium on Sal- ivary Glands and Their Secretion. University of Michigan Press, Ann Arbor, 125 pp. . 1978. Salivary glands and the secretory pro- cess, pp. 547-592. In Shaw, J. H., E. A. Sweeney, C. C. Cappuccino, and S. M. Meller, eds., Textbook of Oral Biology. W. B. Saunders Co., Philadelphia, 1 178 pp. Tandler, B., and R. P. Cohan. 1984. Ultrastructure of the parotid gland in the little brown bat. Anatomical Record, 210:491-502. Tandler, B., T. Nagato, and C. J. Phillips. 1983. Ultrastructure of the submandibular gland in four species of fruit bats (Artibeus sp.). Anatomical Record, 205: 198A. . 1986. Systematic implications of comparative ultrastructure of secretory acini in the submandibular salivary gland in Artibeus (Chiroptera: Phyllostomi- dae). Journal of Mammalogy, 67: 81-90. Tandler, B., and C. J. Phillips. 1985. Unusual se- cretory granules in the accessory submandibular gland of long- winged bats. Anatomical Record, 211: 192A. Tandler, B., C. J. Phillips, and T. Nagato. In press. Parotid salivary gland ultrastructure in an omnivorous Neotropical bat: Evolutionary diversity at the cellular level. Zoologica Scripta. Tandler, B., and R. J. Walter. 1977. Epon-Maraglas embedment for electron microscopy. Stain Technol- ogy, 52: 238-239. TuTTLE, M. D., AND M. J. Ryan. 1981. Bat predation and the evolution of frog vocalizations in the Neo- tropics. Science, 214: 677-678. Venable, J. H., AND R. CoGGESHALL. 1965. A simpli- fied lead citrate stain for use in electron microscopy. Journal of Cell Biology, 25(no. 2, pt. 1): 407-408. Wallach, D., R. Tessler, and M. Schramm. 1975. The proteins of the content of the secretory granules of the rat parotid gland. Biochimica et Biophysica Acta, 382: 552-564. Walton, D. W., and G. W. Walton. 1968. Com- parative osteology of the pelvic and pectoral girdles of the Phyllostomatidae (Chiroptera; Mammalia). Journal of the Graduate Research Center, Southern Methodist University, 37: 1-35. WiMSATT, W. A. 1956. Histological and histochemical observations on the parotid, submaxillary and sublin- gual glands of the tropical American fruit bat Artibeus jamaicensis Leach. Journal of Morphology, 99: 1 69- 209. 228 FIELDIANA: ZOOLOGY Appendix Specimens Examined Voucher specimens of each species and each locality are deposited in the collections of the Car- negie Museum of Natural History and The Mu- seum, Texas Tech University. All specimens ex- cept those of Leptonycteris sanborni (Jalisco, Mexico) and Ariteus flavescens and Eptesicus lynni (Jamaica) were collected in Brokopondo Province, Suriname. Pteronotus parnellii, 10 (5 S6, 5 29); Tonatia bi- dens, 3 {2 SS, 1 2); T. sylvicola, 2 {66); Trachops cirrhosus, 2 (1 3, 1 2); Phyllostomus elongatus, 3 (\6,2 22); P. latifolius, 1 (2); Glossophaga soricina, 3 (2 $6, 1 2); Leptonycteris sanborni, 2 (3<5); Sturnira lilium, 5 (2 35, 3 22); Artibens jamaicensis, 3 (22); Ariteus flavescens, 1 (2); Eptesicus lynni, 2 {$$); E. brasiliensis, 3 (22); Tadarida brasiliensis, 1 (S); Mo- lossus molossus, 8 (4 66, 4 22). PHILLIPS ET AL.: SALIVARY GLANDS IN BATS 229 Distribution of the Species and Subspecies of Cebids in Venezuela Roberta Bodini and Roger Perez-Hernandez ABSTRACTS Thirteen species of Primates representing nine genera in the family Cebidae are found in Venezuela. The geographic distribution of these species exhibits four main patterns. Alouatta and Cebus are widely distributed in all parts of the country. Four genera, Saimiri, Callicebus, Cacajao, and Chiropotes are restricted to south-central Venezuela; their distribution is centered in the Amazonian lowlands. Aotus and Ateles are each represented by one species or subspecies in south-central Venezuela and another in northwestern Venezuela. Pithecia is mainly restricted to the Guianan highlands in eastern Venezuela, with a single known outlying locality in south- central Venezuela. Of the nine cebid genera in Venezuela, all occur in south-central Venezuela, four occur in northwestern and north-central Venezuela, and only three occur in eastern Ven- ezuela. Distribution maps for the 1 3 species in Venezuela are presented with exact localities of specimens. Nueve generos de cebidos representados por trece especies se hallan en Venezuela. La re- particion geografica de las especies se resuelve en cuatro patrones geograficos principales. De este modo, Alouatta y Cebus son extensamente repartido por todo el pais. La distribucion de los generos, Saimiri, Callicebus, Cacajao y Chiropotes esta restringida al sur-central de Vene- zuela con concentracion en las tierras bajas Amazonicas. Aotus y Ateles son representatados, cada cual, por una especie (o subespecie), en el sur-central, y otra especie in el noroeste del pais. Pithecia se reparte en las alturas guayanas venezolanas del este, y es conocido, a la vez, por un solo dato de captura del sur-central venezolano. En resumen, la totalidad de los nueve generos esta presente en el sur-central de Venezuela, cuatro de ellos en el noroeste y norte- central, y tres en la Venezuela oriental. La reparticion de las trece especies esta documentado por mapas. Treze especies de Primatas, representando nove generos da familia Cebidae, sao encontradas na Venezuela. A distribuifao geografica destas especies exibem quatro padroes principais. Al- ouatta e Cebus espalham-se por todas partes do pais. Quatro generos, Saimiri, Callicebus, Cacajao, e Chiropotes, limitam-se ao centro-sul de Venezuela; a Bacia Amazonica sendo seus focos de distribuifoes. Aotus e Ateles sao representados por uma especie ou subespecie cada no centro-sul da Venezuela, e por outra especie no noroeste do pais. Pithecia limita-se ao planalto Guianense, no leste da Venezuela, com apenas uma localidade conhecida no centro- sul do Pais. Em suma, todos nove generos de Cebidae ocorrem no centro-sul da Venezuela, quatro ocorrem no centro-norte e noroeste do Pais, e apenas tres ocorrem no regiao leste da Venezuela. Apresentam-se mapas de distribui96es das 13 especies na Venezuela, com locali- dades exatas dos especimes colecionados. From the Institute de Zoologia Tropical, Universidad Central de Venezuela, Apartado 47058, Los Chaguara- mos, Caracas 1041 -A, Venezuela. BODINI & PEREZ-HERNANDEZ: CEBIDS IN VENEZUELA 231 Introduction Primates of the family Cebidae are some of the most conspicuous mammals in the Neolropics, as they are diurnal, often forage in large troops, and may be quite vocal. However, we actually know very little about the exact distributions of most species of cebids. In Venezuela, several authors have discussed cebids as part of generic revisions (see Elliot, 1912; Hershkovitz, 1949, and later works; HiU, 1960, 1 962; Kellogg & Goldman, 1944) and in works on the status or collections of specific species (i.e., Bodini, 1983; Handley, 1976; Mon- dolfi & Eisenberg, 1979; Rudran & Eisenberg, 1982). Herein we summarize the distribution of cebids in Venezuela based on recent collections, speci- men records from several museums, and the lit- erature. Specimens reported on are housed in the following collections: Estacion Biologica Rancho Grande (EBRG); Museo de Biologia, Universidad Central de Venezuela (MBUCV); Museo de Cien- cias Naturales (MCN); and Museo de Historia Natural La Salle (MHNLS). Our objectives are to provide accurate locality records and distribution maps that may be utilized in future investigations. Generic Distributions The nine genera of cebids that inhabit Venezuela exhibit four main patterns of geographic distri- bution (figs. 1-7). (1) Two geneT^—Alouatta and Cebus— arc widely distributed in all paris of the coimtry. (2) Four genera— Sa/m/r/, Callicebus, Cacajao, and Chiropotes—UTe restricted to south- central Venezuela, centering on the Amazonian lowlands (see Eisenberg & Redford, 1 979). (3) Two genera— .4orw5 and Ateles—are each represented by one species or subspecies in south-central Ven- ezuela and another in northwestern Venezuela or northwestern and north-central Venezuela, which is the region of the northeastern spurs of the An- dean chain and the enclosed Maracaibo basin. (4) One genus— Pithecia— is mainly restricted to the Guianan highlands in eastern Venezuela, with a single known outlying locality in south-central Venezuela. In summary, of the nine cebid genera in Venezuela, all occur in south-central Venezuela (including the outlier record of Pithecia), four oc- cur in northwestern and north-central Venezuela, and only three occur in eastern Venezuela. Specific and Subspecific Distributions Saimiri In a recent revision of the squirrel monkeys, Hershkovitz (1984) recognized four species: Sai- miri boliviensis, S. oerstedi. S. sciureus, and S. ustus. All those squirrel monkeys found in north- em South America he referred to the single species S. sciureus (fig. 1 ). Squirrel monkeys from Ama- zonian Venezuela and adjacent Brazil and Colom- bia he regarded as 5. sciureus cassiquiarensis. Hershkovitz (1984) plotted several localities in Territorio Federal Amazonas, but listed only a single specific locality: "Casiquiare, Rio (mouth), 2°01'N, 67"X)7'W." Hill (1960) previously had mapped the distribution of 5. sciureus as occurring throughout all of Venezuela, although he listed only a single locality in the state of Bolivar (Camp Canaracuni, 4''36'N, 64''10'W). The name cassi- quiarensis Lesson is based on Humboldt's descrip- tion of a captive female from the banks of the Rio Casiquiare, Amazonas, Venezuela. Specimens Examined— Total 29. Bolivar: Ca- naracuni (4°36'N, 64«'10'W). Territorio Federal Amazonas: Campo Cacuri (4°49'N, 65°26'W); Cano Yureba, Rio Ventuari (3''35'N, 66°46'W); Rio Pu- runame, 40 km from union with Rio Orinoco (3°19'N, 65°15'W); Rio Ventuari (3°59'N, 67'^2'W); San Fernando de Atabapo (4°02'N, 67»37'W); San Juan de Manapiare (5°14'N, 66'^2'W). Aotus Night monkeys are found from Panama throughout much of Amazonian South America to Paraguay. Historically, it has generally been as- sumed that all night monkeys represented the sin- gle species Aotus trivirgatus. However, in a recent revision of the genus, Hershkovitz (1983) recog- nized nine allopatric species. A tenth species, Aotus hershkovitzi from Colombia, has recently been proposed by Ramirez-Cerquera (1983). Two species of Aotus, A. lemurinus and A. trivirgatus. have been reported from Venezuela (fig. 2). Aotus lemurinus griseimembra Elliot is known in Venezuela only fi-om the extreme northwestern region, the states of Cojedes and Merida (Hersh- kovitz, 1983), and the states of Tachira, Trujillo, and Zulia (Handley, 1976). Specimens Examined— Total 10. Zulia: Campo a Rosario, EHstrito Catatumbo (l''44'N, 67°03'W); 232 FIELDIANA: ZOOLOGY BODINI & PEREZ-HERNANDEZ: CEBIDS IN VENEZUELA 233 234 HELDIANA: ZOOLOGY BODINI & PEREZ-HERNANDEZ: CEBIDS IN VENEZUELA 235 Reserva San Manuel, Los Canaguatos (9*t)4'N, 71''56'W); Kunana, Rio Negro, Sierra de Perija (9*75'N, 72*25'W); Rio Tocuco (72''25'W). Aotus trivirgatus trivirgatus Humboldt is known from the region south of the Rio Orinoco in Ter- ritorio Federal Amazonas and the state of Bolivar. The Venezuelan national collections contain specimens from Cano Manapiare and Rio Anta- vari, and two literature records report specimens from the right bank of the Rio Caroni (INP- ARQUES, 1982;MARNR-DGS-POA-SFS, 1982). Our recent discovery of Aotus trivirgatus in Guyana to the east of the Rio Caura, well outside of the Rio Orinoco basin, led to an investigation of geographic variation in the species. Preliminary results suggest that these A. trivirgatus represent a new geographic race (Bodini & Ferreira, in prep.). Specimens Examined— Total 1 1. Bolivar: Cano Manapiare (7»1 1'N, 66°40'W); Rio Antavari (S'^O'N, 63°10'W). Territorio Federal Amazonas: Alto Manapiare (S'^S'N, 66°02'W); San Juan de Manapiare (S^M'N, 66°02'W); near Isla Cigarron, Rio Negro (r44'N, 67°03'W); recently collected in La Neblina, east of Rio Varia (0°59'N, 66°10'W). Callicebus The genus Callicebus is the sole representative of the subfamily Callicebinae and is found from northern South America south to northern Para- guay. Hershkovitz (1963) and Kinzey (1982) rec- ognized three species, Callicebus moloch, C per- sonatus, and C. torquatus, of which only C. torquatus is found in Venezuela. Callicebus torquatus lugens Humboldt is known in Venezuela south of the Rio Orinoco (fig. 3). Hill (1960) originally suggested that C. torquatus is found as far east as Guyana, although this was questioned by both Handley (1976) and Hersh- kovitz (1963), who reported Venezuelan speci- mens only from Territorio Federal Amazonas. Handley (1976) reported 31 specimens from southern Territorio Federal Amazonas. Bodini (1981) reported specimens from Maripa, Camp Canaracuni, and the Rio Antavari in Venezuela. Recent reports by Kinzey ( 1 982) and Bodini (1981) confirm the presence of C torquatus in Guiana Region. Specimens Examined— Total 26. Bolivar: Cana- racuni (4'36'N, 64°10'W); Maripa, 150 km from Ciudad Bolivar (7*22'N, 65'^9'W); Rio Antavari (5»20'N, 63M0'W); Alto Paragua (4°30'N, 63*WW). Territorio Federal Amazonas: Alto Cano Caname, Departamento Atabapo (3*'33'N, 67"06'W); Alto Ventuari (3°50'N, 67°04'W); Boca Cano Maica, Rio Ventuari (66°30'W); Cano Ya- gua, Cerro Cucurito, Departamento Atabapo (3''38'N, 66»25'W); Cacuri, Alto Ventuari (4°49'N, 65"26'W); La Esmeralda (3"X)8'N, 65°32'W); Rio Cunucunuma (3°10'N, 66°0rW); Rio Puruname (3°19'N, 65''15'W); La Neblina, east of Rio Varia (0°59'N, 66°10'W). Alonatta Alouatta is the only genus within the subfamily Alouattinae and is represented by some six species widely distributed in Central and South America. A single species, Alouatta seniculus, is abundant and widely distributed in Venezuela. Hill (1962) recognized three subspecies of A. seniculus as oc- curring in Venezuela. Alouatta seniculus seniculus Linnaeus is foimd in extreme northwestern Ven- ezuela, primarily in the states of Apure, Tachira, and Zulia, with a single record from Barinas (Ti- coporo Forest, 8'X)6'N, 70*'40'W). Alouatta seni- culus arctoidea Cabrera, called the "Caracas howler monkey" by Humboldt, inhabits all the coastal region from Falcon to the state of Miranda. Ca- brera (1 958) proposed Caracas as the type locality. A third subspecies, A. seniculus stramineus Hum- boldt, inhabits all of Venezuela south of the Rio Orinoco in Territorio Federal Amazonas and the state of Bolivar (fig. 4). In addition to these three subspecies, we believe a fourth, previously unrecognized, form exists and is widely distributed throughout the Venezuelan llanos. This undescribed subspecies is character- ized by coloration and size. The southern limit of its distribution is clearly defined by the Rio Ori- noco, but its northern and western limits in the Andean piedmont are as yet undetermined. Howler monkeys are extremely adaptable to a wide array of environments, and we believe they are found throughout Venezuela. Their apparent absence in certain regions probably reflects lack of collecting rather than true distributional gaps, as our records demonstrate for the state of Anzoa- tegui. The problem of current and historical dis- tribution and systematic relationships of the var- ious populations in Venezuela is in need of study. Specimens Examined— Total 80. Anzoategui: Los Cocos, Rio Caris (8''30'N, 64'^5'W). Apure: Caiio San Agustin, Selva de San Camilo (7"'19'N, 7r57'W); Hato El Frio (7°44'N, 68°54'W); Las Raicitas, El Saman (7°55'N, 68°40'W). Aragua: Asentamiento Los Castillos, Turagua (10°09'N, 67»3rW); Cumbre de Guacamaya (10°2rN, 67'40'W); Los Picachos, Rancho Grande ( 1 0*2 1'N, 236 FIELDL\NA: ZOOLOGY BODINI & PEREZ-HERNANDEZ: CEBIDS IN VENEZUELA 237 6T4l'W). Barinas: Reserva Forestal Ticoporo (8°09'N, 70°40'W). Bolivar: Guayoba, Rio Caura, 4 km from Maripa (7°20'N, 65°10'W); Hato Bella Vista, El Palmar (8°00'N, 62°00'W); Represa del Guri, Operacion Rescate (7°40'N, 63°00'W); San Martin de Turumban, Anacoco-Cuyuni (6°42'N, 61°02'W). Carabobo: Las Quiguas, San Esteban (10''25'N, eS^OrW). Cojedes: Hato Barbasco (9°04'N, 68°08'W); Hato Itabana, 38 km from Las Vegas (9°17'N, 68*'13'W); Las Queseras, El Baul (8'*25'N, 68°17'W); Montaria Las Lomas, San Car- los (9°38'N, 68°34'W). Distrito Federal: Hacienda El Limon (10°28'N, 67''1 7'W). Falcon: Sanare, De- partamento Silva (10°23'N, 68°25'W); Sierra San Luis (1 1°15'N, 69°10'W). Guarico: Corozo Pando (8°14'N, 67°17'W); Hato Mapurite, 40 km N of Calabozo (9°17'N, 67°24'W); Manapire, near San Antonio (9°1 7'N, 66''1 1 'W); Rio Tiznado (8°1 8'N, 67°48'W). Lara: Cumbre de Las Trojas, 45 km S of Cabudare (9°45'N, 69°07'W). Miranda: Rio Ne- gro (10°20'N, ee'lS'W); La Guzmanera, Guatopo (10'X)0'N, 66°15'W). Portuguesa: Agua Blanca (9°40'N, 69°07'W). Tachira: Cerro El Teteo, Bura- gua (7°30'N, 7 r57'W); La Fria (8°1 3'N, 72°14'W); Paramo Tama (7°27'N, 72°26'W). Territorio Fed- eral Amazonas: Cacuri, Alto Ventuari (4°49'N, 65°26'W); Caiio Yureba, Ventuari (3°35'N, 66°46'W); Rio Hacha, Alto Ventuari (3°47'N, 65''38'W); Rio Puriname, 40 km from confluence of Rio Orinoco (3°19'N, 65°15'W); San Juan de Manapiare (9°05'N, 66°02'W); Cerro La Neblina, E of Rio Varia (0°59'N, 66°10'W). Territorio Fed- eral Delta Amacuro: Caiio Caneima (9°05'N, 60°55'W); Guiniquina (9°10'N, 61°06'W). Zulia: Laguna de Manaties, Departamento Catatumbo (9°27'N, 72°02'W); La Victoria, Rio Negro (9°36'N, 72°1 5'W); Rio Guasare ( 1 1°02'N, 72°05'W); Sierra de Perija (9°00'N, 72°00'W). Chiropotes The bearded sakis comprise two species limited to northern South America (Cabrera, 1958; Hersh- kovitz, 1977;Mittermeier«Sc.Coimbra-Filho, 1981). One species is found in southern Venezuela rep- resented by a single subspecies, Chiropotes satanas chiropotes Humboldt (fig. 5). All of our records and those reported by Handley (1976) are from Territorio Federal Amazonas. Cruz Lima (1945) and Rudran and Eisenberg (1982) proposed the occurrence of this species in the state of Bolivar on the basis of Humboldt's description; Mondolfi (1976) reported observations of Chiropotes at Maripa and Caiio Maniapure (Bolivar), but we are aware of no specimens from this region. Specimens Examined— Total 29. Territorio Federal Amazonas: Cacuri, Alto Ventuari (4°49'N, 65''26'W); Caiio Yureba, Departamento Atabapo (3°3rN, 66''44'W); Caiio Yagua, Cerro Cucurito (3''3 1 'N, 66°44' W); Laguna de Chiripo, Caiio Blan- co (3°27'N, 66°40'W); Rio Ocamo, Alto Orinoco (2°44'N, 65°11'W); Puerto Ayacucho (5°36'N, 67''35'W); Rio Orinoco, S of San Fernando de Ata- bapo (4'^0'N, 67°38'W); San Fernando de Ata- bapo (4°02'N, 67°37'W); San Juan de Manapiare (5°14'N, 66'^2'W). Cacajao The genus Cacajao, or uakaris, contains two species which are found in northern South Amer- ica (Cabrera, 1958; Hershkovitz, 1972; Mit'ter- meier & Coimbra-Filho, 1981). One species, Ca- cajao melanocephalus, is found in Venezuela restricted to the upper Orinoco region of southern Territorio Federal Amazonas (fig. 5). Although few specimens exist in collections, they appear to be abundant. Specimens Examined— Total 2. Territorio Fed- eral Amazonas: Alto Cano Atacavi, Departamento Casiquiare (3°05'N, 67'>02'W); La Nebhna, E of Rio Varia (0°59'N, 66°10'W). Cebus The capuchin monkeys are found from Hon- duras south through Central America and the northern two-thirds of South America. Four species currently are recognized (Cabrera, 1958; Hersh- kovitz, 1972). Three species of Cebus are repre- sented in Venezuela: Cebus albifrons, C apella, and C. «/^rm//a/z^s (Hershkovitz, 1949, 1958). Three subspecies of Cebus albifrons are found in Venezuela (fig. 6). Cebus albifrons adustus Hershkovitz was described on the basis of three specimens from "near head of Rio Cogollo (Apon) at eastern base of Sierra de Perija, about 5 kilo- meters northwest of Machiques, Lake Maracaibo region, Zulia" (Hershkovitz, 1949, p. 369). We report additional specimens from Rio Guasare and Kasmera. This subspecies is restricted to the Sierra de Perija of extreme northwestern Venezuela and adjacent Colombia. Cebus albifrons leucocephalus Gray is found in extreme western Venezuela, in the region south of the Lago de Maracaibo basin. 238 HELDIANA: ZOOLOGY BODINI & PEREZ-HERNANDEZ: CEBIDS IN VENEZUELA 239 and the states of Apure, Merida, Tachira, and Zu- lia. Cebus albifrons unicolor Spix is found in ex- treme southern Venezuela in Territorio Federal Amazonas. Hershkovitz (1949) reported speci- mens from Marimonda, Rio Orinoco, and from Solano, Rio Casiquiare. Handley (1976) reported specimens from Rio Mavaca and Tamatama. All locality records for this subspecies in Venezuela are south of the Rio Ventuari. Specimens Examined— Total 15. Apure: Caiio San Augustin, Selva de San Camilo (7°19'N, 7r57'W). Merida: Palmichoso, S of Las Virtudes (9'^9'N, 70°57'W). Tachira: La Fria (8°13'N, 72°14'W). Territorio Federal Amazonas: Cano Yagua, Cerro Cucurito, Departamento Atabapo (B^SS'N, 66''25'W); near Boca Padamo, left side of Rio Orinoco (3°02'N, 65°13'W). Zulia: Kasmera, Perija (10°05'N, 72°45'W); Kunana (9°36'N, 72°15'W); Rio Bravo, Distrito Catatumbo (9°05'N, 72°22'W); Rio Guasare (1 1°02'N, 72°05'W). Cebus apella is represented in Venezuela by two subspecies (fig. 6). Cebus apella apella Linnaeus is restricted to Amazonian Venezuela, the Territorio Federal Amazonas, and is found along both banks of the upper Orinoco. Cebus apella margaritae HoUister is endemic to and restricted to Margarita Island. The 800-km gap between the ranges of these two subsp>ecies is striking and unexplained. Specimens Examined— Total 10. Nueva Espar- ta: Sierra de Copey (1 1°03'N, 63°56'W). Territorio Federal Amazonas: Alto Caiio Caname (3''22'N, 67°08'W); Caiio Yapacana (3°30'N, 66°45'W); San Fernando de Atabapo (4°02'N, 67°37'W). Cebus nigrivittatus is widely distributed in Ven- ezuela and represented by perhaps five subspecies (fig. 6) (Hershkovitz, 1949; Cabrera, 1958). Cebus nigrivittatus apiculatus was described by Elliot (1912) on the basis of specimens from La Union, Rio Caura, near its confluence with the Rio Ori- noco. It is distributed throughout central Vene- zuela south of the Orinoco between the Rio Caroni and the Rio Ventuari. Cebus nigrivittatus brunneus was described by J. A. Allen (1914) from specimens from Aroa, a station on the Bolivar Railway, Yaracuy, north- western Venezuela. Hershkovitz (1949) reported an additional specimen from the Paria Peninsula. These records plus our specimens suggest that C. n. brunneus is continuously distributed through- out the Cordillera de la Costa of extreme northern Venezuela. Cebus nigrivittatus nigrivittatus Wag- ner is restricted in Venezuela to the Amazonian region of Territorio Federal Amazonas. Cebus ni- grivittatus olivaceus Schomburg is found in south- eastern Venezuela. Hershkovitz (1949, p. 348) re- ports the type locality as "Vicinity of 'Our Village,' said to be situated at latitude 4°57'N., 61°rW., altitude 3,100 feet above sea level, southern foot of Mount Roraima." We believe that an undescribed subspecies, Ce- bus nigrivittatus subsp., is widely distributed throughout central and northern Venezuela north of the Orinoco. The status of this population is currently under study. The Orinoco Delta region, which Eisenberg and Redford ( 1 979) excluded from their consideration of biogeographic regions due to insufficient data, is now shown by the distributions of Alouatta, Cebus, and Pithecia to be clearly aligned with the Guyana highlands. The Llanos region now extends up to the western edge of the delta, but does not include it, as demonstrated by the distributions of Alouatta and Cebus. Specimens Examined— Total 75. Anzoategui: Los Cocos, Rio Caris (8°30'N, 64°05'W); 10 km W of Laguna de Unare (10°02'N, 65°12'W); Mor- ichal Largo [between Anzoategi and Monagas] (8°18'N, 63°15'W). Aragua: Rancho Grande (10°10'N, 67°19'W). Barinas: Reserva ForestalTi- coporo, on Barinas-Pedregal road (8°03'N, 70°18'W). Bolivar: Caiio La Urbina (7°15'N, 66°25'W); Carretera Caicara-S. Juan de Mana- piare, km 1 75 (6°02'N, 66°29'W); Carretera El Do- rado-Santa Elena, km 33 (6°12'N, 6ri4'W); Cur- aima. El Palmar (8°01'N, 61°26'W); El Dorado- Santa Elena, km 121 (5°18'N,6ril'W);Guayopo, Rio Caura, 14 km from Maripa (7°09'N, 65°10'W); Canaracuni (4°17'N, 64°05'W); Guri, Operacion Rescate (7°18'N, 63°00'W); Rio Antavari (5'i09'N, 63°05'W); Rio Marajano, Meseta de Jaua (4°08'N, 64°1 1'W); Rio Villacoa, 4 km N of mouth (6°16'N, 67'^5'W); San Martin de Turumban, Rio Cuyuni (6"'19'N, 61'^9'W). Carabobo: Bahia de Patanemo (10"'12'N, 67''26'W); Urama (10°12'N, 68°08'W). Cojedes: Cerro Azul, La Blanquera (8°26'N, 68°07'W); Montana Las Loma, San Carlos (9°1 7'N, 68°16'W); Pica, Las Vegas (9°1 5'N, 68°1 7'W). Dis- trito Federal: El Avila, Caracas (10° 14'N, 66°13'W); Hacienda El Limon (lO^B'N, 67°08'W). Falcon: Sanare, Distrito Silva (8''23'N, 68°12'W). Guarico: Hato Flores Moradas, Calabozo (8°23'N, 67"! 3'W); Hato Mapurite, 40 km N of Calabozo (9°08'N, 67°irW); Parmana (7°28'N, 65°18'W); San Jose de Tiznados (9°16'N, 67''16'W). Lara: La Pastora, 1 1 km SSW of Sanare (9°2rN, 70°07'W). Miran- da: Rio Negro (10''20'N, 66°17'W); La Guzma- nera, Guatopo (10°00'N, 66°15'W). Territorio Federal Amazonas: Alto Manapiare (5''13'N, 66°0rW); Alto Ventuari (4°45'N, 65°20'W); Caiio Yureba, Rio Ventuari (3°16'N, 66''21'W); Cano 240 FIELDIANA: ZOOLOGY ■ * m ** T r 2 5 ^ :: 2 • > Z Q. • =» « e e e e e u u u o o o ■ o o n ♦ < c o 3 •c BODINI & PEREZ-HERNANDEZ: CEBIDS IN VENEZUELA 241 < 242 HELDIANA: ZOOLOGY Mayaba, Rio Ventuari (4°07'N, 66°16'W); Cano Morrocoy, Alto Ventuari (5°08'N, 66°02'W); La Esmeralda (S-^S'N, 65°32'W); Ocamo, Rio Oca- mo (2°20'N, 65°15'W); Puruname, 40 km from Rio Orinoco (3°19'N, 65°15'W); La Neblina, E of Rio Varia (0°59'N, 66°10'W). Territorio Federal Delta Amacuro: Cano Araguabisi (9°12'N, 60*'27'W); Guiniquina (9''10'N, 61'X)3'W); Tobe- juba, Guayo (9°09'N, 6r25'W). Yaracuy: Agua Negra (10°14'N, 68''14'W); Carretera Boca de Aroa, 20 km from Palmasola (lO^Ol'N, 69°27'W). Ateles Spider monkeys are widely distributed from northeastern Mexico throughout tropical South America. Four species are recognized, of which only Ateles belzebuth occurs in Venezuela (fig. 7). Ateles belzebuth belzebuth GeofFroy is found in southern Venezuela, south of the Rio Orinoco; most records are from Territorio Federal Ama- zonas. Kellogg and Goldman ( 1 944) indicate a wide distribution in Guyana and report the Venezuelan localities of La Union, Rio Mato, and El Llagual (on both banks of the Rio Caura). Ateles belzebuth hybridus Geoffroy is known from northern and western Venezuela (Hershko- vitz, 1949; Cabrera, 1958; Kellogg & Goldman, 1944). Handley (1976) reported specimens from Apure and Trujillo in western Venezuela. Our specimens are from the states of Barinas, Tachira, and Zulia. Mondolfi and Eisenberg (1979) report- ed it from Cupira and Guatopo, state of Miranda, suggesting a discontinuous distribution on the coast. Specimens Examined— Total 16. Barinas: Re- serva Forestal de Ticoporo, Sabana de Anare (8°06'N, 70°40'W). Bolivar: Canaracuni (4°06'N, 64°10'W). Miranda: Cupira (10°10'N, 65°44'W). Tachira: La Fria (8°13'N, 72°14'W). Territorio Federal Amazonas: Cacuri, Rio Ventuari (4°49'N, 65°26'W); Rio Ocamo, Alto Orinoco (2°44'N, 65°11'W); Salto del Oso, Alto Ventuari (4''55'N, 65°25'W); San Juan de Manapiare (S^M'N, 66°02'W). Zulia: Rio Guasare ( 1 1°02'N, 72°05'W). Lagothrix Woolly monkeys have not been collected in Venezuela; however two subspecies of Lagothrix lagothricha are to be expected: L. lagothricha la- got hricha should be found in Territorio Federal Amazonas south of the Rio Ventuari, and L. I. lugens, in the Selva de San Camilo, state of Apure (Fooden, 1963; Hemandez-Camacho & Cooper, 1976). Pithecia Sakis are found only in northern South America; Hershkovitz (1979) recognized four monotypic species, of which only Pithecia pithecia Linnaeus is found in Venezuela. P. pithecia is found in Ven- ezuela south of the Rio Orinoco and throughout the Guianas and northeastern Brazil. In Venezuela specimens have been reported primarily from the extreme northeastern region, the state of Bolivar and Territorio Federal Delta Amacuro (fig. 5), with a single outlier locality record for Belen, Rio Cunu- cunuma, Territorio Federal Amazonas (3°39'N, 65°46'W) (Handley, 1976). If the distribution pro- posed by Hershkovitz (1979) and Mittermeier and Coimbra-Filho (198 1) is correct, P. pithecia should inhabit the entire region between the upper Rio Orinoco and the Rio Caroni, an enormous area for which no specimens have been recorded. Specimens Examined— Total 24. Bolivar: Rio Curumo (7°15'N, 61°20'W); Rio Grande (8°16'N, 6 ri 7'W); Gurisoco, El Palmar (8°62'N, 6 1°26'W); La Trinidad, El Palmar (7°12'N, 6r23'W); Guri, Operacion Rescate (7°18'N, 63°00'W); Rio Boto- namo, near Rio Cuyuni (6°59'N, 6rirW); San Martin de Turumban, Rio Cuyuni (6''59'N, 61°02'W). Territorio Federal Delta Amacuro: Al- tiplanicie de Nuria (7°50'N, 61°18'W); Yotacuay, SW of Cupiare (8°30'N, 61°00'W). Literature Cited Allen, J. A. 1914. New South American Monkeys. BviUetin of the American Museum of Natural History, 33: 653. BoDiNi, R. 1981. Musculatura locomotora de la viudita (Callicebus torquatus lugens). Sus implicaciones fun- cionales y iilogeneticas. Memoria de la Sociedad de Ciencias Naturales La Salle, 16(XL): 1-165. . 1983. Distribuci6n, status de la investigacion y conservaci6n de los cebidos en Venezuela, p. 1 39. In Symjwsio sobre Primatologia en Latinoamerica. IX Congreso Latinoamericano de Zoologia, Arequipa, Peru. Cabrera, A. 1957(1958]. Calalogo de los mamiferos de America del sur. Revista Museo Argentine de Cien- cias Naturales, "Bernardino Rivadavia," 4(1): xviii + 138 pp. Cruz Lima, E. 1945. Mammals of Amazonia. I. Gen- eral introduction and Primates. Contribution from the BODINI & PEREZ-HERNANDEZ: CEBIDS IN VENEZUELA 243 Museu Paraense Emilio Goeldi de Historia Natural e Etnografia, Belem do Para, Rio de Janeiro, Brasil, 274 pp., 42 pis. EiSENBERG, J. F., AND K. Redford. 1979. A biogeo- graphic analysis of the mammalian fauna of Vene- zuela, pp. 31-36. In Eisenberg, J. F., ed., Vertebrate Ecology in the Northern Neotropics. Smithsonian In- stitution Press, Washington, D.C., 271 pp. Elliot, D. G. 1912. A Review of the Primates. Vols. I and II. American Museum of Natural History, New York. FooDEN, J. 1963. A revision of the woolly monkey (genus Lagothrix). Journal of Mammalogy, 44: 213- 247. Handley, C. O., Jr. 1976. Mammals of the Smith- sonian Venezuelan project. Brigham Young Univer- sity Sciences Bulletin, Biological Series, 20(5): 1-91. Hernandez-Camacho, J., and R. W. Cooper. 1976. The nonhuman Primates of Colombia, pp. 35-69. In Thorington, R. W., Jr., and P. G. Heltne, eds.. Neo- tropical Primates: Field Studies and Conservation. National Academy of Sciences, Washington, D.C. Hershkovitz, P. 1949. Mammals of northern Colom- bia. Preliminary report No. 4: Monkeys (Primates), with taxonomic revisions of some forms. Proceedings of the United States National Museum, 98(3232): 323- 427. . 1958. Type localities and nomenclature of some American primates, with remarks on secondary hom- onyms. Proceedings of the Biological Society of Wash- ington, 71: 53-56. 1963. A systematic and zoogeographic account of the monkeys of the genus Callicebus (Cebidae) of the Amazonas and Orinoco river basins. Mammalia, 27: 1-79. . 1972. The Recent mammals of the Neotropical region: A zoogeographic and ecological review, pp. 31 1-431. In Keast, A., F. C. Erk, and B. Glass, eds.. Evolution, Mammals, and Southern Continents. State University of New York Press, Albany, N.Y. 543 pp. 1977. Living New World Monkeys (Platyrrhi- ni) with an Introduction to Primates. University of Chicago Press, Chicago, 1117 pp. 1 979. The species of sakis, genus Pithecia (Ce- bidae, Primates), with notes on sexual dichromatism. Folia Primatologica, 31: 1-22. 1983. Two new species of night monkeys, ge- nus Aotus (Cebidae, Platyrrhini): A preliminary report on Aotus taxonomy. American Journal of Primatol- ogy, 4: 209-243. . 1984. Taxonomy of squirrel monkeys genus Saimiri (Cebidae, Platyrrhini): A preliminary report with description of a hitherto unnamed form. Amer- ican Journal of Primatology, 7: 155-210. Hill, W. C. O. 1 960. Primates. Comparative Anatomy and Taxonomy. IV. Cebidae, Part A. Edinburgh Uni- versity Press, London, 523 pp. . 1962. Primates. Comparative Anatomy and Taxonomy. V. Cebidae, Part B. Edinburgh University Press, London, 537 pp. INPARQUES. 1982. Guia de los Parques Nacionales y Monumentos Naturales de Venezuela. Ediciones Fundacion de Educacion Ambiental, Caracas. Kellogg, R., and E. A. Goldman. 1944. Review of the spider monkeys. Proceedings of the United States National Museum, 96(3186): 1-45. Kjnzey, W. G. 1982. Distribution of primates and for- est refuges, pp. 455-482. In Prance, G. T., ed.. Bio- logical Diversification in the Tropics. Columbia Uni- versity Press, New York, 7 1 4 pp. MARNR-DGS-POA-SFS. 1982. Informe Nacional de Fauna Silvestre. Ministerio del Ambiente y de los Re- cursos Naturales Renovable, Caracas. Mittermeier, R. a., and A. Coimbra-Filho. 1981. Systematic: Species and subspecies, pp. 29-109. In Coimbra-Filho, A. F., and R. A. Mittermeier, eds.. Ecology and Behavior of Neotropical Primates. Aca- demia Brasileira de Ciencias, Rio de Janeiro. MoNDOLFi, E. 1976. Fauna silvestre de los bosques humedos tropicales de Venezuela, pp. 113-181. In Hamilton, L. S., J. Steyermark, J. P. Veillon, and E. Mondolfi, eds., Conservacion de los bosques humedos de Venezuela. Sierra Club— Consejo de Bienestar Ru- ral, Caracas. MoNDOLR, E., AND J. F. EiSENBERG. 1 979. Ncw records for A teles belzebuth hybridus in northern Venezuela, pp. 93-96. In Eisenberg, J. F., Vertebrate Ecology in Northern Neotropics. Smithsonian Institution Press, Washington, D.C, 271 pp. Ramirez-Cerquera, J. 1983. Reporte de una nueva especit de Primates del genero Aotus de Colombia, p. 146. Symposio sobre Primatologia en Latinoamerica. IX Congreso Latinoamericano de Zoologia. Arequipa, Peru. RuDRAN, R., AND J. F. Eisenberg. 1982. Conservation and status of wild primates in Venezuela, pp. 52-59. In Olney, P. J. S., ed.. International Zoo Yearbook, Vol. 22. The Zoological Society of London, London, 488 pp. 244 FIELDIANA: ZOOLOGY Host Associations and Coevolutionary Relationships of Astigmatid Mite Parasites of New World Primates I. Families Psoroptidae and Audycoptidae Barry M. OConnor ABSTRACTS Coevolutionary patterns among mites of the families Psoroptidae and Audycoptidae and New World Primates are reviewed. Host records for primate parasites originally described from artificial situations are compared with field collections from Peruvian primates, with most host associations verified. A new sp)ecies of Audycoptidae, Saimirioptes hershkovitzi, is described from Cebus apella. The psoroptid subfamily Cebalginae is hypothesized to be a monophyletic group based upon 10 synapomorphies. Phylogenetic relationships within the Cebalginae are reviewed, with cospeciational histories supported for the genera Alouattalges and Schizopodalges and their hosts, and a more diffuse cospeciational pattern observed for the genera Cebalgoides, Cebalges, and Fonsecalges and their hosts. Historical relationships of the genus Procebalges remain problematical. Patrones coevolutivos entre acaros de las familias Psoroptidae y Audycoptidae y primates del Nuevo Mundo son revisadas. Registros de huespedes para parasitos de primates originar- iamente descritos de situaciones artificiales son comparados con colecciones de campo de primates peruanos, con la mayoria de asociaciones de huesped verificadas. Una nueva especie de Audycoptidae, Saimirioptes hershkovitzi, es descrita de Cebus apella. Los psoroptidos de la subfamilia Cebalginae son hipotetizados ser un grupo monofiletico basado en 10 sinapo- morfias. Relaciones filogeneticas entre los Cebalginae son revisadas, con historias coesp>ecia- cionales soportadas F>or el genero Alouattalges y Schizopodalges y sus huesp>edes, y un patron coespeciacional mas difuso observado en los generos Cebalgoides, Cebalges y Fonsecalges y sus huesjjedes. Relaciones historicas del genero Procebalges p>ermanecen problematicas. Padroes coevolucionarios entre os acarinos das familias Psoroptidae e Audycoptidae, e os primatas do Novo Mundo, sao revisados. Os registros de hospedes primatas de parasitas, que foram descritos em situa^oes artificials (cativeiros), sao comparados com cole96es de campo de primatas peruanos, e a maioria das associagoes hospedeiras atualmente registradas foram averiguadas. Uma nova esjjecie de Audycoptidae, Saimirioptes hershkovitzi, encontrada em Cebus apella e descrita. Baseando-se num estudo de 1 0 sinapomorfias, prop6e-se ser a subfamilia psoroptidea, Cebalginae, um grupo monofiletico. As rela96es filogeneticas entre os Cebalginae sao revisadas, e as historias de coesF>ecializa9ao entre os generos Alouattalges e Schizopodalges, e seus respectivos hospedes, sao confirmadas. Os padroes de coespecializa9ao entre os generos Cebalgoides, Cebalges e Fonsecalges, e seus hospedes, sao mais difusos. Os relacionamentos historicos do genero Procebalges permanecem incertos. From the Museum of Zoology and Department of Bi- ology, The University of Michigan, Ann Arbor, MI 48 109. OCONNOR: MITE PARASITES OF NEW WORLD PRIMATES 245 Introduction Coevolutionary patterns among hosts and par- asites have been the focus of much recent study and discussion (Brooks, 1979, 1981, 1985; Brooks & Glen, 1982; Futuyma & Slatkin, 1983; Nitecki, 1983). Phylogenetic analysis of evolutionary re- lationships among parasite groups or host groups can provide additional data sets (e.g., treating par- asite distributions as character states of their hosts), or such analyses can be used as tests for hypotheses regarding evolutionary relationships of the other lineages of associated organisms. These phyloge- netic analyses are especially useful when the par- asite groups are host specific and indicate little history of secondary colonization of new hosts. The associations between mites and primates are particularly amenable to such analysis because a number of acarine groups are specifically associ- ated with primates, and these lineages exhibit enough within-group diversity to allow the con- struction and comparison of phylogenetic hypoth- eses. I have previously detailed such hypotheses for several groups of astigmatid mites parasitic on primates (OConnor, 1984). Formulation of hypotheses regarding the his- tory of associations between primates and their associated mites requires three steps. First, the parasite taxa must be described and their natural host and geographic ranges discovered. Second, taxa above the species level in classically derived classifications must be tested for naturalness (i.e., monophyly). Finally, phylogenetic relationships among all taxa must be elucidated. In the present paper, each of these questions will be addressed for the associations among certain groups of as- tigmatid mites and New World Primates. For the reader interested in summary information regard- ing known host-parasite relationships for New World primates, an exhaustive list of literature records may be found in Hershkovitz (1977). One major difficulty in the application of this methodology to the study of the history of mite- primale associations is the scanty knowledge of the distribution and identity of the parasite sp)ecies and their natural host ranges. Although many aca- rine parasites of primates have been described, much of the material has originally come from zoos and primate research centers where the f>os- sibility of unnatural interspecific contact between host species makes the transfer of parasites a real problem. Many other sjjecies have been described from preserved host specimens in museum col- lections where contamination in the field or in the museum may also have been a problem. Finally, the actual field locality from which either parasites or hosts were collected is known for extremely few of the known primate parasites. The uncertainties involved in the host range and geographic distri- bution of so many of the known species of primate parasites render the phylogenetic analyses pro- posed earlier (OConnor, 1984) subject to some doubt. In 1 98 1 , 1 was invited by Philip Hershkovitz to collect parasites from specimens of a number of primate species which had been collected or ob- tained during field studies in Peru in 1 980. A very large number of mites was collected, providing a unique survey of the primate parasites in a small area in Peru and a test for previously reported host-parasite associations. These collections yield- ed specimens belonging to three families of astig- matid mites: Psoroptidae, Audycoptidae, and Atopomelidae. In the present paper, the collec- tions of Psoroptidae and Audycoptidae will be discussed. The collections of Atopomelidae, con- sisting of a number of described and undescribed species of the genus Listrocarpus, will be studied separately. r Materials and Methods Primate specimens were collected in the field by Hershkovitz or obtained from local individuals. These specimens were prepared by removing and simply drying the skins at the time of collection. Upon their arrival in the United States, one or more skins of each species collected were made available to me for parasite removal before the skins were sent for tanning. The following species, all identified by Hershkovitz, were examined (number examined in parentheses): Cebus apella (3); C. albifrons (1); Lagothrix lagothricha (1); L. flavicauda ( 1 ); Alouatta seniculus ( 1 ); Pithecia hir- suta (3); Callicebus moloch (1); Aotus nancymai (4); and Saimiri sciureus (5). Parasites were removed in two ways. First, all skins examined were vigorously brushed over white paper, with the dislodged parasites collected under a dissecting microscop>e. Finally, one skin of each species was soaked in water and mild soap until soft (the single specimen oi L. flavicauda was not soaked). These skins were then gently washed, the wash water filtered through a 200-mesh sieve (mesh 246 HELDIANA: ZOOLOGY openings 75 micrometers), and the residue ex- amined under a dissecting microscope. Mites col- lected were preserved in 70% ethanol for subse- quent study. In the laboratory, mites were cleared in lacto- phenol and mounted in Hoyer's medium, with some specimens retained in alcohol in the cases of large series. Voucher specimens of nominal species are deposited in Field Museum of Natural History, Chicago, and, when available, will be placed in the following institutions: Museum of Zoology, The University of Michigan, Ann Arbor; The United States National Museum of Natural History, Washington, D.C.; LTnstitut Royale des Sciences Naturelles, Brussels, Belgium; and the collection of F. S. Lukoschus, Katholieke Uni- versiteit, Nijmegen, Netherlands. Species Accounts tritonymphs. Among 1 1 protonymphs in the pres- ent collection, eight exhibit characteristics similar to the male tritonymphs (i.e., posterior opistho- somal lobes sclerotized; coxal fields III well scler- otized) while three exhibit characteristics of female tritonymphs (posterior lobes unsclerotized; scler- otization of coxal fields III much weaker). I inter- pret these differences as evidence for sexual di- morphism at the protonymphal instar in this species. Material Examined— Total 133. Twenty-nine females, 43 males (of which 3 1 were in tandem with female tritonymphs), 38 female tritonymphs, 8 male tritonymphs, 3 female protonymphs, 8 male protonymphs, 4 larvae from Cebus albifrons. PERU, Loreto: Nauta, Rio Tigre, 6 km above Rio Tigrillo; 18 December 1980; P. Hershkovitz (9264). Host now a tanned skin (fmnh 122795). Mites labeled bmcx: 81-081 1-3. No specimens were re- covered from Cebus apella. Family PSOROPTIDAE Seven species of mites in the family Psoropti- dae, subfamily Cebalginae, are known to parasitize New World primates. These species, with their known hosts and distributions, are listed below, with new records from the Peruvian collections of Hershkovitz indicated under "Material Exam- ined." Keys to most of these species may be found in Fain (1963c). Cebalgoides cebi Fain, 1963 Cebalgoides cebi Fain, 1 963, Bull. Ann. Soc. Roy. Ent. Belg., 99: 331. Fain, 1963, Bull. Inst. Roy. Sci. Nat. Belg., 39(32): 91. This species was briefly diagnosed by Fain (1963a) from specimens collected from Cebus al- bifrons which originated in "Amerique du Sud" and died in the Antwerp (Belgium) Zoo. It was more thoroughly described and illustrated by Fain (1963c), who listed specimens from Cebus albi- frons from Venezuela (type collection), C apella from "Amerique du Sud," and Leontocebus (Oed- ipomidas) oedipus (= Saguinus oedipus) from Co- lombia. All these hosts had died in the Antwerp Zoo. Fain (1963c) noted sexual dimorphism in tri- tonymphs of C cebi. The two protonymphs he examined exhibited the characteristics of the male Alouattalges corbeti Fain, 1963 Alouattalges corbeti Fain, 1963, Bull. Inst. Roy. Sci. Nat. Belg., 39(32): 122. Fain, 1966, Acarologia 8: 103. Rosalialges cruciformis Lavoipierre, 1 964, Acarologia, 6: 348. This species was briefly described from the ho- lotype female collected from a preserved specimen of Alouatta seniculus macconnelli which had been collected at Paramaribo, Surinam and preserved in the British Museum (Natural History) (Fain, 1 963c). Although only the holotype was described, several specimens were apparently recovered from this host. Almost simultaneously, Lavoipierre (1964a) described and figured the female based upon two specimens collected from an "Aotes'''' (sic) sp. which had died in San Francisco, Cali- fornia, after its importation from Peru. Fain (1966) provided illustrations of the female and illustrated but did not describe the male beyond length and width measurements. Material Examined— Total 15. Five females, 8 males, 2 female tritonymphs from Alouatta se- niculus. PERU, Loreto: Nauta, Rio Samiria; 18 November 1980; P. Hershkovitz (9050). Host now a tanned skin (fmnh 122789). Mites labeled bmoc 81-0809-5. No specimens were recovered from Aotus nancymai, suggesting the possibility that the two specimens collected by Lavoipierre (1964a) represent contamination. OCONNOR: MITE PARASITES OF NEW WORLD PRIMATES 247 Schizopodalges lagothricola Fain, 1 963 Schizopodalges lagothricola Fain, 1963, Bull. Ann. Soc. Roy. Ent. Belg., 99: 469. Fain, 1963, Bull. Inst. Roy. Sci. Nat. Belg., 39(32): 100. This species was described from numerous spec- imens from two juvenile Lagothrix lagothricha which died in the Antwerp Zoo. The origin of the hosts was stated as "Amerique du Sud" (Fain, 1 963b). Fain ( 1 963c) provided illustrations of male and female and indicated that the hosts originated in Colombia. Material Examined— Total 30. Eleven fe- males, 1 1 males, 3 tritonymphs, 2 protonymphs, 3 larvae from Lagothrix lagothricha. PERU, Lo- reto: Nauta, Rio Samiria; 15 November 1980; P. Hershkovitz (9032). Host now a tanned skin (fmnh 122790). Mites labeled bmcx: 81-0809-10. Cebalges gaudi Fain, 1 962 Cebalges gaudi Fain, 1962, Rev. Zool. Bot. Afr., 66: 160. Fain, 1963, Bull. Inst. Roy. Sci. Nat. Belg., 39(32): 81. This species was briefly described from numer- ous sp)ecimens collected from a preserved speci- men of Cebus capucinus with no locality infor- mation (Fain, 1962). Full descriptions and figures were provided later (Fain, 1963c). Material Examined— Four females from Ce- bus apella. PERU, Loreto: Nauta, Rio Samiria; 18 November 1980; P. Hershkovitz (9049). Host now a tanned skin (fmnh 122792). Mites labeled bmoc 81-0811-2. Fonsecalges johnjadini Fain, 1962 Fonsecalges johnjadini Fain, 1 962, Rev. Zool. Bot. Afr., 66: 161. Fain, 1963, Bull. Ann. Soc. Roy. Ent. Belg., 99: 468. Fain, 1963, Bull. Inst. Roy. Sci. Nat. Belg., 39(32): 86. This species was briefly described from several Caliithrixjacchus which had died in captivity (Fain, 1 962). Fain ( 1 963b) provided an illustration of the male, and Fain (1963c) gave a complete descrip- tion with figures of both male and female. The hosts, which were listed as eight Hapale jacchus (= Callithrix jacchus) imported from the "bassin de I'Amazone," died in Antwerp. Additional sjiec- imens were recorded from a specimen of Hapale jacchus leucocephalus (= Callithrix jacchus geof- froyi) collected in Bahia, Brazil, and preserved in Brussels. Material Examined— None. Fonsecalges saimirii Fain, 1963 Fonsecalges saimirii Fain, 1963, Bull. Ann. Soc. Roy. Ent. Belg., 99: 330. Fain, 1963, Bull. Inst. Roy. Sci. Nat. Belg., 39(32): 90. Fain, 1966, Acarologia, 8: 107. Dunnalges lambrechti Lavoipierre, 1 964, Acarologia, 6: 343. This species was briefly described from numer- ous specimens collected from two Saimiri sciureus from "Amerique du Sud" and which died in the Antwerp Zoo in 1959 and 1963 (Fain, 1963a). Fain (1963c) provided further descriptive infor- mation and indicated the original hosts had come from "Amazonie" and had died shortly after their arrival at the Antwerp Zoo. He also mentioned additional specimens from "Tamarins spp. ori- ginaires d' Amazonie." The inexact locality infor- mation leaves the actual identity of the type host in doubt following the revision of the genus Sai- miri by Hershkovitz (1984). Lavoipierre (1964a) described and illustrated this species as Dunnalges lambrechti, from "a long series of sjiecimens com- prising all stages" from several Tamarinus nigri- collis (= Saguinus nigricollis) from eastern Peru which died in San Francisco, California. Fain ( 1 966) illustrated parts of this species from the type specimens. Fain (1963c) did not observe sexual dimor- phism in the nymphs of this species, as only two tritonymphs and some shed tritonymphal cuticles were examined. Sexual dimorphism was noted in the tritonymphs during the present study. All lar- vae and protonymphs examined bear three pairs of sclerotized apophyses in the ventrolateral re- gion: a rounded apophysis immediately posterior to trochanter II, a pointed apophysis lateral to the base of leg III, and a pointed apophysis poster- iolaterally near the posterior lobes. Tritonymphs in which these apophyses are retained are here interpreted as males, while those in which all apophyses are lost are interpreted as females. Material Examined— Total 59. Three females from Saimiri sciureus macrodon. PERU, Loreto: Nauta, Rio Tigre, 5 km above Rio Tigrillo; 17 December 1980; P. Hershkovitz (9257). Host now 248 FIELDIANA: ZOOLOGY a tanned skin (fmnh 122810). Mites labeled bmoc 81-0809-15. Two females, 1 male from same host species and locality; P. Hershkovitz (9258, fmnh 122811, BMOC 8 1 -0809- 1 6). Twenty-four females, 8 males, 5 female tritonymphs, 8 male trito- nymphs, 5 protonymphs, 1 larva from same host species. PERU, Loreto: Nauta, Rio Tigre, 6 km above Rio Tigrillo; same date; P. Hershkovitz (9268, FMNH 1 228 1 6, bmoc 8 1 -0809- 1 8). Two fe- males from same host sjjecies and locality; P. Hershkovitz (9267, fmnh 1 228 1 5, bmoc 8 1 -0809- 19). Procebalges pitheciae Fain, 1963 Procebalges pitheciae Fain, 1963, Bull. Ann. Soc. Roy. Ent. Belg., 99: 332. Fain, 1963, Bull. Inst. Roy. Sci. Nat. Belg., 39(32): 96. This species was briefly described from speci- mens collected from a Pithecia monachus im- ported from "Amerique du Sud" and which died in the Antwerp Zoo (Fain, 1963a). The descrip- tions were completed and illustrations provided later (Fain, 1963c). Material Examined— Total 242. Nineteen fe- males, 28 males (7 in tandem with female trito- nymphs), 15 female tritonymphs, 1 protonymph from Pithecia hirsuta. PERU, Loreto: Nauta, Rio Samiria; 30 November 1980; P. Hershkovitz (9 1 1 5). Host now preserved as a tanned skin (fmnh 122797). Mites labeled bmoc 81-0809-7. Forty- seven females, 8 1 males ( 1 7 in tandem with female tritonymphs), 22 female tritonymphs, 8 male tri- tonymphs, 13 protonymphs, 8 larvae from same host and locality; P. Hershkovitz (9088, fmnh 122796, BMOC 81-0809-9). Family AUDYCOPTIDAE Three species of hair follicle inhabiting mites of the family Audycoptidae have been previously de- scribed from New World primates, all from squir- rel monkeys identified as Saimiri sciureus. With the recent recognition of several valid species in the genus Saimiri (Hershkovitz, 1984), the exact identification of the reported hosts is problemat- ical. Specimens representing a fourth sr>ecies were recovered from Cebus apella from the present col- lections. Audycoptes greeri Lavoipierre, 1 964 Attdycoptes greeri Lavoipierre, 1 964, Ann. Natal Mus., 16: 194. This species was described from females col- lected from the sinus-hair follicles of Saimiri sci- ureus collected in eastern Peru and kept in captiv- ity in California (Lavoipierre, 1964b). In the absence of more detailed collection information, the true host may have been either S. sciureus or S. boliviensis, as both occur in eastern Peru (Hersh- kovitz, 1984). Material Examined— Five females from "5fl/- muri sciurea" (sic), without further collection data, from the Lavoipierre collection. University of Cal- ifornia, Davis. Audycoptes lawrencei Lavoipierre, 1964 Audycoptes lawrencei Lavoipierre, 1964, Ann. Natal Mus., 16: 199. This species was described from the same hosts and habitat as Audycoptes greeri (Lavoipierre, 1964b), so the actual specific identity of the host remains uncertain, as indicated for greeri. Material Examined— Two females from "^a/- muri sciurea"" (sic), without further collection data, from the Lavoipierre collection, University of Cal- ifornia, Davis. Saimirioptes paradoxus Fain, 1968 Saimirioptes paradoxus Fain, 1968, Acarologia, 10: 286. This species was described from a single female, containing a larva, collected from a Saimiri sci- ureus which had died in the Antwerp Zoo (Fain, 1968). No information concerning the geographic origin of the host was given, making the specific identity of the host uncertain. Material Examined— None. Saimirioptes hershkovitzi, new species In the following description, all measurements are given in micrometers and are presented as ho- lotype (range of three measured specimens). Female (figs. 1-2)— Body elongate, cylindrical OCONNOR: MITE PARASITES OF NEW WORLD PRIMATES 249 Fig. 1 . Saimirioptes hershkovitzi, female. A, dorsum; b, venter. to somewhat flattened dorsoventrally; length in- cluding gnathosoma 4 1 5 (4 1 5—42 1 ), width at level of transverse coxal apodemes III 146 (135-146); entire body with transverse striations which are very thick anteriorly, thin posteriorly. Dorsum (fig. la)— Prodorsal sclerite narrow an- teriorly, much widened medially, narrowed pos- teriorly and fused internally with apodemes of cox- al fields II in most posterior region. Sclerite traversed by anterior transverse striations later- ally. Length of sclerite 70 (70-77). Paired dorsal protuberances present mesal to setae h and sh. with striations encircling protuberances along most of their length; lengths 29 (28-29). Idiosomal chae- totaxy as follows: scapular setae filiform, in a transverse line posterior to prodorsal sclerite, lengths sci 4 (4-5), see 39 (29-39); setae J, and /, very short and rounded, subdivided into two dis- tinct lobes; setae h and sh filiform, lengths 9 (6- 9); setae di subdivided into three parts, a ventral filiform part, length 26 (23-26), and two dorsal rounded lobes; setae d^ similar to ^2 but filiform part shorter, length 19 (14-19), and middle lobe more elongate; remaining setae filiform; /j 29 (28- 250 FIELDIANA: ZOOLOGY < > OCONNOR: MITE PARASITES OF NEW WORLD PRIMATES 251 29); /, ventrally positioned, 18 (18-19); ^^4 24 (19- 24); ^5 21 (no variation); l^ ventral, 5 (4-5); l^ ventral, 30 (27-32); a, ventral, 14 (no variation); a, 30 (27-30). Four pairs of idiosomal cupules present: ia dorsal between setae h and sh; im dorsal between and lateral to setae d2 and d^; ip ventral slightly anterior to seta /j; ih ventral, between and mesal to setae U and /j. Opisthosomal glands pres- ent, small, opening between and lateral to setae d^ and /,. Anus terminal. Opening to bursa copulatrix terminal, above anal opening. Venter (fig. lb)— Coxal apodemes well devel- oped. Anterior apodemes of coxal fields I straight, length 66 (57-66), fused medially to form a ster- num of length 23 (23-24). Anterior apodemes of coxal fields II curving mesally, length 75 (75-79); posterior apodemes of coxal fields II transverse, fused ventromedially with anterior apodemes of coxal fields III, extending dorsally, length of ven- tral portion 50 (46-50), length of dorsal portion 31 (28-31). Anterior apodemes of coxal fields III extending almost longitudinally from trochanters III to F>osterior apodemes of coxal fields II, then bending at right angles to fuse with the latter, length of longitudinal portion 49 (49-51), transverse portion 39 (39-44). Anterior apodemes of coxal fields IV very wide, anteriomedially directed, length 50 (50-53). Ovipore located between coxal fields III and IV; a crescentic epigynal apodeme posi- tioned anterior to ovipore; genital valve elongate, disappearing posteriorly under transverse flap; a pair of small thickenings along posterior, internal portion of valves; two pairs of vestigial genital papillae lateral to genital ojiening. Three pairs of filiform coxal setae present: cxI at posterior end of coxal fields I between coxal apodemes II, length 12 (9-12); cxIII on median apex of coxal apo- demes IV, length 9 (6-9); cx/P'mesad of the base of leg IV, length 7 (6-7). Genital setae absent. Gnathosoma (fig. 2g)— Gnathosoma elongate, somewhat widened posteriorly, then tapering to a point ventrally, length (excluding chelicerae) 62 (59-62), maximum width 26 (24-26). Chelicerae elongate, with small, toothed digits, length 46 (44- 46). Palps apparently two-segmented; basal seg- ment bearing two very short, filiform setae, 1 prox- imodorsal, 1 medioventral; distal segment with short solenidion and 5 sclerotized, pxjinted pro- jections. Subcapitulum with pair of short, filiform setae medioventrally; with pair of sclerotized, pointed projections positioned slightly anterior to subcapitular setae, and unpaired projection ante- riomedially; median rutellar lobes with sclerotized points medially. Legs (figs. 2a-f) — Legs I-II similar in structure, with all segments free, lengths 84 (80-90). Femora I-II bearing two ventral apophyses, one filiform seta vF. lengths: I, 9 (8-9); II, 33 (30-33). Genua I-II each with large ventral apophysis; setae cG expanded basally, then tapering, lengths 22 (20- 25), setae mG filiform, lengths 43 (40-45). Tibiae I-II each with large ventral apophysis; setae gT short, filiform, lengths 6 (5-6); solenidia 0 api- codorsal, lengths: I, 20 (18-20); II, 22 (21-23). Tarsi I-II each with pointed dorsal apophysis, two distal hooked apophyses; chaetotaxy of tarsi I-II similar; setae d as strongly hooked, apical claws, lengths 5 (no variation); e filiform, lengths 40 (43- 50); /filiform, lengths 7 (6-7); la filiform, lengths 6 (5-6); wa very short, not measurable; ra filiform, lengths 6 (5-6); solenidia «, blunt apically, lengths: I, 14 (no variation); II, 26 (24-26). Pretarsi I-II similar, total lengths 40 (37-40); pretarsi divided into long ambulacral stalk and rounded ambula- cral disc; ambulacral disc with condylophore guide and ventral "rays" (fig. 2b); condylophores atten- uate, appearing to divide at base of ambulacral disc. Legs III-IV similar, each with fused trochanter- femur bearing a large, ventral apwaphysis. Genua III-IV glabrous. Tibiae III-IV each with setae kT enlarged, bifurcate, and deeply rooted; tibia III with solenidion 0, length 6 (6-7), 0 IV absent. Tarsi III-IV each with apical, pointed apophysis, three setae: fi^ apical, lengths: III, 100(91-100); IV, 97 (86-97); r filiform, lengths 7 (6-7); w filiform, lengths 3 (no measurable variation). Types— Holotype and two paratype females from Cebus apella. PERU, Loreto: Nauta, Rio Samiria; 18 November 1980; P. Hershkovitz (9049). Host now a tanned skin (fmnh 122792). Mites labeled BMOC 81-0811-2. Holotype and 1 paratype de- posited in Field Museum of Natural History, Chi- cago; 1 paratype in Museum of Zoology, Univer- sity of Michigan, Ann Arbor. Systematic Position— Saimirioptes hershko- vitzi shares with 5". paradoxus, the tyF>e-sp)ecies and only other member of the genus, the presence of dorsal lobes between setae ^1 and t/j^ the similar form of setae t/, and /, (rounded, bifurcate, and without filiform part), and the expansion of the bases of setae cG of genua I-II. The new sF>ecies differs from the type-species in the relative lengths of the scapular setae (sce:sci 5-6:1 in S. hershko- vitzi, 1.5:1 in S. paradoxus); the shorter filiform fKJrtion of setae f/, (not reaching dy in S. hersh- kovitzi, extending past d^ in S. paradoxus); the shorter lengths of setae 4 (29 in S. hershkovitzi. 252 HELDIANA: ZOOLOGY 67 in S. paradoxus); the presence of setae U (absent in S. paradoxus); the longer pretarsi I-II (37-40 in S. hershkovitzi, 17 in S. paradoxus); and the greater lengths of setae d on tarsi III-IV (at least twice as long as the entire leg in S. hershkovitzi, less than half the leg length in S. paradoxus). Discussion The Peruvian collections reported upon here support the hypothesis that most of the records of primate-Cebalginae associations previously re- ported reflect natural host-parasite associations. The association between Alouattalges corbeti and Aotus species was not verified and remains ques- tionable. As there were no field collected repre- sentatives of the Callitrichidae examined during this study, the occurrence of cebalgine mites on these hosts remains generally untested in natural situations, with only Fonsecalges johnjadini re- corded from noncapti ve callitrichids (Fain, 1 963c). Having established that most records of pri- mate-cebalgine associations reflect natural asso- ciations, two additional hypotheses must be pro- posed and tested before any coevolutionary hypotheses may be tested. These preliminary hy- potheses concern the monophyly of the Cebalginae and the phylogenetic relationships of the taxa within the group. I have previously discussed these questions (OConnor, 1 984) but, due to space con- straints, was unable to detail the reasoning behind my conclusions. Monophyly of the Cebalginae Before any hypothesis of historical associations may be tested, at least one of the lineages must be hypothesized to be monophyletic (Brooks, 1981). In this study, the psoroptid subfamily Cebalginae, which comprises all psoroptid mites parasitizing New World Primates, must be tested for mono- phyly. In last defining this subfamily. Fain ( 1 963c) listed 21 character states for the group. Because the comprehensive morphological studies of Fain (1963c) were carried out before the methods of phylogenetic systematics became widely discussed and utilized, no distinction between ancestral and derived states was made in the diagnosis of the Cebalginae. In order to test whether the Cebalgi- nae represents a monophyletic group, I have ex- amined the characters listed by Fain (1963c) using outgroup comparison to polarize the states. I have previously hypothesized that the taxa comprising the four subfamilies of Psoroptidae which para- sitize the Primates form a monophyletic group (OConnor, 1984). These taxa include the Makial- ginae (sensu OConnor, 1984; i.e., including the Cheirogalalginae and Galagalgidae of Fain), par- asites of the Strepsirrhini; the Paracoroptinae, par- asites of African Cercopithecidae and Hominidae; the Nasalialginae, parasites of Asian Cercopithe- cidae; and the Cebalginae, from New World pri- mates. This grouping of taxa is regarded as the ingroup in the following analysis. Outgroups used in defining the character state polarities were the other subfamilies of Psoroptidae and earlier de- rivative groups in the Astigmata. Character Analysis 1 . Presence of retrograde apophyses on coxa! fields III. This state is unique to the cebalgine genera, with such apophyses not occurring in other taxa in the ingroup or outgroup. 2. Male body size much smaller than female. In the outgroups and other ingroups, males are similar in size to females or somewhat larger or smaller. The substantial reduction in body size of male Cebalginae is unique. 3. Female opisthosoma more or less squared posteriorly and laterally, reinforced with sclero- tized areas. This condition is exhibited by all Ce- balginae, but not in any outgroup. Among the in- groups, the squared body is unique to the Cebalginae, but sclerotized reinforcement is also found in the Nasalialginae and some Paracorop- tinae (Pangorillalges). Among outgroup taxa, sclerotized reinforcement is present in some Pso- ralginae (Edentalges), but again without the squared body form. 4. Male with legs III very modified, with 3-4 terminal segments fused and bearing medially di- rected projections. The modification of the third pair of legs in male Cebalginae is unique and pres- ent in all taxa. No similar modification occurs anywhere in ingroup or outgroup. 5. Reduction or loss of paranal suckers in males. Paranal suckers are present in males in the out- groups and ingroups. In the Cebalginae, the suck- ers are very reduced or absent. 6. Tarsus I with 2 apical solenidia. I regard the apical displacement of solenidion oj, in the Ce- balginae as derived. In the other ingroups and most outgroups, this solenidion is median or basal on OCONNOR: MITE PARASITES OF NEW WORLD PRIMATES 253 the tarsus. Apical displacement of this solenidion also occurs in the subfamily Psoroptinae and the monobasic Marsupialginae, conditions I regard as convei^ent. 7. Loss of dorsal seta ^,. Seta d^ is absent in all Cebalginae and retained in the other ingroups and most outgroups. This seta is also lost in the Lis- tropsoralginae and the psoroptid parasites of ro- dents (Echimyalges. Myoproctalges, and Coen- dalges), conditions I regard as convergent. 8. Reduction or loss of apophyses from the pos- terior tarsi of the female. Apophyses are present on the posterior tarsi of some of the outgroups such as the Listropsoralginae and are retained in some ingroups, the Makialginae and Paracorop- tinae. I consider the presence of these apophyses to be plesiomorphic for the primate-associated psoroptid lineage. In the Cebalginae, taxa in which females have well-developed posterior legs retain vestiges of these apophyses, while taxa in which the legs are reduced retain no traces. I consider this reduction to be a derived state for the Cebal- ginae and regard the loss of these structures in the Cebalginae and Nasalialginae as convergent. 9. Posterior edge of female opisthosoma with 2 pairs of long, strong setae. In most ingroups and psoroptid outgroups, seta /, is long and strong while seta di is shorter and thinner. Of the ingroups, setae l^ and d^ are equally well developed in the Cebalginae and Nasalialginae. However, in the Ce- balginae, setae /, and d^ are closely associated, usu- ally on a single projection, while in the Nasalial- ginae these setae are separate, on distinctly different projections. I regard these two conditions as con- vergent. I also regard as convergence the elonga- tion of seta di in some Psoralginae. 10. Dorsal seta l^ sometimes absent. Seta I4 is absent in all Cebalginae except Cebalgoides cebi. The seta is retained in that species, the other in- groups, and most outgroups. The loss of seta I4 could be regarded as a synapomorphy defining a group containing all Cebalginae except Cebal- goides. However, this hypothesis conflicts with groupings suggested by all other characters (see page 256). At this point, it is more parsimonious to regard the loss of seta U as a synapomorphy for the Cebalginae, with a reversal in Cebalgoides. 1 1 . Loss of retrograde apophyses on the ante- rior legs. Retrograde apophyses are present on the anterior legs in some of the outgroups, notably the Listropsoralginae and Marsupialginae in the Pso- roptidae and in the related families Audycoptidae, Rhyncoptidae, and Myocoptidae. Among the in- groups, these apophyses are retained in the Ma- kialginae but lost in the Cebalginae, Paracorop- tinae, and Nasalialginae. I regard this loss as characterizing these latter three groups as a natural unit and thus plesiomorphic for the Cebalginae. 12. Female bursa copulatrix subterminal or ventral. In the outgroups, the female bursa cop- ulatrix is terminal, as it is in the Cebalginae, Ma- kialginae, and Nasalialginae among ingroups. I re- gard the dorsal position of the bursa in the Paracoroptinae as derived. Thus, this character retains the plesiomorphic condition in the Cebal- ginae. 13. Base of the gnathosoma with retrograde apophyses. In some outgroups as well as in the Cebalginae, Makialginae and some Paracordpti- nae (i.e., Pangorillalges), the base of the gnatho- soma bears retrograde apophyses. These are ab- sent in certain outgroups (e.g., the Psoroptinae) and in Paracoroptes and Nasalialges among in- group taxa. I regard the retention of these apoph- yses in the Cebalginae as plesiomorphic. 14. Absence of retrograde apophyses on coxae I-II. Retrograde apophyses occur on coxae I-II only in the Makialginae. They do not occur on both coxal fields in any other ingroup or outgroup taxa, although the Listropsoralginae and Myocop- tidae have apophyses on coxae II. I regard, these structures as synapomorphies characterizing the Makialginae, and their absence in the Cebalginae as plesiomorphic. 15. Well-developed "claws" on tarsi I-II. Clawlike apophyses on the apices of tarsi I-II are present in many outgroup taxa, including both mammal and bird parasites. I regard the presence of this character state as ancestrally characterizing all Psoroptidia, and thus plesiomorphic for the Cebalginae. 16. Tarsus II with solenidion apically dis- placed. Solenidion oj is apically displaced in the Cebalginae, Paracoroptinae, and Nasalialginae. In most outgroups and in the Makialginae, the so- lenidion retains its ancestral, basal position. I re- gard this displacement as diagnosing a natural group consisting of the above three primate-as- sociated subfamilies and thus plesiomorphic for the Cebalginae. 17. Female tarsal chaetotaxy 7-7-6—4 or -5. Possession of 7 setae on tarsi I-II involves the loss of seta ba in the Paracoroptinae and Cebalginae and probably the Nasalialginae. Retention of 8 setae on these segments (including ba) is charac- teristic for most outgroup taxa and the Makial- ginae. I regard this state as diagnosing a natural group comprising the Cebalginae, Paracoroptinae, 254 HELDIANA: ZOOLOGY and Nasalialginae and thus plesiomorphic for the Cebalginae. 1 8. Angles of female opisthosoma generally with strong hooks. Such hooks are found in the cebal- gine genera Cebalges, Cebalgoides. and Fonse- calges, but not in other taxa in the outgroup or ingroup. I regard the presence of these structures as characterizing a monophyletic unit within the Cebalginae but not the group as a whole. 19. Posterior legs of female normal or atro- phied. Legs III-IV of the female are reduced in size and may exhibit fusion of segments in the genera Cebalges and Fonsecalges but not in other Cebalginae nor other members of the ingroup. Well-developed legs are characteristic of most out- group taxa, although similar reduction of the pos- terior legs occurs in some Psoroptinae and Pso- ralginae. I regard the atrophied legs of some Cebalginae as characterizing a smaller monophy- letic unit within the group and convergent with the outgroup taxa noted above. 20. Posterior legs of nymphal stages normal or short and atrophied and bearing a long seta. The reduction of the legs in the nymphs is found in all cebalgine genera except Procebalges. It is also found in a few outgroup taxa (e.g., the Psoroptinae and Psoralginae), but not in most outgroups or in other ingroup taxa. I regard this state as diagnosing a group within the Cebalginae and not characteriz- ing the group as a whole. 2 1 . Dorsal seta d^ sometimes absent. Seta d^ is absent in Schizopodalges lagothricola and Fon- secalges johnjadini (but not F. saimirii), and pres- ent in other Cebalginae, other ingroups, and most outgroups. The distribution of this derived state within the Cebalginae strongly suggests the inde- pendent loss of this seta in the two species sharing the state. This character at best diagnoses a group within the Cebalginae, but is more likely a case of convergence. In no way does this character diag- nose the group as a whole. Conclusions On the basis of the above analysis, I conclude that, of the 21 character-states listed by Fain (1963c) as diagnosing the Cebalginae, five repre- sent unique synapomorphies which diagnose the Cebalginae and occur in no other group (character- states 1-5), five are synapomorphies which diag- nose the Cebalginae but which also occur as con- vergent states in certain other taxa (character-states 6-10), seven are symplesiomorphies diagnosing larger groups which include the Cebalginae (char- acter-states 11-17), and four represent within- group apomorphies diagnosing smaller groups within the Cebalginae (character-states 1 8-2 1 ). The ten synapomorphies diagnosing the Cebalginae leave no doubt that the group is a natural one. Phylogenetic Relationships Within the Cebalginae I have previously presented a hypothesis of phy- logenetic relationships among genera in the Ce- balginae based upon 17 characters (OConnor, 1 984). This cladogram is reproduced here (fig. 3), with numbers on the cladogram referring to the derived states of the characters listed below. In the earlier study, space limitations prevented discus- sion of the different states of these characters and the reasons for interpreting their polarity. Addi- tionally, an error in character 1 2 appeared in the list of character-states. As I have hypothesized that the Cebalginae form a monophyletic group within a larger lineage comprising the Makialginae, Par- acoroptinae, and Nasalialginae, taxa in these three lineages were used as outgroups to polarize the character-states within the Cebalginae. 1. Female with coxal apodemes III-IV fused. In all taxa in the outgroups, coxal apodemes III- IV end freely. Within the Cebalginae, these apo- demes are fused together on either side in the gen- era Procebalges, Schizopodalges, and Fonsecalges. Although I regard the fused condition to be de- rived, conflicts with many other characters suggest that these apodemes have fused independently in the three genera. 2. Female epigynum fused with coxal apo- demes I. In most outgroup taxa, the female ovi- pore is located between coxal fields II and III. There has been a trend toward anterior displace- ment of the ovip)ore in a number of psoroptid lineages (e.g., Psoralginae, Listropsoralginae, Pso- roptinae), and the more anterior p>osition occurs in certain taxa among the primate associated lin- eage as well (e.g., Lemuralges in the Makialginae). Within the Cebalginae, this derived state occurs in the genera Procebalges and Schizopodalges. Be- cause this character-state distribution conflicts with many other characters, I regard the presence of this state in the two genera as convergence. 3. Male with paranal suckers lost. In all out- groups, the male retains paranal suckers. Within the Cebalginae, the suckers are retained only in OCONNOR: MITE PARASITES OF NEW WORLD PRIMATES 2SS V) O O O) ,^*— i...^. D >< o l« a -c o o N OD -C D o ~~j oo . ^ VI -2 < «/> 3 C 3 D5 D 00 O ^ O^ to 13 D -Q -Q Q) 0) U O 0) ^ — ■ D) • "^ C o o 00 u^ — ^ ■ 17 ■ 16 ■ 1 Fig. 3. Phylogenetic relationships among genera in the subfamily Cebalginae. Numbers refer to derived states of characters discussed in the text. Hosts of each mite taxon given in parentheses. Procebalges. I regard the loss of these structures as derived and characterizing the sister group of Procebalges. 4. Immature stages with pretarsi III-IV lost. Described immature stages of outgroup taxa all retain the pretarsus on legs III-IV. Within the Ce- balginae, only Procebalges retains this ancestral state. In all other Cebalginae, pretarsi III-IV are lost in the immature stages, a condition I regard as derived. 5. Gnathosomal apophyses lost in both sexes. As discussed in the previous section, I regard the presence of gnathosomal apophyses as ancestral in the Cebalginae. These structures are lost in both sexes in the genera Schizopodalges and Alouat- talges within the Cebalginae, a condition I regard as derived. I regard this state as a synapomorphy for these two genera and convergent with the sim- ilar state occurring in the genera Lemuralges (Makialginae), Paracoroptes (Paracoroptinae), and Nasalialges (Nasalialginae). I have treated loss of these apophyses in the female only as a separate character (16, below). 6. Female with seta 5 of tarsi I-II in the form of a strongly hooked claw. In the outgroup taxa, seta s is simple and filiform. Within the Cebalgi- nae, this state is retained in all taxa except Schi- zopodalges and Alouattalges where the seta is en- larged and hooked. I regard the latter state as derived and a synapomorphy for the two genera. 7. Male with a large spur on tarsus III. This condition is unique to the cebalgine genera Schi- 256 FIELDIANA: ZOOLOGY zopodalges and Alouattalges. In the outgroups and other ingroup taxa, such a spur does not exist. I regard the presence of this character-state as in- dicative of common ancestry of these two genera. 8. Male with ridges on leg III restricted to tibial element. Ridges on the distal portion of leg III of the male characterize the Cebalginae (see above). Such ridges do not occur in any outgroup. I regard the restriction of the ridges to a spur extending from the tibial portion of the leg as a derived state defining the group Schizopodalges + Alouattalges because this condition forms a functional complex with the spur on tarsus III (character 7) and a dorsal spur on femur IV. This latter character was not considered in the previous analysis (OConnor, 1984). The conjunction of the three spurs serves to lock legs III and IV together into a unit. 9. Female with recurved hooks on posterior border of opisthosoma. Strong hooks are present in this position in females of Cebalgoides, Ce- balges. and Fonsecalges. Weaker development in this area is characteristic of other cebalgine genera, while in the outgroups, no projections exist. I re- gard the possession of strong hooks as a derived condition defining a lineage comprising the three genera noted above. 10. Male with pretarsus III lost. Males retain a pretarsus on leg III in most outgroup taxa. This pretarsus is lost in the genera Galagalges and Chei- rogalalges (Makialginae), and also in the cebalgine genera Cebalgoides, Cebalges, and Fonsecalges. I regard the loss of pretarsus III as a synapomorphy for the latter three genera within the Cebalginae, with independent loss in the makialgine lineage comprising the former two genera. 1 1 . Female without opisthosomal sclerite. A median sclerite is present in females of most out- group taxa and is retained in all Cebalginae except Schizopodalges. I regard this loss as derived in the latter genus. Convergence in this character occurs with some outgroup taxa. This sclerotization is also lost in Gaudalges caparti, and Lemuralges (Makialginae), and Nasalialges (Nasalialginae). 12. Female with seta s of tarsi III-IV enlarged and clawlike. In the previous study (OConnor, 1984), a lapsus occurred in that the state "seta 5 of tarsi III-IV reduced" was listed as a derived condition for the genus Alouattalges. In fact, pos- session of a small, filiform seta s on these tarsi must be regarded as the ancestral condition, as it occurs in all outgroup taxa as well as in most Ce- balginae. In the genus Alouattalges, seta 5 is en- larged and recurved on tarsi III-IV, a condition I regard as the true derived state. 13. Male with opisthosomal lobes widely spaced. In most outgroup taxa, the opisthosomal lobes of the male are close together, a condition also found in most Cebalginae. These lobes are widely spaced in the genus Cebalgoides within the Cebalginae, a condition I regard as derived. Con- vergence occurs, with this state also occurring in Nasalialges (Nasalialginae). 1 4. Male with apodeme between genital and anal region strongly reduced. In the outgroup taxa, males do not possess a transverse apodeme between the genital and anal regions. A large apodeme is pres- ent in this position in most male Cebalginae. In the prior analysis (OConnor, 1 984), I stated that this apodeme was absent in Cebalgoides as indi- cated by Fain (1963c) and regarded this as a re- versal. Closer examination of a number of speci- mens indicates that a very small apodeme is present in some males of this taxon. I regard the presence of the apodeme as a derived condition for the Cebalginae, and its reduction or loss in Cebal- goides as a further derived state. 1 5. Female with tibiae-tarsi III-IV fused. In the outgroups, the tibia and tarsus of legs III-IV are freely articulated. This condition is present in most Cebalginae as well. In the genera Cebalges and Fonsecalges, these segments are fused in the fe- male, a condition I regard as derived within the Cebalginae. Similar fusions occur in some other psoroptid subfamilies but not within the primate- associated lineage. 16. Female with gnathosomal apophyses lost. Gnathosomal apophyses are retained in females of most outgroup taxa and are retained in some Cebalginae. These apophyses are lost in the female but retained in the male in the genus Fonsecalges. I regard this loss as independent of the loss of apophyses in the Schizopodalges-Alouattalges lin- eage, where the apophyses are lost in both sexes (character 5). 17. Female with pretarsi III-IV lost. Pretarsi III-IV are retained in the female in all outgroup taxa and in all Cebalginae except Fonsecalges. I regard this loss as derived and convergent with the similar loss in females in other psoroptid groups (e.g., some Psoroptinae, Psoralginae). History of Primate-Cebalgine Evolution The present knowledge of the diversity and host associations of the family Audycoptidae is not suf- ficient for formulating hypotheses concerning the OCO^fNOR: MITE PARASITES OF NEW WORLD PRIMATES 257 history of their host associations. However, given the phylogenetic relationships among cebalgine taxa presented in the cladogram (fig. 3), two meth- ods exist for using these relationships to make hy- potheses regarding the history of the associations between the New World primates and the Cebal- ginae. The first method would involve comparing the phylogenetic hypothesis for the parasite group with a similar hypothesis for the hosts. The hy- pothesis tested by this comparison is whether the current pattern of host-parasite associations di- rectly results from strict cospeciation between the hosts and their parasites. Given a phylogenetic hypothesis for each group, such a hypothesis is easily tested. The second type of conclusion which might be drawn from the parasite cladogram is a phylogenetic hypothesis concerning host relation- ships. This method is dependent upon the as- sumption of cospeciation, or at least noncoloni- zation, between the two lineages. The current state of knowledge concerning the phylogenetic relationships among New World pri- mates is somewhat confused. Many early hypoth- eses were based upon classical methodology in which ancestral and derived character-states were not differentiated in proposing hypotheses of re- lationships. Schwartz et al. (1978) summarized the state of knowledge only a few years ago by saying that "Platyrrhini apF>ears in and of itself to be a natural group, although both its wider relation- ships and the relationships among its members remain unclear" (p. 128). Much new information relating to this question was presented in symp>osia in 1978-1979 and published in a volume edited by Ciochon and Chiarelli (1980). In the following discussion, I refer to phylogenetic hypotheses for New World primates suggested by Rosenberger (1977) and contributors to the Ciochon and Chi- arelli volume as representing the most modem phylogenetic thinking regarding Platyrrhine rela- tionships. Returning to the question of cospeciation be- tween primates and cebalgine mites, the first dif- ficulty encountered is the lack of a consensus con- cerning phylogenetic relationships among all New World primates. Phylogenetic hypotheses based on the following types of data sets are in strong disagreement: dentition (Rosenberger, 1977; 3 separate hypotheses); integumentary characters (Perkins & Meyer, 1980); karyology (Chiarelli, 1980); immunological methods (Sarich & Cronin, 1980); and other immunological data and protein sequence data (Baba et al., 1 980). Despite the strong disagreement among these data sets, some patterns are common to several or all of these. These may be compared with the cladogram of mite relation- ships to test the cospeciational hypothesis. One host relationship which is supported by al- most all data sets is the relationship between Al- ouatta and the Atelinae (including Lagothrix). Of the hypotheses cited above, only the karyological evidence did not support this relationship, and then only because the karyotype of Lagothrix is so derived that Chiarelli ( 1 980) made no hypoth- esis as to its relationships. This relationship be- tween Alouatta and Lagothrix is mirrored by the sister group relationship between their parasites, Alouattalges and Schizopodalges, which is among the most strongly supported relationships among the Cebalginae. A second host relationship which is partially mirrored in the mile phylogeny is the relationship between the genera Cebus and Saimiri, which was supported by all data sets except the integumen- tary data (Perkins & Meyer, 1980). In this case, the mite phylogeny suggests a cospeciational pro- cess, but some additional hypotheses are required to explain the distributions of the genera Cebal- goides, Cebalges, and Fonsecalges. The relation- ships among these three parasite genera are com- patible with a cospeciational scenario if an early sjjeciation event between the Cebalgoides and the Cebalges-Fonsecalges lineages occurred in con- junction with a common ancestor of Cebus and Saimiri. Extinction of the Cebalgoides lineage on the Saimiri line, with cospeciation of the Cebalges- Fonsecalges lineage in both host lines, leads to the present distribution on the Cebidae. In all cases, colonization events are required to explain the presence of both of these lineages on the Callitrich- idae, given the probable monophyly of that taxon. Certain other relationships among the parasite taxa are not mirrored by host phylogenies. The sister group relationship between Procebalges and all other cebalgines is not reflected in any hyp)oth- esis of host relationships. The relationship of Pi- thecia to the Alouatta-Lagothrix lineage was sug- gested by both dental data (Rosenberger, 1977) and karyology (Chiarelli, 1 980). Interestingly, the integumentary data (Perkins &. Meyer, 1980) sug- gest that Pithecia retains the most plesiomorphic skin characteristics of any mite-bearing New World primate (Aotus was regarded as even more plesio- morphic). As cebalgine mites are skin inhabitants, it might be suggested that plesiomorphic skin re- tains plesiomorphic mites, a hypothesis which would require the ancestral possession of a Pro- cebalges lineage on other cebids with subsequent 258 HELDIANA: ZOOLOGY extinction on all but Pithecia in order to retain a basically cospeciational history. Discovery of ce- balgine mites on hosts related to Pithecia (Chi- ropotes, Cacajao) will provide a test of these hy- potheses. The utility of using the proposed mite phylogeny to infer a host phylogeny at higher levels than mentioned above depends upon the degree to which cospeciational patterns outweigh colonizations or extinctions in the hypothesized historical relation- ships. Among the entire psoroptid lineage para- sitizing all Primates, cospeciational patterns ap- pear to be supported in a large majority of cases (OConnor, 1984). However, noncosj)eciational patterns such as the distribution of the genus Le- muralges in the Makialginae, or the cebalgine par- asites of the Callitrichidae, make the use of these mites as consistent indicators of host phylogeny at least somewhat suspect. Given the large dis- parity in phylogenetic hypotheses generated from subsets of the overall character matrix for the New World Primates, this parasite data deserves to be at least considered by future workers in this area. Acknowledgments I would like to thank Philip Hershkovitz, with- out whose cooperation and interest this study could not have been attempted. I also thank Robert Timm and Bruce Patterson, Field Museum of Nat- ural History, for their hospitality and cooperation during the processing of host specimens and for their critical review of the manuscript. I thank J. H. S. Klompen, University of Michigan, for his comments on the manuscript. I thank Margaret van Bolt, Museum of Zoology, University of Michigan, for assistance with the illustrations. Literature Cited Baba, M., L. Darga, and M. Goodman. 1980. Bio- chemical evidence on the phylogeny of Anthropoidea, pp. 423-443. In Ciochon, R. L., and A. B. Chiarelli, eds.. Evolutionary Biology of the New World Monkeys and Continental Drift. Plenum Press, New York. Brooks, D. R. 1979. Testing the context and extent of host-parasite coevolution. Systematic Zoology, 28: 299- 307. studying the evolution of ecological associations. An- nals of the Missouri Botanical Garden, 72: 660-681. Brooks, D. R., and D. R. Glen. 1982. Pinworms and Primates: A case study in coevolution. Proceedings of the Helminthological Society of Washington, 49: 76- 85. Chiarelli, A. B. 1980. The karyology of South Amer- ican Primates and their relationship to African and Asian species, pp. 387-398. In Ciochon, R. L., and A. B. Chiarelli, eds.. Evolutionary Biology of the New World Monkeys and Continental Drift. Plenum Press, New York. Ciochon, R. L., and A. B. Chiarelli, eds. 1980. Evo- lutionary Biology of the New World Monkeys and Continental Drift. Plenum Press, New York, 528 pp. Fain, A. 1 962. Diagnoses d'acariens nouveaux. Revue de Zoologie et de Botanique Africaines, 66: 154-162. . 1 963a. Nouveaux acariens psoriques parasites de marsupiaux et de singes sud-americains (Psoralgi- dae: Sarcoptiformes). Bulletin et Annales de Societe Royale d'Entomologie de Belgique, 99: 322-332. . 1 963b. Un nouvel acarien producteur de gale chez un singe sud-americain. Bulletin et Annales de Societe Royale d'Entomologie de Belgique, 99: 467- 470. 1 963c. Les acariens producteurs de gale chez les lemuriens et les singes avec une etude des Psorop- tidae (Sarcoptiformes). Bulletin du Institut Royal des Sciences Naturelles de Belgique, 39(32): 1-125. 1 966. Les acariens producteurs de gale chez les lemuriens et les singes II. Nouvelles observations avec description d'une espece nouvelle. Acarologia, 8: 95- 114. . 1968. Notes sur trois acariens remarquables . 1981. Hennig's parasitological method: A pro- posed solution. Systematic Zoology, 30: 229-249. . 1985. Historical ecology: A new approach to (Sarcoptiformes). Acarologia, 10: 276-291. FuTUYMA, D. J., AND M. Slatkin, EDS. 1983. Cocvo- lution. Sinauer Associates Inc., Sunderland, Mass., 555 pp. Hershkovitz, P. 1977. Living New World Monkeys (Platyrrhini), vol. 1 . University of Chicago Press, Chi- cago, 1117 pp. . 1984. Taxonomy of squirrel monkeys genus Saimiri (Cebidae, Platyrrhini): A preliminary report with description of a hitherto unnamed form. Amer- ican Journal of Primatology, 7: 155-210. Lavoipierre, M. M. J. 1964a. A note on the family Psoralgidae (Acari: Sarcoptiformes) together with a description of two new genera and two new species parasitic on primates. Acarologia, 6: 342-352. . 1964b. A new family of acarines belonging to the Suborder Sarcoptiformes parasitic in the hair fol- licles of Primates. Annals of the Natal Museum, 16: 191-208. NiTECKi, M. H., ED. 1983. Coevolution. University of Chicago Press, Chicago, 392 pp. OConnor, B. M. 1984. Co-evolutionary patterns be- tween astigmatid mites and primates, pp. 186-195. In Griffiths, D. E., and C. E. Bowman, eds., Acarology VI, vol. 1. Ellis Horwood Ltd., Chichester, England. Perkins, E. M., and W. C. Meyer. 1980. The phy- logenetic significance of the skin of primates: Impli- cations for the origin of New World monkeys, pp. OCONNOR: MITE PARASITES OF NEW WORLD PRIMATES 259 331-346. In Ciochon, R. L., and A. B. Chiarelli, eds., R. L., and A. B. Chiarelli, eds., Evolutionary Biology Evolutionary Biology of the New World Monkeys and of the New World Monkeys and Continental Drift. Continental Drift. Plenum Press, New York. Plenum Press, New York. RosENBERGER, A. L. 1977. A'e«o//?r/jc and ceboid phy- Schwartz, J. H., I. Tattersall, and N. Eldredge. logeny. Journal of Human Evolution, 6: 461-481. 1978. Phylogeny and classification of the Primates Sarich, V. M., AND J. E. Cronin. 1 980. South Amer- revisited. Yearbook of Physical Anthropology, 21: 95- ican mammal molecular systematics, evolutionary '33. clocks, and continental drift, pp. 399-422. In Ciochon, 260 HELDIANA: ZOOLOGY Notes on Bolivian Mammals 2. Taxonomy and Distribution of Rice Rats of the Subgenus Oligoryzomys Nancy Olds and Sydney Anderson ABSTRACTS There are at least three kinds of small, long-tailed rice rats of the subgenus Oligoryzomys (genus Oryzomys) in Bolivia. We use the names Oryzomys microtis (including O. fornesi), O. chacoensis, and O. longicaudatus for these taxa. The correctness of these names is less certain and resolution of the nomenclatorial questions awaits study of specimens from outside Bolivia. The ranges of O. chacoensis and O. microtis are at low elevations and probably overlap to some degree geographically. The range of O. longicaudatus is in the highlands. No sample from one locality includes specimens of more than one species; therefore, ecological and microgeographic differences at places where any two of the three species meet are unknown. Further study may reveal other and more cryptic species within Bolivia. No one measurement or other characteristic that we have studied will unequivocally distinguish all adult specimens of any one of the three kinds from Bolivia. Geographic variation probably occurs within as well as beyond Bolivia in at least two of the three species (the most uniform seems to be O. chacoensis), but more material is needed to describe such patterns. We refrain from using subspecies names in consideration of an ignorance of both geographic patterns of variation and the status of available names. The only name in the subgenus with a Bolivian type locality, O. chaparensis, is tentatively considered a synonym of O. microtis. Existen en Bolivia por lo menos tres clases de pequefias ratas arroceras de cola larga del subgenero Oligoryzomys (genero Oryzomys). Nosotros usamos los nombres de Oryzomys mi- crotis (incluyendo O. fornesi), O. chacoensis, y O. longicaudatus para estos taxa. La exactitud de estos nombres es menos cierta y la resolucion de problemas de nomenclature necesita estudio de esp>ecimenes de fuera de Bolivia. Los rangos de distribucion de O. chacoensis y O. microtis estan en bajas elevaciones y probablemente se superponen geograficamente en algun grado. Ninguna muestra de una localidad contiene especimenes de mas de una especie. Diferencias ecologicas y microgeograficas son desconocidas en lugares donde cualquiera de las dos o tres especies se encuentran. Ulterior estudio puede revelar otras y mas cripticas especies en el interior de Bolivia. Ninguna medida u otra caracteristica que nosotros hemos estudiado separara cla- ramente todos los especimenes adultos de Bolivia. Variacion geografica ocurre probablemente dentro asi como fuera de Bolivia en al menos dos de las tres especies (la mas uniforme parece ser O. chacoensis), pero mas material es necesario para describir tales patrones de variacion. Nos abstenemos de usar nombres subespecificos en consideracion de la ignorancia de los patrones de variacion y el estado de los nombres disponibles. El unico nombre en el subgenero con una localidad tipica en Bolivia, O. chaparensis, es tentativamente considerado como O. microtis. From the Department of Mammalogy, American Mu- seum of Natural History, New York, NY 10024. OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 261 Existem no minimo tres tipos de pequenos ratos-de-arroz, de cauda longa, pertencentes ao subgenero Oligoryzomys (genero Oryzomys) na Bolivia. A estes taxa damos os nomes de Ory- zomys microtis (incluindo O. fornesi), O. chacoensis, e O. longicaudatus. A precisao destes nomes e incerta, e resolu^oes da nomenclatura aguardam estudos de especimes nao bolivianos. O. chacoensis e O. microtis ocorrem em eleva?6es baixas, e as duas esjiecies provavelmenle coocorrem em algumas areas. Oryzomys longicaudatus ocorre em areas montanhosas. Amostras de um so local nao incluem mais do que uma esF)ecie, e nao se conhecem diferen9as ecologicas ou microgeograficas em areas onde duas ou mais especies possam coocorrer. Com futuros estudos, novas especies, mais ocultas, poderao ser encontradas na Bolivia. Nenhum unico carater que estudamos pode inequivocamente separar especimes adultos dos tres tipos de Oligoryzomys na Bolivia. Varia96es geograficas provavelmente ocorrem na Bolivia, bem como em outras areas, mas mais material e necessario para poder documentar quaisquer padroes geograficos (O. chacoensis parece ser a esp>ecie mais uniforme em aparencia). Dada a falta de conhecimento, ambos dos padroes geograficos, e da disponibilidade dos nomes, nao usamos nomes de subespecies. O unico subgenero com localidade de tipo na Bolivia, O. chaparensis, e aqui lentativamente considerado como sinonimo de O. microtis. Introduction Bolivian landscapes range from less than 300 m to more than 4000 m, and habitats range from the humid lowland Amazonian tropical forests and subtropical savannahs to the high barren plains and snow-capped peaks of the Andean altiplano (fig. 1 ). Habitats may change abruptly, often within only a few kilometers. The mammals of Bolivia are also diverse, and provide excellent opportu- nities for ecological and taxonomic studies of broad scope. However, the animals are poorly known (Mares & Genoways, 1982); before satisfactory general conclusions can be reached, the Bolivian species need to be clearly delimited, both mor- phologically and geographically. Mice of the genus Oryzomys occur throughout South America and are important members of small mammal communities (Myers & Carleton, 1981; Mares et al., 1981; Alho, 1982; O'Connell, 1982;Streilein, 1982a-c; Viega-Borgeaud, 1982). We examined critically one subgenus of Oryzo- mys, Oligoryzomys, in Bolivia to determine how many species are present and where they occur. For a general description of the subgenus, see Myers and Carleton (1981, pp. 9-12). The subgenus Oligoryzomys needs revision. There are few published studies of the more than 45 named forms (Tate, 1932; Ellerman, 1941; Ca- brera, 1961; Myers & Carleton, 1981; Honacki et al., 1982). Myers and Carleton (1981) studied Oligoryzomys from Paraguay, where they recog- nized three species: Oryzomys nigripes. O. cha- coensis, and O. fornesi. They also clarified no- menclatorial questions relating to the 'name Oryzomys nigripes. Since Paraguay borders Boliv- ia, this recent study was used as our starting point. We used the same measurements and comparable analyses. We assumed that O. chacoensis and O. fornesi occurred also in southeastern Bolivia, near the Paraguayan border. If O. nigripes occurs in Bolivia, the most probable place for it is in eastern Santa Cruz, from which no specimens are now available. At least six names have been used in the liter- ature or in museum collections for Bolivian sjdcc- imens of the subgenus Oligoryzomys: Oryzomys longicaudatus and O. stolzmanni (of the high- lands); O. chaparensis (from the lowlands of Co- chabamba); O. nigripes (used for all forms); O. delicatus (used for a few specimens at middle el- evations); and 0./7ave5ce/ts (lowland). These names have been used with considerable uncertainty in the past (see summaries in Tate, 1932, and espe- cially Myers & Carleton, 1981). One cause of this problem is that the original descriptions are vague or apply equally well to more than one species of Oligoryzomys. The taxonomic confusion can be resolved by detailed study of adequate numbers of museum specimens, and the nomenclatorial confusion re- solved by comparisons with type specimens. Re- newed interest in South American mammals has resulted in more specimens, which will help in these tasks. 262 FIELDIANA: ZCX)LOGY Brasi Chile Argentina V _ - II Fig. 1 . Map of Bolivia showing the departments and the 500- and 3000-m contour lines. Be, Beni; Ch, Chuquisaca; Co, Cochabamba; LP, La Paz; Or, Onaro; Pa, Pando; Po, Potosi; SC, Santa Cniz; Ta, Tarija. OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 263 Methods We examined specimens in the collections of the American Museum of Natural History (AMNH); Academy of Natural Sciences in Phila- delphia (ANSP); California Academy of Sciences (CAL): Field Museum of Natural History (FMNH); Museum of Vertebrate Zoology, University of Cal- ifornia (MVZ); Museum of Zoology, University of Michigan (UMMZ); and United States National Museum of Natural History (USNM). Measurements were obtained as follows: exter- nal measurements are from the original labels or field notes, when available, or remeasured (fluid- preserved specimens only). The following cranial measurements were taken to the nearest 0.0 1 mm on a craniometer (see Anderson, 1 968) following Myers and Carleton (1981) and Musser (1979): ZN = depth of zygomatic notch LR = length of rostrum from tip of nasals to posterior edge of zygomatic notch GLS = greatest length of skull ZB = zygomatic breadth BB = breadth of braincase BIC = breadth of interorbital constriction LIF = length of incisive foramen LPB = length of palatal bridge LMl, LM2, LM3 = crown lengths of upF>er mo- lars WMl, WM2, WM3 = crown widths of upjjer molars MM = greatest breadth across molars (labial edges) LMX = crown length of upp>er toothrow LNP = length of nasal projection LD = length of diastema LB = length of bulla HB = height of bulla TL = total length, including tail T = tail length HP = length of hind foot, including claw E = length of ear, from notch Specimens were sorted by age, using dental cri- teria as outlined here (following Myers & Carleton, 1981): Age Class I: M' not erupted or newly erupted, M- unworn. Age Class II: M^ slightly to moderately worn, but not flat; M- slightly worn; enamel island formed by the isolation of the internal part of the mesoflexus of M-. Age Class III: M ' flat or slightly concave; enam- el island of above well isolated; M' and M- substantially worn. Age Class IV: M^ concave; enamel island oblit- erated; teeth well worn, but main cusps still discernible. Age Class V: M ' and M- flat or concave; folding pattern obliterated. Individuals in age class I were examined, mea- sured, and included in mapping geographic ranges, but were excluded from statistical treatment of character variation. Sp)ecimens examined are listed by locality and museum catalogue number in the Appendix. Lo- calities are plotted in Figure 2. Statistical analyses were done using the computer facilities of the City University of New York (CUNY) and programs from SAS (Statistical Analysis System) Institute, Inc. (1982). Taxonomy Our knowledge of geographic variation and the status of some of the names is sketchy at best, so we do not use subspecific names. The name Ory- zomys (Oligoryzomys) longicaudatus stolzmanni was first used by Hershkovitz (1940, p. 81), by inference for Bolivian populations, but we are not certain that O. stolzmanni (Thomas, 1894; type locality Huambo, 3700 ft, department of Ama- zonas, Peru) and O. longicaudatus (type locality restricted to Valparaiso, Chile) are conspecific or that the Bolivian specimens are conspecific with either. The resolution of these problems awaits further study beyond Bolivia. We noticed no diflerence between Bolivian spec- imens from Beni and Brazilian SF)ecimens from the vicinity of the typ>e locality of O. microtis (Low- er Rio Solimoes, 50 mi above mouth), and thus the subspecific name Oryzomys microtis microtis might be applied to Bolivian specimens. However, we need to know more about geographic variation in the species (see Remarks under O. microtis). Some other names that may refer to consjsecific populations and thus be relevant as possible sub- specific epithets are as follows: 1. Oryzomys destructor from lowland eastern Peru has been assigned to longicaudatus but may prove to be consp)ecific with microtis; if so, de- structor is the senior synonym. If destructor and microtis are conspecific, the northern Bohvian mice 264 HELDIANA: ZOOLOGY Fig. 2. Distribution of Bolivian Oligoryzomys: • = Oryzomys microtis; O = O. longicaudatus; X = O. chacoensis. might be known as Oryzomys destructor destruc- Bolivia on geographic and ecological grounds, al- tor, or perhaps O. destructor microtis. though no specimens are presently available. Myers 2. Oryzomys fornesi from northern Argentina and Carleton (1981) referred specimens from San might also be expected to occur in southeastern Joaquin, Beni, to O. fornesi. OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 265 3. Oryzomys mattogrossae from Brazil just to the east of Bolivia is conspecific with O. microtis and has its type locality nearest the department of Beni, where most of the Bolivian specimens have been taken; thus, for geographic reasons, it must be considered in any future subspecific determi- nations. The uncertain specific status of utiariten- sis from the same type locality as mattogrossae is noted below. If it is conspecific with mattogrossae, utiaritensis should be regarded as a synonym thereof at the subspecies level also. We concur with Myers and Carleton (1981) in the recognition of two species groups of Oligory- zomys. These are the smaller-bodied, small-toothed Oryzomys microtis. O.Jlavescens, and O. delicatus, and the larger-bodied, larger-toothed O. chacoen- sis, O. longicaudatus, and O. nigripes. In the Species Accounts section, synonymies cover only Bolivian records. Other relevant names are discussed elsewhere in the text. Results of Statistical Analyses Sex and Age Variation We examined sex and age variation in Oryzomys microtis. O. longicaudatus. O. chacoensis. and O. flavescens (from Uruguay). Males are slightly larg- er in general; the average size differences, consid- ering all characters for each species, are 2.7%, 1 .6%, 2.5%, and 2. 1%, respectively. We examined sexual dimorphism in all measurements for each species by one-way and two-way analyses of variance of sex and of sex and age for each species. Results of two-way analyses of variance of sex and age (using all specimens of tooth wear class II or greater) on specimens of O. microtis from the department of Beni, Bolivia, are presented in Table 1 . Our results are roughly comparable to those of Myers and Carleton (1981). We did not separate sexes in fur- ther statistical analyses, although we watched for unbalanced sex ratios in samples when interpret- ing results. Age variation is more difficult to assess, as no detailed study has been published. Mice of this subgenus probably continue to grow for most of their lives (see Myers &. Carleton, 1981), although the rate slows with age. We performed least-squares regression analyses (General Linear Models pro- cedures of SAS) on five variables: greatest length of skull (GLS), total length (TL), zygomatic breadth (ZB), length of the first upper molar (LMl), and length of the hind foot (HF), by species, to plot i graphically the relationship between relative age I and size. Based on study of these graphs, we de- cided to include age classes II-V in further anal- yses, as did Myers and Carleton (1981). Most spec- I imens were juveniles (age class I) or young adults (age classes II-III). Few specimens of age classes IV-V were present among the species we exam- ined. See Tables 2 and 3 for the mean adult ages of the specimens studied. Principal Components We analyzed principal components (with the Princomp procedure of SAS on a correlation ma- trix) using measurements taken on individuals of all sp)ecies (the approximate numbers of specimens are in tables 2 and 3, some specimens were ex- cluded because of missing measurements). When plotted, the first principal component tends to sep- arate the smaller species {O. microtis and O. fla- vescens) from the larger species {O. chacoensis and O. longicaudatus). The percentage of the totJil vari- ance accounted for by the first three components is 68.7%. Discriminant Analysis We performed several different discriminant analyses (using the SAS programs Discrim, Step- disc, and Candisc). A stepwise discriminant anal- ysis chose the following characters (in order of selection): tail length, breadth of interorbital con- striction, diastema length, length of incisive fo- ramina, rostral length, bullar length, length of hind foot, molar breadth, length of nasal projec- tion, length of zygomatic notch, total length, zy- gomatic breadth, length of palatal bridge, and length of maxillary toothrow. Using the SAS program Discrim, we were able to test the posterior probability of group mem- bership. In all analyses, we used as "known" groups samples of the species Oryzomys microtis from Beni, Bolivia; O. flavescens from Uruguay; O. lon- gicaudatus from the department of La Paz. Boliv- ia; and O. chacoensis from Paraguay. Plots of the individual mice on the first two canonical corre- lates showed little overlap among the four species. We assumed that the following "unknowns" be- longed to one of the four "known" groups. We submitted as "unknowns" the holotypes of O. mi- 266 HELDIANA: ZOOLOGY crotis, O. mattogrossae, and O. chaparensis; the paratype of O. chaparensis; and two specimens from Beni assigned by Myers and Carleton (1981) to O. chacoensis (see Remarks under O. microtis). The holotypes of O. microtis and O. mattogrossae were assigned to the O. microtis species sample, with a posterior probability of more than 95%. The holotype of O. chaparensis was also assigned to O. microtis (P = 0.987). The paratype of O. chaparensis was assigned to O. microtis (P = 0.985). Several other specimens that we wanted to test, including the type of O. delicatus, had missing values and could not be used in the analysis. We recognize that the validity of the taxonomic con- clusions based on these analyses depends on the correctness of the initial assumption, and that it needs further testing. Not all measurements were available for all skulls, but in each comparison the largest possible subset was used. Species Accounts Subgenus Oligoryzomys Bangs, 1 900 Diagnosis— Within the genus Oryzomys, the subgenus Oligoryzomys is distinguished by small size and delicate structure throughout, tail rela- tively long, hind foot long and slender; skull small, delicate, interorbital region narrow, outer edges of frontals squarish but unbeaded, braincase smooth and unridged, zygomatic plate narrow and with slight forward projection, molar teeth and incisors small and delicate but with cusp pattern like other Oryzomys (description adapted from Bangs, 1 900). Oryzomys of the subgenus Microryzomys are also small and delicate, but they differ from Oligory- zomys in having a more slender rostrum, shallow- er zygomatic notch, shorter and more rounded braincase, sphenofrontal foramen and squamo- soalisphenoid groove present, and karyotype with a low FN/2n ratio (Myers & Carleton, 1981, p. 12). Oryzomys chacoensis Myers and Carleton, 1981 Oryzomys chacoensis Myers and Carleton, 1 98 1 , p. 19 (typ)e locality "419 km by road NW Villa Hayes [alongside the Trans Chaco Highway], Dept. Bo- queron, Paraguay"). Diagnosis— C>ryzc>my5 chacoensis Myers and Carleton (1981, p. 20) was diagnosed as "A me- Table 1 . Nongeographic variation in Oryzomys mi- crotis. Results of analysis of variance of Bolivian O. mi- crotis. Char- Error F (inter- acter d.f. F(sex) F (age) action) ZN 98 3.75 0.44 0.52 LR 89 26.31*** 1.87 1.35 GLS 81 22.69*** 1.77 0.75 ZB 86 13.65*** 1.19 1.33 BB 92 6.57* 0.41 2.55* BIC 96 0.74 1.79 1.22 LIF 95 4.64* 1.71 0.51 LPB 93 4.32* 0.81 1.01 LMl 98 0.08 2.38* 0.76 LM2 99 1.38 5.37*** 5.57*** LM3 95 4.23* 1.93* 1.82 WMl 98 0.51 1.38 1.68 WM2 99 0.82 1.43 1.06 WM3 95 0.17 1.90* 1.24 MM 95 3.07 3.08** 1.81 LMX 94 1.53 2.38* 1.36 LNP 87 15.05*** 1.26 1.73 LD 96 10.87** 1.23 1.58 LB 92 5.45* 0.48 0.71 HB 91 1.77 0.41 0.83 TL 91 5.36* 1.38 1.21 T 91 0.02 1.65 2.34* HF 92 6.41* 0.95 0.67 E 48 0.02 1.53 0.84 * P < 0.05; ** P < 0.01; *** P < 0.001. dium-sized species of the subgenus Oligoryzomys unique in its whitish underside with hair white to the base on the chin and throat, relatively long ears having hairs on inner surface with unusually short or absent dark basal bands, small but dis- tinctive tufts of orangish hairs anterior to the ears, and karyotype with 2n = 58, FN = 74." Distribution in Bolivia— We examined spec- imens (see Appendix) from the departments of Cochabamba, Santa Cruz, and Tarija (fig. 2). Myers and Carleton (1981) reported two specimens from Beni. These specimens have been restudied and are here reassigned to Oryzomys microtis (see Re- marks under that species). Habitat of O. chacoen- sis is grassland and thomscrub. The highest known elevation is 640 m (Rio Lipeo). General Description and Comparisons— Cranial and external measurements are listed in Table 2. The dorsal pelage is rufous and heavily lined with black hairs. The venter is white and sharply distinguished from sides, which are clearer than dorsum. Frequently a thin orange line sep- arates the sides from the belly. The cheeks are slightly paler than the rest of the face. The tail is dark gray, weakly bicolored, and long relative to the body. The hind feet are whitish above. Juve- OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 161 Table 2. Mean, standard deviation, and range (in mm) for Oryzomys chacoensis and O. longicaudatus. O. chacoensis O. longicaudatus Bolivia Paraguay Argentina Bolivia Character (N = 15) (N = 16) (N = 5) (N = 34) ZN 1.15 ± 0.28 1.30 ± 0.20 1.15 ± 0.12 1.02 ± 0.20 (0.88-1.59) (0.90-1.61) (0.94-1.24) (0.57-1.36) LR 7.27 ± 0.45 7.39 ± 0.57 7.67 ± 0.34 7.14 ± 0.72 (6.62-7.94) (6.57-8.86) (7.25-8.07) (4.49-8.15) GLS 24.90 ± 0.82 24.52 ± 1.09 26.12 ± 0.32 25.01 ± 1.39 (23.80-26.69) (22.52-27.12) (25.79-26.43) (21.06-27.29) ZB 13.10 ± 0.59 12.76 ± 0.62 13.64 ± 0.09 13.17 ± 0.65 (11.94-13.98) (11.91-14.27) (13.54-13.71) (11.55-13.98) BB 11.33 ± 0.41 11.16-0.34 11.39 ± 0.06 11.25 ± 0.37 . (10.42-11.87) (10.59-11.59) (11.35-11.45) (10.58-11.96) BIC 3.89 ± 0.20 3.76 ± 0.10 3.86 ±0.13 3.63 ± 0.22 (3.52-4.27) (3.64-3.95) (3.76-4.04) (3.35-4.17) LIF 4.62 ± 0.26 4.71 ± 0.39 5.07 ± 0.29 4.68 ± 0.27 (4.25-5.20) (4.17-5.56) (4.64-5.31) (4.23-5.26) LPS 4.14 ± 0.22 3.98 ± 0.26 4.22 ± 0.31 3.97 ± 0.21 (3.65-4.54) (3.50-4.65) (3.75-4.52) (3.49-4.41) LMl 1.68 ± 0.10 1.63 ± 0.07 1.74 ± 0.5 1.59 ± 0.09 (1.47-1.84) (1.54-1.79) (1.70-1.83) (1.42-1.81) LM2 1.09 ± 0.06 1.10 ± 0.07 1.14 ± 0.07 1.05 ± 0.04 (0.96-1.18) (1.00-1.28) (1.08-1.26) (0.97-1.15) LM3 0.84 ± 0.05 0.80 ± 0.04 0.84 ± 0.04 0.81 ± 0.05 (0.74-0.92) (0.73-0.88) (0.78-0.88) (0.67-0.92) WMl 1.07 ± 0.07 1.03 ± 0.03 1.11 ± 0.05 1.04 ± 0.06 (0.93-1.24) (0.96-1.08) (1.05-1.17) (0.91-1.15) WM2 1.02 ± 0.07 1.00 ± 0.04 1.04 ± 0.06 0.99 ± 0.06 (0.86-1.14) (0.93-1.06) (0.97-1.11) (0.87-1.09) WM3 0.87 ± 0.05 0.85 ± 0.03 0.90 ± 0.05 0.83 ± 0.06 (0.78-0.97) (0.78-0.92) (0.84-0.94) (0.67-0.95) MM 4.66 ± 0.20 4.57 ± 0.17 4.70 ± 0.13 4.61 ± 0.27 (4.32-4.99) (4.34-5.09) (4.58-4.87) (4.03-5.08) LMX 3.60 ±0.12 3.50 ± 0.13 3.71 ± 0.15 3.46 ± 0.14 (3.29-3.78) (3.23-3.68) (3.61-3.98) (3.20-3.77) LNP 1.43 ± 0.24 1.59 ± 0.17 1.49 ± 0.17 1.40 ± 0.25 (1.04-1.77) (1.35-2.07) ~ (1.29-1.64) (0.99-2.04) LD 5.66 ± 0.34 5.63 ± 0.44 5.89 ± 0.24 6.04 ± 0.47 (5.17-6.33) (5.02-6.44) (5.63-6.18) (4.96-7.07) LB 3.54 ± 0.14 3.60 ± 0.14 3.75 ± 0.03 3.47 ± 0.20 (3.28-3.81) (3.28-3.82) (3.72-3.78) (3.07-3.84) HB 2.75 ± 0.18 2.73 ± 0.15 2.72 ± 0.06 2.76 ± 0.32 (2.43-3.04) (2.40-2.96) (2.66-2.76) (2.23-3.32) TL 227.93 ± 12.45 219.50 ± 12.29 229.00 ± 8.34 217.68 ± 16.88 (204-252) (194-241) (215-237) (172-252) T 134.27 ± 10.26 126.06 ± 8.46 138.00 ± 5.70 124.15 ± 12.18 (112-150) (112-139) (130-145) (87-145) HF 25.33 ± 1.33 24.38 ± 1.54 23.40 ± 2.30 25.13 ± 1.20 (23-28) (23-29) (21-26) (22-28) E 16.70 ± 1.33 15.07 ± 1.69 18.40 ± 1.14 15.03 ± 1.72 (14-18) (12-17) (17-20) (12-18) AGE 2.65 ± 0.82 2.31 ± 0.68 2.55 ± 0.87 3.04 ± 0.88 (2-4) (2-4) (2-4) (2-5) 268 HELDIANA: ZOOLOGY niles are grayer, as are all young Oligoryzomys in Bolivia, but this is especially noticeable on the venter, which has mixed gray and white hairs. In- cisive foramina extend posteriorly to the anterior edge of M' or slightly beyond. Alisphenoid strut (a strut of the alisphenoid bone that covers the lateral part of the alisphenoid canal, see Musser, 1 982, p. 29) is generally absent (table 4). The sides of the interorbital constriction are divergent pos- teriorly (less parallel-sided than in O. longicau- datus) (fig. 3). The following quotes are from Myers and Carleton (1981, first and second quotes, p. 2 1 , third quote, p. 24): Oryzomys chacoensis differs from O. cha- parensis Osgood (1916) primarily in color pattern: the type of chaparensis is much darker and less hispid dorsally, yellowish ventrally, lacks the orange tufts anterior to the ears, and has a grayish throat. The distal portions of the nasals of the holotype flare laterally to an extent not seen in chacoensis. amined the two darker ones. When skulls are com- pared, the two darker mice are among the older and larger individuals in the series. They seem to have relatively broader braincases than most, but there is no character shared by these two that is not also seen in one or more of the others. In a discriminant analysis (SAS program Discrim) the posterior probability of membership (in the four reference species) allies these two mice with O. chacoensis (ansp 18187, P = 0.997, and ansp 18188, P = 0.954 with chacoensis). We have as- signed these mice to O. chacoensis. Since Rio Li- pto is in the area where these two species meet, further study there should reveal whether sym- patry exists and if species differences remain dis- tinct. Specimens from Argentina (listed in the Ap- pendix) are similar in coloration to both Para- guayan and Bolivian specimens, but are woolier. This sample extends the known range of O. cha- coensis into the department of Jujuy, northwestern Argentina. Oryzomys chacoensis can be distinguished from fornesi, with which it occurs sympat- rically, by its larger size (maxillary toothrow usually > 3.3mm, ears usually > 15mm), characteristic karyotype, lack of preputial glands, and in most specimens by its lack of buff" on the belly. . . . the hind feet of chacoensis are relatively short compared to those of the more terres- trial fornesi. In reference to the last point, however, our cal- culations of length of hind foot relative to length of head and body (using data from tables 2 and 3) are about 26% for both O. chacoensis and O. mi- crotis (including fornesi). Comparison of speci- mens also reveals no noticeable difference. Remarks— This species is clearly distinct from O. microtis (here including O. fornesi). In most morphometric characters, O. chacoensis grossly resembles O. longicaudatus from Bolivia and O. nigripes from Brazil. Our specimens from Santa Cruz show some variation in coloration: two specimens (amnh 247772-247773) are pale. The specimens from Tarija resemble the Paraguayan samples more closely than do the specimens from Santa Cruz. A series of 13 sp)ecimens from Rio Lipeo in- cludes two with darker pelage that resemble O. longicaudatus. We measured these mice and ex- Oryzomys longicaudatus (Bennett, 1832) Mus longicaudatus Bennett, 1832, p. 2 (type locality "In trees in Chile," restricted to Valparaiso by Ca- brera, 1961, p. 391). Oryzomys longicaudatus: Thomas, 1898, p. 3 (Aguai- renda Mission, San Francisco, perhaps not O. lon- gicaudatus); Thomas, 1926, p. 194 (Tupiza). Oryzomys Stolzmanni: Thomas, 1902, p. 130 (Cha- ruplaya, Choro); Neveu-Lemaire and Grandidier, 1 9 1 1 , p. 9 (Charuplaya, Choro). Oryzomys ^X>;flavescens groMTp: Thomas, 1925, p. 578 (Carapari, perhaps not O. longicaudatus). Oryzomys stolzmanni stolzmanni: Sanborn, 1950, p. 2 (Rio Aceramarca, Cocapunco, Nequejahuira, Okara, Pitiguaya, Pongo). Diagnosis— Bolivian specimens referred to this species differ from other Bolivian Oligoryzomys in that they are larger (especially in the size of the teeth) than O. microtis and lack the buffy wash on the venter; are grayer-bellied, darker, and less griz- zled dorsally than O. chacoensis; and occur at gen- erally higher elevations (at least in northern Bo- livia). Distribution in Bolivia— O. longicaudatus is found in the valleys and mountains of the Andes from at least middle elevations (1200 m, Entre Rios) up to 3720 m (Poopo). Specimens from near Camiri and Cuyambuyo in southern Bolivia are from elevations of 780 to 1000 m. Specimens (see Appendix) have been reported or examined from OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 269 Fig. 3. Dorsal and ventral views of skull of (left), Oryzomys longicaudatus stolzmanni; (middle), O. chacoensis, and (right), O. microtis microtis. Specimens in the American Museum of Natural History. Scale at lower right represents 10 mm. the departments of Chuquisaca, Cochabamba, La Paz, Oruro, Polosi, Santa Cruz, and Tarija (see map, fig. 2). General Description and Comparisons— Cranial and external measurements are presented in Table 2. The belly is gray, rarely washed with buff; upper parts are a dark brown lined with black hairs, and often the sides are more rufous. A thin orange lateral line may be present. Nose and face are dark with paler cheeks. Often there are pale spots just behind the ears. Hind feet are pale above. The tail is brown above, weakly to strongly bi- colored, and relatively long compared to that of any other Bolivian Oligoryzomys. Incisive foram- ina generally extend to the anterior edge of M' or slightly behind. Alisphenoid strut is generally ab- sent (table 4). The interorbital area is slightly more constricted on the average than that of O. cha- 270 HELDIANA: ZOOLOGY coensis, and has relatively parallel sides (see fig. 3). Remarks— Oryzomys longicaudatus may be a composite of more than one species. Oryzomys destructor and O. stolzmanni were considered sub- species of 6>. longicaudatus by Cabrera ( 1 96 1 ) and full species by Soukup (1961). Gardner and Patton (1976) found four karyotypic variants in speci- mens they assigned to O. longicaudatus, and com- mented that these may represent four separate species. One of their karyotypic variants, no. 2, is represented by 1 7 specimens from Balta, depart- ment of Loreto (now Acayali), Peru, at an eleva- tion of about 300 m. Myers and Carleton (1981, p. 26) mentioned that the karyotype of these spec- imens as reported by Gardner and Patton (1976) agreed better with results they obtained for O.for- nesi (= O. microtis) than for either O. chacoensis or O. longicaudatus. We have examined these specimens and judge that they are clearly different from O. longicaudatus and belong with either O. microtis or O. destructor (Balta is near the type locality of O. destructor). The inclusion of karyo- type no. 3 under O. longicaudatus, represented by one female from Venezuela, also needs verifica- tion. We have five specimens from Caracato, de- partment of La Paz (amnh 248977-248981), that are quite different from other Bolivian Oryzomys longicaudatus. The pelage closely resembles that of O. nigripes from Paraguay in color and texture; it is paler and more obviously lined with black hairs dorsally than the pelage of typical O. lon- gicaudatus. They are also slightly larger bodied, but four of the five are old individuals (age classes IV-V) and this may account for the larger size. Because of the small sample size and the single locality, we chose not to distinguish this popula- tion taxonomically. It is possible that Andean populations may oc- cupy areas that are rather widely scattered and physically isolated in different valleys or side can- yons. If so, there may be considerable geographic variation. More specimens are needed to detect and describe such patterns. The sjjecimens from near Cuyambuyo are at comparatively low elevations (980 to 1000 m) for O. longicaudatus, although the species occurs at progressively lower elevations farther south (Os- good, 1916). The two localities near Cuyambuyo are only about 60 km from Rio Lipeo, where O. chacoensis occurs. The SAS Discrim program analysis allied the largest adult specimen from near Cuyambuyo (ummz 155891) with O. longicauda- tus (F = 0.956). There are problems in distinguish- ing O. longicaudatus and O. chacoensis solely by cranial morphology in all areas. Oryzomys microtis Allen, 1916 Oryzomys (Oligoryzomys) microtis Allen, 1 9 1 6, p. 525 (type locality Lower Rio Solimoes, 50 mi above mouth [80 km from its confluence with Rio Negro, Amazonas, Brazil]). Oryzomys chaparensis Osgood, 1 9 1 6, p. 205 (holotype and paratype, type locality Todos Santos, on Rio Chapare, Department of Cochabamba, Bolivia); Gyldenstolpe, 1932, p. 25 (Todos Santos); Myers and Carleton, 1981, p. 38 (Todos Santos); all in reference to the same specimens. Oryzomys fornesi Massoia, 1973, p. 22 (type locality Naineck, Dept. Rio Pilcomayo, province of For- mosa, Argentina); as used by Myers and Carleton, 1981, p. 25 for five specimens fVom San Joaquin, Beni, Bolivia. Diagnosis— Allen described Oryzomys microtis as being readily distinguished from other Olig- oryzomys by its pale coloration, relatively small ears, and tail less than half total length. However, the tail of the holotype is more than half of the total length, a discrepancy noted by Goodwin (1953). Massoia (1973) diagnosed O. fornesi, here considered a junior synonym of O. microtis, as the smallest species of the subgenus in Argentina, hav- ing the ears covered with short ochraceous hairs, the pterygoids short (shorter than the molar se- ries), the interorbital constriction narrow, and the incisive foramina not extending to the molar se- ries. In Bolivia, it can be distinguished by its small body size, toothrow generally less than 3.3, gray- buffy color of the belly, and relatively short tail in proportion to body. Distribution in Bolivia— Specimens (see Ap- pendix) from the lowlands of Bolivia, in the de- partments of Beni, Cochabamba, Pando, La Paz, and Santa Cruz have been examined. Habitat in- cludes marshes and wet forests up to an elevation of 1 800 m (Guanay, see fig. 2). General Description and Comparisons — Smallest of the Bolivian Oligoryzomys, with small teeth, toothrow generally less than 3.3 (averaging 3.10). Massoia ( 1 973) gave general measurements useful in distinguishing this species from other Ar- gentine species of Oryzomys (Oligoryzomys) as fol- lows: length of hind foot generally less than 24, length of ear generally less than 13, breadth of braincase less than 10.8, and length of incisive foramina generally less than 4.5. All of his mea- OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 271 Table 3. Mean, standard deviation, and range (in mm) for Oryzomys microtis and O. flavescens. O. microtis O. flavescens Bolivia Brazil Peru Uruguay Character (N = 67) (N = 20) (N = 26) (N = 20) ZN 1.01 ± 0.17 0.98 ±0.18 0.84 ±0.13 1.07 ± 0.15 (0.53-1.46) (0.78-1.48) (0.61-1.18) (0.79-1.32) LR 6.82 ± 0.51 6.82 ± 0.42 6.60 ± 0.41 6.75 ± 0.43 (5.17-8.16) (6.16-7.46) (5.96-7.36) (5.89-7.39) GLS 23.54 ± 1.06 23.79 ± 0.88 23.27 ± 0.82 23.41 ± 0.94 (21.24-25.78) (22.47-25.04) (21.75-24.53) (21.66-24.81) ZB 12.28 ± 0.51 12.40 ± 0.44 11.86 ± 0.53 12.05 ± 0.49 (11.42-13.73) (11.59-13.32) (11.01-12.93) (11.25-12.76) BB 10.78 ± 0.36 10.71 ± 0.23 10.47 ± 0.32 10.70 ± 0.26 • (9.38-11.78) (10.35-11.06) (9.49-10.87) (10.29-11.22) BIC 3.76 ± 0.17 3.72 ± 0.14 3.63 ±0.11 3.35 ±0.17 (3.38-4.19) (3.44-3.93) (3.45-3.94) (2.95-3.77) LIF 3.96 ± 0.27 4.01 ± 0.22 4.00 ±0.18 4.59 ± 0.35 (3.38-4.53) (3.65-4.37) (3.73-4.33) (3.54-5.20) LPB 4.04 ± 0.32 4.06 ± 0.22 3.98 ± 0.26 3.77 ± 0.19 (3.14-4.76) (3.64-4.38) (3.42-4.51) (3.51-4.06) LMl 1.46 ± 0.10 1.43 ± 0.07 1.51 ± 0.06 1.47 ± 0.06 (1.22-1.64) (1.32-1.56) (1.39-1.66) (1.37-1.60) LM2 0.95 ± 0.06 0.95 ± 0.05 0.93 ± 0.08 1.01 ± 0.06 (0.81-1.09) (0.84-1.06) (0.67-1.04) (0.83-1.09) LM3 0.70 ± 0.07 0.68 ± 0.05 0.66 ± 0.05 0.73 ± 0.04 (0.54-0.85) (0.59-0.78) (0.55-0.77) (0.66-0.80) WMl 0.94 ± 0.06 0.93 ± 0.06 0.99 ± 0.04 0.97 ± 0.04 (0.78-1.08) (0.80-1.04) (0.91-1.10) (0.93-1.07) WM2 0.88 ± 0.06 0.89 ± 0.06 0.93 ± 0.04 0.93 ± 0.05 (0.70-1.01) (0.78-0.99) (0.82-1.00) (0.86-1.03) WM3 0.73 ± 0.06 0.74 ± 0.06 0.80 ± 0.05 0.78 ± 0.04 (0.55-0.85) (0.58-0.85) (0.71-0.93) (0.70-0.85) MM 4.33 ± 0.21 4.39 ± 0.23 4.33 ± 0.18 4.26 ±0.15 (3.83-4.93) (3.99-4.81) (4.07-4.70) (3.98-4.50) LMX 3.10 ± 0.17 3.07 ± 0.16 3.14 ± 0.13 3.22 ±0.12 (2.64-3.43) (2.75-3.34) (2.93-3.39) (2.96-3.47) LNP 1.32 ± 0.22 1.25 ± 0.32 1.20 ± 0.20 1.33 ± 0.14 (0.91-1.74) (0.75-1.70) ■ (0.81-1.62) (1.16-1.68) LD 5.75 ± 0.41 5.78 ± 0.36 5.32 ± 0.40 5.35 ± 0.35 (4.82-6.62) (5.13-6.51) (4.49-6.01) (4.67-5.90) LB 3.29 ± 0.14 3.18 ± 0.19 3.01 ± 0.12 3.30 ±0.15 (2.94-3.57) (2.90-3.63) (2.75-3.27) (2.91-3.53) HB 2.46 ± 0.20 2.10 ± 0.23 2.22 ±0.12 2.64 ± 0.23 (1.96-2.87) (1.72-2.45) (1.96-2.38) (2.02-2.94) TL 185.88 ± 11.12 185.00 ± 10.60 179.77 ± 9.16 198.56 ± 17.66 (165-214) (164-202) (160-196) (160-223) T 101.00 ± 6.67 95.44 ± 7.88 101.69 ± 8.13 110.37 ± 11.05 (87-116) (75-105) (77-115) (85-127) HF 22.81 ± 1.21 21.68 ± 1.34 21.37 ± 1.60 25.21 ± 1.85 (19-25) (20-25) (18-25) (20-28) E 14.49 ± 1.04 (12-16) 13.97 ± 1.05 (12-16) AGE 2.57 ± 0.73 2.35 ± 0.59 2.97 ± 0.88 2.18 ± 0.35 (2-5) (2-4) (2-5) (2-3) 272 FIELDIANA: ZOOLOGY surements fall within the extremes of our values for O. microtis, except for measurements of the toothrow (LMX), in which his are on the upper end of our range. Values of cranial and external measurements are listed in Table 3. The venter is white mixed with gray to gray-bufFy and clear huf- fy. The dorsum is rufous brown, lined with black hairs, and has paler sides. Often there is no clear demarcation between sides and belly. There are no white spots behind ears. Juveniles are similar to adults, but show a grayer belly. The tail is dark above and weakly bicolored. Occasionally orange- tipped hairs lie anterior to the ear. Hind feet are white above. Incisive foramina generally extend posteriorly to the anterior edge of M' and not beyond. In comparison with adults of O. cha- coensis and O. longicaudatus, O. microtis has a shorter skull, longer braincase, shorter incisive fo- ramina, a relatively broader interorbital constric- tion with divergent sides, and smaller, narrower teeth; also, an alisphenoid strut is more often pres- ent (table 4). Remarks— This species is similar to Oryzomys flavescens of Uruguay and Argentina (see Myers & Carleton, 1981). Specimens of both species were reported from the same locality, Capitan Solari, in the province of Chaco, northern Argentina, by Contreras and Berry (1983). Measurements of 6>. flavescens from Uruguay are listed in Table 3 (see also Langguth, 1 963). In comparison with O. mi- crotis, specimens of O. flavescens have slightly larger teeth, larger bullae, longer incisive forami- na, and are generally larger. Oryzomys flavescens needs to be more clearly defined. Massoia and Fomes ( 1 967) once synonymized O. flavescens with O. nigripes, but according to the later analyses of Massoia (1973) and Myers and Carleton (1981) these taxa are not consjiecific. Uruguayan speci- mens oi O. flavescens are clearly different from O. nigripes from eastern Paraguay. Future studies may reveal additional relationships, including the pos- sibility that O. microtis and O.fiilvescens of Cen- tral America and northern South America are con- specific (Handley, 1976, referred Venezuelan specimens to O. fiilvescens). The name Oryzomys microtis flyrnesi might be used to reflect the conspecific status oiflyrnesi and microtis. This subspecific name would apply to at least the five specimens identified by Massoia (1973), pending more detailed studies that would test whether geographic differences warrant con- tinued subspecific recognition. Until the relationships of the Andean ""longi- caudatus-group" are clarified, it is also possible Table 4. Presence or absence of alisphenoid strut in five species of Oryzomys (Oligoryzomys) expressed as a percentage; number of specimens in parentheses. Alisphenoid strut Species Present Absent O. chacoensis 2.4(1) 97.6(41) O. longicaudatus 5.6 (2) 94.4 (34) O. microtis 35.8 (54) 64.2 (97) O. flavescens 3.2(1) 96.8 (30) • O. nigripes 8.3 (3) 91.7(33) that one or more of the earlier names currently assigned to this group actually belong with Ory- zomys microtis, and have priority. For example: O. stolzmanni (named in 1894, type locality listed under synonymy) and O. destructor (named in 1 844, type locality: "the house mouse of the 'Plan- tagen at the border of the forest," eastern Peru, above 6000 ft, according to Tate, 1932, p. 9; re- stricted by Cabrera, 1961, to the haciendas along the Rio Chinchao in the department of Huanuco, between 900 and 1000 m) have been associated with O. longicaudatus. These type localities are at low elevations, which suggests the possible affinity of these forms with microtis rather than with lon- gicaudatus. We have examined a series from low elevation in the department of Pasco, Peru, that is clearly O. microtis. This locality is near the type locality of destructor (Cole, 1984). If future study of the holotype of destructor proves it to be conspecific with O. microtis, the name of the species should be O. destructor instead of O. microtis. We have examined the type specimens of Ory- zomys microtis (amnh 37091, type locality Lower Rio Solimoes, 50 mi above mouth, Brazil), O. mattogrossae (amnh 37542, type locality Utiarity, Mato Grosso, Brazil), O. utiaritensis (amnh 37541, type locality Utiarity, Rio Papagaia, Mato Grosso, Brazil), and O. delicatus (amnh 7317/5925, type locality Trinidad, West Indies). We judge that the small lowland Bolivian specimens are O. microtis and also represent the same species as specimens referred by Myers and Carleton to O. fornesi (O. microtis has priority). The Bolivian specimens dif- fer slightly from Massoia's description of O. for- nesi in having a shorter average toothrow, other- wise in being slightly larger. If O. fornesi and O. microtis are conspecific, it is possible that this dif- ference in measurements reflects his small sample size (N = 5), or that there may be significant geo- graphic variation between Argentine and Bolivian populations. We further treat O. mattogrossae as a synonym of O. microtis (O. mattogrossae is not OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 111 included in the synonymy above because the name has never been used in print for a Bolivian spec- imen). As mentioned by Myers and Carleton ( 1 98 1 ), O. utiaritensis may belong with O. nigripes. More specimens from Mato Grosso are needed to establish the characteristics of the species there. Oryzomys microtis was considered a synonym of O. delicatus by Alho ( 1 982) and Pine ( 1 973); how- ever, because of the slightly damaged condition of the skin and skull of the holotype of O. delicatus and the remoteness of its type locality (on the is- land of Trinidad, Allen & Chapman, 1 897), we defer any decision on this question. We have reexamined the two specimens from Beni that Myers and Carleton (1981) referred to Oryzomys chacoensis (usnm 39 1 297, 46074 1 ) and have assigned them to O. microtis. These speci- mens are young adults (age class II), have external measurements that fit well with O. microtis and that are rather small for O. chacoensis (TL 200, 200; T 111, 110; HF 19, 24, respectively), and have toothrows measuring 3.34 and 3.40, respec- tively. The toothrows fall on the upper end of the range for O. microtis and the lower end of the range for O. chacoensis (tables 2-3). We then compared their measurements through a discriminant anal- ysis with the samples ofO.flavescens, O. microtis, O. chacoensis. and O. longicaudatus. The posterior probability of membership for both was greatest for O. microtis (usnm 391297, P = 0.937; usnm 460741,/' = 0.953). We refer the holotype (fmnh 21330) and the one paratype (amnh 40787) o^ Oryzomys chapa- rensis to Oryzomys microtis. The holotype is the largest and one of the oldest individuals we have identified as O. microtis. Its greatest skull length is 26.8, zygomatic breadth 13.4, breadth of brain- case 1 1.2, crown length of maxillary toothrow 3.5, tail length 111, and length of hind foot 26. The skull length, the toothrow, and hind foot length are larger than other Bolivian O. microtis (see table 3). The pelage coloration of this specimen falls within the range of variation in O. microtis. For a comparison of pelage coloration between O. cha- parensis and O. chacoensis, see the account of O. chacoensis under General Description and Com- parisons. Geographic Variation— We were unable to detect significant differences in p>elage coloration between our samples from Brazil, Bolivia, and Peru. The Brazilian specimens are much like the Bolivian specimens. A / test showed only three variables that differ significantly: bullar height (/ = 5.45, P < 0.001), tail length (/ = 2.61, P < 0.05), 274 and length of hind foot (/ = 3.24, P < 0.01); all have greater average values in Bolivian specimens, | but the ranges of values overlap. One possible con- founding factor is that the Brazilian sample is com- posed of mostly males (ca. 95%) and has a lower average adult age than the Bolivian or Peruvian ' samples (see tables 2-3). If the degree of sexual dimorphism in this species is considered (2.72%, see p. 266), the difference may actually be less than i noted in our tests. For these reasons the subspecies ! name Oryzomys microtis microtis might be ap- I propriate for specimens from northern Bolivia. More differences exist between Peruvian and Bolivian samples. The Bolivian specimens are generally larger: the zygomatic notch is deeper, diastema longer, braincase broader, zygomatic breadth greater, and interorbital constriction broader. However, Bolivian specimens have sig- nificantly narrower teeth. Brazilian specimens, as noted above, are similar to Bolivian specimens and differ from the Peruvian material about as much as Bolivian specimens do. The Peruvian mice used in this comparison come from a lowland site in Pasco, the department just south of Huanuco, which is near the type locality of Oryzomys de- structor. One skin (amnh 247776) from near Villa Tunari in Cochabamba is dorsally paler and ventrally whiter (not bufly) than any skin from Pando or Beni. A skin (amnh 246809) from near Buena Vis- ta in Santa Cruz is comparable to the darker spec- imens from Beni, but is ventrally less bufl[> . Concluding Remarks Given the considerable pelage differences noted in reference to the few available specimens from Cochabamba and Santa Cruz in the accounts of Oryzomys chacoensis and O. microtis, the prob- lems noted in reference to sF>ecimens from south- em Tarija in the accounts of O. chacoensis and O. longicaudatus, the difficulty of identification that results from overlapping ranges of measurements, the sketchy knowledge of geographic variation, and the small samples from most critical areas, the taxonomy of Bolivian Oligoryzomys will probably prove to be more complicated than that presented here. We hojie our summary will provide infor- mation useful in further taxonomic work on the subgenus. All of our data and detailed analyses are deposited in the Department of Mammalogy, FIELDIANA: ZOOLOGY American Museum of Natural History, and are available for the use of other interested workers. More than 400 additional specimens, many in- cluding karyological and biochemical prepara- tions, have been collected in 1984, 1985, and 1986 since this paper was written. Study of these will certainly help to clarify remaining taxonomic problems. Acknowledgments This study was partially funded by a grant from the Undergraduate-Graduate Research Fund Pro- gram, which is supported by the Susan Greenwall Foundation and administered by the American Museum of Natural History. We thank Dr. Leslie F. Marcus for his statistical advice. Masaaki Yone- da kindly provided us with a list of SF>ecimens in the Museo Nacional de Historia Natural, La Paz, Bolivia (MNHN). We are grateful to the curators of the collections referred to here, especially Dr. Philip Myers, Dr. Guy G. Musser, and Dr. Michael D. Carleton, for their generous advice and assis- tance in various ways. We also thank Dr. Robert S. Voss, Dr. James L. Patton, and Dr. Robert M. Timm for reviewing the manuscript. Literature Cited Alho, C. J. R. 1982. Brazilian rodents: Their habitats and habits, pp. 143-166. In Mares, M. A., and H. H. Genoways. eds., Mammalian Biology in South Amer- ica. Special Publication Series, Pymatuning Labora- tory of Ecology, University of Pittsburgh, 6: 1-539. Allen, J. A. 1916. New mammals collected on the Roosevelt Brazilian Expedition. Bulletin of the Amer- ican Museum of Natural History, 35: 523-530. Allen, J. A., and F. Chapman. 1897. On a second collection of mammals from the island of Trinidad, with descriptions of new species, and a note on some mammals from the island of Dominica, W.I. Bulletin of the American Museum of Natural History, 9: 13- 30. Anderson, S. 1968. A new craniometer and sugges- tions for craniometry. Journal of Mammalogy, 49: 22 1- 228. Bangs, O. 1900. List of the mammals collected in the Santa Marta region of Colombia by W. W. Brown, Jr. Proceedings of the New England Zoological Club, 1: 87-102. Bennett, E. T. 1832. Characters of a new species of Otter {Lutra, Erxl.), and of a new species of Mouse {Mus, L.), collected in Chili by Mr. Cuming. Proceed- ings of the Zoological Society of London, part 2: 1-2. Cabrera, A. 1961. Catalogo de los mamiferos de America del Sur. Revista del Museo Argentino de Ciencias Naturales "Bernardino Rivadavia," Ciencias Zoologicas, 4: 309-732. Cole, Y.I. 1984. Genetic structure of a wild population of a Neotropical rodent (Oryzomys [Oligoryzomys] longicaudatus; Muridae, Cricetinae). M.A. thesis. New York University, New York, 165 pp. CoNTRERAS, J. R., AND L. M. Berrv. 1 983. Notas acer- ca de los roedores del genero Oligoryzomys de la Pro- vincia del Chaco, Republica Argentina (Rodentia, Cri- cetidae). Historia Natural, 3(15): 145-148. Ellerman, J. R. 1941. The Families and Genera of Living Rodents. Vol. 2, Family Muridae. British Mu- seum (Natural History), London, xii + 690 pp. Gardner, A. L., and J. L. Patton. 1976. Karyotypic variation in oryzomyine rodents (Cricetinae) with comments on chromosomal evolution in the Neo- tropical cricetine complex. Occasional Papers of the Museum of Zoology, Louisiana State University, 49: 1-48. Goodwin, G. G. 1953. Catalogue of type specimens of Recent mammals in the American Museum of Nat- ural History. Bulletin of the American Museum of Natural History, 102: 207-412. Gyldenstolpe, N. 1 932. A manual of Neotropical sig- modont rodents. Kungl. Svenska Vetenskapsakade- miens Handlingar. Tredje Serien, Bd. 11, no. 3: 1- 164, 17 plates. Handley, C. O., Jr. 1976. Mammals of the Smith- sonian Venezuelan Project. Brigham Young Univer- sity Science Bulletin, Biological Series, 20: 1-91. Hershkovitz, p. 1 940. Four new oryzomyine rodents from Ecuador. Journal of Mammalogy, 21: 78-84. HONACKI, J. H., K. E. KiNMAN, and J. W. KOEPPL, EDS. 1982. Mammal Species of the World. A Taxonomic and Geographic Reference. Allen Press, Inc. and the Association of Systematics Collections, Lawrence, Kansas, 694 pp. Langguth, A. 1963. Las especies Uruguayas del ge- nero Oryzomys. Comunicaciones zoologicas del Mu- seo de Historia Natural de Montevideo, v. 7, no. 99: 1-19. Mares, M. A., AND H. H. Genoways, EDS. 1982. Mam- malian Biology in South America. Special Publication Series, Pymatuning Laboratory of Ecology, University of Pittsburgh, 6: 1-539. Mares, M. A., R. A. Ojeda, and M. P. Kosco. 1981. Observations on the distribution and ecology of the mammals of Salta Province, Argentina. Annals of Car- negie Museum, 50: 151-206. Massoia, E. 1973. Descripcion de Oryzomys fornesi, nueva especie y nuevos datos sobre algunas especies y subespecies argentinas del subgenero Oryzomys {Oligoryzomys) (Mammalia-Rodentia-Cricetidae). Revista de Investigaciones Agropecuarias, INTA, Buenos Aires, Serie 1 , Biologia y Produccion Animal, 10:21-37. Massoia, E., and A. Fornes. 1967. El estado siste- matico, distribucion geografica y datos etoecologicos de algunos mamiferos neotropicales (Marsupialia y Rodentia) con la descripcion de Cabreramys, genero OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 275 nuevo (Cricetidae). Acta Zoologica Lilloana, 23: 407- 430. MussER, G. G. 1979. Results of the Archbold Expe- ditions. No. 102. The species of Chiropodomys, ar- boreal mice of Indochina and the Malay Archipelago. Bulletin of the American Museum of Natural History, 162: 379-445. . 1982. Results of the Archbold Expeditions. No. 108. The definition of Apomys, a native rat of the Philippine Islands. American Museum Novitates, 2746: 1-43. Myers, P., and M. D. Carixton. 1981. The species of Oryzomys {Oligoryzomys) in Paraguay and the iden- tity of Azara's "Rat sixi^me ou Rat a Tarse Noir." Miscellaneous Publications of the Museum of Zool- ogy, University of Michigan, 161: 1-41. Neveu-Lemaire, M., and G. Grandidier. 1911. Notes sur les mammiferes des hauts plateaux de I'Amerique du Sud. Le Sondier, Paris, viii + 127 pp. O'CoNNELL, M. A. 1982. Population biology of North and South American grassland rodents: A compara- tive study, pp. 167-185. In Mares, M. A., and H. H. Genoways, eds.. Mammalian Biology in South Amer- ica. Special Publication Series, Pymatuning Labora- tory of Ecology, University of Pittsburgh, 6: 1-539. Osgood, W. H. 1916. Mammals of the Collins-Day South American Expedition. Field Museum of Nat- ural History, Zoological Series, 10(14): 199-216. Pine, R. H. 1973. Mammals (exclusive of bats) of Be- lem. Para, Brazil. Acta Amazonica, 3: 47-79. Sanborn, C. C. 1959. Small rodents from Peru and Bolivia. I*ublicaciones del Museo de Historia Natural "Javier Prado," serie A, Zoologica, 5: 1-16. SAS Institute, Inc. 1982. SAS User's Guide: Statis- tics. Gary, North Carolina, 584 pp. SouKUP, J. 1961. Materiales para el catalogo de los mamiferos peruanos. Biota, 3: 240-276. Streilein, K. E. 1982a. Ecology of small mammals in the semiarid Brazilian caatinga. I. Climate and faunal composition. Annals of Carnegie Museum, 51: 79- 107. . 1 982b. Ecology of small mammals in the semi- arid Brazilian caatinga. II. Water relations. Annals of Carnegie Museum, 51: 109-126. 1 982c. Ecology of small mammals in the semi- arid Brazilian caatinga. III. Reproductive biology and population ecology. Annals of Carnegie Museum, 51: 251-269. Tate, G. H. H. 1932. The taxonomic history of the South and Central American cricetid rodents of the genus Oryzomys. Part 2. Subgenera Oligoryzomys, Thallomyscus, and Melanomys. American Museum Novitates, 580: 1-17. Thomas, O. 1 894. Descriptions of some new. Neo- tropical Muridae. Annals and Magazine of Natural History, (6), 14: 346-366. . 1 898. On the small mammals collected by Dr. Borelli in Bolivia and northern Argentina. BoUetino dri Musei di Zoologia ed Anatomia Comparata, R. Universita di Torino, 13: 1-4. . 1902. On mammals from Cochabamba, Bo- livia, and the region north of that place. Annals and Magazine of Natural History, (7), 9: 125-143. . 1925. The Spedan Lewis South American Ex- ploration. I. On mammals from southern Bolivia. An- nals and Magazine of Natural History, (9), 15: 575- 582. . 1926. The Spedan Lewis South American Ex- ploration. IV. List of mammals obtained by Sr. Budin on the boundary between Jujuy and Bolivia. Annals and Magazine of Natural History, (9), 18: 193-195. Vieoa-Borgeaud, T. 1982. Donnees ecologiques sur Oryzomys nigripes (Desmarest, 1819) (Rongeurs; Cricetides) dans le foyer natural de peste de Barracao des Mendes (Etat de Rio de Janeiro, Bresil). Mam- malia, 46: 335-359. 276 HELDIANA: ZOOLOGY Appendix ^ Specimens Reported or Examined The latitude and longitude are abbreviated; 1657/6523 means 16*'57'S, 65''23'W, for example. Where references are given (to Thomas), the specimens have not been examined by us. As noted, there are a few other specimens that we have not seen. O. chacoensis Specimens: 88. Boquer6n 4 1 9 km by road NW Villa Hayes, 1 amnh (248397). BOLIVIA Santa Cruz 1623/6059, San Ignacio, province of Velasco, 10 USNM (390120-390121, 390664-390666, 391522-391526). 1808/6312, 7 km E and 3 km N Ingeniero Mora, 13 AMNH (247758-247761, 247765-247773). Tarua 2056/6321, 2 km S and 10 km E of Tiquipa, 4 amnh (246799-246801, 246822). 21 19/6325, 8 km S and 10 km E of Villa Monies, 8 AMNH (246802-246806, 246810, 146912, 246914). 2128/6317, 35 km by road SE Villa Montes, Ta- ringuiti, 5 ummz (155937-155938, 156332- 156334). 2241/6426, Rio Lipeo, 13 ansp (18176-18188). PARAGUAY Presidente Hayes Rio Pilcomayo, 15 mi W Rio Paraguay, 3 amnh (143892-143894). - - NuEVA AsuNa6N 19 km by road WSW, km 588 Trans Chaco, 8 amnh (249255-249262). Chaco 50 km WNW Fortin Madrejon, 9 amnh (248398- 248406). ARGENTINA JUJUY Yuto, 10 AMNH (167855, 179976, 179980, 182570- 182571, 182738-182739, 183312, 185226- 185227, 185269). Sta. Barbara, 4 amnh (185224-185225, 185228, 186954). O. longicaudatus Specimens: 134. BOLIVIA Chuquisaca 1927/6407, Tola Orko, Tomina Province, 6 usnm (271588-271590, 545226-545227, 545229). 1929/6433, Horcus, 80 km SE Sucal, 3 mvz (134654-134656). 1931/6409, Monte Canto, 1 usnm (271591). ?, Chuyayacu, 1 fmnh (72889). OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS zm. CCXTHABAMBA 1620/6645, Yungas, 1 cm (5276). 1654/6642, El Choro, 1 amnh (1 19405), 12 fmnh (74902-74913), 1 bm (Thomas, 1902). 1714/6541, Incachaca, 19 amnh (38525-38542, 38550, not seen), 2 cm (5081-5082). 1734/6621, Parotani, 2 amnh (38543-38544), 1 fmnh (21668). 1742/6509, 20 mi E Totora, 1 mvz (1 19916). 1807/6509, Aiquile, 2 fmnh (50970-50971). 22 1 2/6436, 8 km by road N of Cuyambuyo, 2 UMMZ (156326-156327). 2213/6436, 4 km by road N of Cuyambuyo, 10 UMMZ (155889-155891, 156312-156313, 156315-156317, 156319-156320). Department Uncertain ?, Zapial, 1 usNM (270911). La Paz 1530/6824, Cocapunco, 1 amnh (72644). 1535/6843, Tacacoma, 1 amnh (91540). 1535/6843, Tacacoma-Sorata, 2 amnh (91541- 91542). 1539/6824, Okara, 2 amnh (72704-72705). 1543/6840, 10 km by road N of Sorata, 3 ummz (156301-156303). 1547/6840, Sorala, at base of Ml. Sorata, 4 amnh (91536-91539). 1600/6516, Charuplaya, 16 bm (Thomas, 1902). 1618/6753, Rio Aceramarca, 2 amnh (72693- 72694). 1619/6752, Nequejahuira, 3 amnh (72722, 72724- 72725). 1620/6808, Mt. Chacaltaya, 1 ummz (1 1571 1). 1620/6756, Pongo, 15 amnh (72702-72703, 72706-72715, 72726, 81283, 241612). 1621/6747, Pitiguaya, 4 amnh (72716, 72729- 72731). 1659/6749, Caracato, 5 amnh (248977-248981). 1823/6659, Poopo, 1 mnlp (not seen). POTOSi 1 844/6609, 3 km SE of Pocoata, 1 amnh (255946). 2127/6543, Tupiza, 1 bm (Thomas, 1926). Santa Cruz 1754/6429, 5 mi W Comarapa, 1 mvz (1 19917). 2005/6334, nr. Camiri, 1 cal (13803). Tarua 2131/6445, 10 mi NW Tarija, 1 mvz (1 19918). 2 1 32/64 1 2, Entre Rios, 3 USNM (27 1 4 1 1-27 1 4 1 2, 271432). O. microtis Specimens: 230. BOLIVIA Beni 1049/6525, Guayaramerin, 2 amnh (210050- 210051). 1110/6522, 4 km below Santa Cruz, 1 "amnh (211727). 1 142/65 1 6, 4 km S Santa Rosa, 4 amnh (2 1 1 729- 211730, 211791-211792). 1 200/6502, Rio Itenez, 20 km from mouth, 2 amnh (211756-211757). 1200/6506, Puerto More, Rio Itenez, 3 amnh (211758-211760). 1 2 1 3/65 1 3, Mamore River on bank opposite Cas- cajal, 16 amnh (211724-211725, 211754, not seen, 211762-211774). 1226/6511, Mamore River, 2 amnh (211721, 211761). 1229/6417, Rio Itenez, opposite Costa Marques, Brazil, 11 amnh (210122, 210038-210045, 210365-210366). 1229/6418, Rio Itenez, 1.5 km below Costa Mar- ques, Brazil, 1 amnh (210364). 1230/6415, Pampa de Meio, 2 amnh (210046- 210047). 1 230/64 1 8, Baures River mouth, 6 amnh (2 1 0028- 210031, 210382, 210383, not seen). 1240/6330, Curiche River mouth, 6 amnh (210032-210037). 1241/6432, 15 km above Horquilla on Rio Ma- chupo, 1 amnh (210053). 1244/6318, Versalles, 1 amnh (210052). 1244/6428, Las Perias, 3 fmnh (1 17079-1 17081), 1 USNM (460741). 278 FIELDL\NA: ZOOLOGY 1300/6515, Mamore River, 1 amnh (21 1722). 1304/6449, San Joaquin, 1 fmnh (1 17075), 6 usnm (364735, 364738, 391299, 460273, 460742- 460743). 1312/6410, Cachuelita, 1 usnm (460739). 1 3 1 2/65 1 5, 8 km N Exaltacion, 6 amnh (2 1 1 775- 211780). 1313/6221, 20 km W Larangeira, Bahia de los Casara, 2 amnh (210048-210049). 1313/6409, Boroica, 1 usnm (460740). 1322/6520, Palacios, province of Yacuma, 1 usnm (461082). 1324/6442, Chaco Lejo, 20 km SE San Ramon, 2 usnm (391295-391296). 1325/6435, Tacuaral, 1 usnm (391297). 1343/6521, Puerto Caballo, 2 amnh (211785, 214597). 1420/6455, 10 km W San Pedro, on Mamore Riv- er, 5 AMNH (21 1786-21 1790). 1428/6734, Rurrenabaque, 2 amnh (247774- 247775). 1434/6455, 23 km W San Javier, 1 amnh (214760). 1437/6457, Ibare River mouth, 2 amnh (21 1783- 211784). 1446/6451, Ibare River, 24 km from mouth, 1 amnh (21 1755, not seen). 1454/6422, 6 km W of Casarabe, 7 amnh (255947- 255953). 1503/6658, 1 km E of La Embocada, 2 ummz (155940, 156292). 1515/6415, El Triunfo, 1 usnm (391298). 1550/6441, 5 km NW Grande River mouth, 2 AMNH (21 1781-21 1782). ?, Centenela, 1 fmnh (1 17074). ?, "Beni" only, 1 amnh (232699). Pando 1117/6855, Rio Nareuda, 11 amnh (248982- 248990. 248993-248994). Santa Cruz 1 703/6335, 7 km N Santa Rosa, 1 amnh (25460 1 ). 1724/6344, 7 km N and 17 km W of Buena Vista, 3 AMNH (246809, 246820, 246935). BRAZIL Amazonas Solimoes, 10 amnh (37088-37097, 37091 = ho- lotype of O. microtis). Guatsue, 1 amnh (37100). Lower Solimoes, 1 amnh (37157). ParA Capim, 150 mi SE Belem, 1 amnh (188964). Capim, Est. B.R. 14, km 97, 1 amnh (203400). Rio Xingu; Porto de Moz, 1 amnh (95983). Rio Xingu; Villarinho do Monte, 4 amnh (95984- 95986, 95997). GoiAs: Rio Madeira COCHABAMBA 1648/6508, Todos Santos, 1 amnh (paratype of O. chaparensis, 40787), 1 fmnh (holotype, 21330). 1657/6523, 2 km E of Villa Tunari, 4 amnh (247662-247664, 247776). Auara Igarape, 5 amnh (91874, 91876-91878, 94245). Sta. Antonio du Uayara, 4 amnh (92258-92261). Rosarinho, Lago Miguel, 1 1 amnh (92705-92715). Rosarinho, Lago Sampaio, 2 amnh (9271 6-927 1 7). Mato Grosso La Paz 1 5 1 5/68 1 0, Mapiri, 5 amnh (7272 1 , 730 1 1 , 72732, 72697-72698). 1528/6752, Guanay, 1 amnh (72701). 1528/6818, Ticunhuaya, 1 amnh (72700). 1540/6742, 4 km NW Alcoche, 2 ummz (126777, 127167). Utiarity, Papagaio River, 3 amnh (37540-37542; 37541 = type of O. utiaritensis, 37542 = type of 6). mattogrossae). ? Amazonas S bank R. Amazon: Villa Bella Imperatriz, 2 amnh (91899-91900). OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 279 PARAGUAY Caaguazu 24 km NNW Carayao, 1 ummz (133821). Presidente Hayes 24 km NW Villa Hayes, 3 UMMZ (133830-1 33831, 133833). PERU Ayacucho Hda. Luisiana on Rio Apurimac, 1 amnh (242484). Pasco Ca. 1 0 km N Puerto Bermudez, 38 amnh (24555 1- 245588). Maldonado Barra del Arrojo Maldonado, 9 km ENE Punta del Este, 1 amnh (205994). Montevideo Rio Santa Lucia, 1 km SE Santiago Vasquez, 2 AMNH (205995-205996). ROCHA 22 km SE Lascano, 18 amnh (205997-206009, 206011-206015). San Jose Dept. only, 3 amnh (232216-232218). Soriano 3 km E Cardona, 1 amnh (206016). O. flavescens Specimens: 33. Treinta y Tres 16 km SSW Boca del Rio Tacuari, 1 amnh (206017). URUGUAY Artigas 6 km NNW Belen, 2 amnh (205986-205987). Cerro Largo Sierra de Vaz, Rio Tacuari, 20 km SE Melo, 1 amnh (205988). Lavalleja 12 km WSW Zapican, 4 amnh (205989-205905). O. nigripes Specimens: 31. BRAZIL GOIAS Anapolis, 7 amnh (134528-134530, 134532, 134534, 134538, 134540). 280 FIELDIANA: ZOOLOGY MiNAS Gerais PARAGUAY (Serra do Caparao), Rio Caparao, 8 amnh (80369- Ainambay 80370, 80372, 80375-80379). (Serra do Caparao), Boa Espera, 1 amnh (80391). 4 km by road SW Cerro Cora, 1 ummz (125523). Mato Grosso CaaguazO Maracaju, 9 amnh (134541-134542, 134544- N of Coronel Oviedo, 1 ummz (124212). 134547, 134551, 134838, 134902). 24 km NNWCarayao, 3 ummz (133835-133837). SAo Paulo Piquete, 1 amnh (36496). OLDS & ANDERSON: RICE RATS OF SUBGENUS OLIGORYZOMYS 281 New Records and Current Status of Euneomys (Cricetidae) in Southern South America Jose L. Yaftez, Juan C. Torres-M iira, Jaime R. Rau, and Luis C. Contreras ABSTRACTS New records of Euneomys from central and southern Chile are given. Specific assignment of this material is difficult, owing to the lack of good diagnostic characters, small number of specimens, and spotty distribution. Absence of clear-cut differences between Euneomys localities suggests that there is only a single species, E. chinchilloides. Subspecific assignments are deferred, because at present they can only be made geographically, and the gaps between localities seem to be more apparent than real. Se citan nuevos registros de Euneomys para Chile Central y Sur. La determinacion es- pecifica de este material es dificil debido a la carencia de buenos caracteres diagnosticos, al bajo niimero de especimenes y a la distribucion localizada con grandes hiatos intermedios. La ausencia de claras diferencias entre Euneomys de distintas localidades sugiere la existencia de tan solo una especie, E. chinchilloides. Las nominaciones subespecificas se posponen, ya que al moment© estas solo se pueden realizar basadas en las localidades geograficas y los hiatos parecen ser mas aparentes que reales. Sao citados novos registros de Euneomys para o Sul e Centro do Chile. Dada a falta de bons carateres diagnosticos, o baixo numero de especimes, e a distribui^ao localizada com hiatos intermediarios, e dificil a determina9ao esfiecifica deste material. A ausencia de caracteristicas distintas entre Euneomys de localidades diferentes sugere a existencia de uma especie apenas: E. chinchilloides. Nao foram designadas nomina96es subespecificas ja que, no momento, estas so poderiam ser realizadas na base das localidades geograficas, e os hiatos podem ser mais aparentes do que reais. Introduction The genus Euneomys is one of the least studied cricetid rodents in southern South America. These herbivorous, volelike rodents have a relatively large, heavy body and a short tail and ears (Os- good, 1943). Euneomys has close affinities with the phyllotine and sigmodontine groups (Hersh- From Museo Nacional de Historia Natural, Casilla 787, Santiago, Chile (Yanez); Dei>artamento Biologia y Quimica, Facultad de Ciencias, Universidad de Talca, Casilla 747, Talca, Chile (Torres-Mura and Contreras); and Estacion Biologica de 'Doiiana' (Sevilla), Apartado 1056, Sevilla 41013, Espana (Rau). kovitz, 1 962; Mann, 1978). The genus has a known distribution from 33°00'S to Cape Horn (fig. 1), mainly along and to the east of the Andes (Hersh- kovitz, 1962; and subsequent reports). In Chile Euneomys is known from Lago Fag- nano and Hermite Island (Tierra del Fuego), Punta Arenas (Magallanes), Torres del Paine and Laguna Lazo (Ultima Esperanza), Puerto Ibaiiez (General Carrera), Campo Bandera (Aysen), Pino Hachado (Malleco), and Lo Valdes and Farellones (Santia- go) (Osgood, 1943; Mann, 1944, 1978; Markham, 1971; Greer, 1965; Reise & Venegas, 1974; Pine et al., 1979). In Argentina Euneomys is certainly known from Bahia del Buen Suceso (Tierra del Fuego), upper Rio Chico (Santa Cruz), Paso de YANEZ ET AL.: NEW RECORDS AND STATUS OF EUNEOMYS 283 Fig. 1. Collecting localities of Euneomys. 1, Farellones; 2, Lo Valdes; 3, Bartos del Flaco; 4, Pino Hachado; 5, Lastarria; 6, Puerto Ibariez; 7, Torres del Paine; 8, Rio Baguales and Cueva del Milodon; 9, Estrecho de Magallanes and Palli Aike; 10, Punta Arenas; 11, Lago Fagnano; 12, Hermite Island; 13, Grevy Island; 14, San Rafael; 15, Cueva Traful; 16, Cerro Leones; 17, Paso de los MoUes, Pilcaniyeu; 18, Upper Rio Chico, Santa Cruz; 19, Bahia del Buen Suceso. Open circles represent new localities. 284 FIELDIANA: ZOOLOGY Table 1. Mandibular measurements of Euneomys found in bam owl pellets from Lastarria and Baiios del Flaco and from animals trapped in Farellones near Santiago. Character Farellones* BaAos del Flaco Lastarria MH-1 4.07 ± 0.12(3.8-4.1) 4.1, 3.9 2.6,t4.0, 3.1t MHC 8.8 ± 0.44(8.1-9.1) 9.2, 9.1 7.9,6.1 MAL 6.07 ± 0.08 (5.9-6.2) 6.1, 6.1 5.9,6.0,6.1 MDL 4.03 ± 0.24 (3.6-4.4) 3.6, 4.1 3.6,3.9,3.1 SL 6.24 ± 0.16(6.1-6.5) 6.2, 6.3 5.8, 6.0, 5.5 MH- 1 = Mandibular height at the first molar; MHC = mandibular height from condyle to angular process; MAL = mandibular alveolar length; MDL = mandibular diastema length; SL = symphysis length. Values from Farellones near Santiago are X ± 2 SE (and range; N = 7). Individual measurements are given for Baiios del Flaco and Lastarria. * LCM collection, f Juvenile, the last cheektooth is not fully erupted. Los MoUes, Pilcaniyeu, and Cerro Leones (Rio Negro), Cueva Traful (Neuquen), and San Rafael (Mendoza) (Hershkovitz, 1962; Pine et al., 1978; Pearson & Pearson, 1982; Massoia, 1982). In this paper we report new records of Euneomys in Chile and review the scant literature on the taxonomy and geographic distribution of this poorly known genus. New Records and Current Status Some new records are from bam owl (Tyto alba) pellets from Lastarria (near Temuco), 39°14'S, 70°40'W, and from Bafios del Flaco (near San Fer- nando), 34°57'S, 70°26'W. In addition, six other specimens of Euneomys were recently captured on Grevy Island (Tierra del Fuego; 55°32'S, 67°37'W) and at Palli Aike (Magallanes; 52°25' S, 69''42'W) and Rio Baguales (Ultima Esperanza; 51°02'S, 72°30'W); all are deposited in the Museo del Ins- tituto de la Patagonia (MIP), Punta Arenas. Data obtained from these new specimens have been in- corporated by locality into Table 2. Determination of this material to genus was not difficult, using Reise's (1973) key and studies by Hershkovitz (1962) and Mann (1978). However, determination to species is difficult, owing to a lack of good diagnostic characters, the small num- ber of specimens, and their spotty distribution. Osgood recognized two species of Euneomys in Chile: E. chinchilloides (with two subspecies) in Tierra del Fuego and Magallanes and E. petersoni in Ultima Esperanza and Aysen. Subsequently, Mann ( 1 944) described E. noei from Lo Valdes in the Andes near Santiago. Hershkovitz (1962) con- sidered E. noei as doubtfully separable from E. mordax, with type locality at San Rafael (Men- doza Province, Argentina), and thought that mor- daxwas probably a subspecies of £". chinchilloides. Greer (1965) referred his four Malleco specimens to E. chinchilloides petersoni, despite their greater similarity to noei or mordax in measured char- acters. Miller and Rottmann (1976) identified all their Euneomys from near Santiago as E. mordax. Mann (1978) thought all Chilean specimens rep- resented a single polytypic species: E. chinchil- loides chinchilloides from Grande de Tierra del Fuego Island and the adjacent mainland, E. c. petersoni from the eastern Andes of Ultima Es- peranza and Aysen, and E. c. noei from the Andes outside Santiago. Subsequently, Pine et al. (1979) suggested that two sympatric species of Euneomys occur in Farellones near Santiago, but declined to identify the other species until comparisons could be made with specimens representing the entire range of the genus. Karyotypes of nine specimens from Farellones show insignificant differences (Spotomo, pers. comm.). Thus, Hershkovitz ( 1 962) and Mann (1978) rec- ognized only one polytypic species, whereas Pine et al. (1979) and Tamayo and Frassinetti (1980) regarded specimens from different areas as differ- ent species. Tamayo and Frassinetti ( 1 980) thought Greer's Malleco specimens also represented Eu- neomys sp. Unfortunately, this taxonomic uncer- tainty is difficult to dispel, as only seven intact skulls of Euneomys from the Andes outside of Santiago are deposited in Chilean collections (La- boratorio Citogenetica de Mamiferos [LCM], Fac. Medicina, Universidad de Chile), and none is de- posited in collections in Mendoza or San Rafael, Argentina (R. A. Ojeda, pers. comm.; H. A. La- giglia, pers. comm.). Euneomys material recently obtained from owl pellets and deposited in the Museo Nacional de Historia Natural de Chile is incomplete, consisting YANEZ ET AL.: NEW RECORDS AND STATUS OF EUNEOMYS 285 ■5 E p I e 3 c .t: ea — u 3 00 5 e — a u 0^ J= SJ^ « o g « « aS 1^ 1^ §2 « « * B « e 0Q « k. sn a> ae •2 H g 1^ , ou sgood, 1943; Hcrshkovitz, 1962; Pineetal. 1978; MIP col- i Pine tion c g 1 |8 ann, 1944; : etal.. 1979; LCM collec "3 c O o O 2 du X •* o vO r- o Is Is 1 5! IE 1 c 00 q 2 3; r-' q 2 is 1 S r-' 1 2 2 ctS 4S op >r> 1 = o c "? 00- ^ £ E| « -5 — -c i r% \0 *Ys 7^ 7° >0 ^^ CT> ^^ O 3 vO o o — — fS Si's 5 ^ o ^ U o 3 S (N ,_ O ^ ^' r<^ 1 r<^ 00 >0 ^^ CM 4- E? CO ^ o J:; c 5 286 HELDIANA: ZOOLOGY mainly of mandibles and maxillae. The limited samples and their disintegration make a statistical analysis impossible. Comparisons of the owl pellet material with the seven specimens collected near Santiago (table 1) show no differences. The same results are obtained when comparisons are ex- tended to include all known Chilean specimens (table 2). Using these criteria it is impossible for us to assign specimens from Lastarria and Banos del Flaco to either E. c. mordax or E. c. petersoni. Conclusion The absence of clear-cut differences between lo- calities in Table 2 suggests that there is only a single species represented, E. chinchilloides. Be- cause subspecific assignments at present can only be made geographically, treatment of subspecies is deferred until an adequate sample of specimens is available. The fact that Euneomys are not taken in traps at the same localities where they are found in owl pellets (e.g., Reise & Venegas, 1976; Pear- son &. Pearson, 1982; present study) suggests that current gaps in the distribution o^ Euneomys might be attributable to sampling error rather than to truly disjunct distributions. The new records re- ported here fill both morphological and geograph- ical gaps. Acknowledgments We thank Angel Spotomo and Claudio Venegas for making specimens of Euneomys available to us. Bruce D. Patterson, Ronald H. Pine, and an anonymous reviewer made valuable comments. Mrs. Veronica Aguirre typed the manuscript, and Luz Uribe assisted with the figure. This report is dedicated to P. Hershkovitz in recognition of his pioneering contributions to our understanding of Neotropical mammals. Literature Cited Greer, J. K. 1965. Mammals of Malleco Province, Chile. Publications of The Museum, Michigan State University, Biological Series, 3: 51-151. Hershkovitz, P. 1962. Evolution of Neotropical cric- etine rodents (Muridae). Fieldiana: 2kx)logy, 46: 1- 524. Mann, G. 1 944. Dos nuevas especies de roedores. Bio- logica (Santiago), 1: 95-126. . 1978. Los pequenos mamiferos de Chile. Gay- ana: Zoologia, 40: 1-342. Markham, B. 1971. Catalogo de los anfibios, reptiles, aves y mamiferos de la Provincia de Magallanes, Chile. Publicaciones Instituto de la Patagonia, Serie Mono- grafias N2. 3, 64 pp. Massoia, E. 1982. Restos de mamiferos recolectados en el paraje Paso de Los Molles, Pilcaniyeu, Rio Negro. Revista Investigaciones Agropecuarias INTA, Buenos Aires, 17: 39-53. Miller, S. D., and J. Rottmann. 1976. Guia para el reconocimiento de mamiferos chilenos. Ed. Gabriela Mistral, Santiago, 200 pp. Osgood, W. H. 1943. The mammals of Chile. Field Museum of Natural History, Zoological Series, 30: 1- 268. Pearson, O. P., and A. K. Pearson. 1982. Ecology and biogeography of the southern rainforests of Ar- gentina, pp. 129-142. In Mares, M. A., and H. H. Genoways, eds.. Mammalian Biology in South Amer- ica. A Symposium Held at the Pymatuning Laboratory of Ecology, May 10-14, 1981. Sf>ecial Publication Se- ries, Pymatuning Laboratory of Ecology, University of Pittsburgh, 6: 1-539. Pine, R. H., J. P. Angle, AND D. Bridge. 1978. Mam- mals from the sea, mainland, and islands at the south- em tip of South America. Mammalia, 42: 105-1 14. Pine, R. H., S. D. Miller, and M. L. S. Schamberger. 1979. Contributions to the mammalogy of Chile. Mammalia, 43: 339-376. Reise, D. 1973. Clave para la determinacion de los craneos de marsupiales y roedores chilenos. Gayana: Zoologia, 27: 1-20. Reise, D. L., and W. Venegas. 1974. Observaciones sobre el comportamiento de la fauna de microma- miferos en la Region de Puerto Ibaiiez (Lago General Carrera), Aysen, Chile. Boletin Sociedad de Biologia, Concepcion, Chile, 47: 71-85. Tamavo, M., and D. Frassinetti. 1980. Catalogo de los mamiferos fosiles y vivientes de Chile. Boletin Museo Nacional de Historia Natural, Chile, 37: 323- 399. YANEZ ET AL.: NEW RECORDS AND STATUS OF EUNEOMYS 287 Morphological Variation, Karyology, and Systematic Relationships of Heteromys gaumeri (Rodentia: Heteromyidae) Mark D. Engstrom, Hugh H. Genoways, and Priscilla K. Tucker ABSTRACTS Morphological variation was assessed within and among populations of Heteromys gaumeri using univariate and multivariate statistical analyses of external and cranial measurements. Although patterns and amount of nongeographic variation in H. gaumeri were similar to other heteromyines, geographic variation was relatively conservative. Mean values of most characters were statistically homogeneous among localities and spatially unpattemed. Consequently, no association was found between levels of within- and among-sample variation for individual characters (the "Kluge-Kerfoot phenomenon"). Populations of//, gaumeri were chromosomally monomorphic. The lack of morphological and chromosomal variation in //. gaumeri contrasts sharply with patterns in other heteromyines. Heteromys gaumeri is morphologically and chro- mosomally distinct from the //. desmarestianus species group (to which it is currently assigned) and appears to share some primitive characters with Liomys (the sister group of Heteromys). We recommend that H. gaumeri be removed from the H. desmarestianus group. La variacion morfologica intra e interpoblacional de Heteromys gaumeri fue evaluada usando analisis estadisticos univariados y multivariados de medidas extemas y craneales. A pesar de que los patrones y cantidad de variacion intrapoblacional en //. gaumeri fue similar a la de otros heterominos, la variacion geografica fue relativamente conservadora. Los valores pro- medio de la mayoria de los caracteres fueron estadisticamente homogeneos entre las localidades, sin mostrar ningun patron de variacion espacial. En conservencia, no se encontro asociacion alguna entre los niveles de variacion intra e interpoblacional para caracteres individuals ("fen- omeno Kluge-Kerfoot"). Las poblaciones de //. gaumeri fueron monomorficas cromosomica- mente. La falta de variacion tanto morfologica como cromosomica en H. gaumeri contrasta marcadamente con los patrones encontrados anteriormente para otros heterominos. Heteromys gaumeri es morfologica y cromosomicamente distinguible del grupo //. desmarestianus (al cual se asigna actualmente) y aparentemente comparte algunos caracteres primitivos con Liomys (el grupo hermano de Heteromys). Nosotros recomendamos que se remueva a H. gaumeri del grupo //. desmarestianus. Avalia-se a variafao morfologica intra- e interpopulacional de Heteromys gaumeri, atraves de analises estatisticas uni- e multivariadas de medidas extemas e craniais. Apesar dos padroes, e da quantidade de varia^ao intrapopulacional em H. gaumeri serem similares aos de outros heteromideos, a varia9ao geografica e relativamente conservadora. Os valores medios da maior parte dos carateres examinados sao estatisticamente homogeneos entre as localidades, e nao From Department of Biology, Angelo State Univer- sity, San Angelo, TX 76909 (Engstrom); Section of Mammals, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pittsburgh, PA 1 5206 (Genoways; pres- ent address. University of Nebraska State Museum, 212 Morrill Hall, Lincoln, NE 68588); and Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, TX 77843 (Tucker, present address, The Jackson Laboratory, Bar Harbor, ME 04609). ENGSTROM ET AL.: HETEROMYS GA UMERI 289 surgiu nenhum padrao de variances locals. Consequentemente, nao foram encontradas asso- cia96es entre os niveis de varaiafoes intra- e interpopulacionais para carateres individuals (o "fenomeno KJuge-Kerfoot"). Popula^oes de H. gaumeh mostraram-se cromossomicamente monomorficas. A falta de varia9ao morfologica ou cromossomica em H. gaumeri e altamente contrastante aos padroes encontrados em outros heteromideos. Heteromys gaumeri dlstingue- se tanto morfologica quanto cromossomicamente do grupo H. desmarestianus. ao qual esta atualmente deslgnado, e aparentemente possue carateres primitivos em comum com Liomys— grupo irmao de Heteromys. Recomendamos que H. gaumeri seja removido do grupo H. des- marestianus. Introduction Spiny mice of the genus Heteromys (Hetero- myidae: Heteromyinae) are Neotropical rodents that typically occur in rainforest and cloudforest habitats from east-central Mexico south to north- em South America. Currently, the genus is par- titioned into two subgenera {Xylomys and Hetero- mys), and two species groups {desmarestianus and anomalus groups) are recognized in the nominate subgenus (Goldman, 1911; Hall, 1981; Rogers & Schmldly, 1982). Hall and Kelson (1959) and Hall (1981) noted that taxonomlc relationships within Heteromys and in particular the H. desmaresti- anus species group were problematical and in need of revision. Recently, Rogers and Schmldly ( 1 982) reviewed morphological variation among repre- sentatives of the H. desmarestianus species group from northern Middle America, exclusive of H. gaumeri, and concluded that only two species (//. desmarestianus and H. goldmani) were represent- ed in the material they examined. Heteromys gaumeri, the third species currently recognized in the desmarestianus group, is endem- ic to the Yucatan Peninsula. Ecologically this species characteristically occurs in relatively dry deciduous and subdeciduous-subperennlal forest. Systematic relationships of H. gaumeri are enig- matic. In his review of the subfamily Heteromyi- nae, Goldman (1911) placed H. gaumeri in his desmarestianus group for convenience, but noted (p. 29) that it was "not closely related to any known species." In this paper, we review morphological and chromosomal variation within H. gaumeri and comment on the systematic relationships of this species to the H. desmarestianus species group and the subgenus Heteromys. Materials and Methods A total of 322 specimens o{ Heteromys gaumeri was examined in the morphological analyses. Specimens examined are listed in the Appendix, and institutions housing those specimens are as follows (abbreviations for North American col- lections follow Choate &. Genoways, 1 975): Amer- ican Museum of Natural History (amnh); British Museum (Natural History), London (bmnh); Field Museum of Natural History (fmnh); Museum of Natural History, University of Kansas (ku); James Ford Bell Museum of Natural History, University of Minnesota (mmnh); Royal Ontario Museum (ROM); Texas Cooperative Wildlife Collections, Texas A«&M University (tcwc); The Museum, Texas Tech University (ttu); Museum of Zoology, University of Michigan (ummz); Unlversidad Na- cional Autonoma de Mexico (unam); and National Museum of Natural History (usnm). Each specimen was assigned to one of five pre- sumptive age classes (I-V, from youngest to oldest) based on pelage characteristics and state of erup- tion and relative wear of the maxillary toothrow, following Genoways (1973). For each adult and selected subadult specimens, the following four external (recorded from specimen labels) and 10 cranial measurements (measured to the nearest 0. 1 mm using dial calipers) were taken, as defined by Genoways (1973): total length (TL); length of tail (TV); length of hind foot (HF); length of ear (LE); greatest length of skull (GLS); zygomatic breadth (ZB); interorbital constriction (IOC); mastoid breadth (MB); length of nasals (LN); length of ros- trum (LR); length of maxillary toothrow (MTR); depth of braincase (DBC); Interparietal width (IW); and interparietal length (IL). Each measure- ment was chosen because it was examined in pre- vious studies of variation in other heteromyine rodents and was geographically variable in some taxa (Goldman, 1911; Genoways, 1973; Rogers & Schmldly, 1982). Initially, four qualitative cranial characters and dorsal coloration also were examined, as described by Genoways (1973) for Liomys (the sister group of Heteromys). Unlike patterns found in species of Liomys, each of these characters was equally or 290 FIELDIANA: ZOOLOGY more variable within populations of H. guumeri as among populations, and the characters varied in no apparent geographic pattern. Consequently, qualitative cranial characters and dorsal color- ation were not analyzed further. Nongeographic variation in the 14 mensural characters was examined in one sample of 94 spec- imens collected near Campo Experimental Fores- tal "El Tormento," 7.5 km W Escarcega, Cam- peche. These specimens were collected within a two square kilometer area of transitional tropical evergreen-tropical deciduous forest and were here considered to represent a single population. All calculations were performed using subprograms of the Statistical Analysis System (SAS; SAS Insti- tute, Inc., 1982). Standard statistics (mean, range, standard deviation, standard error, coefficient of variation, skewness, and kurtosis) were calculated for each variable within each subgroup (MEANS and UNIVARIATE procedures). Student's / test (or an approximation of variances were unequal) was used to test for significant differences between sexes within each age class (TTEST procedure). A model I, one-way analysis of variance (ANOVA) was used to test for significant differences among age classes with sexes pooled (GLM procedure). Subsequently, a Duncan's multiple range test (DUNCAN) was used to determine maximally nonsignificant subsets of age classes. Straney (1978) criticized the use of F tests in unbalanced ANOVAs to determine patterns of nongeographic variation. To augment hypothesis testing, variance partitioning of a model II, two- way ANOVA (VARCOMP procedure) was used to estimate the relative contributions of sex, age, sex by age interaction, and error to within sample variation. Age and sex were considered random factors (see discussion in Leamy, 1983) and the percent contribution of each factor was estimated from variance components. The main effects (sex and age) in the two-way ANOVAs were not in- dependent because the data were unbalanced (Searle, 1971). Consequently, ANOVAs were run with sex entered into the model first, then again with age entered first. For analysis of geographic variation of mensural data, adult specimens were assigned to one of 1 1 grouped localities to increase sample size (fig. 1). In no instance did a grouped locality cross a major physiographic or previously recognized taxonomic boundary. The specific geographic composition of samples is as follows: Group 1— Chuntuqui and Laguna de Sotz (= Zotz), Peten, Guatemala; 103 km SE Escarcega (= Francisco Escarcega), Cam- peche, Mexico. Group 2 — 7.5 km W Escarcega, Campeche, Mexico. Group 3— Apazote, 7 km N, 5 1 km E Escarcega, and La Tuxpeiia, Campeche, Mexico. Group 4— Dzibalchen and San Jose Car- pizo, Campeche, Mexico. Group 5— Esmeralda and Santa Rosa, Yucatan, Mexico. Group 6— Chichen Itza and Piste, Yucatan, Mexico. Group 7— Ti- zimin and Tunkas, Yucatan, Mexico. Group 8 — La Vega, Pueblo Nuevo Xcan, and Puerto Mo- relos, Quintana Roo, Mexico. Group 9— Felipe Carrillo Puerto, Quintana Roo, Mexico. Group 10— Bacalar, Quintana Roo, Mexico. Group 1 1 — Kate's Lagoon and Rockstone Pond, Belize, Be- lize. For each mensural character, we analyzed two aspects of geographic variation: (1) statistical het- erogeneity of mean values among geographic sam- ples, and (2) significant departure of means from spatial randomness (see Sokal & Oden, 1978). The null hypothesis of homogeneity of means among grouped localities for each character was tested using a model I, one-way ANOVA (SAS:GLM). By ANOVA, the variance of each character was partitioned into among and within (error) locality effects, and the percentage of variation attributable to each effect was estimated from variance com- ponents (SAS: VARCOMP). Homogeneity among grouped localities across all characters was ex- amined using a multivariate analysis of variance (SAS:GLM, MANOVA). The null hypothesis of no geographic pattern among grouped locality means was examined by testing for significant association between geo- graphic and phenetic distance matrices. For each character, a phenetic distance matrix was con- structed in which the elements were calculated as the absolute differences between means for all pairs of localities (Sokal, 1979). Multivariate phenetic matrices of taxonomic distance (Sneath & Sokal, 1973) were calculated for all 14 mensural char- acters and a set restricted to those showing sig- nificant heterogeneity, using the NT-SYS library of computer programs (Rohlf et al., 1982). Ele- ments of the geographic distance matrix were the actual map distances (in km) between all pairs of grouped localities (taken from the center of each grouped locality); all connections between pairs of localities were maintained because there are no obvious physiographic or ecological barriers to gene flow among populations of//, gaumeri on the Yu- catan Peninsula. Three test statistics (Mantel's Z; Spearman's rho, R; and a component of Kendall's tau, K^) were used to test for significant association between each phenetic distance matrix and the ENGSTROM ET AL.: HETEROMYS GAUMERI 291 90 21 18 87 "T" KILOMETERS 0 50 100 I H r-" 25 50 MILES ± 21 18 90 87 Fig. 1 . Geographic distribution of Heteromys gaumeri and approximate geographic areas included in the 1 1 grouped locahties used in analyses of morphological variation. Closed circles denote specimens used in analyses of geographic variation and open circles denote specimens examined but not included in statistical analyses. geographic distance matrix (Dietz, 1983). Values of P associated with each statistic were estimated from 2,000 random permutations using a FOR- TRAN program supplied by E. J. Dietz. In the chromosomal analysis, standard karyo- types were examined for 26 specimens oi Hetero- mys gaumeri. Additionally, karyotypes of one specimen of H. desmarestianus and 20 specimens 292 HELDIANA: ZCX)LOGY of H. anomalus were examined for comparison. Standard karyotypes were prepared from speci- mens sampled from natural populations using the in vivo bone marrow technique of Patton (1967), as modified by Lee (1969). Terminology regarding relative chromosome arm ratios is that of Patton (1967). In the calculation of Fundamental Num- bers (FN), relative chromosome arm ratios were scored conservatively. Several chromosomes re- corded as acrocentric here had telomeric knobs of chromatin in elongated preparations. Only chro- mosomes which consistently displayed second arms of chromatin, regardless of state of contrac- tion of the preparation, were scored as biarmed. The following voucher specimens are deposited in the Texas Cooperative Wildlife Collection, Texas A&M University, and the Carnegie Museum of Natural History (sample size in parentheses): Het- eromys gaumeri (total 26)— MEXICO. Campeche: 7.5 km W Escarcega (8). Quintana Roo: 2 km N, 8 km W Bacalar (8); 8 km NNE Felipe Carrillo Puerto (3); 2.5 km NNE Felipe Carrillo Puerto (1). Yucatan: Chichen Itza (1); Cenote Seco, 2 km E Chichen Itza (5). Heteromys desmarestianus (total l)-MEXICO. Chiapas: 9.4 km S Palenque (1). Heteromys anomalus (total 20)— VENEZUELA. Miranda: 25 km N Altagracia de Oricuto (6); 24 km N Altagracia de Oricuto (1). Monagas: Cari- pito (4). Sucre: 40 km SW Caripito (9). Results Morphological Variation NoNGEOGRAPHic VARIATION— The Sample of Heteromys gaumeri (N = 94) from 7.5 km W Es- carcega, Campeche, was used to estimate within- sample variation of the 1 4 mensural characters. Two approaches were taken, hypothesis testing and estimation of variance components. Initially, / tests were used to test for significant differences between sexes in each of age classes I- IV. Males average larger than females for most characters in most age classes; however, significant differences between sexes were found for only one measurement in age class I (length of rostrum) and seven measurements in age class III (table 1). Results of ANOVA among age classes with sexes pooled are given in Table 1 . Separate analyses of the age effect within each sex gave results similar to those in Table 1 and are not presented. Signif- icant variation with age was found in each of the 14 measurements. Each measurement tended to increase with age, although this pattern was not pronounced for length of maxillary toothrow. In the a posteriori DUNCAN analysis, age class V averaged significantly larger than the other age classes for two characters (interorbital constriction and interparietal length), whereas age classes IV- V formed a homogeneous subset for the remaining 12 characters. Patterns of variation among age classes I-III were less consistent, although age classes I-III differed significantly from each other and age classes IV-V in six of 10 cranial mea- surements. To complement hypothesis testing, the relative contributions of sex, age, sex by age interaction, and error (residual variation) to a two-way AN- OVA were estimated from variance components (table 2). Separate analyses with either sex or age entered into the model first generally yielded re- sults that differed by 1% to 4%. Because of the similarity of these estimates (and because sex most often is analyzed before age in studies of nongeo- graphic variation), only the results in which sex was entered into the model first are presented. Most of the variation (average 97%) in the AN- OVA was attributable to the effects of age and error. Age contributed the largest proportion of variance for most characters (average 53%); error was nearly as important, contributing an average of 44%. For length of hind foot, length of maxillary toothrow, interparietal width, and interparietal length, age contributed a relatively small propor- tion of the variance, and variation was mostly attributable to error. Homogeneity of estimates across characters was examined using z-transfor- mations and a subsequent c/z/-square test (Sokal &, Rohlf, 1981). Estimates of age and error were significantly heterogeneous across characters. Sex and interaction were relatively unimportant factors in the ANOVA for most characters, con- tributing an average of 1% and 2% of the variance, respectively. The only noteworthy exception to this pattern was interparietal length, for which inter- action accounted for 23% of the variance. Despite this exception, estimates of the effects of sex and interaction were statistically homogeneous across characters. Given that only a small proportion of the variance of each character was attributable to the effect of sex, the significant sexual dimorphism found for seven of 1 4 characters in age class III in preliminary / tests (table 1) probably was due to trivial differences accentuated by large sample size (N = 47). For most characters, the largest pro- portion of variance was attributable to age and ENGSTROM ET AL.: HETEROMYS GAUMERI 293 Table 1. Age variation in 14 external and cranial measurements of //er^romy^^aM/wer/ from 7.5 km WEscarcega, Campeche, Mexico. Age class N Mean (range) ± 2SE CV Total Length V 2 265.0 (237.0-272.0) + 10.00 2.7 IV 14 261.2(235.0-288.0) ± 9.82 7.0 III* 42 242.6(186.0-275.0) + 5.87 7.8 II 1 17 220.6 (200.0-252.0) + 6.70 6.3 I 1 4 190.3(179.0-210.0) ± 13.60 7.1 F= 19.45*** Length of Tail IV 14 145.5(137.0-159.0) ± 4.58 5.9 V 2 139.5(139.0-150.0) ± 1.00 0.5 HI* 42 132.5 (87.0-152.0) ± 3.65 8.9 U 17 120.1(101.0-136.0) ± 4.62 7.9 I 1 4 99.8 (96.0-117.0) ± 6.85 6.9 F = 20.27*** Length of Hind Foot V 3 35.0 IV 16 34.2 (32.0-36.0) ± 0.66 3.9 in* 47 33.2 (30.0-35.0) ± 0.40 4.1 n 19 33.0 (30.0-35.0) ± 0.70 4.6 " I 1 6 31.0 (29.0-34.0) ± 1.71 6.8 F = 6.76*** Length of Ear * V 3 17.7 (17.0-18.0) ± 0.67 3.3 IV 16 17.4 (17.0-19.0) ± 0.31 3.6 in 47 16.1 (14.0-18.0) ± 0.26 5.6 u 19 15.6 (14.0-18.0) ± 0.44 6.1 I 1 6 14.8 (14.0-16.0) ± 0.61 5.1 F= 15.77*** Greatest Length of Skull V 2 35.8 (35.5-36.0) ± 0.50 1.0 IV 15 35.2 (33.9-37.5) ± 0.49 2.7 HI* 1 47 32.9 (30.2-34.5) ± 0.26 2.7 n 1 19 , 31.2 (29.6-34.2) ± 0.50 3.5 I 1 6 28.8 (27.8-29.8) ± 0.65 2.7 F= 71.20*** Zygomatic Breadth V 3 16.3 (16.1-16.5) ± 0.23 1.2 IV 16 16.1 (15.3-16.7) ± 0.19 2.4 ni* 1 44 15.1 (14.0-16.3) ± 0.14 3.1 n 1 .19 14.4 (13.7-15.1) ± 0.20 3.0 I 1 6 13.5 (12.9-13.8) ± 0.33 3.0 F = 59.28*** Interorbital Constriction V 1 3 9.3 (9.2-9.5) ± 0.20 1.9 IV 1 16 8.7 (8.3-9.0) ± 0.10 2.4 ni* 1 47 8.2 (7.5-9.3) ± 0.12 4.8 n 1 19 7.8 (7.2-8.6) ± 0.16 4.3 I 1 6 7.5 (7.0-7.7) ± 0.20 3.3 F= 29.71*** Mastoid Breadth V 3 15.5 (15.3-15.8) ± 0.30 1.7 IV 15 15.3 (14.8-15.9) ± 0.18 2.3 III 1 47 14.7 (14.0-15.4) ± 0.10 2.3 294 FIELDIANA: ZOOLOGY Table 1. Continued. Age class N Mean (range) ± 2 SE CV 19 6 F = 27.94*** 14.4 (13.4-15.2) ± 0.21 3.2 13.7 (13.1-14.1) ± 0.29 2.6 V IV III II I Length of Nasals 2 14.5 (14.0-15.0) ± 1.00 4.9 16 14.3 (13.2-15.6) ± 0.37 5.1 47 13.0 (11.8-14.2) ± 0.18 4.7 19 11.9 (10.8-13.4) ± 0.29 5.3 6 10.5 (10.0-10.9) ± 0.26 3.0 F = 59.06*** V IV III* II I* Length of Rostrum 2 15.6 (15.1-16.0) ± 0.90 4.1 16 15.2 (14.2-16.3) ± 0.27 3.6 47 14.0 (12.7-15.3) ± 0.18 4.3 19 12.9 (11.9-13.9) ± 0.23 3.9 6 11.6 (11.0-12.1) ± 0.37 3.9 F = 65.77*** IV II V III I Length of Maxillary Toothrow 16 4.9 (4.7-5.2) ± 0.07 3.0 17 4.8 (4.5-5.1) ± 0.08 3.3 3 4.8 (4.7-4.8) ± 0.07 1.2 46 4.7 (4.3-5.2) ± 0.05 3.4 5 4.7 (4.5-4.8) ± 0.17 2.8 F = 3.72** V IV III II I Depth of Braincase 3 9.1 (8.9-9.4) ± 0.29 2.8 16 9.0 (8.5-9.5) ± 0.13 3.0 46 8.8 (8.3-9.7) ± 0.08 3.3 19 8.5 (7.9-8.8) ± 0.13 3.3 6 8.5 (8.3-8.8) ± 0.16 2.3 F= 12.68*** V IV II III I Interparietal Width 3 9.3 (8.9-9.8) ± 0.54 5.1 15 8.8 (8.0-10.0) ± 0.31 6.8 19 8.6 (7.6-9.4) ± 0.21 5.3 45 8.5 (7.3-9.7) ± 0.17 6.7 6 8.0 (7.3-8.6) ± 0.36 5.6 F=4.11** V IV III II I Interparietal Length 3 5.5 (5.1-6.1) ± 0.61 9.6 15 5.1 (4.2-5.8) ± 0.21 8.0 45 5.0 (4.2-5.7) ± 0.10 6.8 19 4.9 (4.2-5.3) ± 0.14 6.3 6 4.7 (3.8-5.3) ± 0.45 11.9 F= 3.23* Vertical lines alongside age classes denote nonsignificant subsets. Asterisks after F statistics indicate levels of significance (* P < 0.05; ** P < 0.01; *** P < 0.001). Males and females in age classes I-IV were used to test for significant mean differences due to sex (/ test). Asterisks following age classes indicate significant sexual dimorphism {P < 0.05). ENGSTROM ET AL.: HETEROMYS GAUMERI 295 Table 2. Percentage of total variation attributable to sex (S), age (A), sex by age interaction (S x A), and error (E) for 14 external and cranial measurements of Hetero- mys gaumeri from 7.5 km W Escarcega, Campeche, Mexico. Char- Variance components acters* S A S X A E TL 0 61.3 1.5 37.2 TV 0 63.6 0 36.4 HF 4.8 29.9 0 65.3 LE 0 50.3 0 49.7 GLS 0 84.1 0 15.9 ZB 0 77.5 4.6 17.9 IOC 0 66.7 0.5 32.8 MB 0 64.1 2.0 33.9 LN 0 79.7 1.4 18.9 LR 0 82.1 0.8 17.1 MTR 2.2 21.1 0 76.7 DBC 1.2 42.4 0.6 55.8 IW 3.9 19.1 0 77.0 IL 0 1.1 22.7 76.2 Mean 0.9 53.1** 2.4 43.6** * Abbreviations of characters are defined in text. Es- timates were calculated from variance components (con- sidering sex as a random factor). ** Significant heterogeneity {P < 0.01) of estimates among characters. (TL) of the total variance. For those characters that were significantly heterogeneous, an average of 1 7% of the variance was attributable to locality. In a MANOVA across all mensural characters, locaHties were marginally heterogeneous {P < 0.05) by two test statistics (Hotelling-Lawley Trace and Wilk's Criterion), but homogeneous by a third test statistic (Pillai's Trace), suggesting an overall lack of mensural differentiation among localities. Because significant spatial patterning is theo- retically possible (although unlikely) even with sta- tistically homogeneous means (Sokal & Oden, 1 978), each character was tested for significant de- partures from spatial randomness. We found no significant association between matrices of phe- netic and geographic distance for any character by any of three test statistics. Multivariate taxonomic distance matrices calculated by using all characters and using only those characters that were signifi- cantly heterogeneous among localities were also incongruent with geographic distance. According- ly, mean values of most mensural characters were homogeneous among grouped localities and not spatially patterned as a simple function of geo* graphic distance. error, and the main effects in the model (sex and age) were independent. Variance components also were estimated from a two-way ANOVA restricted to age classes IV and V (results available on request). In this anal- ysis, the pattern of variance partitioning changed considerably. For each character, error contrib- uted the largest proportion of variance (average 64%), whereas sex, age, and interaction had small- er average contributions ( 1 0%, 1 2%, and 1 4%, re- spectively). These estimates, however, should only be regarded as approximations because of the small sample size of age class V (N = 3). Estimates de- rived from larger subsets of age classes included a large average contribution (> 40%) attributable to the effect of age. Based on these results, geographic analyses were restricted to age classes IV-V with sexes pooled. Geographic Variation— Standard statistics were calculated for each mensural character in each grouped locality. By ANOVA, the variance of each character was partitioned into among and within (error) locality effects. Only six of 14 characters were significantly heterogeneous among grouped localities (table 3). For seven characters (HF, ZB, IOC, MB, DBC, IW, IL), the among-locality vari- ance component was zero and for the remaining characters, locality accounted for 6% (LE) to 30% Karyology Heteromys gaumeri (2n = 56, FN = 76; fig. 2a)— The autosomal complement comprises grad- ed series of 1 1 pairs of large- to small-sized meta- centric and submetacentric chromosomes and 1 6 pairs of large- to small-sized acrocentric elements. The X chromosome is large and submetacentric and the Y is medium-sized and subtelocentric. No variation was noted among individuals. Heteromys desmarestianus (2n = 60, FN = 66; fig. 2b)— The autosomal complement com- prises four pairs of submetacentric and metacen- tric chromosomes, two large, one medium-sized, and one small, and 25 acrocentric pairs graded from large to small. The X chromosomes are pre- sumed to be large and submetacentric (no males were examined). Heteromys anomalus (2n = 60, FN = 68; fig. 2c)— The autosomal complement comprises two large, one medium-sized, and two small pairs of submetacentric and metacentric chromosomes and 24 acrocentric pairs graded from large to small. The X chromosome is large and submetacentric and the Y is medium-sized and subtelocentric. No variation was noted among individuals. 296 FIELDL\NA: ZOOLOGY Discussion r Patterns of nongeographic variation in mensural characters of Heteromys gaumeri are generally concordant with those observed in other hetero- myine rodents (Genoways, 1973; Rogers & Schmidly, 1982; see also Straney, 1978). In all species examined, age contributes a large propor- tion of within-sample variance for most charac- ters; only a minor component of total variance is attributable to sex. Male heteromyines generally average larger than females, however, and with large sample size secondary sexual dimorphism often appears significant for some characters (Gen- oways, 1973; age class III, this study). I Relative levels of variability of individual char- acters among species of heteromyines were com- pared using coefficients of variation of adults for the 12 external and cranial measurements com- mon to each study of intralocality variation (total length, length of tail, and the 10 cranial measure- ments included in this study; see Genoways, 1 973; Rogers & Schmidly, 1982). Coefficients of varia- f tion (CVs) appeared congruent among species for individual characters, although CVs appeared het- erogeneous among characters within species (ex- ternal measurements and interparietal width and length consistently were more variable than other characters). There was no indication of a reduction of within-sample variation for any character in H. gaumeri relative to other heteromyines. Although patterns and level of intralocality vari- ation in Heteromys gaumeri appear similar to oth- er heteromyines, geographic variation in H. gau- meri is relatively conservative. In H. gaumeri, mean values for most characters were homoge- neous among localities and geographically unpat- temed; only a small proportion of variance (av- erage 8%) was attributable to interlocality variation. We found no positive statistical correlation be- tween levels of within- and among-sample vari- ance (the "Kluge-Kerfoot phenomenon," KJuge & Kerfoot, 1 973) for mensural characters in H. gau- meri (tested using Kendall's rank correlation be- tween the W, and A, statistics suggested by Sokal, 1976; but see Rohlf et al., 1983). These data con- trast with studies of geographic variation within species of Liomys (Genoways, 1973) and other members of the H. desmarestianus species group (Rogers & Schmidly, 1982), in which population samples were statistically heterogeneous and ap- peared spatially patterned. The relative lack of interlocality variance in Heteromys gaumeri might be attributable to a re- stricted geographic distribution, to relative envi- ronmental homogeneity on the Yucatan Penin- sula, and/or to a lack of genetic divergence among populations. Compared to wide-ranging species of Liomys and Heteromys, the geographic area oc- cupied by H. gaumeri is relatively small, with little topographic or climatic relief Plant communities grade gradually from lowland tropical evergreen forest in the south and east, to deciduous forest in the northwestern and north-central portions of the peninsula, to a scrub zone bordering the northern coast (Paynter, 1955). For statistically heteroge- neous but spatially unpattemed characters, char- acter distributions might be determined mainly by stochastic factors (e.g., genetic drift) in the absence of strong migration or selective gradients (Sokal &. Oden, 1978). Peromyscus yucatanicus, which has a similar geographic range and occupies sim- ilar habitats to H. gaumeri, is geographically vari- able in color and cranial size (Lawlor, 1 965; Huck- aby, 1980), and at least color closely tracks vegetational changes on the peninsula. Conse- quently, lack of interlocality differentiation in col- or, qualitative, and mensural characters in H. gau- meri might not result solely from selective responses or stochastic processes in a homoge- neous environment. The reduced level of geographic variation in Heteromys gaumeri also is consistent with a hy- pothesis of little genetic divergence among pop- ulations. If populations of H. gaumeri are genet- ically similar, similarity probably is not a product of panmixia, but more likely of a reduction of genetic variation through a genetically depauper- ate founding population (see Johnston, 1976; Johnston &. KJitz, 1977; Baker, 1980) or genetic bottleneck. Although we have little direct evidence of genetic variation among populations of//, gau- meri, individuals sampled appear chromosomally monomorphic. In //. desmarestianus, interlocality polymorphism in fundamental number is pro- nounced (D. S. Rogers, pers. comm.). More sen- sitive estimation of the level of interpopulational genetic divergence of //. gaumeri awaits study of genie variation. Systematic Relationships of H. gaumeri MorpholcxjV— Allen and Chapman (1897) de- scribed Heteromys gaumeri from seven individ- uals from Chichen Itza, Yucatan. In his review of ENGSTROM ET AL.: HETEROMYS GAUMERI 297 Table 3. Geographic variation in six external and cranial measurements of Heteromys gaumeri. Grouped locality Percentage of variation N Mean (range) ± 2 SE Locality Error 1 2 3 4 S 6 7 8 9 10 11 3 16 10 S 9 17 4 3 3 2 4 288.7 261.7 276.9 272.0 280.0 286.8 294.5 281.7 263.3 256.5 270.5 Total Length 276.0-298.0) ± 13.13 235.0-288.0) ± 8.63 245.0-302.0) ± 10.98 262.0-286.0) ± 8.83 263.0-295.0) ± 7.51 265.0-324.0) ± 8.29 288.0-300.0) ± 4.93 275.0-290.0) ± 8.82 250.0-280.0) ± 17.64 253.0-260.0) ± 7.00 250.0-290.0) ± 18.65 23.7 76.3 3.75* Length of Tail 1 3 154.7 ( 148.0-166.0) + 11.39 13.6 86.4 2.02* 2 16 144.8 ( 129.0-159.0) + 4.12 3 10 150.4( 124.0-173.0) + 9.73 4 5 154.4( 141.0-166.0) + 9.48 • 5 9 161.0( 144.0-193.0) + 9.22 6 17 154.4( 135.0-183.0) ± 5.53 7 4 162.8 ( 160.0-166.0) ± 2.75 - 8 3 152.7 ( 151.0-155.0) ± 2.40 V- 9 3 145.7 ( 140.0-150.0) + 5.93 10 2 147.5 ( 147.0-148.0) + 1.00 11 4 146.5 ( 129.0-160.0) + 12.92 Greatest Length of Skull 1 3 35.8 (35.2-36.2) + 0.61 15.6 84.4 2.30* 2 17 35.3 (33.9-37.5) ± 0.44 9 35.0 (33.5-36.3) + 0.57 S 34.9 (33.6-36.2) ± 0.88 9 35.7 (34.3-36.7) + 0.52 23 35.9 (34.1-38.2) + 0.41 6 36.4 (35.8-37.0) ± 0.41 8 3 35.2 (33.8-36.6) ± 1.62 9 4 35.5 (34.2-36.6) ± 1.10 10 3 34.4 (34.0-34.7) + 0.40 11 2 34.6 (34.3-34.8) + 0.50 Length of Nasals 1 3 15.2 (14.7-15.7) + 0.58 13.8 86.2 2.19* 2 18 14.3 (13.2-15.6) + 0.34 10 13.9 (12.8-14.5) + 0.40 5 14.2 (13.6-15.0) + 0.61 9 14.6 (13.7-15.5) + 0.39 25 14.8 (13.6-15.8) + 0.27 6 14.8 (14.0-15.5) + 0.47 4 14.6 (13.7-15.8) + 0.87 4 14.6 (13.9-15.2) ± 0.54 10 3 13.8 (13.2-14.5) ± 0.75 11 2 13.9 (13.3-14.5) ± 1.20 Length of Rostrum 3 15.4 (14.8-16.3) ± 0.92 10.5 89.5 2.16* 18 15.4 (14.2-16.3) ± 0.25 10 15.1 (14.1-15.8) ± 0.38 5 15.2 (14.2-16.0) ± 0.61 8 15.5 (14.8-16.0) ± 0.33 25 15.6 (14.6-16.6) ± 0.23 6 16.0 (15.1-16.7) ± 0.62 8 4 15.4 (14.4-16.4) ± 0.84 298 FIELDIANA: ZOOLOGY Table 3. Continued. Grouped N Mean (range) + 2 SE Percentage of variation locality Locality Error F 9 10 11 4 3 2 15.6 (15.2-16.0) ± 0.37 14.8 (14.2-15.5) ± 0.75 14.6 (14.4-14.8) ± 0.40 Length of Maxillary Toothrow 18.9 1 3 4.9 (4.7-5.1) ± 0.23 2 19 4.9 (4.7-5.2) ± 0.07 3 10 4.9 (4.6-5.2) ± 0.13 4 5 5.0 (4.9-5.1) ± 0.06 5 8 4.8 (4.6-5.0) ± 0.08 6 26 5.0 (4.4-5.3) ± 0.08 7 6 5.2 (4.9-5.6) ± 0.21 8 4 4.9 (4.7-5.1) ± 0.17 9 4 4.8 (4.6-5.0) ± 0.17 10 4 4.9 (4.8-5.1) ± 0.13 11 2 5.2 (5.0-5.4) ± 0.40 81.1 2.76** Asterisks following F statistics indicate levels of significance (see footnote to Table 1). Percentages of variation are based on variance components estimated from the entire data set. Grouped localities are defined in text and outlined in Figure 1. the Heteromyinae, Goldman (1911) placed H. gaumeri in his H. desmarestianus group, but noted that it was aberrant and not closely related to any known species. Although H. gaumeri possesses characters diagnostic for Heteromys, including up- per and lower molars with three lophs and lower permanent premolars with three lophs, it also shares some characters with the sister group Lio- mys not shared with other Heteromys, including posterior sole of hind foot haired as in Liomys rather than naked as in other Heteromys and early disappearance of the enamel island between cin- gulum and metaloph(id) of upper and lower mo- lars, a condition somewhat intermediate between the genera (Goldman, 191 l;Genoways, 1973). As- suming Heteromys is monophyletic, the characters shared between H. gaumeri and Liomys probably are primitive (based on the "ojierational rule" out- group procedure outlined by Watrous & Wheeler, 1981). Dorsal coloration of H. gaumeri also is unusual in that the dorsum has a definite ochra- ceous cast with a bright ochraceous lateral line resembling Liomys pictus and L. spectabilis more than other, darker Heteromys. However, dorsal coloration in H. gaumeri might be convergent on L/omys through their common occupation of xeric forest and thorn scrub habitats rather than more mesic forest characteristic of most other species of Heteromys. Karyology — Compared to the nominate species in the Heteromys desmarestianus and H. anomalus species groups, the karyotype of//, gau- meri also is divergent (see fig. 2). Superficially, the karyotypes of //. desmarestianus (2n = 60, FN = 66) and //. anomalus (2n = 60, FN = 68) appear similar and might differ by a single rearrangement (heterochromatic addition/deletion or pericentric inversion), whereas the karyotype of //. gaumeri (2n = 56, FN = 76) differs by a minimum of seven rearrangements from //. desmarestianus and six from //. anomalus. The karyotype reported herein for //. desmarestianus differs from that reported by Genoways (1973; 2n = 60, FN = 82) in having a larger number of acrocentric elements. Popula- tions of //. desmarestianus are variable in fun- damental number and the karyotype reported herein is among the lowest FN karyotypes known for the species (D. S. Rogers, pers. comm.). Conclusions— We agree with Goldman (1911, p. 30) that Heteromys gaumeri "is a somewhat aberrant species, presenting characters which set it off from all the others [species of the H. des- marestianus group]." The distinctive morpholog- ical, ecological, and karyotypic features of//, gau- meri distinguish it from other members of the //. desmarestianus group (including //. goldmani; D. S. Rogers, pers. comm.). Shared (probably prim- itive) characters with Liomys suggest that H. gau- meri might represent an early branch of the lineage leading to other Heteromys (which share probable derived states for these characters). At present, we believe the distinctness and unresolved phyloge- ENGSTROM ET AL.: HETEROMYS GAUMERI 299 XI &{ii SI M >(» 99 MM AiS Mr A« A^it/s. OA Af> A* OH Awn. /in Ml Aft r%^ AA AA 4 X Y iri B Kn .. 11 a it 9A tl DC «» ai t* It im DA lid t\ »t tt) U M AA OA nn 60 Oft Aft »» X X 10 n QO AA A9 00 to A« i^a i<( (in ft A #0 #11 ak aq HA (fS llW 411 HA #14 jf^ X X Fig. 2. Representative karyotypes of Heteromys: A, karyotype of a male H. gaumeri (2n = 56, FN = 76) from 7.5 km W Escarcega, Campeche, Mexico; B, karyotype of a female H. desmarestianus (2n = 60, FN = 66) from 9.4 km S Palenque, Chiapas, Mexico; C, karyotype of a female H. anomalus {2n = 60, FN = 68) from Caripito, Monagas, Venezuela. Insert shows sex chromosomes of a male. 300 FIELDIANA: ZOOLOGY netic position of H. gaumeri would best be em- phasized by removing it from the H. desmaresti- anus group, and recognizing it as a divergent species at the same cladogenic level as the species groups in the subgenus Heteromys. Further investigation of heteromyines might indicate that H. gaumeri warrants subgeneric recognition. Species Account I Heteromys gaumeri Allen and Chapman, 1897 Heteromys gaumeri Allen and Chapman, 1897, Febru- ary. Bull. Amer. Mus. Nat. Hist., 9: 9. HoLOTYPE— Adult male, skin and skull, amnh 12028/10461; from Chichen Itza, Yucatan, Mex- ico. Type examined. Distribution— Northern Belize; Peten, Gua- temala; and Campeche, Quintana Roo, and Yu- catan, Mexico (see fig. 1); subdeciduous-subpe- rennial tropical rain forest, tropical deciduous forest, and thorn scrub forest from sea level to 100 m. Description— Dorsal coloration of adults ranges from dark to medium gray, with heavy admixture of orange buff hairs lending an overall ochraceous cast to the otherwise gray dorsum (adults in worn pelage appear more ochraceous and molting in- dividuals often have a 'salt and pepper' appear- ance); rich orange buff lateral line, usually broad and conspicuous, extending from cheeks to base of tail, ochraceous hairs often extending onto dor- sal and ventral surfaces of ankle; margins of fore- arm orange buff, interrupted on dorsal surface by white line; venter and feet white; tail well haired, grayish brown above, dull white below, with con- spicuous terminal tuft of hairs; ears dusky, lightly edged with dull white; sole of hind foot haired, posteriorly. Subadults medium to dark gray above, dorsum without ochraceous hairs; ochraceous lat- eral line faint and narrow. Juvenile pelage similar to that of subadults, but spiny hairs on dorsum absent. Tail longer than head and body; soles of hind feet with six tubercules; body size medium for the genus. Skull size medium, with relatively large auditory bullae; lower permanent premolar with three lophs; upp>er and lower molars with three lophs, enamel island formed between meta- loph(id) and cingulum disappearing quickly with wear. 2n = 56, FN = 76. Couv\v.\so^s— Heteromys gaumeri is geo- graphically isolated from all other heteromyines except H. desmarestianus, which it might contact at the southern edge of the Yucatan Peninsula (see Jones et al., 1 974). Heteromys gaumeri differs from H. desmarestianus in averaging smaller in most external and cranial measurements; having a broad, bright ochraceous lateral line, extending onto cheeks and ankles (a narrow, pale ochraceous lat- eral line often is present in H. desmarestianus, but seldom extends onto cheeks or ankles); having soles of hind feet haired posteriorly (this area is naked in H. desmarestianus and all other Heteromys); having a relatively well-haired tail with terminal tuft (the tail is sparsely haired in H. desmaresti- anus, without a conspicuous terminal tuft); having relatively large auditory bullae; and in having a diploid number of 56 chromosomes (compared to 60 in H. desmarestianus). For additional com- parisons with other heteromyines, see Goldman (191 1) and Genoways (1973). Remarks— Laurie (1957, p. 387) assigned eight specimens from three localities in the state of Yu- catan, Mexico (Tekom, 2; X-Cala-Koop, 1; Chi- chen Itza, 5) to Heteromys desmarestianus. One of us (HHG) reexamined her material, which is stored in alcohol; based on size, coloration, and a haired posterior sole of the hind foot, all are as- signable to H. gaumeri. Consequently, the north- ernmost locality for H. desmarestianus on the Yu- catan Peninsula is 85 km W Chetumal, Quintana Roo, Mexico (Jones et al., 1974). Two other spec- imens, from Kate's Lagoon, Belize, identified by Laurie (1957, p. 387) as H. desmarestianus, are referable to H. gaumeri. These specimens, along with additional material from Honey Camp La- goon (reported by Izor &. McCarthy, 1984) and Rockstone Pond reported here, suggest that H. gaumeri is distributed throughout northern Belize. Acknowledgments Permits to collect specimens in Mexico were kindly provided by the Director General of the Departmento de Conservacion de la Fauna Sil- vestre. Especial thanks are extended to Biol. Fran- cisco Rodriguez Gallegos, Ing. Antonio Sanchez, and personnel of Campo Experimental Forestal "El Tormento" for their hospitality during field- work near Escarcega, Campeche. We also thank the many curators who permitted examination of specimens in their care. E. Jaquelin Dietz provided the computer program for testing the association between distance matrices. Kenneth Schoenly pre- pared the final copy of Figure 1. D. S. Rogers ENGSTROM ET AL.: HETEROMYS GAUMERI 301 provided information and helpful discussion on character-states within the Heteromyinae. Por- tions of this study were funded by the Texas Ag- ricultural Experiment Station (project H-1977 to D. J. Schmidly) and a Faculty Organized Research Grant, Angelo State University. Fieldwork in Ven- ezuela was supported by NSF Grant DEB 79-2 1519 to J. W. Bickham and the Carnegie Museum of Natural History through the M. Graham Netting Research Fund, established by a grant from the Cordelia S. May Charitable Trust. Literature Cited Allen, J. A., AND F. M. Chapman. 1897. Onmammals from Yucatan, with descriptions of new species. Bul- letin of the American Museum of Natural History, 9: 1-12. Baker, A. J. 1980. Morphometric differentiation in New Zealand populations of the house sparrow {Passer domesticus). Evolution, 34: 638-653. Choate, J. R., AND H. H. Genoways. 1975. Collec- tions of Recent mammals in North America. Journal of Mammalogy, 56: 452-502. DiETZ, E. J. 1983. Permutation tests for association between two distance matrices. Systematic Zoology, 32:21-26. Gaumer, G. F. 1917. Monografia de los mamiferos de Yucatan. Departmento de Taleres Graficos de la Se- cretaria de Fomento, Mexico, 33 1 pp. Genoways, H. H. 1973. Systematics and evolutionary relationships of the spiny pocket mice, genus Liomys. Special Publications, The Museum, Texas Tech Uni- versity, 5: 1-368. Goldman, E. A. 1911. Revision of the spiny pocket mice (genera Heteromys and Liomys). North Ameri- can Fauna, 34: 1-70. Hall, E. R. 1981. The Mammals of North America, vol. 1. John Wiley and Sons, New York, 600 + 90 pp. Hall, E. R., and K. R. Kelson. 1959. The Mammals of North America, vol. 1 . Ronald Press, New York, 546 + 79 pp. Hatt, R. T., H. I. Fisher, D. A. Langebartel, and G. W. Brainerd. 1953. Faunal and archeological re- searches in Yucatan caves. Bulletin of the Cranbrook Institute of Science, 33: 1-119. HucKABY, D. G. 1 980. Species limits in the Peromys- cus mexicanus group (Mammalia: Rodentia: Muro- idea). Contributions in Science, Natural History Mu- seum of Los Angeles County, 326: 1-24. IzoR, R. J., and T. J. McCarthy. 1 984. Heteromys gaumeri (Rodentia: Heteromyidae) in the Northern Plain of Belize. Mammalia, 48: 465-467. Johnston, R. F. 1976. Estimating variation in bony characters and a comment on the KJuge-Kerfoot effect. Occasional Papers of the Museum of Natural History, University of Kansas, 56: 1-8. Johnston, R. F., and W. J. Klitz. 1977. Variation and evolution in a granivorous bird: The house spar- row, pp. 15-51. In Pinowski, J., and S. C. Kendeigh, eds., Granivorous Birds in Ecosystems, Their Evolu- tion, Populations, Energetics, Adaptations, Impact and Control. Cambridge University Press, London. Jones, J. K., Jr., H. H. Genoways, and T. E. Lawlor. 1974. Annotated checklist of mammals of the Yu- catan Peninsula, Mexico II. Rodentia. Occasional Pa- pers, The Museum, Texas Tech University, 22: 1-24. Kluge, a. G., and W. C Kerfoot. 1973. The pre- dictability and regularity of character divergence. American Naturalist, 107: 426-442. Laurie, E. M. O. 1957. Rodents from British Hon- duras, Mexico, Trinidad, Haiti, and Jamaica collected by Mr. I. T. Sanderson. Annals of the Magazine of Natural History, series 12, 6: 382-394. Lawlor, T. E. 1965. The Yucatan deer mouse, Pero- myscus yucatanicus. University of Kansas Publica- tions, Museum of Natural History, 16: 421-438. Leamy, L. 1983. Variance partitioning and effects of sex and age on morphometric traits in randombred house mice. Journal of Mammalogy, 64: 55-61. Lee, M. R. 1969. A widely applicable technique for direct processing of bone marrow for chromosomes of vertebrates. Stain Technology, 44: 155-158. Patton, J. L. 1967. Chromosome studies of certain pocket mice, genus Perognathus (Rodentia: Hetero- myidae). Journal of Mammalogy, 48: 27-37. Paynter, R. a., Jr. 1955. The omithogeography*of the Yucatan Peninsula. Bulletin of the Peabody Mu- seum of Natural History, Yale University, 9: 1-347. Rogers, D. S., and D. J. Schmidly. 1982. Systematics of spiny pocket mice (genus Heteromys) of the des- marestianus species group from Mexico and northern Central America. Journal of Mammalogy, 63: 375- 386. RoHLF, F. j., a. j. Gilmartin, and G. Hart. 1983. The Kluge-Kerfoot phenomenon— A statistical arti- fact. Evolution, 37: 180-202. RoHLF, F. J., J. KiSHPAUGH, AND D. KiRK. 1982. Nu- merical Taxonomy System of Multivariate Statistical Programs. State University of New York at Stony Brook, Stony Brook, New York, 98 pp. Ryan, R. M. 1960. Mamiferos colectados en Guate- mala en 1954. Acta Zoologica Mexicana, 4: 1-19. SAS Institute, Inc. 1982. SAS User's Guide: Statis- tics, 1982 ed. SAS Institute, Cary, North Carolina, 584 pp. Searle, R. S. 1971. Linear Models. John Wiley and Sons, New York, 535 pp. Sneath, p. H. a., and R. R. Sokal. 1973. Numerical Taxonomy. W. H. Freeman, San Francisco, 573 pp. Sokal, R. R. 1976. The Kluge-Kerfoot phenomenon reexamined. American Naturalist, 110: 1077-1091. . 1979. Testing statistical significance of geo- graphic variation patterns. Systematic Zoology, 28: 232-238. Sokal, R. R., and N. L. Oden. 1978. Spatial auto- correlation in biology. 2. Some biological implications and four applications of evolutionary and ecological 302 HELDIANA: ZOOLOGY interest. Biological Journal of the Linnean Society, 10: 229-249. SoKAL, R. R., AND F. J. RoHLF. 1981. Biometry. W. H. Freeman, San Francisco, 859 pp. Straney, D. O. 1978. Variance partitioning and non- geographic variation. Journal of Mammalogy, 59: 1- 11. Watrous, L. E., and Q. D. Wheeler. 1 98 1 . The out- group comparison method of character analysis. Sys- tematic Zoology, 30: 1-11. Appendix Specimens Examined MEXICO: Campeche: Apazote, near Yohaltun, 6 (usnm); Campeche, 1 (usnm); 5 km S Cham- poton, 10 m, 3 (ku); Dzibalchen, 1 (ku); 7 km N, 5 1 km E Escarcega (= Francisco Escarcega), 4 (ku); 7.5 km W Escarcega, 65 m, 94 (15 ku, 79 tcwc); 7 km E Escarcega, 4 (ku); 103 km SE Escarcega, 2 (ku); La Tuxpena, Champoton, 1 2 (usnm); San Jose Carpizo, 3 (fmnh); San Jose Carpizo, 45 km S Campeche, 19 (ummz); San Juan, 4 (fmnh). Quintana Roo: 2 km N, 8 km W Bacalar, 1 1 (tcwc); 8 mi NNE Felipe Carrillo Puerto, 4 (tcwc); 4 km NNE Felipe Carrillo Puerto, 30 m, 9 (ku); 2.5 mi NNE Felipe Carrillo Puerto, 1 (tcwc); La Vega, 5 (usnm); Pueblo Nuevo Xcan, 10 m, 1 (ku); 1.5 km S, 1 km E Pueblo Nuevo Xcan, 1 (ku); Puerto Morelos, 4 (usnm). Yucatan: Calcehtok, 55 km SW Merida, 2 (1 ku, 1 ummz); Chichen Itza, 10 m, 36 (15 amnh, 11 bmnh, 2 fmnh, 1 ku, 7 usnm); Chichen Itza, Cenote Xtoloc, 1 (tcwc); 2 km E Chichen Itza, Cenote Seco, 5 (tcwc); 4 km E Dzi- tya, 1 (mmnh); Esmeralda, Laguna de Chichan- canab, 2 (1 ku, 1 ummz); Esmeralda, 45 km S Peto, 4 (ummz); 1 km SW Casa de la Esmeralda, Laguna de Chichancanab, 2 (1 ku, 1 unam); 2 km SSW Casa de la Esmeralda, Laguna de Chichancanab, 3 (1 AMNH, 2 unam); La Ceiba, 2 km SE Laguna de Chichancanab, 1 (unam); 66 km NE Merida, 1 (ku); 14 km SW Muna, 1 (rru); Oxkutzcab, Ebiztic (cave deposit), 6 (usnm); Peto, 3 (ku); 3 km N Piste, 16 (7 ku, 6 mmnh, 3 ttu); 2 km N Piste, 2 (ku); Piste, 10 m, 2 (ku); Progreso, 1 (usnm); Santa Rosa, 1 (ummz); Santa Rosa, 20 km S Peto, 2 (ummz); Santa Rosa, 25 km S Peto, 1 (ummz); 1 km SSW Santa Rosa, 3 (1 ku, 1 ummz, 1 unam); Tekom, 2 (bmnh); 6 km N Tizimin, 1 (ku); Tun- kas, 7 (usnm). Buena Vista, 1 (fmnh) and X-Cala- Koop, 1 (bmnh) were not located exactly. GUA- TEMALA: Peten: Chuntuqui, 2 (usnm); Laguna de Sotz [= Zotz], 2 (usnm). BELIZE: Belize: Kate's Lagoon, 2 (bmnh); Rockstone Pond, 1 1 (rom). Or- ange Walk: Honey Camp Lagoon, 2 (fmnh). Additional Records (Hatt et al., 1953, p. 64, unless noted otherwise)— MEXICO: Yucatan: Ac- tun Chacaljas, 3 km SSW Calcehtok (cave de- posit); Actun Coyok [= Coyoc], 3.5 km SSE Oxkutzcab (cave deposit); Actun Has, 3.5 km WSW Yokat (cave deposit); Actun Lara, 3 km SW Yokat (cave deposit); Actun Oxkintok, 3 km SW Santa Cruz (cave deposit); Actun Spukil, 4.5 km SSW Calcehtok (cave deposit); Loltun, 5 km SW Oxkutzcab (cave deposit); Xbac (Gaumer, 1917, p. 13); Yaxcach (Gaumer, 1917, p. 13). GUA- TEMALA: Peten: 1 1 km NE JHores, 3.2 km inland SE shore Laguna Peten Itza, ca. 100 m (Ryan, 1960, p. 11). ENGSTROM ET AL.: HETEROMYS GAUMERI 303 Species Groups of Spiny Rats, Genus Proechimys (Rodentia: Echimyidae) James L. Patton ABSTRACTS Nine species groups of the spiny rat subgenus Proechimys, family Echimyidae, are defined on the basis of bacular characters and qualitative features of the cranium. The latter include the structure of the incisive and mesopterygoid foramina, temporal ridge and infraorbital canal development, and counterfold pattern of the cheekteeth. Three groups are apparently monotypic, including the decumanus -group of southwestern Ecuador and adjacent Peru, the canicollis-group of northeastern Colombia and adjacent Ven- ezuela, and the s i mo nsi -group of the western Amazon Basin from Colombia to northern Bolivia. The remaining six groups are polytypic, but the number of species in each remains unclear. The semispinosus-group ranges from Central America south to southwestern Ecuador in the Pacific lowlands; its only Amazonian representative is P. oconnelli from east-central Colombia. The longicaudatus-group ranges from southeastern Colombia through the western Amazon Basin into the northern Parana Basin of Brazil and northern Paraguay. The goeldii-group ranges throughout the Amazon Basin from eastern Peru to eastern Brazil. The guyannensis-group occurs from the coastal Guianan region through the Rio Negro and eastern half of the Amazon Basin in Brazil, with an isolate in Goias and Minas Gerais states. The cuvieri-group has a similar distribution, but extends further up the Amazon into northern Peru, with one isolate in east-central Peru. Finally, the trinitatus-gcoup is found from north-central Colombia eastward across northern Venezuela to Trinidad. Nueve grupos de especies de la rata espinosa subgenero Proechimys, familia Echimyidae, son definidos primariamente en las bases de caracteres baculares que son soportados por razgos cualitativos del craneo. El ultimo incluye la estructura del foramen incisivo y mesopterigoideo, arista temporal y desarroUo del canal infraorbital y patron de contraplegamiento de los dientes postcaninos. Tres grupos son aparentemente monotipicos, incluyendo el grupo decumanus del suroccidente del Ecuador y adyacente Peru, el grupo canicollis del nororiente de Colombia y adyacente Venezuela, y el grupo simonsi del occidente de la cuenca amazonica desde Colombia hasta el norte de Bolivia. Los seis grupos remanentes son politipicos, pero el numero de especies en cada uno p>ermanece obscuro. El grupo semispinosus se extiende desde el sur de Centroamerica hasta el suroccidente ecuatoriano en las tierras bajas del Pacifico; su solo representante de la cuenca amazonica el P. oconnelli del centroriente colombiano. El grupo longicaudatus se dis- tribuye desde el suroriente de Colombia a traves del occidente de la cuenca amazonica hasta el norte de la cuenca del Parana en Brazil y norte de Paraguay. El grupo goeldii se distribuye a traves de la cuenca amazonica desde el Peru oriental hasta el Brazil oriental. El grupo guyannensis ocurre desde la region costera guayanesa a traves de rio Negro y la mitad oriental de la cuenca amazonica en Brazil, con un aislamiento de poblaciones en los estados de Goias From Museum of Vertebrate 2^ology, University of California, Berkeley, CA 94720. PATTON: SPECIES GROUPS OF PROECHIMYS 305 y Minas Gerais. El grupo cuvieri tiene una distribucion similar, pero se extiende mas arriba del Amazonas en el interior del norte del Peru, con un aislamienlo de poblaciones en el centroriente peruano. Finalmente, el grupo trinitatus es encontrado desde el centronorte de Colombia hacia el oriente a traves del norte de Venezuela hasta Trinidad. Nove grupos de especies de ratos-de-espinho, do subgenero Proechimys, familia Echimyidae, sao definidos principalmente na base de carateres baculares que concordam tambem com caracteres qualitativos do cranio. Estes incluem: a estrutura dos incisivos e do forame mesop- terigoideo, o desenvolvimento das temporas e do canal infraorbital, e o padrao dos molares. Tres grupos sao aparentemente monotipicos, incluindo o grupo decumanus do sudoeste do Equador, o grupo canicollis do nordeste da Colombia e das areas adjacentes na Venezuela, e o grupo simonsi da Bacia Amazonica ocidental, desde a Colombia ate o norte da Bolivia. Os seis grupos restantes sao politipicos, mas o numero de especies em cada continua incerto. O grupo semispinosus estende-se da America Central ao sudoeste do Equador, nas planicies do Pacifico. Seu unico representante na Bacia Amazonica e P. oconnelli, do Centro-leste da Colombia. O grupo longicaudatus estende-se do sudeste da Colombia, atraves da Amazonia ocidental, ate o norte da Bacia do Parana no Brasil e no norte do Paraguai. O grupo goeldii ocorre na Bacia Amazonica, do leste do Peru ao leste do Brasil. O grupo guyannensis ocorre da costa guianense, ate o Rio Negro e a regiao oriental da Bacia Amazonica no Brasil, com uma especie isolada nos estados de Goias e de Minas Gerais. A distribuifao do grupo cuvieri e parecida, mas este ocorre tambem ate o norte do Peru, com uma especie isolada no centro-leste do Pais. Por final, o grupo trinitatus e encontrado do centro-norte da Colombia, atraves da Venezuela, ate Trin- idade. Introduction Spiny rats of the genus Proechimys represent one of the most diverse groups of Neotropical rodents; with the possible exception of tuco-tucos, Cteno- mys, the number of taxa of spiny rats is unmatched by any other caviomorph (Woods, 1984). The ge- nus extends throughout lowland forests from Nic- aragua to northern Paraguay and the coastal re- gions of Brazil. Despite this diversity and large geographic range, however, the group is taxonom- ically one of the most poorly understood among all of the Neotropical mammals. Only a few stud- ies have succeeded in recognizing the number of taxa sympatric at any single locality (e.g., Moojen, 1948; Patton & Gardner, 1972), and no study has been able to follow geographic character trends within a clearly defined taxon over any but the shortest distances. Diagnosis of species and hence definition of natural units in Proechimys have been severely hampered by the often extreme level of variability within and between populations for most morphological characters that have been ex- amined (Moojen, 1948; Hershkovitz, 1948; Pat- ton &. Gardner, 1972). Even karyotypes, which have proven useful in differentiating sympatric taxa of spiny rats (Patton & Gardner, 1972), are often highly variable geographically (Reig & Useche, 1976;Reigetal., 1 980; Gardner & Emmons, W84). In this report I will challenge some of these pre- cepts of character instability (see Thomas, 1928, p. 262) by using the structure of the baculum and specific characters of the cranium to define major taxonomic groups of spiny rats. Taxa can be di- agnosed, despite both within- and between-pop- ulation variation, and the patterns of character variation over geography are coherent, permitting a consistent view of these taxa throughout their range. Some of the more traditional characters that have been used to make taxonomic decisions in spiny rats (e.g., counterfold patterns on the cheek- teeth) are not chaotic in their variation patterns, but are quite helpful in defining units. This paper will consider only members of the subgenus Proechimys, excluding entirely the group of species found along the Atlantic highlands of Brazil that compose the subgenus Trinomys. Species Groups of Proechimys In the section below I provide the basic species groups of spiny rats, subgenus Proechimys, listing 306 HELDIANA: ZOOLOGY Fig. 1 . Geographic distribution of taxa of the simonsi- and guyannensis-^ovii>s. Type locaHties for each included taxon are indicated by stars; dots represent other locahties hsted in the Appendix. those named forms I consider as component parts. In recognizing these groups and their membership I make no conclusions here as to the specific, sub- specific, or other status of these names. Because of the plethora of names available and the con- fusion with which each has been applied to the genus over the past century, this synopsis is pro- vided first to allow for coherent discussion; the documentation upon which these decisions are based follows. I recognize nine species groups within the sub- genus Proechimys. Five of these are widespread, while the remaining ones are more restricted in their ranges. Maps of the distribution of each group, with localities of included holotypes, are presented in Figures 1-4 (see Specimens Examined for lists of localities). Unless otherwise stated, allocation of any given holotype to a specific species-group is based on my examination of that specimen. The groups are defined by a combination of palatal (particularly incisive foramina and mesopterygoid fossa) characters, counterfold patterns of the cheekteeth, temporal ridge development, infraor- bital notch development, and bacular characters (see below). In each case, the group name is taken from the oldest assignable name for that unit. guyannensis-%Tou^ Named forms in this group include: guyannensis (E. Geoffroy, 1 803) cherriei Thomas, 1 899 roberti Thomas, 1 90 1 vacillator Thomas, 1 903 oris Thomas, 1 904 warreni Thomas, 1 905 PATTON: SPECIES GROUPS OF PROECHIMYS 307 SNmi (Mkm Fig. 2. Geographic distribution of taxa of the goeldii-group (dots) and decumanus-group (triangles). Type localities of taxa are indicated by stars. boimensis Allen, 1916 arescens Osgood, 1 944 riparum Moojen, 1948 ar abupu Moo}tn, 1948 Comments— On the basis of septal patterns in the bullae, Gardner and Emmons (1984) included these taxa in their brevicauda-group, an all-inclu- sive unit combining taxa that are here allocated to six separate groups. As will be apparent below, my guyannensis-group only shows close similarity to the taxa included in the 5/wo/J5/-group. Mem- bers of these two groups share virtually no char- acters with the remaining taxa listed by Gardner and Emmons (1984) in their brevicauda-group. This group is confined in its distribution to the Guianan region and southern Venezuela through the central Amazon Basin of Brazil, with an isolate (roberti) in Minas Gerais and Goias states in Brazil (see map, fig. 1). It is sympatric with members of the cuvieri-group in the Guianan region (see Petter, 1978) and with those of both the cuvieri- and goel- dii-groups in the central Amazon Basin. goeldii-group Included are the following named forms: goeldii Thomas, \ 90 5 steerei Goldman, 1911 kermiti M\Qn, 1915 pachita Thomas, 1923 hilda Thomas, 1924 rattinus Thomas, 1 926 quadruplicatus Hershkovitz, 1 948 liminalis Moojen, 1948 amphichoricus Moojen, 194S hyleae Moojen, 1 948 nesiotes Moojen, 1 948 leioprimna Moojen, 1 948 308 HELDIANA: ZOOLOGY Comments— I have not examined the holotypes of liminalis Moojen or hyleae Moojen; their in- clusion here is based on the original descriptions (Moojen, 1948). This group is distributed throughout the Ama- zon Basin, from the most western margins in northern Bolivia, eastern Peru, Ecuador, and southeastern Colombia to southern Venezuela east along the central Amazon to the lower Rio Tapajos in Para state, Brazil (see map, fig. 2). Members of this group are sympatric with those of the guy- annensis-group and CMv/er/-group in the central and eastern Amazon Basin, and with those of the cuvieri-, longicaudatus-, and 5/ mo/u/ -groups in the western parts of the Basin. longicaudatus-^ronp Named forms in this group include: longicaudatus (Rengger, 1830) brevicauda (Gunther, 1877) boliviensis Thomas, 1 90 1 securus Thomas, 1 902 gularis Thomas, 1911 leucomystax Ribeiro, 1914 elassopus Osgood, 1 944 villacauda Moojen, 1 948 ribeiroi Moojen, 1 948 Comments— The taxa leucomystax Ribeiro, vil- lacauda Moojen, and ribeiroi Moojen are included on the basis of descriptions given in Moojen ( 1 948); I have not examined the holotypes. The longicaudatus-group is confined to the western and southwestern parts of the Amazon Basin and northern Parana Basin, from northern Paraguay and adjacent Brazil west and northwest through Bolivia, eastern Peru, eastern Ecuador, and southeastern Colombia (see map, fig. 3). In this region it is sympatric with members of the goeldii-, cuvieri-, and simonsi-groups. stmonsi-^Toup Included members are: simonsi Thomas, 1 900 hendeei Thomas, 1 926 nigrofulvus Osgood, 1 944 Comments— This is perhaps the most readily recognizable of all groups of Proechimys; the level of differences with sympatric taxa of the goeldii-, longicaudatus-, or CMv/er/-groups is geographically consistent and quite sharp. The simonsi-gToup is geographically restricted to the western margins of the Amazon Basin from northern Bolivia through eastern Peru and Ecua- dor to southeastern Colombia (see map, fig. 1 ). In this region, it ranges to higher elevations than any other species in the genus, occurring as high as 2000 m. cuvten-group This group includes only the nominate form: CM v/mPetter, 1978 Comments— Specimens assigned to this species are relatively few in number and are known from localities scattered from the coastal Guianan re- gion and along the Amazon River from near its mouth to northern Peru (see map, fig. 4). In the Guianas, cuvieri is sympatric with guyannensis; in central Brazil, with guyannensis- and goeldii-gjroup taxa; and in northern Peru, with simonsi-, longi- caudatus-, and goeldii-group members. Despite the paucity of widely scattered locality records, these specimens share common bacular, palatal, coun- terfold, and karyotypic characters, the former being particularly divergent from other taxa in the genus Proechimys. Because of similar bacular (but not karyotypic) features, the form from eastern Peru referred to P. guyannensis by Patton and Gardner (1972) is in- cluded here. This form is an enigma at the mo- ment, and its placement must be considered pro- visional; it is not known with certainty from any locality other than Balta, Rio Curanja, Ucayali, Peru (see map, fig. 4, and Patton & Gardner, 1 972). Gardner and Emmons (1984) consider it closely related to P. guyannensis, perhaps even conspe- cific, but it does not share bacular or incisive fo- raminal characters with members of that group (see below). trinitatus-group Named forms in this group include: trinitatus (Allen and Chapman, 1893) chrysaeolus (JhonvdiS, 1898) mmca^ (Allen, 1899) MHc/i/ (Allen, 1899) PATTON: SPECIES GROUPS OF PROECHIMYS 309 Fig. 3. Geographic distribution of taxa of the longicaudatus-group (dots) and trinitatus-group (triangles). Type localities are indicated by stars. guairae Thomas, 1 90 1 ochraceous Osgood, 1912 poliopus Osgood, 1914 hoplomyoides Tate, 1 939 magdalenae Hershkovitz, 1 948 Comments— Gardner and Emmons (1984) in- cluded magdalenae in their brevicauda-group and chrysaeolus in their semispinosus-gxoup based on similarities of bullar septal patterns. Bacular and other characters, however, align these two taxa with others of the trinitati/s-groxip as here defined. This group includes all members referred to as the "guiarae complex" by Reig and co-workers, based on karyotypic data (see Benado et al., 1979; Reig el al., 1980; Reig, 1981), or as the guairae-group by Gardner and Emmons ( 1 984). Members of this group are distributed across northern South America, from Trinidad through the coastal mountains and upper llanos of Vene- zuela and Colombia, and including the isolated northern Andean valleys of Colombia (see map, fig. 3). Only hoplomyoides occurs south of the Rio Orinoco in southeastern and southern Venezuela (see Gardner & Emmons, 1 984). The included taxa are largely allopatric, or parapatric (see Reig et al., 1980; Reig, 1981), and are separated by rivers (upper llanos of Venezuela) or mountain ridges (northwestern Venezuela and northern Colombia). Sympatric contact between members of this group and other Proechimys occurs in several areas in northern Colombia: chrysaeolus with semispino- sus and both chrysaeolus and mincae with cani- collis. semtspinosus-^oup Membership in this group includes the following named forms: 310 HELDIANA: ZOOLOGY Fig. 4. Geographic distribution of taxa of the CMv/er/-group (dots), semispinosus-sroup (triangles), and canicollis- group (squares). Type localities are indicated by stars. semispinosus (Jorrxts, 1860) centralis (Thomas, 1 896) rosa Thomas, 1 900 chiriquinus Thomas, 1 900 panamensis Thomas, 1 900 burnis Bangs, 1 90 1 gorgonae ^2in%s, 1905 calidior Thomas, 1911 oconnelli Allen, 1913 rwfte//M5 Hollister, 1914 colombianus Thomas, 1914 goldmani Bo\e, 1937 ignotus Kellogg, 1 946 Comments— Gardner (1983) reviewed the membership, distribution, and taxonomic history of this species group, and I concur with him. Gard- ner and Emmons (1984) expanded the group to include oconnelli and chrysaeolus based on com- mon buUar septal patterns. I treat oconnelli as a component part of the semispinosus-group, but place chrysaeolus in the trinitatus-group based on bacular, palatal, and counterfold patterns (see be- low). Members of the semispinosus-group are largely restricted in distribution to Central America and the Pacific lowlands of Colombia and Ecuador (see map, fig. 4). The only Amazonian representative is oconnelli, which is restricted to the limited area of Villavicencio in east-central Colombia, mid- way between the northernmost distributional ex- tensions of members of the goeldii-, simonsi-, and longicaudatus-groups and the westernmost exten- sion of the trinitatus-gxoup. canicoliis-group This group is limited to the nominate species: camco//w (Allen, 1899) PATTON: SPECIES GROUPS OF PROECHIMYS 311 Comments— The species P. canicollis is one of the more readily recognizable in the entire genus Proechimys (see below), although on the basis of bullar septal patterns, Gardner and Emmons ( 1 984) included it within their brevicauda-group. It is lim- ited in its distribution to the coastal forested foot- hills from northern Bolivar, Colombia, to north- western Zulia, Venezuela (see map, fig. 4). It is sympatric with mincae and chrysaeolus of the trin- itatus-group. decumanus-group This group includes only the nominate species: decumanus (Thomas, 1 899) CoMMEhrrs— I consider this taxon to represent a separate species group, although Gardner and Emmons (1984) placed it in their brevicauda-gxoup based on bullar septal patterns and karyotypic similarities. It is, however, readily distinguishable on external, palatal, and bacular grounds from oth- er members of that group. Proechimys decumanus is restricted to the Pa- cific lowland forests of extreme southwestern Ec- uador and adjacent northwestern Peru (see map, fig. 2), where it is sympatric with the named form rosa of the semispinosus-gcoup. Bacular Structure and Characteristics Didier (1962) described and figured bacular variants of echimyid rodents, with an emphasis on variation within the genus Proechimys. Bacular variants of Proechimys were also described by Martin (1970) and Patton and Gardner (1972). In the latter paper, bacular characters and karyotypes permitted the delineation of taxa sympatric at sev- eral localities in eastern Peru, suggesting that such structures could be of use in distinguishing taxa within the genus as a whole. Didier's (1962) ma- terial and the names he used were supplied by Philip Hershkovitz, from specimens in Field Mu- seum collections. I have examined each of the bacula discussed by Didier, as well as the associ- ated skulls and skins. Below, as I describe bacular variation in the context of the species groups rec- ognized in this report, I will emend Didier's group- ings according to this reexamination of materials. A basic bacular type characterizes most species in the genus Proechimys, as well as other echimyid genera, despite differences in overall size. This baculum is an elongated, narrow structure, with a rather rounded and broadened base and a shaft tapering distally. The distal tip shows only a weak- ly developed median depression, if any at all. This bacular type is characteristic of the dactylomyine genera Thrinacodus, Kannabateomys, and Dac- tylomys, species in the genera Echimys, Makalata, Mesomys, Diplomys, and Isothrix (see Didier, 1962; Patton & Emmons, 1985), and most species oi Proechimys. It is also characteristic of most oth- er caviomorphs examined to date (e.g., Cavia, Abrocoma, Ctenomys, Agouti, Dasyprocta [Didier, 1962; Hooper, 1961]). There are, however, both subtle as well as more marked differences among bacula of this general form, and these will be de- tailed below in the discussion of variation in Pro- echimys. Overall size of the baculum of Proechimys de- pends strongly on age, although age does not no- ticeably affect shape. Hence, in the descriptions and summarized measurements given below, analyses are restricted to those individuals con- sidered adults on the basis of tooth wear patterns (age classes 8 through 10 of Patton & Rogers, 1983). Bacula of each species group of Proechimys are illustrated in Figures 5-11, and measurements for geographic representatives of each group are summarized in Table 1 . Figure 1 2 illustrates dif- ferences among the species groups in proportions of length and width. It is clear from this figure that there are two major classes of bacular variants in the subgenus Proechimys. Members of the guy- annerisis-, simonsi-, trinitatus-, goeldii-, decu- manus-, and canicollis-groups have long and nar- row bacula (despite differences in relative size and other characters), while taxa of the semispinosus-, cuvieri-, and longicaudatus-groups have massively long and broad bacula with well-developed distal apical wings or extensions. guyannensis-group (Figure 5a-g) Despite Didier's (1962) use of the name guy- annensis, no sp>ecimen he examined can be re- ferred to this group as defined here or as recog- nized by Gardner and Emmons (1984). The four bacular types Didier (1962, pp. 408-415) referred to Proechimys guyannensis in fact represent mem- bers of five different species groups, as follows: Type I (part semispinosus-group, part cuvieri- group), Type II (longicaudatus-group). Type III (P. 312 FIELDIANA: ZOOLOGY n Fig. 5. Representative bacula of members of the guyannensis-gjroup (a-g) and the simonsi-gfoup (h-k); scale = 5 mm. a, bm[nh] 52.1 124— Suriname: Zanderig. b, fmnh 95726— Suriname: Brokopando; Saramacca River, Loksie Hatti. c, AMNH 75803— Venezuela: Bolivar; Arabupu, Mt. Roraima (topotype of arabupu Moojen). d, amnh 7545 1 — Brazil: Roraima; Rio Cotingo, Limao. e, usnm 554847— Brazil: Amazonas; 72 km N Manaus. f, usnm 555653 — Brazil: Para; Altar da Chao, Rio Tapajos. g, bm[nh] 1.11 .3.64— Brazil: Minas Gerais; Araguay, Rio Jordao (topotype of roberti Thomas), h, amnh 71866— Peru: Loreto; Boca Rio Curaray. i, usnm 461305 — Peru: Ucayali; 59 km W Pucallpa. j, AMNH 2 1 3487 — Peru: Pasco; Bermudas de Loma Linda, k, fmnh 8426 1 —Peru: Madre de Dios; Itahuania. chrysaeolus of the trinitatus-group), and Type IV (part goeldii-gToup and part semispinosus-group). The baculum is relatively long and narrow, av- eraging in adult specimens nearly 8 mm long and approximately 2 mm wide at both the proximal and distal extremities (see fig. 5a-g; table 1). The shaft is rather straight, with little dorsoventral cur- vature and only slightly tapered lateral indenta- tions near mid-shaft. The proximal end is usually evenly rounded and paddle-shaped, although sam- ples from every examined locality include bacula with a basal median notch of varying depth. The distal tip shows only slight development of apical wings and a moderate median depression. Except for topotypes of P. roberd from Minas Gerais, Bra- zil, there is no demonstrable geographic variation in length and width measurements among bacula of the same age class from different geographic regions (see table 1 ). The few bacula examined of P. roberti are smaller in length and width relative to others of the group. However, roberti is a rather small animal (see Thomas, 1901), and the pro- portions of its baculum are similar to those of other members of the group (see fig. 1 2). simonsi-group (Figure 5h-k) Bacula of specimens referred to this group were described and figured by Didier (1962, pp. 416- 417, 419, 422) as Proechimys guyannensis brevi- cauda and P. hendeei, and by Patton and Gardner (1972) as P. hendeei. Didier's supposed specimen of guyannensis brevicauda (fmnh 62095) is clearly PATTON: SPECIES GROUPS OF PROECHIMYS 313 Table 1 . Measurements of length, distal width, and proximal width of bacula (mean ± SD) of spiny rats, subgenus Proechimys. Data are presented only for adult individuals (age classes 8 through 10 of Patton & Rogers, 1983). Species group/ region or taxon Age N Length Distal width Proximal width ^>'a/i/i^/t$/5-group Venezuela 8 2 6.38 ± 1.21 1.99 ± 0.12 2.17 ± 0.60 Guianas 8 8 7.44 ± 0.85 2.00 ± 0.30 2.06 ± 0.41 9 12 8.09 ± 0.58 2.17 ± 0.35 1.94 ± 0.51 10 9 8.87 ± 1.04 2.38 ± 0.36 1.83 ± 0.32 Central Brazil 8 8 6.81 ± 1.45 1.89 ± 0.23 1.87 ± 0.39 9 4 6.91 ± 1.35 2.28 ±0.18 2.29 ± 0.31 10 4 8.12 ± 0.86 2.16 ± 0.16 2.37 ± 0.24 Southern Brazil 8 1 6.08 1.28 1.74 [roberti] 10 1 5.43 1.79 2.27 goeldii-group Colombia-Bolivia 8 24 7.21 ± 1.05 2.46 ± 0.41 2.45 ± 0.43 9 31 7.11 ± 1.18 2.79 ± 0.45 2.57 ± 0.41 10 21 8.17 ± 1.02 3.11 ± 0.42 2.93 ± 0.45 Venezuela-Brazil 8 5 6.58 ± 1.01 2.08 ± 0.20 2.06 ± 0.44 9 9 7.44 ± 0.55 2.36 ± 0.44 2.58 ± 0.47 10 3 8.36 ± 0.81 2.48 ± 0.68 2.62 ± 0.20 longicaudatus-group Colombia-Northern 8 33 10.81 ± 1.73 5.13 ± 0.76 4.28 ± 0.62 • Peru 9 18 11.43 ± 1.61 5.75 ± 0.80 4.68 ± 0.69 10 11 11.67 ± 1.90 5.87 ± 1.02 5.12 ± 0.87 Central Peru 8 8 7.79 ± 0.65 4.57 ± 0.55 3.72 ± 0.92 9 11 9.30 ± 1.25 4.61 ± 0.45 4.10 ± 0.80 10 7 10.11 ± 0.66 5.02 ± 0.68 4.08 ± o.6:j Southern Peru-Bolivia 8 4 8.66 ± 1.45 4.85 ± 0.50 4.41 ± 0.55 9 6 9.60 ± 1.12 4.74 ± 0.79 4.46 ± 0.62 SE Bolivia-Brazil 8 2 10.17 ± 0.84 5.22 ± 0.82 4.43 ± 0.72 9 2 11.01 ± 0.49 5.13 ± 0.40 3.20 ± 0.30 10 4 11.08 ± 1.02 5.10 ± 0.32 5.02 ± 0.47 CMv/en-group Guianas 8 7 6.27 ± 0.49 5.51 ± 0.56 5.30 ± 0.50 9 6 7.21 ± 0.84 5.94 ± 0.82 5.72 ± 0.84 10 6 8.61 ± 0.80 7.09 ± 0.45 6.70 ± 0.58 Central Brazil 8 3 5.95 ± 1.29 5.18 ± 1.24 4.69 ± 1.08 9 3 7.26 ± 0.41 6.02 ± 0.54 5.54 ± 0.45 10 2 8.09 ± 1.45 6.14 ± 1.39 5.31 ± 1.05 Northern Peru 8 5 6.13 ± 0.85 5.82 ± 0.74 5.38 ± 0.50 9 3 6.96 ± 0.45 6.88 ± 1.29 6.29 ± 0.45 10 3 7.41 ± 0.75 7.40 ± 0.05 6.53 ± 0.49 2n = 40, Balta 8 1 7.27 5.76 4.45 9 1 7.39 4.99 4.71 5imo/i5/-group 8 8 8.42 ± 1.44 1.51 ± 0.21 1.93 ± 0.23 9 11 9.02 ± 1.06 1.89 ± 0.38 2.05 ± 0.27 10 11 10.09 ± 1.18 1.77 ± 0.22 2.50 ± 0.32 semispinosus-group semispinosus 8 9 8.54 ± 0.95 5.11 ± 0.91 4.21 ± 1.11 9 11 9.16 ± 1.26 6.11 ±0.67 4.67 ± 0.54 10 9 9.66 ± 1.01 6.19 ±0.31 4.94 ± 0.67 oconnelli 9 1 8.43 5.26 4.94 10 1 8.16 4.63 4.93 trinitatus-group trinitatus 9 1 9.97 3.14 3.10 10 1 10.46 3.32 3.15 magdalenae 8 2 8.07 ± 0.81 2.37 ± 0.41 2.56 ± 0.22 10 2 10.47 ±0.16 2.62 ± 0.30 2.88 ± 0.07 314 HELDIANA: ZOOLOGY Table 1. Continued. Species group/ region or taxon Age N Length Distal width Proximal width guairae 8 9 10 8.62 9.25 10.01 2.79 2.78 3.34 2.93 3.57 3.50 hoplomyoides 10 11.64 1.79 2.47 mincae 8 7.29 2.42 2.04 9 3 9.98 ± 0.99 2.94 ± 0.18 2.17 ± 0.25 chrysaeolus 8 6 9.01 ± 0.94 2.34 ± 0.27 2.74 ± 0.23 9 7 10.03 ± 0.73 2.34 ± 0.32 2.97 ± 0.44 10 3 10.97 ± 1.35 2.63 ± 0.66 3.11 ±0.31 decumanus-group 8 1 10.63 3.84 3.41 9 2 9.16 ± 1.44 2.93 ± 0.64 2.80 ±0.52 10 3 11.83 ± 1.30 3.08 ± 0.34 2.92 ±0.42 canicollis-group 9 1 7.74 2.80 2.55 10 1 8.11 2.64 2.78 referable to P. simonsi by the cranial characters that distinguish this group from P. brevicauda (see below). The baculum is elongate and narrow, with a rounded and slightly broadened base. In older specimens the base is often laterally expanded with thin wings of bone. The weakly expanded distal end is usually characterized by a small lateral plat- form on each side separated by a shallow median depression. In general aspects, the baculum of this group is similar to that described for the guyan- nensis-group, although it averages longer and nar- rower (see table 1). No geographic variation is ap- parent, but the samples are not adequate to document this. goeldii-group (Figure 6) Included in this group are bacula described and figured by Didier (1962) as Proechimys guyan- nensis Type IV and as P. quadruplicatus, both from Caqueta, Colombia. Specimens from Peru referred to P. brevicauda by Patton and Gardner (1972) also belong to this unit (see Gardner & Emmons, 1984, for the correct allocation of these specimens to P. steerei). One specimen figured by Martin (1970, p. 8; fig. 4d), from Riberalta, El Beni, Bo- livia, is also referable to this group. There is virtually no distinguishable geographic variation in the baculum of goeldii-group mem- bers among samples which range from southern Venezuela to Bolivia and from the Rio Tapajos to eastern Colombia and Peru (see fig. 6; table 1). In general form, this baculum is similar to that described for members of the guyannensis-group. At similar cranial ages, it is nearly the same length but slightly wider both basally and distally, giving the baculum a somewhat stouter appearance. The base varies in shape from rounded to bilobed with a median notch, the sides are parallel to only slightly concave, and the tip shows only faint development of apical wings and a median depression. In lateral view, the baculum is straight to slightly convex dorsally and concave ventrally. trinitatus-group (Figure 7) Didier ( 1 962, pp. 4 1 2, 4 1 7^ 1 8) described bacula of specimens referred to this group as Proechimys guyannensis Type III (which represent P. chry- saeolus) and as P. guyannensis mincae. The baculum is long (averaging over 10 mm in adults; table 1) and narrow, but except for hoplo- myoides it is considerably stouter than that of 5/- wo«5/-group members which resemble them in overall length (see fig. 1 2). The lateral margins are only slightly concave; the base is broadened and rounded, usually with a distinct median notch; and the distal tip has only slightly developed apical wings and a median depression (fig. 7). decumanus-group (Figure 8a-b) The baculum of Proechimys decumanus has ap- parently not been described before. It is similar in general size and shape to that of the trinitatus- group (fig. 12; table 1), being elongate yet stout, and with rather parallel sides. The base is some- what rounded and the distal tip only slightly ex- PATTON: SPECIES GROUPS OF PROECHIMYS 315 Fig. 6. Representative bacula of members of the goeldii-gxoup; scale = 5 mm. a, fmnh 7 1 1 78 —Colombia: Caqueta; Florencia. Mantanito. b, mvz 157955 — Peru: Amazonas; La Poza, Rio Santiago, c, amnh 71592— Peru: Loreto; Boca Rio Curaray. d, amnh 76268— Peru: Loreto; Sarayacu. e, usnm 530931— Peru: Madre de Dios; Rio Manu. f, amnh 97017— Venezuela: Amazonas; Esmeralda, g, usnm 4151 17— Venezuela: Amazonas; Capibara, Casiquiare Canal, h, AMNH 92272— Brazil: Amazonas; Rosarinho, Rio Madiera, i, amnh 96828 — Brazil: Para; Ilha do Taiuno, Rio To- cantins. panded, with a weak median depression (fig. 8a- b). canico/Zis-group (Figure 8c-d) The baculum of Proechimys canicollis was de- scribed and figured by both Didier ( 1 962, p. 419) and Martin (1970, p. 8). It is most similar to that of the goeldii-group in both shape and size, being relatively short and stout with a rounded base, weakly concave sides, and a rather flat distal tip with only weakly developed apical wings (see fig. 8c-d; table 1). longicaudatus-group (Figure 9) Bacula of this group were figured by Didier (1962, p. 410, fig. 2, p. 412) as Proechimys guyannensis Type II (specimens from Caqueta, Colombia, and Santa Cruz, Bolivia), and by Martin (1970, p. 8, fig. 4c,e-k), also as P. guyannensis, from south- eastern Peru, Bolivia, and southwestern Brazil. Fatten and Gardner (1972) described and figured the bacula of Peruvian specimens of this group as P. longicaudatus (now referred to P. brevicauda [see Patton & Rogers, 1983; Gardner & Emmons, 1984]). In general aspect, the baculum is elongate and broad, with well-developed apical wings (see fig. 9). The margins are concave and the proximal and distal ends are usually about equal in width. In some specimens, the proximal end bears a median indentation of variable depth; in others, the prox- imal base is evenly rounded. The shaft is arched dorsally from base to tip and transversely concave along its entire ventral length. While overall length varies considerably, as do width measures to a lesser extent (table 1), the uniform and character- 316 FIELDIANA: ZOOLOGY Fig. 7. Representative bacula of members of the trinitatus-group; scale = 5 mm. a, P. chrysaeolus. bm[nh] 1 2.4.2.4— Colombia: Santander; Margarita, b, P. chrysaeolus, fmnh 69109— Colombia: Bolivar; San Juan Nepomuceno. c, P. guairae, mvz 168945— Venezuela: Portuguesa; Sto. Domingo, d, P. magdalenae, usnm 4997 19— Colombia: Antioquia; 22 km S and 22 km W Zaragoza. e, P. trinitatus, mvz 168946— Trinidad: Chaguaramas. f, P. hoplomyoides, amnh 75827— Venezuela: Bolivar; Arabupu, Mt. Roraima. istic shape of the baculum renders members of this group easily identifiable. Geographically, samples allocated to P. brevicauda average larger and broader in the northern (e.g., southern Co- lombia, Ecuador, and northern Peru) than in more southern localities (e.g., southeastern Peru and ad- jacent Bolivia); samples from southeastern Bolivia and Brazil referred to P. longicaudatus approach the general size of northern samples of P. brevi- cauda (see table 1). cifvf'm-group (Figure 10) Didier (1962, p. 411) recorded one specimen (fmnh 1 8 198) of Proechimys cuvieri from Guyana in his P. guyannensis Type I bacular group. This group is otherwise made up of specimens of P. semispinosus (see below), although the bacula of P. cuvieri and some populations of semispinosus have similarities of shape in common (see figs. 10- 11). Patton and Gardner (1972) figured and de- FiG. 8. Representative bacula of the decumanus-group (a-b) and canicollis-group (c-d); scale = 5 mm. a, fmnh 82023— Peru: Piura; Laguna Lamadero. b, fmnh 81199— Peru: Tumbes; Matapalo. c, usnm 2801 13— Colombia: Magdalena; Rio Cesar, opposite El Orinoco, d, usnm 2801 14— Colombia: Magdalena; Rio Cesar, opposite El Orinoco. PATTON: SPECIES GROUPS OF PROECHIMYS 317 Fig. 9. Representative bacula of members of the longicaudatus-group; scale = 5 mm. a, fmnh 7 1 1 74— Colombia: Caqueta; Florencia, Mantanita. b, mvz 155034— Peru: Amazonas; Huampami, Rio Cenepa. c, mvz 157934— Peru: Amazonas; La Poza, Rio Samiago. d, amnh 71877 — Peru: Loreto; Boca Rio Curaray. e, mvz 157854— Peru: Madre de Dios; Lago Sandoval, f, fmnh 1 19356— Bolivia: El Beni, San Pedro, g, bm[nh] 28.2.9.48— Bolivia: Santa Cruz; Buenavista. h, bm[nh] 3.7.7.94 — Brazil: Malo Grosso; Serra de Chapada. scribed bacula of specimens from eastern Peru with 2n = 40 karyotype that they referred to P. guy- annensis, and that have this type of baculum. The baculum is massive, with a broad shaft and a thickened and expanded base (see fig. 10; table 1). In cross section, the proximal two-thirds is convex dorsally and deeply concave ventrally. The distal end has a pair of diverging apical extensions separated by a wide median depression of varying depth. This is the most distinctive of any of the bacular types in Proechimys. It characterizes spec- imens from widely scattered localities in the Guianas and along the entire length of the Amazon River, as well as the karyotypically differentiated 2n = 40 form from Balta, Rio Curanja, Ucayali, Peru (see Patton & Gardner, 1972). These latter specimens were considered to be conspecific with P. guyannensis by Gardner and Emmons (1984) on karyotypic grounds. In bacular characters, how- ever, they are clearly different from guyannensis- group members and are placed here solely because of these uniquely shared bacula. The true taxo- 318 HELDIANA: ZOOLOGY Fig. 10. Representative bacula of members of the CMv/m-group; scale = 5 mm. a, bm[nh] 7.6.20.5— Guyana: Hyde Park. Demarara River, b, amnh 140540— Guyana: Kamakusa. c, amnh 96844— Brazil: Para; Ilha do Taiuno, Rio Tocantins. d, amnh 96867 — Brazil: Para; Ilha do Taiuno, Rio Tocantins. e, mvz 157874— Peru: Amazonas; La Poza, Rio Santiago, f, amnh 7409 1 — Peru: Loreto; Orosa, Rio Amazonas. nomic position of these specimens remains an enigma, as material clearly assignable to this form has not been found elsewhere. In the meager samples available, there appears to be little geographic variation in bacular size within P. cuvieri. although specimens from the Guianas average slightly larger than those from Brazil or Peru (table 1 ). semispinosus-^oup (Figure 11) Didier (1962, pp. 409-411) defined his Pro- echimys guyannensisType I baculum based largely on specimens from western Colombia which rep- resent the semispinosus-gToup as defined herein. Interestingly, three of the individuals he included in this group (fmnh 69063, 69064, 69071) were also listed as members of his guyannensis Type IV complex (Didier, 1962, p. 415), which is oth- erwise composed of specimens here referred to the goeldii-group. All cranial characters, as well as bacular ones (see fig. 1 1 b), show that these three sp)ecimens represent P. semispinosus. Patton and Gardner (1972, pp. 16-17) also described and fig- ured specimens of P. semispinosus from Costa Rica. The additional specimens examined here do not differ importantly from the descriptions provided in these two papers. In general aspects, the baculum is intermediate between the longicaudatus and cuvieri groups. The shaft is long and massive, with deeply concave margins, a broadly expanded and thickened base, and a wide distal portion with well-developed api- cal wings separated by a median depression. In cross section, the proximal and distal portions are convex dorsally and deeply concave ventrally. The width across the distal portion of the baculum typically exceeds that of the proximal portion (ta- ble 1 ), which is also characteristic of the cuvieri- PATTON: SPECIES GROUPS OF PROECHIMYS 319 Fig. 1 1 . Representative bacula of members of the semispinosus-group {P. semispinosus [a-e] and P. oconnelli [f- g]); scale = 5 mm. a, usnm 592694— Panama: Canal 2^ne. b, fmnh 69063— Colombia: Bolivar; Socorre, upper Rio Sinu. c, FMNH 70085— Colombia: Choco; Unguia. d, fmnh 89525— Colombia: Narino; La Guayacana. e, fmnh 90150— Colombia: Cauca; La Boca, f, mvz 99680— Colombia: Meta; Villavicencio. g, fmnh 88050— Colombia: Meta; Los Micos, San Juan de Aramas. group. However, specimens from the northern (Costa Rica and western Panama) and southern limits (Cauca, Colombia, and southward) of the range of P. semispinosus, as well as those of P. oconnelli (table 1), tend to be more symmetrical, those from central Panama south through Choco, Colombia, more expanded distally (compare fig. 1 la-d with 1 le-g). Qualitative Cranial Characters The usual morphometric approach to specific and infraspecific taxonomy of small mammals has met with little success in studies of Proechimys. In part, this is due to the large age-related com- ponent of character variation within localities that obscures any geographic patterns and species dif- ferences (see Patton & Rogers, 1983). Use of qual- itative characters has proven more successful, but only Patton and Gardner (1972) and Gardner and Emmons (1984) have marshalled such features as palatal structure, bullar septal pattern, temporal ridge development, and counterfold pattern on the cheekteeth into coherent patterns that identify geographically overlapping forms. Characters of the palate, in particular, proved to be concordant with bacular and karyotypic differences in delin- eating taxa in Peru (Patton & Gardner, 1972). Here, I focus on the qualitative description of five cranial features that prove useful in the dis- crimination of sympatric taxa and in the definition of homogeneous regional units. These include (see Patton & Gardner, 1972; Gardner & Emmons, 320 FIELDIANA: ZOOLOGY [ E ^ e ■D O 7- 5- 3- 2- cuvieri \ 2n=40 semispinosus I oconnelli I brevicauda I longicaudatus \roberfi guairoe ^trinitatus sfeerei 1 • •__| 1 canicollis^ magdalenae^ \ T ]guyonnensis \ A Ah decumanus Simons I chrysaeolus hop lomyo ides T- 8 9 — 1- 10 II T- 12 13 Bacular Length (mm) Fig. 1 2. Bivariate plot of distal bacular width and bacular length measurements for taxa of spiny rats, subgenus Proechimys. AH individuals represent age class 10 of Patton and Rogers (1983). Circles indicate means, vertical and horizontal lines indicate standard deviations. See text for the allocation of taxa into species groups. 1984): (1) shape and structure of the incisive fo- ramen; (2) angle and depth of mesopterygoid fossa; (3) degree of development of a bony groove in the floor of the infraorbital foramen; (4) degree of de- velopment of the temporal ridges across the pa- rietals; and (5) the counterfold pattern of the upper and lower cheekteeth. These structures were used by Moojen ( 1 948) with limited success to segregate sympatric taxa in Brazil, but he failed to use them to group regional samples into consistently de- fined morphological entities. Hence, his analysis appears to show that these features are much more variable, and thus of less utility, than actually proves to be the case. Other qualitative features of the cranium, such as the size and shape of the bullae, the size and position of the hamular processes of the ptery- goids, and the degree of lateral indentation of the paroccipital processes, need further examination (see, e.g., Patton & Gardner, 1972). My prelimi- nary analysis of each of these features indicates strong concordance with those characters used here to define taxonomic limits. Shape and Structure of the Incisive Foramen Specimens of each recognized species group were analyzed for the following characteristics of the incisive foramen: (1) general shape and size [lyre- shaped, constricted posteriorly; oval; evenly ta- pered posteriorly or parallel sided]; (2) presence or absence of grooves extending onto the anterior portion of the palate; (3) flanged or flat postero- lateral margins of the foramen; (4) degree of de- velopment of the maxillary and premaxillary por- PATTON: SPECIES GROUPS OF PROECHIMYS 321 Fig. 1 3 (top). Representative incisive foramina of specimens of P. brevicauda of the longicaiddatus-group; scale = 5 mm. a, mvz 1 53596 — Peru: Amazonas; Huampami, Rio Cenepa. Arrows indicate median palatal ridge and elevated flange marking posterolateral foraminal margins, b, Same specimen as in a, emphasizing the strongly keeled maxillary portion of the foraminal septum (arrow), c, mvz 153607 — Peru: Amazonas; Huampami, Rio Cenepa. d, mvz 157855 — Peru: Amazonas; La Poza, Rio Santiago. The premaxillary (pm), vomerine (v), and maxillary (m) portions of the foraminal septum are identified; arrows indicate sutures between these elements. Fig. 14 (bottom). Representative incisive foramina of sf>ecimens of the CMv/er;-group {P. cuvieri [a-c] and the 2n = 40 karyotypic form from Balta, eastern Peru [d]); scale = 5 mm. a, fmnh 95720— Suriname: Brokopando; Saramacca River, Loksie Hattie. Arrow indicates direct contact between premaxillary and maxillary portions of the septum. Note that only a small part of the vomerine portion is visible ventrally. b, mvz 160091 —Venezuela: Bolivar; 69 km S Rio Cuyuni. An expanded vomerine portion of the septum is evident (arrows identify vomerine contact with the premaxilla and maxilla), c, mvz 157874— Peru: Amazonas; La Poza, Rio Santiago. The vomerine portion of the septum is visible ventrally, widely separating the premaxillary and maxillary components (arrows), d, lsu 14425- Peru: Ucayali; Balta, Rio Curanja. Note the small size of this specimen relative to the others, and the elongated premaxillary portion of the septum. tions of septum; (5) whether the vomerine portion of septum is visible ventrally; (6) whether or not the maxillary portion of the septum is keeled; and (7) whether the anterior portion of the palate has a median ridge. Descriptions of the incisive fo- ramina for representatives of each of the sp>ecies groups are given below and are illustrated in Fig- ures 1 3-20; examples were chosen to express the 322 FIELDIANA: ZOOLOGY full range of character variation for each group, regardless of the specific localities from which specimens were collected. that of P. cuvieri. The maxillary portion is well developed and appears to contact directly the pre- maxillary portion. longicaudatus-ffroup (Figure 13a-d) The general features of this type of incisive fo- ramen were given by Patton and Gardner (1972, p. 10). Figure 13 illustrates the range of form typ- ical for members of the group. The most diagnostic features include: a lyre-shaped foramen, usually with a strongly constricted posterior portion; the maxillary terminus of the foramen deeply grooved onto the anterior palate; the posterolateral margins of the foramen strongly flanged; an expanded, long premaxillary portion of the septum, usually ex- tending more than one-half its length; a well-de- veloped and strongly keeled maxillary portion of the septum (see fig. 13b), the maxillary keel ex- tending onto the anterior palate resulting in a well- developed median ridge; and a vomerine portion of the septum exposed ventrally between the pre- maxillary and maxillary components. This is one of the more consistently recognizable types of incisive foramina within the genus Pro- echimys, varying mostly in the degree of constric- tion at the posterior margins, hence in the degree of the general lyre-shape. goeldii-group (Figure 15a-d) The general features of this foraminal type were provided by Patton and Gardner (1972, p. 4) un- der P. brevicauda. The foramen is usually only weakly lyre-shaped, or with margins tapering slightly posteriorly or parallel-sided. The premax- illary portion of the septum is short, usually one- half or less of the length of the foramen; the max- illary portion varies greatly, being usually rather weak and attenuate, often not in contact with the premaxillary portion (fig. 15d) but sometimes broadly spatulate and filling much of the foramen (fig. 15a). Nevertheless, the vomer is only rarely exposed ventrally, being completely enclosed in the premaxillary sheath. The maxillary portion of the septum often exhibits a median vacuity (fig. 1 5b); it may be slightly ridged, but is never strongly keeled, and seldom does this ridge extend onto the anterior palate (fig. 15b). Thus, there are only moderately developed grooves onto the anterior palate, and the posterolateral margins of the fo- ramen are only moderately flanged. cuvieri-group (Figure 14a-d) The incisive foramen of Proechimys cuvieri is most similar in structure to that of the longicau- datus-gToup. The general conformation is weakly to strongly lyre-shaped with strongly developed posterolateral flanges. The anterior palate, how- ever, is only weakly to moderately grooved re- sulting in a slight median ridge. The premaxillary portion of the septum is strongly developed, ex- tending more than one-half its length; the maxil- lary portion varies from stout to attenuate, but is always short and is only weakly keeled. The vomer is varyingly exposed ventrally (compare fig. 14a with 14b-c). The 2n = 40 specimens from Balta which have bacula similar to P. cuvieri share only some fo- raminal characters with that taxon (fig. 1 4d). The foramen is weakly lyre-shaped and the postero- lateral margins are only weakly flanged, hence the anterior palate is scarcely grooved. Nevertheless, the premaxillary portion of the septum is elon- gated and broad, similar in shape and structure to semispinosus-group (Figure 16a-d) Specimens of this group from Costa Rica were described by Patton and Gardner (1972, p. 15). Foraminal shape varies from rather evenly tapered margins to moderately lyre-shaped ones. The pos- terolateral margins are usually strongly flanged, creating deep grooves extending onto the anterior palate despite only moderate development of a medial ridge (compare fig. 1 6b with 1 6c). The pre- maxillary portion of the septum is dominant, usu- ally broadly filling the foramen and extending well over one-half its length. The maxillary portion varies from moderately developed to attenuate, but is almost always in direct contact with the premaxillary portion. The vomer is completely en- cased within the premaxilla and thus is not visible in ventral asp>ect. simonsi-gToup (Figure 17a-d) Again, the incisive foramen of the simonsi-group was described fully by Patton and Gardner (1972, PATTON: SPECIES GROUPS OF PROECHIMYS 323 Fig. 1 5 (top). Incisive foramina of P. steers i oi ihe goeldii-groxip. All are from La Poza, Rio Santiago. Amazonas. Peru; scale = 5 mm. a, mvz 157949. Note the enlarged maxillary' portion of the septum (m) and the direct contact (arrow) between it and the premaxillary portion (pm). b, mvz 1 57956. Note the moderately developed posterolateral flange (arrow), c, mvz 157861. d, mvz 157869. Note the attenuate maxillary portion of the septum and the lack of contact between it and the premaxilla portion. Fig. 16 (bottom). Incisive foramina off. semispinosus of the semispinosus-group; scale = 5 mm. a, mvz 165794— Panama: Panama; 0.8 km N Paraiso. b, fmnh 90169— Colombia: Choco; Rio Baudo. Note the direct contact between the premaxillary (pm) and maxillary (m) portions of the septimi. c, fmnh 90177— Colombia; Choco; Rio Baudo. d, fmnh 70080— Colombia; Choco; Unguia. Note the well-developed posterolateral flange (arrow). p. 19). This is a distinctive foramina] type, as all specimens examined were consistent in most fea- tures despite variation in overall shape. The fo- ramen is oval in general shape, although often asymmetrical in anteroposterior direction (fig. 17b). The premaxillary^ portion of the septum is rather short, usually no more than one-half the length of the foramen. The maxillary portion is usually weak and atteniiate, only rarely in contact with the pre- maxillary portion. When the septum is complete (fig. 1 7a), the vomer is either completely enclosed by the premaxilla or barely visible (fig. 1 7c). The posterolateral margins are flat, not flanged, and no groove extends onto the anterior palate. Rather, the palate is noticeably flat and smooth, without a medial ridge. guyannensis-group (Figure 18a-d) This foraminal type is virtually indistinguish- able firom that described for the simonsi-group. 324 RELDIANA: ZOOLOGY I If 1l 11 Fig. 17 (top). Incisive foramina of P. simonsi of the simonsi-group; scale = 5 mm. a, mvz 155045 — Peru: Amazonas; headwaters of Rio Kagka. Note lack of posterolateral flanges or palatal grooves, b, mvz 157914— Peru: Amazonas; La Poza, Rio Santiago, c, mvz 136654— Peru: Ucayali; Balta, Rio Curanja. Note slightly exposed vomer (y) and attenuate maxillary portion of septum, d, mvz 168955— Peru: Madre de Dios; Albergue, Rio Madre de Dios. Fig. 18 (bottom). Incisive foramina of the guyannensis-group (a-b, arabupu; c, oris; and d, roberti); scale = 5 mm. a, amnh 139741— Venezuela: Bolivar; Auyantepui. Note lack of posterolateral flanges and anterior palatal grooves, and the attenuate maxillary portion of the septum, b, mvz 160094— Venezuela: Bolivar; 4 km E El Pauji. Note the lack of contact between the premaxillary and maxillary portions of the septum, c, amnh 93997 — Brazil: Para; Faro, north bank Rio Amazon. Note weakly developed posterolateral flanges, d, amnh 134309— Brazil: Goias; Anapolis. The shape is oval, although often unequal (fig. 1 8d). The anterior palate is flat, without grooves or a median ridge, and the posterolateral margins of the foramen are not flanged, or only weakly flanged (fig. 18c shows maximal development of flanges). The premaxillary portion of the septum is relatively short, usually less than one-half the length of the foramen, and the maxillary portion is attenuate, usually not in contact with the pre- maxillary portion. The vomer generally does not contribute to the ventral aspect of the septum. trinitatus-group (Figure 19a-d) Members of this species group generally exhibit the most enlarged foramina within the subgenus Proechimys. Specimens referred to mincae, polio- PATTON: SPECIES GROUPS OF PROECHIMYS 325 ^ r%Si Fig. 19 (top). Incisive foramina of the trinitatus-group; scale = 5 mm. a, P. mincae, fmnh 13203— Colombia: Magdalena; Minca (topotype). b, P. ochraceous, fmnh 1 8688— Venezuela: Zulia; El Panorama, Rio Aurare (topotype). c, P. guairae. fmnh 92588— Colombia: Arauca; Rio Cobaria. d, P. chrysaeolus. fmnh 69037— Colombia: Bolivar; San Juan Nepumoceno. Fig. 20 (bottom). Incisive foramina of the decumanus- and canicollis-g^oups; scale = 5 mm. a, P. decumanus, fmnh 82024— Peru: Piura; Laguna Lamadero. b, P. canicollis. fmnh 691 1 1— Colombia: Bolivar; San Juan Nepu- moceno. pus, and ochraceous have smoother, less ridged palates and ovoid foramina lacking posterolateral flanges (fig. 1 9a-b); those referred to guairae, trin- itatus. urichi, hoplomyoides, and chrysaeolus have somewhat more lyre-shaped foramina with weakly to moderately flanged posterolateral margins which define grooves extending onto the anterior palate (fig. 1 9c-d). In all forms the premaxillary portion of the septum is enlarged, usually extending one- half or more of the length of the foramen, while the maxillary pKJrtion is attenuate, most often not in direct contact with the premaxillary portion. Only in specimens referred to guairae and trini- tatus does the maxillary portion of the septum show a medial ridge (fig. 1 9c). H/a/i«-group chrysaeolus 5 0.25 0.75 mincae 19 0.05 0.95 trinitatus 12 0.08 0.92 urichi 8 1.00 guairae 3 1.00 ochraceous 1 1.00 poHopus 1 1.00 hoplomyoides 1 1.00 canicoUis-group 13 1.00 decumani4s-gFoup 17 ■ 0.59 0.29 0.12 annensis; simonsi-, trinitatus-, and canicollis- groups characteristically do not (table 2). Ventral Canal of the Infraorbital Foramen The infraorbital ntrvt courses near the medial floor of the infraorbital foramen, producing a canal of varying distinctness in many caviomorph ro- dents, including Proechimys (Woods, 1984). Both Moojen (1948) and Gardner and Emmons (1984) suggested that the presence and degree of devel- opment of this canal or groove has taxonomic sig- nificance in the genus. To evaluate this view, sev- eral grades of notch development were scored for representatives of each species group: 1. No groove present (fig. 22c) 2. Groove present, with moderately developed lateral flange (fig. 22b) 3. Groove present, with extreme development of a lateral flange (fig. 22a) Intermediate levels between each of these classes were also recognized, providing five scores ranging from 1.0 to 3.0 in half increments. ^ The degree of groove development for geo- graphic and taxonomic representatives of each species group is given in Table 3. While groove development varies v^adely within the genus, each species group displays a relatively narrow range. Members of the longicaudatus-, trinitatus-, cuvie- ri-, and guyannensis-groups exhibit the least de- velopment of a groove, the longicaudatus-group rarely showing any groove at all. The goeldii- and 5/won5/-groups show moderate development of the groove, while a notch is most strongly developed in the semispinosus-group. Specimens of P. ocon- nelli of the semispinosus-group consistently dis- played the most extensive lateral flange develop- ment. Geographic variation in the expression of groove development within each species group is virtually absent. The longicaudatus- and semispinosus- groups, with the lowest and highest mean scores, respectively, are the most uniform geographically; the guyannensis- and goeldii-gfoups are most vari- able (table 3). considerable overlap in the expression of this fea- ture, although decumanus-, longicaudatus-, goel- dii-, and CMv/er/-group members consistently show some ridge development while taxa of the guy- Angle and Depth of the Mesopter>'goid Fossa The angle formed by the anterior margins of the mesopterygoid fossa was measured with a pro- 328 FIELDIANA: ZOOLOGY Table 3. Mean scores and ranges for the develop- ment of the infraorbital foramen canal of spiny rats, subgenus Proechimys. See text for explanation of scoring system. Fig. 22. Degrees of development of the canal or groove on the medial floor of the infraorbital foramen that accommodates the infraorbital nerve; see text for description of scoring system used; scale = 5 mm. a, Score = 3.0, P. oconnelli, mvz 99685— Colombia: Meta; Villavicencio. b, Score = 2.0, P. oconnelli, mvz 99684— Colombia: Meta; Villavicencio. c. Score = 1 .0, P. quad- ruplicatus, ummz 80069— Ecuador: NapK); San Francis- co, Rio Napo. Species group/ region or taxon N Mean Range longicaudatus-group Colombia-Ecuador 6 1.1 1.0-1.5 Northern Peru 145 1.1 1.0-1.4 Central Peru 39 1.1 1.0-1.5 SE Peru-Bolivia 35 1.1 1.0-1.5 Southern Bolivia-Brazil 32 1.0 1.0-1.5 cuvieri-group Guianas 3 1.4 1.0-1.5 Central Amazon 37 1.3 1.0-2.0 Peru 4 1.4 1.0-2.0 2n = 40, Balta 5 1.4 1.0-2.0 semispinosus-group Central America 82 2.6 1.0-3.0 Northern Colombia 63 2.7 1.5-3.0 Central Colombia 31 2.6 2.0-3.0 Northern Ecuador 19 2.7 2.0-3.0 Southern Ecuador 29 2.7 1.5-3.0 oconnelli 20 2.8 2.0-3.0 goeldii-group Northern Peru 104 2.1 1.0-3.0 Central Peru 31 1.8 1.0-3.0 SE Peru-Bolivia 17 1.5 1.0-2.0 Venezuela-Brazil 21 1.8 1.0-3.0 Central Amazon 63 1.5 1.0-2.5 Eastern Amazon 123 1.9 1.0-3.0 guyannensis-group Goias, Brazil 20 1.2 1.0-1.5 Para, Brazil 50 1.1 1.0-1.5 Central Amazon 31 2.0 1.0-3.0 NW Brazil 34 2.0 1.5-3.0 SE Venezuela-Brazil 95 1.5 1.0-3.0 Suriname 1 2.5 simonsi-group Colombia-Ecuador 5 2.0 Northern Peru 32 2.1 1.5-3.0 Central Peru 15 1.6 1.0-2.0 SE Peru-Bolivia 27 1.8 1.0-2.5 trinitatus-group chrysaeolus 7 1.7 1.0-2.0 mincae 19 1.8 1.0-2.0 trinitatus 12 1.9 1.5-3.0 urichi 6 1.3 1.0-2.0 guairae 3 2.0 1.5-3.0 ochraceous 1 1.5 poliopus 1 1.5 hoplomyoides 1 1.0 canicollis-group 13 1.2 1.0-1.5 decumanus-group 17 1.4 1.0-2.0 PATTON: SPECIES GROUPS OF PROECHIMYS 329 Fig. 23. Angle and extent of the mesopterygoid fossa for representative taxa of Proechimys; scale = 5 mm. a, P. brevicauda, mvz 157854— Peru: Amazonas; La Poza, Rio Santiago, b, P. steerei, mvz 157888— Peru: Amazonas; La Poza, Rio Santiago, c, P. simonsi, mvz 157950— Peru: Amazonas; La Poza, Rio Santiago, d, P. guyannensis, amnh 75820— Venezuela: Bolivar; Arabupu, Mt. Roraima (topotype of arabupu). tractor to the nearest degree, and the maximal penetration of the fossa into the palate was scored relative to the cheekteeth, as follows: 1 . Not extending to the posterior margins of M3 2. Extending to the posterior one-half of M3 3. Extending to anterior one-half of M3 4. Extending to posterior one-half of M2 5. Extending to anterior one-half of M2 These two characters, angle and depth, are cor- related in that the greater the depth usually the more acute the angle (fig. 23). Table 4 provides data for the mesopterygoid fossa characters for representatives for each rec- ognized species group of Proechimys. Members of the longicaudatus-group consistently have the broadest angle with the most shallow fossa (fig. 23a); those of the simonsi-group have the most acutely-angled fossa and, with members of the guyannensis-group, the deepest penetration into the palate (fig. 23c-d). Most other groups show moderate angles and degree of penetration. Except for the guyannensis-group, where samples referred to P. oris (Para state, Brazil) and P. roberti (Goias and Minas Gerais states, Brazil) are quite different from other samples examined (table 4), there is little geographic variation in the expression of me- sopterygoid fossa characters within each species group. Counterfold Pattern of the Cheekteeth Early attempts to establish systematic relation- ships within Proechimys placed considerable em- phasis on variation in counterfolds of the cheek- teeth, both in number and pattern (e.g., Hershkovitz, 1948; Moojen, 1948). In general, these earlier studies indicated that fold number and pattern are quite variable geographically with- in taxa. As a result, counterfolds have been used primarily to recognize taxa sympatric at given lo- calities rather than as a character complex capable of uniting distinct populations into cohesive and 330 FIELDIANA: ZOOLOGY Table 4. Mesopterygoid fossa (mpf) angle (in degrees) and depth scores of spiny rats, subgenus Proechimys. See text for detailed descriptions of character-states. N MPF angle MPF depth Species group/region or taxon Mean ± SD Range Mean Range longicaudatus-%xo\ix> Colombia-Ecuador 6 79.0 ± 8.9 64-89 2.0 1-3 Northern Peru 145 79.6 ± 5.8 64-107 2.0 1-3 Central Peru 39 78.7 ± 7.1 64-93 1.9 1-3 SE Peru-Bolivia 35 72.7 ± 6.3 60-90 1.9 1-3 Southern Bolivia-Brazil 32 77.6 ± 8.5 60-91 2.2 1-3 cuv/er/-group Guianas 3 69.7 ± 4.9 64-73 2.0 Central Brazil 37 72.8 ± 5.6 57-92 2.2 1-3 Peru 4 66.7 ± 5.7 59-71 2.0 1-3 2n = 40, Balta 5 56.4 ± 3.2 52-60 3.2 2-4 semispinosus-ffoup Central America 82 56.4 ± 6.4 45-73 2.6 1-3 Northern Colombia 63 57.7 ±6.1 45-72 2.5 2-4 Central Colombia 31 57.1 ± 5.6 46-62 2.8 2-4 Northern Ecuador 19 62.6 ± 6.8 53-75 3.0 Southern Ecuador 29 62.1 ±4.8 53-71 2.8 2-4 oconnelli 20 63.3 ± 5.3 56-72 2.9 2-4 goeldii-group Northern Peru 104 60.6 ± 7.0 49-80 2.7 1-4 Central Peru 31 65.3 ± 7.1 45-76 2.6 2-3 SE Peru-Bolivia 17 67.9 ± 7.0 55-79 2.4 2-3 Venezuela-Brazil 21 68.2 ± 6.2 56-80 2.7 2-4 Central Brazil 63 62.1 ± 5.1 52-78 2.9 2-4 Eastern Brazil 123 65.3 ± 5.1 50-80 2.6 1-4 guyannensiS'group Goias, Brazil 20 67.7 ± 6.1 54-78 2.8 2-3 Para, Brazil 50 64.2 ± 6.2 44-82 3.0 2-4 Central Brazil 31 56.1 ± 4.9 45-66 3.7 3-5 Venezuela-Brazil 129 47.5 ± 6.3 34-67 3.9 2-5 Suriname 1 58.0 ... 4.0 simonsi-group Colombia-Ecuador 5 49.3 ± 6.9 39-53 3.6 3-4 Northern Peru 32 51.1 ± 3.3 47-62 4.1 2-5 Central Peru 15 53.3 ± 4.6 46-62 3.8 3-5 SE Peru-Bolivia 27 51.2 ± 4.5 45-64 3.2 2-4 trinitatus-group chrysaeolus 7 59.3 ±9.9 53-67 2.3 1-3 mincae 19 56.7 ±3.2 53-64 3.7 3-4 trinitatus 12 52.8 ± 4.9 48-65 3.4 3-4 urichi 6 56.3 ± 4.9 48-64 3.3 3-4 guairae 3 55.7 ± 3.1 53-58 3.3 3-4 ochraceous 2 48.5 ± 3.5 46-51 4.0 poliopus 2 47.0 ± 1.4 46-48 3.5 3-4 hoplomyoides 1 40.0 4.0 canicollis-gTOup 13 54.2 ± 2.9 50-61 3.0 decumanus-group 17 52.9 ± 3.5 46-59 3.1 2-4 definable groups. While there is geographic vari- ation in counterfold pattern and number within taxa of spiny rats, these characters can be used to form groups which are consistently separable from others. In this respect, the counterfold pattern and number reinforces group membership delineated by other morphological features, such as bacula and palatal characters. Illustrations of counterfold patterns for both up- per and lower toothrows for each species group PATTON: SPECIES GROUPS OF PROECHIMYS 331 Fig. 24. Upper (left) and lower (right) toothrows of P. semispinosus of the semispinosus-gro\np of Proechimys; scale = 5 mm. a, fmnh 90143— Colombia: Cauca: Rio Saija. b, fmnh 70072— Colombia: Choco; Unguia. (with the exception of the decumant4s-group) are presented in Figures 24-30. As with other char- acter complexes I examined, these examples were chosen to illustrate the range of variation. Geo- graphic variation within each group is summa- rized in Table 5. Counterfold number and, to a lesser extent, pat- tern change with increasing age. Obviously, folds become obliterated in advanced age, but even in moderately aged individuals smaller folds ^easily can become lost, and coalescence or isolation of folds occurs (see Moojen, 1 948). These age-related ^i^S^^ Fig. 25. Upper (left) and lower (right) toothrows of P. steerei of the goeldii-group; all specimens from La Poza, Rio Santiago, Amazonas, Peru; scale = 5 mm. a, mvz 157871. b, mvz 157863. c, mvz 157861. 332 HELDIANA: ZOOLOGY Fig. 26. Upper (left) and lower (right) toothrows of P. brevicauda (a-b) and P. longicaudatus (c) of the longicau- datus-gxoup; scale = 5 mm. a, mvz 157584— Peru: Amazonas; La Poza, Rio Santiago, b, mvz 157855 — Peru: Ama- zonas; La Poza, Rio Santiago, c, JRS 222— Paraguay: Chaco; 54 km E Agua Dulce (specimen to be deposited in National Museum, Asuncion, Paraguay). phenomena create difficulties in counting folds and clearly are partly responsible for some of the vari- ability observed within and between samples. To minimize this extraneous variation, the data as- sembled here are based on individuals in age classes 8 or 9 (as defined by Patton & Rogers, 1 983) where folds are still mostly confluent with the sides of each tooth. Taxa of the semispinosus-group (fig. 24; table 5), followed by those of the goeldii-group (fig. 25), Fig. 27. Upper (left) and lower (right) toothrows off. cuvieri of the CMv/er/-group; scale = 5 mm. a, mvz 1 57874— Peru: Amazonas; La Poza, Rio Santiago, b, mvz 160091— Venezuela: Bohvar; 69 km S Rio Cuyuni. PATTON: SPECIES GROUPS OF PROECHIMYS 333 Table 5. Counterfold patterns of the cheekteeth of spiny rats, subgenus Proechimys. Folds are given as frequencies for each tooth, based largely on specimens of the age classes 8 and 9 (Patton & Rogers, 1983). dPM* M' M' Species group/ region or taxon N 2 3 4 2 3 4 2 3 4 5em/5p/>J05M5-group Central America 76 1.00 0.92 0.08 0.70 0.30 Northern Colombia 36 1.00 0.89 0.11 0.71 0.29 Central Colombia 12 0.67 0.33 0.58 0.42 0.30 0.70 Northern Ecuador 10 0.80 0.20 0.30 0.70 1.00 Southern Ecuador 29 1.00 0.97 0.03 0.05 0.85 0.10 oconnelli 19 1.00 1.00 1.00 goeldii-group Colombia-Ecuador 9 0.25 0.75 0.12 0.88 0.08 0.92 Northern Peru 91 0.36 0.64 0.30 0.70 0.10 0.90 Central Peru 13 0.85 0.15 0.75 0.25 0.55 0.45 SE Peru-Bolivia 18 0.92 0.12 0.78 0.22 0.22 0.78 Brazil-Venezuela 21 0.79 0.21 0.58 0.42 0.25 0.75 Central Amazon 61 0.86 0.14 0.72 0.28 0.57 0.43 Eastern Amazon 108 1.00 1.00 0.98 0.02 longicaudatus-group Colombia-Ecuador 7 1.00 1.00 1.00 Northern Peru 134 1.00 1.00 0.97^ 0.03 Central Peru 19 1.00 1.00 1.00 SE Peru-Bolivia 36 1.00 1.00 0.95 0.05 Southern Bolivia-Brazil 37 1.00 1.00 1.00 cuvieri-group Guianas 3 1.00 1.00 1.00 Brazil 24 1.00 1.00 1.00 Peru 20 1.00 1.00 0.80 0.20 2n = 40, Balta 5 1.00 1.00 1.00 simonsi-group Colombia-Ecuador 12 1.00 1.00 1.00 Northern Peru 39 1.00 1.00 0.88 0.12 Central Peru 24 1.00 0.98 0.02 0.76 0.24 SE Peru-Bolivia 31 1.00 1.00 0.68 0.32 guyannensis-gxoup Goias 20 1.00 1.00 1.00 Para 50 1.00 1.00 0.98 0.02 Central Amazon 31 1.00 1.00 1.00 NW Brazil- Venezuela 34 1.00 1.00 1.00 SE Venezuela 95 1.00 0.02 0.98 0.03 0.97 Suriname 17 1.00 1.00 1.00 /r/«/7a/M5-group chrysaeolus 12 1.00 1.00 1.00 mincae 19 1.00 1.00 0.16 0.84 trinitatus 12 1.00 1.00 1.00 urichi 6 1.00 1.00 1.00 guairae 2 1.00 1.00 1.00 hoplomyoides 1 1.00 1.00 1.00 decumanus-group 17 1.00 1.00 1.00 canicollis-group 13 1.00 1.00 1.00 * Five folds occur on dpm4 with frequencies of 0.02 and 0.0 1 , respectively exhibit the most complex counterfold patterns. Specimens of P. semispinosus from central Colom- bia south into northern Ecuador have the highest average number of folds per tooth, particularly of the maxillary row (table 5); this number decreases slightly to the north and south. Proechimys ocon- nelli shows the lowest counterfold number for this group. In the goeldii-group, specimens from Co- 334 FIELDIANA: ZOOLOGY Table 5. Continued. M' dpin4 m, nij IHj 2 3 4 2 3 4 2 3 4 2 3 4 2 3 4 0.91 0.09 • 0.62 0.38 0.01 0.99 • 0.05 0.90 0.05 0.18 0.81 0.01 0.77 0.23 • 0.44 0.56 1.00 • 0.97 0.03 0.20 0.73 0.07 0.83 0.17 1.00 1.00 1.00 1.00 • 1.00 1.00 0.90 0.10 0.83 1.00 0.17 0.05 0.85 0.10 • 0.25 0.75 1.00 • 1.00 0.48 0.52 0.04 0.96 0.28 0.72 1.00 1.00 0.18 0.82 • 1.00 • 0.22 0.78 1.00 0.85 0.15 1.00 0.14 0.86 0.98* 0.95 0.05 0.93 0.07 1.00 0.23 0.77 1.00 1.00 • 1.00 1.00 0.58 0.42 1.00 1.00 • 1.00 1.00 0.54 0.46 1.00 1.00 • 1.00 0.07 0.93 0.70 0.30 0.06 0.93* 1.00 • 0.02 0.98 0.05 0.95 0.06 0.92 0.02 • 0.33 0.67 0.01 0.99 0.10 0.90 0.50 0.50 1.00 • 0.70 0.30 1.00 • 1.00 0.23 0.77 0.98 0.02 ■ 0.74 0.26 1.00 • 0.02 0.98 0.26 0.74 1.00 • 0.75 0.25 1.00 • 0.11 0.89 0.46 0.54 0.98 0.02 • 0.71 0.29 0.02 0.98 0.02 0.98 0.43 0.57 1.00 • 0.98 0.02 0.14 0.86 0.10 0.90 0.69 0.31 1.00 ■ 0.67 0.33 1.00 • 0.18 0.82 0.33 0.67 1.00 • 0.94 0.06 0.02 0.98 0.24 0.76 0.33 0.67 0.98 0.02 • 0.25 0.75 1.00 • 1.00 0.95 0.05 1.00 • 0.79 0.21 1.00 • 1.00 1.00 0.04 0.96 • 0.75 0.25 1.00 • 1.00 1.00 0.97 0.03 • 0.70 0.30 1.00 • 0.01 0.99 0.15 0.84 0.01 1.00 • 0.42 0.58 0.03 0.97 • 1.00 0.21 0.79 0.79 0.21 • 0.29 0.71 1.00 • 1.00 0.11 0.89 0.08 0.92 1.00 0.52 0.48 • 0.75 0.25 1.00 0.11 0.89 • 0.85 0.15 0.33 0.67 • 0.39 0.61 0.96 0.04 0.11 0.89 ■ 0.91 0.09 0.75 0.25 • 0.75 0.25 0.92 0.08 0.12 0.88 • 0.90 0.10 0.67 0.33 ■ 0.62 0.38 0.84 0.16 0.36 0.64 1.00 1.00 1.00 0.78 1.00 0.22 ■ 0.85 1.00 0.15 0.97 1.00 0.03 1.00 • 0.97 0.03 1.00 • 0.35 0.65 0.50 0.50 0.62 0.38 1.00 0.98 0.02 • 0.98 0.02 0.97 0.03 0.05 0.95 1.00 0.18 0.82 • 0.22 0.78 0.30 0.70 0.21 0.95 • 1.00 0.25 0.75 • 0.30 0.70 0.71 0.29 1.00 1.00 1.00 1.00 1.00 1.00 • 1.00 1.00 1.00 1.00 0.04 0.96 1.00 0.04 0.96 • 0.63 0.37 0.98 0.02 1. 00 1. 00 ••• 1.00 1.00 1.00 lombia south to northern Peru and east into Ven- ezuela and adjacent Brazil have the highest fold counts (table 5); those from Ecuador represent quadruplicatus, which was named by Hershkovitz (1948) for its high fold number. Counts decrease slightly to the south into Bolivia and central and eastern Brazil (table 5). The longicaiidatus-group members have a con- PATTON: SPECIES GROUPS OF PROECHIMYS 335 Fig. 28. Upper (left) and lower (right) toolhrows of P. simonsi of the s/mo/ts/ -group; scale = 5 mm. a, mvz 157950— Peru: Amazonas; La Poza, Rio Santiago, b, mvz 166035— Peru: Madre de Dios; Albergue, Rio Madre de Dios. sistent fold count for the upper teeth throughout their ranges (fig. 26; table 5). The lower cheekteeth, however, show a decrease in fold number from north to south: samples referred to P. longicau- datus are characterized by only two folds on the last molar; specimens referred to P. brevicauda typically have the first and second medial folds on mj displaying degrees of coalescence (see fig. 26a- b). The CMv/en-group shows a fold coimt similar to that of the longicaiuiatiis-group (table 5), and these two groups cannot be distinguished in pat- tern (compare fig. 26 with fig. 27). Taxa of the 5/mo/is/-group are somewhat intermediate be- tween the goeldii- and longicaudatm-groups in fold number (fig. 28). The upper cheekteeth, particu- larly M^, show a shght increase in number of folds from north to south, as does dpm4 (tal)le 5). This geographic pattern is the reverse of that seen in both the goeldii- and longicaudatiis-groups over the same part of the western Amazon Basin. Taxa of the guyannensis-group are uniform in number and pattern of counterfolds throughout their range (fig. 29; table 5). They are characterized by having three folds on most teeth, often with two folds on the lower molars, and only rarely four folds on dpm4. Members of the trinitatus-, decu- manus-, and canicollis-groups display the lowest counterfold number (table 5) and, hence, the sim- plest pattern (fig. 30). Specimens of P. canicollis have the least complex cheekteeth of any taxon in the subgenus Proechimys. with two folds on each tooth the general rule. Fig. 29. Upper (left) and lower (right) toothrows of P. guyannensis (a) and P. roberti (b) of the guyannensis-ffo\x'p\ scale = 5 mm. a, amnh 130737— Venezuela: Bolivar; Auyantepui. b, amnh 134309— Brazil: Goias; Anapolis. 336 HELDIANA: ZCX)LC>GY V^ilfrS-' Fig. 30. Upper (left) and lower (right) toothrows of representatives of the trinitatus (a-c) and canicoltis (d) groups; scale = 5 mm. a, P. chrysaeolus, fmnh 69039— Colombia: Bolivar; San Juan Nepumoceno. b, P. mincae, fmnh 13203— Colombia: Magdalena; Minca (topotype). c, P. guairae, fmnh 92588— Colombia: Arauca; Rio Cobaria. d, P. canicollis, FAiNH 69109— Colombia: Bolivar; San Juan Nepumoceno. Remarks and Prospectus Nine species groups of spiny rats (subgenus Proechimys) are defined herein, and 59 of the 67 names which have been proposed are allocated to one or another of these groups. Although the de- fined groups differ from those proposed recently by Gardner and Emmons (1984), it is reassuring that we have grouped taxa similarly, with only minor exceptions, despite our use of different suites of characters. Our common conclusions indicate that characters are not hopelessly chaotic geo- graphically (see, for example, Thomas, 1928), but that consistent patterns are recognizable. The preceding discussion, however, neither sug- gests the number of species that are likely present in each of the groups defined, nor comments on the phyletic relationships among them. I would like here to summarize my opinions as to the likely number and distribution of the species contained within each group. It remains for future work, both in the field and in the museum, to verify the ac- curacy of these hypotheses. The decumanus-, canicollis-, and simonsi-&o\xp& are considered monotypic; certainly the restricted ranges and uniform character distributions of both Proechimys decumanus and P. canicollis sup- port this view. Despite a much broader geographic range, character variation among populations as- signed to the simonsi-group is either negligible or clinal in nature. Indeed, this is perhaps the most consistently recognizable group of spiny rats be- cause of this character uniformity (see also Gard- ner & Emmons, 1984). Even the karyotype is in- variant throughout the species range, based on samples available from southern Colombia (Reig PATTON: SPECIES GROUPS OF PROECHIMYS 337 & Useche, 1 976) to southern Peru (Patton & Gard- ner, 1972; Gardner & Emmons, 1984). By se- niority, the single species in the simonsi-group should be recognized as P. simonsi Thomas, with the names hendeei Thomas and nigrofulvus Os- good considered synonyms. The guyannensis-group members are close to P. simonsi in most salient features described here. Their bacula are nearly indistinguishable, as are characters of the incisive foramina and mesoptery- goid fossa. Differences exist in counterfold pattern, but these could represent nothing more than geo- graphic variation. Only the analysis of samples from the hiatus in western Brazil between the known distributions of these taxa (see fig. 1 ) will permit such a determination. Certainly, of the groups defined herein, the simonsi- and guyan- nensis-gToups are more similar in examined char- acters than any of the others. Future work may indicate that these groups should be condensed into one. More than one species is likely present, how- ever, in my guyannensis-group. Karyotypic vari- ation is large, as diploid number ranges from 44 to 30 among the limited geographic samples (see Gardner & Emmons, 1984). Moreover, there are some seemingly striking geographic differences in some of the characters examined, although not in all. For example, specimens from eastern Para, Goias, and Minas Gerais states of Brazil (referred to oris Thomas and roberti Thomas) have broader and shallower mesopterygoid fossae and a less de- veloped infraorbital canal than do those from else- where in the group's range. Clearly, a more refined and critical examination of detailed geographic variation in these characters is needed; the view provided here is simply too general to judge ad- equately the significance of this variation. The goeldii-gvoup varies more over its geo- graphic range than any other; nevertheless, much of the variation in counterfold pattern, for ex- ample, appears clinal, and abrupt character shifts which might signal species-level demarcations are not readily apparent in the characters I examined. Known karyotypic variation also apF>ears limited, with samples examined from southern Venezuela (2n = 26, FN = 42; Reig & Useche, 1976), Ecuador and northern Peru (2n = 28, FN = 42-44; Gardner & Emmons, 1984), and central and southern Peru (2n = 24, FN = 42; Patton & Gardner, 1972; Gardner & Emmons, 1984). Specimens from the western Amazon Basin and from the Casiquiare region of southern Venezuela appear fairly ho- mogeneous in counterfold and pelage color char- acters. There does, however, appear to be rela- tively sharp transition of some characters, particularly those of the pelage (not examined in this report), in the central Amazon Basin between the lower Rio Negro and the Rio Tapajos. It is probable that at least two species are present in this group, a western one to which the name steerei Goldman would apply, and an eastern one to which the senior name goeldii Thomas applies. The character summaries provided in this paper tend to minimize the difficulties that I had in as- signing individual specimens to species groups, and thus may provide a sense of false security. This is particularly true for specimens of the goel- dii and guyannensis groups from the eastern parts of both ranges, primarily in Para slate, Brazil. There are more individual question marks regard- ing group assignments for specimens from this re- gion than for any other area or group, and much more detailed effort is necessary to corvfirm the character differences described herein. Within the longicaudatus-group there appears to be at least two species, Proechimys longicau- datus Rengger from eastern Bolivia east through adjacent Brazil into northern Paraguay; and P. brevicauda Gunther, which occupies the remain- der of the group range as depicted in Figure 3. An area of rather sharp character transition, particu- larly in pelage color and color pattern but also in bacular measurements, for example, occurs in the upper Rio Itenez and Rio Mamore of southern El Beni and Santa Cruz in Bolivia. The limited sam- ples off*, longicaudatus examined show little vari- ation throughout its range. Samples of P. brevi- cauda from southern Colombia to northern Bolivia are, however, quite variable, and more than one species may be represented here. Gardner and Em- mons (1984) suggested that the Ecuadoran pop- ulations referred to gularis are specifically distinct from northern Peruvian brevicauda based on karyotypic differences (2n = 30, FN = 48 without large subtelocentric autosomes, versus 2n = 30, FN = 48 with two pairs of large subtelocentrics). They also suggested that the central and southern Peruvian populations might represent a valid sub- sjjecies of P. brevicauda, to which the name elas- sopus would apply, based both on karyotypic (2n = 28, FN = 50) and color pattern differences. A thor- ough analysis of geographic variation within this group is certainly warranted. No more than two species appear to be repre- sented in the semispinosus-gvonp; P. semispinosus (Tomes) is distributed from Nicaragua south along the Pacific lowlands to southern Ecuador, and P. 338 FIELDIANA: ZOOLOGY oconnelli is restricted to the western llanos in the vicinity of Villavicencio, Colombia. Karyotypic variation occurs in P. semispinosus, but such is minimal (2n = 30, FN = 50-54); P. oconnelli dif- fers only by a single fusion/fission (Gardner & Em- mons, 1 984). A more detailed examination of geo- graphic variation in P. semispinosus is needed before the intraspecific status of the large number of named forms referred to this group can be prop- erly evaluated. As mentioned previously, the cwv/er/-group is clearly divisible into two biological units. The widespread species P. cuvieri Petter is uniform in its characters, including karyotype, from the Guianas to northern Peru. The status of the 2n = 40 karyotypic form from Balta in eastern Peru, however, is an enigma at present. It is clearly spe- cifically distinct from cuvieri and is only placed in this group because of similar bacular design. I do not believe that it is a relative of guyannensis- group taxa, as suggested by both Patton and Gard- ner (1972) and Gardner and Emmons (1984). I have also not been able to identify this form with certainty anywhere except at Balta. Finally, I have treated the various taxa assigned to my trinitatus-group as though they were species, primarily because I have examined relatively few specimens from only scattered localities. Never- theless, it is likely that a number of species exist in this group. Karyotypic variation is extensive (see reviews by Reig et al., 1980; Reig, 1981), and both karyotypic and electromorphic data (Benado et al., 1979) differentiate a guairae superspecies (including guairae, poliopus, ochraceous, and min- cae [see Gardner & Emmons, 1984]) and a trini- tatus superspecies (composed of trinitatus and uri- chi). Proechimys hoplomyoides is clearly a species distinct from the above, as indicated by Gardner and Emmons ( 1 984). These latter authors included chrysaeolus in their semispinosus-%ro\xx) and mag- dalenae in their brevicauda-group, positions which are not supported by the bacular characters cov- ered here. If their true relationships do lie with the trini tat US-group, they too are probably separate species. Certainly, chrysaeolus is the most distinc- tive member of my trinitatus-group in incisive foraminal and counterfold characters. Acknowledgments Alfred Gardner introduced me to Proechimys in the field 19 years ago. My continued interest in these animals results solely from his own infec- tious curiosity and our long-lasting friendship; I value both immeasurably. I am grateful to J. E. Hill and P. D. Jenkins of the British Museum (Nat- ural History); G. G. Musser and S. Anderson of the American Museum of Natural History; A. L. Gardner and C. O. Handley, Jr. of the National Museum of Natural History; P. W. Freeman, R. M. Timm, and B. D. Patterson of Field Museum of Natural History; M. S. Hafner and J. P. O'Neill of the Museum of Zoology, Louisiana State Uni- versity; and P. Myers of the Museum of Zoology, University of Michigan for the opportunity to ex- amine materials in their respective collections. Special appreciation is extended to A. L. Gardner and L. H. Emmons for their constant willingness to share information and ideas on Proechimys and thus to work toward a common understanding. In the same vein, I also thank O. A. Reig for coop- erative interactions over the past decade. Aid in the field has been generously provided by C. P. Patton, A. L. Gardner, J. E. Cadle, M. A. Barros, M. D. Robinson, J. P. O'Neill, P. Myers, and O. B. Berlin. This research has been supported by the National Science Foundation (BNS 76- 1 7485), the National Geographic Society, and the Museum of Vertebrate Zoology. Fieldwork has been facilitated by the Direccion General Forestal y de Fauna, Ministerio de Agricultura, Lima, Peru. Literature Cited Benado, M., M. Aguilera, O. A. Reig, and F. J. Ayala. 1 979. Biochemical genetics of chromosome forms of Venezuelan spiny rats of the Proechimys guairae and Proechimys trinitatus superspecies. Genetica, 50: 89- 97. DiDiER, R. 1 962. Note sur I'os penien de quelques ron- geurs de I'Amerique du Sud. Mammalia, 26(3): 408- 430. Gardner, A. L. 1983. Proechimys semispinosus (Ro- dentia: Echimyidae): Distribution, type locality, and taxonomic history. Proceedings of the Biological So- ciety of Washington, 96: 134-144. Gardner, A. L., and L. H. Emmons. 1984. Species groups in Proechimys (Rodenlia, Echimyidae) as in- dicated by karyology and bullar morphology. Journal of Mammalogy, 65(1): 10-25. Hershkovitz, P. 1948. Mammalsof northern Colom- bia. Preliminary report no. 2: Spiny rats (Echimyidae), with supplemental notes on related forms. Proceedings of the United States National Museum, 97: 125-140. Hooper, E. T. 1961. The glans penis in Proechimys and other caviomorph rodents. Occasional Papers of PATTON: SPECIES GROUPS OF PROECHIMYS 339 the Museum of Zoology, University of Michigan, no. 623, 18 pp. Martin, R. E. 1970. Cranial and bacular variation in populations of spiny rats of the genus Proechimys (Ro- dentia: Echimyidae) from South America. Smithson- ian Contributions to Zoology, 35: 1-19. MoojEN, J. 1 948. Speciation in the Brazilian spiny rats (genus Proechimys. family Echimyidae). University of Kansas Publications, Museum of Natural History, 1(19): 301-406. Patton, J. L., AND L. H. Emmons. 1985. A review of the genus Isothrix (Rodentia, Echimyidae). American Museum Novitates, 2817: 1-14. Patton, J. L., and A. L. Gardner. 1972. Notes on the systematics of Proechimys (Rodentia: Echimyi- dae), with emphasis on Peruvian forms. Occasional Papers, Museum of Zoology, Louisiana State Univer- sity, 44: 1-30. Patton, J. L., and M. A. Rogers. 1983. Systematic implications of non-geographic variation in the spiny rat genus Proechimys (Echimyidae). Zeitschrifl fiir Saugetierkunde, 48: 363-370. Petter, F. 1978. Epidemiologic de la leishmaniose en Guyane fran^aise, en relation avec I'existence d'une espdce nouvelle de Rongeurs Echimyides, Proechimys cuvieri sp. n. Comptes Rendus Hebdomadaires des Seances de I'Academie des Sciences, series D, Sciences Naturelles, 287: 261-264. Reig, O. a. 1981. Modelos de especiacion cromoso- mica en las casiraguas (genero Proechimys), pp. 149- 190. In Reig, O. A., ed.. Ecology and Genetics of An- imal Speciation. Equinoccio, Caracas, Venezuela, 295 pp. Reig, O. A., M. Aguilera, M. A. Barros, and M. UsECHE. 1980. Chromosomal speciation in a Ras- senkreis of Venezuelan spiny rats (genus Proechimys, Rodentia, Echimyidae), pp. 291-312. In Vorontsov, N. N., and J. M. Van Brink, eds.. Animal Genetics and Evolution. Dr. W. Junk B. V., The Hague, Neth- erlands, 383 pp. Reig, O. A., and M. Useche. 1976. Diversidad cari- otipica y sistematica en poblaciones venezolanas de Proechimys (Rodentia, Echimyidae), con datos adi- cionales sobre poblaciones de Peru y Colombia. Acta Cientifica Venezolana, 27: 132-140. Thomas, O. 1901. On mammals obtained by Mr. Al- phonse Robert on the Rio Jordao, S. W. Minas Geraes. Annals and Magazine of Natural History, ser. 7, 8: 526-536. . 1928. The Godman-Thomas expedition to Peru. VII. The mammals of the Rio Ucayali. Annals and Magazine of Natural History, ser. 10, 2: 249-265. Woods, C. A. 1 984. Histricognath rodents, pp. 389- 446. In Anderson, S., and J. K. Jones, Jr., eds., Orders and Families of Recent Mammals of the World. John Wiley & Sons, New York, 686 pp. Appendix: Specimens Examined Repositories for specimens examined in this study are as follows: American Museum of Natural History (amnh), British Museum (Natural History) (bm[nh]), Field Museum of Natural History (fmnh), Louisiana State University Museum of Zoology (lsu), University of California Museum of Vertebrate Zoology (mvz), University of Michigan Museum of Zoology (ummz), and National Museum of Natural History (usnm). caiiiC0//i5-group COLOMBIA: Atlantico: Cienaga de Guajaro, Sabana Larga ( 1 , usnm). Bolivar: San Juan Nepu- moceno (9, fmnh). Magdalena: Honda (29, amnh; 1, fmnh; 8, usnm; 2, bm[nh]); Honda, Finca Ve- racruz (2, usnm); Mamatoca (11, amnh; 7, fmnh); Minca Road (1, amnh); El Libano plantation (6, amnh); Santa Marta (11, amnh; 1, fmnh); Rio Guaimaral, Valledupar (1 3, usnm); Aguas Verdes, Valledupar (26, usnm); Parmarilo, Valledupar (2, usnm); El Orinoco, Rio Cesar, Valledupar (14, usnm); Villanueva, Valledupar (33, usnm). VENEZUELA: Zulia: Perija, Rio CogoUo (7, fmnh); Rio Cachiri (1, mvz). cifvieri-group P. CUVIERI BRAZIL: Para: Ilha do Taiuno, Rio Tocantins (25, amnh); Vila Hela Imperatriz, Serra de Parin- tins, Rio Amazonas (1, amnh). GUYANA: Kamakusa ( 1 , amnh); Kalacoon ( 1 , amnh); Kartabo (8, amnh); Minehaha Creek (1, amnh); Samin Island, Mazarani River (6, amnh); Maracai Creek, Demarara River ( 1 , bm[nh]); De- marara River (2, bm[nh]); Supinaam River (2, bm[nh]). PERU: Amazonas: La Poza, Rio Santiago (1, mvz). Loreto: Pebas, Rio Amazonas (5, bm[nh]); 340 FIELDIANA: ZOOLOGY Orosa, Rio Amazonas (14, amnh); Pto. Indiana, Rio Amazonas ( 1 , amnh); Santa Luisa, Rio Nanay (3, fmnh); Sarayacu, Rio Ucayali (3, amnh). SURINAME: Carolinakreek (5, fmnh); Wil- helmina Mts., West River (1, fmnh); Finisanti, Saramacca River (1, fmnh); Lelydorpplan (2, fmnh); La Poule (2, fmnh); Dirkshoop (3, fmnh). 2n = 40 PERU: Ucayali: Balta, Rio Curanja (5, lsu). decumanus-group ECUADOR: El Oro: Santa Rosa (2, amnh). Guayas: Chongon (4, bm[nh]): Chongoncito (10, amnh); Cerro Manglar Alto (3, amnh); Cerro Baja Verde (1, amnh); Los Pozos (25, amnh). Los Rios: Vinces, Hda. Pijigual (2, amnh). Manabi: Bahia de Caraques, Rio Briseno (7, amnh). PERU: Piura: Quebrada Bandarrango ( 1 , fmnh); Laguna Lamadero (2, fmnh). Tumbes: Matapalo (6, fmnh). goeldii-group BOLIVIA: El Beni: Riberalta, Vaca Diez (13, usnm); 13 km W Riberalta (11, usnm); 3.5 km NW Riberalta (2, usnm); 5 km NW Riberalta (3, usnm); Rio Mamore, 4 km below Santa Cruz (2, amnh); 6 km S Buena Hora (1, amnh); Rio Ma- more, 7 km N Lagionha (2, amnh); Rio Mamore, 5 km S Guayaramarin ( 1 , amnh); Rio Mamore, 5 km S Guayaramarin (1, amnh); Rio Mamore (2, amnh); Rio Mamore, opposite Cascajal ( 1 , amnh); Rio Mamore, 17 km NNW Nuevo Berlin (1, amnh). BRAZIL: Acre: Sena Madureira, Mandel Ur- bano ( 1 , usnm); Rio Branco, 3-4 km S Rio Branco (2, usnm). Amazonas: Faro, Rio Yumunda [= Nhamunda] (2, bm[nh]); Faro, Paraiso (2, amnh); Acajutuba, Rio Negro (2, bm[nh]); Mirapinima, Rio Negro (5, amnh); Cacauo Pereira Igarape, Rio Negro (9, amnh); Yucabi, Rio Negro (1, amnh); Tatu, Rio Uaupes (3, amnh); Itamarati, Rio Uaupes (1, amnh); Tahuapunta, Rio Uaupes (2, amnh); Manacapuru, Rio Solimoes (2, bm[nh]); Maturaca Mission, northern Amazonia (1, usnm); Humaita, km 886-990, Br 230 (3, usnm); Sao Antonio de Amatari (1, amnh); Borba, Rio Madeira (2, amnh); Auara Igarape, Rio Madeira (15, amnh); Sao An- tonio de Uayara (9, amnh); Rosarinho, Rio Ma- deira (11, amnh); Ipixuna, Rio Purus (1, usnm); Rio Purus, Hyutanahan (4, usnm). Mato Grosso: Serra da Chapada (4, bm[nh]); Utiariti, Rio Pa- pagaio (1, amnh). Para: Cameta, Rio Tocantins (2, bm[nh]); Manapiri Island, Rio Tocantins (2, bm[nh]); Ilha do Taiuno, Rio Tocantins (57, amnh); Mocajuba, Rio Tocantins (1, amnh); Baiao, Rio Tocantins (3, amnh); Urucum de Corumba (1, fmnh); Tuary, Rio Tapajos (1, fmnh; 1 1, amnh); Aramanay, Rio Tapajos (6, amnh); Piquiatuba, Rio Tapajos (7, amnh); Igarape Amorim, Rio Ta- pajos (30, amnh); Igarape Brabo, Rio Tapajos (2, amnh); Farinicatuba, Rio Tapajos (1, amnh); Aquiatuba, Rio Tapajos (1, amnh); Limoal, Rio Tapajos (1, amnh); Inajatuba, Rio Tapajos (1, amnh); Fordlandia, Rio Tapajos (5, amnh); San- tarem (1, bm[nh]); km 84, Santarem-Cuiaba hwy (27, usnm); km 212, Santarem-Cuiaba hwy (4, usnm); km 216, Santarem-Cuiaba hwy (3, usnm); Itaituba, Rio Tapacurazinho (9, usnm); Mojui Dos Campos (1 5, usnm); km 19, Itaituba-Jacareacanga hwy (12, usnm); km 25, Itaituba-Altamira hwy (4, usnm); Itaituba (6, usnm); 54 km S, 150 km W Altamira (3, usnm); Agrovila, km 43 Itaituba-Al- tamira rd (2, usnm); Vila Bela Imperatriz, south bank Rio Amazonas (3, amnh); Serra de Parintins, Vila Bela Imperatriz, south bank Rio Amazonas (7, amnh); Porto de Moz, Rio Xingu (2, amnh); Vilarinho do Monte, Rio Xingu (4, amnh). Ron- donia: Pista Nova, 8 km N Porto Velho (4, usnm). COLOMBIA: Caqueta: La Tagua, Tres Tron- cos, Rio Caqueta (13, fmnh); Rio Mecaya (1, fmnh); Florencia, Mantanito (3, fmnh); La Mu- relia, Rio Bodoquera (2, fmnh). ECUADOR: Napo: San Francisco, Rio Napo (8, ummz); Llunchi, Rio Napo (5, ummz). PERU: Amazonas: La Poza, Rio Santiago (128, Mvz). Huanuco: San Antonio, Rio Pachitea (1, bm[nh]); Port Leguia, Rio Pachitea (1, bm[nh]). Loreto: Yurimaguas ( 1 , fmnh); Orosa, Rio Ama- zonas (14, amnh); Boca Rio Curaray (2, amnh); Pto. Indiana, Rio Amazonas (27, amnh); Rio Ma- zan (2, amnh); Rio Panduro (6, amnh); Pampa Chica, Iquitos (1, amnh); Santa Rita, Iquitos (5, fmnh); Santa Luisa, Rio Nanay (9, fmnh); Santa Elena, Rio Samiria (11, fmnh); Rio Samiria (10, fmnh); San Lorenzo, Rio Maraiion (1, fmnh; 4, bm[nh]); Boca Rio Peruate, Rio Amazonas (2, fmnh); Lagunas (10, fmnh; 1, bm[nh]); Quista- cocha, Maynas (4, fmnh); Rio Tigre, 1 km above Rio Tigrillo (7, fmnh); Pebas, Rio Amazonas (1, PATTON: SPECIES GROUPS OF PROECHIMYS 341 bm[nh]); Sarayacu, Rio Ucayali (16, amnh); San Jeronimo, Rio Ucayali ( 1 , bm[nh]); Cantamana (3, bm[nh]); Lago Mirano, Rio Napo (6, bm[nh]). Madre de Dios: La Pastora, Maldonado ( 1 , fmnh); Tambopata, Puerto Maldonado (3, usnm); Ita- huania ( 1 , fmnh). Ucayali: Yarinacocha (6, fmnh; 9, Lsu); Chicosa, upper Rio Ucayali (6, bm[nh]); 59 km W Pucallpa ( 1 , usnm); Pucallpa (2, amnh); Santa Rosa, Rio Ucayali (12, amnh); Fernando Stahl Mission (14, amnh); Cumaria (5, bm[nh]); Tushemo, Masisea, Rio Ucayali ( 1 , bm[nh]); Balta, Rio Curanja ( 1 , lsu). VENEZUELA: Amazonas: 68 km SE Esmeral- da (9, usnm); Rio Orinoco, Tamatama ( 1 3, usnm); Casiquiare Canal, Capibara ( 1 4, usnm); 30 km SSE Puerto Ayacucho, Coromoto (8, usnm); 1 8 km SSE Puerto Ayacucho (2, usnm); Mt. Duida, Rio Ca- siquiare, Quemapure (2, amnh); Mt. Duida, 8 mi from Rio Orinoco (2, amnh); Mt. Duida, Esme- ralda (11, amnh); Mt. Duida, Caiio Seco ( 1 , amnh); Mt. Duida, El Merey (2, amnh); Rio Orinoco, Par- ipari ( 1 , amnh); Rio Orinoco, Boca del Rio Ocamo (3, amnh). guyannensis-group BRAZIL: Amapa: Serra do Navio (3, usnm); Calicoene (1, usnm); Capoeira (1, usnm). Ama- zonas: Hd. Rio Tucaro ( 1 , usnm); Serra de Neblina (1, usnm); Rio Uaupes, Tauarate (1, amnh); Rio Uaupes, Tahuapunta ( 1 0, amnh); Rio Negro, Tatu ( 1 , amnh); Rio Negro, Uacara ( 1 , amnh); Rio Ne- gro, Camanaos (1, amnh); Rio Negro, Pira-pocu (1, amnh); Rio Negro, Manaus (1, amnh); Faro, north bank Rio Amazonas (10, amnh); Rio Par- atucu (2, amnh); Rio Nhamunda, Castanhal (20, amnh); Rio Nhamunda, Sao Jose (3, amnh). Goias: Fazenda Cangalha ( 1 , usnm); Anapolis (79, amnh). Minas Gerais: Rio Jordao, Araguari ( 1 , fmnh; 10, bm[nh]). Para: Providencia (1, fmnh); Cameta, Rio Tocantins ( 1 , fmnh; 1 , amnh); Ilha do Taiuno, Rio Tocantins ( 1 , amnh); Baiao, Rio Tocantins ( 1 , amnh); Manapiri Island, Rio Tocantins ( 1 , amnh); Maranhao, Alto Pamaiba ( 1 , fmnh); km 84, San- tarem-Cuiaba hwy (53, usnm); km 2 1 2, Santarem- Cuiaba hwy (1, usnm); km 217, Santarem-Cuiaba hwy (4, usnm); km 19, Itaituba-Jacareacanga hwy (2, usnm); Rio Tapacurazinho (10, usnm); Agrov- ila, Altamira (6, usnm); km 43, Itaituba-Altamira hwy (5, usnm); Maraba, Serra Norte (7, usnm); Jatobal (11, usnm); Itupiranga (1, usnm); Belem (46, usnm; 1, amnh); Igarape-A9u (13, amnh); Tury-Agu, Maranhao (1, bm[nh]); Abaete (8, bm[nh]); Patagonia (1 5, amnh); Capim (10, amnh); Igarape Amorim, Rio Tapajos (17, amnh); Ina- jatuba, Rio Tapajos ( 1 , amnh); Igarape Brabo, Rio Tapajos (4, amnh); Limoal, Rio Tapajos ( 1 , amnh); Vila Bela Imperatriz, south bank Rio Amazonas, Lago Andina ( 1 , amnh); Vila Bela Imperatriz, south bank Rio Amazonas, Boca Rio Andina (3, amnh); Vila Bela Imperatriz, south bank Rio Amazonas, Serra de Parintins (3, amnh). Roraima: Uaico, Rio Uraricoera (3, amnh); Rio Cotingo, Limao (64, amnh). SURINAME: Brokopando: Carolinakreek (4, fmnh); Lawa Mission, Lawa River ( 1 , amnh); Lok- sie Hattie, Saramacca River (4, fmnh); Finisanti, Saramacca River (8, fmnh). VENEZUELA: Amazonas: 68 km SE Esmeral- da, Mavaca (2, usnm); 68 km SE Esmeralda, Boca Masiaca (1, usnm); Rio Canucunuma, Belen (26, usnm); Casiquiare Canal, Capibara (3, usnm); San Juan, Rio Manapiari (17, usnm); Rio Orinoco (2, usnm); Rio Orinoco, Boca del Rio Ocamo (5, amnh); Rio Casiquiare, El Merey (6, amnh); Rio Casiquiare, Buena Vista (4, amnh); Rio Casi- quiare, Solano (2, amnh); Rio Casiquiare ( 1 , amnh); Rio Casiquiare, 2 mi W Tamasu (1, amnh); Mt. Duida, foothills camp (1 , amnh); Mt. Duida, mid- dle camp (7, amnh); Mt. Duida, Valle de los Mon- os (2, amnh); Mt. Duida, Playa del Rio Base (6, amnh); Mt. Duida, Cario Seco ( 1 , amnh); Mt. Dui- da, Pie del Cerro (1, amnh); Mt. Duida, La Lajo, Rio Orinoco (1, amnh). longicaudatus-g^oup BOLIVIA: El Beni: 6 km S Buena Hora (1, amnh); Rio Machupo, 1 5 km above Horquilla ( 1 , amnh); San Ignacio (47, usnm); 3.6 km NNE San Ignacio (9, usnm); Riberalta (2, fmnh); Riberalta, Vaca Diez (4, usnm); Fortaleza (2, usnm); San Marco, 3.2 km SW San Joaquin (3, usnm; 22, fmnh); San Joaquin (7, usnm; 90, fmnh; 4, amnh); 20 km S San Joaquin, Est. Yutiole ( 1 , amnh); Caf- etal, 20 km SE San Ramon (4, usnm); Rio Itenez, opposite Principe da Beira (3, amnh); Rio Ma- more, 5 km NE Rio Grande mouth (5, amnh); Rio Mamore, 1 mi NW Guayaramarin (5, amnh); Guayaramarin (1, amnh); Rio Mamore, 5 km S Guayaramarin (2, amnh); Rio Mamore ( 1 , amnh); Rio Mamore, opposite Cascajal (2, amnh); Rio Mamore, 2 km SE Puerto Siles (5, amnh); 10 km E San Antonio ( 1 , amnh); Rurrenabaque ( 1 , amnh); 342 HELDIANA: ZOOLOGY Lago Victoria ( 1 , usnm; 1 4, fmnh); La Esperanza, 42 km NE San Joaquin (1, usnm); Est. Barran- quita, 20 km S San Joaquin (3, fmnh); El Carmen (20, fmnh); Azunta (27, fmnh); Santo Dios (10, fmnh); San Pedro (2, fmnh); Caravana ( 1 7, fmnh); San Pablo (8, fmnh); Filadelfia (6, fmnh); Aca- pulco (1, fmnh); Buena Vista (1, fmnh); Arruda (2, fmnh); Centenela ( 1 , fmnh); Cinco (4, fmnh); Las Pavas (6, fmnh); Providencia ( 1 , fmnh); Puer- to Siles ( 1 , fmnh); San Andres (2, fmnh); San Juan (3, fmnh); Tapera Jorillo (3, fmnh); Veintedos (1, fmnh); Huchulu ( 1 , usnm); Las Penas ( 1 , usnm); Pampitas (1, usnm). Cochabamba: El Mojon (4, FMNH); San Rafael, 19 km SW Villa Tunari (2, usnm); 4 km SE Villa Tunari (1, usnm); 2 km E Villa Tunari (2, amnh); Todos Santos (3, fmnh; 13, amnh); Mission San Antonio, Rio Chimore (8, amnh); El Palmar (5, fmnh); Charuplaya, upper Rio Secure (5, bm[nh]). La Paz: 5 km SE Guanay, Rio Challana (2, ummz); Caranavi (4, ummz); Ma- piri (4, amnh; 3, bm[nh]); Ticunhuaya (5, amnh); San Ernesto (2, bm[nh]). Pando: Rio Nareuda (2, amnh). Santa Cruz: Buenavista (4, fmnh; 10, bm[nh]); Ascencion de Guarayos (12, fmnh); Rio Surutu (1, bm[nh]); Rio Ichilo, 54 km S Boca Rio Chapare (12, amnh); Rio Ichilo, 52 km S Boca Rio Chapare (2, amnh); Rio Ichilo, 34 km S Boca Rio Chapare (4, amnh); Rio Ichilo, 30 km S Boca Rio Chapare (1, amnh); Rio Mamore, 2 km from Boca Rio Chapare (2, amnh); Wames (2, usnm); 1 .3 km NE Wames (8, usnm); 1 km NNW Wames (9, usnm); 3 km SW Wames, Santa Rosita (6, usnm); Florida, near Floripondio (2, fmnh); Cerro Hosana (1, fmnh). BRAZIL: Acre: Sena Madureira, Mandel Ur- bano ( 1 , amnh); Rio Branco, 3-4 km S Rio Branco (2, usnm). Mato Grosso: Tapirapua, Rio Siputuba (2, amnh); Umcum (2, amnh); Serra da Chapada (6, bm[nh]); Fazenda Acurizal (2, usnm); Ari- puana, Humboldt-Aripauna (8, usnm); Corumba (7, usnm); 7 km SE Commba (1, usnm); 22 km S Corumba (2, usnm); Sta. Theresa, 7 km WSW Umcum (7, usnm); Cuiaba, 10 km N Cuiaba (1, usnm); Limao, 48 km W Caceres, Rio Jauru (15, usnm). Rondonia: Pista Nova, 8 km N Porto Velho (6, usnm); Porto Velho ( 1 , fmnh). COLOMBIA: Caqueta: Rio Mecaya (2, fmnh); Florencia, Mantanito (17, fmnh); Florencia (17, amnh); La Murelia, Rio Bodoquera (15, amnh). ECUADOR: Napo: San Francisco, Rio Nape (17, ummz); Intillama, Rio Napo (2, ummz); Llun- chi, Rio Napo (2, ummz); near Rio Napo, Oriente (4, bm[nh]); San Jose Abajo (6, amnh); Rio Suno Abajo (2, amnh). Pastaza: Canelos, Rio Bobonaza (2, bm[nh]); Rio Pastaza (2, bm[nh]); Rio Pindo Yacu (2, fmnh); Rio Yana Rumi (1, fmnh); Rio Capihuara (3, fmnh); Rio Copataza (3, fmnh); Rio Lipuno (1, amnh); Sarayacu (4, amnh); Canelos (1, amnh). PERU: Amazonas: Huampami, Rio Cenepa (173, mvz); La Poza, Rio Santiago (38, mvz). Huanuco: Port Leguia, Rio Pachitea (3, bm[nh]); San Antonio, Rio Pachitea (1, bm[nh]); 35 km NE Tingo Maria, Sta. Elena (2, lsu); Tingo Maria (5, bm[nh]; 1 , lsu; 9, fmnh); Chinchavita ( 1 0, bm[nh]). Loreto: Yurimaguas (2, usnm; 2, bm[nh]; 25, fmnh); Pebas, Rio Amazonas (10, bm[nh]); Boca Rio Cur- aray (32, amnh); Iquitos, Rio Amazonas (1, bm[nh]); Santa Luisa, Rio Nanay (1, fmnh); San Fernando, Rio Yavari (1, fmnh); Cantamana (3, bm[nh]). Madre de Dios: Tambopata, Puerto Mal- donado (23, usnm); Lago Sandoval, Rio Madre de Dios (3, mvz); La Pastora, Puerto Maldonado (5, fmnh); Albergue, Rio Madre de Dios (8, mvz). Pasco: Nevati Mission (54, amnh); San Pablo (32, amnh). Puno: Santo Domingo [= Inca Mines] (1, bm[nh]; 5, amnh; 6, fmnh). San Martin: Achin- amiza, Rio Huallaga (1, amnh). Ucayali: Yari- nacocha (1, fmnh); Pucallpa (1, fmnh); 59 km W Pucallpa (21, usnm); Fernando Stahl Mission (2, amnh); Balta, Rio Curanja (18, lsu; 10, mvz). semispinosus-group P. semispinosus COLOMBIA: Antioquia: Uraba, Villa Arteaga, Rio Cumlao (23, fmnh). Cauca: Rio Saija (16, fmnh); El Papayo, Rio Saija (3, fmnh); La Boca, Rio Saija (3, fmnh); San Jose ( 1 , fmnh; 1 1 , amnh); Novita trail, western Andes (1, amnh). Choco: Condoto ( 1 , bm[nh]); Rio Docampado ( 1 2, fmnh); Rio Saudo ( 1 6, fmnh); Unguia (24, fmnh); Bagado (4, amnh); Andaqueda (1, amnh). Cordoba: So- corre, upper Rio Sinu (4, fmnh). Nariflo: La Guay- acana (15, fmnh); La Candelilla (4, fmnh); Isla Gorgona (5, bm[nh]; 2, fmnh); Barbacoas (8, amnh). Valle de Cauca: Sabaleta 2, fmnh). COSTA RICA: Alajuela: San Carios (2, fmnh). Limon: Cariari (5, lsu); Finca La Lola (1 , lsu); 4.6 km W Limon (2, mvz). Puntarenas: San Geronimo (2, fmnh; 2, amnh); Rincon de Osa (1, lsu); Pal- mar Sur (2, lsu); Palmar (28, amnh). San Jose: 16.3 km SE San Isidro (2, mvz); 34.7 km SE San Isidro (2, mvz); 1 .6 km W Villa Colon (2, mvz); PATTON: SPECIES GROUPS OF PROECHIMYS 343 2.8 km W Villa Colon (1, Mvz); 14.5 km N Quepos, Rio Damitas ( 1 , lsu); Caspirola ( 1 , lsu). ECUADOR: Esmeraldas: San Javier (6, bm[nh]; 1, fmnh); Esmeraldas (3, amnh). Manabi: Rio Mongaya (2, fmnh). Pichincha: Santo Domingo (9, bm[nh]). El Oro: Santa Rosa (3, bm[nh]; 4, amnh); Pasaje (4, amnh). Los Rios: Bucay (1, amnh); Puente de Chimbo, Bucay (2, amnh); Ca- gue. El Destino (6, amnh); Limon, Balsapampa to Babahoyo road (7, amnh); Ventura (1, amnh). NICARAGUA: Rivas: Rio Grande (11, amnh). Zelaya: Toro Rapids (2, amnh); Bluefields (5, mvz). PANAMA: Canal Zone: Barro Colorado Island (3, amnh); Gatun (20, amnh); Maxim Ranch (3, amnh); Buena Vista Peninsula (6, lsu); Rio Chagres (2, amnh); Balboa ( 1 , amnh). Chiriqui: Boqueron (1, bm[nh]; 12, ftvinh; 56, amnh); Bugaba (6, bm[nh]). Darien: Cituro (4, amnh); El Real (8, amnh); Boca de Cupe (4, amnh); Tapaliza (3, amnh); Tacarcuna (2, amnh). Panama: Tocumen (3, bm[nh]); Gobemador Island (7, bm[nh]); Sa- vanna near Panama (3, bm[nh]); 0.8 km N Paraiso (4, mvz); Cebaco Island (5, bm[nh]); San Miguel Island (3, fmnh; 5, amnh). P. OCONNELU COLOMBIA: Meta: Quaicaramo (3, amnh; 1 5, usnm); Mambita (2, usnm); La Aguadita (1 , amnh); Barrigona (2, amnh); Restrepo (12, amnh); Vil- lavicencio (26, amnh; 2, usnm; 10, ummz; 4, mvz); 3 km N Villavicencio (1, usnm); Los Micos, San Juan de Arama (16, fmnh). simonsi-ffoup Rio Santiago (3, mvz); Yambrasbamba (1, fmnh; 5, bm[nh]). Cajamarca: Huarandosa (1, amnh). Cuzco: 40 km E Quincemil above Rio Marcapata (2, lsu); Cosiiipata, Hda. Villa Carmen (4, fmnh); Urubamba (1, bm[nh]). Junin: Rio Perene (1, bm[nh]). Loreto: Yurimaguas (3, fmnh); Santa Luisa, Rio Nanay (5, fmnh); Boca Rio Curaray (22, amnh); Orosa, Rio Amazonas ( 1 , amnh); Cer- ro Azul, Cantamana, Rio Ucayali (4, bm[nh]); Cantamana ( 1 , bm[nh]). Madre de Dios: Itahuania (4, fmnh); Tambopata, Puerto Maldonado (4, usnm); Aguas Calientes, Rio Alto Madre de Dios (10, mvz); Hda. Erika, Rio Alto Madre de Dios (3, mvz); Albergue, Rio Madre de Dios (7, mvz). Pasco: Mairo, Rio Palcazu (2, bm[nh]); San Juan (1, usnm); Bermudas de Loma Linda (13, amnh); San Pablo (3, amnh). San Martin: Puca Tambo, 50 mi E Chachapoyas ( 1 0, bm[nh]). Ucayali: Balta, Rio Curanja (2, mvz; 17, lsu); Yarinacocha (1, fmnh); 59 km W Pucallpa (33, usnm). trinitatus-group P. chrysaeolus COLOMBIA: Antioquia: Puri, above Caseres (8, fmnh); Medellin (1, bm[nh]). Bolivar: San Juan Nepumoceno (26, fmnh); Coloso (20, fmnh); Mar- garita (3, bm[nh]). Boyaca: Muzo (3, fmnh; 2, bm[nh]). Cauca: Rio Chili (2, bm[nh]). Cordoba: Catival, upper Rio San Jorge (4, fmnh); Socorre, upper Rio Sinu (18, fmnh). Tolima: Santana (3, bm[nh]). VENEZUELA: Tachira: San Cristobal (1, bm[nh]). BOLIVIA: El Beni: Rio Mamore (1, amnh). Cochabamba: Yungas ( 1 , amnh). Pando: Rio Na- reuda ( 1 , amnh). COLOMBIA: Caqueta: Rio Mecaya (9, fmnh); La Murelia, Rio Bodoquera (1, amnh). ECUADOR: Napo: Intillana, Rio Napo (5, ummz); near Rio Napo ( 1 , bm[nh]); San Jose Abajo (3, amnh); Rio Suno Abajo (1, amnh). Pastaza: Rio Pindo Yacu (4, fmnh); Rio Bobonaza, Mon- talvo (3, fmnh); Rio Bobonaza (2, bm[nh]); Rio Yana Rumi (1, fmnh); Rio Capihuara (2, fmnh); Rio Pastaza (5, bm[nh]); Rio Tigre (4, bm[nh]). Zamora: Gualaquiza ( 1 , bm[nh]). PERU: Amazonas: Huampami, Rio Cenepa (3, mvz); headwaters Rio Kagka (2, mvz); La Poza, P. GUAIRAE COLOMBIA: Arauca: Rio Cobaria (22, fmnh); Rio Bojaba (5, fmnh); Rio Arauca (18, fmnh); Fatima, Rio Cobaria ( 1 2, fmnh). Boyaca: La Ar- gentina, Rio Cubugon (4, fmnh); El Porvenir, Rio Cubugon (3, fmnh). VENEZUELA: Barinas: Guaquitas (1, mvz). Portuguesa: Sto. Domingo (1, mvz). P. HOPLOMYOIDES VENEZUELA: Bolivar: Mt. Roraima, Arabupu ( 1 , AMNH); Mt. Roraima, Rondon camp ( 1 , amnh). 344 HELDIANA: ZOOLOGY p. MAGDALENAE P. POLIOPUS COLOMBIA: Antioquia: near La Providencia, SW Zaragoza (1, usnm); 25 km S and 22 km W Zaragoza (48, usnm). Bolivar: Norosi, Mompos, Rio San Pedro (18, usnm). VENEZUELA: Tachira: San Juan de Coldn (1, fmnh). Zulia: Kasmera ( 1 , mvz). P. TRINITATUS p. MINCAE COLOMBIA: Magdalena: Minca (2, bm[nh]; 6, usnm; 78, amnh); Bonda (12, amnh; 2, usnm); Onaca (4, amnh); Buritaca ( 1 , amnh); Don Dago (1, amnh); Cuaco (1, amnh); Masinga Vieja (1, amnh); Manzanares (2, usnm); Colonia Agricola de Caracolicito (1, usnm); El Salado (5, usnm). TRINIDAD: Caparo (2, fmnh); Princeslown (2, fmnh; 1 , amnh); Turure Forest (2, amnh); Cumaca (1, amnh); Oropuche Heights (2, fmnh); Chag- uaramas ( 1 , mvz). VENEZUELA: Monagas: 2 km N and 4 km W Caripe (1, usnm). Sucre: 5 km S and 25 km E Carupano (1, usnm). P. ochraceous VENEZUELA: Zulia: El Panorama, Rio Au- rare (2, fmnh). P. URICHI VENEZUELA: Sucre: San Esteban (2, fmnh); Quebrada Seca ( 1 , fmnh; 1 , amnh); Campo Alegre (1, amnh); Los Palmales (1, amnh). PATTON: SPECIES GROUPS OF PROECHIMYS 345 An Assessment of the Systematics and Evolution of the Akodontini, with the Description of New Fossil Species of Akodon (Cricetidae: Sigmodontinae) Osvaldo A. Reig ABSTRACTS The taxonomy and systematics of the tribe Akodontini (Cricetidae: Sigmodontinae) are dis- cussed and revised. Bolomys is distinguished from Akodon, and Cabreramys is considered a synonym of Bolomys; a diagnosis of Bolomys and a Hst of its species are given. Akodon is thought to comprise five subgenera, namely, Akodon, Abrothrix, Chroeomys, Deltamys, and Hypsimys; Thaptomys is included within the subgenus Akodon, and Thalpomys is considered a synonym of Bolomys. Distinctive character states of each subgenus are described, including a detailed description of cranial and dental features of Akodon and Abrothrix. Microxus is distinguished from Abrothrix and is retained as a genus. Blarinomys, Oxymycterus, Lenoxus, Juscelinomys, and Podoxymys are all accorded generic status within the Akodontini. The status of the long-clawed fossorial akodontines of southern South America is discussed; Chelemys, Geoxus, and Notiomys each deserve generic recognition and are diagnosed. New species of Akodon from the Plio-Pleistocene of Buenos Aires Province, Argentina, are described, and their significance for the evolution of the tribe is noted. The Pliocene Akodon {Abrothrix) kermacki Reig, 1978 is described in detail, with the new species A. (Abrothrix) magnus and A. (Akodon) lorenzinii from the Lower Pleistocene, and A. (Ak.) johannis from the Middle Pleistocene. Also described are remains of A. (Ak.) cf cursor and A. (Ak.) cf iniscatus from the Middle and Upper Pleistocene, respectively. A tentative scenario of the origin and evolutionary deployment of the Akodontini is presented, including an origin in the Puna by Middle Miocene times and an indication of probable dispersal corridors. Se discute aqui y se revisa la situacion taxonomica y las relaciones sistematicas de los roedores cricetidos sudamericanos de la tribu Akodontini (Cricetidae: Sigmodontinae). Se distingue a Bolomys de Akodon como un genero pleno, y se considera a Cabreramys sinonimo del primero, del que se proporciona una diagnosis y la lista de sus especies. Se reconocen dentro del genero Akodon cinco subgeneros: Akodon, Abrothrix, Chroeomys, Deltamys e Hypsimys; a Thaptomys se lo incluye dentro del subgenero Akodon, y a Thalpomys se lo considera como un sinonimo de Bolomys. Se describen los estados de caracteres distintivos de cada uno de los subgeneros de Akodon y se proporciona una descripcion detallada de los rasgos craneanos y dentarios de Akodon y Abrothrix. Se plantea la necesidad de distinguir a Microxus de este ultimo y de conferirle rango pleno de genero. Se reconoce tambien rango generico pleno dentro de los Akodontini a Blarinomys, Oxymycterus, Lenoxus, Juscelinomys y Podoxymys. Se discute tam- bien la situacion taxonomica de los akodontinos hipogeicos de uiias largas de la region austral de America del Sur; se reconoce la distincion generica de Chelemys, Geoxus y Notiomys y se proporcionan diagnosis de los tres. Se describen nuevas especies de Akodon del Plio-Pleistoceno From the Departamento de Ciencias Biologicas, F.C.E. y N., Universidad de Buenos Aires, Pabellon 2, Ciudad Universitaria Nunez, (1428) Buenos Aires, Argentina. Dr. Reig is a CONICET Career Investigator. REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 347 de la provincia de Buenos Aires, Argentina, comentandose su significado para la comprension de la evolucion de la Iribu. Se describe en detalle la especie del Plioceno Akodon (Abrothrix) kermacki Reig, 1978, y las nuevas especies A. (Abrothrix) magnus y A. (Akodon) lorenzinii del Pleistocene inferior, y A. (Ak.) johannis del Pleistoceno medio. Se incluye asi mismo la de- scripcion de restos de A. (Ak) cf. cursor y A. (Ak.) cf. iniscatus del Pleistoceno medio y superior, respectivamente. Se esboza tambien un panorama interpretativo del origen y de la historia evolutiva de los Akodontini en su conjunto, en el que se sostiene su origen en la Puna hacia el Mioceno medio, y se indican las posibles rutas de la dispersion de sus distintos taxones. A taxonomia e a sistematica de tribo Akodontini (Cricetidae: Sigmodontinae) e discutida. Justificam-se a separa^ao de Bolomys do genero Akodon e a sinonomia de Cabreramys com Bolomys. Uma diagnose do genero Bolomys e oferecida, com una lista de suas especies. O genero politipico Akodon indue 5 subgeneros: Akodon, Abrothrix, Chroeomys, Deltamys e Hypsimys. Thaptomys nao e reconhecido como genero ou subgenero, e sua especie linica, nigrita, e referida a Akodon s.s. Considera-se Thalpomys como sinonimo menor de Bolomys. Provi- denciam-se os caracteres distintos de cada subgenero, com uma descri^ao detalhada do cranio e das estruturas dentais de Akodon e de Abrothrix. Justifica-se o rango de genero a Microxus e salientam-se as diferen^as entre ele e Akodon (Abrothrix). Blarinomys, Oxymycterus, Lenoxus, Juscelinomys, e Podoxymys sao generos akodontinos que nao requerem agrupamento especial, ou colocacao numa tribo separada. Discutese o rango dos akodontes fossoriais com longas unhas que habitam a regiao sul do continente; Chelemys, Geoxus, e Notiomys sao reconhecidos como generos validos. Novas esF>ecies de Akodon sao descritas do Plio-Pleistoceno da Provincia de Buenos Aires na Argentina, e sua importancia na historia evolutiva e discutida. Akodon (Abro- thrix) kermacki Reig, 1978, do Plioceno, e descrito em detalhe, e A. (Abrothrix) magnus e A. ( Ak.) johannis (do Pleistoceno medio) sao descritos como especies novas. Descreve-se tambem OS restos de A. (Ak.) cf. cursor e A. (Ak.) cf. iniscatus do Pleistoceno medio e superior, respe- tivamente. Um possivel cenario da evolu9ao dos Akodontinos e apresentado. Justifica-se o origem dos Akodontini na Puna, durante o Mioceno medio, e as possiveis rotas de dispersao sao reconstruidas. Introduction In a recent paper (Reig, 1984), I advanced some interesting features of the frequency distribution of diversity among living South American mam- mals. One of 1 2 orders of living terrestrial mam- mals, the Rodentia, comprises 43.2% of the total extant mammalian species of this continent. Most rodent species belong either to the Myomorpha (2 1 .9%) or to the Caviomorpha (19.1 %). However, species of the latter are distributed among 1 1 dif- ferent families, whereas South American living species of myomorphs are all grouped within a single family, Cricetidae. The extant South Amer- ican cricetids include about 250 species and 53 or 54 genera. Except for two species belonging to two genera of North American affinities, all South American cricetids belong to the subfamily Sig- modontinae. This subfamily may itself be subdi- vided into seven well-defined tribes, namely Oryzomyini, Ichthyomyini, Akodontini, Scapter- omyini, Phyllotini, Wiedomyini, and Sigmodon- tini (Reig, 1980, 1981). Reinforcing the observed tendency for a log-normal distribution of species diversities among South American mammals (which seems to be a general pattern of diversity in nature; see Williams, 1964), an overwhelming number of sigmodontine species (44.2%) belong to one tribe, the Oryzomyini. The other two tribes showing high species frequencies are the Akodon- tini (24.9%) and the Phyllotini (1 8. 1%). Thus, three of seven tribes of sigmodontine rodents include 87.1% of the species. The Akodontini comprise around 63 species distributed in 1 1 different genera (see later). They are predominately Andean in distribution, al- though many species are widely distributed in temperate, subtropical, and to a lesser degree, tropical lowlands (Reig, 1984). Despite extensive cytogenetic studies (see particularly Bianchi et al., 1971; Bianchi & Merani, 1984; Rodriguez et al., 1983; Vitullo et al., 1986), the systematics of Ako- dontini is still confusing in several important re- spects, including the taxonomic status and the rank 348 HELDIANA: ZOOLOGY of some supraspecific taxa; the generic allocation of several species; and the status, distribution, and geographical variation of many of their species. Moreover, little is known of their evolutionary history and fossil record, although their present diversity and distribution and a few fossil remains permit a plausible explanation of their age, place of origin and diversification, and phylogenetic re- lationships (Bianchietal., 1971;Reig, 1978, 1980, 1981, 1984). I started to revise the Akodontini at the British Museum about 1 5 years ago. Although the work remains uncompleted because of the pressure of other, more urgent duties, I gathered a consider- able amount of information which is relevant to a better understanding of the systematics of this diverse tribe of sigmodontine rodents. Moreover, I have had the chance to study an important col- lection of fossil akodontine remains from different levels of the Argentinian Plio-Pleistocene strati- graphic column, partly described elsewhere (Reig & Linares, 1969; Reig, 1978). In this paper, I pres- ent and briefly document the more important con- clusions of my revision in progress of the living Akondontini and describe all the known fossil rep- resentatives of the genus Akodon. I also present a tentative explanation of the origin and evolution- ary and biogeographic deployment of these South American mice. The first purpose is treated un- evenly, as some genera (i.e., Bolomys and Akodon) are covered in more detail than others. This is intentional, as clarification of the limits between these two genera is considered essential for an assessment of the whole tribe, and as an elucida- tion of the limits Akodon and its subgenera is a prerequisite to study its fossil representatives. Materials and Methods This systematic survey of the Akodontini was primarily based on direct examination of museum specimens, including in most cases the type spec- imens and original series. The collection of the British Museum (Natural History) in London (BMNH) was the main source of information for this study. However, specimens of other collec- tions have been also extensively studied, partic- ularly those of the American Museum of Natural History (AMNH), the United States National Mu- seum (USNM), the Museum of Comparative Zo- ology, Harvard University (MCZ), and the Mu- seum of Vertebrate Zoology, University of California (MVZ). Moreover, specimens from South American collections provided substantial information, especially those of the Museo Mu- nicipal de Ciencias Naturales de Mar del Plata "Lorenzo Scaglia" (MMP), the Universidad Cen- tral de Venezuela in Caracas (MBUCV), the Uni- versidad Austral de Chile in Valdivia (UACH), and the Facultad de Ciencias Exactas y Naturales of Buenos Aires University (FCM). The fossil specimens belong primarily to the rich collection gathered by Galileo J. Scaglia and collaborators, which is deposited at the Museum of Mar del Pla- ta, but also include material from the Museo de La Plata (MLP). Skull and molar morphology has been the main source of information in the systematic revision of the living forms, and, indeed, the only source available in the study of the fossil forms. Skull measurements were taken from adult sp>ecimens with dial calipers graduated to 0. 1 mm. The mea- surements of the teeth were taken through the re- ticule eyepiece of a Wild M-5 stereomicroscope. Only the greatest lengths and widths of the teeth are given. The names of the enameled components of the crown of the molar teeth follow my proposed unified nomenclature (Reig, 1977). Other aspects of nomenclature of tooth morphology follow Hershkovitz (1962, 1967). The drawings were made by me, with the aid of a drawing tube. Although several conceptual tools of cladistic methodology are used, I follow a syncretic evo- lutionary approach within which paraphyletic taxa are not rejected (Reig et al., in press). Supraspecific taxa are treated as class-concepts and, as in pre- vious papers (Reig, 1970, 1982), I use the logical terms "intension" and "extension" to refer to the set of attributes that determine the taxon-concepts and the set of subordinate taxa that are members of them, respectively. Taxa are also considered to be polythetic. Thus, their intension is defined by reference to a set of character-states which are not exclusive of the taxon, and no claim is made that membership to the taxon requires sharing all char- acters used in defining the taxon's intension (Sokal ifeSneath, 1963). Historical Shaping of the Concept of Akodontini Before the recognition of the Akodontini as a formal taxon of tribal rank of South American cricetid rodents (Vorontzov, 1959; Reig, 1980, 1981), the assemblage of genera and subgenera REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 349 making up this tribe was usually treated informally as the akodont rodents or the akodont group. The belief that those genera and subgenera formed a natural group arose from the studies of Thomas, but the concept of this taxonomic group long re- mained rather vague and controversial, both in intension and in extension. In 1916, Thomas wrote on the taxonomic status of a group of related species, which he allocated in seven different gen- era, namely Akodon (with Chalcomys as a sub- genus), Thalpomys, Thaptomys, Bolomys, Chroeomys. Abrothrix, and Zygodontomys. Sub- sequently (Thomas, 1918), he added the genera Hypsimys and Deltamys to the same group. He recognized that other supraspecific taxa, such as Blarinomys, Geoxus, Notiomys, Chelemys, Mi- croxus, Oxymycterus, and Lenoxus, were also re- lated to the above group. Osgood (1925) was prob- ably the first to revise Thomas's akodont genera, lumping Notiomys, Chelemys, and Geoxus into a single genus, Notiomys, a contention which was criticized by Thomas (1927). The monotypic ge- nus Podoxymys was later added to the group by Anthony (1929). Tate (1932b) presented a preliminary revision of the whole group, which introduced the desig- nation "akodont rodents." As is true of his other revisions of South American rodents, Tate's work is a useful and careful historical and bibliograph- ical review of the taxa involved rather than a re- vision based on actual collections, but it did not result in a clarification of the intension of the group. He recognized generic rank for most of Thomas's supraspecific groups, but followed Osgood regard- ing Notiomys; he also withdrew Zygodontomys from the akodont assemblage, placing it in the oryzomyine group (Tate, 1932a). The next comprehensive study of the whole group was that of Gyldenstolpe (1932), which did not separate the group from other "sigmodont ro- dents," equivalent to my subfamily Sigmodontin- ae. Gyldenstolpe accorded generic rank to all taxa of akodont rodents created by Thomas, and he contributed to their definition. He also agreed with Thomas in recognizing the generic status of A^o- tiomys, Geoxus, and Chelemys. Ellerman (1941) made a thorough reappraisal of Thomas's akodont mice. He claimed that most of the genera created by Thomas in 1916, as well as the later Hypsimys and Deltamys, were best treated as subgenera oi Akodon. He excluded Zy- godontomys from this grouping, regarding it as a distinct genus related to Akodon. However, he re- tained Microxus. Oxymycterus, Lenoxus, and No- tiomys as genera; following Osgood, the latter in- cluded Geoxus and Chelemys. Ellerman's balanced judgment and overall experience influenced the work of subsequent authors, and his concept of the akodont genera was essentially followed by Cabrera (1961). Vorontzov (1959) coined the name Akodontini for the tribe containing Akodon (sensu Ellerman), Zygodontomys, Microxus, Podoxymys, Lenoxus, Oxymycterus, Blarinomys, and Notiomys. Hersh- kovitz (1962) later withdrew Zygodontomys from the akodonts, placing it in his phyllotine group, close to Calomys. Hooper and Musser (1964), in their discussion of the bearing of phallic morphology on the in- terrelationships of cricetids and allied genera, con- cluded that Oxymycterus is distinct enough in the characters of the glans penis to be recognized as a group distinct from, but allied to the akodont group. Moreover, they stated that Notiomys is as distinct from Akodon as is Oxymycterus, that Zygodon- tomys is annectent between akodonts and oryzo- myines (not closer to the phyllotines as claimed by Hershkovitz), and that Calomys, and especially Eligmodontia, are to be placed nesiv Akodon. Many of the conclusions of these authors were criticized by Hershkovitz (1966) on methodological grounds, namely that phallic evidence coming from few taxa, each represented by a few individuals, should not be the basis for introducing major changes in a taxonomic arrangement based on a complex of character states from different organ systems and a large number of genera. Hershkovitz, however, followed Hooper and Musser in splitting an oxy- mycterine group (including Oxymycterus, Podox- ymys, Lenoxus, and Abrothrix; he considered Mi- croxus to be a synonym of Abrothrix) from the akodont group. A further basis for splitting Oxy- mycterus from the akodonts is suggested by spe- cializations of its stomach and intestines for an insectivorous diet, as illustrated in the detailed studies of Vorontzov ( 1 967; see also Tullberg, 1 899; Echave Llanos &. Vilches, 1964; Carleton, 1973). However, in view of the incomplete and scattered information on the anatomy of the digestive sys- tem in cricetids, it does not seem wise to put much weight on this sort of evidence. The complex stom- ach of Oxymycterus may well represent an exclu- sive autopomorphy. To complete this picture of the Akodontini, Massoia and Fomes ( 1 967) proposed a new genus, Cabreramys, for a group of species formerly re- ferred to Akodon and Zygodontomys. They allege it to be transitional between akodontine and phyl- 3S0 FIELDIANA: ZOOLOGY lotine cricetids. Additionally, Moojen (1965) de- scribed a new genus, Juscelinomys, which is clearly related to Oxymycterus, and Reig (1978) created the fossil Dankomys, which is related to Bolomys. The Akodontini remain one of the most obscure groups of South American rodents. There is no agreement as to its extension (viz., Zygodontomys), to the rank of its supraspecific groupings, or even to its unity as a taxon. This situation is obviously due to the lack of comprehensive studies and to the failure of authors to draw conclusions from the partial evidence. Although we are still far from the goal of achieving an accurate knowledge of this group, I hope that recent advances in chromo- somal systematics and certain conclusions about several critical genera and species can serve to improve our understanding of this complex array of rodents. Various chromosome studies on several ako- dontine taxa have been undertaken by various au- thors in the last 15 years (Barquez et al., 1980; Barros & Reig, 1979; Bianchi & Contreras, 1967; Bianchi & Merani, 1984; Bianchi etal., 1969, 1971, 1973, 1979; Gallardo, 1982; Gardner & Patton, 1976; Gentile de Fronza, 1970; Kiblisky et al., 1970, 1976; Lobato et al., 1982; Maia & Lang- guth, 1981; Pearson, 1984; Rodriquez et al., 1983; Sbalqueiro et al., 1984; Spotomo & Fernandez, 1976; Vitullo etal., 1986; Yonenaga, 1972, 1975, 1979; Yonenaga et al., 1975). These studies dem- onstrate that species of Akodon, Abrothrix, Bolo- mys, Microxus, Chroeomys, Thaptomys, and Oxy- mycterus resemble one another in their karyotype, but that Zygodontomys is distinct from this group. The first group is characterized by karyotypes of no more than 54, mostly telocentric chromo- somes, with several species showing diploid num- ber reduction by Robertsonian fusions and tan- dem translocations. In contrast, Zygodontomys is exceptional among mammals, but resembles the Oryzomyini, in having karyotypes of 2n = 84-88, mostly subtelocentric chromosomes. These results strongly suggest the unity of the oxymycterine and akodontine groups. An almost indistinguishable 2n = 52 karyotype occurs in Akodon {Ak.) oliva- ceus, A. (Ak.) xanthorhinus, A. {Ak.) andinus, A. {Abrothrix) sanborni, A. {Ak.) nigrita, A. {Chroeo- mys) jelskii, Chelemys macronyx, and Geoxus val- divianus. The karyotype of Oxymycterus proved to be of 2n = 54 chromosomes, quite similar to these in gross morphology. At the same time, these results distinguish Zygodontomys quite apart from the Akodontini. As suggested by Gardner and Pat- ton (1976), it may represent a separate group or tribe of its own. Pending more complete evidence, I prefer to treat Zygodontomys as incertae sedis within the Sigmodontinae. Once Oxymycterus and allied genera are includ- ed within the Akodontini, and Zygodontomys is excluded from this tribe, it is convenient to discuss the status of the taxa of generic and subgeneric rank which remain as its members. Status of Bolomys and Cabreramys By definition, the most typical akodontine is Akodon. It is also the most polytypic, the most complex, and the most problematic of the Ako- dontini. However, some recent conclusions help to understand its limits. As I advanced and partially substantiated else- where (Reig, 1978), and in agreement with the independent results of Maia and Langguth (198 1), I exclude from Akodon a series of species which have been variously grouped under Akodon prop- er, the "southern group" of species of Zygodon- tomys (Hershkovitz, 1 962), or under the genus Ca- breramys (Massoia & Fomes, 1 967). Those species deserve separate generic status by virtue of clear- cut morphological and karyological differences. Comparisons of character-states of this group with those of Akodon amoenus Thomas, 1 900, which is the type species of Bolomys by original desig- nation, show this species belongs to the same group, and therefore all these species must be grouped under the genus Bolomys Thomas, 1916. How- ever, the matter is still confusing to some and deserves further substantiation. This taxonomic confusion started from the very beginning; when he erected Bolomys as a generic name, Thomas (1916) selected Akodon amoenus as its type species, but included also within the same taxon Akodon albiventer and A. berlepschii, two species which proved to be quite inseparable from the typical Akodon {A. boliviensis, fig. 4A), or other allied species, such as A. andinus (fig. 1 B). Afterward, Thomas (1926b) referred lactens, ne- grito (sic), and orbus, which are probably conspe- cific but different from amoenus, to Bolomys. However, these taxa share peculiar character-states with amoenus that strikingly differentiate the group from any other typical species of Akodon. As clearly observed in the type species A. amoenus (figs. lA, 2C-D), Bolomys shows char- acteristic cranial and dental features which can be REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 351 B Fig. 1 . Skulls of species of Akodon and Bolomys. Left row, lateral view, right row, dorsal view. A, Bolomys amoenus (Thomas); BMNH 1.1.1.12; Sanguero Puno, Peru. B, Akodon (Akodon) andinus (PhiUppi); female; type of Akodon gossei Thomas, BMNH 98.3.21.5; Puente del Inca, Mendoza, Argentina. C, Akodon (Akodon) albiventer Thomas; male; BMNH 21.1 1.1.51; Sierra de Zenta, Jujuy, Argentina. D, Bolomys obscurus (Waterhouse); sex un- known; lectotype, BMNH 55.12.24.161; Maldonado, Uruguay. used in its diagnosis: braincase broad and deep; occipital region short; rostrum rather short and markedly tapering forward in lateral view; upper profile of skull gradually sloping forward from the middle of parietals; nasals short, with anterior bor- ders well posterior to the level of the anterior bor- der of incisors; frontals long, always longer than nasals; parietals short, less than half the length of frontals, and extending forward anterolaterally by means of narrow spines penetrating between fron- tals and temporals; interparietal noticeably re- duced anterop>osteriorly and transversely; occiput short and truncated; interorbital area with well- formed, anteriorly convergent borders; posterior palate moderately long and wide, the median pos- terior border of palatines behind the posterior bor- 3S2 FIELDIANA: ZOOLOGY B Fig. 2. Molar teeth of Bolomys obscurus and Bolomys amoenus. Upper row, left upper molar series; lower row, left lower molar series. A, Bolomys obscurus (Waterhouse); sex unknown; lectotype, BMNH 55.1 2.24. 161; Maldonado, .1 Uruguay. B, Bolomys obscurus (Waterhouse); male; holotype of Akodon bene/actus Thomas, BMNH 16.10.3.35; Laguna Alsina (Bonifacio), Partido de Guamini, Buenos Aires Province, Argentina. C, Bolomys amoenus (Thomas); male; holotype, BMNH 0.10.1.77; Rio Colca, north of Sum bay, Peru (molars with advanced stage of wear). D, Bolomys amoenus (Thomas); male; BMNH 22. 1 . 1 .97; Huarconda, Peru. REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 353 der of M^ zygomatic plate broad and strong, with anterior border straight or sHghtly concave, per- pendicular to diastema; upper incisors orthodont or proodont; molars mesodont, terraced with moderate wear, broad and robust; upper molars with lophs almost completely transverse, and me- soloph usually completely coalesced with para- loph; procingulum of M' simple, with anterome- dian flexus absent or only slightly developed; lower molars with lingual cusps somewhat anterior to the labial ones, with mesolophid remnants and mesostylids usually absent. This combination of cranial and dental char- acter-states is quite distinctive and clearly differ- entiates Bolomys amoenus and B. lactens (fig. 31) from species of Akodon and Zygodontomys (re- stricted to Z. brevicauda and northern relatives) and from any other recognized genus of akodon- tine mice (fig. 3). Meaningfully, that set of char- acter-states is also shared by Akodon obscurus (figs. 1 D, 2 A-B), which is the type species of Cabrera- mys Massoia & Fomes, 1967 by original desig- nation. Akodon bene/actus (fig. 3H) and A. len- guarum (fig. 3J), two other nominal species referred to Cabreramys by Massoia and Fomes ( 1 967), also share that set of character-states. Therefore, there is no doubt that Cabreramys Massoia & Fomes, 1967 must be placed under the synonymy of Bo- lomys Thomas, 1916. Massoia and Fomes rec- ognized the distinctiveness of the species they grouped in Cabreramys but, unfortunately, failed to realize that A. amoenus belonged to the same group. The latter being the type species of Bolo- mys, it was not necessary to propose a new generic name for this group of species. It is also evident (see also Maia & Langguth, 1981) that the same set of character-states is also present in Mus lasiurus Lund, Akodon fuscinus Thomas (fig. 3G), Zygodontomys pixuna Moojen, Zygodontomys tapirapoanus Allen, and inferen- tially, "Hesperomys" brachiurus Wagner. All these nominal taxa, together with A. lenguarum, were placed by Hershkovitz (1962), probably inspired by Tate (1932a), in the "southem group" of Z;^^o- dontomys as subspecies of Z. lasiurus. In addition to the cogent arguments of Maia and Langguth (1981), the fact that lasiurus does not belong to Zygodontomys (fig. 3C-D) has been recently sub- stantiated by Voss and Linzey (1981) on ventral prostate morphology as well as on some dental characters (and the chromosomal evidence dis- cussed below). Similarities of lasiurus and species of Akodon in the presence of smaller medial than lateral prostates led Voss and Linzey to assign that species to Akodon and to withdraw it from Zygo- dontomys. This part of their conclusions agrees with the previous discussion. However, the evi- dence from prostate morphology should not be taken as an indication that lasiurus belongs to Ako- don. The former shares enough derived character states in the skull, dentition, and chromosomes with amoenus, obscurus, lactens, and bene/actus to make convincing its assignment to Bolomys. Derived prostate morphology in lasiurus and species of Akodon found by these authors is better interpreted as a suggestion that Akodon and Bo- lomys share a synapomorphous character-state which distinguishes them from the remaining gen- era of akodontines. , Hesperomys arviculoides Wagner also belongs to Bolomys, and it is a junior synonym of Mus la- siurus Lund. The type specimen of arviculoides, illustrated by Ximenez and Langguth (1970), is undoubtedly a Bolomys in all character-states vis- ible in the illustration, and additionally, it cannot be differentiated from the type of lasiurus illus- trated by Winge (1887) and in the Museum of Copenhagen. Moreover, it cannot be differentiated from a topotype of lasiurus (BMNH 88. 1 .9.4.) (fig. 3F), which I studied in the British Museum. Thus, I cannot agree with Maia and Langguth when they claim, without giving reasons, that arviculoides does not belong to Bolomys. The name arviculoides has been freely applied in recent cytogenetic literature to Brazilian akodontines with 2n = 14, 2n = 16, and 2n = 24 chromosomes (Cestari & Imada, 1 968; Maia & Langguth, 1981; Yonenaga, 1972, 1979; Yonenaga et al., 1975) showing a complex system of intra- and interpopulational chromosomal vari- ation, in fact, it has never been demonstrated that they agree phenotypically with the type of arvi- culoides, which seems inseparable from the type of lasiurus on morphological grounds. As speci- mens of lasiurus matching the character-states of Lund's type specimen have been shown to have a distinctive and different 2n = 34 karyotype, there are reasons to surmise that Brazilian forms with low chromosome numbers must be placed else- where. Thanks to the kindness of Dr. A. Langguth, I recently had the opportunity to examine speci- mens of so-called arviculoides from Pemambuco showing the 2n = 1 6 karyotype described by Maia and Langguth. I concluded that they belong to Akodon s.s., probably to a new species related to A. cursor. The status of the Sao Paulo and Rio de Janeiro forms reported by Yonenaga and associ- ates is still unsettled, although the chromosomal evidence indicates that they may belong to the 354 FIELDIANA: ZOOLOGY Fig. 3. Skulls in lateral view of species ofBolomys, Akodon {Akodon), and Zygodontomys. A, Akodon {Ak.) dolores Thomas; male; type specimen, BMNH 1 6. 1 .6.38; Villa Dolores, Cordoba, Argentina. B, Akodon (Ak.) cursor (Winge); female; BMNH 66.1874; Puerto Gisela, Misiones, Argentina. C, Zygodontomys thomasi Allen; female; BMNH 14.9.1.60; El Trompillo, Carabobo, Venezuela. D, Zygodontomys microtinus Thomas; female; holotype, BMNH 66.8.1 1.10; Surinam. E, Akodon {Ak.) varius Thomas; female; holotype, BMNH 2.1.1.67; Cochabamba, Bolivia. F, Bolomys lasiurus lasiurus (Lund); male; topotype, BMNH 88. 1 .9.4; Lagoa Santa, Minas Gerais, Brazil. G, Bolomys lasiurusfuscinus Thomas: male; holotype, BMNH 94.4.1.3; Soure, Ilha de Marajo, Para, Brazil. H, Bolomys obscurus bene/actus Thomas; male, holotype, BMNH 16.10.3.35; Laguna Alsina (Bonifacio), Partido de Guamini, Buenos Aires Province, Argentina. I, Bolomys lactens lactens Thomas; female; holotype, BMNH 18.1.1.37; Leon, Jujuy Province, Argentina. J, Bolomys lenguarum (Thomas); sex unknown; holotype, BMNH 98.5.14.1; Waikthlatingmay- alwa, northern Chaco, Paraguay. REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 355 same group, and that at least the 2n = 24-25 poly- morphic form described by Yonenaga as Akodon sp. from specimens captured in Sao Paulo may represent another, probably synmorphic new species. To make this story even more complex, Lang- guth ( 1 975) demonstrated that Mus lasiotis Lund, 1838 is based on a specimen which belongs to the same species as the holotype of Mus lasiurus Lund, 1837. Therefore, Langguth placed lasiotis under the synonymy of lasiurus, which seems correct. However, the specimens described by Winge (1887) as Habrothrix lasiotis belong certainly to a species distinct from lasiurus. One of the specimens iden- tified by Winge as lasiotis was sent to the British Museum and was the basis of lasiotis Thomas, 1916, being selected as the type species of his genus Thalpomys. Langguth proposed a new name, Ako- don reinhardti, for Winge's and Thomas's lasiotis, but as Thalpomys was based on the species lasiotis as described by Lund, and as lasiotis Lund is a junior synonym of lasiurus Lund, and as lasiurus belongs to Bolomys, Thalpomys becomes a sub- jective synonym of Bolomys. This nomenclatorial issue means that, if lasiotis as conceived by Winge and Thomas is really distinctive enough from Ako- don to deserve generic or subgeneric distinction, a new name should be proposed to replace Thal- pomys for the taxon containing reinhardti Lang- guth. However, this species is now too poorly known to justify this action, and I prefer to keep reinhardti in Akodon s.s. Recently, Massoia (1982) described a new species, Cabreramys temchuki from Misiones, Ar- gentina. Later, Contreras (1982) referred temchuki to Bolomys and described two new subspecies from northern Corrientes Province and northeastern Chaco Province, Argentina: B. t. elioi and B. t. liciae. The previous conclusions on Bolomys and its species, based on morphological analysis, are also supported by cytogenetic data. Karyotypes of 2n = 34 identical in gross morphology have been de- scribed or recorded in Akodon amoenus (Gardner & Patton, 1976; O. Pearson, pers. comm.), Ako- don obscurus (Bianchi et al., 1971), Akodon ar- viculoides (Yonenaga & Ricci, 1969), Zygodon- tomys lasiurus (Yonenaga, 1975), Zygodontomys sp. (probably B. lenguarum) (Gardner & Patton, 1976). More recently, banding patterns have been published for Bolomys obscurus (Bianchi et al., 1976) and B. lasiurus (Maia & Langguth, 1981), and we have recently confirmed that B. temchuki shares the same G-banded karyotype with the lat- ter (VituUo et al., 1986). As with Oxymycterus. Bolomys shows karyotypic homogeneity, although differences in banding pattern are not unexpected when more detailed comparative studies are com- pleted. Species of Akodon are variable in karyo- types, but none of them have the 2n = 34 karyo- type reported for obscurus, lasiurus, and allies. However, B. obscurus was found to be more closely related in karyotype and G-banding to some species of Akodon (e.g., A. dolores and A. azarae) than other species of Akodon are related to each other (Bianchi & Merani, 1984). This must be inter- preted as an indication that Bolomys evolved amidst the radiation of Akodon, which is consis- tent with the results from prostate morphology. Once differentiated from Akodon, Bolomys diver- sified as a clade of its own by adaptive divergence that did not involve significant chromosomal evo- lution. Therefore, the karyotypic distinction and uniformity of amoenus, lasiurus, lenguarum, ob- scurus, and temchuki must be taken as additional evidence that this group of species (together with lactens for which we lack chromosomal infor- mation) belongs to a separate group for which the name Bolomys Thomas, 1 9 1 6 is available and has priority. If, as it seems most likely, Bolomys got its identity as a genus within the cladogenesis of Akodon, the latter should be considered as a para- phyletic taxon, which I consciously admit. Bolomys is polytypic and widespread, compris- ing several different species and subspecies dis- tributed in the Central Andes, the Chacoan and Pampean regions, northeastern Argentina, eastern and southern Brazil, and Uruguay. It is not yet clear how many sp)ecies must be recognized within this genus. The specific or subspecific status of the various nominal forms described are still prob- lematical, and new forms may be recognized in poorly explored regions, as shown by the recent discoveries of Massoia (1982) and Contreras (1982). As a tentative arrangement, I propose the following: Bolomys amoenus (Thomas, 1900). Living in highlands and Altiplano of southeastern Peru. Bolomys bonapartei Reig, 1978. Fossil from the Pliocene Monte Hermoso Formation of south- em Buenos Aires Province. Bolomys lactens (Thomas, 1918) (fig. 31) (includ- ing Akodon orbus Thomas, 1919 and Bolomys negrito Thomas, 1 926 as junior synonyms; Ako- don leucolimnaeus Cabrera 1926 may deserve recognition as a subspecies). Living in highlands and Pampean mountains of Jujuy, Catamarca, and Tucuman, Argentina. 356 HELDIANA: ZOOLOGY B Fig. 4. Skulls of typical representatives of Akodon (Akodon), Akodon {Abrothrix), Microxus, and Oxymycterus. Left row, dorsal view; right row, lateral view. A, Akodon (Ak.) boliviensis; male; MBUCV 1.1889; 30 km N.W. OUantaytambo, Cuzco, Peru. B, Microxus mimus Thomas; female; holotype, BMNH 1.1.1.48; Limbane, Puno, Peru. C, Akodon (Ab.) longipilis CWalerhouse); male; topotype, BMNH 98.2.2.2; Valparaiso, Chile. D, Oxymycterus platensis (= Oxymycterus rufus platensis?); male; holotype, BMNH 99.10.4.1; Ensenada, Rio Santiago, La Plata, Buenos Aires Province, Argentina. Bolomys lasiurus (Lund, 1 838) (including Mu5 la- siotis Lund, 1838, Hesperomys arviculoides Wagner, 1842, and Hesperomys brachyurus Wagner, 1 845, as junior synonyms; Akodon fus- cinus Thomas, 1 897, and Zygodontomys pixuna Moojen, 1943, may deserve recognition as sub- species). Living in eastern and southern Brazil, including the states of Para, Ceara, Pemambuco, Paraiba, Sergipe, Bahia, Minas Gerais, Sao Pau- lo, and vicinities. Bolomys lenguarum (Thomas, 1 898) (probably in- cluding Zygodontomys tapirapoanus J. A. Al- len, 1916 as a junior synonym). Living in Par- aguayan, Bolivian, and (probably) Argentinian Chaco, and Planalto de Mato Grosso, Brazil. Contreras (pers. comm.) says that specimens from the Argentinian Chaco may belong to a different species. Bolomys obscurus (Waterhouse, 1837) (including Akodon bene/actus Thomas 1919 probably as a distinct subspecies). Living in southern Uru- guay, northwest of Buenos Aires Province, south of Cordoba and Santa Fe, and east of La Pampa, Argentina. Bolomys temchuki (Massoia, 1982) (including Bolomys t. elioi Contreras, 1982 and B. t. liciae Contreras, 1982, as subspecies). Living in Mi- siones, north of Corrientes and northeast of Chaco, Argentina. Bolomys innom. sp. (reported as Akodon obscurus and Zygodontomys obscurus). Living in south- em Buenos Aires Province. It is a larger and REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 357 darker form found at Partidos de General Puey- rredon, Balcarce, and General Juan Madariaga and at Sierra de La Ventana, to be described in a forthcoming paper. As advanced above, Akodon albi venter Thomas (fig. IC) and A. berlepschii Thomas, originally in- cluded by Thomas (1916) in Bolomys, do not be- long to that genus. In a previous paper (Bianchi et al., 1971), the former was considered to be a Bolomys by uncritical adoption of Thomas's opin- ion. The two species are probably conspecific (see Thomas, 1 902). As already pointed out by Osgood (1943), albiventer shows several similarities with Akodon andinus (Philippi) (fig. IB), and it does not exhibit any of the distinctive character-states of Bolomys. Moreover, albiventer has a 2n = 40 karyotype very similar to that of A. boliviensis and A. varius (Bianchi et al., 1 97 1), which is easily distinguished from the 2n = 34 karyotype of species of Bolomys. Akodon and Its Allied Supraspecific Taxa Once Bolomys is excluded, the limits of Akodon as a genus appear more precise. However, it is still mostly a matter of sense and balance to decide whether Abrothrix, Akodon s.s., Chalcomys, Chroeomys, Del t a my s, Hypsimys, Microxus, Thalpomys, and Thaptomys should be considered as separate genera, different subgenera within Ako- don, or simply synonyms of Akodon. Some time ago (Reig et al., 1971), following Cabrera (1961), I included Chalcomys as a syn- onym of Akodon. I also advocated (in Bianchi et. al., 197 1) generic rank for the names listed above, excluding Chalcomys and Deltamys. Afterward (Reig, 1978, 1981), I reevaluated that view, as further study indicated more strongly that species referred to most of these names represent a com- plex array of allied forms, more closely related to each other than with other distinctive genera of akodontines, such as Bolomys and Oxymycterus. This is certainly a subjective feeling which badly needs to be substantiated by careful quantitative systematic analysis of morphological, chromo- somal, and allozymic evidence. However, until such study is accomplished, the available evidence suggests that they are more likely to constitute a single unitary, although paraphyletic, taxon better treated as a large central genus within which some distinctive species or set of species may be distin- guished as subgenera. However, as substantiated below, species referred to Microxus seem better excluded from Akodon, Thalpomys is a subjective synonym of Bolomys (as explained above), and Chalcomys seems not to deserve recognition. I also believe that separation of Thaptomys from Ako- don s.s. is unwarranted. Actually, I could not find differences distinctly beyond the limits of varia- tion within Akodon s.s. to keep A. nigrita Lich- tenstein, the type and single species of Thaptomys (see Cabrera, 1 96 1 ; Massoia, 1 963a) as a separate subgenus. Cranially and dentally (figs. 10E,I), ni- grita is a typical Akodon, and its alleged fossorial adaptations are too incipient to deserve any special taxonomic treatment. Its only distinction in male accessory glands is the presence of a single pair of ventral prostates in A. nigrita (Voss & Linzey, 1981), but the existing survey of species is still too small to give much weight to this isolated dis- tinctive feature. Thus, I tentatively recognize a central genus Akodon subdivided into five subgenera, namely Abrothrix, Akodon s.s., Chroeomys, Deltamys, and Hypsimys. 1 do not deny, however, that more de- tailed and comprehensive further studies could eventually elevate some of these taxa to generic level, but this action does not seem warranted by the present state of knowledge. Recently, Massoia (1981b) and Contreras and Rosi (1981) advocated generic status for Abro- thrix, whereas other modem authors accord it only subgeneric status (Yanez et al., 1978; Patterson et al., 1984). Actually, species of Abrothrix seem to be rather distinctive in skull and dental morphol- ogy (figs. 4-6). They differ from Akodon s.s. in the more robust and elongated skulls; long nasals which are clearly longer than frontals and that exceed backward the frontomaxillary suture and project also forward; long and rounded braincase; longer and more slender muzzle; slight inflation of the anterodorsal frontal sinuses; longer palate; low and elongated mandible with a projecting incisor cap- sule; broader molars, with obsolete or completely absent anteromedian flexus in M' and antero- median flexid in the m,; bulging of entoconid in m, and mj; and total absence of ectolophids and rare presence of ectostylids. It also differs from typical Akodon in the woolly and longer pelage, in the relative sizes of ampullary glands and pros- tates, and in the presence of two pairs of preputial glands (Voss & Linzey, 1 98 1). However, in overall resemblance, species of Abrothrix are clearly more closely related to species of Akodon s.s. than species of conventionally distinctive genera as Oxymyc- terus, Bolomys, or Blarinomys. Thus, it seems wis- 358 FIELDIANA: ZOOLOGY er to keep it as a subgenus of Akodon. Abrothrix is polytypic, but it is not clear how many species it contains. Typically, longipilis, lanosus, and san- borni and their subspecies (see Yanez et al., 1978) belong to Abrothrix. All inhabit the southern An- des and adjacent lowlands and show an identical 2n = 52 karyotype (Bianchi et al., 1971; Spotomo & Fernandez, 1976; Gallardo, 1982; Reig, unpubl. data). Akodon illutea Thomas, 1925 (fig. 7D) is an Abrothrix in molar and skull morphology. How- ever, living in the mountains of Tucuman in northwestern Argentina, it is isolated in distri- bution from the former group and was reported (Bianchi et al., 1971; Dulout et al., 1976) to have a polymorphic 2n = 4 1 karyotype. However, Bar- quez et al. (1980) claimed that the identification of the specimens bearing those karyotypes was in error, and that they may represent Akodon varius, which seems quite probable. Akodon xanthorhi- nus has also been alleged to belong to Abrothrix (Bianchi et al., 197 1), primarily on karyotypic and distributional grounds. It inhabits the same gen- eral region of longipilis and allies, and shows a 2n = 52 karyotype quite similar to that of longi- pilis (see Rodriguez et al., 1983). However, the same karyotype is present in A. andinus, A.jelskii, A. nigrita, A. olivaceus, Chelemys macronyx, and Geoxus valdivianus (Yonenaga, 1975; Gardner & Patton, 1976; Spotomo & Fernandez, 1976; Pear- son, 1 984) and seems to represent a plesiomorphic character-state for akodontines as a group (Vitullo et al., 1986). Although the symplesiomorphy in karyotypic character-states hinders inclusion of xanthorhinus in Abrothrix, this conclusion is sug- gested by other attributes. Actually, xanthorhinus shares with species of Abrothrix an elongated ros- trum, long nasals, rather inflated braincase, and similar molar patterns, including obliterated an- teromedian flexi in the upper first molars. In this connection, it is quite relevant to take into account the recent parasitological results of Dolores del C. Castro. In her studies of the lice Hoplopleura (Ano- plura), she erected the species H. andina, of the travassosi group (Castro, 1981). This group of species are typical ectoparasites of oryzomyines, whereas akodontines are more typically infected by lice of the H. aikeni species group. Hoplopleura andina was found on Akodon olivaceus, A. andi- nus, and A. xanthorhinus (Castro, 1 982) and more recently also on A. longipilis (Castro, pers. comm.). Thus, xanthorhinus shares the same sp>ecies of lice with a typical Abrothrix (A. longipilis) and with other southern species of Akodon s.s. having 2n = 52 chromosomes. Besides, the hce belong to a group which is more typical of the more primitive ory- zomyines. Needless to say, a definite conclusion on the relationships of the southern akodonts needs more detailed studies, and especially an adequate evaluation of polarity in character-states. By over- all resemblance, however, it seems that the avail- able evidence should advise placing xanthorhinus in Abrothrix. Recently, Patterson et al. (1984) maintained xanthorhinus in Akodon s.s., but they described a new species, A. hershkovitzi, which they claim to be closely related to xanthorhinus. From the illustrations and description of these au- thors, I am also inclined to place hershkovitzi in Abrothrix, in keeping with the view of Patterson and associates that hershkovitzi represents an is- land derivative of xanthorhinus. In addition to the species discussed above, Akodon {Abrothrix) man- soensis was recently described by de Santis and Justo (1980). The status of this name does not seem clear to me, and from the illustrations and description, I surmise that it may be a local form of ^. olivaceus, a view which is also agreed upon by O. Pearson (pers. comm.). However, a com- prehensive revision of the southern Akodon has not yet been conducted, and my conclusions on xanthorhinus, hershkovitzi, and mansoensis are only tentative. Meanwhile, and in the following treatment, the concept of Abrothrix is centered around the typical species longipilis, sanborni, lanosus, and illutea. Chroeomys is certainly a well-differentiated off- shoot from Andean akodonts. Akodon jelskii, its type species, can be easily distinguished from species of Akodon s.s. by the broadened braincase; enlarged and moderately swollen bullae; shorter palate; M' lacking a distinct anteromedian flexus, but with a well-developed and deeply infolded an- teroflexus; reduction or absence of mesoloph rem- nants and mesostyles; deeply infolded anterome- dian flexus of m,; absence of mesolophids, ectostylids, and mesostylids. Chroeomys also dif- fers from typical Akodon in having two pairs of preputial glands instead of one pair and in the large size of the median prostates, which are reduced in species of the subgenus Akodon (Voss & Linzey, 1981). Again, it is more like members of that genus in the sum of its characters than are members of other akodontine genera. Following the revision of Sanborn (1947), the five species recognized by Thomas are lumped into a single polytypic species, A. jelskii. Hypsimys is also a well-differentiated subgenus of Akodon. Superficially, A. budini, its type species, may be confused with a Bolomys, but the confu- i REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 359 sion rapidly disappears when one examines the skull, which resembles that of Akodon in overall morphology. However, it differs neatly from Ako- don s.s. in its long, round, and broad braincase; shorter palate; more elongated incisive foramina; more slender mandible and, especially, in the clearly more hypsodont molars, which are also more elongated and narrower than those in Ako- don, Chroeomys, or Abrothrix. My associates and I have recently studied the karyotype of topotyp- ical specimens of ^. budini, finding that it has a peculiar 2n = 38 karyotype quite distinctive in its banding pattern (Vitullo et al., 1986). After com- parison of the type specimens in the British Mu- seum, I agree with Cabrera (1961) in considering deceptorThovciSiS, 1921 as a junior synonym of ^. budini Thomas, 1918. Thus, Akodon (Hypsimys) is also monotypic. Deltamys has also been recently accorded ge- neric status (Massoia, 1981b; Gentile de Fronza et al., 1979). The absence of a small pair of meta- centrics in its chromosome complement was given as an argument for its separation, but this is not convincing. Hypsimys also lacks this minute meta- centric characteristic of akodontines, and Zygo- dontomys microtinus shows the character, being quite different from akodontines in many other respects. Cranially and dentally, Deltamys shares many more character-states with Akodon s.s. and other discussed subgenera than with any other ge- nus. It is, however, quite distinctive as a subgenus in its elongated and narrow skull, with weak and low zygomatic plate; longer narrowed braincase; very reduced interparietal; shorter palate; low and elongated mandible with projecting incisor cap- sule; elongated and narrow, rather more hypso- dont molars (less so than in Hypsimys); inclined molar flexi and flexids; M' with well-developed anteromedian flexus, but anteromedian flexid of m, obsolete; large m,, almost as long as mj. Up to now, only one species A. (D.) kempi is recog- nized, along with two subspecies (Massoia, 1981b). Repeating the log-normal distribution of species mentioned in the Introduction section, Akodon s.s. is by far the most polytypic of all Akodon. Al- though the number of recognized species is a mat- ter of debate, there are at least 23 species in this subgenus. It is probably the most generalized of the akodontines, representing the stock from which several episodes of akodontine diversification started in different places and probably at different times (see later). Both in morphology and in cy- togenetics, it is a complex taxon, and it may prove convenient in the future to distinguish within it various species groups and superspecies. A formal description of its skull and dental characters and the present status of its species is considered be- low. The situation of Microxus remains to be dis- cussed. Described as a genus by Thomas (1909), it was accorded generic rank in his later synthesis (Thomas, 1916). It was also recognized as a genus by Tate (1932b) and Ellerman (1941). However, Cabrera (1961) treated Microxus as a subgenus of Akodon, a view followed by Arata ( 1 967) and Voss and Linzey (1981). Hershkovitz (1966) included Microxus under Abrothrix on the argument that M. mimus, the type species, is actually an Abro- thrix, a view echoed by Gardner and Patton ( 1 976). I studied the type specimen of M. mimus (BMNH 1.1.1 .48) and of M bogotensis (BMNH 95. 1 0. 1 4.2) and several specimens referred to the latter from the Andes of Venezuela, and I concluded that Mi^ croxus belongs neither to Akodon nor to Abrothrix and that it is a distinctive genus of Akodontini. Including M. mimus (fig. 4B) and M. bogotensis, Microxus differs from Akodon and Abrothrix as well as from other akodontines by its inflated short and deep braincase; short and low zygomatic plate with anterior border gradually sloping forward; muzzle narrow and elongated; strongly reduced interparietal; long frontals; interorbital region rather broad; molars relatively large and broad, with a simplified enamel pattern and cusps well opposed to each other, and noticeably reduced up- per and lower M3. Microxus bogotensis is also peculiar in possessing a single pair of ventral pros- tates (Voss & Linzey, 1981). Additionally, and as discussed above, species of Akodon (Abrothrix) share an identical 2n = 52 karyotype, whereas specimens assigned to M. bogotensis collected at Mucubaji, Merida (Venezuela), showed polymor- phic karyotypes of 35-37 chromosomes (FN = 48) (Barros & Reig, 1 979). Further differences and dis- tinctive character-states of Microxus in the skull and in different organ systems are provided by Vorontzov (1982). Thus, all lines of argument in- dicate that, at the present state of knowledge, this taxon is better kept as a genus within the Ako- dontini. Other Genera Recognized Within the Akodontini As discussed above, Oxymycterus is not sepa- rable from the Akodontini on chromosomal grounds. Moreover, it is clearly an akodont in cra- nial and tooth morphology and intestine and lung 360 HELDIANA: ZOOLOGY anatomy (Vorontzov, 1967, 1982). While there are no grounds to separate it from the Akodontini, it is I indeed a distinct genus within the tribe. It sharply differs from the remaining akodont genera by its larger size, elongated skull with anteriorly ex- panded and trumpet-like nasals (fig. 4D), reduced zygomatic plate, simplified enamel pattern of mo- lar teeth, diet (Kravetz, 1973), stomach morphol- ogy (Echave Llanos & Vilchez, 1 964; Vorontzov, 1967, 1982; Carleton, 1973), phallic morphology (Hooper & Musser, 1964), and male accessory glands (Voss & Linzey, 1981). Oxymycterus is rather polytypic, although the status and nomenclature of some of its species are still dubious. Cabrera (1961) recognized eight species and 13 subspecies. Honacki et al. (1982) accorded specific status to nine nominal forms. The genus needs revision. I examined the type specimens in the British Museum and can advance some results. I Following Massoia and Fomes (1969), I distin- guish O. nasutus from O. rufus. At present, it is questionable whether O. platensis represents a subspecies of O. rufus or a distinct species. I am inclined to the former alternative; in any case, platensis is fully separated from nasutus (Massoia & Fomes, 1969). Regarding rufus, I follow Lang- guth (in Honacki et al., 1982, p. 459) in using Fischer's name instead of rutilans Olfers. It is at present unsettled whether delator is different from rufus. Oxymycterus paramensis and O. akodontius are reported to show some differences in skull and color characters, but it is unclear whether they are different sp)ecies or merely phenotypic variants of a single species. Oxymycterus incae (including dor- is, inca, iris, and juliacae as subspecies) is quite distinct from paramensis or akodontius. As rec- ognized by Massoia (1963b), iheringi is a good species of Oxymycterus, not a Microxus; this au- thor is wrong, however, in allocating Akodon san- borni to Oxymycterus. I follow Cabrera ( 1 96 1 ) in placing judex, misionalis, and quaestor under O. hispidus. However, this action is tentative until I the type of hispidus is re-studied; but judex and quaestor are similar to each other and distinctive from rufus and the other mentioned species. Oxy- i> mycterus angularis and O. roberti are also ten- tatively accepted as valid species. The karyotypes of four studied species of Oxymycterus (Vitullo et al., 1986) are identical in number (2n = 54), mor- phology, and in the known cases, in G- and C-banding patterns of chromosomes. Thus, kar- yology may not afford good markers to clarify the status of the dubious species. Lenoxus, known from the single species L. ap- icalis (with boliviae as a subspecies), is just an exaggerated Oxymycterus in size and skull mor- phology. The skull, as illustrated by Vorontzov (1982, fig. 1 12), is much like that oT Oxymycterus, but can be distinguished by the broadened incisive foramina, larger palate, broader interorbital re- gion, and greater development of the interparietal. The molar teeth show a less simplified enamel pattern, are more brachyodont, and have more sharply defined cusps; in the upper teeth, the lin- gual cusps are more displaced backward relative to the labial ones, and the anteromedian flexus and flexid are much more developed and pene- trating than those in species of Oxymycterus (Vo- rontzov, 1 982, p. 228, fig. 1 92). Additionally, Len- oxus can be distinguished from Oxymycterus in the less-developed foreclaws, the softer fur, and the longer tail. Juscelinomys, which Moojen (1965) erected for J. candango from the "cerrado" near the city of Brasilia, and which probably also includes Oxy- mycterus talpinus V/inge, 1888 (^e Moojen, 1965) from Minas Gerais, is certainly an akodontine closely related to Oxymycterus and Lenoxus. From its description and illustration, its generic distinc- tion seems also warranted, as ensues from the di- agnosis, which, taken with modifications from Moojen (1965, p. 283), is as follows: An akodontine cricetid close to Oxymycterus in dental character states, body form, and fos- sorial way of life. It differs from Oxymycterus in external characters in less elongated muz- zle; tail very thick and densely covered with hairs that totally hide the tail scales; ventral region of fur, with rufous hairs whitish in their base. Skull stronger built; bullae swollen; ros- trum short and broad; premaxillae not ex- tending beyond the level of upper incisors; nasals reaching slightly in front of the incisors, but not trumpet-like; posterior border of in- fraorbital foramen in front of the level of first molars; zygomatic plate more anterior; dia- stema shorter; incisive foramina F>enetrating between the first upper molars beyond their protocone. Upper incisors slightly grooved, with the groove closer to their lateral borders; molars stronger; procingulum of first upper molar with a deep anteromedian flexus. Four pairs of mammae. Podoxymys, erected by Anthony (1929) for the single species P. roraimae from the tepuis of the Guiana region, has also been alleged to stand near REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 361 Oxymycterus (Hershkovitz, 1 966), although Eller- man ( 1 94 1 ) pointed out that, from the description, it seems to be very close to Microxus. In fact, the description is rather vague, but doubtless it is a distinctive genus of the Akodontini. Its stomach is unilocular and hemiglandular (Carleton, 1973; Vorontzov, 1 982) as in most akodontines, but dif- fers in having a reduced area of glandular epithe- lium. Its molar teeth differ from those o^ Microxus and Oxymycterus in being much simplified, al- most as much as in the otherwise quite different genus Geoxus. Blarinomys is an akodontine highly modified for fossorial life and insectivorous diet (Vorontzov, 1982): extremely reduced eyes; tail short; ear very small; broad hind foot with fairly strong claws; hand with long claws. The skull is characterized by the absence of the interparietal, a broad inter- orbital region, and a very narrow zygomatic plate. The molars are highly modified in enamel pattern, and the mj is fairly reduced in size. A point of clarification is required on the status of the fossorial southern akodonts currently re- ferred to Notiomys. Thomas advocated that the burrowing, short-tailed mice of this group inhab- iting southern Chile and Argentina represented three genera, namely Notiomys Thomas, 1 890 (with Hesperomys edwardsi Thomas, 1890, as the type species), Geoxus Thomas, 1919 (with Notoxusfos- sor Thomas, 1919, as the type species), and Che- lemys Thomas, 1903 (with Hesperomys megalo- nyx Waterhouse, 1 844, as the type species). Later, Osgood (1925) proposed combining all three gen- era under Notiomys, a contention which, in spite of Thomas's (1927) criticisms, was accepted by most later authors, including, recently, Honacki et al. (1982). However, Gyldenstolpe (1932) accept- ed the generic rank of all three taxa. On the basis of the study of the types and series of specimens in the British Museum and the American Museum of Natural History, I concluded that two genera may be distinguished, namely Notiomys (including Geoxus) and Chelemys (Reig, 1981). More re- cently, Pearson (1983) treated Chelemys and Geoxus as different genera, and said that he con- sidered it untenable to assign Geoxus to Notiomys. In a subsequent paper (Pearson, 1984), he sup- ported his view with evidence from ecology, food habits, and external and craniodental morphology. With the new information of Pearson, and after a reconsideration of the information from my for- mer studies, I am ready to accept the Thomas- Pearson view. It is in order, therefore, to afford an up-to-date definition of these three genera. Notiomys Thomas, 1 890 Size small: head and body less than 100 mm. Tail shorter than half the length of head and body, even shorter than in Chelemys; front claws long, less strong than in Geoxus; claws of the hind feet shorter than in Geoxus; fur not molelike, brightly colored, with whitish underparts and with bright rufous nose; margins of the hind feet with a shaggy fringe of hairs; ears with extremely thin and small pinnae, hidden in the fur; nose tipped with a dark leathery button. Skull rather heavy in shape, much shorter and wider than in Geoxus; muzzle shorter, wider and conical; incisive foramina short, not reaching the anterior borders of M'; palate long, extending backward far behind the level of last molars, with small posterior palatal pits; pterygoid region shorter than in Geoxus, with rounded an- terior border of mesopterygoid fossa; zygomatic plate reduced, moderately slanting to almost ver- tical; interparietal small and narrow; frontals not inflated, interorbital region broad, with sharply squared supraorbital ridges; mandible not as slen- der and proodont as in Geoxus, with long coronoid process, condyloid process not bent inward, and small capsular projection. Incisors slender, the up- per ones orthodont; molars brachyodont, small, fairly narrow and elongated; M' with procingulum reduced, without anteroflexus but with a very shal- low anteromedian flexus; upper molars with para- and metaflexus moderately oriented backward; mesoflexus present, deep and narrow on M ' . Me- sostyle and enterostyle missing; M' simple and ■ strongly reduced, less than Vz the length of M^; procingulum of m, narrow, with a shallow an- teromedian flexid; lower molars with meso- and \ posteroflexids not deeply infolded and lophids ori- ] ented rather transversely in position; mesolophid ^. remnants visible on m , ; hypoflexid shallow; meso- \ stylid and ectostylid absent; m, strongly reduced, rounded and simple, V2 the length of mj; cecum I absent; stomach and male glands unknown. Included species: edwardsi. i Geoxus Thomas 1919 Size small: head and body length less than 100 mm; tail shorter than half the length of head and ! body, even shorter than in Chelemys; front claws J stouter and longer than in Notiomys and Chele- j mys; hind foot with long claws, and without a shaggy fringe of hairs on the margins; fur molelike, uniform in color; ears small, but pinnae easily 362 HELDIANA: ZOOLOGY visible; nose normal. Skull slender and rather del- icately built, longer and narrower than in Notio- mys; muzzle long and narrow; incisive foramina long, surpassing backward the anterior borders of M'; palate long, extending past the level of last molars, without posterior palatal pits; pterygoid region long; mesopterygoid fossa squarely open in front and nearly parallel-sided; zygomatic plate reduced, narrow and slanting; interparietal rather narrow and short, not extremely reduced; frontals rather inflated; interorbital region smooth and rounded, the supraorbital edges not square-shaped; mandible slender and proodont, with short coro- noid process, condyloid process not bent inward, and small capsular projection. Incisors rather slen- der, the upper ones orthodont; molars small and narrow, simple and brachyodont; M' with reduced procingulum with a shallow anteroflexus and no anteromedian flexus; upper molars with trans- versely oriented para- and metaflexus, mesoflexus absent and para- and mesoloph coalesced; meso- style and anterostyle missing; M^ strongly reduced, less than % the length of M^, ring-shaped; procin- gulum of m, narrow, without anteromedian flexid; lower molars with transverse lophids and flexids little infolded; mj reduced, less than % the length of mj, and with a simplified enamel pattern. Ce- cum missing; stomach unilocular-hemiglandular, with glandular epithelium more extended than in Akodon; medial ventral prostates large, and bul- bourethral gland much larger than typical condi- tion. Included species: valdiviantis (including valdi- vianus, fossor, chiloensis, bullocki, and bicolor as subspecies) and, probably, tnichaelseni (see Pear- son, 1984, p. 233). backward; zygomatic plate comparatively strong and broad, its anterior border nearly vertical or slightly slanting forward; interparietal normally developed, not reduced, although relatively nar- row; frontals not inflated, interorbital region smooth and rounded, with supraorbital edges not square-shaped; mandible rather strongly built; ra- mus much deeper than in Geoxus or Notiomys, with long coronoid process, condyloid process bent inward, and a medium-sized capsular projection. Incisors thick and robust, the upper orthodont, the lower less proodont than in Notiomys and Geoxus; molars rather large and broad, relatively hypso- dont; M' with moderately reduced procingulum lacking any trace of anteromedian flexus and an- teroflexus; upper molars with para- and metaflexus strongly oriented backward; mesoflexus complete- ly reabsorbed by complete fusion of para- and me- soloph; mesostyle normally absent, enterostyle to- tally missing; M' relatively well developed, about % the length of M^; m, with short procingulum without anteromedian flexid; lower molars with ento- and posteroflexids oblique, metaflexid more transverse; mesoloph and paraloph coalesced, mesostylid and ectostylid missing; mj sigmoid- shaped, not reduced, more than % the length of mj; cecum present; stomach and male glands un- known. Included species: macronyx (including macro- nyx, alleni, fumosus, and vestitus as subspecies), megalonyx (including megalonyx and microtis as subspecies), and probably, delfini, a species from Punta Arenas whose status is dubious and which might represent a third subspecies of megalonyx. For removal ofangustus from Chelemys, see Pear- son (1984, p. 231). Chelemys Thomas, 1 903 Size larger: head and body length longer than 1 20 mm; tail shorter than half the length of head and body; front claws stout and long, but com- paratively less developed than those in Geoxus; hind foot with moderately long claws, without a shaggy fringe of hairs on their margins; fur short and dense, uniform in color; ears small, but pinnae visible; nose normal; skull robust and broadly built; muzzle broad and rather short; incisive foramina long, extending backward beyond the anterior bor- ders of M'; palate short, extending backward to the level of last molars, with small posterior palatal pits; mesopterygoid fossa deep, with rather squared front edge, and lateral borders slightly divergent Defining the Tribe Akodontini As discussed in the previous sections, 1 1 extant genera are recognized within the Akodontini; namely, Akodon, Blarinomys, Bolomys. Chele- mys, Geoxus, Juscelinomys, Lenoxus, Microxus, Notiomys, Oxymycterus, and Podoxymys. The ex- tinct Dankomys is also included. The various genera to which the term Akodon- tini is applied constitute in one sense its meaning. This is the referential approach to the assessment of the meaning of a concept (Alston, 1 964). This sense of meaning is called the extensional or deno- tative meaning. Since defining is an operation by which we explain the meaning of a term, assessing REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 363 in this way the meaning of the Akodontini is equivalent to formulating an extensional defini- tion of that taxon-concept. However, this deno- tative definition is only one meaning of the taxon and is not convenient for further taxonomic pur- poses. Although frequently the elucidation of the extension of a taxon-concept precedes the assess- ment of the properties that are shared by the mem- bers of the concept (the intensional or connotative meaning of the term), a complete definition of concepts requires that the common properties or attributes which qualify a given taxon for mem- bership be defined. This definition of a taxon by its connotative or intensional meaning is an unavoidable aim in tax- onomy, as it enables the zoologist to grasp the basic features which characterize the taxon as a natural product of evolution, and to decide wheth- er as yet unknown or problematic taxa deserve membership in the taxon. To arrive at this definition satisfactorily would require revision of the whole tribe and other re- lated groups of the subfamily Sigmodontinae. However, we can agree on a definition of the term Akodontini by ascertaining from the present body of knowledge its conventional intension (Copi, 1953, p. 102). This definition must be polythetic (Sokal & Sneath, 1963; Bechner, 1959 called the same kind of definition polytypic) in that the Akodontini are defined by reference, not to a set of attributes the common possession of all of which is both nec- essary and sufficient for membership in its exten- sion, but to a set of attributes the common pos- session of a large (but unspecified) number of which is sufficient for belonging to its extension. Needless to say, the same approach is applicable to the def- inition of genera and subgenera within the tribe, and it permits defining the intensional meaning of various taxa by commonality of character-states, abjuring the typological claim of exclusive sharing of one or another of all the alternative states of the characters used in defining taxa of the same rank and which belong to the same taxon of im- mediate higher rank. The intensional meaning of the polythetic con- cept of the Akodontini follows. Akodontini Vorontzov, 1959 Sigmodontine cricetids of small to medium size, with omnivorous to insectivorous digestive sys- tem, without specializations for plant feeding; stomach normally of the unilocular-hemiglandu- lar type, exceptionally of the unilocular-discoglan- dular type; large intestines short, usually less than 15% the length of small intestines; cecum small or absent, never enlarged; molar teeth subhypsodont to mesodont, rarely brachyodont, and crested, ter- raced, or secondarily planed; mesoloph and me- solophid reduced or vestigial when present, often fully or partially coalesced with paraloph or en- tolophid, and only shown as terminal remnants usually united with mesostyle or mesostylid; pos- teroloph coalesced with metaloph, and postero- flexus usually obsolete; skull with zygomatic plate little to moderately developed, never very high and strongly projecting before the antorbital bridge; palate broad, from short to moderately long; in- cisive foramina usually large and reaching back- ward to or beyond the anterior plane of M'; pre- putial gland usually single; medial ventral prostate usually reduced or absent; glans penis complex; diploid karyotype never with more than 54 chro- mosomes. The Akodontini are certainly more derived than the Oryzomyini in molar structure, skull, and male accessory glands (Voss & Linzey, 1981; Voron- tzov, 1982); they are probably direct descendants of the latter. At the same time, they are more primitive than the Phyllotini in molar pattern and digestive system (Vorontzov, 1982) and may have been their ancestors. The Akodontini are predom- inantly Andean in distribution (Reig, 1984), and most of them are inhabitants of open land, al- though some of their representatives may dwell in forested tropical, subtropical or temperate habi- tats, and some of their genera— especially Blari- nomys, Chelemys, Geoxus, and Not iomys— ex- ploited the subterranean niche and might be considered the South American counterparts of the moles. Most species studied are quite omniv- orous (Meserve, 1981; Meserve & Glanz, 1978; Pearson, 1983), but some became more special- ized for an insectivorous diet, such as Oxymyc- terus (Kravetz, 1973), Notiomys, and Geoxus (Pearson, 1983, 1984). Although Chelemys eats mostly mushrooms, and species of Bolomys have been found to be partly vegetarians (Scaglia & Ve- lazquez, pers. comm.), the remaining akodontines did not invade especially the herbivorous niches, which are heavily exploited by their probable de- rivatives, the phyllotines. Description of New Plio-Pleistocene Akodon The study of the fossil material described and discussed in this section necessitated evaluation of the character-states of the genus Akodon and its 364 HELDIANA: ZOOLOGY subgenera Abrothrix and Akodon s.s. in cranial, mandibular, and dental morphology. The full de- scription of the character-states of these taxa is essential for the identification and description of fossil specimens. The following description of the morphological attributes oi Akodon is based in the study of types and series of about 70% of the in- cluded species. It is intended, therefore, to apply to the whole genus. Akodon is, however, rather varied and, as already discussed, the distinction of various subgenera seems appropriate. There- fore, in all those cases where exceptions to the next description are known, the corresponding char- acter-states have been qualified as "usually" pres- ent. Departures from those states in included sub- genera are assessed in the description of morphological attributes of Akodon s.s. and Abrothrix provided afterward. Genus Akodon Meyen, 1833 Type Sv^c\es— Akodon boliviensis Meyen, 1 833, by original designation. Distribution of Living Species— Andean val- leys, highlands, and fringes of mountain forests of Argentina, Bolivia, Chile, Colombia, Ecuador, Peru, and Venezuela; temperate-zone meadows, grasslands and brushlands of Argentina, Bolivia, SE Brazil, Chile, Paraguay, and Uruguay. Cranial Character-States— Skull usually slender, with a typically fairly narrow and rounded braincase, a fairly elongated occipital region, and a rostrum of regular shape, not tapering forward in lateral view. Upper profile of the skull sloping forward and backward from the posterior part of the frontals. Zygomata slightly expanded and not markedly convergent anteriorly. Nasals usually longer than, or as long as, the frontals, their an- terior border passing forward beyond the anterior plane of the incisors, but not projected or ex- panded to form a trumpet-shaped opening. Posterior borders of the nasals usually tapering backward and projecting beyond the fronto-pre- maxillary suture. Frontals long, usually with a nar- row and transversely convex interorbital region with more or less sharply squared edges not de- fining a supraorbital ridge. Frontoparietal suture angular or crescentic in shape. Parietals relatively long, their length, in the midline, usually more than half the length of the frontals, not extending forward through lateral processes between frontals and temporals. Interparietal moderately reduced in width and length. Zygomatic plate moderately developed, relatively narrow and low, but usually with anterior border vertical, straight or slightly concave, not slanting gradually backward from its lower root to the upper border. Upper comer of the zygomatic plate rounded, not projecting for- ward. Incisive foramina wide and elongated, nar- row behind, usually penetrating well beyond the anterior plane of the first molars to reach or slightly surpass the level of the protocone of M'. Posterior palate long and wide, its median posterior border usually slightly behind the posterior plane of the M'. Palatal surface relatively simple, without marked ridges and with shallow grooves. Meso- pterygoid fossa narrow and less than width of parapterygoid fossa. Bullae usually small, less fre- quently moderately large. Mastoid not noticeably inflated. Occipital region somewhat elongated, its posterior border rounded and continuous with the line of the braincase when viewed from the lateral side. Mandible somewhat slender, its height at m, usually shorter than diastema length. Lower mas- seteric crest present, but not strong, reaching for- ward to the level of the anterior half of m,. Upper masseteric crest rather long and usually as strong as the lower one. Coronoid process rather short, with anterior border gradually slanting backward. Condyloid process relatively low, elongated, and projected backward. Articular surface of the con- dyle extending dorsally and slightly posteriorly. Capsule of incisor root normally not projected as a well-developed tubercle, lying on the anterior half of the sigmoid notch. Angular process longer than high. Dental Character-States— Upper incisors usually opisthodont, less frequently orthodont, never proodont. Molar rows parallel-sided. Mo- lars relatively small, usually with moderately de- veloped hypsodonty, with crested and with bilevel occlusal surface in slightly worn teeth, terraced with advanced wear. M' four-rooted. Upper molars with labial (para- cone-metacone) main cusps only slightly posterior to the level of the lingual ones (protocone-hypo- cone), and main lophs transverse in position. M' and M^ usually trilophodont in moderately worn teeth. Mesoloph reduced, usually only partially fused with the paraloph, so that a lingual remnant of it is usually present, forming a bifurcated broad median loph marked by a shallow mesoflexus in M' and M^. Metaloph united with the posteroloph and not reaching the hypocone, almost completely coalesced with posteroloph, so that posteroflexus is very narrow or completely absent in moderately worn teeth. Paraflexus and metaflexus somewhat directed backward, hypoflexus and protoflexus REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 365 somewhat forward, the opposite flexi shghtly al- ternating. Enterostyle and enteroloph usually completely absent. Mesostyle often present in M' and M^, usually united with the remnant of me- soloph. M^ usually much longer than wide. M' much reduced, cylindriform in moderately worn teeth. Procingulum of M' moderately simple, slightly oblique in position, usually clearly bicon- ulate by the presence of a well-developed antero- median flexus. Procingulum united to protocone through an anteroposteriorly oriented anterior mure. Protoflexus absent or undivided by a pro- tostyle, anteroflexus usually present but not deeply infolded. Anteroloph well defined and usually united to parastyle. Protoflexus of M^ evident in moderately worn teeth, absent on M' of similar stage of wear. Lower molars with lingual cusps (metaconid and entoconid) placed fairly anterior to the labial ones (protoconid-hypoconid), with metalophid, as well as entolophid and posterolophid usually oblique, directed slightly forward from the lingual border to the labial one. Mesolophid almost completely fused with entolophid, but a weak lingual remnant, usually united with a mesostyle, is often present in most species. Ectolophid and ectostylid fre- quently present in m,, occasionally in mj, rarely in mj. In m, and mj, hypoflexid broad and trans- verse, mesoflexid directed obliquely forward from outside. Posteroflexid well developed, oblique and parallel to the mesoflexid, absent in mj. m, tetra- lophodont, with a somewhat complex procingu- lum, defined by a usually well-defined, well-in- folded metaflexid, and protoflexid; anteroflexid normally absent. Protostylid and anterolabial cin- gulum usually present. Anteromedian flexid nor- mally well developed in moderately worn teeth, mj trilophodont, well longer than wide, with pro- toflexid usually well marked in moderately worn teeth, mj relatively large, but even smaller than m2, usually bilophodont and sigmoid-shaped in outline, with protoflexus frequently present in moderately worn teeth. Subgenus Akodon, Meyen Akodon Meyen, 1833, Nova Acta Leopold. 16, pt. 11: 599. Hesperomys, Wagner, 1 839, Schrebers Saugeth. Suppl. 3: 516 (in part); Leche, 1886, Zool. Jahrb. 1: 687 (in part). Habrothrix. Thomas, 1 884, Proc. Zool. Soc. London: 450. Chalcomys. Thomas, 1916, Ann. Mag. Nat. Hist. (8)18: 338. Thaptomys. Thomas, 1916, Ann. Mag. Nat. Hist. (8)18: 339. Bolomys, Thomas, 1916, Ann. Mag. Nat. Hist. (8)18: 339 (in part). Type Spejcies— Akodon boliviensis Meyen, by original designation. Known Distribution— As for the genus. Included Species— aerosus, albiventer, andi- nus, azarae, boliviensis, brachiotis, cursor, dolores, iniscatus, markhami, molinae, mollis, nucus, oli- vaceus, orophilus, pacificus, puer, reinhardti, ser- rensis, surdus, tolimae, urichi, varius, and two un- named species from Brazil, one with 2n = 14-16 chromosomes, the other with 2n = 24-25 chro- mosomes (Yonenagaetal., 1975; Yonenaga, 1979; Maia & Langguth, 1981). Additionally, two fossil species which are described below. [The number of living species is certainly larger, mostly because of the complex nature of A. varius. Philip Myers (pcTS. comm.) recognizes three distinct species in what is usually considered to be A. varius; namely, A. varius, A. toba, and A. simulator. He is also inclined to treat A. neocenus as a full species and to recognize A. dayi as a distinct species.] Characters— Skull normally built and usually somewhat elongated behind. Nasals longer to slightly shorter than frontals. Zygomatic plate nor- mal, with anterior border usually vertical in po- sition. Braincase moderately long, usually not broadened, its breadth as large or slightly shorter than '/2 the condylobasal length. Interorbital region usually fairly narrowed, without supraorbital ridges. Anterodorsal frontal sinuses not inflated. Interparietal normal to much reduced. Bullae usu- ally not enlarged. Incisive foramina usually reach- ing the protocone of M' or slightly before it. Pos- terior border of palate behind the posterior border of M'. Mandible relatively high and stout, more slender in the smaller species, with masseteric crest normally developed and reaching the middle of the m,. Incisor capsule usually not projected as a definite tubercle. Upper incisors normally opis- thodont, less frequently orthodont. Molars not markedly elongated and narrowed, with a mod- erately developed tubercular hypsodonty. Molar crowns usually bilevel, terraced to planate with wear, cusps neither noticeably tuberculate, nor with noticeably inclined enamel walls. M' usually with an anteromedian flexus and an anteroflexus. Par- aflexus of M^ directed lingually, the anteroloph in normal position. Mesoloph remnants usually united to mesostyle, typically on M' and M^. Ento- conid wide, but not noticeably bulging laterally in m, and m2. Mesoflexid and posteroflexid of m, and m.2 normally inclined and well developed. An- 366 nELDIANA: ZOOLOGY teromedian flexid of m, frequently present, but only exceptionally deeply infolded, metaflexid moderately to scarcely infolded. Mesolophid rem- nants, ectolophids, and ecto- and mesostylids fre- quently present, mj long, but clearly shorter than m2. Subgenus Abrothrix Waterhouse Mus (Abrothrix) Waterhouse, 1837, Proc. Zool. See. London: 21. Habrothrix Wagner, 1 843, Schrebers Saugeth. Suppl. 3 (in part). Acodon Thomas, 1895, Ann. Mag. Nat. Hist. (6)14: 369 (in part). Abrothrix Thomas, 1916, Ann. Mag. Nat. Hist. (8)18 (considered as a full genus). Akodon {Abrothrix), EUerman, 1941, The Families and Gen. of Living Rodents. Vol. 2: 409, 416; Osgood, 1943, Field Mus. Nat. Hist., Zool. Ser. 30: 184-198. Type Species— ^^o^o/j {Abrothrix) longipilis Waterhouse (by original designation). Distribution— Lowlands and mountain val- leys of southern and central Chile; Andean slopes and low valleys of Argentina from Tierra del Fuego to Mendoza; mountain valleys of Tucuman, Ar- gentina. Included SPEC\E&—hershkovitzi, illutea, lano- sus, longipilis, sanborni, xanthorhinus. (For the provisional assignment of xanthorhinus and hershkovitzi to Abrothrix, see above; see also Pat- terson et al. [1984], who specifically rejected this assignment.) Additionally, the fossil A. kermacki and a new fossil species as described below. Another species referred to Abrothrix, A. mansoensis (de Santis & Justo, 1980), is a dubious form. It is not evident that it belongs to Abrothrix or even that it is a valid sp)ecies of the genus Akodon. Characters— Skull strong and elongated, with a rather long and slender muzzle. Nasals well long- er than frontals, exceeding backward the fronto- maxillary suture and slightly projecting forward. Zygomatic plate relatively deep and short, with anterior border vertical or slightly inclined back- ward. Braincase relatively long, rounded and slightly broadened, its breadth as large as '/z the condylobasal length. Interorbital region of median breadth, without supraorbital ridges and with smoothly rounded edges. Anterodorsal frontal si- nuses slightly inflated, its dorsal surface rounded. Interparietal normally reduced. Incisive foramina elongated but scarcely reaching the protocone of M'. Posterior border of the palate well behind the posterior border of the M^. Bullae not enlarged. Mandible moderately low and elongated, with up- per masseteric crest better marked than the lower masseteric crest, slightly surpassing the middle of the m,. Incisor capsule projected as a tubercle. Upper incisors orthodont, rather strong. Molars comparatively broad, with a moderately well-de- veloped tubercular hypsodonty and a slight crown hypsodonty. Molar crowns bilevel, terraced to planate with advanced wear. Cusps not noticeably tuberculate, with somewhat inclined enamel walls. Anteromedian flexus of M' completely obsolete or barely noticeable. Anteroflexus present, but shal- low. Paraflexus of M^ directed lingually, antero- loph normal. Mesoloph remnants usually united to mesostyles in M' and M^. Mesoflexid and pos- teroflexid of m, and mj well developed, the former scarcely inclined and the latter smaller and nearly transverse in position. In m, and mj entoconid typically bulging lingually. m, with anteromedian flexid obsolete or occasionally present in an incip- ient stage as a shallow and open notch, metaflexid little to moderately infolded. Mesolophid rem- nants constant on m,-m3, but poorly developed and projecting from the anterolateral border of the entoconid, usually connected with mesostylids. Ectolophids completely absent, ectostylids some- times present on m,, very rarely so in m2. m, long, but smaller than m,. Akodon (Abrothrix) kermacki Reig, 1978 HoLOTYPE— MMP S-32 1 , almost complete right and left lower jaws with the entire lower dentition; portion of the left maxilla including the three up- per molars (figs. 5A,E; 6A). Collected by G. J. Scaglia in stratum IX of the Chapadmalal For- mation (Kraglievich, 1952), 500 m north of "Ba- jada del Vivero" (Punta Loberia), Atlantic cliffs of the Partido de General Pueyrredon, SE of Bue- nos Aires Province, Argentina. Figured by Reig and Linares (1969) as Akodon sp. Hypodigm— The holotype and the following: MMP S-222— Almost complete left mandibu- lar ramus, with the entire dentition. Collected by G. J. Scaglia in stratum VIII or IX (Kraglievich, 1952) of the Chapadmalal Formation, 650 m north of "Bajada del Vivero" (Punta Loberia). Figured by Reig and Linares (1969) as Akodon sp. MMP M- 1067— Anterior part of left lower jaw with the whole dentition. Collected by G. J. Scaglia and Mr. Prima in Stratum IX of Chapadmalal Formation at "Bajada del Vivero" (Punta Lo- beria). MMP M- 1 07 1 —Anterior part of right lower jaw with the whole dentition. Collected by G. J. Scaglia REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 367 ■a •c o o 4> s 1 5 I > -CJ M ■« 3 CQ E a X v.. O K U u _« S <«. ^ ■« o ^ E u ^.^ u ^ ;« 3 X E < « ^ ra X o »' _« d ^ « _o c « 3 s % 1 u >< 'C o b » 'S •o « >^ c > ■« CQ ^^ o « c O E 3 1 1 >< >v u CQ X c •o ^^ c oq' » ep ^^ _o '53 V) % '■% a. •c 1 2 o S ^ > ."i? >^ — . -^- E ■^ -x^ T .2P -i«: a o ^ 5T o ^ ^ ^■^ a. ■u3S^ ° eft" M^ CO b ^ O •- _ . 368 HELDIANA: ZOOLOGY in the lower levels of Barranca Lobos Formation (Kraglievich, 1952) in the cliffs north of "Bajada del Vivero" (Punta Loberia). MLP 62.VII.27.84-Almost complete left low- er jaw with the whole dentition. Collected by G. J. Scaglia in association with the holotype. MM? M- 1 1 54— Incomplete right lower jaw with the whole dentition. Collected by G. J. Scaglia in association with the holotype. (For former num- bers of the last two specimens, see Reig, 1978, p. 175.) Known Distribution— Chapadmalalan (Up- per Pliocene) and Uquian (Lowermost Pleisto- cene) ages, SE of Buenos Aires Province, Argen- tina (see Marshall et al., 1983, 1984). Diagnosis— A species of Abrothrix close to A. longipilis; size larger than in A. I. longipilis. Incisor stronger and deeper, M' with a shorter and wider, noticeably oblique procingulum, with a visible, although weakly developed, anteromedian flexus. m, relatively shorter and mj larger than in A. I. longipilis. M' without any evidence of a metafos- setus. Lower jaw with the capsular projection for the base of the incisor more developed than usual in the subgenus. Description— The only known part of the skull, a piece of left maxilla including the cheekteeth which belongs to the type specimen, is too frag- mentary to afford useful information about the structure of the palate. Moreover, it does not in- clude any part of the usually diagnostic zygomatic plate. From the bone tissue preserved posteroin- ternal to the M^, it can be inferred that the pos- terior border of the palate was behind the posterior border of the last upper molar, as it is in Abrothrix. The mandible (fig. 5) is very well preserved in the holotype, MLP 62.VII.27.84, MMP M-1 154, and MMP S-222. It is relatively slender, moder- ately low and elongated, as in Abrothrix and some species ofAkodon s.s. (e.g., A. cursor). The diaste- ma has approximately the same length as the com- bined length of m, and mj, and the depth of the horizontal ramus below m, is less than the dia- stema length in all specimens except MMP M-1 071, in which it is slightly larger. The lower border of the ramus bends gently upward and backward behind the level of the middle of the m2 and descends again behind the m,, shaping a con- cave line, as is usual in Abrothrix. The border of the ramus immediately in front of the m, descends rather abruptly downward, making a slightly ob- tuse angle with the upper border of the symphysis. The symphysis is relatively long and moderately low, and the uppermost anterior point of the di- astema is almost at a level with the alveolar row. The lower masseteric crest is smooth, but well marked, rather high in position, and better de- veloped than is usual in A. longipilis; it reaches to a level anterior to the middle of the m,, but behind its anterior border. The upper masseteric crest is not so well defined as the lower one, and is less developed than in A. longipilis. The mental fora- men is normally developed and opens on the dor- solateral surface of the diastema. The anterior edge of the coronoid process originates at the level of the middle of the mj and slopes gradually upward, with most of the mj visible in lateral view when the mandible is seen perpendicular to the plane of the symphysis. The coronoid process is relatively short, and the condyloid process is low and elon- gated, slightly projected backward, resembling closely the situation found in A. longipilis. The capsular projection, which lies at the level of the anterior part of the sigmoid notch, is stronger than is usual in Abrothrix, reaching a development sim- ilar to that in Deltamys, but less developed than in Bolomys. The greater development of the capsule of the incisor root is obviously a consequence of the more strongly developed lower incisor. This is unusually deep for Akodon sensu lato, and it is absolutely and proportionally deeper than in the living species of Abrothrix. In all the available specimens ofker- macki, the mean depth of the incisor is almost as large as the length of the mj (length m,- 100/depth incisor = 0.993), and in Vi of the available indi- viduals, it surpasses the m, (see table 3, fig. 9). In a sample of 19 ^. /. longipilis from Valparaiso, Chile, in the British Museum, the same index is 0.927, and the length of the m, is, in all the in- dividual cases, longer than the depth of the incisor. The molar teeth agree in all respects with the characters of Abrothrix as stated above in the di- agnosis. In all six known specimens, the masti- catory surface shows an appreciable, but not ad- vanced, degree of wear, corresponding to wear stages 2-3 ofAkodon azarae as described by Pear- son (1967). Therefore, most of the details of the enameled structures of the crown can be observed. The upper dentition is only known from the type specimen. The total length of the upper molar row (4.91 mm in crown length) places A. kermacki among the largest species ofAkodon s.s. (A. urichi, A. varius) and within the range of variation in the available sample of ^. /. longipilis from Valparai- so, which is the largest subspecies of longipilis. In molar morphology, some differences are apparent, which might be of diagnostic value. In M' and M^ REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 369 370 nELDIANA: ZOOLOGY Table 1. Measurement (in mm) of the known specimens of Akodon (Abrothrix) magnus n. sp. and of Akodon (Abrolhrix) kermacki Reig (some of the specimens of A. kermacki were reported with slightly different values for some of the measurements in Reig & Linares, 1 969; the new values given here result from the adoption of conventions used in the present study). Akodon kermacki Barranca Akodon magnus. Chapadmalal Formation Lobes Vorohue rormanon Type, MMP MLP Forma- Type, MMP tion MMP S-321 62-7- MMP MMP MMP MMP Variate M-551 S-107 (r) 25-84 M-1154 S-222 M-1067 M-107I Mandibular condyle- symphysis length 17.00 16.55 16.80 Lower diastema length 4.22 4.61 4.03 4.03 3.99 4.42 3.65 Ramus depth at middle m, 3.96 4.48 3.71 3.78 3.82 4.03 3.84 mi-mj (alveolar) length 6.40 6.14 5.06 5.04 5.57 5.25 5.69 5.57 m.-m, (coronal) length 6.02 5.03 4.86 5.38 4.93 5.63 5.37 m, length 2.42 2.32 1.98 1.89 1.95 1.95 2.29 2.05 m, width 1.55 1.47 1.33 1.30 1.36 1.37 1.58 1.33 m, length -^ 1.93 1.80 1.58 1.46 1.64 1.52 1.74 1.64 m, width 1.61 1.52 1.31 1.30 1.47 1.35 1.55 1.30 m, length ... 1.80 1.43 1.39 1.58 1.49 1.59 1.53 m, width ... 1.25 1.15 1.12 1.20 1.14 1.39 1.12 Lower incisor depth 1.36 1.34 1.43 1.47 1.44 1.54 1.70 1.41 Lower incisor width (thickness) 0.87 0.77 0.86 0.87 0.82 0.84 0.99 0.87 M'-M' (alveolar) length 5.37 M'-M' (coronal) length 4.91 M' length 2.45 M' width ... ... 1.39 M^ length ... ... 1.49 M- width ... ... 1.30 M^ length ... 1.05 M^ width 1.12 the main cusps are nearly opposed, the paracone and metacone being only slightly posterior to the protocone and hypocone, respectively. As in Abrothrix and in Akodon s.s., the M^ is clearly longer than wide, and the M' is strong and com- paratively broad. The procingulum of the M' is short and wide, more so than is usual in A. lon- gipilis and A. illutea, and it is more strongly oblique in position, the anterolingual conule being more anterior than the anterolabial conule. The antero- median flexus is distinct, although it is only very slightly infolded; its presence is also indicated in Opposite Page: Fig. 6. Upper and lower molar teeth of Akodon {Abrothrix) kermacki Reig, and Akodon (Abrothrix) magnus n. sp. A, Right upper molar series and B, Right lower molar series of ^. {Ab.) kermacki; holotype, MMP S-321; Chapadmalal Formation, Partido de General Pueyrredon, Buenos Aires Province, Argentina (Upper Pliocene). C, Left lower molar series of A. (Ab.) kermacki; MMP M-1153; Chapadmalal Formation; found in association with S-321. D, Left lower molar series of A. (Ab.) kermacki; MMP S-222; Chapadmalal Formation, Partido de General Pueyrredon, Buenos Aires Province, Argentina. E, Right m, and mj of A. (Ab.) magnus, n. sp.; holotype, MMP M-55 1 ; Vorohue Formation (Lower Pleistocene), Chapadmalal region, Partido de General Pueyrredon, Buenos Aires Province, Argentina. F, Right lower molar series of A. (Ab.) kermacki; MMP M-1067; Chapadmalal Formation, Partido de General Pueyrredon, Buenos Aires Province, Argentina. G, Right lower molar series of A. (Ab.) kermacki; MMP M-1071; lower levels of Barranca Lobos Formation (lowermost Pleistocene), Partido de Creneral Pueyrredon, Buenos Aires Province, Argentina. H, Right lower molar series of ^. (Ab.) kermacki; MMP M-1 154; Chapadmalal Formation; found in association with S-321 (A-B) and M-1 153 (C). REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 371 Table 2. Length of upper molar row of difTerent species and subspecies of Akodon (Abrothrix). Species and subspecies N Ref.« Akodon (Ab.) kermacki Akodon iAb) longipilis longipilis Akodon (Ab.) longipilis castaneus Akodon (Ab.) longipilis francei Akodon (Ab.) longipilis suffusa Akodon (Ab.) longipilis apta Akodon (Ab.) longipilis nubila Akodon (Ab.) sanborni Akodon (Ab.) lanosus Akodon (Ab.) mansoensis Akodon (Ab.) hershkovitzi 1 5.37 1 19 5.13 1 16 4.58 2 11 4.18 2 4 3.88 2 27 4.27 2 14 4.07 2 18 4.03 2 4 3.63 2 14 3.48 3 1 3.90 4 * 1 = This paper; 2 = Yanez et al., 1 978; 3 = de Santis & Justo, 1980; 4 = Patterson et al., 1984. the anterior surface of the crown by a shallow groove descending to the alveolus. The anteroloph is barely defined by a very shallow anteroflexus, much as in the type specimens of /I. /. longipilis and A. illutea (fig. 7C-D). As in them, the proto- flexus is wide and moderately infolded to the cen- ter of the tooth, its innermost point reaching a level anterior to the level of the innermost point of the opposed paraflexus. The metaflexus is rather transverse in position, scarcely inclined backward, and it is at a level posterior to the main axis of the hypoflexus, which is wide and slightly oriented forward. The lingual surface of the crown is par- tially broken at the walls of the medial loph (par- aloph + mesoloph), but the presence of a free lingual remnant of the mesoloph is clearly indi- cated, though it cannot be ascertained if a meso- style was also present. In the M^, the mesoloph remnant is evident, but there is no distinguishable mesostyle. The protoflexus is obsolete both in the M^ and in the M'. The M^ is very similar to that in the types of illutea and longipilis, but the para- flexus is better indicated, probably because of less advanced wear, although the metaflexus is less re- entrant than in those specimens. The M' is sub- cylindrical in outline, with traces of the lingual flexi. There is no trace, however, of a metafossetus in this tooth, whereas such an internal enamel is- land is present in the type specimens of A. I. lon- gipilis. A. I. hirtus, A. I. nubilus, and A. illutea, and in all the specimens of the sample of A. I. longipilis from Valparaiso. Even in the most worn M' of ohxTwod Abrothrix, the presence of a metafossetus is constant; its absence in A. kermacki is a diag- nostic feature. The lower molars can be studied in the six avail- able specimens. They are typical Abrothrix lower molars in the bulging of the entoconid and the small, anteriorly directed mesolophid remnant and the elongated and oblique median murid of m, and mj. As it is usual in Akodon s.l., the main cusps are disposed nearly in echelon, and the meta- conid and entoconid are placed at a level anterior to the protoconid and hypoconid, respectively. The procingulum of the m, is wider than is usual in Abrothrix and bears a very shallow, but distin- guishable, anteromedian flexid in five of six avail- able individuals (~83%). In the studied sample of A. I. longipilis from Valparaiso, it was observed only in 25% of the cases, and then only as a very* shallow notch. There is a well-developed antero- labial cingulum, similar to that in other species of Abrothrix, and, as in them, there is no evidence of a division of the protoflexid into an anterior and a posterior portion, although the lateral sur- face of the crown at the procingulum shows a pro- nounced concavity anterior to the anterolabial cin- gulum in some cases. As in other Abrothrix, the mesoflexid is obliquely directed anteriorly from the outside in m, and m,; and the posteroflexid is less oblique, almost completely transverse with more advanced wear. In all the specimens, a well-defined, but weak mesolophid remnant con- nected with a mesostylid is apparent in the m,. It grows out from the anterior border of the ento- lophid, and directs forward and outward, defining a very shallow entoflexid. In the mj, this structure is even weaker, being obsolete in three of six known specimens. This development and disposition of Opposite Page: Fig. 7. Molar teeth (occlusal views) of representative species ofliving species of Akodon (Abrothrix). Upper row, right lower molar series; lower row, left upper molar series of same individuals. A, Akodon (Ab.) longipilis longipilis (Waterhouse); male; BMNH 97.5.1.6; Valparaiso, Chile. A rather young specimen of the sample of 20 individuals, showing internal remnants of the mesoloph in M' and M- in the form of a persisting mesofosettus and a persisting meuflexid in the m,. B, Akodon (Ab.) sanborni Osgood; male; MBUCV 1-2025; Mehuin, Valdivia, southern Chile. Rather young specimen, showing persisting mesofosettus on M'. C, Akodon (Ab.) longipilis longipilis (Waterhouse); holotype, BMNH 55.12.24.177; Coquimbo, Chile. D, Akodon (Ab.) illutea Thomas; female; type specimen, BMNH 28.10.14.2; Aconquija, Tucuman, Argentina. 372 HELDIANA: ZOOLOGY tO 1 .. ZniM REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 373 Table 3. Statistics of the sample of Akodon (Abrothrix) kermacki Reig compared with a sample of Akodon (Ahrothrix) longipilis longipilis from a living population at Valparaiso, Chile. Akodon {Ab.) kermacki. Upper Pliocene and lowermost Pleistocene SE Buenos Akodon (Ab.) longipilis, BMNH sample Aires Prov. (N = 6) from Valparaiso (N = = 19) Variate Range jC s Range i s m,-mj (alveolar) length 5.04-5.69 5.35 0.28 4.80-5.57 5.18 0.21 m,-mj (coronal) length 4.86-5.63 5.19 0.30 4.73-5.31 5.02 0.19 m, length 1.89-2.29 2.02 0.14 1.95-2.23 2.09 0.08 m, width 1.30-1.58 1.38 0.10 1.27-1.39 1.32 0.04 m. length 1.46-1.74 1.60 0.10 1.43-1.62 1.53 0.05 m, width 1.30-1.55 1.38 0.10 1.21-1.39 1.30 0.05 m, length 1.39-1.59 1.50 0.08 1.27-1.49 1.37 0.05 Incisor depth 1.41-1.70 1.50 0.11 1.14-1.39 1.27 0.07 M'-M* (alveolar) length (5.37) 4.67-5.69 5.13 0.24 M'-M' (coronal) length (4.91) 4.35-5.18 4.79 0.20 the mesolophid remnant is also typical of other species oi Abrothrix. As is also the case in other species of this subgenus, the ectolophid is absent in all the observed specimens, and the ectostylid is present in the m, in only one of the six speci- mens, but in none of them was there any evidence of it in the mj. In the observed type specimens of living species of Abrothrix, there is no trace of ectostylid either on m, or mj, and in the studied sample of ^4. /. longipilis from Valparaiso, an ec- tostylid was found in eight of 19 cases in the m,, and in two of 19 cases in the m,. However, no specimens showed any trace of an ectolophid. The ectolophid is usually present in the species oi Ako- don s.s. which are comparable in size to kermacki. Following our records, the presence of an ectolo- phid in the m, has a frequency of 92% in Akodon tolimae (N = 40), 83% in Akodon urichi saturatus (N = 48), 100% in Akodon urichi venezuelensis (N = 27), and 64% in Akodon azarae (N = 58). An ectolophid is completely absent, however, in some small species oi Akodon as A. iniscatus. In the m2 the protoflexid is well defined, though it may be completely eroded by wear (cf. M- 1071). It disappears earlier by wear than in the m,. This tooth varies with wear from sigmoid-shaped to nearly eight-shaped. In one case (M- 1071), the me- soflexid is completely obliterated by advanced wear. In size, the m, is relatively longer as regards the m, than in A. I. longipilis (fig. 7), and from what can be inferred in the type specimens, than in the other forms of the subgenus. As regards metrical differences of A. kermacki in comparison with A. I. longipilis and large-sized species of Akodon s.s.. Figures 7 and 8 and Tables 1 through 4 show the corresponding data. Discussion— From the morphological charac- teristics of molar teeth and mandible, it seems mandatory to place A. kermacki in the subgenus Abrothrix. It matches all studied character-states of this subgenus, and the differences found are merely indicative of a clear-cut distinction at the species level. As regards the metrical analysis, the evidence is conclusive for statistical differences in some, but not in all, the studied variables, as ex-« pected in this kind of analysis. Metrically, ker- macki is quite distinctive in the depth of the in- cisor, the relative shortness of the m,, and the relatively longer mj. These two differences are complementary, and the result is that the mean crown length of the molar row is not statistically different from living^. /. longipilis. This illustrates the care which must be taken in inferring system- atic kinship or distance from overall statistical dif- ferences. Combining the results of the morpho- logical studies and the metrical analysis, there can be little doubt that we are dealing with a distinctive species of Akodon (Abrothrix). The fossil species is more similar in size to the living A. (Ab.) longipilis and A. (Ab.) illutea than to A. (Ab.) sanborni, A. (Ab.) lanosus, the dubious A. (Ab.) mansoensis, A. (Ab.) xanthorhinus, and A. (Ab.) hershkovitzi (for the assignment of the two last species to Abrothrix. see above). An exami- nation of Table 2 clearly supports this assertion. The fossil species is slightly larger than the largest living subspecies of A. (Ab.) longipilis, namely, A. (Ab.) I. longipilis (tables 2-4), and it belongs, along with A. (Ak.) varius simulator, A. (Ak.) urichi sa- turatus, and A. (Ab.) I. longipilis (table 5, fig. 8), to the group of the largest species of Akodon s.l. Within the limits of this genus, its size is only surpassed by the new species described next. In the larger size of the incisor and the correlated 374 FIELDIANA: ZOOLOGY X> M U 4.a *A 4.« 4.1 ».♦ I 1 1 1 1 1 < 1 1 1 — t" — I 1 V . V . V A. azarae. F.zeiza . N= 94 A. cursor, Misiones. N- 20 A. uric hi vcnezuelensis, Avila- N- 39 A.urichi venezuelensis, Orinnte. N= 17 A. urichi saturatus, S. Venezuela. N«47 A. varius simulator, Tucuma'n. A. kermacki, Chapadmalal and Bca. Lobos Ftions. N« 40 N«6 A. longipilis, Valparaiso. N» 19 L E N G T H Mj.-Mg. fC R O \>' N) Fig. 8. Dice-grams of the variation in the length of the lower molar row (coronal) in various species of medium- and small-sized Akodon {Akodon) and Akodon (Abrothrix). The diagram shows the mean, the range, two standard errors to each side of the mean (black squares), and one standard deviation to each side of the mean (open squares + black squares). greater development of the capsular projection of the base of the incisor, kermacki is more highly modified than the living species. If this is indic- ative of evolutionary divergence, kermacki could not be the ancestor of any living species o{ Abro- thrix. This conclusion, however, is based on frail evidence, and more specimens and further study of other characters are necessary to evaluate its evolutionary significance. In any case, Abrothrix in the Upper Pliocene and lowermost Pleistocene of SE Buenos Aires Province occurred almost 1,000 km eastward of the present distribution of this subgenus. Actually, the living representatives of Abrothrix are now limited to the lowlands and low valleys of central and southern Chile and the eastern Andean slopes of Mendoza and Patagonia, to Tierra del Fuego, plus ihe isolated A. illutea of Tucuman. The dif- ferent distributions of living and fossil Abrothrix indicate a reduction of the range of this subgenus from a much more extended area to its present limits, a phenomenon which could have been caused by the climatic changes that occurred dur- ing the Pleistocene. The case of Abrothrix is not isolated, and other mammals presently restricted to Chilean or Andean distributions were also pres- ent in the Upper Pliocene of Buenos Aires Prov- ince. One is the fossil caviomorph Pithanotomys, which is hardly separable from the living Acon- aemys, restricted now to a few isolated populations in southern and central Chile (Osgood, 1 943) and high valleys of Mendoza and Neuquen, Argentina (Pearson, 1984), but abundant in the fossil de- Table 4. Student's / test for differences between means in six selected variates o{ Akodon {Abrothrix) lon- gipilis and Akodon (Ab.) kermacki (from data in tables 1 and 3). A. (Ab.) longipilis-A. (Ab.) kermacki (df = 23) Variate t P m,-m3 (alveolar) length 1.62 >0.05 m|-m, (coronal) length 1.59 >0.05 m, length 1.36 >0.05 mj length 1.91 <0.05 m, length 3.94 <0.001 Incisor depth 5.41 <0.001 REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 375 L E N G T H M M. X O Z 2.1 A. loni A. magnut A.kcrmcicki -• 1— « 1 ' 1- 1.3 1.4 1J 1.6 WIDTH ■ A.(Abrothrix) 1. longipilis ▼ A.CAbrothrix) kermacki • A.(Abrothrix) magnus Fig. 9. Scattergrams of measurements of difTerent teeth in living and fossil species of Akodon (Abrothrix). 376 FIELDIANA: ZOOLOGY posits of the Pliocene and Lx)wer Pleistocene of Buenos Aires Province. The marsupial subfamily Caenolestinae, which is now restricted in southern South America to the south temperate forests of southern Chile and possibly Argentina, is also rep- resented by the fossil genus Pliolestes in the Plio- cene of Buenos Aires Province (Reig, 1955). Also the living genus Abrocoma, now restricted to the Andean region, had representatives in the Pliocene of Buenos Aires Province, and a specimen in the collection of the Museum of Mar del Plata (MMP- 1059) from the Atlantic cliffs of the Chapadmalal region dates from the early Pleistocene. Akodon (Abrothiix) magnus, n. sp. HOLOTYPE-MMP M-551 (figs. 5D, G-H,; 6E): left lower jaw with incisor, m, and mj, lacking mj and condyloid, coronoid and angular processes; right lower jaw with incisor and m,, lacking mj and mj, coronoid and condyloid processes; left femur broken in the middle of the shaft; right cal- caneum (the postcranial bones are only tentatively associated with the mandibles). Found by G. J. Scaglia in Vorohue Formation (Kraglievich, 1952), at the Atlantic slopes of the Chapadmalal region near Baliza San Andres, Partido de General Puey- rredon, SE of Buenos Aires Province, Argentina. These sp)ecimens were found associated with MMP M-869, the holotype of Cholomys pearsoni Reig, 1980, and MMP M-897 and M-868, the last two belonging to a new species of the subgenus Akodon described below. Hypodigm— The holotype and MMP S-407: right fragmentary lower jaw of an old individual, bearing the incisor and the three molar teeth, the latter greatly worn. Found by G. J. Scaglia in Vo- ^ rohue Formation, at the sector of Atlantic slopes stretching from south of Arroyo Loberia, Cha- padmalal region, Partido de General Pueyrredon, SE of Buenos Aires Province, Argentina. Diagnosis— A very large species of Akodon (Abrothrix), exceeding in size A. (Ab.) kermacki; mandible slender; incisor relatively much weaker; m, with a distinct metastylid. Known Distribution— Vorohuean subage of the Uquian age (Lower Pleistocene), SE of Buenos Aires Province, Argentina (see Marshall et al., 1984). Description— The description is based on the holotype, as specimen MMP S-407 is only ten- tatively included in the species. The mandible is distinctly larger than that of A. kermacki, but more slender. The diastema is, how- ever, a little shorter, its length being less than that of the two first molars. The horizontral ramus is shorter, with a height less than the diastema length, and its lower border bends slightly upward from the level of the anterior part of the mj. The border of the ramus immediately in front of the m, is as in A. kermacki, but the upper border of the dia- stema is slightly more concave. The symphysis, while shorter, is lower and more slender. The low- er masseteric crest is somewhat higher in position than in A. kermacki. The mental foramen is as in kermacki. The anterior border of the coro- noid process is only partially preserved, but it is clear that it originates further posteriorly than in kermacki, at the level of the middle of the al- veolus of m,. The condyloid process is not pre- served, but the remaining parts of the ascending ramus show a great deal of the sigmoid notch and suggest that the condyle was rather low and well projected backward. The capsular projection has broken walls in the two rami of the holotype, and it is slightly less pronounced than in kermacki, the root of the incisor lying further forward, be- tween the coronoid process and the beginning of the sigmoid notch. The incisor is much weaker than in kermacki, and its proportions are as normal in normal A. longipilis. Its absolute size in depth is less than in kermacki, even when magnus shows greater values for all the remaining measurements of the denti- tion. The m, and mj are very similar to those of kermacki, the main distinction being absolute size. However, the m, clearly shows a metastylid, a character which has not been observed in any oth- er specimen of Abrothrix examined by me, and which may be considered diagnostic. However, a larger sample is necessary to evaluate the con- stancy of this character-state. As is typical of ker- macki and other Abrothrix, the mesolophid rem- nant is weak and grows out from the middle of the entolophid in the m,, and the entoconid makes a noticeable bulge on the lingual face of the tooth. In the mj the mesolophid remnant is rudimentary, but the protoflexid is better marked than in spec- imens of kermacki. The roots of the mj show that this tooth was large, probably relatively larger than in kermacki. This is confirmed by the very worn mj of specimen MMP S-407. Discussion— Although the main distinction of A. magnus from A. kermacki is one of size, the slender mandible and weaker incisor confirm that they are different species. The difference in abso- REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 377 -if 378 FIELDIANA: ZOOLOGY lute size is obvious at first sight (figs. 21-22), but one is tempted to wonder whether this specimen might not be an extremely large individual within the size range of ^. kermacki (cf. fig. 21). One specimen oi A. kermacki (MMP M-1067) fi'om the Chapadmalal Formation has a first molar near- ly as long as that of specimen MMP S-407, at- tributed to A. magnus. The Chapadmalal speci- men, however, has a much shorter mj, and agrees with kermacki in the mandible and relative size of the incisor (table 1). Another specimen oi ker- macki from the Chapadmalal Formation (MMP M- 1 1 54) has an mj which approaches the size of the m2 in specimen MMP S-407, but its m, is much shorter and it also agrees with kermacki in incisor and mandible characters (table 1). In any case, there is no overlap in absolute size, and the specific distinction seems to be validated by the sum of all studied characters. It could be alleged, however, that the transitions in size between organs in sjjecimens of the two species might be a reflection of a real transition between the two taxa, and that a process of spe- ciation by gradual transformation is involved here. Such alleged cases of phyletic speciation have been described in the echimyid rodent Eumysops from the same sequence (Kraglievich, 1965) and sur- mised in the case of the didelphoids Thylatheri- dium (Reig, 1959) and Sparassocynus (Reig & Simpson, 1972), also from the same sediments. Although this possible interpretation can only be substantiated by more material, I believe that the greater development of the incisor in kermacki precludes the possibility that this species is the direct ancestor of magnus. Moreover, the single specimen of kermacki known from the interme- diate Barranca Lobos Formation (MMP M- 1 07 1 ), does not show intermediate character-states, but is closer to the holotype of A. kermacki in size than the other specimens referred to the same SF)ecies. In any case, even if it were demonstrated that there is a direct phyletic link between the two species, there is still enough evidence to maintain A. magnus as a well-distinguished species; its size differences from kermacki are not compatible with the known range of size variation in species of the subgenus Abrothrix (figs. 8-9). Akodon (Akodon) johannis, n. sp. Holotype- MMP M-742 (fig. 10F,H,J): right lower jaw with incisor and m|-m2, lacking the cor- onoid and angular processes and the mj; left lower jaw with m,-m2, with incisor broken, and lacking mj and the same processes; left maxilla with M '- M^; portion of right maxilla with M'-M^; the two tibiae, the right incomplete; incomplete right and left femora; right humerus and cubitus; portions of scapula and of left pelvis; two vertebrae. Found by G. J. Scaglia in Miramar Formation (Kragliev- ich, 1952) at the Atlantic slopes south of "Bajada de San Andres," Chapadmalal region, Partido de General Pueyrredon, SE of Buenos Aires Province, Argentina. Etymology— The species name, johannis, is given for Juan Brkljacic, a close collaborator of G. J. Scaglia in the latter's work at the Museum of Mar del Plata. Brkljacic has been responsible for a great deal of progress by that institution in the recent past. Hypodigm— The holotype is the only known specimen. Diagnosis— A small species of Akodon {Ako- don) the size of A. andinus or A. nigrita; moder- ately strong mandible with a low symphysis, a relatively deep incisor, and a fairly well-developed capsular projection. Incisive foramina almost level Opposite Page: Fig. 10. Lower jaws, maxillae, and molar teeth of fossil and living species of small Akodon (Akodon). A, Later view of left lower jaw of living A. (Ak.) cursor montensis Thomas; female; BMNH 1874; Puerto Gisela, Misiones, Argentina. B, Lateral view of left lower jaw of /I. (Ak.) cf cursor (Winge); MLP 66. VII. 27. 95 (a); Miramar Formation (Ensenadan Stage), vicinity of Camet, Partido de Mar Chiquita, SE of Buenos Aires Province, Argentina (Middle Pleistocene). C, Crown view of left m, of A. {Ak.) cf cursor (Winge); MLP 66. VII. 27. 95. D, Crown view of left m, of living A. {Ak.) cursor montensis Thomas; type specimen, BMNH 4.1.5.3.36; Sapucay, Paraguay. E, Lateral view of right lower jaw of living A. nigrita Lichtenstein; male; BMNH 3.7.1.74; Ro9a Nova, Parana, Brazil. F, Lateral view of right lower jaw of ^4. {Ak.) johannis n. sp.; holotype, MMP M-742; Miramar Formation (Ensenadan stage), Chapadmalal region, Partido de General Pueyrredon, Buenos Aires Province, Argentina (Middle Pleistocene). G, Crown view of right lower m, and mj of A. {Ak.) johannis n. sp.; holotype, MMP M-742. H, Lateral view of left maxilla of ^. {Ak.) johannis n. sp.; holotype, MMP M-742. 1, Lateral view of left maxilla of y4. {Ak.) nigrita, BMNH 3.7.1.74. J, Palatal view of left and right maxillae of A. {Ak.) johannis n. sp.; holotype, MMP M-742. K, Crown view of upper molar series of A. {Ak.) johannis n. sp.; holotype, MMP M-742. REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 379 Table 5. Statistics of the coronal length of m,-mj in 10 samples of 10 species and subspecies of the genus Akodon. Species, subspecies, and locality Statistic N Range SD SE A. puer (Peru and Bolivia) A. iniscatus (Chubut) A. azarae (Ezeiza, Buenos Aires) A. cursor (Misiones, Argentina) A. urichi venezuelensis (Avila) A. urichi venezuelensis (Oriente) A. urichi saturatus (Tepuyes) A. varius simulator (Tucuman) A. I. longipilis (Valparaiso) A. kermacki (Pliocene, Buenos Aires) 14 20 94 20 39 17 47 40 19 6 3.58-^.16 3.39-4.22 3.84-4.60 4.23-4.73 4.54-5.02 4.54-5.21 4.80-5.50 4.38-5.39 4.73-5.31 4.86-5.63 3.76 3.97 4.24 4.40 4.74 4.82 5.06 5.01 5.01 5.19 0.176 0.186 0.154 0.138 0.128 0.179 0.135 0.167 0.186 0.299 0.040 0.041 0.014 0.031 0.019 0.043 0.019 0.023 0.042 0.118 X = arithmetic mean; SD = standard deviation; SE = standard error. with the anterior borders of M'. Zygomatic plate strong, wider than the length of the M', with a rounded and slightly forward-projecting anterior border. Molars relatively broad. M' with a wide and oblique procingulum showing a moderately developed anteromedian flexus, but without an- teroflexus. Lower molars without mesolophid remnants and mesostylids; ectolophids and ecto- stylids also absent. Known Distribution— Ensenadan age. Middle Pleistocene of SE of Buenos Aires Province, Ar- gentina. Description— Of the skull, only the maxilla and the middle palatal region can be studied. The pal- ate is long and wide; the space between the internal borders of the crowns of the M' is greater than the length of the M'. There is no direct evidence of the position of the posterior border of the palate, but the maxillary bone surroimding the M' clearly indicates that the border was slightly behind the posterior border of those molars. The posterior limits of the incisive foramina (fig. lOJ) are clearly indicated in the two portions of maxillae. The fo- ramina scarcely surpass the anterior border of the M', and they are even less expanded behind than in Akodon nigrita, a living species with rather short incisive foramina. In fact, the posterior position of these foramina resembles that in Notiomys and Microxus (in which they scarcely surpass the an- terior border of the M') more than the usual con- dition in Akodon s.s.. in which they usually reach the level of the protocone of the M'. In the zy- gomatic plate, however, johannis stands quite apart from Notiomys and Microxus, and shows an un- usually strong and wide plate, with an anteropos- terior length greater than the length of the M', as is also the case in A. nigrita and A. andinus. In most other species of small Akodon, the length of the M' either exceeds the anteropKJSterior diam- eter of the plate, as is the case in A. puer, A. bo- liviensis, and A. azarae, or the two measurements are roughly equivalent, as is the case in A. inis- catus. The anterior border of the zygomatic plate is quite upright, and it slightly projects forward at its rounded upper comer, and it is not sharply cut off above, the upper comer being very slightly turned into the anterior border. The mandible is also characterized by its low symphysis and the very procumbent incisor. This * is reflected in the anterior median point of the diastema, which is well below the level of the al- veolar row, even more so than in A. cursor (fig. lOA), a species with a particularly low symphysis. In A. (Deltamys) kempi and in A. nigrita (fig. lOE) and A. andinus (fig. IID), the symphysis is also low, but less markedly so than that in A. johannis; in most of the other species of Akodon s.l., the symphysis is more upturned and, consequently, the incisor is less procumbent. The mandibular ramus is relatively deep: although the depth of the ramus at the m, is less than the diastema length, the ramus is higher than in species of similar size, such as A. nigrita, A. iniscatus, and A. puer, and it is longer than the combined mi-m, length. The lower masseteric crest is high and moderately marked, more clearly so than the upper masseteric crest, and the two crests reach forward nearly to the anterior border of the m,. The tip of the cor- onoid process is broken in the two mandibles, but its anterior border is partially preserved, and it slopes backward somewhat abmptly. The condyle is well posterior and fairly high in position, and the capsular projection is well develop)ed, as com- pared to usual Akodon s.s. The incisor is comparatively strong, markedly more so than in nigrita, aruiinus, and puer, and it 380 FIELDIANA: ZOOLOGY Table 6. Measurements of lower jaw and lower molars of fossil and living specimens of Akodon (Akodon) cursor. Akodon cf. cursor, Miramar Akodon c. montensis rormanon Female, BMNH 66-1874, Misiones, . Sample of Akodon c. mc MLP 62- VII- 27-95 ntensis MLP 62-VII- MLP 62-VII- Holotype, BMNH in BMNH from Puerto Gisella, Misiones, Argentina Variate 27-95 (a) 27-95 (b) (c) 4.1. 5 J5 Argentina Range i s Mandibular condyle- anterior border diastema length 14.20 13.60 14.12 14.41 Condyle-anterior border m, length 11.39 11.39 10.99 Diastema length 3.20 2.94 3.32 3.46 ... Condyle-posterior border alveolar mj length 6.46 5.82 6.21 Mandible depth at m, 3.26 3.26 3.27 3.07 2.56-3.64 3.01 0.36 m,-mj (alveolar) length 4.67 4.74 4.54 4.55 4.61 4.35-4.93 4.65 0.18 m, length 1.86 1.80 1.77 1.71 1.75-1.90 1.82 0.04 m, width 1.18 1.02 0.99 1.08 1.02-1.24 1.11 0.06 m, length 1.39 1.43 1.37 1.33-1.49 1.40 0.05 m, width 1.03 0.99 1.05 0.99-1.22 1.09 0.07 mj length 1.24 1.24 1.22-1.49 1.31 0.08 mj width 0.93 0.99 0.87-0.99 0.94 0.04 Lower incisor depth 1.19 1.05 1.08 1.04 is comparable in relative depth to the incisor of iniscatus (fig. 11 A). As already indicated, the in- cisor is characteristically procumbent. The molar teeth (fig. 10G,K) are broad and rather short, and they look similar to, but are slightly more heavily built than, the molars of nigrita. They differ from molars of that species, however, in the lack of any indication of remnants of mesolophids, ectolo- phids, and ectostylids in the m, and mj; the upper molars of the two species are more closely com- parable in morphology and proportions and in the advanced reduction of the mj. Both in the upper and in the lower series, the posterior border of the first molars is partially cut off, as is, although less markedly, the anterior border of the second mo- lars. This is probably an individual anomaly. Discussion— .4 /ccx/ow johannis appears to be clearly distinctive from the living small-sized species oi Akodon s.l. It is obviously distinct from the contemporary^, cf. cursor, which is described next. Among living species, it seems to be more closely related to A. nigrita than to any other com- pared species. Doubtless it should be allocated in the subgenus Akodon s.s., and it probably repre- sents an extinct lineage among the extensive di- versification of the subgenus. The dubious Necro- mys conifer Ameghino, which Ameghino (1889) mentions as being represented in the coetaneous Ensenadan stage of the north of Buenos Aires Province, does not seem to have anything in com- mon with A. johannis. Although the illustrations and the description are obscure, the drawings giv- en by Ameghino (1889, Atlas, table IV, figs. 17- 1 8) show a mandible with an upturned symphysis and a nonprocumbent incisor. Hershkovitz ( 1 962) considers Necromys a mere synonym of Calomys, a contention that does not seem warranted by Ameghinos's data, even given the faulty nature of the illustrations. Akodon (Akodon) cf. cursor (Winge) Referred Specimens- MLP 62.VII.27.95 (a) (fig. lOB-C): left lower with incisor and m, of an old individual; MLP 62.VII.27.95 (b): right lower jaw with the incisor and the alveoli of the molar teeth; MLP 62.VII.27.95 (c): fragment of right lower jaw with extremely worn m, and m2. These three specimens were found in association with each other and with remains ofReithrodon auritus, Nectomys squamipes, and Ctenomys sp. in a bone conglomerate probably representing fossil owl pel- lets. The bone conglomerate was extracted from a rocky block from the Atlantic cliffs 5 km N of Colonia Camet (about 1 5 km N of the city of Mar del Plata), Partido de Mar Chiquita, SE of Buenos Aires Province, Argentina. The cliffs at this point REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 381 expose sediments of the Miramar Formation (Kraglievich, 1952), and this geological prove- nance can be ascertained for the fossils found in the fallen block. Description— The morphology and measure- ments (table 6) of the mandibles indicate that these represent Akodon closely allied, if not identical, to the living species A. (Ak.) cursor (Winge). As in the latter species, the mandible is elongate and slender relative to other species of intermediate- sized Akodon. The symphysis is elongated and low, the anterior median point of the diastema being at a level below that of the alveolar row. The length of the diastema is as long as m, and mj, and the depth of the ramus below m, is slightly greater than the diastema length. The lower border of the ramus is gently concave behind the m,. The lower masseteric ridge is relatively well marked and rath- er high in position, the upper masseteric ridge being scarcely noticeable and somewhat parallel to the alveolar border. The coronoid process is low, and its anterior border slopes upward very gently. The condyloid process is also low and projected well backward. There is a fairly well-developed cap- sular projection of the base of the incisor, slightly stronger than usual in living specimens of cursor examined, but the difference is not really very marked. The incisor is, as in cursor, well devel- oped, and its depth is a little greater than that of the holotype and other specimens examined of /I. cursor montensis. The molar teeth are too worn to show many details of structure. However, the m , of specimen (a) is slightly less worn (fig. 1 OC) and shows a clear indication of an anteromedian flexid and an over- all shape and development of the procingulum which matches perfectly with the procingulum of cursor. A well-marked anteroflexid is present in the holotype of A. cursor montensis (fig. lOD) and in each of 15 animals from Puerto Gisela, Mi- siones, in the British Museum referred to at sub- species. The shape of the enamel walls at the hy- poflexid suggests that an ectostylid was present. This element is absent in the holotype of /I. c. montensis, but is present in 7 of 1 4 specimens of the above-mentioned sample. From the enamel wall of the mesoflexid, no mesolophid remnant is evident in this fossil specimen. This structure is present in 80% of the modem sample, but it is almost completely absent in the type of /I. c. mon- tensis. In length and width, the m, falls within the limits of variation in a sample of living cursor (table 6), and there is also a complete correspon- dence in the length of the molar series between the fossil specimens and the living sample. Discussion— Apart from Akodon cursor, only A. azarae, now living in the same locality where the fossils were found (Reig, 1 964), is a plausible relative. Akodon azarae agrees with the fossil spec- imens in being a medium-sized species with a rath- er elongated mandible. It is, however, significantly smaller than the fossil specimens (fig. 1 2) and has a stronger mandible. A sample of 55 A. azarae from Ezeiza, close to Buenos Aires, now in the Museum of Mar del Plata, shows an alveolar length of the lower series significantly smaller (f < 0.001) than the studied sample of A. cursor from Puerto Gisela (fig. 8). The alveolar molar length of the three fossil specimens here described have the same mean value as the A. cursor sample. Moreover, azarae is characterized by narrower molar teeth, as it is evident for the m, in the diagram of Figure 1 2. Thus, a close relationship of the fossil speci- mens with azarae must be ruled out. The mor- phological resemblance and the agreement in size with cursor is such that the fossils from Camet represent a form probably conspecific with living cursor. Akodon cursor was first described by Winge (1887) from living and subfossil specimens from Lagoa Santa, Minas Gerais, Brazil, as a member of Habrothrix (Thomas [1884] had placed under Habrothrix, a misspelling of Abrothrix Water-* house, all Akodon-\\)/it mice from South America). Afterward Thomas (1902) placed cursor in Ako- don, and he later (1913) erected A. arviculoides montensis, which he compared with cursor. As already discussed (pages 354, 356) arviculoides Wagner is not an Akodon, but a synonym of Bolo- mys lasiurus (Lund); montensis Thomas is most likely a subsp)ecies of cursor Winge, as proixjsed by Ximenez and Langguth (1970). Thus, two subspe- cies of cursor may be recognized: A. c. cursor Winge, which following Vieira (1955), extends over Minas Gerais, Espirito Santo, Guanabara (Rio de Janeiro), Sao Paulo, and Parana; and A. c. montensis, which is known from Paraguay (Thomas, 1913), Misiones (Massoia & Fomes, 1962), and central Uruguay (Ximenez & Langguth, 1 970). However, these sub- species are not well defined, and our fossil sample is too small and fragmentary to attempt a compar- ison with living subspecies or to place it in a sub- species of its own. The present evidence indicates only that they represent Akodon cf cursor and that this species extended its range in Middle Pleisto- cene times at least 600 km south of the known limits of its present distribution. Akodon cf cursor from the Miramar Formation cannot be confused with Akodon johannis, found also in the same strata. The latter is much smaller, 382 HELDIANA: ZOOLOGY has a relatively stronger mandible and a relatively deeper incisor, a more upturned and higher cor- onoid process, and a much lower symphysis. Its m, is proportionally much wider, and its procin- gulum lacks a well-defined anteromedian flexid. The differences between these two species are as great as the differences between living A. cursor and A. nigrita in areas of sympatry. Akodon (Akodon) lorenzinii, n. sp. HoLOTYPE-MMP M-1081 (figs. 1 II, 13H): the two lower jaws with incisors and molar teeth. Found by Mr. S. Lorenzini in the Atlantic slopes 5 km N of the city of Miramar customarily known as "Barranca Parodi," Partido de General Alvara- do, SE of Buenos Aires Province, Argentina. The fossils were found in strata of San Andres For- mation (Kraglievich, 1952; see also Marshall et al., 1 984), as confirmed in the field by G. J. Scaglia and the late J. Zetti. Hypodigm— The holotype; and MMP M-867 (figs. 11 G; 1 3G): incomplete right maxilla with m , , broken rcvj, and m,, and partially broken zygo- matic plate. Found by G. J. Scaglia in the Vorohue Formation (Kraglievich, 1952; Marshall et al., 1984) at the Atlantic slopes of the Chapadmalal region, close to "Baliza San Andres," Partido de General Pueyrredon, SE of Buenos Aires Province, Argentina. Found in association with next speci- men and with MMP M-551, holotype o^ Akodon (Abrothrix) magnus (see above), and MMP M-869, holotype of Cholomys pearsoni Reig (see Reig, 1980). MMP M-868 (fig. 1 IJ): left lower jaw with in- cisor and all molar teeth, partially lacking coro- noid, condyloid, and angular processes. Found in association with MMP M-867. M-867 and M-868 may belong to the same individual, but the lower molar teeth look more worn than the upper ones. Therefore, I prefer to treat them as belonging to two different individuals. MLP 52-X-4-44 (a) (fig. IIM): most of right lower jaw broken in front of the middle of the symphysis, and at the posterior processes, with a broken incisor and all molar teeth, the second mo- lar partially broken; fragment of right maxilla in- cluding the first molar and the posterior half of the zygomatic plate; left femur; portion of left tib- ia; right upper incisor. Found by the late J. Fren- guelli in association with MLP 52-X-4-44 (b), a specimen referred to Scapteromys sp., in beds of San Andres Formation (= "Prebelgranense" in Frenguelli's stratigraphic nomenclature), in the Atlantic slopes extending S of Punta Hermengo, Miramar, Partido de General Alvarado, SE of Bue- nos Aires Province, Ai^entina. Etymology— The species is named for Mr. Sil- vio Lorenzini, discoverer of the type specimen and other remarkable fossil cricetids, and an active collaborator of the Museum of Mar del Plata. Diagnosis— A small species oi Akodon the size of A. puer. Mandible relatively short and high. Zygomatic plate moderately wide. Incisive foram- ina well behind the anterior border of first molar, but not reaching their protocones. M' with a well-marked anteromedian flexus, a shallow an- teroflexus, and a projecting, narrow mesoloph remnant united to a mesostyle. Lower molars rel- atively narrow, without indication of mesolophid remnants, mesostylids, or ectolophids, and with somewhat oblique entolophids and posterolo- phids. Procingulum of m, narrow, with a shallow anteromedian flexid and a well-marked metaflex- id. mj relatively small. Known Distribution— Vorohuean and San Andresian subages. Lower Pleistocene (see Mar- shall et al., 1984), SE of Buenos Aires Province, Argentina. Description— The skull fragments of speci- mens MMP M-867 and MLP 52-X-4-44 (a) afford only a few indications of the maxillary and palatal region. They show a rather robust zygomatic plate, probably wider than in A. iniscatus and as wide as in A. puer. The anterior border of the zygomatic plate is not preserved in either specimen, but its lower limit can be observed in MMP M-867, sug- gesting that it is slightly wider than the length of the M'. The incisive foramina (fig. 1 IG) are ex- panded backward far more than in johannis, and they extend to the middle of the protoflexid of the M', not reaching to the protocone. In this respect, lorenzinii resembles iniscatus and andinus more than puer, in which the incisive foramina extend slightly behind the protocone of the M' in all 20 individuals I examined from different localities. The mandible (fig. 1 1 1-J) is much shorter than in puer, iniscatus, andinus, or johannis. It differs markedly from the slender and elongated man- dible oi puer (fig. 1 IC), and in proportions it is closer to the mandible of iniscatus, although no- ticeably smaller. The symphysis is fairly upturned, as the middle anterior point of the diastema reach- es the level of the molar alveolar rows, differing from puer, and obviously from johannis, to ap- proach more closely the condition found in inis- catus. The masseteric crests are located midway up the sides of the ramus, as in iniscatus, whereas in puer and andinus, they are placed higher. They REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI ^3 384 FIELDIANA: ZOOLOGY are less marked than in iniscatus, and the lower one is smooth, although stronger than the upper one. The depth of the mandible at the m, is greater than the diastema length, but it is less than the combined length of m,-m2. In iniscatus, the depth of the mandible is less than both the diastema length and the combined length of the first two molars, whereas in puer the diastema is longer than the depth of the mandible. The coronoid process slopes rather abruptly backward, more or less as in iniscatus, and more so than in puer. The process itself is short and low, so that the condyle is at a level higher than the tip of the coronoid process. The condyloid process is high, and it is not mark- edly projected backward. The capsular projection is moderately developed, but it is stronger than in puer and andinus, even a little stronger than in iniscatus. The incisor is comparatively deep, clearly more so than in puer, and it is also slightly deeper than in iniscatus. Its depth equals or exceeds the length of the m3 (fig. 14). The upper molars of MMP M-867 show little wear, while the M' of MLP 52-X-4-44 (a) is mod- erately worn. They are very similar in morphology to upper molars of both puer and iniscatus. The only significant difference lies in the mesoloph remnant which unites to the mesostyle in the two available specimens and is projected further lat- erally than in puer or iniscatus. In this respect, lorenzinii resembles more closely A. andinus, but differs from it in the stronger procingulum of M', which has a well-marked anteromedian flexus and a less projecting parastyle. The M^ is broken in its external half in the only specimens that show this organ (MMP M-867), but it is evident that it was relatively narrow, as in puer and iniscatus. The lower molars are more distinctive in show- ing a little marked anteromedian flexid and a rather narrow procingulum in the m , , without indi- cation of a protostylid. There is, however, a well- developed anterolabial cingulum, but it does not contribute to the shape of the crown enamel pat- tern of the procingulum, as in other species. The metaflexid is also more re-entrant than in puer and iniscatus, but as in them, there is no trace of a mesolophid remnant or a mesostylid, and the sim- ple entolophid is rather oblique in position, es- pecially in the m,. The posterolophid is even more oblique, so that the posteroflexid is noticeably wide. A protoflexid is well marked in the three lower molars, and the m,, as in puer, is relatively small and has a sigmoid shape. No trace of ectolophid is shown in any of the three lower molars, but a tiny ectostylid is observed in the m2 of the type specimen. Discussion— /4/cot/o/i lorenzinii is a very small species of Akodon showing a distinctive combi- nation of characters. It seems to be more closely related to ^. puer and A. iniscatus than to any other species of the subgenus Akodon, and the balance of similarities would favor a closer relationship with iniscatus. This is also expected on biogeo- graphic grounds. In fact, puer is a widespread species, but is restricted to the Andean and pam- pean mountains from Peru to Tucuman in Argen- tina. Thomas ( 1 902) originally described puer from specimens of Chaquecamata in west-central Bo- livia, and he subsequently identified as puer spec- imens I examined from south and central Peru. I Opposite Page: Fig. 1 1 . Lower jaws and maxillae o^ Akodon {Akodon) lorenzinii n. sp., and Akodon {Akodon) cf. iniscatus Thomas and of related living Akodon species. A, External view of right mandible oi A. iniscatus Thomas; female; holotype, BMNH 3.7.9.64; Valle del Lago Blanco, Chubut, Argentina. B, External view of right mandible of A. albiventer (Thomas); male; BMNH 21.1 1.1.51; Sierra de Zenla, Jujuy, Argentina. C, External view of right mandible o^ A. puer Thomas; female; holotype, BMNH 2.1.1.78; Choquecamate, Bolivia. D, External view of right lower mandible oi A. andinus (Philippi); female; holotype oi A. gossei Thomas, BMNH 98.3.21.5; Puente del Inca, Mendoza, Argentina. E, External view of fragment of left maxilla with M' of ^. lorenzinii n. sp.; MLP 52.X.4.44 (a); San Andres Formation, Miramar, Buenos Aires Province, Argentina (San Andresian, uppermost Lower Pleistocene). F, External view of incomplete left maxilla with molar teeth of A. lorenzinii n. sp.; MMP M-867; Vorohue Formation, Chapadmalal region, Partido de General Pueyrredon, SE of Buenos Aires Province, Argentina/Vorohuean (Lower Pleistocene). G, Reconstructed palatal view of ^. lorenzinii n. sp., based on MMP M-867 (the right half is an inverted drawing of the original left half). H, Palatal view of the skull of A. iniscatus Thomas; male; BMNH 13. 11. 1. 5; Pampa Central, Argentina. I, External view of left mandible of A. lorenzinii n. sp.; type specimen, MMP M-1081; San Andres Formation, Barranca Parodi, Miramar, Buenos Aires Province, Argentina; San Andresian (Lower Pleistocene). J, External view of left mandible of A. lorenzinii n. sp.; MMP M-868; found in association with MMP M-867. K, External view of right mandible of A. cf iniscatus: MMP S-640; Vorohue Formation, south of Arroyo Loberia, Chapadmalal region. L, External view of right mandible of a. iniscatus Thomas; male; BMNH 13.1 1.1.5. M, External view of incomplete mandible of A. lorenzinii n. sp.; MLP 52.X.4.44 (a). REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 385 o M I a c i \ I c c I d V) a ■J? ** » ,- >< u T3 •h "o C CA C « O c o E o (O o i-i CO eg n £ «5 3 ■•3 6 3 .2 c 00 c I. c E ? ^ s 2 •5 J= e" g't •S rt « 4> *-• fl) O XJ 00 > u _ ^EEEii^«^«^«22«^«j«^ I E" E' E" e" e" E E E E ^ S S S S S S 5 S j= j3 JZ 00 c !> a ^_^ u T5 7) CO c 8 o 386 FIELDIANA: ZOOLOGY 4.0 4.1 4.2 I 1 I I L_ 4.3 4.4 4.5 4.6 4.7 4.8 4.9 _J 1 I I ■ I 1 L_ N»SS A. (Akodon) azarae N=15 A. (Akodon) cursor li«» A. (Akodon) cf. cursor LENGTH M1-M3 CA L V/) ^^■ 11- III U • A. asaras ^ A. cursor ■ A. cf. cursor a9 10 1.1 u WIDTH OF M]. Fig. 12. Dice-grams and scattergram for measurements of dentition of species of Akodon (Akodon). He later (Thomas, 1918) described caenosus from the mountains at Leon, Jujuy Province, in north- west Argentina, as a subspecies ofpuer. However, he later (Thomas, 1 920) recorded more specimens from San Salvador de Jujuy, proposing specific status for caenosus. The extension of its range to Tucuman was recorded by Thomas (1926b) and Barquez et al. (1980). On examination of the cor- responding types and of all the specimens referred to puer and caenosus in the British Museum, I could not find any reasonable basis to accept spe- cific or even subspecific recognition for caenosus, and I treat it as a junior synonym of puer (see also Vitullo et al., 1986). Akodon iniscatus is based on an animal caught in the Andean region of Patagonia, from southwest of Chubut Province, but Thomas referred to it specimens from northern Patagonia to central La Pampa Province, believing (Thomas, 1 9 1 9, p. 205) that it extended to the south of SE of Buenos Aires Province. Akodon iniscatus colli nus was described (Thomas, 1919, p. 206) as a subspecies from northwestern Patagonia, and A. nucus, described as a full species (Thomas, 1 926a) from specimens of western Neuquen and southern Mendoza, was considered as a subspecies of iniscatus by Cabrera (1961). Examination of the holotypes and fairly large series in the British Museum shows A. nucus is obviously different from the typical iniscatus. It is a much larger form, and I believe that it must be considered as a distinct sjiecies. The subsF)ecific distinction of collinus from typical iniscatus is not at all evident, and I prefer to treat the former as a synonym of the latter. I examined specimens in the collection of the British Museum from central La Pampa Province (fig. 1 1 H) which match per- REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 387 388 FIELDIANA: ZOOLOGY fectly the type of A. iniscatus (fig. 11 A), and I therefore agree with Thomas that iniscatus is widespread, reaching northeast to the south of Buenos Aires Province. This latter extension of its range seems to have been recently documented, as O. A. Scaglia and C. Velazquez (p)ers. comm.) reportedly found this species in a grassland field near Balcarce (Buenos Aires Province) where it occurs in sympatry with A. azarae. Therefore, the presence of a form apparently related to iniscatus in the Lower Pleistocene of SE Buenos Aires Province is not surprising. Wheth- er lorenzinii can be thought of as an ancestor of iniscatus or as a member of an independent, but related, lineage is a matter that cannot be settled now. The second alternative is more likely, as a form more closely related to the living A. iniscatus than A. lorenzinii was contemporaneous with the latter, as I shall present next. Akodon (Akodon) cf iniscatus Thomas Referred Specimens— MMP S-640 (fig. UK): right lower jaw with incisor and very worn m, and mj; broken at the tip of the coronoid process and lacking condyloid and angular processes. Found by G. J. Scaglia in stratum II of Vorohue For- mation (Kraglievich, 1952; Marshall et al., 1984) at the Atlantic slopes S of Arroyo Loberia, Cha- padmalal region, Partido de General Pueyrredon, SE of Buenos Aires Province, Argentina. Description and Discussion— This specimen cannot be included either in the coeval species A. lorenzinii or in A. cf. cursor or A. johannis, which immediately follow in the stratigraphic succession. It differs from the former mostly in size and in all the characters I discussed in comparing lorenzinii with iniscatus. It cannot be confused wiihjohannis because of its somewhat greater size and more upturned symphysis. In all the observed charac- ters, this specimen matches the states in studied mandibles of iniscatus, making plausible its ref- erence to the living SF>ecies. The alveolar length of the lower molar row is, however, somewhat great- er (table 7) than in the type of iniscatus, but the difference obviously falls within the range of geo- graphic variation of that species (fig. 1 2). Here again, A. azarae must be considered as a possible relative of the fossil specimen. Although a close relationship with azarae could be eventually dem- onstrated by new material, I believe that it is un- likely and unsupported by the present evidence. Akodon azarae shows a less marked capsular pro- jection, a more elongated mandibular ramus, a less upturned symphysis, and a relatively deep incisor; and in all these characters, specimen MMP S-640 agrees more closely with iniscatus (fig. 1 1 L). Un- fortunately, the molar teeth are too worn to show details of the enamel pattern, which is quite dif- ferent in azarae and iniscatus, as is shown by the high frequency in the former of ectolophids and ectostylids, mesolophid remnants, and meso- stylids, which are almost completely absent in the latter. The mj of the fossil specimen, although very worn down, shows the external border of the me- soflexid fairly clearly, without trace of a mesolo- phid remnant or of a mesostylid, thus confirming a closer resemblance to iniscatus. The Meaning of the Fossil Akodontini One of the striking features of the fossil Ako- dontini known from the Pliocene and Lower and Middle Pleistocene of Argentina is that they rep- resent diverse species closely related to some living ones or else advanced extinct taxa, such as Dan- komys. The same conclusion emerged from the study of the Pliocene Phyllotini, described in another paper (Reig, 1978). Table 8 summarizes our present knowledge of the fossil species of the tribe and their chrono- il Opposite Page: Fig. 1 3. Occlusal crown views of right upper and left lower molar teeth ofliving species oi Akodon {Akodon) and of the Lower Pleistocene Akodon {Akodon) lorenzinii n. sp. A, Upper molars and B, lower molars ofliving A. andinns (Phil.); type oi A. gossei Thomas; female; BMNH 98.3.21.5; Puente del Inca, Mendoza, Argentina. C, Upper molar teeth and D, lower molar teeth of /I. puer Thomas; female; holotype, BMNH 2. 1 . 1 .78; Choquecamata, Bolivia. E, Upper molar teeth and F, lower molar teeth oi A. iniscatus Thomas; female; holotype, BMNH 3.7.9.64; Valle del Lago Planco, Chubut, Argentina. G, Upper molar teeth of /I. lorenzinii n. sp.; MMP M-867; Vorohue Formation, Lower Pleistocene, Partido de General Pueyrredon, Buenos Aires Province, Argentina. H, Lower molar teeth oi A. lorenzinii n. sp.; holotype, MMP M-1081; San Andres Formation, Barranca Parodi, Miramar, Partido de General Alvarado, Buenos Aires Province, Argentina; Lower Pleistocene. I, Right M' of ^. lorenzinii n. sp.; MLP 52. X. 4.44 (a); San Andres Formation, south of Punta Hermengo, Miramar, SE of Buenos Aires Province, Argentina; Lower Pleistocene. J, Lower molar teeth of A. lorenzinii n. sp., MMP M-868; found in association with MMP M-867. REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 389 a^a 3,4 SB ae a? as a» 4^o 4^t 4^a 4a A. iniscatus N» 20 A. puer N= 14 A. lorenzinii N* 3 L E N G T H MpM3 (CROWN) i.t u 1.1* •.t •.I It 1, 1.1 DEPTH LOWER INCISOR 12 • A. azarae V A. iniscatus « A A. puer ■ A. cf. iniscatus ▼ A. lorenzinii Fig. 14. Dice-grams and scattergrams of measurements of molar teeth in Akodon (Akodon) azarae, Akodon (Akodon) cursor, and a fossil sample referred to the latter. stratigraphic distribution in the Plio-Pleistocene column of the south of Buenos Aires Province (Marshall et al., 1984). Included are representa- tives of Bolomys and Dankomys which are still undescribed, but which will be the subject of forth- coming papers. As discussed in the previous sys- tematic part, the Upper Pliocene {A. kermacki) and Lower Pleistocene {A. magnus) species of the subgenus Abrothrix cannot be considered either ancestral to or more primitive than the living species of the subgenus. Bolomys bonapartei from the Lower Pliocene, far from representing primi- tive conditions in its character-states, is better in- terpreted as part of the spatiotemporal diversifi- 390 FIELDIANA: ZOOLOGY Table 8. Chronostratigraphic distribution of the known species of fossil Akodontini, as reported in this paper andinReig(1978). PHo-Pleistocene ages and subages Species Pliocene s. Uquian (Lower Pleistocene) Si I Bolomys bonapartei Bolomys sp.* Bolomys sp.f Dankomys simpsoni Dankomys sp4 Akodon (Ab.) kermacki Akodon {At.) magnus Akodon (Ak.) of cursor Akodon {Ak.) lorenzinii Akodon {Ak.) johannis Akodon {Ak.) of iniscatus * From ML? 52. X. 4. 30 (a), undescribed specimen from San Andres Formation in the vicinity of Miramar, Partido de General Alvarado, SE of Buenos Aires Province. t From MMP M-642 (b), undescribed M^ from Miramar Formation at Sta. Helena, Partido de Mar Chiquita, SE of Buenos Aires Province. t From MMP M-1064, undescribed jaws, partial skull and postcranial bones found in Vorohue Formation in the vicinity of Punta Loberia, Partido de General Pueyrredon, SE of Buenos Aires Province. This sp)ecimen is the basis of a new species which I shall describe in a forthcoming paper, and which is also represented by several other specimens from the Vorohue and San Andres formations. cation of a relatively advanced akodontine genus (Reig, 1 978, p. 1 69). Dankomys simpsoni from the Upper Pliocene is related to Bolomys, but shows more advanced adaptations to an herbivorous diet; it is best thought of as a derivative of the latter. The genus is represented in the Lower Pleistocene by another, more advanced species (table 8). Fossil representatives of the subgenus Akodon from the Lower (A. lorenzinii) and the Middle {A. johannis) Pleistocene are neither ancestral nor more prim- itive than related living species, and two extant species {A. cursor and A. iniscatus) were present in the Middle Pleistocene. Thus, the fossil evi- dence from the pampean region does not indicate an early stage in akodontine evolution, but rather suggests that they had already attained a high de- gree of evolution and differentiation at the generic and subgeneric level in the Pliocene. It follows that the Akodontini started to differentiate and to di- versify in times earlier than the Lower Pliocene, that is, during Miocene times. Several lines of rea- soning support the parsimonious hypothesis that this differentiation took place in those times in South America from oryzomyine South American ancestors (Reig, 1984). Fossil cricetids have not been found, however, in the rather rich deposits of the Upper Miocene (Chasicoan and Huayquerian) sediments of the pampean region. They are also absent in Miocene deposits elsewhere in Argentina or in South Amer- ica. In view of the intrinsic incompleteness of the fossil record, the absence of akodontines (and oth- er sigmodontines as well) in the known Miocene deposits may be just a matter of sampling, and they could eventually be discovered in those de- posits after more careful collecting. It might be argued that the small size of cricetid remains makes their discovery less probable than those of larger rodents, which have actually been found in rela- tive abundance in Miocene beds of Patagonia, the pampas and west of Argentina. In this sense, it may be meaningful that fossil cricetids have been mostly found in the Plio-Pleistocene outcrops of the Mar del Plata-Miramar region. This region has been exploited continuously and with careful scru- tiny during the last 40 years by a collector, Galileo J. Scaglia, who was especially well trained in hunt- ing tiny fossil remains. Although these arguments are reasonable and do not discount the eventual discovery of fossil REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 391 cricetids in the rich Miocene mammal-bearing for- mations of Argentina, I am inclined to the alter- native hypothesis that the absence of cricetids in those formations is not a result of faulty sampling, but that it represents a real absence of mice in that time and place. This absence may reflect the bio- geographic history of akodonts in particular and of the Sigmodontinae in general. There is sugges- tive evidence that the evolutionary history of sig- modontines was tightly linked with the history of the changing environments of the Andean region (Reig, 1984). As inferred from distributions of liv- ing species. South American sigmodontines are much more frequent in Andean highlands (7 1 .6%) than in the remaining lowlands (28.4%; Reig, 1 984). The pattern of species occurrences and the ende- mism of genera suggest the hypothesis that the Oryzomyini (which likely included the akodontine ancestors) had the northern Andes of Ecuador, Colombia, and Venezuela as its area of original differentiation (aod). For the Akodontini, 71.2% of total reported occurrences are strictly Andean, and the south-central Andes and north-southern Andes between parallels 30°S and 40°S show the greatest diversity of genera and the greatest fre- quency of species occurrences, as defined and dis- cussed in Reig (1984). Thus, it can be postulated that the aod of the Akodontini was located in the general area of the present southern altiplano. The akodont rodents may have radiated from this area of original differentiation, expanding gradually to the northern Andes, the southern Andes of Ar- gentina and Chile, and eventually to the lowlands of Bolivia, Paraguay, Brazil, Argentina, and Uru- guay. Thus, the colonization of the pampas by ako- dontines may have been preceded by a p>eriod of evolution in other geographic areas within South America, and the same may hold for the time of colonization of Patagonia. Therefore, the absence of cricetids in the Miocene deposits of the pam- pean region and Patagonia may reflect the fact that they had not yet colonized those areas. This being the case, the earliest occurrence of fossil akodon- tines in the Lower Pliocene deposits of Monte Her- moso would be interpreted as signaling the time of the first establishment of akodont rodents in the pampean plains. It can be argued that this explanation remains pure speculation until fossils are found in Miocene formations of the Andes. However, the explana- tion is not mere guesswork, as it is based on strong empirical data coming from patterns of diversity and distribution of the living fauna. Nevertheless, I agree that the discovery of Miocene Andean fos- sil akodontines would be a critical additional cor- roboration to the hypothesis. Several Miocene fossil faunules from the altiplano have been discovered in the last 10 years by Dr. R. Hoffstetter and collaborators (see a review in Marshall et al., 1983). So far, these faunules yielded remains of large- to medium-sized mammals, but this may be a mere reffection of the exploratory stage of their study. It must also be recognized that pres- ervation of tiny rodents demands special tapho- nomic conditions, which may or may not be found in the Andean Miocene deposits. A Tentative Scenario of the Evolutionary Deployment of the Akodontini My views on the origin, antiquity, and evolu- tionary history of the Sigmodontinae have been presented and discussed in previous papers (Reig, 1978, 1980, 1981, 1984) and are not the subject of further comment here. Instead, I shall base this argument on the main premises of my theory, namely: (1) The Sigmodontinae are a separate subfamily of the family Cricetidae that evolved in South America from North American ancestors; (2) their probable ancestors are the generalized cricetids of the North American Oligocene fep- resenting the subfamily Eucricetodontinae {sensu Martin, 1980); (3) the more primitive living sig- modontines belong to the tribe Oryzomyini, which represents the direct or indirect ancestral stock of the remaining tribes; (4) the Sigmodontinae start- ed its evolutionary deployment in South America from a proto-oryzomyine ancestor which entered that continent by overwater dispersal by the early Miocene or late Oligocene and which became es- tablished in the northern Andes of Colombia; (5) the main episodes of the differentiation of the Sig- modontinae from the ancestral oryzomyine stock occurred within the Andes and were followed by successive invasions to the eastern lowlands; and (6) after the establishment of the Panamanian land bridge, several sigmodontine lineages that had dif- ferentiated in South America invaded Middle and North America in different dispersal episodes. To complete and update the picture, I now consider the tylomyines as a separate subfamily of the Cri- cetidae which independently evolved in Middle America from eucricetodontine ancestors (see Reig, 1984, commenting on Carleton, 1980). Within this theory, the Akodontini are consid- 392 FIELDIANA: ZOOLOGY ered as a group directly descended from the Ory- zomyini, which differentiated in the area of the present altiplano. The reasons supporting this conclusion are more extensively given in Reig (1984) and were alluded to in the previous sec- tions. Let us examine the probable picture of their origin and further differentiation. First, of the 64 extant species of akodontines recognized in this paper, 35 (54.7%) belong to a single genus, Akodon, of the 1 1 recognized living genera in the tribe. Akodon, Bolomys, and Oxy- mycterus comprise 81.3% of living species. The remaining eight genera are either monotypic or comprise two or three species. Thus, although an explanation of the evolution of the latter is im- portant, a picture of the evolutionary history of the akodontines must necessarily focus on Ako- don, Bolomys, and Oxymycterus. Species o^ Akodon live in the puna, the paramos, in montane tropical and subtropical forests, in grassy pampas, dry montane Andean valleys, semidesert Patagonian tablelands, and cold south- em Andean forests. The frequency of localities from which species of Akodon are reported shows that 82% of the occurrences belong to Andean en- vironments. Seventy-two percent of species oc- currences of the subgenus Akodon are Andean, while 100% of the living Abrothrix, Chroeomys, and Hypsimys are so distributed. Bolomys shows only 46% of Andean occurrences, inhabiting both the highest altitudes of the altiplano, and the Cha- coan, pampean, and Brazilian lowlands. Oxymyc- terus also exhibits 46% Andean occurrences, rep- resented there by three of the 10 tentatively recognized species. Its species live in puna local- ities, subtropical mountain and lowland forests, the Argentinian Mesopotamian region, and the grassy pampean steppes. The habitat versatility of species of Akodon, Bolomys, and Oxymycterus contrasts with the stenotopic nature of other gen- era: Podoxymys (restricted to the high tepuis of the Guaianan region), Notiomys and Geoxus (fosso- rial and restricted to the south temperate forests and neighboring areas), or Microxus (only inhab- iting the Andean heights). The evolutionary history of the akodonts is one of successful dispersal, which was surely fueled by the habitat and trophic versatility of the more spe- ciose genera. In fact, they are broadly distributed in South America despite their overwhelming pre- dominance in Andean and montane habitats. The ancestral akodontine may have been a gen- eralized Akodon-\\\ie form of North Andean origin which colonized the area of the puna from the north in Middle or Late Miocene times before the altiplano reached considerable heights. Elevation of the altiplano began in the Middle Pliocene (Ahlfeld, 1970). This ancestral form may have en- countered adequate conditions in the southern proto-puna, and from here local differentiation may have developed as a response to heterogeneous environments in the changing Andes. Compara- tive cytogenetics suggests that this ancestral hy- pothetical akodontine might have possessed a karyotype of 58 pairs of telocentric autosomes plus the sexual pair (Vitullo et al., 1986). After a first branching which separated Oxymycterus and allies (see later) from the remaining akodontines, the earliest main radiation may have centered around what is now the genus Akodon. Akodon is the most diversified and the most generalized of the ako- dontines, and is therefore likely the original stock from which most of the remainder of the tribe radiated. A non-SF)ecialized member of the sub- genus Akodon with 2n = 52 (autosomal FN = 58) chromosomes (which is most likely the primitive karyotype of Akodon and other related genera; see Bianchi &. Merani, 1984; Vitullo et al., 1986; see also Gardner «&. Patton, 1976, and the discussion above), similar to A. andinus, may have been gen- eralized enough to live in different habitats and to settle either in montane forests of the eastern slopes of the rising Andes or in the dry high mountain valleys and open semidesertic heights. Akodon (Chroeomys) jelskii, which retains the primitive 2n = 52 karyotype, may represent a well-differ- entiated Akodon offshoot which adapted to arid heights and remained endemic to the rising alti- plano. An early main branch of the diversification within the subgenus Akodon acquired the derived karyotype of 2n = 40 (autosomal FN = 40), as found in the puna in A. boliviensis and A. albi- venter. Akodon puer (= A. caenosus, see Vitullo et al., 1 986, and above), which shows a derived 2n = 34, but keeps the same autosomal FN (Barquez et al., 1980; Vitullo et al., 1986), must have origi- nated in the puna from this branch and expanded to more southern Andean valleys and the lowlands of Tucuman. Bolomys may also have differen- tiated from the same branch early in the puna, as indicated by the generalized features ofB. amoen- us and, dating the time of its origination, by the occurrence of ^. bonapartei in the Lower Pliocene. The derived 2n = 34, FN = 34 karyotype oi Bolo- mys is more likely to have evolved from a 2n = 40 karyotype than from a 2n = 52 karyotype (Bianchi & Merani, 1984). Akodon (Hypsimys) budini, which shows a peculiar 2n = 38, FN = 42 REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 393 karyotype with few ann-to-arm homologies in banding pattern with the 2n = 40 karyotypes (Vi- tullo et al., 1 986), may represent either an isolated offshoot of the same main branch or a direct de- rivative of the earlier 2n = 52 branch which be- came endemic to the high Andean valleys. Oxy- mycterus is likely to be the result of the first dichotomy which occurred in the proto-puna from the hypothetical early akodontine ancestor. Its 2n = 54, FN = 58 karyotype seems to represent an in- dependent derivation from the hypothetical 2n = 60 ancestor (Vitullo et al., 1986). Oxymycterus may have evolved in adaptation to an animal diet and more humid Andean slojies. Some of its forms became secondarily adapted to puna habitats (as represented by some subspecies of 6>. paramensis), but the main body of the genus eventually spread into the eastern lowlands. Lenoxus is certainly a well-differentiated Oxymycterus which is more likely to have evolved from the latter in the humid Andean slopes. A complete picture of the evolutionary deploy- ment of the Akodontini, however, must account for their high diversity in regions and habitats oth- er than the original center of diversification. From what we have already said, it must be granted that much of their early diversification took place in what is now the southern puna region. From there, however, they must have migrated in different di- rections and radiated further in other areas. Three main directions of dispersal must be assumed: one to the north and another to the south, both fol- lowing the Andean axis, and a third to the south- eastern lowlands. The southern dispersal took root in the gener- alized 2n = 52 Akodon and reached the southern Andes, which served as a secondary dispersal cen- ter. This branch may be based in Akodon andinus, which reaches south to the Andes of Mendoza and is now represented by species of the subgenus ^A:<9- don (A. olivaceus, A. brachiotis, and A. markhami) and species of the subgenus Abrothrix (including A. xanthorhinus and A. hershkovitzi). It is of in- terest to note here that xanthorhinus and longipilis share the same species of the parasitic louse Ho- plopleura andina with andinus and olivaceus (Castro, 1981; pers. comm.), which might be in- terpreted as indicating persistence of an early host- parasite relation. The specialized genera Chele- mys, Notiomys, and Geoxus. two of which are known to retain the 2n = 52 karyotype, the second being unknown in its chromosomes (Pearson, 1984), may easily be interpreted as independent offshoots of this southern branch which special- ized in habits and diet. Abrothrix originated cer- tainly in the Pliocene, as indicated by its presence in the Upper Pliocene (Chapadmalalan) of south- em Buenos Aires Province. The presence of both fossil Abrothrix in the Upper Pliocene and Lower Pleistocene of the pampean region, and of the ex- tant A. longipilis and A. xanthorhinus) in the Pa- tagonian tablelands, indicates that this southern Andean dispersal subsequently spread into eastern stepjjes. Significantly, all southern akodonts so far known in their chromosomes retain the primitive 2n = 52 karyotype without meaningful modifi- cations. It is unclear whether living Patagonian species as A. iniscatus and A. nucus, the chro- mosomes of which having not been described yet, derived from this branch or from the third dis- persing branch described below. In view of the affinity of these species with puer and azarae, as discussed above, we are inclined to the second alternative. A dispersal to the north from the original south- em puna differentiation area is necessary to ex- plain the distribution oi A. aerosus, A. orophilus, A. mollis, A. tolimae, and A. urichi, as well as of species of Microxus and Podoxymys. Given the presence of most of these taxa in the northern Andes and connected mountain ranges, where they could have originated directly from the oryzo- myines, it is necessary to postulate that the north Andean akodontines migrated there from a south- central Andean region in order to maintain the monophyly of the tribe. Chromosomal evidence is highly suggestive of the derived condition of the northern species of Akodon, with A. orophilus (2n = 26), A. mollis (2n = 22), and A. urichi (2n = 18) indicating a northward decrease in chromosome number (chromosomal data from Bianchi & Mer- ani, 1984; Gardner & Patton, 1976; Reig et al., 1971). Unpublished chromosomal counts from Colombian Akodon, which may represent A. to- limae (fTom Santander; C. Ramirez, pers. comm.), with 2n = 24 chromosomes, and A. urichi (from Villavicencio; A. Gardner, pers. comm.), with 2n = 18 chromosomes, suggest that the pattern of decrease may not be regular. The less reduced karyotyp)e of species of Microxus (as known in the polymorphic 2n = 35-37, FN = 48 of Microxus bogotensis from the Venezuelan paramos; see Bar- ros & Reig, 1979) suggests a separate origin from the primitive 2n = 52 Akodon stock, consistent with its generic distinction. Podoxymys is an en- demic genus from the tepuis that is likely to have been derived from the same main branch; its karyotype is unknown. 394 FIELDIANA: ZOOLOGY The third main direction of dispersal, from the southern puna area of original differentiation di- rectly toward the eastern lowlands, is necessary to explain the present distribution of lowland rep- resentatives ofAkodon s.s., Oxymycterus. and Bo- lomys, as well of the exclusively lowland Akodon {Deltamys) of the genera Blarinomys, Juscelino- mys, and fossil Dankomys. The itinerary of Bo- lomys may be inferred from the distribution of the living species, starting with a B. amoenus-hke an- cestral form in the puna, with B. lactens in pam- pean range valleys of northwestern Argentina, and B. lenguarum in the lowlands of Bolivia and Par- aguay. Two diverging lines of dispersal can be in- ferred from the last region: one toward the pam- pean region (represented by B. bonapartei, B. obscurus, including bene/actus, and a new unde- scribed species), the other toward southern and eastern Brazil (B. lasiurus), passing through the Argentinian Chaco and the northern Mesopota- mian region {B. temchuki). A similar disp)ersal pat- tern may be postulated for Oxymycterus, which shows a similar pattern of SF)ecies distribution in the lowlands of Paraguay, Argentina, Brazil, and Uruguay, as well as in the high valleys of north- western Argentina. Dankomys may represent a lo- cal pampean derivative of Bolomys, and Jusceli- nomys may have been derived from a Brazilian branch of Oxymycterus. Regarding the lowland differentiation within the genus Akodon, it is in- teresting to realize that A. {Ak.) boliviensis, a typ- ical inhabitant of the present puna, but also rep- resented at lower altitudes in sites surrounding the puna and even in the lowlands of Tucuman, has 2n = 40 chromosomes. Most species of Akodon of the lowlands of central Argentina, the pampean region, the Mesopotamia, and Uruguay show either an identical 2n = 40 karyotype (A. varius complex) or a rather similar 2n = 42-43 (A. molinae), 2n = 38 (/I. azarae, A. dolores), or 2n = 37 (/i. kempi) karyotyijes, sharing many resemblances to each other (Bianchi & Merani, 1984). Unfortunately, we still do not know the karyotyjje of Akodon cursor, which might belong to the same group. These Argentinian forms are then likely to rep- resent an evolutionary group stemming from a puna ancestor probably closely related to A. bo- liviensis. Akodon nigrita poses an interesting prob- lem, as its 2n = 52 karyotype precludes its mem- bership in the same clade. The place within this picture of the unnamed Brazilian species with 2n = 14-16 and 2n = 24-25 chromosomes, as well as of the poorly known A. serrensis, A. reinhardti, and Blarinomys breviceps, is still unclear. Needless to say, the above scenario is quite ten- tative and must be taken only as a set of working hypotheses open to partial or overall modification after the test of more detailed morphological, pa- leontological, cytogenetical, and biochemical studies. However, I believe that in its present ver- sion, it may represent a reasonable explanation of the most probable biogeographic and evolutionary events connected with the radiation of the ako- dontines and, therefore, a heuristic framework for further advances in the knowledge of the evolution of these rodents. Acknowledgments I thank S. Anderson, G. Corbet, K. Kermack, R. Pascual, G. J. Scaglia, and O. A. Scaglia for facilitating access to collections under their care. O. P. Pearson, A. Spotomo, N. O. Bianchi, and A. Langguth are also acknowledged for fruitful dis- cussions. I thank Bruce Patterson and two anon- ymous referees for suggestions and a careful read- ing of the manuscript. This paper is also the outcome of a John Simon Guggenheim fellowship, and this Foundation is heartily acknowledged for its continuous support. Literature Cited Ahlfeld, F. 1 970. Zur Tektonik des andinen Bolivien. Geologischer Rundschau, 56: 1 124-1 140. Alston, W. F. 1 964. Philosophy of Language. Pren- tice-Hall, Inc., Englewood Cliffs, N.J., 1 12 pp. Ameghino, F. 1889. Contribucion al conocimiento de los mamiferos fosiles de la Republica Argentina. Actas de la Academia Nacional de Ciencias de Cordoba, 6: 1-1208 and Atlas. Anthony, H. E. 1 929. Two new genera of rodents from South America. American Museum Novitates, 388: 1-9. Arata, A. 1967. Muroid, gliroid, and dipodoid ro- dents, pp. 226-253. In Anderson, S., and J. K. Jones, Jr., eds.. Recent Mammals of the World. A Synopsis of Families. The Ronald Press Co., New York, viii + 453 pp. Barquez, R. M., D. F. Williams, M. A. Mares, and H. H. Genoways. 1980. Karyology and morpho- metries of three species of Akodon (Mammalia: Mu- ridae) from northwestern Argentina. Annals of Car- negie Museum, 49: 379-403. Barros, M. A., AND O. A. Reig. 1979. Doble poli- morfismo robertsoniano en Microxus bogotensis (Ro- dentia, Cricetidae) del Paramo de Mucubaji (Merida, REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 395 Venezuela). Acta Cientifica Venezolana, 30(Suple- mentoN" 1): 106. Beckner, M. 1959. The Biological Way of Thought. Columbia University Press, New York, 200 pp. BiANCHi, N. O., AND J. R. CoNTRERAS. 1967. Mcca- nismo de determinacion sexual en el raton de campo, Akodon azarae (Rodentia, Cricetidae). Revista Ar- gentina de Endocrinologia y Metabolismo, 13: 157- 167. BlANCHl, N. O., J. R. CONTRERAS, AND F. N. DULOUT. 1969. Intraspecies autosomal polymorphism and chromosomal replication in Akodon azarae (Roden- tia, Cricetidae). Canadian Journal of Genetics and Cy- tology, 11: 233-243. BlANCHl, N. O., AND S. Merani. 1984. Cytogenetics of South American akodont rodents (Cricetidae). X. Karyological distances at generic and intrageneric levels. Journal of Mammalogy, 65: 206-219. BlANCHl, N. O., S. Merani, and M. Lizarralde. 1979. Cytogenetics of South American akodont rodents (Cri- cetidae). V. Segregation of chromosome Nr. 1 poly- morphism in Akodon molinae. Experientia, 35: 1438- 1439. BlANCHl, N. O., O. A. Reig, J. Molina, and F. N. DuLOUT. 1971. Cytogenetics of South American akodont rodents (Cricetidae). I. A progress record of Argentinian and Venezuelan forms. Evolution, 21: 724— 736. BlANCHl, N. O., L. ViDAL RiOJA, AND M. S. A. BlANCHl. 1973. Constitutive heterochromatin, G-bands and robertsonian rearrangements in the chromosomes of Akodon molinae (Rodentia, Cricetidae). Canadian Journal of Genetics and Cytology, 15: 855-861. . 1976. Cytogenetics of South American ako- dont rodents (Cricetidae). II. Interspecific homologies of G-banding patterns. Cytologia, 41: 139-144. Cabrera, A. 1961. Catalogo de los mamiferos de America del Sur. II (Sirenia, Perissodactyla, Artiodac- tyla, Lagomorpha, Rodentia). Revista del Museo Ar- gentino de Ciencias Naturales "Bernardino Rivada- via"; (Zoologia) 4: 309-732. Carleton, M. D. 1973. A survey of gross stomach morphology in New World Cricetinae (Rodentia, Mu- roidea), with comments on functional interpretations. Miscellaneous Publications Museum of Zoology, Uni- versity of Michigan, 146: 1-43. . 1980. Phylogenetic relationships in neotom- ine-peromyscine rodents (Muroidea) and a reappraisal of the dichotomy within New World Cricetinae. Mis- cellaneous Publications Museum of 2toology, Univer- sity of Michigan, 157: 1-146. Castro, M. del C. 1981. Contribucion al conoci- miento de los Anoplura neotropicales. Revista Socie- dad Entomologica Argentina, 40: 231-236. . 1982. Las especies del grupo travassosi del genero Hoplopleura Enderlein, 1904 en la Republica Argentina. Revista Sociedad Entomologica Argentina, 41: 171-182. Cestari, a. N., AND J. Imada. 1968. Os cromossomos do roedor Akodon arviculoides cursor Winge, 1888 (Cricetidae, Rodentia). Ciencia e Cultura, 20: 758- 772. Contreras, J. R. 1 982. Notas acerca de Bolomys tem- chuki (Massoia 1982) en el noreste Argentino, con la descripcion de dos nuevas subspecies (Rodentia, Cri- cetidae). Historia Natural (Corrientes), 2: 174-176. , and M. I. Rosi. 1981. Notas sobre los Ako- dontini argentinos (Rodentia, Cricetidae). I. Abrothrix longipilis moerens Thomas, 1919, en el Parque Na- cional Nahuel Huapi. Historia Natural (Corrientes), 1: 209-212. Copi, I. M. 1953. Introduction to Logic. The Mac- Millan Co., New York, 472 pp. de Santis, L. J. M., AND E. R. JusTO. 1980. Akodon (Abrothrix) mansoensis, sp. nov., un nuevo "raton la- noso" de la Provincia de Rio Negro, Argentina (Ro- dentia, Cricetidae). Neotropica, 26: 121-127. DuLOUT, F. N., R. M. PoNSSA, A. E. Schugurensky de Adris, and L. L. Dulout. 1976. Multiformidad y polimorfismo cromosomico en Akodon {Abrothrix) il- lutea Thomas (Rodentia, Cricetidae). Mendeliana, 1: 49-52. Echave Llanos, J. M., and C. A. Vilchez. 1 964. Ana- tomia microscopica del estomago del raton hocicudo {Oxymycterus rutilans). Revista de la Sociedad Ai^gen- tina de Biologia, 40: 187-192. Ellerman, J. R. 1941. The Families and Genera of Living Rodents. Vol. II. Family Muridae. British Mu- seum (Natural History), London, xii + 690 pp. Gallardo, M. H. 1982. Chromosomal homology in southern ^A:orfo/i. Experientia, 38: 1485-1487. Gardner, A. L., and J. L. Patton. 1976. Karyotypic variation in oryzomyine rodents (Cricetinae) with comments on chromosomal evolution in the Neo- tropical cricetine complex. Occasional Papers, Mu- seum of Zoology, Louisiana State University, 49: 1- 48. Gentile de Fronza, T. 1970. Cariotipo de Oxymyc- terus rutilans platensis (Rodentia, Cricetidae). Phypis, 30: 343. Gentile de Fronza, T., R. L. Wainberg, and G. E. HuRTADO DE Catalfo. 1979. Cromosomas sexuales multiples en Deltamys kempi (Rodentia, Cricetidae): estadio preliminar a la instauracion del sistema XYiYj/ XX. Mendeliana, 4: 28-35. Gyldenstolpe, N. 1 932. A manual of Neotropical sig- modont rodents. Kungliga Svenska Vetenskapsaka- demiens Handliger, 11(3): 1-164; 18 pi. Hershkovitz, P. 1962. Evolution of Neotropical cri- cetine rodents (Muridae), with special reference to the phyllotine group. Fieldiana: Zoology, 46, 1-524. . 1966. South American swamp and fossorial rats of the Scapteromyine group (Cricetinae, Muridae), with comments on the glans penis in murid taxonomy. Z«itschrifl fiir Saugetierkunde, 31: 81-149. 1 967. Dynamics of rodent molar evolution: A study based on New World Cricetinae, family Mu- ridae. Journal of Dental Research, 46 (Supplement to No. 5): 829-842. HoNACKi, J. H.,K. E. Kinman, andJ. W. Koeppl. 1982. Mammal Species of the World. Allen Press and As- sociation of Systematic Collections, Lawrence, Kan- sas, 694 pp. Hooper, E. T., and G. G. Musser. 1964. The glans penis in Neotropical cricetines (family Muridae) with 396 FIELDIANA: ZOOLOGY comments on classification of muroid rodents. Mis- cellaneous Publications Museum of Zoology, Univer- sity of Michigan, 123: 1-57. Kjblisky, p., N. O. Bianchi, and M. S. Merani. 1976. Citogenetica de los roedores sudamericanos akodon- tinos (Cricetidae). III. Polimorfismos cromosomicos intraespecificos en Akodon dolores Thomas. Mende- liana, 1: 97-103. Kjblisky, P., I. Lobig, and O. A. Reig. 1970. Estudio cromosomico en diversas poblaciones del raton Ua- nero, Zygodontomys microtinus Thomas. Acta Cien- tifica Venezolana, 21 (Suplemento 1): 35. Kraglievich, J. L. 1952. El perfil geologico de Cha- padmalal y Miramar, Provincia de Buenos Aires. Re- sumen preliminar. Revista del Museo Municipal de Ciencias Naturales y Tradicional de Mar del Plata, 1: 8-37. . 1965. Speciation phyletique dans les rongeurs fossiles du genre Eumysops (Echimyidae: Heteropso- myinae). Mammalia (Paris), 29: 258-265. Kravetz, F. O. 1973. Estudio de regimen alimen- tario, periodos de actividad y otros rasgos ecologicos en una poblacion de "raton hocicudo" {Oxymycterus rufus platensis) de Punta Lara. Acta Zoologica Lil- loana, 29: 201-212. LoBATO, L., G. Cantos, B. Araujo, N. O. Bianchi, and S. Merani. 1982. Cytogenetics of South American akodont rodents (Cricetidae). VIII. Akodon mollis: A species with XY females and B-chromosomes. Ge- netica, 57: 199-205. Langguth, A. 1975. La identidad de A/iis /as/o/w Lund y el status de genero Thalpomys Thomas (Mam- malia, Cricetidae). Papeis Avulsos de Zoologia (Sao Paulo), 29: 45-54. Maia, v., and a. Langguth. 1981. New karyotypes of Brazilian akodont rodents, with notes on taxonomy. Zeitschrift fiir Saugetierkunde, 46: 241-249. Marshall, L. G., A. Berta, R. Hoffstetter, R. Pascual, O. a. Reig, M. Bombin, and A. Mones. 1 984. Geochronology of the continental mammal- bearing Quaternary of South America. Palaeoverte- brata, Memoire Extraordinaire, 1984: 1-80. Marshall, L. G., R. Hoffstetter, and R. Pascual. 1983. Geochronology of the continental mammal- bearing Tertiary of South America. Palaeovertebrata, Memoire Extraordinaire, 1983: 1-93. Martin, L. D. 1980. The early evolution of the Cri- cetidae in North America. The University of Kansas Paleontological Contributions, 102: 1—42. Massoia, E. 1963a. Sobre la posicion sistematica y distribucion geograhca de Akodon {Thaptomys) nigrita (Rodentia, Cricetidae). Physis (Buenos Aires), 24(67): 73-80. . 1963b. Oxymj'c/erus //imn^' (Rodentia, Cri- cetidae), nueva esf)ecie para la Argentina. Physis (Bue- nos Aires), 24(67): 129-136. 1981a. Nuevos datos sobre Akodon, Deltamys y Cabreramys, con la descripcion de una especie y de una subespecie nuevas (Mammalia, Rodentia, Crice- tidae). Historia Natural (Corrientes), 1: 179-180. . 1981b. Notas sobre los cricetidos mendocinos. . 1982. Diagnosis previa de Cabreramys tem- chuki, nueva especie (Rodentia, Cricetidae). Historia Natural (Corrientes), 2: 91-92. Massoia, E., and A. FoRNES. 1962. Un cricetido nuevo para la Argentina: Akodon arviculoides montensis Thomas (Rodentia). Physis (Buenos Aires) 23(65): 1 83- 194. . 1967. El estado sistematico, distribucion geo- grafica y datos etoecologicos de algunos mamiferos neotropicales (Marsupialia y Rodentia), con la de- scripcion de Cabreramys, genero nuevo (Cricetidae). Acta Zooloogica Lilloana, 23: 407-430. 1969. Caracteres comunes y distinitivos de Oxymycterus nasutus (Waterhouse) y O. iheringi Thomas (Rodentia, Cricetidae). Physis (Buenos Aires), 28(77): 315-321. Meserve, p. L. 1 98 1 . Trophic relationships among small mammals in a Chilean semiarid thorn scrub com- munity. Journal of Mammalogy, 62: 304-314. Meserve, P. L., and W. Glanz. 1978. Geographical ecology of small mammals in the northern Chilean arid border. Journal of Biogeography, 5: 135-148. Moojen, J. 1 965. Novo genero de Cricetidae do Brasil Central (Glires, Mammalia). Revista Brasileira de Biologia, 25: 281-285. Osgood, W. H. 1925. The long-clawed South Amer- ican rodents of the genus Notiomys. Field Museum of Natural History, 2^oological Series, 12: 1 13-125. . 1943. The mammals of Chile. Field Museum of Natural History, Zoological Series, 30: 1-268. Patterson, B. D., M. H. Gallardo, and K. E. Freas. 1984. Systematics of mice of the subgenus Akodon (Rodentia, Cricetidae) in southern South America, with the description of a new species. Fieldiana: Zoology, n.s., 23: 1-16. Pearson, O. P. 1967. La estructura por edades y la dinamica reproductiva en una poblacion de ratones de campo, Akodon azarae. Physis (Buenos Aires), 27: 53-58. . 1983. Characteristics of mammalian faunas from forests in Patagonia, southern Argentina. Journal of Mammalogy, 64: 476—492. 1984. Taxonomy and natural history of some fossorial rodents of Patagonia, southern Ai^gentina. Journal of Zoology (London), 202: 225-237. Reig, O. A. 1955. Un nuevo genero y especie de ce- nolestinos del Plioceno de la Provincia de Buenos Aires (Rep. Argentina). Revista de la Asociacion Geologica Argentina, 10: 70-71. . 1959. Comunicacion preliminar sobre nuevas especies del genero Thylatheridium Reig (Mammalia, Didelphidae). Neotropica, 4: 89-95. -. 1964. Roedores y marsupiales del Partido de General Pueyrredon y regiones adyacentes (Provincia de Buenos Aires, Republica Argentina). Publicaciones del Museo Municipal de Ciencias Naturales de Mar del Plata, 1: 203-224. . 1970. The Proterosuchia and the early evo- Historia Natural (Corrientes) 1: 205-208. lution of the Archosauria; An essay about the origin of a major taxon. Bulletin of the Museum of Com- parative Zoology at Harvard University, 139: 229- 292. REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 397 . 1 977. A proposed unified nomenclature for the enamelled components of the molar teeth of the Cri- cetidae (Rodentia). Journal of Zoology (London), 181: 227-241. . 1978. Roedores cricetidos del Plioceno supe- rior de la Provincia de Buenos Aires. Publicaciones del Museo Municipal de Ciencias Naturales de Mar del Plata, 2: 164-190. . 1 980. A new fossil genus of South American cricetid rodents allied to Wiedemys, with an assess- ment of the Sigmodontinae. Journal of Zoology (Lon- don), 192:257-281. 1981. Teoria del origen y desarollo de la fauna de mamiferos de America del Sur. Monographiae Na- turae, Museo Municipal de Ciencias Naturales "Lo- renzo Scaglia," 1: 1-161. 1982. The reality of biological species: A con- ceptualistic and a systemic approach, pp. 479—499. In Los, J., and H. Pfeifer (eds.). Proceedings of the 6th International Congress of Logic, Methodology, and Philosophy of Science. D. Reidel Publishing Com- pany, Dordrecht, Netherlands. -. 1984. Distribu^ao geograiica e historia evo- lutiva dos roedores muroideos sulamericanos (Crice- tidae: Sigmodontinae). Revista Brasileira de Genetica, 7: 333-365. Reig, O. a., p. Kjblisky, and O. J. Linares. 1971. Datos sobre Akodon urichi venezuelensis Allen (Ro- dentia: Cricetidae), un raton de interes para la inves- tigacion citogenetica. Acta Biologica Venezuelica, 7: 157-189. Reig, O. A., Kirsch, J. A. W., and Marshall, L. In press. Systematic relationships of the living and Neo- cenozoic opossum-like marsupials (suborder Didel- phimorpha). In Archer, M., ed.. Possums and Opos- sums. Royal Society of New South Wales, Sydney, Australia. Reig, O. A., and O. J. Linares. 1969. The occurrence of Akodon in the Upper Pliocene of Argentina. Journal of Mammalogy, 50: 643-647. Reig, O. A., N. Olivo, and P. Kjblisky. 1971. The idiogram of the Venezuelan vole mouse, Akodon uri- chi venezuelensis Allen (Rodentia, Cricetidae). Cyto- genetics, 10: 99-114. Reig, O. A., and G. G. Simpson. 1972. Sparassocynus (Marsupialia, Didelphidae), a peculiar mammal from the Late Cenozoic of Argentina. Journal of Zoology (London), 167:511-539. Rodriquez, M., R. Montoya, and W. Venegas. 1983. Cytogenetic analysis of some Chilean species of the %enus Akodon Meyen (Rodentia, Cricetidae). Caryolo- gia, 36: 129-138. Sanborn, C. C. 1947. Geographic races of the rodent Akodon jelskii. Fieldiana: Zoology, 31: 137-142. Sbalqueiro, I. J., M. A. Mattevi, and L. F. B. Oliveira. 1984. An X.X.X^Xj/X.XjY mechanism of sex de- termination in a South American rodent, Deltamys kempi (Rodentia, Cricetidae). Cytogenetics and Cell Genetics, 38: 50-55. SoKAL, R. R., AND P. H. A. Sneath. 1963. Principles of numerical taxonomy. W. H. Freeman, San Fran- cisco. Spotorno, a., AND R. Fernandez. 1976. Chromo- some stability in southern Akodon (Rodentia, Crice- tidae). Mammalian Chromosomes Newsletter, 17(4): 13-14. Tate, G. H. H. 1932a. The taxonomic history of the South and Central American oryzomine genera of ro- dents (excluding Oryzomys): Nesoryzomys, Zygodon- tomys, Chiiomys, Delomys, Phaenomys, Rhagomys, Rhipidomys, Nyctomys, Oecomys, Thomasomys, In- omys. Aepeomys. Neacomys, and Scolomys. American Museum Novitates, 581: 1-28. . 1 932b. The taxonomic history of the South and Central American akodont rodent genera: Thal- pomys, Deltamys, Thaptomys, Hypsimys, Bolomys, Chroeomys, Abrothrix, Microxus, Podoxymys, Len- oxus, Oxymycterus. Notiomys and Blarinomys. Amer- ican Museum Novitates, 582: 1-32. Thomas, O. 1 884. On a collection of Muridae from central Peru. Proceedings of the Zoological Society of London, 1884: 447-458. . 1900. New Peruvian species of Conepatus, Phyllotis, and Akodon. Annals & Magazine of Natural History, (7)6: 466-469. -. 1902. On mammals collected by Mr. Perry O. Simons in the Southern part of the Bolivian Plateau. Annals & Magazine of Natural History (7)9: 222-230. . 1909. Notes on some South American mam- mals, with description of new species. Annals & Mag- azine of Natural History (8)4: 230-242. -. 1913. NQv/form^oi Akodon and Phyllotis, and a new genus for "'Akodon" teguina. Annals & Maga- zine of Natural History, (8)11: 404-409. -. 1916. The grouping of the South American Muridae commonly referred to Akodon. Annals & Magazine of Natural History, (8)18: 336-340. 1918. On small mammals from Salta and Jujuy collected by Mr. E. Budin. Annals & Magazine of Nat- ural History, (9)1: 186-193. 1919. On small mammals collected by Sr. E. Budin in North Western Patagonia. Annals & Maga- zine of Natural History, (9)3: 199-212. . 1920. A further collection of mammals from Jujuy. Annals & Magazine of Natural History, (9)5: 188-198. 1926a. Two new mammals from northern Ar- gentina. Annals & Magazine of Natural History, (9)17: 311-312. . 1 926b. The Spedan Lewis South American ex- ploration. III. On mammals collected by Sr. Budin in the Province of Tucuman. Annals & Magazine of Nat- ural History, (9)17: 602-609. . 1927. On a further collection of mammals made by Sr. E. Budin in Patagonia. Annals & Magazine of Natural History, (9)19: 545-554. Tullberg, T. 1899. Ueber das System der Nagetiere; eine phylogenetische Studie. Nova Acta Regiae Socie- tatis Scientiarum Upsalensis (3)18: 1-514. ViEiRA, C. 1955. Roedores e lagomorfos do Estado de Sao Paulo. Arquivos de Zoologia do Estado de Sao Paulo, 8: 129-167. ViTULLo, A. D., M. S. Merani, O. A. Reig, A. E. Kajon, O. Scaglia, M. B. Espinosa, and A. Perez-Zapata. 1986. Cytogenetics of South American akodont ro- dents (Cricetidae): New karyotypes and chromosomal 398 FIELDIANA: ZOOLOGY banding patterns of Argentinian and Uruguayan forms. Journal of Mammalogy, 67: 69-80. VoRONTzov, N. N. 1959. Sistema khomiakov (Cricet- inae) mirovoi fauny i ikh iilogeneticheskie sviazi. Biu- letin Moskovskogo Obshtschestva Ispitately Prirody, Otdel Biologia, 44: 134-137. . 1967. Evoliutzia pishtshevaritelnoi sistemy gryzunov (mysheobraznye). Isdatelstvo "Nauka," No- vosibirsk, 240 pp. . 1982. Nizshie khomiakoobraznye (Cricetidae) mirovoi fauny. Chast I. Morfologiia i Ekologiia. In Fauna S.S.S.R. G^ov. Seria No. 125), Mlekopitaiu- shtchie, T. Ill, Vyp. 6. Izdatielstvo "Nauka," Lenin- grad, 451 pp. Voss, R. S., AND A. V. LiNZEY. 1981. Comparative gross morphology of male accessory glands among Neotropical Muridae (Mammalia: Rodentia) with comments on systematic implications. Miscellaneous Publications Museum Zoology, University of Michi- gan, No. 159: 1-41. Williams, C. B. 1 964. Patterns in the Balance of Nature. Academic Press, London and New York, 324 pp. WiNGE, H. 1887. Jordfunde og nulevende Gnavere (Rodentia) fra Lagoa Santa, Minas Geraes, Brasilien. E Museo Lundii, 1(3): 1-200. XiMENEZ, A., AND A. Langguth. 1970. Akodon cursor montensis en el Uruguay (Mammalia, Cricetinae). Co- municaciones Zoologicas del Museo de Historia Nat- ural de Montevideo, 128: 1-7. Yaisiez, J., W.SiEFELD, J. Valencia, AND F.Jaksic. 1978. Relaciones entre la sistematica y la morfometria del subgenero Abrothrix (Rodentia: Cricetidae) en Chile. Anales del Instituto de la Patagonia, 9: 185-197. Yonenaga, Y. 1972. Chromosomal polymorphism in the rodent Akodon arviculoides ssp. (2n = 1 4) resulting from two pericentric inversions. Cytogenetics, 11: 488- 499. . 1975. Karyotypes and chromosome polymor- phism in Brazilian rodents. Caryologia, 28: 269-286. . 1 979. New karyotypes and somatic and germ- cell banding in Akodon arviculoides (Rodentia, Cri- cetidae). Cytogenetics and Cell Genetics, 23: 24 1-249. Yonenaga, Y., S. Kasahara, E. J. C. Almeida, and A. L. PiRACCHi. 1975. Chromosomal banding patterns in Akodon arviculoides (2n = 14), Akodon sp. (2n = 24 and 25), and two male hybrids with 19 chromo- somes. Cytogenetics and Cell Genetics, 15: 388-399. Yonenaga, Y., AND F. Ricci. 1969. Estudos cromos- somicos em especies de roedores. Cien9a e Cultura, 21: 249. REIG: SYSTEMATICS AND EVOLUTION OF AKODONTINI 399 Biogeography of Octodontid Rodents: An Eco-Evolutionary Hypothesis Luis C. Contreras, Juan C. Torres-Mura, and Jose L. Yaflez ABSTRACTS The family Octodontidae (Rodentia, Hystricognatha) is an old group of low diversity, cur- rently found on both sides of the Andean mountains between 16°S and 41°S. Information on the geographic distribution of the octodontid genera is presented and discussed, and the sys- tematic status of each species and subspecies is given. An explanation is also proposed for the present distribution of the family, considering geo- logical, climatic, floristic, faunistic, and ecological events that occurred after the first appearance of octodontids in the Deseadan age (early Oligocene) in Bolivia and Patagonia. The uplift of the Andes, the formation of Patagonian pampas, the disappearance of echimyids from the Patagonian Subregion, and the appearance of ctenomyids seem to be the most important factors determining the present distribution of octodontids. La familia Octodontidae (Rodentia, Hystricognatha) es un antiguo grupo de roedores poco diversificados que se encuentra a ambos lados de la Cordillera de los Andes entre los 1 6° y los 41°S. En este trabajo presentamos y discutimos la informacion sobre la distribucion de los generos de octodontidos y hacemos comentarios sobre el estatus sistematico de cada especie y subespecie. Proponemos una explicacion para la distribucion actual de esta familia considerando eventos geologicos, climaticos, floristicos, faunisticos y ecologicos que ocurrieron despues de la aparicion de los octodontidos en el registro fosil en el Deseadano (Oligoceno) de Bolivia y Patagonia. El levantamiento de los Andes, la formacion de las Pampas Patagonicas, la desaparicion de los echimidos de la Subregion Patagonica y la aparicion de los ctenomidos, parecen ser los factores mas importantes en la estructuracion de la distribucion actual de la familia Octodon- tidae. A familia Octodontidae (Rodentia, Hystricognatha) e um antigo grupo de pouca diversidade, que se encontrado em ambos os lados das Cordilheiras dos Andes, entre 16° e 41°S. Apresen- lamos informa96es sobre as distribui96es geograficas dos generos octodontinos, e discuti-se as categorias sistematicas de cada especie ou subespecie. Prop6e-se uma explica9ao para a atual distribui^ao desta familia, considerando-se os eventos geologicos, climaticos, floristicos e ecologicos que ocorreram apos o aparecimento dos octo- dontinos no registro fossil, durante a Era Deseadana (no Oligoceno inferior) da Bolivia e da Patagonia. A eleva9ao dos Andes, a formafao dos Pampas da Patagonia, o desaparecimento dos equimideos da Subregiao Patagonica, e o aparecimento dos ctenomideos, parecem ser os fatores mais importantes determinando a atual distribuijao dos octodontinos. From the Departamento de Biologia y Quimica, Fa- cuhad de Ciencias, Universidad de Talca, Casilla 747, Talca, Chile (Contreras and Torres-Murra); and Museo Nacional de Historia Natural, Casilla 787, Santiago, Chile (Yanez). CONTRERAS ET AL: BIOGEOGRAPHY OF OCTODONTID RODENTS 40 1 Introduction The order Rodentia is represented in South America by two main groups: the hystricognaths (sensu Woods, 1 982) and the cricetids. The former had an independent evolutionary history lasting at least 35 million years while South America was an island continent. About 3.5 million years before present (MYBP), the Isthmus of Panama was formed, allowing the other main group of rodents, the cricetids, as well as other mammals, to invade from North America (Reig, 1981; Patterson & Wood, 1982; Webb & Marshall, 1982). South American hystricognaths are characterized by a relatively large body mass (80 g to 50 kg), low species diversity, long gestation period, and small litter size (Rowlands & Weir, 1974). The opposite is characteristic of South American cricetids (Pear- son, 1958; Hershkovitz, 1962). Octodontidae Waterhouse (excluding Cteno- mys) is one of 1 1 families of hystricognath rodents found in South America. The family has a small number of lineages (Woods, 1 982; Mares & Ojeda, 1982). It is distributed along the Andean moun- tains from 16°S to 41°S, especially on the western slopes. As a result of this, a large proportion of the species are endemic to or found mainly in Chile (seven of nine species). Despite its low phyletic diversity, this family is highly diverse in forms, ranging from the relatively generalized Octodon degus, capable of digging and climbing shrubs, to the specialized, fossorial Spalacopus cyanus (Glanz, 1977). The octodontids are probably the least-studied family of South American hystricognaths. The phylogenetic relations of its genera are not clear, almost all species are poorly known from an eco- logical standpoint, and the geographic ranges of most species are uncertain (Mares & Ojeda, 1 982). Here we analyze information published in the last few years and new data concerning the distribution and systematics of this family. Geographic Distribution and Systematic Status Genus Octodon Bennett, 1832 According to Osgood (1943) three monotypic species are recognized in this genus: O. degus Mo- lina, 1782; O. bridgesi Waterhouse, 1844; and O. lunatus Osgood, 1943. Of these, the first two can be differentiated by external morphology and be- havior (Osgood, 1943; Ipinza et al., 1971). Octo- don lunatus was distinguished only by the absence of an indentation on the inner border of the last upper molars, all other features being the same as those in O. bridgesi. However, intrapopulational variability of this dental character is high, not only within O. bridgesi but also within O. degus (pers. obs.; Simonetti, pers. comm.), rendering it useless for taxonomic diagnosis. Although this does not invalidate O. lunatus, we believe it is a species of questionable status. Unfortunately, this uncertain- ty cannot be resolved due to scarcity of existing material. Octodon degus is one of the best-studied small mammals of central Chile in terms of its ecology, behavior, and physiology (Woods & Boraker, 1975; Rosenmann, 1977; Yaiiez & Jaksic, 1978; Con- treras & Rosenmann, 1982; Meserve et al., 1984; and references therein). The degu is thought to be distributed south of Huasco Province (28°28'S) to Curico (35°00'S) (Mann, 1 978; Tamayo «fe Fras- sinetti, 1 980). However, we are not aware of any specimens collected south of Santiago (fig. 1). The distributional limits of this abundant (Jaksic et al., 1981) and diurnal (Rosenmann et al . , 1981) species seem correlated with those of the mediterranean shrubland formation known as matorral (Mann, 1978), whose northern and southern limits are de- termined by a scarcity and superabundance of water, respectively (Mann, 1 964). The altitudinal limits of O. degus seem to be determined by its poor tolerance to low oxygen partial pressure (Ro- senmann & Morrison, 1975) and also by altitu- dinal limits of its preferred habitat. This altitu- dinal limit is probably lower at higher latitudes because of temperature effects. Although O. degus can burrow, it feeds mainly on grasses and forbs (Meserve et al., 1983) above ground, which are covered by snow much of the year at high altitudes in the Andes. The highest capture record for this species is 2,000 m at 30°S. At 33°S (around San- tiago) the altitudinal limit is probably below 1,200 m. Octodon bridgesi is known to be distributed south from Cachapoal Province (34°15'S), along the foothills of the Andes to Malleco Province (38°40'S) (Greer, 1965; Tamayo «&Frassinetli, 1980) (fig. 1). Octodon lunatus is reportedly found along the Cordillera Occidental (Cordillera de la Costa) from La Dormida (33°04'S) in Quillota Province to as far north as lllapel (31°30'S) (fig. 1). These limits 402 FIELDIANA: ZOOLOGY o O. degui A Q bridgesi • O.lunatus Fig. 1 . Localities for Octodon degus (O), O. bridgesi (A), and O. lunatus (•). The "?" symbols represent uncertain and tentative identifications of captured specimens (see text). clearly indicate a distributional overlap of at least O. degus and O. lunatus. The distribution pattern of these three species has been obscured by the uncertain identification oi Octodon specimens captured in Nuble and Cau- quenes provinces. These have been tentatively as- signed to O. lunatus (cf. Tamayo & Frassinetti, 1 980) and O. bridgesi (Rodriquez &. Herrera, 1 983), respectively (fig. 1). In our opinion this confusing situation reflects the lack of good diagnostic char- acters to separate O. lunatus from O. bridgesi. If the reported identifications prove correct, they would indicate that the distribution of O. bridgesi includes the Cordillera Occidental (Cordillera de la Costa), at least between 35°30'S and 37°00'S and/or that the southernmost limit of O. lunatus extends farther south along the coastal range than previously recognized. A third possibility is that O. lunatus and O. bridgesi are but a single species with a much larger geographic range. Octodon bridgesi and O. lunatus seem not to burrow as much as O. degus (Greer, 1965; Ipinza et al., 1971), and their distributions are associated with denser, more humid scrub than that where O. degus is found. Octodon bridgesi is nocturnal, and preliminary data indicate that it has a greater evaporative water loss than O. degus (F. Bozino- vic, pers. comm.). Consequently, factors related CONTRERAS ET AL.: BIOGEOGRAPHY OF OCTODONTID RODENTS 403 to water availability may be important in deter- mining the distribution of both O. bridgesi and O. lunatus. Genus Octodontomys Palmer, 1903 The soco, Octodontomys gliroides (Gervais & D'Orbigny, 1 844), is a monotypic species, found only in Andean and sub-Andean zones of south- western Bolivia from La Paz to Potosi, in north- western Argentina from Jujuy to La Rioja (Ca- brera, 1961), and in northeastern Chile only in Tarapaca Province (Mann, 1945; Pine et al., 1979) (fig. 2). We believe that the apparently dis- junct distribution of this species is due only to inadequate sampling. The soco has nocturnal hab- its according to Ipinza et al. (197 1), but is diurnal according to Mann (1978). It inhabits very dry areas characterized by cacti and rock piles where it digs short burrows connected by superficial run- ways. This species eats succulent plants and the bark of resinous shrubs (Mann, 1945). In many characteristics, such as a silky coat, a plantar surface with fine granulations, and an en- larged rostrum, Octodontomys is similar to Abro- coma ( Abrocomidae), perhaps indicating adaptive convergence. Genus Octomys Thomas, 1920 This is certainly the least-known octodontid ge- nus. Woods (1982) included Tympanoctomys bar- rerae in the genus Octomys, so that the genus in- cludes two species: O. mimax and O. barrerae. Octomys inhabits mountainous regions in northwest Argentina in Catamarca, La Rioja, San Juan, and Mendoza provinces (Cabrera, 1 96 1 ). The genus resembles Octodontomys, and the distri- butions of the two overlap in the northern prov- inces of Argentina. Octomys lives in desert scrub habitats and is nocturnal, a burrower, and an her- bivore (Mares & Ojeda, 1982). Genus Aconaemys Ameghino, 1891 According to Pearson ( 1 984) two species are rec- ognized in this genus: A.fuscus Waterhouse, 1 84 1 , and A. sagei Pearson, 1 984. The former is consid- ered to have two subspecies: A. f. fuscus Water- house, 1841, found in the slopes of the south- central Andes, and A. f. porteri Thomas, 1917, found in the southern extreme of the known geo- graphic range of the genus (fig. 3). Unfortunately, type specimens of both forms lack reliable local- ities (Pearson, 1984). Available, geographically reliable representatives of these forms consist of two specimens of fuscus we captured at the confluence of Rios Vergara and Nascimiento (35°08'S, 70°28'W) and eight specimens oi porteri reported by Pearson ( 1 984) from Ruca Malen. The southern form, porteri, is in part distinguished from fuscus by its bicolored tail. A word of caution on the validity of these is pertinent, because besides uncertain type localities (Osgood, 1 943), the main diagnostic character has some variability not yet quantified. The tails of the specimens from Taica attributed to fuscus are bicolored, to some extent approaching the condition described for ^. / por- teri (Pine et al., 1979), and the Ruca Malen spec- imens have moderately bicolored tails, not as the type of y4. / porteri (Pearson, 1984). Aconaemys sagei, a smaller species than A. fus- cus, was recently described from Neuquen Prov- ince, Argentina (Pearson, 1984). According to this author, some specimens from Chile, previously listed under A. fuscus by Osgood (1943), Greer (1965), and Pine et al. (1979), may also belong to this species. However, after examination of 21 specimens from Curico, Nuble, and Malleco (east and west) provinces, we concluded that none of the specimens can certainly be assigned to A. sa- gei. We found considerable overlap between spec- imens from these localities. Thus, we consider all known Chilean specimens as belonging to one species, A.fuscus, which is found along the Andes between 35°S and 41°S and in the coastal Cordi- llera de Nahuelbuta (fig. 3). Aconaemys is fossorial, although to a lesser ex- tent than Spalacopus cyanus. Its timnel systems are more superficial, and its runways resemble somewhat those of voles (Microtus) of the north- em hemisphere (Greer, 1965). This species seems to have a wide habitat tolerance, considering that it has been foimd in Nothofagus and Araucaria forests in regions with high rainfall (Osgood, 1 943; Greer, 1965; Mann, 1978), to above timberline in close association with bunch grasses {Festuca and Stipa spp.) (Contreras & Torres-Mura, pers. comm.). Competitive exclusion is thought to be common among fossorial mammals (Nevo, 1979). This might be the case with Spalacopus and Aconaemys (fig. 3), provided that the latter species replaces the former to the south. However, the range of Aconaemys seems to overlap that of other fossorial 404 HELDIANA: ZOOLOGY Fig. 2. Range of distribution (stippled) of Octodontomys gliroides in Bolivia, Chile, and Argentina. herbivorous rodents, such as Ctenomys maulinus and Ctenomys sp. (see Gallardo, 1979; Pearson, 1984). At present the lack of data does not permit assessment of the actual fine-grain distributions of these two genera. They might be parapatric, but if so, it could be difficult to resolve whether their separation is caused by differences in microhabitat preferences, by competition, or by historical fac- tors of colonization. Genus Spalacopus Wagler, 1832 This genus is monotypic, and the taxonomic validity of its subspecies is debatable (Mann, 1 978; Tamayo & Frassinetti, 1 980). We agree with Mann (1978) that the distinction between S. c. cyanus (Molina, 1 782) from the coast of central Chile and S. c. maulinus Osgood, 1943 from south-central Chile should be carefully reconsidered, because it is based on subtle cranial traits of a small number of specimens, and because individuals vary great- ly. In contrast, the distinction between S. c. cyanus and S. c. poeppigii Wagler, 1832 from the Andes seems to be valid. Andean populations of Spala- copus are phenotypically distinct from coastal ones. Those from the Andes are larger (Reig et al., 1 972; Yanez & Ztilch, 1981), probably in response to thermal factors and food availability (Contreras, 1983, 1986). These two forms also seem to differ in skull, tooth morphology, and color pattern (Reig et al., 1972). Because of the small number of Spa- lacopus in the Lx)ngitudinal Valley (Central Valley) and adjacent areas (see below), we believe that gene flow between the coastal and mountain pop- ulations is probably low, although these rodents do not show differences in G- and C-banding of chromosomes (Ziilch et al., 1982) or in the elec- trophoretic patterns of six blood proteins (Woods & Kilpatrick, pers. comm.). Spalacopus tabanus Thomas, 1 925 was treated as a subspecies of S. cyanus by Osgood (1943). Because the single known specimen has a provenance of "South of Chile" and is quite large, we agree with Reig et al. CONTRERAS ET AL.: BIOGEOGRAPHY OF OCTODONTID RODENTS 405 • S cyanus A A sagei o A fuscus Fig. 3. Localities for Aconaemys fuscus (A), A. sagei (A), and Spalacopus cyanus (•). (1 972) and Tamayo and Frassinetti ( 1 980) that the animal probably represents S". c. poeppigii from the Andes, rather than S. c. cyanus as proposed by Mann (1978). The low variability within Spalacopus has been attributed to the great mobility of its colonies (Reig, 1970). However, studies of its home ranges by radioactive tagging indicate that these are very sta- ble areas (Torres-Mura & Contreras, 1983). Taking the new data presented here into con- sideration, Spalacopus cyanus is probably the best known octodontid in terms of its geographic range (fig. 3). Its populations are found along the Pacific coast from Caldera (27°03'S) down to Quirihue (36°17'S) in Nuble Province and also along the Andes from Alicahue (32°19'S, 70°39'W) to Los Cipreses (34'^rS, 70°29'W) up to above 3,000 m. Small populations are also found in ravines drain- ing into the Longitudinal Valley (Central Valley) from the Cordillera Occidental (Cordillera de la Costa) and the Andes. The altitudinal limit of S. cyanus seems to be set by lack of food rather than by low oxygen critical pressure that allows it to tolerate severe hypoxic conditions resulting from the combination of high altitude and burrowing habits (Contreras, 1983). Within its geographic range, S. cyanus inhabits areas with shrub cover of no more than 60%, which allows the develop- ment of an herb stratum, containing the geophytes and hemicryptophytes that form its main food. 406 FIELDIANA: ZOOLOGY The northernmost population found along the coast is Quebrada Pajonales, located 22 km north of Caldera. In this locality we have found extensive unoccupied burrow systems, similar to those in- dicated by Osgood (1943). Although we failed to capture specimens, we found skull remains in bur- rowing owl pellets. The northern populations of Spalacopus are mainly restricted to the coast, but can also be found inland in the valley of some rivers (fig. 3). These populations can live there because the extreme desert conditions of the Ata- cama Desert are ameliorated along the coast by the formation of fog banks (Tricart, 1 969), and also because extensive marine terraces with alti- tude lower than 300 m exist there, permitting the development of suitable vegetation for Spalaco- pus. This condition disappears along the coast north of 26°30'S where the land sharply rises from sea level to 1 ,000 m or more. Extreme desert-like con- ditions may also determine the northernmost limit of S. cyanus populations along the Andes, a limit which is farther south than along the coast (fig. 3). This may be due to the fact that in this area (about 29*S to 30°S) the desert penetrates the Andes and crosses to Argentina. There it continues southward along the eastern side of the Andes, producing sharp climatic and vegetational changes along the Andes (Arroyo et al., 1982; Villagran et al., 1983) which are not suitable for Spalacopus. The southernmost distributional limit of S. cy- anus may be determined by two factors: (1) an increasing scarcity of open habitats along the coastal ranges approaching the south temperate rain forest and (2) the presence of Aconaemys fuscus on the southern Andes (fig. 3). Spalacopus seems absent from the valleys of Rios Teno and Tinguiririca in the Andes, although the habitat there is not dif- ferent from that farther north. However, Aco- naemys fuscus occurs in these two valleys, there- fore we postulate that the Andean southern limit of S. cyanus is somewhere between the basins of Rios Cachapoal and Tinguiririca, which coincides with the northernmost limit of A. fuscus. Perhaps competitive exclusion is occurring between these species. Discussion The distribution of living organisms is deter- mined by a combination of historical, evolution- ary, and ecological factors. We propose that the present distribution of octodontids is related to the eco-evolutionary history of South America, especially to that of the Patagonian Subregion (Hershkovitz, 1972). The earliest records of South American hystri- cognaths date from the E>eseadan Oligocene (35 MYBP) of Patagonia and Bolivia (Wood & Patter- son, 1959; Patterson & Wood, 1982). The earliest forms were represented by the families Octodon- tidae, Echimyidae, Eocardiidae, Dasyproctidae, Dinomyidae, Chinchillidae, and Erethizontidae. Of these, octodontids appear to be the most prim- itive (Patterson & Wood, 1 982). Early octodontids and echimyids were probably similar to each oth- er, at least in molar structure (Wood & Patterson, 1959). Since the postcranial skeleton of the early octodontids is similar to that of the present day Octodon, we infer that both octodontids and echi- myids were generalized ground-dwelling forms living in the woodland habitats that covered much of southern South America. At that time the Andes had yet to rise, and the landscape was relatively flat and homogeneous from east to west (Webb, 1978), with a small thermal gradient from north to south (Simpson, 1983). During the early Miocene, the Patagonian Subregion still had a humid climate supporting widespread woodlands (Menendez, 1961). From the Miocene this pattern changed drastically, but not suddenly, with the Andean orogeny, the for- mation of the Humboldt Current and of the Pacific Anticyclone. The uplift of the Andes permitted the maintenance of the humid forest on the west- em slope, and the resulting rain shadow brought about the replacement of forest by savannas, grass- land, and steppes to the east of the Andes, thus producing the present pampas. On the other hand, the Humboldt Current brought about the desert- ification of southern Peru and northern Chile, leading to the formation of the Atacama Desert, the southern retreat of forest, and the establish- ment of mediterranean scrub and sclerophyllous forest in central Chile. As a result of these events, two relatively xeric areas and two mesic areas were formed: (1) a desert scrub in the Atacama region, (2) a steppe vegetation in Patagonia, (3) a high- altitude steppe in the altiplano, and (4) an in- creasingly mesic gradient south of the Atacama Desert, with vegetation ranging from scrublands to temperate rain forests. Since that time, except for fluctuations due to Pleistocene glaciations, this pattern has remained basically the same. The co-occurrence of octodontids and echi- myids in the Patagonian Subregion persisted through the Miocene (Pascual et al., 1965). At this CONTRERAS ET AL.: BIOGEOGRAPHY OF OCTODONTID RODENTS m 10"- -30" Octodontidae Ctenomyidae so* 90^ 50° 40» 20» Fig. 4. Map of southern South America, showing the primarily eastern and western distributions of the Cteno- myidae and Octodontidae, respectively. time octodontids were still similar to the earliest known forms in the family, but echimyids had clearly diverged from the primitive condition and had diversified into at least three groups (Wood & Patterson, 1959; Patterson & Wood, 1982). Un- fortunately, echimyid fossil remains consist only of cranial fragments, with no postcranial elements. However, they were probably in a morphologically intermediate stage between the early generalized condition and the more scansorial-arboreal forms of today. The fossil record indicates that echimyids dis- appeared from Patagonia during the early Plio- cene. Subsequently, octodontid diversity in- creased markedly during the Pliocene in the pampas region (Patterson & Pascual, 1972). The reasons why echimyids disappeared, while octo- dontids remained in the Patagonian Subregion, are unclear. It is likely that echimyids had become more closely associated to tropical or subtropical forests and retreated to the north with them. This would explain the absence of echimyids from the temperate rain forest that became established in the south, west of the Andes. This forest developed from the most cold-adapted elements of the aus- tral-antarctic flora (Van der Hammen & Cleef, 408 HELDIANA: ZOOLOGY 1 983). A link between echimyids and tropical hab- itats is also suggested by their establishment in tropical North America after the Great American Interchange, but their failure to occupy the tem- perate zones of that continent (Patterson & Pas- cual. 1972). During the Upper Pliocene, fossorial octodon- tid forms, not previously known, appear on the steppes. These fossorial forms belong to the family Ctenomyidae and were derived from octodontids (see Reig & Kiblisky, 1969; Reig, 1970). The genus Ctenomys first appears in the early Pleistocene of Argentina (Chapadmalalian), but it more likely originated in the late Pliocene (Reig & Kiblisky, 1 969). The several Pleistocene species of Ctenomys and the roughly 30 species at present denote a genus with a high rate of diversification and expansion from its probable center of origin (Chapadmalal?). Ctenomys now extends through- out the Patagonian Subregion and also to the southern part of the Brasilian Subregion of the Neotropics. During its expansion, Ctenomys en- tered Chile at the end of the Pleistocene (Tamayo & Frassinetti, 1980), mainly through mesic areas of the altiplano and the low xeric areas of Pata- gonia. They were thus able to colonize ( 1 ) the mar- ginal and high-altitude zone of northern Chile and southern Peru (an extension of the Bolivian puna), (2) the steppes of Aysen and Magallanes, which are extensions of the Argentinian pampa into Chile, and (3) certain Andean areas in south-central Chile, where Patagonian vegetation interdigitates with montane forest between snow line and tree line. The distribution of Ctenomys thus essentially en- closes the fossorial forms of the Octodontidae (see below) and partially overlaps with some nonfos- sorial octodontids (fig. 4). After a Pliocene expansion, the area occupied by octodontids decreased; they disappeared from the pampas and persisted only along the slopes of the Andes. Unfortunately, there are almost no specimens from the Pleistocene, but some gener- alized ground-dwelling forms of the Holocene are very similar to those of the Oligocene and Miocene (Wood 8l Patterson, 1959). Forms present during the Pleistocene were also probably similar. The low species diversity (monotypy) and the pronounced morphological and physiological spe- cializations to fossorial life found in Spalacopus and Aconaemys indicate that these forms probably arose in situ. Thus, their origin would have been contemporaneous with the development of open areas of sclerophyllous scrub, savanna-like and grasslands vegetation in central Chile following Pleistocene glaciations (Troncoso et al., 1980). Their restricted geographic distribution may be a result of the rapid colonization of suitable habitat by Ctenomys. In summary, the present distribution of octo- dontids seems to be the result of historical inter- actions between the Andean orogeny, the associ- ated climatic and vegetational changes, and the various evolutionary trends of the echimyids, ctenomyids, and the octodontids themselves. Acknowledgments We would like to express our appreciation to F. Jaksic, M. T. Kalin-Arroyo, A. Troncoso, and C. Villagran for valuable comments. We especially thank B. D. Patterson and R. H. Pine for their constructive comments as reviewers. Ms. V. Aguirre and M. L. Uribe provided competent tech- nical assistance. Specimens of Aconaemys were made available to us by A. Spotomo, Universidad de Chile; T. Cekalovic, Universidad de Concep- cion; and M. Gallardo, Universidad Austral de Chile. Partial support was provided by Facultad de Ciencias, Universidad de Talca. This work is in honor of P. Hershkovitz for his great contri- bution to the understanding of South American mammals. Literature Cited Arroyo, M. T. K., C. VillagrAn, C. Marticorena, AND J. Armesto. 1982. Flora y relaciones biogeo- graficas en los Andes del norte de Chile (18''19°S), pp. 71-92. In Veloso, A., and E. Bustos, eds., El ambiente natural y las poblaciones humanas de los Andes del Norte Grande de Chile (Arica, Lat. 18''28'S), Vol. 1. Roslac, Montevideo. Cabrera, A. 1961. Catalogo de los mamiferos de America del Sur. Revista del Museo Argentino de Ciencias Naturales "Bernardino Rivadavia", Ciencias Zoologicas, 4(2): 309-732. CoNTRERAS, L. 1983. EcopHysiology offossorial mam- mals: A comparative study. Ph.D. Diss., University of Florida, Gainesville, 185 pp. . 1986. Bioenergetics and distribution of fos- sorial Spalacopus cyanus (Rodentia): Thermal stress or cost of burrowing? Physiological Zoology, 59: 20- 28. CoNTRERAS, L., AND M. RosENMANN. 1982. Thermo- regulation in the testes oi Octodon degus: A non-scro- tal rodent. Physiological Zoology, 55(2): 144-147. Gallardo, M. 1979. Las especies chilenas de Cteno- CONTRERAS ET AL.: BIOGEOGRAPHY OF OCTODONTID RODENTS 409 my^ (Rodentia, Octodontidae). I. Estabilidad cariotip- ica. Archivos Biologia y Medicina Experimentales, 12: 71-82. Glanz, W. E. 1977. Comparative ecology of small mammal communities in California and Chile. Ph.D. Diss., University of California, Berkeley, 300 pp. Greer, J. K. 1965. Mammals of Malleco Province, Chile. Publications of The Museum, Michigan State University, Biological Series, 3: 51-151. Hershkovitz, P. 1962. Evolution of Neotropical cric- etine rodents (Muridae). Fieldiana: Zoology, 46: 1- 524. . 1972. The recent mammals of the Neotropical Region: A zoogeographic and ecological review, pp. 31 1-432. In Keast, A., F. C. Erk, and B. Glass, eds., Evolution, Mammals, and Southern Continents. State University of New York Press, Albany, 543 pp. Ipinza, J., M. Tamayo, and J. Rottmann. 1971. Oc- todontidae en Chile. Noticiario Mensual Museo Na- cional de Historia Natural, Santiago, Chile, 183: 3- 10. Jaksic, F. M., J. L. YaRez, and E. R. Fuentes. 1981. Assessing a small mammal community in central Chile. Journal of Mammalogy, 62(2): 391-396. Mann, G. 1945. Mamiferos de Tarapaca. Observa- ciones realizadas durante una expedicion al alto norte de Chile. Biologica, 2: 23-134. . 1 964. Compendio de Zoologia. Vol. I. Ecologia y Biogeografla. Santiago, Chile, 37 pp. -. 1978. Los pequeiios mamiferos de Chile. Gay- ana: Zoologia, 40: 1-342. Mares, M. A., and R. A. Ojeda. 1982. Patterns of diversity and adaptation in South American hystri- cognath rodents, pp. 393-432. In Mares, M. A., and H. H. Genoways, eds.. Mammalian Biology in South America. A Symposium Held at the Pymatuning Lab- oratory of Ecology, May 10-14, 1981. Special Publi- cation Series, Pymatuning Laboratory of Ecology, University of Pittsburgh, 6: 1-539. MENfeNDEZ, C. A. 1961. Estipitepetrificadodeunanueva Cyatheaceae del Terciario de Neuquen. Boletin de la Sociedad Argentina de Botanica, 9: 331-358. Meserve, p. L., R. E. Martin, and J. M. Rodrjguez. 1983. Feeding ecology of two Chilean caviomorphs in a central mediterranean savanna. Journal of Mam- malogy, 64(2): 322-325. . 1984. Comparative ecology of the caviomorph rodent Octodon degus in two Chilean mediterranean type communities. Revista Chilena de Historia Nat- ural, 57: 79-89. Nevo, E. 1979. Adaptive convergence and divergence of subterranean mammals. Annual Review of Ecology and Systematics, 10: 269-308. Osgood, W. H. 1943. The mammals of Chile. Field Museum of Natural History, Zoological Series, 30: 1- 268. Pascual, R., J. PisANO, AND E. J. Ortega. 1965. Un nuevo Octodontidae (Rodentia, Caviomorpha) de la formacion Epecuen (Plioceno Medio) de Hidalgo (Pro- vincia de La Pampa). Consideraciones sobre los Cten- omyinae Reig, 1958, y la morfolo^a de sus molari- formes. Ameghiniana, 4: 19-29. Patterson, B., and R. Pascual. 1972. The fossil mammal fauna of South America, pp. 247-309. In Keast, A., F. C. Erk, and B. Glass, eds.. Evolution, Mammals, and Southern Continents. State University of New York Press, Albany, 543 pp. Patterson, B., and A. E. Wood. 1982. Rodents from the Deseadan Oligocene of Bolivia and the relation- ships of the Caviomorpha. Bulletin of the Museum of Comparative 2^ology, 149(7): 371-543. Pearson, O. P. 1958. A taxonomic revision of the rodent genus Phyllotis. University of California Pub- lications in Zoology, 56(4): 391-496. . 1984. Taxonomy and natural history of some fossorial rodents of Patagonia, southern Argentina. Journal of Zoology (London), 202: 225-237. Pine, R. H., S. D. Miller, and M. L. Schamberger. 1979. Contributions to the mammalogy of Chile. Mammalia, 43: 339-376. Reig, O. A. 1970. Ecological notes on the fossorial octodont rodent Spalacopus cyanus (Molina). Journal of Mammalogy, 51(3): 592-601. . 1981. Teoria del origen y desarrollo de la fauna de mamiferos de America del Sur. Monografia Na- turae. Museo Municipal de Ciencias Naturales Loren- zo Scaglia, Mar del Plata, Argentina, 162 pp. Reig, O. A., and P. Kiblisky. 1969. Chromosome multiformity in the genus Ctenomys (Rodeptia, Oc- todontidae). Chromosoma (Berlin), 28: 211-244. Reig, O. A., A. Spotorno, and R. Fernandez. 1972. A preliminary survey of chromosomes in populations of the Chilean burrowing octodont rodent Spalacopus cyanus (Molina) (Caviomorpha, Octodontidae). Bio- logical Journal of the Linnean Society, 4: 29-38. Rodriguez, J. A., and L. A. Herrera. 1983. Evalu- acion del daiio causado por Octodon bridgesi en plan- taciones de Pinus radiata, p. 163. Actas 92 Congreso Latinoamericano de Zoologia, Arequipa, Peru. Rosenmann, M. 1977. Regulacion termica en Octodon degus. Medio Ambiente, 3: 127-131. Rosenmann, M., and P. R. Morrison. 1975. Meta- bolic responses of highland and lowland rodents to simulated high altitude and cold. Comparative Bio- chemistry and Physiology, 51A(4): 523-530. Rosenmann, M., G. Ruiz, and A. Cerda. 1981. Re- laciones de actividad espontanea de Octodon degus y Abrocoma benneti en Chile central. Medio Ambiente, 5(1-2): 125-129. Rowlands, L W., and B. J. Weir (Eds.). 1974. The biology of hystricomorph rodents. Symposium of the Zoological Society of London, 34: 1-482. Simpson, B. B. 1983. An historical phytogeography of the high Andean flora. Revista Chilena de Historia Natural, 56: 109-122. Tamayo, M., and D. Frassinetti. 1980. Catalogo de los mamiferos fosiles y vivientes de Chile. Boletin Museo Nacional de Historia Natural, Chile, 37: 323- 399. Torres-Mura, J. C, and L. C. Contreras. 1983. Ambito de hogar en Spalacopus cyanus (Rodenlia, Octodontidae). Archivos de Biologia y Medicina Ex- perimentales, 16(2): R-223. Tricart, J. 1969. Le modele des region seches. In 410 HELDIANA: ZOOLOGY Tricart, J., and A. Cailleux, eds., Traite de Geomor- phologie, IV. SEDES, Paris, 472 pp. Troncoso, a., C. Villagran, and M. MuSoz. 1980. Una nueva hipotesis acerca del origen y edad del bos- que de Fray Jorge (Coquimbo, Chile). Boletin Museo Nacional de Historia Natural, Chile, 37: 1 17-152. Van der Hammen, T., and A. M. Cleef. 1983. Datos para la historia de la flora andina. Revista Chilena de Historia Natural, 56: 97-107. VillagrAn, C, M. T. Kalin Arroyo, and C. Marti- corena. 1983. Efectos de la desertizacion en la dis- tribucion de la flora andina de Chile. Revista Chilena de Historia Natural, 56: 137-157. Webb, S. D. 1978. A history of savanna vertebrates in the New World. Part II: South America and the Great Interchange. Annual Review of Ecology and System- atics, 9: 393-426. Webb, S. D., and L. G. Marshall. 1982. Historical biogeography of Recent South America land mam- mals, pp. 39-52. In Mares, M. A., and H. H. Geno- ways, eds.. Mammalian Biology in South America. A Symposium Held at the Pymatuning Laboratory of Ecology, May 10-14, 1981. Volume 6, Special Pub- lication Series, Pymatuning Laboratory of Ecology, University of Pittsburgh, Linesville, Pa., 539 pp. Wood, A. E., and B. Patterson. 1959. The rodents of the Deseadan Oligocene of Patagonia and the be- ginning of South American rodent evolution. Bulletin of the Museum of Comparative Zoology at Harvard University, 120(3): 279-428. Woods, C. A. 1982. The history and classification of South American hystricognath rodents: Reflections on the far away and long ago, pp. 377-392. In Mares, M. A., and H. H. Genoways, eds.. Mammalian Biology in South America. A Symp>osium Held at the Pyma- tuning Laboratory of Ecology, May 10-14, 1981. Spe- cial Publication Series, Pymatuning Laboratory of Ecology, University of Pittsburgh, 6: 1-539. Woods, C. A., and D. K. Boraker. 1975. Octodon degus. Mammalian Species, 67: 1-5. YAfiEZ, J. L., AND F. M. Jaksic. 1 978. Historia Natural de Octodon degus (Molina) (Rodentia, Octodontidae). Publicacion Ocasional Museo Nacional de Historia Natural, Santiago, Chile, 27: 3-11. YASez, J., AND R. ZuLCH. 1981. Morfometria de tres poblaciones de Spalacopus cyanus Rodentia. Archivos de Biologia y Medicina Experimenlales, 14(3): R-304. ZuLCH, R., L. E. Walker, and J. Pincheira. 1982. Comparacion de los cariotii>os bandeados G y C de las formas costera y andina de Spalacopus cyanus y Octodon degus (Rodentia, Octodontidae). Archivos de Biologia y Medicina Experimen tales, 15(2): R-164. CONTRERAS ET AL.: BIOGEOGRAPHY OF OCTODONTID RODENTS 411 I Population Dynamics and Ecology of Small Mammals in the Northern Chilean Semiarid Region Peter L. Meserve and Eric Le Boulenge ABSTRACTS Small mammal populations were studied in a northern Chilean semiarid thorn scrub com- munity for 17 months during a period of sparse rainfall (27-76 mm annually, 1973-1975). A permanent 1.4-ha grid was trapped for four consecutive nights bimonthly with mark-and- recapture techniques, and standard data were taken for the estimation of population parameters. Snap-trapping in similar habitat provided reproductive information. Seven species were cap- tured in the field (six rodents, one marsupial), but only four of them yielded sufficient captures for a quantitative study (three sigmodontines, one hystricomorph). Reproduction was strongly seasonal, starting and ending earlier in Octodon degus and Akodon longipilis (July-November) than in Phyllotis darwini and Akodon olivaceus (September-January for the former, September- November for the latter). The latter two species may produce more than one litter per repro- duction period. Seasonal reproduction, generally high trappability, and clear distinction between the weight fluctuations of adults and juveniles— at least during the first two censuses after the latter appear— allowed for a classification of individuals into a cohort of adults, one of young of the year and a "cohort" of probable immigrants. Besides enumeration techniques which give an overall view of population trends, we used the Modified Calendar of Captures technique to describe the dynamics of cohorts and estimate the numbers of nontrappable juveniles from the date of their first capture back to their birth period. Akodon olivaceus had low survival in reproductive months and low recruitment in the second year of study; this resulted in a steady population decline through the study period. Akodon longipilis had a high survival rate and low recruitment in both years, and showed very constant population numbers. Phyllotis darwini and O. degus had a generally lower survival, but showed a high potential rate of reproduction, albeit with low juvenile survival. Dispersal is probably a crucial feature in the demography of these two species. Despite changes in the relative numbers and proportions of the component species, the fauna in this community was remarkably constant in overall numbers during four years of observation. Since the principal species studied here are widely distributed, physio- logically and morphologically unspecialized for arid environments, and represent diverse trophic specializations, ecological adaptations and life history characteristics seem more important for their persistence and success than are biogeographical considerations or evolutionary pread- aptations. Poblaciones de pequenos mamiferos fueron estudiadas en una comunidad semiarida de matorrales espinosos en el norte de Chile por unos 17 meses durante un periodo de escasas From the Department of Biological Sciences, North- em Illinois University, DeKalb, IL 60115-2861 (Me- serve), and Unite d'Ecologie, Universite Catholique de Louvain, Place Croix du Sud, 5, 1348 Lx)uvain-la-Neuve, Belgium (Le Boulenge). MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 413 lluvijas (27-76 mm anuales, 1973-1975). Una red de trampas cubriendo 1.4 ha fue accionada por cuairo noches consecutivas bimensualmente. Los datos tipo para determinar los parametros de poblacion se obtuvieron por la tecnica de captura, marquaje y recaptura. El uso de trampas de golpe en similar habitat proveyo informacion reproductiva. Un total de siete especies fueron capturadas en el campo (seis roedores y un marsupial), pero solo cuatro de ellos produjeron suficientes capturas para un estudio cuantitativo (tres sigmodontines, un hystricomorfo). La reproduccion fue marcadamente estacional, empezando y terminando mas temprano en Oc- todon degus y Akodon longipilis (Julio-Noviembre) que en Phyllotis darwini y Akodon olivaceus (Septiembre-Enero para el primero, Septiembre-Noviembre para el ultimo). Las ultimas dos especies pueden producir mas que una camada por periodo de reproduccion. Estacionalidad de reproduccion, generalmente alta probabilidad de captura y clara distincion entre las fluc- tuaciones de peso de los jovenes y de los adultos durante los dos primeros censos despues de la apariencia de los primeros, permitieron una clasificacion de individuos dentro de una cohorte de adultos, una de los jovenes del ario y una "cohorte" de probables inmigrantes. Ademas de las tecnicas de enumeracion, las cuales dan una vista general de las tendencias de la poblacion, usamos el Calendario Modificado de Capturas para describir las dinamicas de cohortes y estimar los numeros de jovenes no trampeables desde la fecha de su primera captura hacia atras al periodo de nacimiento. Akodon olivaceus tuvo una supervivencia baja en meses reproductivos y bajo reclutamiento en el segundo aiio de estudio, lo que result© en un decrecimiento continuo de sus numeros. Akodon longipilis tuvo una alta tasa de supervivencia y bajo reclutamiento en ambos aiios y mostro numeros muy constantes de poblacion. Phyllotis darwini y O. degus mostraron una supervivencia generalmente baja, pero una alta tasa potencial de reproduccion, sin embargo con alta tasa de mortalidad juvenil. Dispersion es probablemente una caracteristica crucial en la demografia de estas dos especies. A pesar de los cambios en los numeros y proporciones relativas de las especies componentes, la faima de esta comunidad fue notable- mente constante en numeros globales durante cuatro aiios de observacion. Ya que estas especies son ampliamente distribuidas, fisiologicamente y morfologicamente no especializadas para ambientes aridos, y troficamente diversas, adaptaciones ecologicas y caracteristicas de historia de vida parecen mas importantes para su persistencia y exito que consideraciones biogeograficas o preadaptaciones evolutivas. Populafoes de pequenos mamiferos que habitam uma comunidade semi-arida de arbustos espinhosos no Chile, foram estudadas durante 17 meses, num periodo com pouca chuva (27- 76 mm anuais entre 1973 e 1975). Durante quatro noites consecutivas, a cada dois meses, foram armadas armadilhas ao longo de uma grade permanente de 1 ,4 hectares. Foram usadas as tecnicas de marca9ao e recaptura (mark and recapture), e todos dados convencionais foram tomados. Para informa96es reprodutivas, usaram-se al^apoes (snap-traps) em um habitat se- melhante. Um total de sete especies foram capturadas (seis roedores e um marsupial), mas apenas quatro renderam numeros suficientemente altos para uma analise quantitativa (tres cricetideos e um histricomorfo). A epoca reprodutiva de cada especie foi altamente relacionada £te epocas do ano, come^ando e terminando mais cedo para Octodon degus e Akodon longipilis (julho a novembro), e mais tarde para Phyllotis darwini e Akodon olivaceus (setembro a Janeiro e setembro a novembro, respectivamente). As ultimas duas especies podem produzir mais de uma cria por periodo reprodutivo. Dada a existencia de epocas reprodutivas bem-definidas, os pesos claramente diferentes dos jovens capturados nos seus primeiros dois censos, e, em geral, a frequente captura dos animais, foi possivel classihcar individuos em grupos de adultos, de jovens-do-ano, e de provaveis imigrantes. Alem das tecnicas de enumera^ao dos animais, que dao uma ideia dos padroes gerais das populagoes, usamos tambem a tecnica de Calendario Modificado de Capturas (Modified Calendar of Captures), para descrever dinamicas entre os grupos de idades diferentes, e para estimar o numero de jovens nao-capturaveis. Os resultados indicam que A. olivaceus sofreu baixa sobrevivencia durante os meses reprodutivos e baixa recruta durante o segundo ano de estudo, provocando assim um declineo constante na sua popula(ao durante o periodo de observafao. Akodon longipilis teve alta sobrevivencia mas 414 nELDIANA: ZOOLOGY baixa recruta durante os dois anos, mostrando entao um numero de popula9ao muito estavel. Phyllotis darwini e O. degus tiveram um nivel geral de sobrevivencia mais baixo, porem mos- traram um alto potencial reprodutivo embora com baixa sobrevivencia juvenil. Dispersao e provavelmente um fator critico na demografia destas duas especies. Apesar das mudancas nos numeros relativos das popula^oes e nas propor^oes das especies componentes, a fauna desta comunidade semi-arida foi notavelmente estavel durante os quatro anos de observa9ao. Como as especies principais deste estudo representam animais de vasta distribui^ao, naoespecializados fisiologica ou morfologicamente para ambientes aridos, e de niveis troficos diversos, seu sucesso e sua persistencia parecem ser mais ligados a adaptafoes ecologicas, e a caracteristicas biologicas vitais (life history), do que a biogeografia, ou a preadapta^oes evolutivas. Introduction The warm arid and semiarid regions of South America are of relatively recent origin. The Ar- gentine Monte desert was formed in the Miocene- Pliocene period and was strongly influenced by the rise of the Andes (Solbrig, 1976). The western Pa- cific coastal desert and central mediterranean zone are even more recent, having been strongly influ- enced by the progressive cooling of the Pacific Ocean, the maximum rise of the Andes in the late Pliocene, and montane Pleistocene glaciations (Axelrod, 1973; Simpson, 1975a,b; Solbrig, 1976). The small mammal faunas of both the Monte and Pacific coastal deserts are typically depauperate in comparison to other desert faunas, are inhabited by few widely distributed species, and have few autochthonous forms (Osgood, 1 943; Baker, 1967; Mares, 1975a, 1976, 1980). Even more indicative of a recent origin of these faunas, their component species are relatively unspecialized morphologi- cally and few of them have developed the capacity to exist without free water, as is typical in other deserts (Mares, 1975a,b, 1980; Meserve, 1978). This has not prevented the development of con- vergence in ecologically related morphological traits for members of the Sonoran and Monte des- ert faunas (Mares, 1975a, 1976, 1980). The evi- dence for such convergence is less convincing for small mammals of the Chilean and Califomian semiarid and mediterranean zones, perhaps be- cause they may be more recent in origin and have a more generalized morphology and broader hab- itat ranges; also, there may be many alternative ecological strategies in newly formed ecosystems (Glanz, 1977; Glanz & Meserve, 1982). Fulk's pioneering study (1975) of a Chilean semiarid small mammal community, involving qualitative comparisons with similar North Amer- ican communities, emphasized the lower overall diversity and large annual fluctuations of the fauna there. He concluded that there was little evidence for intrinsic mechanisms of population regulation in the species studied. Fulk's study, however, was conducted during a period of exceptionally high rainfall; here, as well as elsewhere in Chile (Glanz, 1977; Pefaur et al., 1979), high population den- sities of many small mammal species were re- corded. In addition, a ubiquitous caviomorph ro- dent, Octodon degus (family Octodontidae), frequently reported from here and elsewhere in north-central Chile, was virtually absent from his site. Here we report the results of a subsequent study of the same small mammal community followed by Fulk (1975), for the years 1973-1975. We em- phasize the patterns of population changes ob- served over some 1 7 months of both live- and snap-trapping, as well as the general implications of these results for viewing population dynamics of the principal species and faunal composition of Chilean semiarid small mammal communities. Study Area and Methods The study site was located in a semiarid thorn scrub community in Parque Nacional Fray Jorge, Coquimbo Province (IV Region), Chile (71°40'W, 30°38'S, 200 m). This area lies on the northern fringe of the Chilean mediterranean zone known as the "Norte Chico" and on the southern edge of the Pacific coastal desert. The site is situated in an interior valley ("Quebrada de las Vacas") on the east side of a coastal range (500 m) and about 5 km east of the Pacific Ocean. Descriptions of the vegetation have been given in Fulk (1975) and Meserve (1981a); briefly, the community is char- acterized by spiny drought-deciduous and ever- MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 415 green shrubs, a sparse herbaceous understory, and open sandy areas between shrubs. The community has been termed the Porlieria chilensis-Proustia pungem-Adesmia bedwellii association for its most characteristic shrubs (Munoz & Pisano, 1 947). The flora of this region combines elements of the dry western Andean slopes and a few species from the Monte on the east side of the Andes (Sarmiento, 1975). The absolute cover of major plant cate- gories as determined from line transect analysis of the live-trap grid conducted in November 1973 was 59.6% shrubs, 21.7% grasses and forbs, and 46.0% bare ground (total exceeds 100% due to overlapping projections of shrub canopies on other categories). A subsequent reanalysis of some sta- tions in September 1974 showed no significant change in total shrub cover. Fulk (1975) reported a total shrub cover of 44% for his site, located approximately 500 m away across a dirt road. A fundamental characteristic of this region is the presence of moderately warm temperatures, high insolation, and extremely variable precipi- tation which falls mostly (90%) between the months of May and September. Prior to 1965, the average annual precipitation was about 127 mm; in 1965, an exceptionally heavy rainfall of 326 mm was recorded (Kummerow, 1966). For the eight-year period of 1969-1976, during which continuous records have been kept in the park, precipitation averaged 68.9 ± 78.5 mm (SD); in the principal year of Fulk's study (1972-1973), precipitation was 255.3 mm. For the three subsequent years dealt with in this study (1973-1976), precipitation was 75.8, 26.5, and 37.0 mm. Kummerow (1966) showed that coastal fog can contribute up to ten times the amount of moisture available from pre- cipitation on higher (500 m) west-facing mountain slopes; in the interior valley (200-m elevation), the impact of coastal fog is considerably less. Mean maximum temperature in the warmest month (January) is 24°C, and mean minimum tempera- ture in the coolest month (July) is 4°C (all weather data are from the park headquarters located about 5 km southeast of the study site and collected by the Oficina Meterologica de Chile, Santiago). Between November 1973 and January 1975, a 1.4-ha live-trapping grid with 48 stations 20 m apart (6x8 configuration) was censused at bi- monthly intervals for four consecutive nights. Two medium Sherman live traps baited with rolled oats were placed within 1.5 m of each station and stan- dard mark-and-recapture techniques employed, using toe-clipping or ear-tagging. Observations of species, body weight (to the nearest 0.5 g), sex. sexual condition (perforate or imperforate for fe- males; scrotal or abdominal testes for males), evi- dence for reproduction (obviously pregnant or lac- tating females), number, and trap location were taken during handling. Traps were closed during the day and reopened in the late afternoon from November 1973 through March 1974; thereafter, traps were left open for at least one day each cen- sus, with twice-daily trap checks. Beginning in September 1973, adjacent snap-trapping was con- ducted during each census in similar habitat at least 400 m from the live-trap grid. Usually, a line of 66 Museum Specials and Victor 4-way rat traps was set in pairs approximately 10 m apart for two to four nights. Traps were checked twice daily and animals autopsied for presence of uterine scars and embryos, and relative testis size and position. Stomachs were also retained for subsequent di- etary analysis (Meserve, 1981a). Evidence for reproduction was based on the presence of scrotal males and pregnant and/6r lac- tating females. Observations from both the live- trap grid and autopsy lines were utilized. Although the presence of perforate females may be consid- ered to be evidence of sexual comp>etence (as is the presence of scrotal testes in males), we utilized this mainly to indicate the minimum weight at which females are capable of reproduction and hence an approximate index to developmental age. For nonreproductive periods, developmental age was assessed by comparison with the last mini- mum weight at which evidence of sexual compe- tence (i.e., perforate vagina in females, scrotal testes in males) was observed. This approach was deemed necessary because none of the species studied have distinct juvenile-subadult-adult molts, and there presently exists no satisfactory method for oth- erwise aging the species studied here in live-trap- ping studies. Even an index developed from snap- trapping results would still be dependent on some morphological character, such as weight taken in the field, in order to age live-trapped individuals. Tamarin ( 1 984) and Dueser et al. ( 1 984) have dis- cussed the possible errors inherent in applying body weight criteria to live-trapped animals without in- dependent verification. Live-trap capture records were codified and re- corded on computer for subsequent analysis with the CMR (capture-mark-recapture) package of Le Boulenge (1985a,b) at the University of Louvain computer facility, Louvain-la-Neuve, Belgium. This package produces data summaries such as the Calendar of Captures (Petrusewicz & Andrze- jewski, 1962), sex ratios, reproductive condition, 416 FIELDIANA: ZOOLOGY Minimum Number Known Alive (Krebs, 1966) tabulations, plots of traps visited, and weights at first capture. In addition, higher level analyses are available for trappability analysis, density estima- tions (Manly-Parr, Hayne, and Zippin methods), home range parameters (using the bivariate nor- mal method of area estimation, Mazurkiewicz, 1969; Jennrich & Turner, 1969), disappearance rates, and weight dynamics. Trappability is esti- mated either running "vertically" through the cap- ture matrix to find the proportion of marked an- imals that are captured in a given time period (usually in a census; this is the Manly-Parr [1968] trappability estimator), or running "horizontally" through the recapture data to find the proportion of capture opportunities in which a given individ- ual or group of individuals was effectively caught ("individual trappability" sensu Le Boulenge & Le Boulenge-Nguyen, 1981). The CMR package is available at cost from the second author upon re- quest. In small mammal studies, "classic" statistical methods very often yield unreliable estimates due to inadequate sample sizes or failure to fulfill un- derlying assumptions. Thus, in addition to the Minimum Number Known Alive method, we have calculated population sizes by the Modified Cal- endar of Captures technique whenever possible (Le Boulenge & Le Boulenge-Nguyen, 1981). Ba- sically, the Modified Calendar of Captures tech- nique is equivalent to the Calendar of Captures or Minimum Number Known Alive method, where an individual is counted as present from the first to last capture (Krebs, 1966); however, this tech- nique includes two additional steps. The first con- sists of adjusting the observed residency time of each individual by adding the "mean section of time between successive captures" {sensu Andrze- jewski, 1963, 1 969) before its first and after its last capture (see details in Le Boulenge & Le Boulenge- Nguyen, 1981). The second step consists of count- ing young individuals as present since their ap- proximate birthdate (i.e., in this case, the closest subsequent census) irrespective of the date of their first capture (a procedure called Age-Projected Number Alive by Ford & Pitelka, 1984). This backward projection of numbers of young is ad- justed by the juvenile survival rate, assuming this to be constant from the date of first capture back to the date of birth (Le Boulenge & Le Boulenge- Nguyen, 1981). Direct survival estimates between successive sessions are biased by trappability fluc- tuations and are imprecise when, as often, sample sizes are small; whence survival values used in the above procedure were derived by modeling trap- pability and survival with the method of Clobert et al. ( 1 985). This consists of finding the minimum sufficient multinomial trappability/survival mod- el which adequately fits the observed data. Good- ness-of-fit is assessed by comparing the candidate minimum model to a reference model in which survival and trappability parameters attached to each census are free from constraints. Comparison of two models is made using a log-likelihood ratio, distributed as a x^ variable with degrees of free- dom equal to the difference in the number of free parameters of the two models. The Modified Calendar of Captures method re- lies on the ability to distinguish age cohorts and to separate older cohorts from young bom in situ. The major assumptions underlying the technique are that no individuals living on the plot escape capture entirely (as is also assumed by the Cal- endar of Captures/Minimum Number Known Alive methods) except for young ones prior to first capture, and that the mortality rate of marked and unmarked animals is similar (Le Boulenge & Le Boulenge-Nguyen, 1981). Definitions of the age cohorts that were utilized in this analysis are spec- ified below for each species. The objective of the analyses was to arrive at a more realistic view of changes occurring in the populations. In addition, insight into the nature of population dynamics may be gained. For ex- ample, when the backward projection procedure gave unrealistic estimates of numbers of young present before first capture (that is, far greater or less than the potential for known resident females to produce them), this could be interpreted as in- dicating a major role, respectively, of immigration or emigration. Results Trends of Minimum Numbers Known Alive Figure 1 presents the Minimum Numbers Known Alive on the grid during the study period, for the four principal small mammal species as well as total. Numbers of Akodon olivaceus showed a constant decline throughout 1 5 months of ob- servation; in contrast, those of A. longipilis were remarkably constant, from two to five individuals. Phyllotis danvini showed fairly regular increases during late spring-summer months and declines afterwards. Lacking diurnal trapping prior to May MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 417 160- 140- = 120 H oiooH 80- 60- ^ 40- 20- I J ^J J ^J rJ I. .J I- Nov. 1973 Jan. T Mar. ♦ r May July 1974 Sept. -* Total number "♦ PhyUotis darwini ■ Octodon degus • Akodon olivaceus O Akodon k3ngipriis r Nov. Jan. 1975 Fig. 1. Minimum Number Known Alive trends for Akodon olivaceus, A. lon- gipilis, PhyUotis darwini, and Octodon degus and total numbers (all species in- cluding Abrocoma bennetti, Oryzomys longicaudatus, and Marmosa elegans) during November 1973-January 1975. 1974, earlier estimates of Octodon degus numbers are probably considerable underestimates; never- theless, minimum numbers of degus were consid- erably greater than reported by Fulk (1975). In May 1974, a large number of new O. degus adults were recorded on the grid; most of these (64% of 39 individuals) were not recaptured in subsequent censuses. As this occurred during the beginning of the degu breeding season, when intense activity was observed on the grid following the first winter rains, these were likely peripheral nonresident an- imals (Meserve et al., 1984). Consequently, al- though these individuals were included in Mini- mum Numbers Known Alive figures, they have been excluded from the following demographic and age analysis. Table 1 . Observed sex-ratios of four rodent species given as percentages of females among the individuals captured in each season. Season Nov. *73- March - July- Nov. '74- Total Species Jan. '74 May '74 Sept. '74 Jan. '75 '73-'74 Akodon olivaceus 0.31* 0.44 0.38 0.44 0.34* (64) (25) (2fi) (18) (88) Akodon longipilis 0.14 0.33 0.33 0.67 0.33 (7) (3) (3) (6) (12) PhyUotis darwini 0.30* 0.24 0.S0 0.42 039 (30) (17) (8) (36) (7« Octodon degus 0.36 0.53 0.73 0.49 0.S0 (25) (32) (11) (127) (169) Numbers in parentheses are numbers of individuals captured; total column includes all individuals caught during the study. * Values differing significantly from a 0.5 ratio (chi-square test). 418 HELDIANA: ZOOLOGY lOOn 80 I 60- I 40- q! 20- 0 Akodon olivaceus 11 10 31 13 18 13 9 9 98 16 13 12 8 8 3 3 6 Phyllotis darwini 11 6 14 8 22 10 13 16 12 5 9 5 6 4 26 10 4 4 SNJMMJSNJ SNJMMJSNJ 1973 1974 1975 1973 1974 1975 100- 0, 80- O) 2 60- c a> o 40- 0) '^ 20- Akodon longipilis 63 52 72 31 21 62 75 33 23 100- 80- 2*60- §40- Q. 20- Octodon degus 3 15 7 1 7 12 6 SNJMMJSNJ 1973 1974 1975 SNJMMJSN J 1973 1974 1975 I scrotal (males) nonscrotal (mates) l-^::^V^^ pregnant/lactating (females) nonpregnant and nonlactating (female) Fig. 2. Reproductive trends for Akodon olivaceus, A. longipilis, Phyllotis darwini, and Octodon degus during September 1973-January 1975, based on live- and snap-trap results. Percentages of reproductively active individuals are indicated by crosshatched areas (males) and dotted areas (females) in columns for each sample period; sample sizes are indicated over respective columns. Other species captured on the grid include the caviomorph rodent Abrocoma bennetti (one to five individuals each census except in March and Sep- tember 1 974); Oryzomys longicaudatus (one to five individuals in five of eight censuses); and the mouse opossum Marmosa elegans (single individuals during three censuses). Table 1 presents the sex-ratio for the four most common species in four seasons: late spring (No- vember 1973-January 1974), summer-fall (March- May 1974), winter-early spring (July-September 1974), and late spring (November 1974-January 1975), based on the actual numbers of males and females caught in each season. Interestingly, the observed ratios are consistently biased in favor of males in the three sigmodontines, although sig- nificantly so only for Akodon olivaceus and Phyl- lotis darwini in late spring of the first year and in the total sample of the former species. Nonsignifi- cant results for Akodon longipilis, which shows the most strongly biased ratio, are probably due to the low sample sizes in this species. Octodon degus in contrast seems to have a balanced sex-ratio. Reproduction Reproductive trends for all four species are sum- marized in Figure 2. For Octodon degus only in- formation on females is presented, as relative testis size and position as determined by palpation is probably an inaccurate index of reproductive ac- tivity in this nonscrotal caviomorph rodent. Pregnant or lactating females of Akodon oliva- ceus were briefly present in September-Novem- ber, reproductively active males somewhat longer (fig. 2). A few scrotal males were present in January 1974 but none in January 1975, suggesting earlier termination of sexual competency. Embryo counts for 1974-1975 females (N = 11) averaged 5.6 ± 1.1 (SD), similar to that reported by Fulk (5.6; 1 975) and Greer (5.5; 1 965) in southern Chile, and MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 419 somewhat higher than reported by Pearson (5.1; 1983) in the southern Argentine temperate rain forest. The maturation time for ^. olivaceus is un- known, but Pearson (1983) and Murua (pers. comm.) have reported breeding males and females about two months old. Field-caught male and fe- male A. olivaceus reached sexual maturity at 22- 24 g, but did not breed in the same season. One female bom probably about November-Decem- ber 1973 bred twice, in the following September and November 1974. Data on Akodon longipilis is more limited due to small sample size. This species appeared to ini- tiate and terminate reproduction earlier than either Akodon olivaceus or Phyllotis darwini (fig. 2). Fulk (1975), however, reported pregnant females in February 1973 and juveniles the following May. Similar to trends for other species, reproduction ceased earlier in 1974 than in 1973. Two females had six and four embryos; Greer (1965) and Pear- son (1983) reported mean embryo counts of 3.7 and 3.8, respectively, for southern Chile and Ar- gentina. Minimum weights for perforate females averaged 39.6 g, for scrotal males 47.5 g; no in- dividual reached sexual maturity in less than five months. Pearson ( 1 983), however, reported breed- ing in the same season as birth, and our obser- vations are probably biased without data prior to September 1973. In general, Phyllotis darwini had the longest pe- riod of potential and actual reproduction; scrotal males were recorded in seven out of nine periods and were only absent in May during the 1 974 sam- ples. Fulk (1975) reported no scrotal males later than January during a wet year in Fray Jorge, but they were present in the Santiago area in March of the same year. Pregnant or lactating females were present between September and January; captive females maintained in the Santiago area bred in virtually all months of the year. Embryo count for 1 1 autopsied females was 5.1 ± 1.0 (SD), close to that reported by Fulk (5.2; 1 975) and Pear- son (5.25; 1975) for an "outbreak" population in southern coastal Peru. Young bom in captivity to field-pregnant females from Fray Jorge reached sexual maturity in 60.0 ± 10.2 days at a weight of 40.8 ± 6.5 g (females, N = 3), and 50.3 ± 7.2 days at 46.3 ± 6.4 g (males, N = 6). The lowest weight for a perforate female in the field was 39.2 g, but the next lowest weight was 47.0 g. Minimum weights for nine scrotal males in the field was somewhat lower than in captivity (X = 44.9 ± 5.9 g). Minimum time to first parturition for six cap- tive females was 112.3 ± 3.3 days; more inter- estingly, minimum interval for second litters fol- lowing birth of the first was 16-18 days and apF>eared to alternate between longer intervals of 30-60 days for continuously breeding P. darwini in captivity (captivity data from Le Boulenge, un- publ. data). Recently, this phenomenon was con- firmed independently in pregnant field-caught fe- males from Fray Jorge which gave birth in captivity to a second litter only 16 days after the first (M. H. Gallardo, pers. comm.). In the Fray Jorge pop- ulation, males probably bom in August-Septem- ber reached sexual competency by November. Octodon degus, as Akodon longipilis females, initiated and terminated reproduction earlier than those of Phyllotis darwini and A. olivaceus. Preg- nant females were found only in June and July; thereafter only lactating ones were captured. With a gestation of 90 days (Weir, 1970, 1974), females probably conceived in May-June and gave birth in August-September. There was no evidence of a postpartum estrus or a second litter, as reported in Fray Jorge for 1972-1973 and in central Chile (Fulk, 1975; Rojas et al., 1977; Meserve et al., 1984). Embryo count for nine individuals was 5.7 ± 1.3, slightly greater than the 5.3 figure given for 1974 Fray Jorge females only, or for Santiago area degus (Meserve et al., 1 984). Woods and Bor- aker (1975) reported litter sizes of 6.8 for degus in laboratory colonies. Young did not breed in the same season of birth in Fray Jorge; the minimum weight for perforate females was 1 27.5 g, and field- caught individuals did not reach this weight in less than 100 days of life (including a minimum esti- mated pre-weaning time of 28 days; Weir, 1970). Cohort Definition As live-trapped animals could not be aged di- rectly using classic age-estimation techniques be- cause of poor information on these species, we relied on several sources of evidence, including the seasonality of reproduction, body weight dis- tributions, and sexual maturity of individuals at first capture. The study started at or near the end of the 1973-1974 reproductive season for most species; thus, young of the year had already reached adult size and/or sexual maturity by November 1973-January 1974 and could not be distin- guished from older adults except in the case of Akodon olivaceus. For this species, there was a clear separation between sexually active, heavier individuals which characteristically lost weight (pregnant females excluded) through January 1 974 420 HELDIANA: ZOOLOGY O K, ,3 Females #K^3 Males It Kj^ Females • K74 Males * K 74 Females D Ko Males ir Kq Females Fig. 3. Body weight trends for co- horts of Akodon: top, A. olivaceus, and bottom, A. longipilis. Meanings of sym- bols are defined in legend; see text for cohort definitions. Only two cohorts were defined for A. longipilis. Vertical line in- dicates interval of one standard error of the mean. 40- 20 0 60 40- 20- K it .t-% *fr J I L f~->>. r- N 1973 --\ ,-+-+'' * . D ~r r M J 1974 J 1975 (fig. 3) and were significantly heavier than a group of smaller, mostly sexually inactive individuals which gained weight steadily; these were desig- nated the K<73 and K73 cohorts, respectively. Thus, K<73 individuals were considered adults probably bom early in the 1973 reproductive season (about September) and reaching weights of over 27.5 g by November 1973-January 1974. The K73 in- dividuals were bom late in the reproductive season and weighed less than 27.5 g then. After January 1974, weights of both cohorts converged, being virtually identical by July (fig. 3). The third cohort (K74) consisted of individuals bom in the 1974 reproductive season about September that entered the trappable population in November 1974-Jan- uary 1975. Individuals caught for the first time at adult weights (> 27.5 g) after January 1974 were classified as unknown (Ko) individuals and prob- able immigrants; such a procedure is conservative, and particularly justified if trappability is high (see next section). For the remaining species, only two cohorts were distinguished— a cohort of adults bom before the beginning of the live-trapping study (K573), and a cohort of younger, lower weight individuals ap- pearing in September 1974-January 1975 (K74). In addition, a cohort of probable immigrants (Ko) was distinguished. Figures 3 and 4 show the weight trends for the remaining three species; in all cases, a clear difference in body weight existed between K573 and K74 individuals. The Ko group deserves some additional com- ments. Single new adults of Akodon longipilis ap- peared only in November 1 974 and January 1 975. On the other hand, new adults of Phyllotis darwini were captured in November 1973 and January 1 974 which were far heavier than the average K573 individuals (fig. 4); although these were probably K573 individuals, they were classified in the Kq cohort, as were all subsequent new adults. Finally, due to inadequate sampling prior to May 1974, all Octodon degi4S individuals marked through May MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 421 150- 100 § w 100- 2 50 D ir ^-.-~..^=z^ZZ---. -+—-■5:— N 1973 ♦ r M J 1974 • J 1975 Fig. 4. Body weight trends: top, co- horts of Octodon degus, and bottom, co- horts ofPhyllotis darwini. Symbols are as in Figure 3. Only two cohorts were de- fined for each of these species. ^ were considered K573 cohort members, and all subsequent new adults (recorded from September on) as Ko individuals and hence probable immi- grants. Trappability Any evaluation of population dynamics in free- living animals requires an estimate of trappability. This may vary between species, sex, age group, and season, and hence should be estimated when- ever possible. Enumeration methods such as the Minimum Number Known Alive have an inherent requirement of high trappability between sessions/ censuses. Other estimation methods, such as the JoUy-Seber and derived models and the Modified Calendar of Captures, allow correction for variable trappabilities during the measurement period. Here, daily trappability and seasonal trappability during each of the four three-month periods out- lined in Meserve (1981b) were determined: late spring (November 1973-January 1974), summer- fall (March-May 1974), winter-early spring (July- September 1 974), and late spring (November 1 974- January 1975). Daily "individual trappabilities" (see Study Area and Methods section) were esti- mated for each sex and cohort in each season; the results are presented in Table 2. Monthly or be- tween-session trappabilities were determined by the method of Manly and Parr (1968). Finally, seasonal variations in daily trappability were eval- uated using the modeling technique of Clobert et al. (1985). Akodon olivaceus trappability might be consid- ered homogeneous only in the summer-fall months (March-May) and in the winter months (July-Sep- tember; x^ = 1-26, 2 d.f., NS); otherwise, monthly values were significantly heterogeneous (x' = 10.9, 3 d.f., P < 0.025). Modeled daily trappability was highest in winter (0.75) and lowest in both January sessions (0.48 and 0.18, respectively). This means that the estimated percentage of animals missing a complete census ranged from 55% in January 1974 to 0.4% in winter censuses. The percentage of animals known alive and caught at least once in each census was generally high, however (X = 84.8 ± 14.8%, 1 SD). For Akodon longipilis, trappability could be pooled for all sessions except during January 1 974 when it was significantly less than in other periods (modeled trappability of 0.27 vs. 0.73, x^ = 3.32, 2 d.f., NS for this model, vs. x' = 9.4, 3 d.f, P < 0.025 when all sessions were pooled). Converted into probability of escaping capture in any given 422 FIELDIANA: ZOOLOGY Table 2. Daily trappability by season for Akodon olivaceus, A. longipilis. Phyllotis darwini, and Octodon degus. Season Species, sex, cohort Nov. '73- Jan. '74 March - May '74 July- Sept. '74 Nov. '74- Jan. '75 Akodon olivaceus Adult males Adult females Young (both sexes) 0.67 (63) 0.68(41) 0.48 (86) 0.66 (56) 0.63 (70) 0.75 (57) 0.50(18) 0.90(10) 0.45(11) Akodon longipilis Adult males Adult females Young (both sexes) 0.63 (8) 0.65 (20) 0.64(11) 0.83(12) 0.88 (8) 0.17(6) 1.00(7) 0.45(11) Phyllotis darwini Adult males Adult females Young (both sexes) 0.68 (28) 1.00(5) 0.61 (66) 0.58(12) 0.70 (23) 0.88 (8) 0.33 (30) 0.65 (20) Octodon degus Adult males Adult females Young (both sexes) 0.40 (10) 0.31 (58) 0.29 (35) 0.11(19) 0.14(35) 0.21(63) 0.26 (53) 0.60(189) Numbers in parentheses are numbers of capture occasions. session, that for January was 0.47, and for the remaining months, 0.0 1 . All individual A. longi- pilis known alive on the grid were captured at least once during each census when present. Phyllotis darwini trappabilities were nonsignifi- cantly lower in spring 1974-1975 than in the re- maining sessions (modeled trappability of 0.36 vs. 0.64, x' = 8.82, 4 d.f., 0.1 < P < 0.05). This was equivalent to 1 7% of the animals escaping capture in the spring 1974-1975 censuses, and 2% in the remaining trap sessions. The mean proportion of P. darwini known alive and caught at least once in each session was X = 81.8 ± 16.9%. Daily trappabilities of Octodon degus were rela- tively homogeneous throughout the study (x^ = 7.67, 5 d.f., P> O.IO, NS) with a low mean value of 0.37. The percentage of individuals escaping capture during a session was estimated at 16%. The relative constancy of daily trappability despite the apparently inadequate sampling utilized prior to May 1974 is surprising. The percentages of an- imals known alive and caught at least once each census was quite low, however (X = 39.5 ± 3 1 .9%). In general then, except for Octodon degus throughout the study, and isolated sessions for the remaining species (January 1974 for Akodon lon- gipilis, January 1975 for Akodon olivaceus, No- vember 1974-Januaryl975 ior Phyllotis darwini), we can assume that less than 2% of the individuals went uncaptured throughout a trapping session. The difference between this figure and that ob- tained only from observing occurrences of animals during consecutive censuses may be due to the temporary absence of individuals during specific (and relatively short) censuses with long between- census intervals (Clobert, pers. comm.). Thus, enumeration methods were fairly reliable for es- timating minimum numbers of animals at risk of capture. Demographic Trends Figures 5-8 show the cohort dynamics as re- vealed by the Modified Calendar of Captures tech- nique, as well as the observed survival rates for the defined cohorts and numbers of recruits into the populations of the four principal small mam- mal species. Due to the variable time intervals between bimonthly censuses (52 to 70 days), sur- vival rates were standardized to monthly values. New adults were considered to have entered the population at the start of their corrected residency time and young recruits were estimated to have entered the population at the time of their birth- dates, based on the presence of pregnant or lac- tating females in the population and approximate maturation times. Akodon olivaceus showed low survival in the spring of 1973, but improved survival, especially during the winter, until the 1974-1975 reproduc- tive period (fig. 5). No K<73 individuals survived MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 423 * K.„Males O K.rjFemales * K^jMales * K^jFemales • K^^Males * K, 4 Females Q Kq Males it Ko Females 1.0-1 ^^ ji ^ ^r 1 ^ ^ •♦ r2 - ' i 1 N H-^ V : n- 1 ^ n^ fl* ? «\ -.T? -0 Fig. 5. Top, Estimated number of an- imals by cohort for Akodotfolivaceus us- ing the Modified Calendar of Captures technique during 1973-1975; and bot- tom, observed survival rates (line graphs) and recruitment (dots connected to base line, cumulated by cohort) during the same period. The K.<73 and K73 survival rates have been combined to simplify comparisons (see text for explanation of cohort definitions). Dashed lines indicate estimated entrance of newborn or im- migrant individuals into the ix>pulation. to November 1974, and only one K73 survived to January 1975. The seasonality of survival was confirmed by the modeling technique of Clobert et al. (1985); four different survival values had to be distinguished— 0.38 between consecutive cen- suses in both November-January periods and also in September-November 1974; 0.62 from Janu- ary-March and March-May 1974; 0.31 from May to July; and 0.46 from July-September (x" = 5.7 1, P > 0.10, simplest nonsignificant model). Seven young were marked in November 1974 and Jan- uary 1975, leading back to a projected estimate of 15 young bom in September 1974. This is con- siderably less than the potential production of six pregnant females detected in September (one of which was pregnant again in November). As four of these females disappeared between September and November, a significant portion of the pro- duction of young for that season may have been lost close to birth. Recruitment in the second re- productive season was insufficient to compensate for disappearance rates of older individuals, and the population experienced a decline from 33 in- dividuals the previous January to only 12 in Jan- uary 1975. Akodon longipilis presented a considerably dif- ferent pattern (fig. 6). The K573 cohort persisted into the last month and females especially had high survival rates. Modeled survival may be consid- ered to have been constant throughout the study (x* = 1.05, P > 0.05). The projected production of six new individuals in August-September (pos- sibly from only one female) was sufficient to en- tirely replace the population, especially with the high indicated survival rate for young ^. longipilis. The overall effect was for the population number to be extremely stable in time. The pattern for Phyllotis darwini was less con- sistent (fig. 7). Although observed survival rates seemed to vary from a low spring value to a winter peak and declined again to very low values in the spring of the following year, in fact they were non- significantly different from January through No- vember 1974. Only the periods November 1973- January 1974 and November 1974-January 1975 had extraordinarily low survival rates (average of 1 3%). Due to this poor survival and low trappabil- ities in November-January of the second year, the number of young bom in that reproductive period as estimated by the backward projection proce- dure is unrealistic— over 300 individuals in Sep- tember. Thus, we have only indicated Minimum Numbers Known Alive for the K74 cohort in Fig- MXt 7. This discrepancy between observed number 424 FIELDIANA: ZOOLOGY * Kj, 3 Females * K 74 Females a Kq Males ■it Kq Females Fig. 6. Top, Estimated number of an- imals by cohort for Akodon longipilis us- ing the Modified Calendar of Captures technique during 1973-1975; and bot- tom, observed survival rates and recruit- ment during the same period. Symbols and meaning of dashed lines are as in Figure 5. Only two cohorts were defined for this species (see text). 2 ft. A.I 1.0- of young (which is a minimum figure) and poten- tial production of young of the two reproductive females detected in September-November 1974 indicates a high immigration rate among younger individuals, a trend also observed in adults. In addition, immigrant females could have contrib- uted significantly to reproduction, and the evi- dence for multiple short-interval pregnancies in captivity suggests a high potential production rate of litters by females in general. In view of the poor survival rates of younger individuals, immigration seems to play a major role in the ability of this species to reach high densities rapidly. The results for Octodon degus (fig. 8) indicate a very large influx of juvenile animals occurred in November 1974, probably from a single episode of reproduction initiated in June with birth in early September and weaning in early October. The fact that 97 young were recorded in November when a maximum of 17 adult females were estimated to be present on the grid (including September animals) indicates that the degus were reproducing maximally. However, unless all of the young bom on the plot survived until their first capture, im- migration must be invoked to account for the large numbers of young actually observed. Survival of Fig. 7. Number of animals by cohort for Phyllotis darwini using the Modified Calendar of Captures technique during 1973-1975: top. Minimum Numbers Known Alive figures for the K74 cohort; no backward projection; and bottom, ob- served survival rates and recruitment during the same period. Symbols and meaning of dashed lines are as in Figure 5. Only two cohorts were defined for this species (see text). 30- >20- - 10- 1.0- 25 2 MESERVE &. LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 425 120- 2 80- 40- 0-^ 1.0-1 H — — r^ — 1^ — r" -f — — i — —t- — h N J M M J s N j 1973 1974 1975 Jte mt « -I d^ rlOO Fig. 8. Number of animals by cohort for Octodon degus using the Modified Calendar of Captures technique during 1973-1975: top, numbers in the K74 co- hort as in Figure 7; and bottom, observed survival rates and recruitment during the same period. Symbols and meaning of dashed lines are as in Figure 5. The bro- ken lines in March-May 1974 indicate uncertainty regarding previous pyopula- tion changes due to inadequate sampling. Only two cohorts were defined for this species (see text). adults was fairly high except in late spring (No- vember-January) and winter months (May-July). Disappearance rate of young once they reach trap- pable age was extremely high; this may reflect high dispersal and/or mortality rates. As in the case of Phyllotis darwini, both immigration and low sur- vival of juveniles led to absurd estimates of total production of young by the backward projection procedure. Discussion Semiarid environments are often heterogeneous in space and time. Desert and semiarid mediter- ranean-type communities have frequent events of local extinction (e.g., MacMillen, 1964; M'Clos- key, 1972; Glanz, 1977). Historically, the small mammals of the Chilean arid zone have not been isolated sufficiently long for speciation to occur, nor is there evidence for zoogeographic differen- tiation due to chance colonizations by members of the potential species pool elsewhere in South America. On a local scale, semiarid and arid com- munities in Chile are extremely heterogeneous in numbers of species encountered from year to year (Glanz, 1977; Meserve &. Glanz, 1978). Overall, species number declines monotonically in a north- erly direction due to the influence of declining pre- cipitation and its concomitant effects on primary production and plant community structure (Me- serve & Glanz, 1978). A similar situation exists along an altitudinal gradient on the western side of the southern Peruvian Andes where mammal species number increases with altitude, which in turn is related to vertical vegetation density (pro- file) and perhaps indirectly to precipitation (Pear- son & Ralph, 1978). The frequent observations of irruptions of mice (= ratadas; Hershkovitz, 1962) following unusual rains and subsequent plant growth are further support for the role of extrinsic factors, such as climate, in this region (e.g., Pear- son, 1975; Glanz, 1977; Pefaur et al., 1979). The 1972-1973 "outbreak" of Phyllotis darwini re- ported by Pearson (1975) in coastal southern Peru following a rainfall of 8 1 mm (in a locale averaging 32 mm annually) coincided with similar obser- vations for this and other species in and near Fray Jorge by Fulk (1975) and Pefaur et al. (1979), and in the central to northern Chilean mediterranean zone by Glanz (1977). In view of these well-documented outbreaks, it is significant that our results obtained in years sub- sequent to the above studies, during periods of meager rainfall, demonstrate the persistence of a permanent small mammal fauna after such out- breaks. As a way of viewing the significant changes that occurred in the fauna between the years of Fulk's study (1972-1973) and ours, we have pre- sented data from four years of sampling using both live- and snap-trapping results and an index of numbers of animals caught per 100 trap-nights of effort (table 3). As a convention, to simplify com- 426 HELDIANA: ZOOLOGY Table 3. Numbers of animals/ 100 trap-nights and relative proportions by species for small mammals trapped during four "rainfall" years (see text) during 1972-1976 in Fray Jorge. Trap Species Year Akodon olivaceus Akodon loHgipilis Phyllotis darwini OctodoH degus Other Total July '72-June '73 Live-trap 12.92 (65.0%) 0.80 (4.0%) 5.75 (29.0%) 0.05 (0.3%) 0.33 (1.7%) 19.85 (100%) July '73-June '74 Live-trap 7.62 (42.4%) 0.98 (5.5%) 3.65 (20.3%) 4.43 (24.6%) 1.30 (7.2%) 17.98 (100%) July '73-June '74 Snap-trap 4.96 (23.8%) 1.68 (8.1%) 10.16 (48.8%) 3.36 (16.2%) 0.64 (3.1%) 20.80 (100%) July '74-June '75 Live-trap 4.03 (18.9%) 1.04 (4.9%) 3.26 (15.3%) 12.36 (15.7%) 0.69 (3.2%) 21.38 (100%) July '74-June '75 Snap-trap 2.45 (14.1%) 2.60 (14.9%) 6.88 (39.4%) 5.35 (30.7%) 0.15 (0.9%) 17.43 (100%) July '75-June '76 Snap-trap 2.65 (14.8%) 3.53 (19.6%) 5.29 (29.4%) 3.24 (18.1%) 3.24 (18.1%) 17.95 (100%) parisons, we have used periods of "rainfall years" which run from July of one year through June of the next, recognizing that virtually all precipitation falls during the May-September period, and that subsequent reproduction and recruitment occur in the population from July onward (Fulk, 1975; this study). While differences in trappabilities between species and between live- versus snap-traps would be expected, it is remarkable that total numbers of animals/ 1 00 trap-nights effort was constant from year to year. In the years most strictly comparable due to similar effort and methodology (1 973-1 974 and 1974-1975), mean total numbers of individ- uals/census (taken from fig. 1) were very similar despite large within-"rainfall year" variation (e.g., 76.5 ± 12.6 [1 SD] vs. 81.0 ± 58.5 individuals, respectively). This indicates that, despite within- year and between-year variations in species num- bers, the major changes that occurred over time were in the relative numbers (and proportions) of faunal members. These changes included (1) a de- cline in the Akodon olivaceus population; (2) a relatively constant or even slightly increasing pop- ulation of Akodon longipilis; (3) seasonal changes in the Phyllotis darwini population; and (4) a large increase in the Octodon degus population. The in- crease in numbers of degus is in part an artifact of trapping methodology, as Fulk (1975) did not often trap during the day for this bimodally diurnal to crepuscular species and used a small locally made trap which probably biased against captur- ing the larger degus. The effect of including degus is particularly significant in biomass estimates; for example, total small mammal biomass for the months of November 1974 and January 1975 was 9.2 and 3.1 kg/ha, respectively, of which 83.4% and 82.5% consisted of O. degus individuals. The former figure exceeds the maximum biomass ob- served by Fulk (1975) in any month by almost 100%. The decline of Akodon olivaceus is intriguing, in view of the stable or even increasing number of y4. longipilis. Akodon olivaceus has been shown to be a relatively omnivorous species preferring habitats with less shrub cover and greater herba- ceous cover, while A. longipilis is more insectiv- orous and prefers habitats with higher shrub and litter cover (Fulk, 1975; Glanz, 1977, 1984; Me- serve, 1981 a,b). Geographically, A. longipilis is the more limited species in the arid zone, being re- stricted to lower elevations from La Serena south- wards. In addition, it maintains a lower metabolic rate and has a lower energetic assimilation effi- ciency than A. olivaceus (Rau et al., 1981). Al- though little information is available on water bal- ance relations, it appears to be a more mesic species than A. olivaceus (Meserve, 1978). Elsewhere, in moist primary and secondary growth temperate rain forests, A. longipilis maintains more constant and often numerically superior populations than A. olivaceus (Pearson & Pearson, 1982; Meserve et al., 1982; Murua & Gonzalez, 1985). Thus, the tendency to maintain more stable numbers is rel- atively independent of the community in which it is found. It is tempting to consider A. longipilis a more "K-selected" species in view of its lower density, relatively constant populations, long sur- vivorship, slower maturation rate, and smaller lit- MESERVE &. LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 427 ter sizes. In addition, we observed stable, non- overlapping home ranges of adult females, as did Fulk(1975). Phyllotis danvini populations showed, between years, fairly predictable patterns of population changes and reproduction which differed relatively little, except in magnitude, from the results of Fulk (1975). As litter size observed in our study was similar to that reported by Fulk (1975) and Pear- son (1975) during an "outbreak" year, rapid in- creases in numbers are apparently achieved by rapid maturation of young to reproduce in the same season as birth, and by multiple short-in- terval pregnancies. With a minimum interval of only 16 days between consecutive litters, a post- partum estrus is indicated. Survival rate of older adults between reproductive seasons appears to be zero, so that the entire breeding population in a given year is made up of immigrant adults and of younger individuals bom in the same season. Sur- vival rates are low, but the species appears to col- onize new habitats successfully and maintains high reproductive rates. Phyllotis danvini is the most ubiquitous species in the semiarid and arid regions of northern Chile and appears to be capable of seasonally adjusting its resistance to desiccation and to utilize seeds and succulents (Meserve, 1978; Meserve & Glanz, 1 978). Interestingly, Fulk (pers. comm.) snap-trapped a dry overgrazed habitat dominated by Baccaris sp. and Pronstia pungens immediately east of Fray Jorge in late May-early June 1973, simultaneously with trapping of irri- gated bean fields near La Serena (about 100 km NE) by Pefaur et al. (1979). Whereas Fulk en- countered 86.6% P. danvini in two nights of trap- ping (total of 82 animals, 51.3% trap success), Pe- faur et al. (1979) reported 86.7% Oryzomys longicaudatus (135 animals, 44.4% success) in two nights. Mares ( 1 977a,b) reported that both of these sf)ecies were F>oor water conservers in the Argen- tine Monte desert. Percentages of the catch com- posed ofAkodon olivacens for the two Chilean sites mentioned above were 14.6 and 1.5, resp)ectively, at a time when this species still made up a majority of the Fray Jorge live-trap grid population (Fulk, 1975). The large numbers and relative importance of Octodon degus in this locality during two very dry years was surprising in view of the relatively steno- thermic physiology and high evaporative water loss rates of this species (Rosenmann, 1977). While degus appear to have high population turnover and low juvenile survival rates under such con- ditions, a few individuals have very high repro- ductive potential after surviving to the next breed- | ing season. There is evidence that precipitation | directly triggers reproduction; in 1974 and 1976, reproduction immediately followed the first an- nual rains before the initiation of herbaceous growth (only 6.2 and 17 mm, respectively). With a gestation of 90 days (Woods & Boraker, 1975), such a strategy may be optimal in the northern \ arid zone. On the other hand, Rojas et al. (1977) ! have argued for the principal role of herbaceous growth in initiating reproduction in central Chile, paralleling the situation documented for hetero- myid rodents by Beatley ( 1 969, 1 976) and Van De GraafrandBalda(1973). Other species present in Fray Jorge included Oryzomys longicaudatus, which appears to be spo- radic and irruptive and was transient in the Fray Jorge population. Interestingly, Fulk (1975) doc- umented pregnant O. longicaudatus females in March 1973, beyond the reproductive period for other sigmodontines. Populations of O. longicau- datus might respond to increased seed production following unusual periods of rainfall, as this species is granivorous (Meserve, 1981a; Murua & Gon- zalez, 1 98 1 ). In the southern temperate rain forest, Gonzalez and Murua (pers. comm.) have suggest- ed a relationship between the fluctuations in an- nual seed crops and those of O. longicaudatus pop- ulations. A second species, Abrocoma bennetti, was recorded in low numbers in most censuses but was never recaptured. It is the largest rodent found in this community (200-280 g) and appears to be a highly specialized herbivore with greatest similar- ity to Octodon degus in various aspects of its ecol- ogy (Meserve, 1981a; Meserve et al., 1983). In conclusion then, we have documented the presence of a permanent and successful assem- blage of small mammals living in what would be considered an extremely arid environment by vir- tually any standard. One might predict, without prior knowledge of the history and zoogeography of the area, that this community would be inhab- ited by highly specialized, granivorous, water-in- dependent forms; instead one encounters an as- semblage composed of relatively generalized, trophically diverse, and mostly water-dependent members. Such cases of "nonconvergence" F>er- haps only emphasize the importance of history, zoogeography, and alternative strategies for sur- vival in harsh environments. While extrinsic en- vironmental factors related to food and water availability may ultimately limit populations in 428 HELDIANA: ZOOLOGY this community, there is remarkable stabihty in overall numbers from year to year; most changes seem to involve relative numbers of component species. It may also be significant that Fray Jorge represents a relatively undisturbed native shrub community surrounded by vast areas of badly de- graded and overgrazed shrublands and marginal cultivation. The extent to which environmental degradation elsewhere in arid Chile influences the patterns of population change, extinction, and community instability reported in previous stud- ies can only be speculated on. Acknowledgments We wish to thank the Laboratorio de Ecologia, Pontifica Universidad Catolica de Chile, for sup- port provided both authors during the time that they were conducting fieldwork. The Laboratoire d'Ecologie Theorique et de Biometrie, Universite Catholique de Louvain-la-Neuve, Belgium, pro- vided computing facilities and logistical support. Several reviewers made valuable comments on the manuscript. Drs. Eduardo R. Fuentes and Robert E. Martin assisted with aspects of fieldwork. Fi- nally, we thank Janet L. Meserve and Paule Le Boulenge-Nguyen for unfailing assistance and forebearance during all aspects of the fieldwork and preparation of the manuscript. Literature Cited Andrzejewski, R. 1963. Processes of incoming, set- tlement and disappearance of individuals and varia- tions in the numbers of small rodents. Acta Theriolo- gica, 7: 169-213. . 1969. Analysis of results of catches of small mammals by the "Calendar of Catches" method. Zeszyty Naukowe, 2: 1-104. (In Polish, English sum- mary.) AxELROD, D. I. 1973. History of the Mediterranean ecosystem in California, pp. 225-277. In di Castri, P., and H. A. Mooney, eds., Mediterranean Type Eco- systems. Springer- Verlag, Berlin, 395 pp. Baker, R. H. 1967. Distribution of Recent mammals along the Pacific coastal lowlands of the Western Hemisphere. Systematic Zoology, 16: 28-37. Beatley, J. C. 1 969. Dependence of desert rodents on winter annuals and precipitation. Ecology, 50: 721- 724. . 1976. Rainfall and fluctuating plant popula- tions in relation to distribution and numbers of desert rodents in southwest Nevada. Oecologia (Berlin), 24: 21-42. Clobert, J., J. D. Lebreton, M. Clobert-Gillet, and M. CoQUiLLART. 1985. Estimation of Survival in bird populations by recaptures or sightings of marked in- dividuals, pp. 197-213. In Morgan, B. J. T., and P. M. North, eds., Statistics in Ornithology. Springer- Verlag, Berlin, 418 pp. Dueser, R. D., R. K. Rose, and J. H. Porter. 1984. A body-weight criterion to identify dispersing small mammals. Journal of Mammalogy, 65: 727-729. Ford, R. G., AND F. A. PiTELKA. 1984. Resource lim- itation in populations of the California vole. Ecology, 65: 122-136. FuLK, G. W. 1975. Population ecology of rodents in the semiarid shrublands of Chile.' Occasional Papers, The Museum, Texas Tech University, 33: 1-40. Glanz, W. E. 1977. Comparative ecology of small mammal communities in California and Chile. Ph.D. diss.. University of California, Berkeley, 300 pp. . 1984. Ecological relationships of two species of Akodon in central Chile. Journal of Mammalogy, 65:433-441. Glanz, W. E., and P. L. Meserve. 1 982. An ecological comparison of small mammal communities in Cali- fornia and Chile, pp. 220-226. In Conrad, C. E., and W. C. Oechel, eds.. Dynamics and Management of Mediterranean-type Ecosystems. U.S. Forest Service, Pacific Southwest Forest Range Experiment Station, General Technical Report PSW-58, 637 pp. Greer, J. K. 1965. Mammals of Malleco Province, Chile. Publications of the Museum, Michigan State University, Biological Series, 3: 59-152. Hershkovitz, P. 1962. Evolution of Neotropical cri- cetine rodents (Muridae), with special reference to the Phyllotine group. Fieldiana: Zoology, 46: 1-524. Jennrich, R. I., and F. B. Turner. 1969. Measure- ment of non-circular home range. Journal of Theo- retical Biology, 22: 227-237. Krebs, C. J. 1 966. Demographic changes in fluctuating populations of Microtus californicus. Ecological Monographs, 36: 239-273. Kummerow, J. 1966. Ajjorte al conocimiento de las condiciones climaticas del bosque de Fray Jorge. Uni- versidad de Chile, Facultad de Agronomia, Estacion Experimental Agronomica, Boletin Tecnico N° 24 (Santiago, Chile), pp. 21-24. Le BouLENofe, E. 1985a. Package CM. R., analysis of capture-recapture data. Package description and user's guide (version 4, January 1985). Mimeo report, Unite d'Ecologie, Universite Catholique de Louvain, Lou- vain, Belgium, IV + 50 pp. . 1985b. Computer package for the analysis of capture-recapture data. Acta Zoologica Fennica, 173: 69-72. Le BOULENOfe, E., AND P. Y. Le BoULENOfe-NoUYEN. 1981. Ecological study of a muskrat population. Acta Theriologica, 26: 47-82. MacMillen, R. E. 1964. Population ecology, water relations, and social behavior of a southern California semidesert rodent fauna. University of California Pub- lications in Zoology, 7: 1-59. MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 429 Manly, B. F. J., and M.J. Parr. 1 968. A new method of estimating population size, survivorship, and birth rate from capture-recapture data. Transactions of the Society for British Entomology, 18: 81-89. Mares, M. A. 1975a. South American mammal zoo- geography: Evidence from convergent evolution in desert rodents. Proceedings of the National Academy of Sciences (USA), 72: 1702-1706. . 1975b. Observations of Argentine desert ro- dent ecology, with emphasis on water relations of Elig- modontia typus, pp. 155-175. In Prakash, 1., and P. K. Ghosh, eds.. Rodents in Desert Environments. Junk, The Hague, Netherlands. -. 1976. Convergent evolution of desert rodents: roedores cricetidos. Medio Ambiente (Valdivia, Chile), 5: 115-124. . 1985. A cycling population of /lA:oJo/io//vaceM5 Multivariate analysis and zoogeographic implications. Paleobiology, 2: 39-63. 1977a. The comparative water relations and habitat requirements of three South American phyl- lotine rodents. Journal of Mammalogy, 58: 514-520. 1 977b. Aspects of water balance of Oryzomys longicaudatus from northwest Argentina. Comj)ara- tive Biochemistry and Physiology, 57A: 237-238. 1 980. Convergent evolution among desert ro- dents: A global perspective. Bulletin of the Carnegie Museum of Natural History, 16: 1-51. Mazurkiewicz, M. 1969. Elliptical modification of the home range pattern. Bulletin de I'Academie Polonaise des Sciences Classe II, 17: 427-431. M'Closkey, R. T. 1972. Temporal changes in popu- lations and species diversity in a California rodent community. Journal of Mammalogy, 53: 657-676. Meserve, p. L. 1978. Water dependence in some Chil- ean arid zone rodents. Journal of Mammalogy, 59: 217-219. . 1981a. Trophic relationships among small mammals in a Chilean semiarid thorn scrub com- munity. Journal of Mammalogy, 62: 304-314. 1981b. Resource partitioning in a Chilean semi- arid thorn scrub community. Journal of Animal Ecol- ogy. 50: 745-757. Meserve, P. L., and W. E. Glanz. 1 978. Geographical ecology of small mammals in the northern Chilean arid zone. Journal of Biogeography. 5: 135-148. Meserve, P. L., R. E. Martin, and J. Rodriguez. 1983. Feeding ecology of two Chilean caviomorphs in a cen- tral mediterranean savanna. Journal of Mammalogy, 64: 322-325. . 1984. Comparative ecology of the caviomorph rodent Octodon degus in two Chilean mediterranean type communities. Revista Chilena de Historia Natu- ral (Santiago, Chile), 57: 79-89. Meserve, P. L., R. MurOa, O. Lopetegui, and J. R. Rau. 1982. Observations on the small mammal fau- na of a primary temperate rain forest in southern Chile. Journal of Mammalogy. 63: 3 1 5-3 1 7. Muftoz, P. C, and E. PiSANO V. 1947. Estudio de la vegetacion y flora de los parques nacionales de Fray Jorge y Talinay. Agricultura Tecnica (Santiago, Chile), 7:71-190. MuRUA, R., AND L. A. GonzAlez. 1981. Estudios de preferencias y habitos alimentarios en dos especies de (Rodentia: Cricetidae) in a Chilean temperate rain for- est. Acta Zoologica Fennica, 173: 77-79. Osgood, W. H. 1943. The mammals of Chile. Field Museum of Natural History, Zoological Series, 30: 1- 268. Pearson, O. P. 1975. An outbreak of mice in the coast- al desert of Peru. Mammalia, 39: 375-386. . 1983. Characteristics of a mammalian fauna from forests in Patagonia, southern Argentina. Journal of Mammalogy, 64: 476—492. Pearson, O. P., and A. K. Pearson. 1982. Ecology and biogeography of the southern rainforests of Ar- gentina, pp. 129-142. In Mares, M. A., and H. H. Genoways, eds., Mammalian Biology in South Amer- ica. Special PubUcation Series, Pymatuning Labora- tory of Ecology, University of Pittsburgh, 6: 1-539. Pearson, O. P., and C. P. Ralph. 1978. The diversity and abundance of vertebrates along an altitudinal gra- dient in Peru. Memorias, Museo de Historia Natural "Javier Prado" (Lima, Peru), 18: 1-97. PfeFAUR, J. E., J. L. YaSez, and F. M. Jaksic. 1979. Biological and environmental aspects of a mouse out- break in the semi-arid region of Chile. Mammalia, 43: 313-322. Petruseavicz, K., and R. Andrzejewski. 1962. Nat- ural history of a free-living population of house mice {Mm musculus Linnaeus) with particular reference to groupings within the ix>pulation. Ekologia Polska, Se- na A, 10:85-122. Rau, J., R. Murua, and M. Rosenmann. 1981. Bio- energetics and food preferences in sympalric southern Chilean rodents. Oecologia (Berlin), 50: 205-209. Rojas, M., O. Rivera, G. Montenegro, and C. Barros. 1977. Algunas observaciones de la reproduccion de la hembra silvestre de Octodon degus, Molina y su posible relacion con la fenologia de la vegetacion. Me- dio Ambiente (Valdivia, Chile), 3: 78-82. Rosenmann, M. 1 977. Regulacion termica en Octodon degus. Medio Ambiente (Valdivia, Chile), 3: 1 27-13 1 . Sarmiento, G. 1975. The dry plant formations of South America and their floristic connections. Journal of Biogeography, 2: 233-25 1 . Simpson, B. B. 1975a. Glacial climates in the eastern tropical South Pacific. Nature. 253: 34-36. . 1975b. Pleistocene changes in the flora of the high tropical Andes. Paleobiology, 1: 273-294. SoLBRiG, O. T. 1976. The origin and floristic affinities of the South American temperate desert and semi- desert regions, pp. 7-49. In Goodall, D. W., ed.. Evo- lution of Desert Biota. University of Texas Press, Austin, 250 pp. Tamarin, R. H. 1984. Body mass as a criterion of dispersal in voles: A critique. Journal of Mammalogy, 65:691-692. Van De Graaff, K. M.. and R. P. Balda. 1973. Im- portance of green vegetation for reproduction in the kangaroo rat, Dipodomys merriami merriami. Journal of Mammalogy, 54: 509-512. 430 FIELDIANA: ZOOLOGY Weir, B. J. 1970. The management and breeding of and B. J. Weir, eds., The Biology of Hystricomorph some more hystricomorph rodents. Laboratory Ani- Rodents. Symposium, Zoological Society of London, mals, 4: 83-97. 34: 1-482. . 1974. Reproductive characteristics of hystri- Wcx>ds, C. A., and D. K. Boraker. 1975. Octodon comorph rodents, pp. 269-299. In Rowlands, L W., degus. Mammalian Species, 67: 1-5. MESERVE & LE BOULENGE: SMALL MAMMAL POPULATIONS OF CHILE 431 Demography and Reproduction of the Silky Desert Mouse (Eligmodontia) in Argentina Oliver Pearson, Susana Martin, and Javier Bellati ABSTRACTS Specimens of Eligmodontia typus collected in semiarid steppe habitat were assigned to age classes according to tooth wear, and the age structure of the population was determined for different times of the year. The reproductive season extends from October to the end of April; average litter size is 5.9. Males and females can reach sexual maturity at approximately I'/z months of age and rarely live longer than 9 months. Adult females were 19% heavier than adult males and 6% longer. Abundance ranged from 0.4 per hectare in spring to 3.5 in autumn. Although Eligmodontia is widely dispersed through many arid habitats, life history features such as small home range, large litters, and brief life span distinguish it from highly adapted desert rodents of other continents. Especimenes de Eligmodontia typus colectados en habitat de estepa semiarida fueron clasi- ficados por edad segun desgaste dentario; la estructura de edades de la poblacion fue determinada para diferentes epocas del aiio. La estacion reproductiva se extiende desde Octubre hasta fines de Abril y el tamaiio promedio de camada es de 5.9. Machos y hembras pueden alcanzar la madurez sexual a los 1 '/2 meses de edad y raramente ven mas de 9 meses. Las hembras resultaron ser de mayor tamaiio que los machos. La abundancia fluctua de 0.4 individuos por hectarea en primavera a 3.5 en otorio. Aunque Eligmodontia es ampliamente distribuido en varios habitats aridos, aspectos del ciclo vital tales como pequeiios territorios, camadas numerosas y vida de corta duracion la distinguen de roedores altamente adaptados de desiertos de otros continentes. Especimes de Eligmodontia typus, coletados nos habitat semi-aridos da estepe, foram desig- nados a classes de idade de acordo com o gasto de seus dentes. A estrutura de idade da popula9ao foi assim determinada durante varias epocas do ano. A epoca reproductiva estende-se de outrubro ao fim de abril, e a ninhada media e de 5,9 crias. Machos e femeas podem atingir maturidade sexual por volta de 1,5 meses de idade, e raramente vivem mais do que 9 meses. As femeas sao maiores que os machos. A densidade dos camundongos foi de 0,4 por hectare na primavera e de 3,5/ha. no verao. Apesar da ampla distribui9ao dos Eligmodontes em varios habitats aridos, aspectos da biolo- From the Museum of Vertebrate Zoology, University of California, Berkeley, CA 94720 (Pearson); and Insti- tuto Nacional de Tecnologia Agropecuaria, Casilla de Correo 277, 8400 Bariloche, Argentina. PEARSON ET AL.: ELIGMODONTIA IN ARGENTINA 433 gia destes animals, como por exemplo o seu uso de area, o seu curto periodo de vida, e o seu grande numero de crias, distinguem estes camundongos dos roedores de outros continentes que sao altamente adaptados k vida arida do deserto. Introduction The silky desert mouse is abundant in arid or semiarid habitats from southern Argentina north to the altiplano of southern Peru. It occupies hab- itats that on other continents are filled by highly specialized rodents such as North American kan- garoo rats, Asian dipodids, and African gerbils. Although Eligmodontia is probably as morpho- logically adapted for desert life as any other genus of mouse now living in South America, it is not highly specialized (Mares, 1980). Several tubercles on the soles of each hind foot are fused into a furry pad that may aid locomotion on sandy soil; the hind feet are somewhat elongate (about 20% longer than those of Peromyscus maniculatus of similar size), and many populations have pale silky fur. But, aside from these few presumed desert adap- tations, Eligmodontia is in general appearance and proportions scarcely distinguishable from the wide- ranging P. maniculatus of North America. It lacks the large, inflated bullae, long tufted tail, and kan- garoo-like hind limbs seen in many desert species on other continents. In physiological properties as well, Eligmodon- tia does not appear to be highly adapted to desert living. It can live in captivity for weeks on a diet of dry seeds without drinking water, but this is true for mice from a variety of habitats. It will drink water if available, and also will utilize the water in succulent vegetation (Mares, 1975a, 1977). It has an unusual capability to drink salt water, which suggests that it could utilize succulent parts of desert plants with a high salt content (Mares, 1 977). As measured by rate of loss of body weight by captives without water, the physiological ca- pacity oi Eligmodontia to resist dehydration is in- termediate to that of desert-adapted pocket mice (Perognathus) and the relatively unspecialized Peromyscus maniculatus (Mares, 1975b). Because Eligmodontia is the most abundant mammal in large areas of arid and semiarid hab- itat in South America and lacks a "desert" mor- phology and physiology, we thought that a study of its ecology and life history might reveal con- vergences with life history features of desert ro- dents elsewhere. Convergence of life history fea- 434 tures of unrelated species living in different arid regions would suggest their adaptive significance. Materials and Methods The Argentine species of Eligmodontia is small- er than the species living in Peru and has a different karyotype (unpubl. obs.). Even within the prov- ince of Rio Negro it shows considerable variation in color, size, and proportions. An isolated pop- ulation 10 km west of Bariloche in an island of steppe vegetation consists of very small, dark in- dividuals. Populations from the center and eastern parts of Rio Negro are paler and have much longer tails. To minimize possible geographic variation of the demographic and reproductive features measured in our study, we have limited our sam- ples to specimens collected within 25 km of Pil- caniyeu Viejo in Rio Negro. They are referred to the subspecies E. typus typus pending resolution of nomenclatural complications pointed out by Hershkovitz(1962). The study area (ca. 4 1°S and 70°30'W) is located within what has been described as the Patagonian Province by Cabrera (1971) or as the Western Mesetas and Sierras by Anchorena (1978). Aver- age monthly temperatures range from a low of 1.4°C. in July to a high of 13.6''C. in January. No period is free of frosts. The mean annual temper- ature for the area is approximately 7.4°C. Precip- itation averages 300 mm per year, almost three- quarters of it in the winter. Relative humidity for the area ranges from about 20% in January to 85% in July. Winds are especially strong in spring, al- though frequent and forceful at all seasons. Land use is characterized by extensive sheep and cattle ranching, with seasonal use of summer and winter grazing areas. Sheep were introduced a cen- tury ago, and overgrazing has since modified the plant coverage and species composition of the area. The study area consists of steppe interrupted by small marshy areas referred to locally as mallines. The vegetation has a mean height of approxi- mately 50 cm and is composed mainly of low shrubs and a variety of grasses. The main shrubs FIELDIANA: ZOOLOGY PEARSON ET AL.: ELIGMODONTIA IN ARGENTINA 435 inguol view onterior iabiol view Tooth wear index = o + b Fig. 2. Upper toothrows of Eligmo- dontia showing the two measurements taken for calculation of the tooth wear index. are neneo{Mulinum spinosum), mata torcida {Stil- lingia patagonica), charcao (Senecio bracteolatus), colapiche (Nassauvia glomerulosa), mamuel choi- que {Adesmia campestris), duraznillo (Coliguaya integerrima), and Senecio neaei. The main grasses are various species of bunchgrass, especially Stipa speciosa var. major, Poa lanuginosa, Festuca ar- gentina, P. ligularis, and S. speciosa var. speciosa. The small-mammal fauna of the study area is dominated by Eligmodontia (nocturnal granivore- omnivore) and another mouse of about the same size, Akodon xanthorhinus (nocturnal and diurnal granivore-omnivore). Other small mammals pres- ent but caught much less frequently were Phyllotis danvini (nocturnal omnivore), Reithrodon auritus (nocturnal herbivore), Ctenomys haigii (fossorial herbivore), Notiomys edwardsii (semi-fossorial in- sectivore-omnivore), Euneomys sp. (herbivore), Microcavia australis (diurnal herbivore), and the introduced Lepus capensis (nocturnal herbivore). In the few patches of mesic or moist vegetation, Akodon longipilis and Auliscomys micropus may be abundant. Almost all of these small mammals range widely in southern Argentina. Some of them occur in much moister habitats. Readers familiar with Mares's (1983) descrip- tion of desert communities will note that the Ar- gentine steppe community described above con- tains many of the elements found in desert communities on other continents. The guinea pig/ ground squirrel, tuco-tuco/pocket gopher, and Eu- ropean hare/jackrabbit similarities are especially close. The small bipedal granivores are notably absent in South America. 436 Two hundred twenty Eligmodontia were col- lected in the study area between November 1981 and January 1984 using Sherman live traps baited with rolled oats. Museum Special kill-traps baited usually with commeal or rolled oats but occasion- ally with other baits, or smaller kill-traps baited with commeal and peanut butter. On several oc- casions traplines were set, alternating Sherman traps baited with rolled oats and Museum Specials baited with commeal. The Museum Specials were more effective. Almost all specimens were dis- sected immediately. Skulls were dried for later cleaning. The reproductive condition of males was as- \ sessed from measurements of the length of one testis, the color and texture of the testes, and the length of the seminal vesicles. The seminal vesicles are J-shaped; the length was considered to be the distance from the base of the bladder to the curve of the J. The measurements were used to classify each male into one of three categories: sexually immature, mature, or postmature. Males in breed- ing condition had pale, firm testes more than 5.5 mm long and seminal vesicles more than 7.5 mm long; the diameter of the epididymis was great enough so that this contorted tube could be seen easily through the sheath of the cauda epididymis. For specimens of uncertain category, the presence or absence of abundant spermatozoa was con- firmed by microscopic examination of a smear of an epididymis. Flabby or dark-colored testes, usu- ally associated with submaximal seminal vesicles, were considered to be an indication of postbreed- ing condition. FIELDIANA: ZOOLOGY i TOOTH WEAR INDEX 0.541 mm ■ anterior 0.346 mm Fig. 3. View of the grinding surfaces of upper toothrows showing the degree of wear associated with three different tooth wear indices. 0.078 mm Females were placed in reproductive categories on the basis of examination of uteri, nipples, and the pubic symphysis. Thin, pale uterine horns without embryos or placental scars indicated nul- liparous females. Females with swellings in the uteri were considered to be pregnant. Females with uterine scars, thick uteri, or large nipples were con- sidered to be parous. If milk could be expressed from a nipple, a female was registered as lactating. Nonpregnant females with an open pubic sym- physis were considered to be parous. Estimation of Age— The relative age of each individual was estimated by measuring the amount of wear on the cusps of two upper molar teeth. The tall, sharp cusps of young individuals wear down through life until, if the individual survives long enough, the tooth surface is smooth. Amount of wear was converted to approximate age by not- ing the progression of wear in successive collec- tions of the first cohort of young each season. Sim- ilar techniques have been used by Pearson (1945, 1967, 1975), Happold (1967), French etal. (1974), and Feito et al. (1981). In this study of Eligmodontia, the height of the middle cusp of the first upper molar was measured from the labial side, the depth of the groove be- tween the two main cusps of the second molar in the opposite toothrow from the lingual side (fig. 2). Measurements were recorded to 0.001 mm us- ing a microscope fitted with crosshairs and an elec- tronic digital readout. Although there was a high positive correlation between the two measurements for each specimen (r = +0.88, n = 176), we assumed that accuracy would be increased by averaging the two mea- surements. This average is the number presented as the "Tooth Wear Index" for each specimen: the larger the number, the younger the individual. The cusps on the first molars are usually taller than those on the second molar (average 0.07 mm taller), but in young individuals the second is sometimes taller than the first. The M^ wears more rapidly, however, and a regression of M ' against M- shows that M' may be expected to have about 0.175 mm of cusp remaining at an age when the cusps on M^ have worn away completely. Measurement of the height of the cusps on two teeth is tedious and requires a special microscope, but the Tooth Wear Index is well correlated with the physical appearance of the teeth. Three ex- amples of different ages are shown in Figure 3. It is possible that a subjective estimate of the amount of wear on the surface of all three molars is more reliable than actual measurement, because the ob- server can integrate information from all of the peaks and valleys of the entire toothrow. To determine whether males and females were of equal size, it was desirable to establish a min- imum age at which individuals could be consid- ered adult. This was done by plotting length of head and body against tooth wear and estimating the point at which the growth curve levels off. PEARSON ET AL.: ELIGMODONTIA IN ARGENTINA 437 Table 1 . Measurements of adult male and female Eligmodontia and of all male and female Eligmodontia from the population studied. Measurement Adalt males Adult females All males All females Weight Mean ± SE 18.22 ± 0.52** 21.72 ± 0.65** 16.36 ± 0.44* ' 17.99 ± 0.61* Range 10-25 13-31 7.2-25 7.5-31 No. of specimens 49 47 88 86 Coefficient of variation 19.9 20.5 ... ... Length of head and body Mean ± SE 82.12 ± 0.72** 87.08 ± 0.90** 79.35 ± 0.79* 82.17 ± 0.88* Range 70-92 75-99 59-96 63-99 No. of specimens 49 47 88 86 Coefficient of variation 6.2 7.1 ... ... Length of tail Mean ± SE 84.29 ± 1.49* 88.51 ± 1.24* 80.45 ± 1.25 82.29 ± 1.29 Range 58-108 68-105 51-108 56-105 No. of specimens 49 47 88 86 Coefficient of variation 12.4 9.6 ... ... Greatest length of skull Mean ± SE 24.24 ±0.12 24.43 ±0.16 ... ... Range 23.17-25.97 22.40-26.40 ... ... No. of specimens 33 29 ... ... Coefficient of variation 2.9 3.5 ... r Length of toothrow Mean ± SE 3.71 ± 0.02 3.68 ± 0.03 ... ... Range 3.51-4.05 3.35-4.05 ... c No. of specimens 33 29 ... ... Coefficient of variation 3.6 4.0 * Difference significant at 0.05 level. ** Difference significant at 0.01 level. Similarly, body weight was plotted against tooth wear. From these graphs. 0.460 mm was chosen as the separation between juveniles and adults. No adult defined in this way had a head and body shorter than 70 mm (although some juveniles were larger than this), and only two adults weighed less than 1 2 g (although numerous juveniles weighed more). Results Size— In general, females weigh more, have longer bodies, and have longer tails than males (table 1 ). When only adult individuals (Tooth Wear Index less than 0.460) are considered, females are statistically larger in all three of these dimensions. In length of skull and length of molar toothrow, however, males and females are of the same size. We have presented the size data in considerable detail because seldom are female mice, on average, demonstrably larger than males (Ralls, 1976). Eligmodontia and the desert dipodid Jaculus jaculus are rare exceptions (Happold, 1967). The coeflScient of variation was notably large for body weight (19.9, 20.5) and length of tail (12.4, 9.6). These coefficients are twice as large as those for tail length in one species of kangaroo rat (Li- dicker, 1960). Sex Ratio— The sex ratio of the entire sample was almost equal (table 1). When the sample is divided into age groups, no statistically significant difference in proportion of sexes appears, nor does the proportion of sexes depart appreciably from equality in spring or autumn. It appears, therefore, that the sexes are indeed equally abundant in the wild population, or else increased susceptibility to capture of one sex exactly compensates for its rel- ative scarcity. Season of Reproduction— The abundance of pregnant and lactating females eariy in November and the absence of young animals in the popula- tion at this time (fig. 4) demonstrate that the re- productive season begins in the Southern Hemi- sphere's spring (October). Reproduction continues 438 HELDIANA: ZOOLOGY FEMALES «g^*^* O'NULUPAROUS ••PREGNANT O = PAROUS 9 = LACTATIN6 ■ •? I N M M Fig. 4. Age and reproductive condition of female Eligmodontia during spring, summer, and autumn. Each double- sized symbol represents five individuals. Age is in tooth wear units, with the oldest animals at the top of the diagram. The stippled band suggests the rate of aging of one summer-bom cohort. during the spring and summer, during which time the proportion of juveniles in the population in- creases enormously (figs. 4-6). In the autumn preg- nant females were caught until the end of April, but none of numerous females caught in the mid- dle of May was pregnant (fig. 4). The male reproductive season coincides with that of the females (fig. 5). Even old, sexually ma- ture males pass out of breeding condition in May. Age of Sexual Maturity— Females bom late in October or early in November may be visibly pregnant, or even lactating, with tooth wear of 0.50 mm in mid-January (fig. 4). They must have been inseminated at not more than six to eight weeks of age. Young males become sexually ma- ture at six to eight weeks of age also (fig. 5). Longevity— The cluster of young breeding mice between tooth wear of 0.40 and 0.52 in January (figs. 4 and 5) must have been bom late in October or early in November and therefore were about 10 weeks old. A different cluster of young mice of that age in May have aged to tooth wear of 0.20- 0.30 in November, 22 weeks later. This indicates an age of 22 + 10 = 32 weeks for mice with tooth cusps 0.20 to 0.30 high. By January, when they have entered the oldest age category, they are about 41 weeks old. Since there is no accumulation of individuals in this old category, they must die soon after reaching that age. It is doubtful that any of the mice in oiu" samples were as much as 1 2 months old. The oldest females and males were capable of breeding. 0 10 _j o 1 20 X ^30 ^40 UJ ^50 O60 70 MALES r • = BREEDING 0= POST -BREEDING Oc IMMATURE e o 0 N D J F M A Fig. 5. Age and reproductive condition of male Eligmodontia. PEARSON ET AL.: ELIGMODONTIA IN ARGENTINA 439 2 °- |..o- c SO- S-JO- *.40- I I-.50- o O rf . 9 I I n ' 30 Fig. 6. wear units. 20 10 10 20 30% NOV. JAN. APR.- MAY Age pyramids and sex ratios oi Eligmodontia in spring, summer, and autumn. Age is represented in tooth The changes in age composition of the popu- lation can be followed easily in Figure 6. The entire population is middle-aged in November. This co- hort is already much reduced in January and has probably disappeared entirely by April-May. Sim- ilarly, the young cohorts that have entered the population by January have almost disappeared before April-May, leaving the overwinter success of the population to the young bom relatively late in the summer. In summary, the reproductive activities of the population proceed as follows: In autumn (May) the population is made up almost entirely of an- imals bom during the summer. The younger fe- males are nulliparous and the middle-aged females parous. The young males have not matured sex- ually and the middle-aged and old males are no longer sexually competent. The overwinter sur- vivors breed in October but die before the end of the summer. In fact, few individuals bom early in the breeding season survive until autumn. The young bom in November breed promptly, and some of their male and female offspring breed be- fore the end of the reproductive season in April. Number of Fetuses— Twenty females were vis- ibly pregnant; each carried between three and nine fetuses (mean 5.90, SE ± 0.39). If number of fe- tuses is plotted against age of the mother, there is a loose but significant positive correlation (r = -J-0.55). The wearing away of 0.10 mm of molar cusps, which requires about six weeks, was accom- panied by an average increase of 0.66 more fetuses. Abut^dance— Eligmodontia is usually the most abundant species in its habitat, although in some places Akodon xanthorhinus outnumbers it. In the springtime, when populations are lowest, 1 7 trap- lines composed of a total of 1,723 trap-nights caught 35 Eligmodontia. giving a trap success of 2%. Nine of the lines caught no Eligmodontia. In the most successful line, 14% of the traps held Eligmodontia. In the autumn, 22 traplines adding up to 1,339 trap-nights produced a trap success of 13%. If trap success is directly proportional to abundance, this suggests that Eligmodontia was about six times as abundant in autumn as in spring. This ratio is confirmed by the census data. As many as 52% of the traps on one trapline in the autumn caught Eligmodontia. Density of Eligmodontia was measured on a study area at the Campo Anexo Pilcaniyeu of the Instituto Nacional de Tecnologia Agropecuaria. A 1 -ha- plot was measured and marked as a grid with a stake at each 10-m intersection. A Sherman live- trap was set within 2 m of each stake (121 traps). At the time of each census, trapping was carried out for three or four nights; captured mice were marked with numbered metal ear tags and then released. The size of the population was estimated by Lincoln Index calculations, and the area oc- cupied by this population was considered to be the 1 -ha grid increased by a border strip the width of the three-night home range oi Eligmodontia (mea- sured by recapture of marked individuals). Trap- ping for supplementary animals for dissection was carried out at the same time a few kilometers away. The census area was chosen to represent a mod- erately grazed sample of Patagonian steppe (fig. 1 ), but subsequent study revealed that it should prob- ably be considered to be heavily grazed. It was also within the hunting range of several domestic cats. The vegetation in the census area was scrub with an abundance of Poa lanuginosa and Stipa speciosa var. major, with a high proportion ofStil- lingia patagonica (Lores et al., 1983). An assessment of the vegetation at alternate stakes on the census grid is summarized in Table 2. The dominant three species at each stake were ranked as first, second, or third. The dominance 440 HELDIANA: ZOOLOGY ranking was based on an estimate of the biomass or of the area covered by each species growing within 1 m of each stake. The percentage of ground covered by vegetation at each stake was also es- timated by imagining a 1 -m hoop with the stake at its center. The lowest percentage of ground cov- er was 10%; the highest was 100% in a dense mat ofBerberis heterophylla. The average coverage was 43%. Table 2 shows that Stillingia, bunchgrass (Poa ligularis and Festuca argentina), Mulinum, and Senecio bracteolatus, in that sequence, were the species most frequently listed as one of the dom- inant three species. Stillingia was dominant at 25 of the stakes, Mulinum at 17, bunchgrass at 1 1 , Senecio bracteolatus at four, and Nassauvia at two. Other species that were among the dominant three species at a few stakes were Cerastium arvense, Adesmia campestris, and Berberis heterophylla. The area was censused for mice on 24-26 No- vember 1981, 8-11 April 1982, and 10-12 No- vember 1982. On each of the spring censuses, only a single Eligmodontia was captured, giving a pop- ulation size of one. On the April census, eight Eligmodontia were captured on the study grid, and Lincoln Index calculations indicate a population of 9.3 individuals. Seven recaptures of five of the mice indicate an average distance moved of 3 1 m. If a border strip of 3 1 m is added to the four sides of the 1-ha trapping grid to allow for the area utilized by the population during three nights, then the density of Eligmodontia was 3.5/ha. The av- erage weight of these Eligmodontia was 15.9 g, which gives a biomass of 56 g/ha. If 3 1 m is assumed to be an appropriate width of a border strip in spring as well as in autumn, then the single captures in November in 1981 and in 1982 represent a density of Eligmodontia of 0.4/ha. This is only one-ninth as abundant as in autumn, a ratio that is in fairly good agreement with the difference in percentage trap success in spring and autumn. The only other mammalian species captured on the census grid was Akodon xanthorhinus; but feces of Reithrodon auritus (rata conejo) were seen at several places on the grid, armadillo burrows were present, and a hare {Lepus capensis) was seen. Diet— Eligmodontia is primarily granivorous. Stomach contents of individuals captured in two different areas (Pichileufu and Los Menucos) were examined. Each sample was made up of contents from 23 individuals captured in the month of May. The samples consisted mainly of seeds which were identified and separated from the remaining Table 2. Dominant vegetation at 6 1 evenly distrib- uted sites on the census grid. Dominance was estimated from the apparent biomass of each species. Among dominant Dominant 3 species at sites at sites Vegetation (N) (N) Mata torcida (Stillingia patagonica) 25 57 Neneo {Mulinum spinosum) 17 37 Bunchgrass (Poa ligularis. Festuca argentina) 11 45 Charcao (Senecio bracteolatus) 4 30 Colapiche (Nassauvia glomerulosa) 2 6 contents. The latter consisted of green vegetation which was identified microscopically using epi- dermal characteristics. The most frequent seeds in the sample from Pichileufu were of Berberis and, in the sample from Los Menucos, Prosopis. The green vegetation was made up exclusively of di- cots, including Mulinum spinosum and Acaena sp. in the sample from Pichileufu and predominantly Lycium sp. in the sample from Los Menucos. Grasses were completely absent in the contents examined. Mares (1977) reported that captive in- dividuals ate moths, a grasshopper, and a locust. Our captives ate, among other things, the acrid seeds of Stillingia and the legs of the common black desert beetles. Discussion We have noted above that the anatomy and physiology of Eligmodontia typus do not seem to be highly adapted for desert life. Mares ( 1 983) has listed various behavioral and ecological traits that one would expect to find in a desert rodent, such as noctumality, nesting in burrows, strong terri- torial behavior, relatively large home range, re- production associated with a rainy season, low fecundity, and long survivorship. He points out that these traits have become associated in our minds with desert existence because a number of well-studied desert species display them, but he cautions that confirmation is needed from many more desert communities before we can safely as- PEARSON ET AL.: ELIGMODONTIA IN ARGENTINA 441 sume that they are indeed adaptive traits fixed by the arid environment. We shall show that Elig- modontia displays some of them but not others, and that it shares more traits with its relatives living nearby in South Temperate forest (Pearson, 1983) than with desert forms elsewhere. Eligmodontia is strictly nocturnal. Its nests and retreats are underground, sometimes in burrows made by tuco-tucos or armadillos. We know noth- ing of the exclusiveness of its territories, but its home range, unlike the expectation, is relatively small (see table 3; compare French et al., 1975). Curiously, E. puerulus in rather similar habitat in Peru and measured by the same methods has a much larger home range (table 3). A survey of the number of fetuses carried by desert mice indicates that the average for Elig- modontia (5.9) is unusually large compared with the highly adapted desert rodents belonging to the Dipodidae, Heteromyidae, and Gerbillinae (table 3; Smith & Jorgensen, 1975; Naumov & Lobach- ev, 1975; French etal., 1975; Conley et al., 1977). Layne (1968) listed litter sizes for 21 species and subspecies of Peromyscus, all between 2.8 and 5.0 and with no obvious relationship between litter size and aridity of habitat. Eisenberg and Isaac (1963) listed average litter sizes between 2.2 and 4.0 for several desert species and pointed out that in arid regions small litters would conserve water during lactation. Phyllotis gerbillns is a species re- stricted to the Sechura Desert of Peru. Ten preg- nant females of that species in the collection of the Museum of Vertebrate Zoology at Berkeley, Cal- ifornia, were carrying an average of only 2.8 fe- tuses (range 1-4). In this feature, therefore, P. ger- billus conforms more to the expected, small-litter pattern of desert mice than does Eligmodontia. Male and female Eligmodontia typus breed in the same season in which they are bom, and fe- males are capable of producing more than one litter in a season, even becoming pregnant at a postpartum estrus. These features are found also in a variety of highly adapted desert rodents (McCulloch & Inglis, 1961; Speth et al., 1968; French et al., 1975; Smith & Jorgensen, 1975; Conley et al., 1977) and contribute to a popula- tion's potential to respond rapidly to favorable (but rare and unpredictable) climatic departures from a stressful norm. The population increase of many highly adapted desert species, however, is impeded by the small size of their litters. Initiation of reproduction in many species of desert mice seems to depend upon the sprouting and growth of seedlings after rains (Chew & But- terworth, 1964; Beatley, 1969; Van De GraafTA Balda, 1973; Reichman & Van De Graaff, 1975). Eligmodontia breeds during the plant growing sea- son and, although it is granivorous and insectiv- orous, its stomach during the breeding season fre- quently contains green vegetation. It is probable, therefore, that the timing and success of repro- duction of Eligmodontia also is controlled by the response of vegetation to temperature and rainfall. Such a response is clearly adaptive because keying the timing and the intensity of reproduction to the sprouting of seeds of annual plants maximizes the probability that there will be an adequate seed crop to support offspring. If rainfall should be unusually abundant or should be repeated for two or more years, populations of mice may increase to very high densities. This has occurred in North Amer- ica (Soholt, 1973; Pearson, 1963), South America (Hershkovitz, 1962; Fulk, 1975; Pearson, 1975), and Africa (Christian, 1977; Poulet, 1978). Populations of Eligmodontia did not reach high densities during the three years of this study. On the most productive traplines, as many as 52% of the traps captured Eligmodontia, but other trap- lines at the same time in similar habitat nearby were much less productive. It seems, therefore, that populations were very local and that we en- countered no regional outbreak of either Elig- modontia or Akodon xanthorhinus. Trap success on the grid was usually lower than on other trap- lines. This suggests that our measurements of den- sity (table 3) are lower than would have been found in the most favorable habitats. Extreme fluctuations of population densities make it difficult to compare small mammal com- munities of different deserts. Only in studies last- ing many years can one be certain whether one is studying a sparse, an average, or an abundant pop- ulation. Nevertheless, we have listed data in Table 3 that permit a crude comparison of densities and biomasses of Eligmodontia with those of other desert rodents. Extreme variation is immediately apparent. Fulk's (1975) data from Chilean deserts were from very abundant populations following two rainy years. The highly adapted desert genera such as Dipodomys, Gerbillurus, Desmodillus, and Taterillus were found at greater densities than Eligmodontia. Of special interest is Eligmodontia puerulus in the high desert of Peru. This habitat is quite similar to the steppe of our study area, and Eligmodontia was living in Peru at densities and biomasses similar to those of Eligmodontia typus. It is also of interest that two species of the North American cricetid Peromyscus, relatively 442 HELDIANA: ZOOLOGY Table 3. A comparison of some characteristics of small rodents in various deserts. The studies cited were chosen because the field data were gathered using techniques comparable with our own. Location Ground cover (%) Species Diam. of home range Density Biomass (m) (no./ha) (g/ha) Aver- age litter size Reference Stepi>e, Argen- tina Tola, Peru Loma, Peru Desert scrub, Peru Mountain scrub, Peru Semiarid shrub. La Rincona- da, Chile Semiarid shrub. La Rincona- da, Chile Semiarid shrub, Fray Jorge, Chile Semiarid shrub. Fray Jorge, Chile Larrea, USA Larrea, USA Larrea, USA Larrea, USA Larrea-yucca, USA Larrea-yucca, USA Larrea-cassia, USA Coastal sage, USA Coastal sage, USA Coastal sage, USA Coastal sage, USA Desert grass- land. South- west Africa Desert grass- land. South- west Africa Dry bush, dunes, Sene- gal 43 Eligmodontia typus 33 Eligmodontia puerulus 37 Phyllotis danvini 1 5 Phyllotis danvini 36 Phyllotis danvini 54 44 44 Phyllotis danvini 54 Akodon olivaceus Phyllotis danvini Akodon olivaceus 23 Dipodomys merriami 23 Onychomys torridus 23 Perognathus flavus 23 Peromyscus eremicus + 27 Dipodomys merriami + 27 Perognathus longimembris 6.6 Dipodomys merriami Dipodomys agilis Perognathus fallax Peromyscus eremicus Peromyscus maniculatus Gerbillurus paeba Desmodillus auricularis Taterillus pygargus 31 92 36 60 54 (0.4-3.5) 1.0 6.5 2.6 2.5 (6-56) 25 206 100 95 77 (3.2-4.4) (184-191) 54 (6-16) (158-502) (36-41) (29-46) (1,798-2,463) (30-45) (30-97) 70 118 62 87 61 64 41 27 11.5 (7.8-15.6) 1.83 (0.6-3.3) 1.1 (0-3.9) 1.1 (0.5-3.3) 0.7-3.7 0.8-1.7 16.2 (12.3-19.5) (1.2-4.8) (0.2-1.3) (0.5-4.0) (0-2.5) (7-12) (992-2,707) 453 (370-590) 40.3 (15-71) 7.3 (0-25) 21.9 (4-67) (459-741) (74-295) (4.2-24.8) (8.5-75.6) (0-51) 23 (12-16) 25 3.0, 8.0 108,288 5.9 This study Pearson & Ralph (1978) Pearson & Ralph (1978); Pearson (1975) Pearson & Ralph (1978) 3.7 Pearson & Ralph (1978); Pearson (1975) ••• Fulk(1975) Fulk(1975) 5.2 Fulk(1975) 5.5 Fulk(1975) Chew & Chew (1970) Chew & Chew (1970) Chew & Chew (1970) Chew & Chew (1970) Chew & Butter- worth (1964) Chew & Butter- worth (1964) 2.2 Soholt(1973) 2.6 MacMillen(1964) (2-4) MacMillen(1964) 2.9 MacMillen(1964) (2-6) 4.3 MacMillen(1964) (2-7) Christian (1977) Christian (1977) :4 Poulet(1972) PEARSON ET AL.: ELIGMODONTIA IN ARGENTINA 443 unspecialized and in general appearance similar to Eligmodontia, were living at densities similar to those o^ Eligmodontia (table 3). Our conclusion after comparing the densities oi Eligmodontia with those of other species in Table 3 and with the densities reported by French et al. (1975), Smith and Jorgensen (1975), Naumov and Lobachev (1975), and Conley et al. (1977) is that the highly adapted desert mice on the other continents main- tain higher densities than does Eligmodontia and other genera inhabiting arid zones in South Amer- ica. No long-term studies have been carried out, however, in South America. One of the most impressive demographic dif- ferences between Eligmodontia and the North American genera Dipodomys and Perognathus is in longevity. Our estimate of longevity for Elig- modontia is made by comparing age pyramids of samples trapped at different seasons of the year. These samples indicated such a rapid turnover of the population that very few individuals could be expected to survive nine months after birth and almost none could be expected to reach one year. All studies of Dipodomys and Perognathus show that these genera survive much longer than this. French et al. (1967) recaptured 25 Perognathus longimembris that had been marked three to five years earlier. Chew and Butterworth (1964) recap- tured up to 1 9% of their marked Dipodomys mer- riami and 31% of their Perognathus longimembris a year later. Chew and Chew (1970) found that at least 50% of their marked Dipodomys merriami were still alive six months later, but 50% of the Peromyscus eremicus haa disappeared in two months. M'Closkey (1972) reported survivals per month of 79%, 79%, and 54% for Dipodomys agil- is, Peromyscus eremicus, and Peromyscus manicu- latus, respectively. In Soholt's (1973) study of D. merriami, at least 42% of the marked animals re- mained 50 weeks later. It is not known how Dipodomys achieves such longevity. The long survival of Perognathus is usu- ally attributed to the fact that it is capable of hi- bernation and thereby avoids both metabolic deg- radation and above-ground dangers. We have no evidence that Eligmodontia hibernates and doubt that it does so because it does not accumulate appreciable quantities of fat in the autumn. In the Namib Desert of South-West Africa, nearly half the Gerbillurus and Desmodillus dis- appeared from the population each month (Chris- tian, 1 977). This suggests that their longevities and their age structure would be more like those of Eligmodontia than of Dipodomys and Perogna- thus. Many jerboas in the Sudan, however, survive more than one year (Happold, 1967). Phyllotis darwini in the coastal deserts of Peru and Chile is almost as short-lived as Eligmodontia (Pearson, 1975; Fulk, 1975). The fact that Eligmodontia has been able to colonize an enormous expanse of arid and semi- arid habitat without having many of the anatom- ical, physiological, reproductive, behavioral, and ecological features traditionally associated with desert rodents indicates either 1) that survival in deserts is not so stressful as to require extreme specializations, or 2) that Eligmodontia possesses very effective, undetected desert adaptations. From our study of Eligmodontia in the Patagonian steppe, we conclude that the first alternative is closer to the truth. Other species of relatively unspecialized mice are able, in the absence of highly-adapted competitors, to survive in much harsher deserts, such as Mus domesticus in Peruvian and Austra- lian deserts or Peromyscus maniculatus in North American deserts. The key consideration is one of population density. Densities of highly adapted desert species are held down by even the usual or "normal" climatic regime of the region that they occupy. This is demonstrated by the population increase that occurs following unusual sequences of rain. Relatively unadapted species respond to rains also, but start their increase from lower den- sities and rely on their greater reproductive po- tential to carry them to high densities. The highly adapted species rely on their greater longevity and their other adaptations to maintain relatively high densities during the "normal" unfavorable years. At high densities (much higher than Eligmo- dontia reached in our study), desert rodents eat as much as 70% (Chew & Chew, 1970) or 95% (So- holt, 1973) of the seed production of some plant species, yet this impact seems supportable by the plant populations. Therefore, unless the mouse population increases to much higher densities than we measured, Eligmodontia is not likely to be a threat to the steppe vegetation. Literature Cited Anchorena, J. DE. 1978. Regiones ecologicas de la Patagonia. Primer Curso de Manejo de Pastizales. In- ternal Report of Instituto Nacional de Tecnologia Agropecuaria, Bariloche. Beatley, J. 1969. Dependence of desert rodents on winter annuals and precipitation. Ecology, 50: 721- 724. 444 HELDIANA: ZOOLOGY Cabrera, A. 1971. Fitogeografia de la Republica Ar- gentina. Boletin Sociedad Argentina Botanica, 14: 1- 42. Chew, R., and B. Butterworth. 1964. Ecology of rodents in Indian Cove (Mojave Desert), Joshua Tree National Monument, California. Journal of Mam- malogy, 45: 203-225. Chew, R., and A. Chew. 1970. Energy relationships of the mammals of a desert shrub {Lxirrea trideniata) community. Ecological Monographs, 40: 1-21. Christian, D. 1977. Effects of fire on small mammal populations in a desert grassland. Journal of Mam- malogy, 58: 423-427. Conley, W., J. Nichols, and A. Tipton. 1977. Re- productive strategies in desert rodents, pp. 193-215. In Wauer, R., and D. Riskind, eds.. Transactions of the SymfKJsium on the Biological Resources of the Chihuahuan Desert Region. United States and Mex- ico. National Park Service Transactions and Proceed- ings Series, No. 3. U.S. Department of the Interior, Washington, D.C. EisENBERG, J., AND D. Isaac. 1963. The reproduction of heteromyid rodents in captivity. Journal of Mam- malogy, 44: 61-67. FErro, R., L. Gonzales, AND R. MoNDioN. 1981. De- termination of age and age classes ofAkodon olivaceus in Southern Chile. Neotropical Fauna and Environ- ment, 16: 201-207. French, N., B. Maza, and A. Aschwanden. 1 967. Life spans of Dipodomys and Perognathus in the Mojave Desert. Journal of Mammalogy, 48: 537-548. French, N., B. Maza, H. Hill, A. Aschwanden, and H. Kaaz. 1974. A population study of irradiated desert rodents. Ecological Monographs, 44: 45-72. French, N., D. Stoddart, AND B. BoBEK. 1975. Pat- terns of demography in small mammal populations, pp. 73-102. In Golley, F., K. Petrusewicz, and L. Ryszkowski, eds.. Small Mammals: Their Productiv- ity and Population Dynamics. International Biological Programme 5. Cambridge University Press, London. FuLK, G. 1975. Population ecology of rodents in the semiarid shrublands of Chile. Occasional Papers of The Museum, Texas Tech University, No. 33, 40 pp. Happold, D. 1967. Biology of the jerboa,yacM/«.J7acM- lus butleri (Rodentia, Dipodidae), in the Sudan. Jour- nal of Zoology, 151: 257-275. Hershkovitz, P. 1 962. Evolution of Neotropical cri- cetine rodents (Muridae), with special reference to the phyllotine group. Fieldiana: Zoology, 46: 1-524. Layne, J. 1968. Ontogeny, pp. 148-253. In King, J., ed.. Biology of Peromyscus (Rodentia). Special Pub- lication No. 2, American Society of Mammalogists, xiii + 593 pp. Lidicker, W., Jr. 1960. An analysis of intraspecific variation in the kangaroo rat Dipodomys merriami. University of California Publications in Zoology, 67: 125-218. Lores, R., C. Ferreira, J. de Anchorena, and A. Mar- golin. 1983. Las unidades ecologicas del campo ex- perimental Pilcaniyeu (Provincia de Rio Negro). Su importancia regional. Gaceta Agronomica, 4: 66 1-69 1 . MacMillen, R. 1964. Population ecology, water re- lations, and social behavior of a southern California semidesert rodent fauna. University of California Pub- lications in 2toology, 71: 1-59. Mares, M. 1975a. Observations of Argentine desert rodent ecology, with emphasis on water relations of Eligmodontia typus. pp. 155-175. In Prakash, I., and P. Ghosh, eds.. Rodents in Desert Environments. Monographiae Biologicae, vol. 28. The Hague, Neth- erlands. 1975b. South American mammal zoogeogra- phy: Evidence from convergent evolution in desert rodents. Proceedings of the National Academy of Sci- ences of the United States of America, 72: 1 702-1 706. . 1977. Water economy and salt balance in a South American desert rodent, Eligmodontia typus. Comparative Biochemistry and Physiology, 56 A: 325- 332. . 1980. Convergent evolution among desert ro- dents: A global f>erspective. Bulletin of Carnegie Mu- seum of Natural History, No. 16, 51 pp. 1983. Desert rodent adaptation and commu- nity structure, pp. 30-43. In Biology of Desert Ro- dents. Great Basin Memoirs No. 7, Brigham Young University, Provo, Utah. McCulloch, C, and J. Inglis. 1 96 1 . Breeding periods of the Ord kangaroo rat. Journal of Mammalogy, 42: 337-344. M'Closkey, R. 1972. Temporal changes in popula- tions and species diversity in a California rodent com- munity. Journal of Mammalogy, 53: 657-676. Naumov, N., and V. Lobachev. 1975. Ecology of des- ert rodents of the U.S.S.R. (jerboas and gerbils), pp. 465-598. In Prakash, I., and P. Ghosh, eds.. Rodents in Desert Environments. Monographiae Biologicae, vol. 28. The Hague, Netherlands. Pearson, O. 1 945. Longevity of the short-tailed shrew. American Midland Naturalist, 34: 531-546. . 1963. History of two local outbreaks of feral house mice. Ecology, 44: 540-549. 1967. La estructura por edades y la dinamica reproductiva en una p>oblacion de ratones de campo, Akodon azarae. Physis, 27: 53-58. . 1975. An outbreak of mice in the coastal desert of Peru. Mammalia, 39: 375-386. . 1983. Characteristics of a mammalian fauna from forests in Patagonia, southern Argentina. Journal of Mammalogy, 64: 476-492. Pearson, O., and C. Ralph. 1978. The diversity and abundance of vertebrates along an altitudinal gradient in Peru. Memorias del Museo de Historia Natural "Javier Prado," Lima, No. 18, 97 pp. PouLET, A. 1972. Recherches ecologiques sur une sa- vane sahelienne du Ferlo Septentrional, Senegal: les mammiferes. Terre et Vie, 26: 440-472. . 1978. Evolution of the rodent population of a dry bush savanna in the Senegalese Sahel from 1969 to 1977, pp. 113-117. In Schlitter, D., ed.. Ecology and Taxonomy of African Small Mammals. Bulletin of the Carnegie Museum of Natural History, No. 6. Ralls, K. 1 976. Mammals in which females are larger than males. Quarterly Review of Biology, 51: 245- 276. PEARSON ET AL.: ELIGMODONTIA IN ARGENTINA 445 Reichman, O., and K. Van De Graaff. 1975. As- sociation between ingestion of green vegetation and desert rodent reproduction. Journal of Mammalogy, 56: 503-506. Smith, D., and C. Jorgensen. 1975. Reproductive biology of North American desert rodents, pp. 305- 330. In Prakash, I., and P. Ghosh, eds.. Rodents in Desert Environments. Monographiae Biologicae, vol. 28. The Hague, Netherlands. SoHOLT, L. 1 973. Consumption of primary production by a population of kangaroo rats (Dipodomys merria- mi) in the Mojave Desert. Ecological Monographs, 43: 357-376. Speth, R., C. Pritchett, and C. Jorgensen. 1 968. Re- productive activity of Perognathtds parvus. Journal of Mammalogy, 49: 336-337. Van De Graaff, K., AND R. Balda. 1973. Importance of green vegetation for reproduction in the kangaroo rat, Dipodomys merriami merriami. Journal of Mam- malogy, 54: 509-512. 446 FIELDIANA: ZOOLOGY Baculum of the Lesser Andean Coati, Nasuella olivacea (Gray), and of the Larger Grison, Galictis vittata (Schreber) Edgardo Mondolfi ABSTRACTS The bacula of Nasuella olivacea (Gray) and Galictis vittata (Schreber) are described for the first time. Despite its smaller size, Nasuella has a larger and more robust baculum than Nasua; the bifid distal tip resembles that of Procyon cancrivorus. The baculum of G. vittata is larger and stouter than that of the smaller Galictis cuja, with a more expanded spatulate tip vaguely resembling that seen in Eira barbara. Mustelid specimens previously described by Didier ( 1 947) are reidentified on the basis of these descriptions. Los baculos de Nasuella olivacea (Gray) y Galictis vittata (Schreber) son descritos por primera vez. A pesar de su talla mas pequeiia, Nasuella tiene un baculo mas largo y robusto que Nasua; la punta distal bifida se parece a la de Procyon cancrivorus. El baculo de G. vittata es mds grande y fuerte el del pequeiio Galictis cuja, con una punta espatulada mas expandida, vagamente similar a la observada en Eira barbara. Especimenes mustelidos previamente descritos por Didier (1947) son reidentificados en base a estas descripciones. Descreve-se, pela primeira vez, os bacula de Nasuella olivacea (Gray) e de Galictis vittata (Schreber). Apesar de seu tamanho menor, Nasuella possue um baculum maior e mais robusto do que Nasua, e a ponta distal bifida assemelha-se a de Procyon cancrivorus. O baculum de G. vittata e maior e mais robusto do que o da especie menor, Galictis cuja, possuindo uma ponta larga e espatulada que lembra a forma da de Eira barbara. Especimes mustelideos previamente descritos por Didier (1947), sao aqui reidentificados, baseado nas descrifoes dos bacula. Introduction The baculum or os penis of different species of Camivora has distinctive morphological charac- teristics, with possible taxonomic value. However, as noted by Ewer (1973), the differences do not always reflect taxonomic relationships in any sim- ple way. The bacula of several species of Procyonidae and Mustelidae have been described by Pohl ( 1 909, From the Embassy of Venezuela, P. O. Box 34477, Nairobi, Kenya. 191 l),Pocock(1918, 1921), Chaine(1925), Didier (1947, 1950), and Burt (1960). However, none of these studies refers to the bacula of Nasuella oli- vacea (Gray) or Galictis vittata (Schreber). The purpose of this paper is to describe the bac- ula of these species, with comments on their tax- onomy. Materials and Methods Five bacula of the larger grison {Galictis vittata) and two of the lesser Andean coati (Nasuella oli- MONDOLH: BACULA OF COATI AND GRISON 447 vaced) were analyzed for shape and measurements. Four bacula of the South American coati, Nasua nasua (Linnaeus), two of the crab-eating raccoon, Procyon cancrivorus (G. Cuvier), one of the lesser grison, Galictis {Grisonella) cuja (Molina), and two of the tayra, Eira barbara (Linnaeus) were available for comparison. All these specimens are in my private collection. Results and Discussion Lesser Andean Coati, Nasuella olivacea (Gray) One of the two bacula is from a skeleton of an old individual salvaged on 1 8 March 1 975 by Omar Linares at the Paramo de Piedras Blancas, 30 km NE of Merida City, state of Merida, in western Venezuela. It was found in Polylepis sericea Wedd. brush at an altitude of 4200 m. This baculum is astonishingly robust for the size of the animal, whose skull has a condylobasal length of 106 mm. It is larger and stouter than the baculum of Nasua nasua and about the same size as that of Procyon cancrivorus, but thicker (figs. 1-2). The shaft is nearly straight, slightly curved dorsally in the dis- tal third and ventrally in the proximal third. The proximal end is very wide, rugose, blunt at the tip, with a deep groove on the dorsal face. On the ventral face of the base, there is a wider but shal- lower groove. The shaft tapers gradually toward the distal end. The dorsal surface has a prominent crest, extending from the end of the groove at the base to almost the distal tip. The ventral face is nearly flat on about two-thirds of the shaft, with a very slight indication of a urethral groove. The shaft is compressed laterally, appearing triangular in cross section. The tip is widened and bifid. It is forked in two thick blunt branches or knobs, one slightly longer than the other, separated by a notch. Measurements: total length, 97.2 mm; maximum height at base, 6.5 mm; maximum width at base, 8.7 mm; width at middle of shaft, 4. 1 mm; width across tips of distal knobs, 9.0 mm; length of distal knobs, 5.4-6.5 mm. The other baculum comes from a specimen, pre- served in formalin, found killed on the road on the Piramo de Piedras Negras, altitude 4,250 m. The total length of this animal is 620 mm. It is similar in shape, but much smaller than the former specimen and considerably thinner. Measure- ments: total length, 65.5 mm; maximum width at base, 3.5 mm; width at middle of shaft, 2.7 mm; width across tips of distal knobs, 6.1 mm. The differences in size and massiveness of the two specimens may be due to age. The first bac- ulum is from an old animal, the second probably belongs to a juvenile. In this respect it could be mentioned that the figure given by Powell (1981) of four bacula of fisher (Martes pennanti) shows progressive changes with age, with bone deposi- tion at the basal end and a more massive appear- ance in the adult. The baculum of Nasua, as described by Pohl (1911), Pocock (1921), Chaine (1925), Didier (1950), and Burt (1960) and as shown by four specimens at hand, has an expanded, flattened, subspatulate, indistinctly bifid distal end (figs. 1- 2). The distal end of the baculum of Procyon can- crivorus is very similar to that of Nasuella olivacea, that is, forked in two divergent condyle-like knobs separated by a medial notch. The largest baculum of Procyon cancrivorus at hand shows two tiny dorsal knobs at the base of the larger condyle-like ones (figs. 1-2). The monotypic genus Nasuella was erected by HoUister ( 1 9 1 5) for the little-known lesser Andean coati, Nasuella olivacea. HoUister designated as genotype Nasuella olivacea meridensis (Thomas) from the Andes of Merida, Venezuela. Nasuella resembles Nasua in general features, but is smaller in size, head and body length, 420-478 mm. The tail is much shorter, 228-270 mm. Weight: 1 ,072- 1,500 g. HoUister (1915) described a series of cra- nial characteristics that clearly distinguishes Na- suella from Nasua. He pointed out that the teeth of Nasuella are similar in many respects to those of Bassaricyon. In regard to the postcranial skeleton of Nasuella, the following features could be pointed out: Nasua nasua is unusual among carnivores in that the slender fibula is fused proximally with the tibia, but articulates distally by a synovial joint (Ewer, 1973). The same condition is present in Nasuella olivacea. Stains (1973) described the calcaneum of Nasua nasua and Nasua narica, but made no reference to that of Nasuella olivacea. Regarding a calcaneum I sent him, he gave (in litt.) the fol- lowing comments: "It appears much smaller than that of Nasua nasua. When compared to other procyonids, it appears to be most similar to Bas- sariscus astutus and is about the same size." In- formation regarding this aberrant coati is very scanty. It inhabits the Paramo Life Belt of western Venezuela, throughout Colombia to Ecuador, at 448 FIELDIANA: ZOOLOGY Fig. 1. Left, Dorsal view of baculum of Nasua nasua; middle, dorsal view of baculum oiNasuella olivacea; right, ven- tral view of baculum ofProcyon cancriv- orus. altitudes of 2300 to 4250 m. In the paramos of tion, and it is often seen on the ground. On the the state of Merida, it is called "guache," and there is a small lagoon that bears the name "Laguna de los Guaches." The habitat preference of the lesser Andean coati seems to be paramo brush vegeta- label of a specimen (USNM- 143658) collected in the Montes de la Culata, state of Merida, by Bri- ceno-Gabaldon, is written "nido en tierra, pare cuatro hijos" (nest on ground, gave birth to four MONDOLFI: BACULA OF COATI AND ORISON 449 450 FIELDIANA: ZOOLOGY Fig. 3. Dorsal views of bacula: left, Galictis (Gri- sonella) cuja; right, Galictis (Galictis) vittata. young). The stomach of the male collected at Pdr- amo de Piedras Negras, which I examined, con- tained insect remains. Three weakly distinguished subspecies have been described: Nasuella olivacea olivacea (Gray) from Santa Fe de Bogota, Colombia; Nasuella olivacea meridensis (Thomas) from the Andes of Merida; and Nasuella olivacea quitensis (Lonnberg) from "mas abajo de Lloa, ladera meridional del Pi- chincha, Ecuador." According to the descriptions, it seems that the separation of these subspecies on details of pelage coloration is not justified. MONDOLFI: BACULA OF COATI AND GRISON 451 Fig. 5. Dorsal views of bacula: left and middle, Galictis vittata; right, Eira barbara. Larger Grison, Galictis (Galictis) vittata (Schreber) Didier (1947) described a baculum of the lesser grison, Galictis (Grisonella) cuja (Molina), under the name Grison (Galictis) vittata, which corre- sponds to the larger grison. Ewer (1973, p. 30, fig. 2. 1 5) reproduced Didier's drawing of the baculum. A detailed description of the baculum of the lesser grison is given by Pocock (1918) under the name of Grison Jurax. The description that follows of the baculum of the larger grison is based on five adult specimens. Three of these were collected in the state of Gua- rico, both north and south of Calabozo, Central Llanos of Venezuela; another comes from south of Ciudad Bolivar, state of Bolivar, eastern Ven- ezuela; a fifth is from a captive animal that died at the El Pinar Zoo in Caracas. For comparison, I used a baculum of an adult lesser grison from the state of Minas Gerais, Brazil, as well as the descriptions given by Pocock and Didier. The baculum of the larger grison is larger and much stouter than that of Galictis cuja and shows some differences in shape (figs. 3-4). The shaft is nearly straight. The proximal half is wider, thicker. 452 FIELDIANA: ZOOLOGY Fig. 6. Lateral views of bacula: top and middle, Galictis vittata; bottom, Eira barbara. and compressed laterally, with a blunt end. The ventral surface is flatter, the dorsal prominently ridged. In cross section the basal half presents a triangular outline. The shaft tapers distally. The distal half is somewhat rounded on its ventral face and is slightly bent ventrally. The distal end is depressed, much widened, and bent ventrally, re- calling a golf club. This spatulate distal end is shaped like a heart or an arrowhead, is flattened on its ventral surface, and is slightly concave on its dorsal one. Close to the neck of the shaft, there is a pair of erect, apically rounded excrescences or knobs pointing backward. There is some variation in the curvature of the shaft: in the specimen from the state of Bolivar, the shaft is bent to the right at its distal third; the one from the zoo specimen has a pronounced downward curvature. Extremes and mean measurements for five bacula are: total length, 54.6-56.9 (55.74) mm; maximum width at proximal extremity, 4.3-5.2 (4.72) mm; length of depressed tip, 6.2-7.5 (6.86) mm; width of de- pressed tip, 6.2-7.4 (7.05) mm; least width of shaft just behind the widened tip, 2.2-3.1 (2.82) mm. The baculum of G. cuja is much smaller and thinner than that of G. vittata, but its shape is similar. The shaft is nearly straight, its distal half slightly bent ventrally. The ventral and dorsal sur- faces are similar to the baculum of the larger gris- on. The widened distal end has the shape of an arrowhead; it is narrower than that in G. vittata; the pair of prominent excrescences pointing up- ward close to the neck, called "horns" by Pocock (1918), are prominent. Measurements: total length, 36.5 mm; maximum width at base, 2.7 mm; length of depressed tip, 3.4 mm; width of depressed tip, 2.7 mm; least width of shaft just behind the horns, 1.3 mm. The baculum of the tayra (Eira barbara) is lon- ger (total length, 75-83 mm) than that of the larger grison (figs. 5-6). It has a horseshoe-shaped distal end, formed by an arched concavity in the middle of its dorsal surface, surrounded by a thick rim. The shaft is straight. The enlarged proximal half is laterally flattened, forming a prominent ridge on its dorsal face. Burt ( 1 960) pointed out in his description and figure of a baculum of Eira bar- bara that it is quite different from the one illus- trated by Didier (1947, pp. 139-140). It is notice- able that the description and the three drawings (Didier, 1947, fig. 1) of the baculum of a mustelid Didier called "Le Grison Taira (Galera barbara (L.), Bresil" matches the Venezuelan specimens of the larger grison, G. vittata, very closely in size and shape, particularly the one with the shaft curved downward on the distal half. It can be de- duced that the specimen studied by Didier was MONDOLFI: BACULA OF COATI AND GRISON 453 incorrectly identified and probably belongs to a larger grison. Galictis vittata is a larger animal than G. cuja. Adult males have an average body and head length of 450-626 (577.6) mm; average tail length, 150- 182 (168.4) mm; average weight (based on seven adult males), 2,600-3,665 (3,076) g. According to Walker ( 1 964), G. cuja has a head and body length of 400-450 mm; tail length, 1 50-1 90 mm; weight, about 1 kg. Nehring (1886), Krumbiegel (1942), and Cabrera (1958) considered the presence of a metaconid on the first lower molar an important characteristic which distinguishes G. vittata from G. cuja. Thomas (1912) erected the subgenus Gri- sonella for the latter species, based on the absence of the metaconid. The distribution of the larger grison ranges from southeastern Mexico through Central America southward to central Peru and southeastern Brazil. Cabrera (1958) listed four subspecies: Galictis vit- tata andina Thomas from Pozuzo, Peru; G. vittata vittata (Schreber) from Surinam; G. vittata brasi- liensis (Thunberg) from Rio de Janeiro; and G. vittata canaster Nelson from Cercanias de Tunkas, Yucatan, Mexico. My preliminary work suggests that a thorough revision may reduce the number of subspecies to only two: G. vittata vittata and G. vittata andina. Acknowledgments I am grateful to Dr. Bruce Patterson for review- ing the manuscript and to Hilary M. Airey for typing the manuscript. Literature Cited Burt, W. H. 1960. Bacula of North American mam- mals. Miscellaneous Publications, Museum of Zool- ogy, University of Michigan, 113: 1-76. Cabrera, A. 1958. Catalogo de los mamiferos de Amferica del Sur. Revista del Museo Argentino de Ciencias Naturales "Bernardino Rivadavia" e Insti- tuto National de Investigaci6n de las Ciencias Natu- rales. Ciencias Zool6gicas, 4: i-xvi; 1-307. Chaine,J. 1925. L'os p6nien, 6tude descriptive et com- parative. Actes de la Soci6t6 Linneenne de Bordeaux, 78: 5-195. DiDiER, R. 1947. Etude syst6matique de I'os p6nien des mammiferes. Carnivores. 2. Famille des Muste- lides (suite). Mammalia, 11: 139-152. . 1950. Etude systematique de I'os penien des mammifferes (suite). Famille de Procyonides et Ur- sides. Mammalia, 14: 78-94. Ewer, R. F. 1973. The Carnivores. Weidenfeld and Nicolson, London. HoLLisTER, N. 1915. The genera and subgenera of rac- coons and their allies. Proceedings of the United States National Museum, 49: 143-150. Krumbiegel, I. 1942. Hyrare und Orisons (Tayra und Grison). Die Saiigetiere der Siidamerika— Expediti- onen Prof E>r. Kriegs. 1 7. 2^oologischer Anzeiger, 139: 81-108. Nehring, A. 1886. Beitrage zur Kenntniss der Galictis- Arten. Zoologische Jahrbiicher Zeitschrifl fur Syste- matik. Geographic und Biologic der Thiere, 1: 177- 221. PococK, R. I. 1918. The baculum or os penis of some genera of Mustelidae. Annals and Magazine of Natural History, ser. 9, 1: 307-312. . 1921. The external characters and classifica- tion of the Procyonidae. Proceedings of the Zoological Society of London, 1921: 389-422. PoHL, L. 1909. Ueber das os penis der Musteliden. Jenaische Zeitschrifl fur Naturwissenschaft, 45: 381- 394. . 1911. Das OS penis der Camivoren einschlies- slich der Pinnipedier. Jenaische /feitschrift fur Natur- wissenschaft, 47: 115-160. Powell, R. A. 1981. Martes pennanti. Mammalian Species, 156: 1-6. Stains, H. J. 1973. Comparative study of the calcanea of members of the Ursidae and Procyonidae. Bulletin Southern California Academy of Sciences, 72(3): 137- 148. Thomas, O. 1912. Small mammals from South Amer- ica. Annals and Magazine of Natural History, ser. 8, 10: 44-48. Walker, E. P. 1964. Mammals of the World, vol. IL Johns Hopkins Press, Baltimore, Md. 454 FIELDIANA: ZOOLOGY Origin, Diversification, and Zoogeography of the South American Canidae Annalisa Berta ABSTRACTS Members of the Canidae are known from the late Pliocene and early Pleistocene (Uquian) through the Recent in South America. Ten genera and 28 species of wolves and foxes are represented. Cladistic analysis supports recognition of four monophyletic groups: ( 1 ) Urocyon (including Vulpes, Urocyon, and Otocyon; (2) Dusicyon (including Pseudalopex, Dusicyon, Pro- tocyon, and Theriodictis); (3) Cerdocyon (including Nyctereutes, Cerdocyon, Atelocynus, and Speothos); and (4) Chrysocyon (including Chrysocyon and Canis). Zoogeographic implications of the cladistic hypotheses presented here are supported by the fossil record, suggesting the origin of canids in North America and their subsequent dispersal and extensive radiation in South America. The extinction of large canids in South America at the end of the Pleistocene is a consequence of extinction of their specialized large herbivorous prey. The current high diversity of South American foxes is, at least in part, the result of an opportunistic feeding strategy that utilizes small prey as well as fruits and grains. Miembros de los Canidae son conocidos del Plioceno tardio y Pleistoceno temprano (Uquian) hasta el Reciente en Sudamerica. Diez generos y 28 especies de lobos y zorros estan represen- tados. El analisis cladistico soporta el reconocimiento de cuatro grupos mayores: (1) Urocyon (incluyendo Vulpes, Urocyon, y Otocyon); (2) Dusicyon (incluyendo Pseudalopex, Dusicyon, Protocyon, y Theriodictis); (3) Cerdocyon (incluyendo Nyctereutes, Cerdocyon, Atelocynus, y Speothos); y (4) Chrysocyon (incluyendo Chrysocyon y Canis). Implicaciones zoogeograficas de la hipotesis cladistica de parentesco presentadas aqui son soportadas por el registro fosil y ellas sugieren el origen de los canidos en Norteamerica y su subsecuente dispersion y extensiva radiacion en Sudamerica. La extincion de canidos grandes en el sur del continente a fines del Pleistoceno es una consecuencia de la perdida de sus presas, los especializados herbivoros grandes. La presente alta diversidad de zorros sudamericanos es, al menos, en parte el resultado de una estrategia oportunista de alimentacion que utiliza pequeiias presas como tambien frutos y cereales. Membros dos canidos sao conhecidos do Pleioceno superior, e desde o Pleistoceno inferior (Uquiano) a epoca Recente, na America do Sul. Dez generos e 28 especies de lobos e raposas estao representados. As analises cladisticas fundamentam o reconhecimento de quatro grupos principals: (1) Urocyon (incluindo Vulpes, Urocyon, e Otocyon); (2) Dusicyon (incluindo Pseu- From the Department of Biology, San Diego State University, San Diego, CA 92182. BERTA: SOUTH AMERICAN CANIDAE 455 dalopex, Dusicyon, Protocyon, e Theriodictis); (3) Cerdocyon (incluindo Nyctereutes, Cerdocyon, Atelocynus, e Speothos); e (4) Chrysocyon (incluindo Chrysocyon e Canis). Implicafoes zoogeograficas, provenientes das hipoteses cladisticas apresentadas neste tra- balho, sao sustentadas tambem pelo registro fossil. O cenario, entao, e dos canideos terem ohginado na America do Norte, dispersando-se, subsequentemente, a America do Sul, onde ter-se-iam radiado amplamente. A extin9ao dos canideos grandes sulamericanos, ao final do Pleistoceno, ocorrera devido a exlin9ao de sua especializada rapina— a dos grandes animais herbivoros. O alto grau de diversifica^ao atual das raposas sulamericanas devese, ao menos em parte, aos seus habitos alimentares oportunistas, que utilizam tanto animais pequenos como tambem diversos graos e fnitos. Introduction Although South American canids are more di- verse than those of any other continent, they are less well known. Living South American foxes or wild dogs include seven genera and eleven species. Most are small to medium in size and predomi- nantly omnivorous. Except for the mostly North American gray fox, Urocyon, South American fox- es occur from Panama to Tierra del Fuego in a wide range of habitats that include rain forests, tree-covered steppes, grasslands, and deserts. In the past, canids attained a much greater diversity in South America and included wolves and wolf- like forms. In addition to foxes, they comprised a major component of the large carnivorous adap- tive zone on that continent (Berta, 1 98 1 , in press). The fossil record suggests that the origin and initial diversification of modem canids took place in North America and possibly Middle America during the late Miocene and early Pliocene. Sub- sequently they entered South America after emer- gence of the Panamanian land bridge, approxi- mately 3 MYBP. Thereafter an impressive Plio- Pleistocene radiation of canids is documented in Uquian (late Pliocene and early Pleistocene) through Recent faunas. This paper summarizes current knowledge of the evolution and radiation of South American canids. Cladistic hypotheses of relationship presented provide a test of pro- posals regarding origin, patterns of dispersal, and changes in diversity through time. Previous Systematic Studies The first reference to canids in South America was Kerr's (1792) description of a large wild dog, Canis australis, from the Falkland Islands, off the eastern coast of Argentina. Since that time the nomenclatural and taxonomic history of living and fossil South American canids has been con- fusing and problematic. Important contributors to the classification of South American canids prior to 1945 include Thomas (1914), Kraglievich (1930), Cabrera (1931), and Osgood (1934). Kraglievich (1930) provided the first comprehensive classification of living and fossil taxa, basing his systematic ar- rangement on traditionally emphasized characters of the skull and teeth. Since Simpson's (1945) monograph, classification of the South American canids has taken a variety of different approaches. Langguth (1969, 1970) proposed systematic ar- rangements based on morphology and ecology. He recognized a forest fox group {^"zorros de monte") which included the genus Cerdocyon and the sub- genera Cerdocyon, Atelocynus, and Speothos. Characters uniting this group included dark-col- ored pelage, short, robust skull and limb propor- tions, and a short, vmcoiled cecum. A second group, the grassland foxes {^"zorros de campo") is com- prised of the genus Dusicyon, with Dusicyon and Pseudalopex as subgenera. Members of this group were characterized by a light-colored pelage, skull and limbs of average proportion, and a long, coiled cecum. In 1975, Langguth revised his classifica- tion and gave generic rank to members of the "dif- ferentiated group," Cerdocyon, Speothos, Lycalo- pex, and Atelocynus. The genus Canis comprised his "generalized group," including "dogs possess- ing features common to the majority of canid species" (1975, p. 193). Within this genus he rec- ognized the subgenus Dusicyon for the Falkland Island Wolf, D. australis, and placed four other species (culpaeus, gymnocercus, griseus, and se- churae) in a second subgenus, Pseudalopex. Uro- cyon was categorized as a "Vulpes-like" fox and its generic distinction maintained. Clutton-Brock et al. (1976) used numerical tax- onomy to assess relationships within members of 456 nELDL\NA: ZOOLOGY the family Canidae based on anatomical and be- havioral characters. They concluded that subfam- ilial level separations suggested by Hough ( 1 948) and Thenius (1954) were not warranted. Chryso- cyon and Speothos were recognized as monotypic, I Urocyon was synonymized with Vulpes, and Pseu- dalopex, Atelocynus, Cerdocyon, and Lycalopex were included in the polytypic genus Dusicyon. This taxonomic arrangement is similar to Simp- son's ( 1 945) but without subgeneric designations. r Van Gelder (1978) proposed a classification based on the nature and extent of hybridization between taxa. He recognized the genus Canis as polytypic. Dusicyon, Pseudalopex, Lycalopex, Atelocynus, and Cerdocyon were arranged as sub- genera of Canis, and Urocyon and Vulpes were considered congeners. Chrysocyon and Speothos were identified as monotypic. Unfortunately, as acknowledged by Van Gelder (1978), too few data from molecular, immunological, or karyological studies are available to establish clear relation- ships among members of this group. The arrangement prescribed here is based on cladistic hypotheses of relationship (see fig. 2, table 3). Seven genera of living South American canids are recognized: Chrysocyon (maned wolf), Speo- thos (bush dog), Cerdocyon (crab-eating fox), Atelocynus (small-eared dog), Dusicyon (Falkland Island wolf), Pseudalopex (South American fox), and Urocyon (gray fox). These genera include eleven species. The extinct large canids, Therio- dictis, Protocyon, and Canis, and fossil species of both foxes and wolves bring the total number of recognized living and fossil South American can- ids to 10 genera and 28 species (table 1). Despite numerous systematic treatments, rela- tively little is known about the behavior and ecol- ogy of canids in South America (reviewed by Lang- guth, 1975). The recent field studies of Brady (1978, 1979) on Cerdocyon thous, Jaksic et al. (1980) on Pseudalopex culpaeus and P. griseus, Crespo ( 1 975) on P. culpaeus and P. gymnocercm, and Dietz (1984) on Chrysocyon brachyurus are notable exceptions and hopefully indicative of a new trend. Fossil Record Canids have been recovered from the fossil lo- calities shown in Figure 1. The chronology and usage of South American Land Mammal Ages is as follows: Uquian (2.5-1 .5 mybp), Ensenadan (1 .5- 0.3 MYBP), and Lujanian (300,000-10,000 ybp), which represent the late Pliocene and early Pleis- tocene, middle Pleistocene, and late Pleistocene, respectively (Marshall etal., 1982, 1984). The ear- liest record of canids in South America is from the Vorohue Formation, Buenos Aires Province, Ar- gentina, which has been considered Uquian in age (Pascual et al., 1 966). Canids recovered from these deposits include the highly derived dhole-like species, Protocyon scagliarium, and the general- ized fox, Pseudalopex gymnocercus (Kraglievich, 1952). Large wolves and wolflike canids were especially well represented in South America during the En- senadan mammal age. Large canids that make their first appearance during this interval include Canis, Theriodictis, and Chrysocyon from diverse local- ities in Argentina and the classic Tarija Basin in Bolivia (fig. 1). The foxes Pseudalopex and Cer- docyon are also recorded from Argentina during the Ensenadan (table 2). Maximum diversity of both large and small canids was attained during the succeeding mammal age, the Lujanian. The best known Lujanian faunas containing canids in- clude those from Talara, Peru; Andean Ecuador; Lagoa Santa Caves, Brazil; and Muaco, Venezuela (fig. 1). Two derived species of the true wolf CaAi/5 and two species of the highly specialized large can- id Protocyon are known. The majority of living South American foxes are also first recorded dur- ing this interval: Dusicyon australis, Pseudalopex culpaeus, P. sechurae, P. vetulus, Cerdocyon thous, and Speothos (table 2). At the end of the Pleisto- cene the large canids except Chrysocyon became extinct. The modem forms Atelocynus microtis, Urocyon cinereoargenteus, and Pseudalopex gris- eus are unknown in the South American fossil record. Phylogeny Cladistic analysis yields the phylogenetic rela- tionships presented in Figure 2. Derived charac- ters that support this arrangement are listed in Table 3 and were based on Tedford and Taylor's (in prep.) comprehensive study of North Ameri- can canids. A more detailed discussion of these characters is provided in the Appendix and else- where (Tedford & Taylor, North American fossil Canidae [Mammalia: Camivora]: Tribe Canini [Caninae], unpubl. data; Berta, in press). The po- larity of characters for all South American canids, BERTA: SOUTH AMERICAN CANIDAE 457 Table 1 . Taxonomic arrangement of genera and species of South American Canidae. Genera Stratigraphic occurrence* Family Canidae Atelocynus Cabrera, 1 940 A. microtis (Sclater, 1 882) Canw Linnaeus, 1758 C. dindst Leidy, 1858 C. gezif L. Kraglievich, 1928 C nehringrt (F. Ameghino, 1902) Cerdocyon Hamilton Smith, 1839 C avins\ Torres and Ferrusquia, 1981 C. emenadensis\ {/Kmt^ino, 1885) C. thous (Linnaeus, 1 766) ?C. new sp.t (Tedford, pers. comm.) Chrysocyon Hamilton Smith, 1839 C. brachyurus Illiger, 1815 C. new sp.t (Tedford, pers. comm.) Dusicyon Hamilton Smith, 1839 D. australis\ (Kerr, 1 792) D. avMjf (Burmeister, 1 864) Protocyon G'lehcl, 1855 P. orcesi^ Hoffstetter, 1952 P. scagliarumf J. Kraglievich, 1952 P. troglodytes^ Lund, 1839b PseudalopexBnrmti^XtT, 1856 P. culpaeus (Molina, 1 782) P. griseids {Gray, 1837) P. gymnocerctds (Fischer, 1814) P. peruanusf (Nordenskiold, 1 908) P. sechurae (Thomas, 1 900) P. vetulus Lund, 1 842 Speothos Lund, 1839a S. pacivorns^ Lund, 1839a S. venaticus (Lund, 1 842) Theriodictis T. platensist Mercerat, 1891 T. tarijensis\ (F. Ameghino, 1902) Urocyon U. cinereoargenteus {Schrcher, 1775) U. progressus^ Stevens, 1965. Rec. L. Pleist. (Lujanian) M.-L. Pleist. (Ensenadan-Lujanian) L. Pleist. (Lujanian) L. Plio. E. Pleist. (Blancan) M. Pleist. (Ensenadan) L. Pleist.-Rec. (Lujanian-Rec.) E.-M. Plio. (L. Hemphillian) M. Pleist.-Rec. (Ensenadan-Rec.) M.-L. Plio. (E.-M. Blancan) Rec. L. Pleist.-Rec. (Lujanian-Rec.) L. Pleist. (Lujanian) L. Plio. & E. Pleist. (Uquian) L. Pleist.-Rec. (Lujanian-Rec.) L. Pleist.-Rec. (Lujanian-Rec.) Rec. L. Plio. & E. Pleist.-Rec. (Uquian-Rec.) L. Pleist. (Lujanian) L. Pleist.-Rec. (Lujanian-Rec.) L. Pleist.-Rec. (Lujanian-Rec.) L. Pleist.-Rec. (Lujanian-Rec.) L. Pleist.-Rec. (Lujanian-Rec.) M.-L. Pleist. (Ensenadan-Lujanian) M.-L. Pleist. (late Ensenadan and/or early Lujanian) Rec. M.-L. Plio. (Blancan) * E. = early, M. = middle, L. = late, Plio. = Pliocene, Pleist. = Pleistocene, and Rec. = Recent, t Extinct. with the exception of Urocyon, was assessed by outgroup comparison with ^''Canis" davisi, a prim- itive North American canid. The bat-eared fox, Otocyon, was the outgroup for Urocyon. With the exception of the gray fox, Urocyon, and the maned wolf, Chrysocyon, the South American canids are a monophyletic group. Urocyon is a sister group of Otocyon, and Chrysocyon is the sister taxon of Cants. Based on cladistic analysis, four groups of South American canids are recognized: (1) Uro- cyon, (2) Dusicyon, (3) Cerdocyon, and (4) Chry- socyon (fig. 2). Urocyon Group This group includes the foxes Vulpes, Otocyon, and Urocyon and their fossil relatives. It is united by the derived characters (5, 6) listed in Table 3. Urocyon is distinguished from other members of this group by canines small relative to cheekteeth, and mandibular condyle above the level of the alveolar border of the cheekteeth. However, both characters also occur in the Dusicyon group, in- terpreted as parallel acquisitions. The Urocyon group is more primitive in lacking characters 7-9 458 FIELDIANA: ZOOLOGY r^ --. / S^EHEZUE^ii K -0 /ecJMPo^ < ^^-^ \.£H^UV^\ flflXZIt / / -20 a^gentinA^"-./^ • / -40 CHtL^E rJ"^^ % / °" 600 1200 km 8 0 ''^'^V^. 40 1 Fig. 1 . Geographic and stratigraphic distribution of South American fossil canids. • = Lujanian: Brazil (Lagoa Santa Caves), Bolivia (Nuapua), Chile (Eberhart Cave), Ecuador (Salinas, La Carolina), Peru (Talara), Venezuela (Muaco); ■ = Ensenadan: Argentina (Buenos Aires, La Plata, and environs), Bolivia (Tarija); ▲ = Uquian: Argentina (near Mar del Plata). which unite the Dusicyon group with more ad- vanced canids. The only living South American representative of this group, the gray fox, Urocyon cinereoargenteus, extends its range from Canada through the eastern and western United States, Mexico, and Central America, into the northern part of Colombia and Venezuela. This species lacks a South American fossil record, although it has been recorded from nearly 40 North American Pleistocene localities, the oldest of which are late Irvingtonian or possibly early Irvingtonian (Kur- ten &. Anderson, 1980). The earliest record of the genus is Urocyon progressus described by Stevens (1965) from the Blancan of Texas, Kansas, and possibly Nebraska. Dusicyon Groap The Dusicyon lineage can be subdivided into two modem genera, the more generalized Pseu- dalopex foxes and the more derived Dusicyon BERTA: SOUTH AMERICAN CANIDAE 459 Urocyon group Dusicyon group Cerdocyon group Chrysocyon group o" ,A .*^o^V ..^' (5,6) (24-26) (27.28) '(29-31) (17) (18-23) *, extinct (12-16) Hi 0-11) (7-9) Hl-4) Fig. 2. Cladogram of proposed relationships among South American canids and related taxa. "wolves." Monophyly of this group is supported peruanus; the Argentine chilla fox, P. griseus; the by the derived characters 12-16 (table 3). The Sechura Desert fox, P. sechurae; Azara's fox, P. Pseudalopex foxes include six species: the modem gymnocercus; and the hoary fox, P. vetulus. Pseu- culpeo fox, P. culpaeus, and its fossil relative, P. dalopex culpaeus (fig. 3) and P. peruanus are coy- Table 2. Stratigraphic ranges of South American canid genera. Genera Uquian Ensenadan Liuanian Recent Atelocynus Canis Cerdocyon * Chrysocyon * Dusicyon Protocyon Pseudalopex Speothos Theriodictis Urocyon * 460 FIELDIANA: ZOOLOGY Table 3. (see fig. 3). Derived characters that support phylogeny No. Character 1 . m2 anterolabial cingulum enlarged. 2. m, posterior cingulum present. 3. m, metaconid enlarged, taller than protoconid. 4. Metatarsal I reduced to proximal rudiment. 5. Upper incisors simple, I'-- cusplets weak or absent. 6. Paroccipital process broad, tip does not extend be- low body of process. 7. Frontal sinus large, penetrating postorbital process. 8. V enlarged, extending markedly below level of I'--. 9. Humerus lacks entepicondylar foramen. 10. Strongly arched zygoma with inverted jugals. 1 1 . Angular process large with expanded fossa for in- ferior/superior branch of medial pterygoideus mus- cle or expanded pterygoid fossa. 1 2. Long palatines extending to or beyond toothrow. 1 3. Coronoid process anteroposteriorly broad and dor- soventrally low. 14. m, with protostylid. 15. m|_2 with mesoconid (fig. 3). 16. m, with strong paracristid (fig. 3). 17. Frontal sinus large, does not penetrate postorbital process. 18. Canines small relative to cheekteeth. 19. Angular process with pterygoid fossa greatly ex- panded. 20. External auditory meatus very short and of small diameter. 2 1 . Cecum short and straight. 22. Ears short. 23. Limbs short. 24. Camassials small relative to cheekteeth. 25. Mandibular condyle above level of alveolar border of cheekteeth. 26. Subangular lobe. 27. Nasals short. 28. Frontal sinus small, does not j)enetrate postorbital process. 29. Frontal sinus large, penetrating postorbital process, extending anteriorly and particularly posteriorly, ul- timately to the frontal-parietal suture. 30. Angular process with large fossa for superior branch of medial pterygoideus muscle. 31. P enlarged with accessory cusps and a strong pos- teromedial cingulum. ote-like in size, proportionately larger and broader than other South American foxes, with a reduced m, metaconid. Pseudalopex sechurae and P. ve- tulus share small size, short rostrum, very small camassials relative to the cheekteeth, and M'"^ very narrow for their length. The earliest recorded fox, Pseudalopex gym- nocercus, presumably diflferentiated into two lin- eages. The large, robust, coyote-like lineage rep- resented by P. culpaeus in Ecuador and P. peruanus in Peru is known from the Lujanian. Pseudalopex culpaeus is now widely distributed throughout the parcrd lesd Fig. 3. Left upper (top) and lower (bottom) dentition of Pseudalopex culpaeus. illustrating dental features characteristic of the Dusicyon lineage, parcrd = para- cristid; mesd = mesoconid; scale = 2 cm. Andes from Colombia to the southern tip of Chile, and is found throughout the Patagonian plateau. A second, smaller lineage comprised of/*, sechurae and P. vetulus is also known from the late Pleis- tocene. The past and present distribution of P. sechurae is restricted to northwestern Peru and southwestern Ecuador. Pseudalopex vetulus was recorded as a fossil in Argentina but survives today in southeastern Brazil (Minas Gerais and Matto Grosso states). Members of the more derived Dusicyon "wolves," the recently extinct Falkland Island wolf, D. australis, and its fossil relative D. avus share 2 short, high-crowned premolars and small M - relative to M-. These taxa are related to the large fossil canids, Theriodictis and Protocyon, in pos- sessing the following derived characters: broad palate, deep zygoma with wide masseteric scar, m, metaconid reduced or absent, and mj metaconid relatively unreduced (Berta, 1 98 1 , in press). Pro- tocyon and D. australis share extreme reduction of P* protocone and have a posterior tilt to the pj crown. Although D. australis is unknown as a fossil and D. avus is only known from the late Pleisto- cene, their closest relatives, Protocyon and Ther- iodictis, are known from the late Pliocene and early Pleistocene through the late Pleistocene and pos- siby into the Recent. Theriodictis, a medium-sized, short-faced canid represented by T. platensis and T. tarijensis, is BERTA: SOUTH AMERICAN CANIDAE 461 Fig. 4. Lateral (top) and occlusal (bottom) views of left mandibular ramus of Protocyon orcesi. Escuela Poli- technica Nacional, Quito, EPN V2871. Scale = 2 cm. recorded from Ensenadan- and Lujanian-aged de- posits in Argentina, Bolivia, and Ecuador. The type species, T. platensis, with its simplification of the m, talonid (metaconid lost and the ento- conid retained as a distinct cusp), is a likely ances- tor for the more derived genus Protocyon. Among Protocyon species these distinctive dental char- acters and others are further modified: both the metaconid and entoconid are lost on m,, as are the hypocones on M'-^. Three species of this well- known canid are recognized, P. orcesi (fig. 4), P. scagliarum, and P. troglodytes (fig. 5), from Uqui- an through Lujanian and possibly Recent deposits in Argentina, Bolivia, Brazil, and Ecuador. De- rived members of the Dusicyon group, with their high-crowned premolars and trenchant m, with very reduced or absent talonid cusps, show a trend toward hyp)ercamivory— that is, increased spe- cialization of the shearing mechanism. Cerdocyon Group The crab-eating fox, Cerdocyon, and the Asian raccoon dog, Nyctereutes, are primitive members of the Cerdocyon clade. They form a sister group defined by derived characters 24-26 (table 3). Cer- docyon is known from late Miocene-early Pliocene (6-3 MYBP) deposits in North America; Nycte- reutes has been recorded from coeval deposits in Europe (R. H. Tedford, pers. comm.). Cerdocyon avius is reported from the Blancan of Baja Cali- fornia, Mexico (Torres & Ferrusquia, 1981), and an undescribed lower jaw is questionably assigned to this genus from the late Hemphillian of the Texas Panhandle (R. H. Tedford, pers. comm.). Its South American fossil record includes two sjjecies, Cerdocyon ensenadensis from the Ensena- dan of Argentina, and the closely related living species C thous (including lydekkeri; see Berta, 462 FIELDIANA: ZOOLOGY Fig. 5. Lateral (top) and occlusal (bottom) views of right maxillary of Protocyon troglodytes. Universitets Zoo- logiske Museum (Peter Wilhelm Lund Collection), Copenhagen, UZML 5697-5698. Scale = 2 cm. 1982) (fig. 6), from Lujanian to Recent deposits in Brazil. Today crab-eating foxes inhabit the sa- vannah and woodland areas of northeastern South America, with a range extending from Colombia, northern Argentina, and Uruguay. Derived members of the Cerdocyon clade, the small-eared dog, Atelocynus, and the bush dog, Speothos, are distinguished by shared derived characters 27-28 (table 3). Speothos, with two known species, is the most derived member of the Cerdocyon c\di6e. Speothos pacivorus, from the late Pleistocene-Recent Lagoa Santa Caves of Brazil, is characterized by large size, presence of a meta- conule and hypocone on M', and double-rooted m^. The living species S. venaticus, also recorded from these cave deposits, is distinguished by its reduced size, loss of metaconule and hypocone on M', and absence of M- and mj (Berta, 1984). The current range of the bush dog extends from Pan- ama throughout the Amazonian basin. No fossils of Atelocynus are known. The living species, A. microtis, occurs in tropical rain forests in the Amazonian basin in Brazil, Peru, Ecuador, and Colombia; the upper Rio Orinoco basin in Colombia and Venezuela; and the upper Rio Para- na basin in Matto Grosso, Brazil (Hershkovitz, 1961). Chrysocyon Group This group includes the maned wolf, Chryso- cyon brachyurus, and three extinct species of South American "true" wolves, Canis gezi, C. nehringi, and C dims. Derived characters shared by these taxa include characters 29-3 1 (table 3). The maned wolf, Chrysocyon, is the most dis- tinctive South American canid. It is distinguished from Canis in having small camassials relative to the cheekteeth, a short, straight cecum, and straight and greatly elongate limbs. In addition to the living C brachyurus, reported from the Ensenadan of BERTA: SOUTH AMERICAN CANIDAE 463 Fig. 6. Lateral (top) and ventral (bottom) views of skull ofCerdocyon thous thous. American Museum of Natural History, New York, AMNH 130475. Scale = 5 cm. Bolivia and Lujanian and possibly Recent deposits in Brazil (Berta, 1981, in press), an undescribed new species is known from the early and middle Blancan of Arizona and Mexico (R. H. Tedford, pers. comm.). The present range of the maned wolf indicates its subsequent southern dispersal into northern Argentina and Paraguay. North American Pleistocene wolves (including C. armbrusteri, C. cf. C. dints, and C. lupus) were ancestors of South American species. The earliest recorded species, Canis gezi from the Ensenadan of Argentina, is a good structural ancestor for the Lujanian species, C.nehringi. Canis nehringi dif- fers from C gezi in the continued trend toward larger size, development of a narrow, triangular supraoccipital shield, and greater complication of cusps on the upper and lower molars. The best- represented and most derived species, C. dirus (fig. 7), is distinguished from C nehringi in its larger size, more massive proportions, and more com- plex construction of the lower molars (Berta, 1981, in press). Canis dirus, widely distributed in North America, has a more limited South American dis- tribution, occurring only in Bolivia, Peru, and Venezuela. Zoogeographic History: Problems of Origin, Dispersal, and Ecology Inherent in any cladistic analysis is the hypoth- esis that two taxa shared a closest common ances- 464 FIELDIANA: ZOOLOGY Fig. 7. Lateral (top) and ventral (bottom) views of skull ofCanis dints. Royal Ontario Museum, Toronto, ROM 4303. Scale = 2 cm. tor. For this reason every cladistic hypothesis entails zoogeographic implications about the tem- poral and spatial distribution of the animals stud- ied. Early consideration of the stratigraphic posi- tion of the fossils can bias a cladistic analysis, which initially should be based on morphological information. However, once a cladogram has been constructed, the stratigraphic and spatial occur- rence of fossil forms can provide additional valu- able information regarding patterns of distribution that can be tested. A North American origin for the canids of South America is supported by the fossil record. The earlier record of Canis, Chrysocyon, Cerdo- cyon, and Urocyon in North America suggests that this group originated as early as the late Miocene and early Pliocene (6-3 mybp, table 4). This same phylogenetic pattern is seen among other "endem- BERTA: SOUTH AMERICAN CANIDAE 465 Table 4. First appearance of South American canid genera in North and South America. Record in Record in Genera North America South America Atelocynus Recent Canis Blancan Ensenadan Cerdocyon Hemphillian Ensenadan Chrysocyon Blancan Ensenadan Dusicyon Lujanian Protocyon Uquian Pseudalopex Uquian Speothos Lujanian Theriodictis Ensenadan Urocyon Hemphillian Recent ic" groups that later dispersed to South America, including the sigmodontine rodents (Jacobs & Lindsay, 1981), equine horses (MacFadden, 1979), and llamas (Webb, 1974). Although four of the 10 recognized South American canid genera have more ancient records in North America, canids probably did not actually arrive on the southern continent until much later. It is suggested that can- ids originated and initially diversified in North America (and possibly Middle America, but data there are lacking). After emergence of the Pana- manian Land Bridge, during the late Pliocene and early Pleistocene (beginning about 3 mybp), canids entered South America during the "Great Amer- ican Interchange" and then radiated to achieve their present diversity. Support for this proposal again comes from the fossil record. Canids are conspicuously absent from late Tertiary deposits in South America, and their first record of occur- rence on the southern continent is from Uquian (2.5-1.5 mybp) deposits. There is no known phys- ical evidence which places canids or their possible ancestors in South America earlier than the late Pliocene. Our knowledge of patterns of diversification and dispersal of canids, once in South America, is limited by their more common occurrence in fossil localities in the southern portion of the continent (fig. 1 ). The known record suggests Argentina as a major center of evolution for this group, but this is at least partly due to the fact that both Quater- nary and Tertiary faunas from Argentina are the best known and most intensively studied (for re- cent reviews see Marshall et al., 1984). Undoubt- edly, the Brazilian highlands were also an impor- tant area of canid differentiation (Langguth, 1 975). The cave faunas of Lagoa Santa and environs which range from the late Pleistocene to the Re- cent provide some measure of this rich sample of past diversity. Among the nearly 40 mammalian genera known (excluding bats) are four genera of cdin\6s— Cerdocyon, Atelocynus, Speothos, and Protocyon. Pleistocene canids and other vertebrates record- ed from the Lagoa Santa Caves and Muaco, Ven- ezuela, presumably followed Webb's ( 1978) "East- em Savanna Route," a major dispersal corridor that extended from the Caribbean perimeter of Central America into eastern Venezuela and there into the Amazon Basin. Savanna habitats predom- inated along this route, the major North-South corridor for amphibians and reptiles. In Argen- tina, this route converged with a second major dispersal corridor, the "Andean Route," which ex- tended from the isthmian region along the Andean chain (Webb, 1 978). The cool, dry unforested hab- itats of this route were occupied by most bird and mammal groups. Fossil canid faunas along this pathway were the Pacific coast fauna of Talara, Peru, and those of coastal and Andean Ecuador and Bolivia. If one considers the current distributions oi Cer- docyon in South America and Nyctereutes in Asia, their hypothesized sister group relationship is puz- zling. Additional stratigraphic and geographic in- formation clarifies this systematic proposal. The fossil record suggests that the common ancestor of those taxa ranged over Eurasia and North America between 10 and 4 mybp. Cerdocyon is known from late Miocene and early Pliocene (6- 3 mybp) deposits in North America, and Nycte- reutes has been reported from coeval deposits in Eurasia (Torres & Ferrusquia, 1981; Soria &, Aguirre, 1976). The disappearance of the Bering Land Bridge between 4.5 and 4.0 mybp can be seen as a vicariant event, isolating Nyctereutes in Eur- asia and Cerdocyon in North America. After es- tablishment of the Panamanian Land Bridge, ap- proximately 3 MYBP, Cerdocyon dispersed from North America to South America, where it is re- corded in late Pleistocene through Recent faunas. One of the more interesting biogeographic prob- lems concerns the origin of the Falkland Island wolf, Dusicyon australis, which has been extinct since 1880. According to Clutton-Brock (1977), the white tail tip, enlarged frontal sinuses, and wide muzzle of D. australis are characters signi- fying domestication; the same traits frequently oc- cur in the Australian dingo. Clutton-Brock argues that D. australis was domesticated by early man and then introduced to the Falklands. However, the geological history of this area and phylogenetic 466 HELDIANA: ZOOLOGY Fig. 8. South American canid diver- sity through time. Genera considered as either "fox" or "wolf" morphotypes. Uquian Ensenadan Lujanian Recent information offer an alternative hypothesis. Cla- distic analysis indicates a close sister group rela- tionship between D. australis and D. avus. Re- mains of D. avus are known from Eberhart Cave in southern Chile. Sloth dung, hide, hair, and bone from this cave have yielded '"C dates ranging from 10,200 ± 200 to 13,500 ± 190 ybp (Long & Mar- tin, 1974), suggesting a minimum time of diver- gence for these species. However, evidence that Dusicyon avus survived even longer is indicated by isolated teeth recovered from the southern coast of Argentina which were referred to this species by Tonni and Politis ( 1 98 1 ). In any event, lowered sea level during glacial times would have brought the mainland and the Falkland Islands closer to- gether and facilitated dispersal of terrestrial mam- mals, including D. australis (and possibly D. avus). The distinctiveness of the Falkland Island wolf, with its high-crowned teeth, reduced molar cusps, and highly modified shearing camassials, is more likely the result of its isolation as the island's only indigenous carnivore rather than its domestica- tion. It has been suggested that during the Pleistocene placental carnivores replaced the doglike marsu- pial family, Borhyaenidae, which occupied the car- nivorous adaptive zones during the late Tertiary (Marshall, 1978). As Marshall (1978, p. 82) noted. Whether this "relay" of various carnivorous groups through time was due to active com- petition between the successive groups fill- ing these roles, or to the disappearance of one carnivorous group (possibly linked with concurrent environmental changes) with subsequent passive replacement by another group which came to fill a similar role in a later fauna, (or whether the faunal changes were the result of a combination of these possibilities or others) is unknown. However, it is clear that before we can understand interactions of competition, extinction, and/or re- placement, the pattern must be known. Figure 8 summarizes the pattern of canid diversity through the Plio-Pleistocene. It is apparent that beginning in the late Pliocene and early Pleistocene, when canids first appear in South America, both adap- tive types are represented, the small-medium fox and the large wolf By the end of the Pleistocene these two adaptive types have shifted dramatically in diversity. The large wolves declined rapidly, with only the maned wolf, Chrysocyon, surviving into the present. It should be noted that morphologically and ecologically Chrysocyon is a large omnivore. The fox morphotype with its generalized omnivorous habits seems to have been favored. Today, among North American, African, and Eurasian canids, foxes are the most numerous and display the great- est diversity. It seems that animals with more flex- ible food habits have been able to adapt more easily to environmental changes, especially those at the end of the Pleistocene. Such an opportunistic feeding strategy has been documented for South American foxes. Field stud- BERTA: SOUTH AMERICAN CANIDAE 467 ies of Pseudalopex culpaeus and P. griseus have shown that increasing amounts of fruit are eaten from spring to winter as rodent densities decrease towards the winter (Yariez & Jaksic, 1978; Jaksic et al., 1980). Brady (1979) observed that Cerdo- cyon thous on the llanos of central Venezuela showed seasonal food shifts. When insects and fruit (primary food sources during the wet season) become scarce during the dry season, crab-eating foxes hunt crabs and vertebrates (lizards, snakes, and rodents). The extinction of large canids in South America at the end of the Pleistocene has been related to the extinction of their large herbivorous prey (Ber- ta, 1 98 1 , in press). Today, the large cats— the puma, Felis concolor, and the jaguar, Felis o/ica— survive as sole occupants of the large carnivorous adaptive zone. Reasons for the success of large felids at the expense of large canids is not known, although their differing feeding and hunting strategies and social behavior probably played a part. Fox (1975) classified the wolf, the dhole, and the African hunt- ing dog "social hunters" which obtain large un- gulate prey by pack hunting. A second group, "sol- itary-social hunters" typified by the coyote and jackal, hunt alone or in pairs depending on the size, abudance, and distribution of prey. A third group, "solitary hunters" which hunt small prey and may also be omnivorous, is exemplified by foxes, mustelids, procyonids, and felids (lions). The large South American canids of the Plio-Pleisto- cene Canis, Theriodictis, and Protocyon were likely both "social hunters" and "solitary-social hunt- ers"; thus they would have been profoundly af- fected by the extinction of large ungulates between 1 5,000 and 8,000 ybp. During the late Pleistocene, the herbivorous megafauna (including ground sloths, glyptodonts, proboscidians, horses, no- toungulates, and litoptems) reached its acme and numbered nearly 50 genera. This stands in sharp contrast to the 1 1 genera represented today. The present-day high diversity seen among the foxes (especially the Pseudalopex complex), procyonids, mustelids, and felids is a consequence of an op- portunistic feeding strategy and the greater avail- ability of small to medium-sized prey (e.g., ro- dents, birds) and fruits and grains. Acknowledgments I am particularly grateful to Dr. Richard H. Ted- ford, whose continuing encouragement, guidance, and extensive knowledge of the Canidae greatly facilitated this study. J. David Archibald and Rog- er Carpenter are acknowledged for critically read- ing the manuscript. Specimen illustrations were skillfully prepared by Nancy Halliday (fig. 6) and Patricia Lufkin (figs. 3-5, 7). Literature Cited Ameghino, F. 1885. Nuevos restos de mamiferos fo- siles oligocenos, recognidos por el professor Pedro Sclabrin: y pertenecientes al Museo provincial de la ciudad de Parana. Boletin de la Academia Nacional de Ciencias, Cordoba, 8: 3-207. . 1902. Notas sobre algunos mamiferos fosiles nuevos o poco conocidos del valle de Tarija. Anales del Museo Nacional de Buenos Aires, 3: 225-261. Berta, a. 1981. Evolution of large canids in South America. Anais II Congresso Latino Americano de Paleontologia, Porto Alegre, 2: 835-845. . 1982. Cerdocyon thous. Mammalian Species, 186: 1-4. . 1984. The Pleistocene bush dog Speothos pa- civorus from the Lagoa Santa Caves, Brazil. Journal of Mammalogy 65(4): 549-559. In press. Quaternary evolution and biogeog- raphy of the large South American Canidae (Mam- malia: Camivora). University of California Publica- tions in the Geological Sciences. Brady, C. A. 1978. Reproduction, growth and parental care in crab-eating foxes {Cerdocyon thous) at the Na- tional Zoological Park, Washington. International Zoo Yearbook 18: 130-134. . 1 979. Observations on the behavior and ecol- ogy of the crab-eating fox {Cerdocyon thous), pp. 161- 1 7 1 . /« Eisenberg, J. P., ed.. Vertebrate Ecology in the Northern Neotropics. Smithsonian Institution Press, Washington, D.C., 271 pp. BuRMEiSTER, H. 1856. Erlauterungen zur Fauna Bra- siliens, enthaltend Abbildungen und ausfiihrliche Be- schreibungen neuer oder ungeniigend bekannter Thier- Arten. Georg Reimer, Berlin, viii +115 pp. . 1864. Fauna Argentina. I. Mamiferos fosiles. Anales del Museo Nacional de Buenos Aires, 1: 123- 142. Cabrera, A. 1931. On some South j\merican canine genera. Journal of Mammalogy, 12: 54-67. . 1940. Notas sobre camivoros sudamericanos. Notas Museo de La Plata, Zoology, 29: 1-22. Clutton-Brock, J. 1977. Man-made dogs. Science, 197: 1340-1342. Cllttton-Brock, J., G. B. Corbet, and M. Hills. 1976. A review of the family Canidae, with a clas- sification by numerical methods. Bulletin of the British Museum (Natural History), 29(3): 120-199. Crespo, J. A. 1975. Ecology of the Pampas gray fox and the large fox (culpeo), pp. 179-191. In Fox, M. W., ed., The Wild Canids. Van Nostrand Reinhold Co., New York. 468 FIELDIANA: ZOOLOGY DiETZ, J. 1984. Ecology and social organization of the maned wolf (Chrysocyon brachyurus). Smithsonian Contributions to Zoology, 392: 1-55. Ewer, R. F. 1973. The Carnivores. Cornell University Press, New York. Fischer, G. von Waldheim. 1814. Zoognosia Tabulis Synopticis lUustrata. Vol. 3, Quadrupedum reliquo- rum, cetoriim et monotrymatum descriptionem con- tinens. Typis Nicolai S. Vsevolozsky, Mosquae, 605 pp. Fox, M. W. 1975. Evolution of social behavior in can- ids, pp. 429-460. In Fox, M. W., ed.. The Wild Canids. Van Nostrand Reinhold Co., New York. GiEBEL, C. G. 1855. Die Saugethiere in zoologischer, anatomischer, und palaeontologischer Beziehung um- fassend dargestellt. A. Abel, Leipzig, 1,108 pp. Gray, J. E. 1837. Descriptions of some new or little known Mammalia principally in the British Museum Collection. Magazine of Natural History (Charles- worth's New Series), 1: 577-587. Hershkovitz, p. 1961. On the South American small- eared zorro Atelocynus microtis Sclater (Canidae). Fieldiana: Zoology, 39(44): 505-523. HoFFSTETTER, R. 1952. Lcs Mammif6r6s Plcistoccnes de la Republique de I'Equateur. Memoires de la So- ciete Geologic, France (nouvelle serie), 31(66): 1-39 1 . Hough, J. R. 1 948. The auditory region in some mem- bers of the Procyonidae, Canidae and Ursidae. Its sig- nificance in the phylogeny of the Camivora. Bulletin of the American Museum of Natural History, 92(2): 70-118. Illiger, K. 1815. Uberlick der Saugethiere nach ihrer Verteilung uber die Weittheile. Abhandlungen der (Koniglichen preussischen) Akademie der Wissen- schaften, Berlin, 1804-1811: 39-159. Jacobs, L. L., and E. H. Lindsay. 1981. Prosigmodon oroscoi, a new sigmodont rodent from the late Tertiary of Mexico. Journal of Paleontology, 55(2): 425-430. Jaksic, F. M., R. p. Schlatter, and J. L. Yanez. 1 980. Feeding ecology of central Chilean foxes, Dusicyon culpaeus and Dusicyon griseus. Journal of Mammal- ogy, 61(2): 254-260. Kerr, R. 1792. The animal kingdom or zoological system, of . . . Linnaeus: Class L Mammalia [Class II. Birds] . . . being a translation of that part of the Sys- tema Naturae, as lately published, by . . . Professor Gmelin. ... 2 pts. in 1 vol. J. Murray and R. Faulder, London, xii [xviii] + 644 pp. Kraglievich, J. L. 1952. Un canido del Eocuartario de Mar del Plata y sus relaciones con otras formas brasilefias y norteamericanas. Revista del Museo Mu- nicipal de Ciencias Naturales y Tradicional de Mar del Plata, 1: 53-70. Kraglievich, L. 1928. Contribucion al conocimiento de los grandes canidos extinguidos de Sud America. Anales de la Sociedad Argentina de Ciencias Natu- rales, 106: 35-66. . 1930. Craneometria y clasificacion de los ca- nidos sudamericanos especialmente los argentinos ac- tuales y fosiles. Physis (Revista de la Sociedad Argen- tina de Ciencias Naturales), 10: 35-73. KuRTEN, B. J., and E. Anderson. 1980. Pleistocene Mammals of North America. Columbia University Press, New York. Langguth, A. 1969. Die sudamerikanischen Canidae unter besonderer Beriicksichtigung des Mahnenwolfes Chrysocyon brachyurus Illiger. Zeitschrift fur wissen- schaftliche Zoologie 179(1, 2): 1-88. . 1970. Una nueva clasificacion de los canidos sudamericanos, pp. 129-143. In Klappenbach, M., et al., eds., IV Congresso Latinoamericano de Zoologia, Actas. Vol. 1. Caracas, Venezuela. . 1975. Ecology and evolution in the South American canids, pp. 192-206. In Fox, M. W., ed., The Wild Canids. Van Nostrand Reinhold Co., New York. Leidy, J. 1858. Notice of remains of extinct vertebrata from the valley of the Niobrara River, collected during the exploring expedition of 1857, in Nebraska, under the command of Lieut. G. K. Warren, U.S. Top Eng., by Dr. F. V. Hayden, Geologist to the Expedition. Proceedings of the Academy of Natural Sciences of Philadelphia, 10: 20-29. Linnaeus, C. 1 766. Systema Naturae, vol. 1 , 1 2th ed. Laurentii Salvii, Uppsala, Sweden, 532 pp. Long, A., and P. S. Martin. 1 974. Death of American ground sloths. Science, 186: 638-640. Lund, P. W. 1839a. Blik paa Brasiliens Dyreverden for sidste jordomvaeltning. Anden Afhandlinger Pat- tedyrene. Danske Videnskabemes Selskabs Naturvi- denskabelig og Mathematisk Afdeling, 8: 61-144. . 1839b. Coup d'oeil sur les esp^ces eteintes de mammiferds du Bresil; extrait de quelques memoires presentes a I'Academie royale de Copenhaguen. An- nales des Sciences Naturelles (Paris), 11: 214-234. 1 842. Blik paa Brasiliens Dyreverden for sidste jordomvaeltning. Fjerde Afhandling: Fortsaettelse af Pattedyrene. Danske Videnskabemes Selskabs Na- turvidenskabelig og Mathematisk Afdeling, 9: 137- 208. MacFadden, B. J. 1979. Diversification and biogeog- raphy of the one-toed horses Onohippidium and Hip- pidion. Postilla, 175: 1-10. Marshall, L. G. 1978. Evolution of the Borhyaenidae, extinct South American predaceous marsupials. Uni- versity of California Publications in the Geological Sciences, 117: 1-89. Marshall, L. G., A. Berta, R. H. Hoffstetter, R. Pascual, O. a. Reig, M. Bombin, and A. Mones. 1984. Geochronology of the continental mammal- bearing Quaternary of South America. Palaeoverte- brata, Memoire Extraordinaire: 1-76. Marshall, L. G., R. F. Butler, R. E. Drake, and G. H. Curtis. 1982. Geochronology of type Uquian (Late Cenozoic) land mammal age, Argentina. Science, 216: 986-989. Mercerat, a. 1 89 1 . Caracteres diagnosticos de algu- nas especies de Creodonta conservados en el Museo de La Plata. Revista Museo de La Plata, 2: 51-56. Molina, G. I. 1782. Saggio sulla Storia Naturale del Chile, 367 pp. NoRDENSKioiD, E. 1908. Ein neuer Fundort fiir Sau- getier fossilien in Peru. Arkiv fiir Zoologi, 4(1 1): 1- 22. BERTA: SOUTH AMERICAN CANIDAE 469 Osgood, W. H. 1934. The genera and subgenera of South American Canidae. Journal of Mammalogy, 15: 45-50. Pascual, R., et ai_ 1966. Paleontografia Bonaerense. Fasciculo IV: Vertebrata. Comision de Investigacion Cientifica, Provincia de Buenos Aires, La Plata, Ar- gentina. ScHREBER, J. C. D. 1775. Die Saugethiere in Abbil- dungen nach der Nature mit Beschreibungen. W. Wal- ther. Leipzig, 3: 281-590. ScLATER, P. L. 1882. Reports on the additions to the society's menagerie in June, July, August, September, and October. Proceedings of the Zoological Society of London, 1882:630-631. Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bulletin of the Amer- ican Museum of Natural History, 85: 1-350. Smith, C. H. 1839. The natural history of dogs, pp. 1- 267. In Jardine, W., ed.. The Naturalist's Library. Mammalia, vol. 9. W. H. Lizars, Edinburgh. SoRiA, D., AND E. Agltirre. 1 976. El canido de Layna: revision de los Nyctereutes fosiles. Trabajos sobre Neogeno-Cuatemario, Seccion de Paleontologia de Vertebrados y Humana (I.L.C., C. S.I.C, Madrid), No. 5, Miscelanea Neogena Alberdi and Aguirre, eds., pp. 83-115. Stevens, M. S. 1965. A new species of Urocyon from the Upper Pliocene of Kansas. Journal of Mammal- ogy, 46(2): 265-269. Thenius, E. 1954. Zur Abstammung der Rotwolfe (Gattung Cuon Hodgson). Osterreichische Zoolo- gischc Zeitschrift, 5(3): 378-387. Thomas, O. 1900. New Soiith American mammals. Annals and Magazine of Natural History, (7)5: 148- 153. . 1914. The generic and subgeneric names of the South American Canidae. Annals and Magazine of Natural History, (8)13: 350-352. ToNNi, E. p., AND G. P. PouTis. 1981. Un gran canido del Holoceno de la Provincia de Buenos Aires y el registro prehispanico de Canis (Canis) familiaris en las areas pampeana y patagonica. Ameghiniana, 18(3- 4): 251-265. Torres, R. V., and I. Ferrusquia, V. 1981. Cerdocyon sp. nov. A (Mammalia, Camivora) en Mexico y su significacion evolutiva y zoogeografia en relacion a los canidos sud americanos. Anais II Congresso Latino- Americano de Paleontologia, Porto Alegre, Brazil, 2: 709-719. Van Gelder, R. G. 1978. A review of canid classifi- cation. American Museum of Natural History Novi- tates, 2646: 1-10. Webb, S. D. 1974. Pleistocene llamas of Florida, with a brief review of the Lamini, pp. 170-213. In Webb, S. D., ed.. Pleistocene Mammals of Florida. Univers- ity of Florida Press, Gainesville. . 1978. A history of savanna vertebrates in the New World. Part II: South America and the Great Interchange. Annual Review of Ecology and System- atics, 9: 393-426. YaSez, J., AND F. Jaksic. 1978. Rol ecologico de los zorros (Diisicyon) en Chile central. Anales del Museo de Historia Natural, 11: 105-1 12. Appendix The following are notes on the evaluations of characters listed in Table 3. Characters 1-4— These derived characters dis- tinguish the living Canini from the extinct Bo- rophaghini, a Miocene-Pliocene lineage of hyaena-like dogs. Characters 5-6— The simple construction of the upper incisors lacking accessory cusps distin- guishes Vulpes. Urocyon. and Otocyon from more advanced canids. Another diagnostic feature of the Urocyon group is the development of a broad paroccipital process with a tip that does not ex- tend below the body of the process. Character 7— The presence and expansion of a frontal sinus is the derived condition among advanced canids (Tedford & Taylor, in prep.; Berta, in press; fig. 9). Character 8— The increase in size of P relative to I'"- is recognized as a derived state, diagnostic of advanced canids. Character 9— The absence of an entepicondylar foramen on the humerus distinguishes all ad- vanced canids from the Urocyon group. Character 10— Primitive canids including "Ca- /2/5" davisi possess a moderately arched or near- ly flat zygoma with an everted jugal. The derived state, a strongly arched zygoma with an inverted jugal, unites advanced canids. Character 1 1— The development of a large an- gular process with expansion of various fossa for branches of the medial pterygoideus muscle 470 HELDIANA: ZOOLOGY is the derived condition among advanced canids (Tedford & Taylor, in prep.; Berta, in press; fig. 10). Character 12— The primitive condition, short palatines extending to or just anterior to the toothrow, is observed in most advanced canids. Long palatines extending to or beyond the toothrow is recognized as one of two derived states of this character and it unites the Dusicyon group. Character 1 3— The broad, low coronoid process diagnostic of the Dusicyon and Cerdocyon groups is a marked departure from the narrow, high coronoid process of other advanced canids (e.g., Canis and Chrysocyon). Characters 14-16— These dental features indi- cate a trend toward greater complication of cusps on the lower molars, characteristic of the Du- sicyon group (Berta, in press; fig. 8). Character 1 7— The frontal sinus, although large, does not penetrate the postorbital process, rec- ognized as the least derived sinus condition. This character unites the Dusicyon group. See com- ments under character 7. Character 18— Canines that are proportionally small relative to the cheekteeth is a derived character which distinguishes the Cerdocyon clade from the Dusicyon clade. Character 19— Further expansion of the ptery- goid fossa is recognized as continued modifi- cation of the trend established with character 11. Characters 20-23— Short extremities are de- rived features of the Cerdocyon clade. Characters 24-26— These derived dental and mandibular characters distinguish Cerdocyon and Nyctereutes from Atelocynus and Speothos. Development of a subangular lobe has been cor- related with an insectivorous diet (Ewer, 1973). Character 27— Short nasals which rarely extend beyond the maxillary-frontal suture is a derived character which \xn\\es Atelocynus and Speothos. Character 28— The marked reduction in size of the frontal sinus in Speothos, Atelocynus, and Nyctereutes is a reversal of the derived condi- tion. See comments under character 7. Character 29— Posterior expansion of the fron- tal sinus is regarded as a further modification of the trend established with characters 7 and 17. Character 30— Expansion of the fossa for the superior branch of the medial pterygoideus muscle is regarded as a further modification of the trend established with character 1 1 . Character 31— The further increase in size and complication of P relative to I'"- observed in Canis and Chrysocyon is recognized as a more derived state of character 8. BERTA: SOUTH AMERICAN CANIDAE 471 Comparative Cytogenetics of South American Deer Angel E. Spotorno, Nadir Brum, and Mariela Di Tomaso ABSTRACTS Karyotypes of a male Hippocamelus bisulcus from Chile had 2n = 70, and those of a female Blastoceros bezoarticus from Uruguay, 2n = 68, both with NF = 74. Most chromosomes were similar, and at least pairs 1 (also NOR-bearing), the X long arms, the small metacentric pair, and Hippocamelus 2 and 4 with Blastoceros 32 p and q were identical in G-band patterns. C-bands were large and paracentromeric, but absent in the small metacentric pair that is shared with other deer. Similarity relationships based on X shapes and lengths compared in a karyo- idiogram are {Pudu- Blastoceros, Hippocamelus-Mazama; Odocoileus) Cervinae. All the other large metacentric chromosome (Blastoceros 32 and Mazama 1 through 9) are inferred to be unique, favoring the hypothesis of centric and tandem fusions. Such metacentrics and the distribution of 2n and NF suggest the occurrence of many parallel and independent fusions throughout the three phyletic lines of living Cervidae. Los cariotipos de un macho Hippocamelus bisulcus de Chile tienen 2n = 70 y los de una hembra Blastoceros bezoarticus de Uruguay 2n = 68; ambos tenian un NF = 74. La mayor parte de los cromosomas fueron similares en ambas especies, y por lo menos los pares 1 (que es portador del sector NOR), los brazos largos de los cromosomas X. El pequeno par metacen- trico, e Hippocamelus 2 y 4 con Blastoceros 32 p y q eran identicos en sus patrones de bandas G. Las bandas C eran grandes y paracentromericas pero estaban absentes en el pequeiio par metacentrico que es compartido con otros ciervos. Las relaciones de similitud basadas en las longitudes y formas comparadas del X en un cario-idiograma son (Pudu- Blastoceros, Hippo- camelus-Mazama; Odocoileus) Cervinae. Todos los otros cromosomas metacentricos grandes (Blastoceros 32 y Mazama 1 al 9) se inhere que son unicos, lo que favorece la hipotesis de fusiones centricas y en tandem. Tales metacentricos y la distribucion de 2n y NF sugieren la ocurrencia de varias fusiones paralelas e independientes en las tres lineas fileticas de Cervidae vivientes. Cariotipos de um macho Hippocamelus bisulcus do Chile contam com 2n = 70, e os de uma femea Blastoceros bezoarticus do Uruguai, com 2n = 68, ambos com NF = 74. A maior parte dos cromossomos sao parecidos. Os pares 1 (tambem contendo NOR), os brazos longos do X, o pequeno par metacentrico; Hippocamelus 2 e 4, e Blastoceros 32 p e q, todos possuem padroes identicos na faixa G (G-band). As faixas C (C-band) sao grandes e paracentromericas porem nao ocorrem no par metacentrico pequeno, comuns nas outras especies de veados. Baseando- From the Departamento Biologia Celular y Genetica, Facultad de Medicina, Universidad de Chile, Casilla 70061, Santiago 7, Chile (Spotorno); and Divisi6n Ci- togenetica, Instituto de Investigaciones Biol6gicas Cle- mente Estable, Av. Italia 3318, Montevideo, Uruguay (Brum and Di Tomaso). SPOTORNO ET AL.: CYTOGENETICS OF SOUTH AMERICAN DEER 473 se nas formas e comprimentos dos seus cromossomos X (comparados atraves de cario-idio- gramas) as semelhan^as sao (Pudu-Blastoceros, Hippocamelus-Mazama; Odocoileus) Cervi- nae. Infere-se que todos outros cromossomos grandes metacentricos {Blastoceros 32 e Mazama 1-9) sao exclusivos, o que sugere ocorrencias de fusoes alinhadas (tandem) e centrals. Tais metacentricos, junto com a distribuifao dos 2n e NF, sugerem a ocorrencia de varias fusoes paralelas e independentes nas tres linhas fileticas dos Cervidae agora existentes. Introduction Materials and Methods Of the 1 1 species of deer now living in South America (Cabrera, 1961), only three species are cytologically known: Odocoileus virginianus (Wurster & Benirschke, 1967), Mazama ameri- cana (Taylor et al., 1969) and Pudu puda (Kou- hscher et al., 1972; Spotomo &. Fernandez, 1975). Furthermore, no detailed chromosome analysis or G- and C-band descriptions have been published yet. We report here cytogenetic descriptions and comparisons of material from Hippocamelus bisul- cus and Blastoceros bezoarticus, which complete representative karyotypic descriptions of all South American genera of deer recognized by most au- thorities (Koopman, 1967). Chromosomes can provide reliable taxonomic characters when they are examined in detail, par- ticularly the specific patterns of G- or R-bands or certain kinds of DNA (C-bands and NOR). More- over, mechanism of change in character-states can eventually be inferred. These analyses have been rarely done on cervid chromosomes. With newly reported data, we think there is now enough data to compare the chromosomes of American deer systematically. An adult male Hippocamelus bisulcus (fig. la), captured in Region de Aisen, Chile, and kept alive by CONAF (Chilean National Forestry Service), was the source of a 10-ml blood sample. Conven- tional 72-hour blood cultures were done in TC- 199 with phytohemagglutinin. Cells were then treated with Colcemid, 0.075 M KCl hypotonic solution, fixed in Camoy, dropped on clean slides, and air-dried. A female Blastoceros bezoarticus (sister of the male shown in fig. lb), captured in Departamento de Sal to, Uruguay, died by accidential injury dur- ing the trip to Montevideo. Immediately, bone marrow samples were obtained. Cells were incu- bated in TC- 1 99 with 0.04% colchicine plus 3 drops of Liquemin for 4 hours and then treated as above. Slides were Giemsa-stained. G- and C-bands and NOR were induced by treating slides with the methods of ChiareUi et al. (1972), Sumner (1972), and Rufas et al. (1982), respectively. A selected number of metaphases were photographed with fine-grain high-contrast copy-film. Measurements were made on enlargements us- FiG. 1 . Specimens examined: a, Hippocamelus bisulcus male in Reserva de Penuelas (photograph by A. Spotomo); b, Blastoceros bezoarticus male in Reserva Maldonado (photograph by H. Cardoso). 474 HELDIANA: ZOOLOGY ii 11 M •• ti li M 8 14 At AA 15 21 22 M Aft •« M M M I 34 X Y l5iiiieooAA<)n« AO H II •• II trt ^^^ •• •• •• •• Aft A*^ A* 15 21 •##i ## M « 10 urn ^32 33 XX Fig. 2. Karyotype of single cells from: top, Hippocamelus bisulcus male; and bottom, Blastoceros bezoarticus female. SPOTORNO ET AL.: CYTOGENETICS OF SOUTH AMERICAN DEER 475 chromosome nomenclature (Levan et al.1964) 95% confidence limits 1 ■ Hippocamelus bisulcus O Pudu puda ▲ Blastoceros bezoarticus ODOCOILEINAE ^ Mazama a. temama * Odocoileus virglnianus ^ Odocoileus hemionus CERVINAE * Platyceros dama I'll ^ 2 3 A Short arm p % Fig. 3. Karyo-idiogram displaying relative chromosome lengths of some deer; chromosome size and shape can be read on diagonals. Numbers and letters are from original karyotype descriptions. Some overlapping chromosomes are not displayed. 476 HELDIANA: ZOOLOGY ing the best single chromatid per pair, and values were transformed into percentages of the total hap- loid plus X set. Such relative lengths are displayed in a scatter diagram called a karyo-idiogram (Spotomo et al., 1979; Spotomo, 1985), a useful device that allows eventual chromosome identi- fication and comparison through two indep)endent variables, total chromosome size (short arm length plus long arm length), and centromeric index ( 1 00 times short arm length divided by total chromo- some length). This procedure assumes the conser- vation of total nuclear material. Although this as- sumption is generally true for mammals, it can be validated by C-banding techniques which detect heterochromatin-containing satellite DNA or by marker chromosomes (Spotomo, 1977). A karyo- idiogram can also portray confidence intervals (Spotomo et al., 1979); we use them here to com- pare some centromeric indices calculated without measuring all chromosomes from a karyotype. Chromosomes of other deer used in comparisons were measured from the following sources: Ma- zama americana temama from Jorge and Be- nirschke ( 1 977, p. 7 1 2), Odocoileus virginiana and O. hemionus from Wurster and Benirschke (1967, p. 275), and Platyceros dama from Wurster and Benirschke, 1967, p. 277). Results The diploid number of Hippocamelns bisulcus was 70 in almost all of the 30 cells analyzed, with 33 telocentric pairs of decreasing size (fig. 2), a submetacentric (here labeled 34), and a hetero- morphic pair of metacentrics. The latter are the largest and the smallest elements of the karyotype, having relative sizes of 4.3% and 0.9% of the total haploid set (means from eight cells measured; see also fig. 4); they probably are the X and Y chro- mosomes, respectively. The total number of chro- mosome arms per cell (FN or fundamental num- ber) was 74. The diploid number of Blastoceros bezoarticus was 68 in six cells analyzed, with 31 telocentric pairs of decreasing size (fig. 2) and three metacen- trics, here labeled 32, 33, and a presumptive X. These had relative sizes of 6.9%, 3.3%, and 5.6% of total haploid set (means from six cells). The FN was also 74. Many similarities and some differences in the length and shape of chromosomes were immedi- ately detected when both karyotypes were com- pared, particularly with the aid of a karyo-idi- ogram constmcted from their measurements (fig. 3). Thus, 31 telocentrics, ranging from 1.6% to 4.6% of haploid set, and one metacentric having a 3%, tended to cluster at the same positions in the karyo-idiogram, indicating that they share similar size and shape. In contrast, the large chromosome 32 of Blas- toceros was not found in Hippocamelus (fig. 3), but the latter had two additional telocentrics whose sizes corresponded to the short and long arms of this chromosome 32. Another difference was de- tected in the X chromosome. The presumptive X of Blastoceros seemed to be larger than that of Hippocamelus, mainly in the short arm. This is reflected in their statistically different centromeric indices, 46.49 ± 3.89 and 38.15 ± 2.60, respec- tively (confidence intervals do not overlap in fig. 3; N = 8 for Blastoceros, N = 1 5 for Hippoca- melus). G-banded karyotypes of the two species (illus- trated in fig. 4) allowed a better identification of chromosome pairs than that displayed in Figure 2. Many gross similarities were found when band- ed chromosomes of the species were compared side by side. This was a difficult task, because cells with similar states of contraction and treatment must be used. Nevertheless, approximate banding patterns were found in Hippocamelus for the fol- lowing Blastoceros chromosomes: 1, 32, 33, and the X long arm (illustrated in fig. 5a). The position of the nucleolar organizing region (NOR) was detected at the largest telocentric pair. Ag-stained plates of Blastoceros exhibited clear terminal dark spots over chromosome 1 only (fig. 5c). The same technique was unsuccessful in Hip- pocamelus, but in the G-banded plates the ho- mologous chromosomes of pair No. 1 were usually associated by faint terminal material (fig. 5b), probably nucleolar material. This suggests the probable position of the NOR. C-bands in almost all chromosomes of both species were paracentric (fig. 5d,e), and one of the largest autosomes (probably the 1 ) and the X chro- mosomes had rather large C-bands near the cen- tromere. This was also shown in the G-bands, where large centromeric light bands were observed at the same positions. The only exception to this pattern was the minute amount of the C-bands in pairs 34 of Hippocamelus and 33 of Blastoceros (fig. 5e). SPOTORNO ET AL.: CYTOGENETICS OF SOUTH AMERICAN DEER 477 II H II II II II ii ii M il •! •• II ti 14 •• M •• II tl •• •# 15 21 •i •• •• M •« N •• I- 34 X Y M nil M #••#•• 1 7 it ta •• il «i •• it 14 •iiii«#iii«»t« 22 llll ^r32 33 10 urn ^^32 33 XX Fig. 4. G-banded karyotypes from: top, Hippocamelus bisulcus male; and bottom, Blastoceros bezoarticus female. 478 FIELDIANA: Zendentes, como ordem, tamanho, nivel trofico e afinidade de habitat, zona de vida, e status atual no Chile. Os resultados indicam rela96es positivas significantes entre o numero de especimes no Museu e a abundancia atual no campo (deduzida de varias maneiras diferentes). Conclue-se assim que, pelo menos quanto a fauna Chilena, cole9des em museus oferecem uma base adequada para dedu9des sobre a fauna de um local. Introduction Of course museum specimens form the basis of scientific nomenclature, but further they figure Much of what is known of the biology of Neo- prominently in studies of evolution, descriptive tropical mammals is based on museum sp>ecimens. and functional morphology, genetics, cytology, and even ecology. Given this essential role, especially From Division of Mammals, Field Museum of Nat- ;» ;„»:„i ^. j:„„ •» •„ • _* . * j ♦ j *u ural History, Chicago, IL 60605-2496 (Patterson); and " '"»*»^' ^^"^1^^' " ^^ '"^PO^ant to understand the Department of Evolutionary Biology, Northwestern limitations of museum collections. Some limita- University, Evanston, IL 60201 (Feigl). tions are well known, as the nature of preservation PATTERSON & FEIGL: FAUNAL REPRESENTATION IN MAMMAL COLLECTIONS 485 (skulls, fluid-preserved) obviously determines the kinds of questions that can be asked. However, the limitation imposed by collection bias, by col- lections which represent only part of a native fau- na, remains unexplored. How representative are museum collections of the natural Neotropical mammal fauna? As examples of the sort of bias we consider here, one can point to the monotypic sigmodon- tine genera Abrawayaomys Cunha and Cruz, 1 979, Scolomys Anthony, 1929, Juscelinomys Moojen, 1965, Podoxymys Anthony, 1929, Galenomys Thomas, 1916, and Anotomys Thomas, 1906, all known from one or a few localities and represented in the world's museum collections by a mere hand- ful of specimens (Nowak & Paradiso, 1983). At a finer level, the ichthyomyine genus Daptomys An- thony, 1929, is comprised of three species, D. ven- ezuelae, D. peruviensis, and D. oyapocki, repre- sented in world collections by three, one, and one specimens, respectively (Nowak & Paradiso, 1 983). Finally, of the 47 recognized species of the murine opossum Marmosa Gray, 1821 {sensu lato; Hon- acki et al., 1982), four are known only from the type localities (M. agricolai, M. andersoni. M. sca- pulata. and M. tatei) and two others from the vi- cinity of their type localities (A/, handleyi and M. cracens). For this short list at least, how different our understanding of these faunas might be if mu- seum collectors had camped at different localities, used different varieties of traps or baits, placed traps in different positions, or had delayed their travel plans by as little as a week. Fieldworkers are aware firsthand of the serendipity involved in certain captures, and many have spent days and even weeks at a locality, trying in vain to secure additional representatives of a new or poorly rep- resented form. However, little attention (beyond the routine caveats which often accompany faunal listings) and no systematic scrutiny have been giv- en to the question these data raise: again, how representative are museum mammal collections? The question is a frustrating one, for it imme- diately leads full circle. If our knowledge of Neo- tropical mammals depends even in part on mu- seum collections, how can the representation of museum collections be independently assessed? The question as posed is tautologous. However, introducing certain assumptions about the native fauna permits an investigation of the question of representation. Other studies have shown that the population density of a species in a given area is related to several morphological and ecological characteristics, namely body size, trophic habits, and habitat affinities. A given area can support fewer animals of large size, other factors being equal, because of the greater energy demands of larger biomass (e.g., Peters & Raelson, 1984). Sim- ilarly, that area supports fewer carnivores than omnivores, and fewer omnivores than herbivores, because of the lower available biomass for con- sumption. Finally, species having catholic habitat affinities can sustain a greater number of individ- uals than species which are highly restricted in the habitats that they can exploit, other factors being equal (e.g., Patterson, 1982). While many other factors are known to affect population density, knowledge of only these three may permit refined inference of patterns of spatial and temporal dis- tribution (e.g.. Brown, 1971; Patterson, 1 984). Ap- plying these assumptions to a faunal list, we can hypothetically specify the relative number of in- dividuals in nature expected for each species, and use this hypothetical expectation to independently assess faunal representation in museum mammal collections. We elected to study the representation of Field Museum's mammals of Chile for several reasons: (1) the Chilean mammal fauna is thoroughly stud- ied and is one of the best known faunas in the Neotropics (Pine, 1982); its faunal list can be con- sidered virtually complete; (2) this fauna was the subject of systematic and expressly faunistic stud- ies by FMNH curator W. H. Osgood, who led two expeditions there in 1922-1923 and 1939-1940; as a result, Field Museum maintains the world's premier Chilean mammal collection (pers. obs.; Yaiiez, 1982); and (3) a listing of Chilean mam- mals that are endangered, vulnerable, rare, or in- adequately known has recently appeared (Miller et al., 1983), yielding an independent criterion for determining natural population numbers. Materials and Methods Species of mammals occurring in Chile were assembled from the primary and secondary liter- ature. Osgood's (1943) treatise summarizes the earlier literature. Discrepancies between Osgood's faunal list and that presented in Table 1 reflect additions to the fauna from more recent collecting, as well as taxonomic reappraisals of previously known forms. Many additions were made shortly after Osgood's work with the publication of Mann's ( 1 945) mammals of Tarapaca, and records from this region were augmented by Spotomo 486 HELDIANA: ZOOLOGY (1976) and Pine et al. (1979). Several additional sp)ecies were added by taxonomic revisions of forms Osgood reported (e.g., Phyllotis xanthopygus, My- otis atacamensis). Documentation for the extend- ed list in Table 1 may be found in Mann (1978), Pine et al. (1979), Pine (1973), Spotomo (1976), Honacki et al. ( 1 982), Patterson et al. ( 1 984), Pear- son (1984), Walker et al. (1984), and literature cited therein. Introduced species (e.g., beaver, rab- bits, hares, European rats) were excluded from the analysis. Finally, a native species reported to occur in Chile, Felis geoffroyi, was deleted from the Chil- ean fauna on the authority of Honacki et al. ( 1 982). For each terrestrial species occurring in Chile, we determined body size, trophic habits, and hab- itat affinities. Body size, as measured by head and body lengths (most taxa) or shoulder heights (un- gulates), was taken from Osgood ( 1 943), the FMNH collection, or primary sources, using means or midpoints of ranges. An average value was used for sexually dimorphic taxa (e.g., mustelids, pin- nipeds). We used only four categories of trophic-habitat characteristics to reduce subjectivity and to gain suitable sample sizes for statistical analyses (cf Eisenberg, 1981): carnivores, insectivorous car- nivores, sanguinivores, piscivores, and omnivores were grouped together as carnivores; insectivores included both insect-eating and insectivorous-fru- givorous taxa; animals subsisting on granivorous, graminivorous, or herbivorous diets were consid- ered herbivores. This category included many sjjecies and was further subdivided into gener- alized (marked by dietary and habitat breadth) and specialized (restricted in dietary and habitat breadth) components. Information on diet and habitat was taken from all available sources, but especially Osgood (1943), Hershkovitz (1962), Mann (1978), and Pearson (1983, 1984). The con- tinuum between generalized and specialized habits of herbivores was assessed relative to the distri- bution of resources in Chile; thus, Auliscomys bo- liviensis, which has broad habitat requirements in Peru (Pearson, 195 1) but occurs in only a minute portion of Chile, was considered a specialized her- bivore. The four trophic-habitat categories were coded 1-4 as carnivores, insectivores, specialized herbivores, and generalized herbivores, respec- tively. The current status of these species in Chile was taken directly from Miller et al. ( 1 983). Taxa were assigned numerical codes 1-5 according to the sta- tus given by Miller et al.: endangered, vulnerable, rare, inadequately known, and secure, respective- ly. Where Miller et al. gave different status as- sessments to subspecies of polytypic forms, the subspecies were assigned numerical codes and an average determined, rounding down to the nearest integer. Species or subspecies not explicitly men- tioned in the account of Miller et al. were pre- sumed to be secure. To determine specimen representation for these taxa in the FMNH mammal collections, we used those numbers reported by Osgood ( 1 943) in the text and "Specimens examined" sections of his species accounts. In a few cases, this number is less than the number collected by Osgood and as- sociates and currently in the collection; in these cases we used the number of Osgood-era speci- mens in the collection. We used specimen numbers reported by Osgood rather than the number currently in the collection for two principal reasons. First, FMNH now has fewer specimens of some taxa than Osgood and associates collected, due to specimen exchange programs and attrition of material on loan. Our primary goal is to assess bias in museum collecting techniques and collections, not the bias in cura- torial exchange programs or collection users. Sec- ond, subsequent fieldwork by R. E. Martin on Octodon and B. D. Patterson on Akodon, Oryzo- mys, and marsupials has greatly augmented FMNH holdings of these taxa; inclusion of specimens ob- tained in these highly focused studies could strong- ly confound the results. The data were analyzed using the Statistical Analysis System (SAS) procedures at the Univer- sity of Chicago Computation Center. Correlation procedures used variable ranks in nonparametric analyses. Because of the lesser efficiency of non- parametric procedures in comparison to para- metric ones, and hence the greater probability of Type II errors, we report values having probability levels between 0.05 and 0. 10 as being "marginally significant." This assumes that larger sample size (i.e., a richer fauna or faunal category) would suf- fice to generate a traditionally significant value {P < 0.05). Results The native Chilean mammal fauna includes six orders, 55 genera, and 93 species (table 1). The taxonomic distribution is as follows: Marsupialia with three genera (5%) and three species (3%); Chi- roptera with seven genera (13%) and 10 species PATTERSON &. FEIGL: FAUNAL REPRESENTATION IN MAMMAL COLLECTIONS 487 3 ^ X) X) A CO js x: 9t T3.S i s £ 2 § 13 «* N e^ -•1 "^1 <-» CO n iS - M "3 "S M s §1 ,2° 1 ^^ 2 w •5-8 e •eg H |l 5 3 E 2 3 •9 2? ■s ^ {J « 3 5 0 c: CS 1 JU 0 £ c U l/> _c n E 3 E 2 cd 0 t I/) 1 0 u ?i C C/1 0 — •c a r a u ,< i'' H n I c I y> M tf) W V u c c c c c c c c c c t <<<<<<<<< 'H'2 s « « 0 0 0 > > > c c c rt cd w 333 I- U k. Ill a <3 O .Si -c Ml 3 3 E b O C 3 £ ;55 ?J c c 73 -O C c c «5 u u a: ^ c 3 3 o.| S 2 E o « « o c £ £ 0^33 B >> Uk.L.UUiUUUl-iUk. fcOOOOOOOOOO > .t. .i. .i, .i. .i. .i. .i. .^ .t. .i. £•■"0000000000 cccccccc C C r\ CCCCCCCCCC CC r\ r\ r\ r\ r\ r\ r\ 00000 > > > > > E E e e e CQ CO CQ CQ CQ uouuu ^ E E e E E E e cdcQcdcdcdcdcocQcscd OUUUiOUUOUOCQ OiOiO.O.O.O.O.OiO.O.cQ 0000000000^ '.C .C .C .C .C .C .H '^ ^ £ -O UUUUUUOUUUtS 0000nO>/^Ov cQ cd CO cd cd CO cQ Sooooooo 2 2 .S; .^ .i. .i. .t, .i. .i. (uuEEEEEEE *^ *o R} cd CO ^ CO CO A UWUUUUOUU CQ CO CO CQ CO £ E E E E cO CO CO cO cO uuuuu ■CI F o o o '2 5 5 3 S3 3:5 I ^ HIS c i: .0 ^ ft. 5 3 ,^ ,^ U'J -^ t*, 1*, E E cd cd u-i o o >n o r~ o w^ t^ On O f*^ 1/^ O r^ >D ir> so f«% ■^ vO 2 2 o o > > EE cd cd «2 .6c O 488 HELDIANA: ZOOLOGY la H U a 5 H S E ^S §33§ n CO C4 ti c« C ■c •c •c •c c« wj U u V u u t t fc t t OUUOU""^^^^ iMW^k^UaliNU^WNk^l-^k^li^ OS 0$ CQ 9] c c -c c c c -c c c -c (4 M B'S (/) M '(/) ^ 'M (A '(/I ^ U) (/) (A C C C C cQ cd cd cS cQ CQ cq c -c OOOOO — OOf*^'* — 00 c/5 cJ5 00 W > > W U « 00 V) c/3 c/5 00 on V3 00 c/5 cio do n3 cJ5 c/5 c/5 c«o c/5 c/5 go «5 do c/5 c?5 „ I- fc" „ > o o > o •g > > -^ > •S X) XI u X) CJ U U fl> u ^ a H x: a c w ? U C/3 V5 v u c. c; S 3 5 ^ W U CJ (L> C^ C^ ■« "o u o a g o > o ,D ■t: X) U U b H 43 Si t/: Cj V5 o o o o > > > > XI XI X) X3 k. u U b. u U U U x: j: ji js u "O •o -o -o u u u u u > -^ •;2 cs PI ^^^ £ V3 C/3 C/3 O) v •> I X) X) t: u u u (O « S •^-^^ U U N II o 6 ^ k« a> 1m w a> cd cd c3 cd cd Im o o ^ W2 (/I O ^ 4> ■•3 « t^O •o -o "S V U N _N _N -^ "« 13 2 So a C/) W3 O o o o o o •= • a • a ^ . a .t .t .>. .t.t 2 « .2 2 « oooooooooo ooooo-t — t^O — in O^ O^ O^ a, a, -^ T PATTERSON & FEIGL: FAUNAL REPRESENTATION IN MAMMAL COLLECTIONS 489 n ^ ^ u a 5 H S e ^S S H S B B B B B 'B B 'B 'B ttttttttfcfctfc C C C o c c -c c c c c c c c c -c -c c fctcatttttttttttCtfc HH(-«J333?> oooS90ooot)ooj^osgOg00otjoooS3uoyJ3 uuuuuuuvvuvv u u u Si 2^ £i £i ««222£gfififi 4> V- u u 0 0 0 0 ^hOOOOOOOO 000000000000 0 9 9 > > > > go>>>>>>>> 0 >>>>>>>>>>>> > > > .s .- .- .« .> •S -S -S ^ £ J3 ^ •a -s ^ ^ lo ^ ^ ^ ic io £)£>£i£>SlS>XiX)X>X)X>X> Xi JO Xi "t: "u T, "u. XiXluuuuuuut: ^ b. U k. u U u u ^^j;j;j;j;j3j3j3j3 x:jijsj:j:j:xij:j:j:j:j: x: x: ^ x: •O-O'O'OTJ'OTJ'O-O-O'O'O •0 -p -p "S "2 "S "S _,_,T3'0'OT3'OT3T3'0 V U V U N N N N uvnnnnnnnn 4> NNNNNNNNNNNN N N N .3 .3 .3 .3 N •3 "5 "5 2 2 2 2 •3'522222222 "w i.K.2.K.^^H.S.S.^^K •r; -r; -r; a a u a 99CCCCCCCC 0 8 C/3C/5C/5C«C«C/5C«(/3(/5C/3C/3C/3 («c«c«0000 co{«00000000 (« — O CdcOCQAO^CSCdcQACSCQOAAACQCQ oSiCCcCocccocomOi,ocoi,o^cc 'i'f'1'f'i'l'Bo'So'io'io'i'i'i's U ^ ^ 490 FIELDIANA: ZOOLOGY (1 1%); Edentata with three genera (5%) and three species (3%); Camivora, including pinnipeds, with 1 1 genera (20%) and 19 species (20%); Artiodac- tyla with four genera (7%) and five species (5%); and Rodentia with 27 genera (49%) and 53 species (57%). Four orders (80%), 34 genera (62%), and 56 species (60%) were represented in the 1943 FMNH collections, totaling 1 ,66 1 specimens. The taxonomic distribution of specimens is as follows: Marsupialia (3%), Chiroptera (6%), Camivora (6%), Artiodactyla (1%), and Rodentia (84%). Taxa missing from the 1 943 Chilean collections at FMNH have the following distribution. No marsupial genus or species is missing. Two genera and three of 1 0 species of bats are missing. All three genera and species of edentates are absent. Six of 1 1 genera of carnivores (including all 5 pin- nipeds) and seven of 1 9 species are unrepresented. One of four artiodactyl genera and two of five species are lacking. Finally, nine of 27 rodent gen- era and 20 of 53 species of rodents are not rep- resented. Three of these groups, the edentates, pin- niped carnivores, and artiodactyls, appear by these figures to be grossly underrepresented in the col- lections, but whether this is due to bias in col- lecting effort and efficiency or to their morpho- logical and ecological characteristics remains to be determined. We consider the evident features of this data set in greater detail below. The first section focuses on the taxonomic, trophic-habitat, and morpho- logical composition of the Chilean mammal fauna, the second on the representation of this fauna in the FMNH Chilean mammal collections. Faunal Analyses Across all taxa, there is the expected correlation between trophic-habitat affinities and order, and a negative correlation between trophic-habitat af- finities and size (r, = —0.22; P = 0.03). Current status in Chile is strongly correlated with body size (r^ = -0.56; P < 0.001), but not with trophic- habitat affinities (r, = 0. 1 7; NS). Contingency table analysis of current status by life zone character- istics indicates a significantly larger proportion of endangered aquatic species than would be ex- pected by chance alone. Results of the contingency table analysis should be considered tentative, how- ever, because of small expected numbers in certain cells. When the taxa are stratified by life zone and correlations made within groups, neither the aquatic or volant groups present interesting rela- tionships, probably because both lack the diversity of habits, habitats, and behaviors requisite for such patterns. However, patterns are evident in the terrestrial group that support those apparent in the fauna as a whole. Trophic habits and size are strongly correlated (r^ = -0.29; P < 0.02). As in the fauna as a whole, current status in Chile is significantly related to body size (r^ = -0.56; P < 0.001); this correlation excludes the large pin- nipeds of conservation concern and the small bats with presumably secure populations. As before, current status is not correlated with the trophic- habitat affinities of any group (r^ = 0.14; NS). When the taxa are stratified by order, several of these patterns disappear. Only the Camivora and Rodentia are considered here because of sample size. Neither group shows a significant correlation between trophic habits and size (carnivores, r, = 0; rodents, r, = 0.25; P < 0.08). However, both groups exhibit significant or marginally significant correlations between life zone and size (camivores, r, = -0.68; P < 0.002; rodents, r, = -0.24; P < 0.09). Neither group shows a relation between sta- tus and life zone, but rodents (and not camivores) show a significant correlation between current sta- tus and body size (r^ = -0.40; P < 0.005), larger rodents being progressively more vulnerable than smaller ones. Collection Analyses For the Chilean fauna as a whole, the number of specimens in the FMNH mammal collection is correlated with trophic-habitat affinities (r, = 0.24; P < 0.02), there being fewer specimens of cami- vores and insectivores than herbivores of both categories. Also, there are significantly fewer spec- imens of large taxa, as shown by the inverse rela- tion of number and body size (r, = —0.23; P < 0.03). Additionally, there is a highly significant relationship between the number of specimens and the current status of taxa in Chile (r^ = 0.25; P < 0.02). Similar results emerge when the fauna is strat- ified by life zone. There is a significant relationship between body size and number of specimens only among aquatic forms— aquatic, r^ = —0.88 (P < 0.005); terrestrial, r, = -0.21 {P < 0.08); and volant, r, = -0.36 (NS)— although the two groups showing this relationship have small sample sizes. For the terrestrial group (N = 75), number of spec- imens and trophic-habitat affinities are related (r, = PATTERSON & FEIGL: FAUNAL REPRESENTATION IN MAMMAL COLLECTIONS 491 (0 z UJ 300 200- O tu a 100 B DO A DOACAAA OLMDDADAAA A A B 600 1800 3000 4200 BODY SIZE Fig. 1. Plot of number of Field Museum specimens of 58 Chilean mammals against body size. Plotting symbols represent number of species: A = one sp)ecies; B = two species, etc. The rank correlation of these variables is -0.32 (P < 0.02). See text for discussion. 0.23; P < 0.05), and number of specimens and current status in Chile are strongly and positively related (r, = 0.29; P < 0.02). When orders are considered separately, nearly all show the inverse relationship between speci- men number and body size, although small sam- ples preclude significance in some cases: marsu- pials, r, = 0.50 (NS); bats, r, = -0.36 (NS); carnivores, r, = -0.47 {P < 0.05); rodents, r, = -0.17 (NS); edentates, r, = 0; and artiodactyls, r, = -0.4 1 (NS). Only the rodents (N = 53) exhibit a significant range of trophic-habitat affinities, and these are correlated with specimen numbers (r^ = 0.47; P < 0.001). Finally, in all represented orders showing variation in current status, there is a pos- itive correlation between status and specimen numbers: marsupials, r^ = 0.87 (NS); carnivores, r, = 0.28 (NS); rodents, r, = 0.22 (NS); and artio- dactyls, r, = 0.43 (NS); however, in none of these cases is the relationship significant. It is noteworthy that these correlations remain if species that are altogether lacking in the FMNH Chilean mammal collection are excluded. Using the 58 species having nonzero numbers of speci- mens in Table 1, specimen number is positively correlated with trophic-habitat affinities (r^ = 0.26; P = 0.05) and negatively correlated with body size (r. = -0.32; P < 0.02). As before, number of specimens is highly correlated with current status in Chile (r, = 0.53; P < 0.001) (figs. 1 and 2). Interestingly, the correlation between specimen number and trophic-habitat affinities disappears when both unrepresented and exceedingly well represented taxa (N > 100) are excluded. Basing correlations on 54 species, this correlation equals 0. 15 (P = 0.27). However, correlations of speci- men number with body size (r^ = —0.26; P < 0.06) and with current status in Chile (r^ = 0.52; P < 0.001) still hold. Discussion In The Mammals of Chile. Osgood (1943) de- scribed the itinerary of the two FMNH Chilean expeditions, undertaken "with the intention of making a survey of the vertebrate fauna of that country" (p. 9). His own appraisal of the resulting collection is as follows: The mammals obtained by these two ex- peditions form a collection vastly larger and more varied than anything previously ex- isting. . . . This collection is still deficient in many respects, but it covers the principal 492 nELDIANA: ZOOLOGY 300 0) z UJ 200 o lU Q. (0 100 V STATUS Fig. 2. Plot of number of Field Museum specimens of 58 Chilean mammals against current status in Chile. Status categories are: S = secure; I = incompletely known; R = rare; V = vulnerable; E = endangered. Plotting symbols represent numbers of species: A = one species; B = two species, etc.; & = 38 species. The rank correlation of these variables is 0.53 {P < 0.001). See text for discussion. faunal areas of Chile and probably furnishes a fairly accurate and nearly complete picture of the whole mammal fauna. This may seem to be a rash statement, not justified by ex- perience in other fields . . . but the main facts seem to be already in hand (pp. 9-10). Results of this analysis suggest that Osgood was correct in his inferences. The two expeditions se- cured most taxa that are reported from the coun- try, and they secured these in numbers broadly indicative of natural abundances. Well represent- ed taxa in the collections tend to represent smaller species at lower trophic levels; these species tend also to have generalized habitat requirements and secure population status. In contrast, unrepre- sented or poorly represented taxa tend to be larger, are restricted to higher trophic levels or specialized habitats, and are often characterized by less secure status. The diversity of collecting techniques and pro- cedures used by Osgood and associates was clearly adequate to sample the Chilean mammal fauna as a whole. Osgood's field notes indicate that he used mouse traps, carnivore traps, shotguns, and rifles in making his collections, and he employed field parties of as many as four coworkers. In addition, he reports having bought Pudu captured by native hunters with dogs and purchased Lutra from ga- teros who hunted them for their pelts. His use of commercial hunters and salvage operations to sample marine mammals accounts for the striking absence of seals and sea lions in the collection. The Chilean pinnipeds had been decimated by sealing operations in the 1 8th and 1 9th centuries, and now occur in highly restricted portions of their former ranges. No taxonomic group presents stronger evidence of collecting bias than do the five pinnipeds miss- ing from the FMNH collections. These indicate the insufficiency of Osgood's sampling of the lit- toral zone. However, even this omission from the FMNH collection pales in comparison to that in- volving forms restricted to northernmost Chile. Most of these are puna species of the central An- des, occurring in southern Peru, Bolivia, and northwestern Argentina, extending into Chile only in the northernmost Tarapaca region. Their ab- sence in the 1 943 collections reflects Osgood's fail- ure to devote sufficient time to sampling these marginal localities, not to biases in the collecting techniques themselves. Lacking such material in his collections, Osgood (1943) did not recognize an important tropical and subtropical component PATTERSON & FEIGL: FAUNAL REPRESENTATION IN MAMMAL COLLECTIONS 493 to the Chilean mammal fauna, although Hellmayr (1932) had earlier recognized such components to the Chilean avifauna. The omission enhanced the apparent endemicity of the Chilean mammal fau- na and did little to aid identification of relation- ships with other regions. Two of the remaining species, Akodon mark- hami and A. hershkovitzi, are island derivatives of mainland species (Pine, 1973; Patterson et al., 1 984) and exhibit restricted geographic ranges that FMNH collecting parties did not sample. A main- land taxon, Aconaemys fuscus, was recently re- stricted by Pearson's revision (1984), so that its known range in Chile is quite small. Similarly, Microcavia australis is a pampas species barely crossing the Chilean frontier in southern Chile, while Chelemys delfini is known only from the holotype. [We dissent with the opinion of Miller et al. (1983), who considered delfini of doubtful validity. The disjunct Geoxus taxon michaelseni is the only other long-clawed rodent occurring near Punta Arenas, and it seems only reasonable to believe that Cabrera could distinguish the sub- stantially different skulls of these forms (cf figures of Geoxus and Chelemys in accounts of '^Notio- mys" in Osgood ( 1 943).] Osgood's failure to secure other taxa, for example Histiotius macrotus, is per- plexing, given the number of localities reported for these species by Mann (1978). For the bats at least, modem records may be attributed to the use of mist nets rather than shotguns to collect bats. The use of trophic-habitat affinities and body size to estimate naturally occurring population densities seems justifiable in view of their contri- butions to other studies. This justification is rein- forced by the strong positive correlations between current status in Chile and the numbers of spec- imens Osgood and associates collected; these cor- relations hold for the fauna as a whole, and also for selected subsets (e.g., stratifications by life zone and by order). The correlations hold in spite of uncontrolled variation in other factors known to affect the current status of Chilean species. For example. Miller et al. (1983) note regional varia- tion in endangerment; species living in central Chile have been most severely impacted by human ac- tivities. The correlation analyses assumed that, during the first half of this century, Osgood and associates should have collected mammals pro- portional to their 1983 abundances. However, the assumption seems approximated in view of the demonstrated causes of species decline, principally exploitation through hunting and habitat deteri- oration. Chile was settled by Europeans in the 1 6th century, and much of its fauna and flora were known by the end of the 1 8th century (e.g., Molina, 1782). Although economic exploitation and hab- itat destruction are accelerating with population growth in many parts of Chile, patterns of species densities are probably broadly comparable to those Osgood encountered. This analysis suggests that it should be possible to predict the species currently lacking from the Chilean mammal inventory. If unbiased collecting techniques sample animal populations propor- tional to their densities, then one could expect that species of carnivores or specialized herbivores might thus far have been overlooked. However, a critically confounding variable in making this in- ference is the geographic range size of species. While carnivores may exist in nature at lower densities than herbivores and thus are sampled less often at a given locality, they tend to have larger ranges (Rapoport, 1982), which means they can be sam- pled at more localities. Averaged across an entire country or across an entire fauna, the expected correlation between numbers of specimens and ex- pected population density holds. But when a fauna has been largely sampled, as is certainly true for Chile's mammal fauna, further additions to the faimal list are apt to be those species with highly restricted geographic ranges, especially those in remote areas (e.g., Akodon hershkovitzi, Patterson etal., 1984). What are the implications of this study for other mammal collections made in other faunas? Chile's mammal fauna is highly peculiar, stemming from its isolation from the remainder of South America by the Atacama Desert to the north and the Andes to the east. Excluding bats, pinnipeds, and forms that barely cross Chile's borders, fully one-third of the genera of Chilean mammals are endemic (Osgood, 1943, p. 36), but the fauna does share higher-level affinities with adjacent areas. The fau- na is also a comparatively depauperate one: the 93 mammal species recorded from Chile stand in contrast to the 1 4 1 species of bats alone that are thought to occur in Colombia (Koopman, 1982). Finally, the Chilean fauna occurs in habitats that are structurally simple by tropical standards and similar in many respects to those of the north tem- perate zone. Thus trapping procedures and expe- riences of north temperate scientists might enable them to sample Chilean habitats more effectively than tropical ones. For these reasons, this study should be repeated in a strictly Neotropical area {sensu Hershkovitz, 1972). Both Suriname and Venezuela have diverse Neotropical mammal fau- 494 FIELDIANA: ZOOLOGY nas that are comparatively well studied (see Hus- son, 1 978; Genoways& Williams, 1979; Handley, 1976, and continuing reports, respectively). Both countries would represent prospective areas for repetition of this study. Acknowledgments This study was an outgrowth of a Senior Linkage Seminar entitled "The biology and conservation of South American mammals" at Northwestern University (EEB C-94) during Spring 1984; we thank classmates for helpful discussion. The senior author thanks Milton Gallardo for fostering his appreciation of the mammals of Chile, and we both thank Ronald H. Pine, Guy G. Musser, and an anonymous reviewer for criticisms of the manuscript. Literature Cited Brown, J. H. 1971. Mammals on mountaintops: Non- equilibrium insular biogeography. American Natural- ist, 105: 467-478. EiSENBERG, J. F. 1981. The Mammalian Radiations: An Analysis of Trends in Evolution, Adaptation, and Behavior. University of Chicago Press, Chicago, 610 pp. Genoways, H. H., and S. L. Williams. 1979. Records of bats (Mammalia: Chiroptera) from Suriname. An- nals of Carnegie Museum, 48: 323-335. Handley, C. O., Jr. 1976. Mammals of the Smith- sonian Venezuelan Project. Brigham Young Univer- sity Science Bulletin, Biological Series, 20(5): 1-89. Hellmayr, C. 1 932. The birds of Chile. Field Museum of Natural History, Zoological Series, 19: 1-472. Hershkovitz, p. 1 962. Evolution of Neotropical cri- cetine rodents (Muridae), with special reference to the phyllotine group. Fieldiana: 2toology, 46: 1-524. . 1972. The Recent mammals of the Neotropical region: A zoogeographic and ecological review, pp. 31 1-431. In Keast, A., F. C. Erk, and B. Glass, eds.. Evolution, Mammals, and Southern Continents. State University of New York Press, Albany, 543 pp. HONACKl, J. H., K. E. KiNMAN, AND J. W. KOEPPL. 1982. Mammal Species of the World: A Taxonomic and Geographic Reference. Allen Press and Association of Systematic Collections, Lawrence, Kansas, 694 pp. HussoN, A. M. 1978. The mammals of Suriname. Zoologische Monographieen van het Rijksmuseum van Natuurlijke Historic, 2: xv + 1-569. KooPMAN, K. F. 1982. Biogeography of the bats of South America, pp. 273-302. In Mares, M. A., and H. H. Genoways, eds.. Mammalian biology in South America. Special Publication Series, Pymatuning Lab- oratory of Ecology, University of Pittsburgh, 6: 1-539. Mann F., G. 1945. Mamiferos de Tarapaca. Obser- vaciones realizadas durante una expedicion al Alto Norte de Chile. Biologica, 2: 23-134. . 1978. Los pequeiios mamiferos de Chile. Gay- ana: Zoologia, 40: 1-342. Miller, S. D., J. Rottmann, K. J. Raedeke, and R. D. Taber. 1983. Endangered mammals of Chile: Status and conservation. Biological Conservation, 25: 335- 352. Molina, G. I. 1782. Saggio sulla storia naturale del Chili. Bologna, 368 pp. NowAK, R. M., and J. L. Paradiso. 1983. Walker's Mammals of the World, vol. II, pp. 569-1362. Johns Hopkins University Press, Baltimore. Osgood, W. H. 1943. The mammals of Chile. Field Museum of Natural History, Zoological Series, 30: 1- 268. Patterson, B. D. 1982. Pleistocene vicariance, mon- tane islands, and the evolutionary divergence of some chipmunks (genus Eutamias). Journal of Mammal- ogy, 63: 387-398. . 1984. Mammalian extinction and biogeogra- phy in the southern Rocky Mountains, pp. 247-293. In Nitecki, M. H., ed.. Extinctions. University of Chi- cago Press, Chicago, 354 pp. Patterson, B. D., M. H. Gallardo, and K. E. Freas. 1984. Systematics of mice of the subgenus Akodon (Rodentia: Cricetidae) in southern South America, with the description of a new species. Fieldiana: Zoology, n.s., 23: 1-16. Pearson, O. P. 1951. Mammals in the highlands of southern Peru. Bulletin of the Museum of Compara- tive Zoology, 106: 1 17-174. . 1983. Characteristics of a mammalian fauna from forests in Patagonia, southern Argentina. Journal of Mammalogy, 64: 476-492. . 1984. Taxonomy and natural history of some fossorial rodents of Patagonia, southern Argentina. Journal of Zoology, London, 202: 225-237. Peters, R. H., and J. V. Raelson. 1984. Relations between individual size and mammalian population density. American Naturalist, 124: 498-517. Pine, R. H. 1973. Una nueva especie de Akodon (Mammalia: Rodentia: Muridae) de la isla de Wel- lington, Magallanes, Chile. Anales del Instituto de la Patagonia, IV(l-3): 423-426. . 1 982. Current status of South American mam- malogy, pp. 27-37. In Mares, M. A., and H. H. Gen- oways, eds.. Mammalian biology in South America. Special Publication Series, Pymatuning Laboratory of Ecology, University of Pittsburgh, 6: 1-539. Pine, R. H., S. D. Miller, and M. L. Schamberger. 1979. Contributions to the mammalogy of Chile. Mammalia, 43: 339-376. Rapoport, E. H. 1982. Areography: Geographical Strategies of Species. Pergamon Press, New York, 269 pp. Spotorno O., a. E. 1 976. Analisis taxonomico de tres PATTERSON & FEIGL: FAUNAL REPRESENTATION IN MAMMAL COLLECTIONS 495 i especies altiplanicas del genero Phyllotis (Rodentia, subspecies of /'/!j'//o//5rfflrw/«/ and their experimental Cricetidae). Anales del Museo de Historia Natural, hybrids. Journal of Mammalogy, 65: 220-230. Valparaiso. 9: 141-161. YASez V., J. 1982. Estado de las colecciones masto- Walker, L. I., A. E. Spotorno, and J. Arrau. 1984. zoologicas nacionales. Publicacion Ocasional, Museo Cytogenetic and reproductive studies of two nominal Nacional de Historia Natural, Santiago, 38: 13-28. 496 FIELDIANA: ZOOLOGY Taxonomic Index Abderitinae 104 Abrawayaomys 486 Abrocoma 312, 377, 404 bennetti S5, 419, 490 cinerea 490 Abrocomidae 404 Abrothrix 350, 351, 357-359, 367 ff. Aconaemys 375, 404, 409 fuscus 404, 406, 407, 490, 494 sagei 404, 406, 490 Agouti 3 1 2 paca 29, 17-20, 23, 25, 29, 37, 43, 50, 62, 69 Akodon 347 ff., 487 aerosus 366, 394 agreste 63 albiventer 35 1, 352, 358, 366, 384, 393, 394 andinus 351, 352, 359, 366, 380 ff., 393, 394, 489 arviculoides 356 azaraeSO, 356, 366, 369, 374, 375, 380 ff., 395 berlepschii 351, 358, 489 boliviensis 69, 351, 357, 358, 365, 366, 380, 393, 395 brachiotis 366, 394 budini 359, 360, 394 canescens 83 CO I i breve 63, 80 cursor 355, 366, 369, 375, 378, 380 ff., 395 dolores 355, 356, 366, 395 hershkovitzi 359, 367, 372, 374, 394, 489, 494 illutea 359, 367 ff. iniscatus 366, 374, 380 ff., 394 jelskii 351, 359, 394 johannis 378 ff. kempi 360, 380, 395 kermacki 367 ff. lanosus 359, 367, 372, 374, 489 longipilis 85, 357, 359, 367 ff., 413 ff., 436, 489 lorenzinii 383 ff. magnus 368, 370, 37 1 , 376 ff. mansoensis 359, 372, 374 markhami 366, 394, 489, 494 Akodon molinae 366, 395 mollis 366, 394 nigrita 351, 358, 359, 378- 381, 386, 395 nucus 366, 387, 394 olivaceus 85, 351, 359, 366, 394, 413 ff., 443, 489 orophilus 366, 394 pacificus 366 /jwfT 366, 380, 383, 385, 386, 390, 393, 394 reinhardti 356, 366, 395 sanborni 35\, 359, 361, 367, 372-374, 489 serrensis 366, 395 sp. 366 surdus 366 tolimae 366, 394 urichi 366, 369, 374, 375, 380, 394 varius 355, 358, 359, 366, 369, ,374, 375, 395 xanthorhinus 83, 85,351, 359, 367, 374, 394, 436, 440, 441, 489 Akodon tini 347 ff. Alces alces 480 Alouatta 20, 46, 232, 236, 240, 256, 258 belzebul 23, 25, 31 caraya 29, 31,36, 57,63 fusca 23, 29, 31,36, 77 seniculus 18, 25, 31, 40, 43, 46,54,55,57,68,74,89, 236, 237, 246 Alouattalges 256-258 corbeti 247, 253 Alouattinae 236, 237 Ametrida 225 Amorphochilus schnablii 185, 488 Andinomys edax 490 Anotomys 486 Anoura caudifer 36, 183 cult rat a 183 geoffroyieS, 183 Antrozous pallidus 152 Aotinae 234 Aotus 232, 253, 258 azarae 31, 57, 63 hershkovitzi 232 lemurinus 232, 234 nancymai 246, 247 nigriceps 68 seniculus 247 sp. 25, 47 trivirgatus 54-57, 232, 234, 236 vociferans 3 1 Arctocephalus 20 australis 20, 75, 489 Argyrolagidae 100, 105-107 Argyrolagoidea 107, 108 Ariteus flavescens 213 ff. Artibeus 147, 163 ff., 187 ff., 225 amplus 164-167 anderseni 168-170, 184, 189, 191, 192 aztecus 169, 170 cinereus6S, 166 ff., 184, 189, 191, 192,223 cinereus group 1 69 concolor 166, 168-170, 223 concolor gvonp 169 frater cuius 184 fuliginosus 164, 184 glaucus 166 ff., 174, 184, 189, 192, 193 glaucus group 1 69, 1 70 gnomus 167 ff., 189, 192 hartii (see also Enchisthenes) 166, 168-170 hartii group 169 intermedius 162 jamaicensis 153, 162, 164- 167, 189, 193, 213 ff. lituratus 36, 64, 153, 162, 164, 184, 189,223 phaeotis 153, 162, 169-171, 189, 193 ff, 207, 223 phaeotis group 1 69, 1 70 planirostris 32, 74, 164, 176, 184 pumilio 166, 171 INDICES 497 Artibeus rosenbergi 166 toltecus 148, 154, 162, 169, 170, 189, 195 ff. toltecus group 169, 170 watsoni 153, 162, 166, 169- 171, 189, 198 ff. Ateles 46, 232, 243 belzebuth 55, 57, 242, 243 paniscus 25, 40, 42, 46, 57, 68 Atelinae 242, 258 Atelocynus 455 ff. microtis 457, 458, 463 Audycoptes greeri 249 lawrencei 249 Audycoptidae 245 ff. Auliscomys boliviensis 487, 489 micropus 83, 436, 489 sublimis 490 Balantiopteryx io 140, 153, 161 Bassaricyon 448 Bassariscus astutus 448 Bathyergidae 108 Bauerus dubiaquercus 152-154, 162 Blarinomys 350, 358, 362-364, 395 breviceps 395 Blastoceros bezoarticus 20, 23, 29, 37, 60, 75,80,81,473 ff. dichotomus 20, 37, 60, 75 Bolomys 350 ff., 358, 363, 364, 369, 390,391, 393, 395 amoenus 351-354, 356, 393, 395 bonapartei 356, 390, 393, 395 lactens 351, 354-356,395 lasiurus 354-357, 382, 395 lenguarum 355-357, 395 obscurus 352-357, 395 sp. 357, 391 temchuki 356, 357, 395 Borhyaenidae 467 Borhyaenoidea 107 Brachyteles arachnoides 23, 31, 36, 57 Bradypus torquatus 20, 29, 37, 70 Bradypus tridactylus 40, 45 variegatus 19, 23, 25, 29, 45, 70 Burramyidae 1 1 4 Cabassous tatouay 63 unicinctus 37, 45, 46, 70 Cacajao 54, 232, 238 melanocephalus 30, 31, 55- 57, 238, 239 Caenolestes 112, 122 Caenolestidae 104, 112 Callicebinae 235, 236 Callicebus 232, 236 amictus 23 cenerascens 3 1 cupreus 3 1 donacophilus 73, 74 moloch 25, 57, 236, 246 personatus 23, 30, 31, 36, 57, 68, 236 torquatus 3\, 55, 57, 68, 235, 236 Callimiconidae 4 Callithrix 256 argentata 23, 25, 57 humeralifer 23, 25, 57 jacchus 23, 25, 26, 29, 31, 36, 42, 57, 63, 248 Callitrichidae 4, 57, 258 Calomys 350, 381 callosus 64 laucha 63, 80 lepidus 489 Caluromys 1 1 7 ff. derbianus 1 1 9 lanatus 60, 66, 69, 119 philander 29, 44 Caluromysiops 1 1 7 ff. irrupt a 1 1 7 ff. Canidae 62, 455 ff. Canis 455 ff. armbrusteri 464 davisi 458, 460, 470 dirus 458, 463-465 familiaris 18 gezi 458, 463, 464 lupus 464 nehringi 458, 463, 464 Capreolus capreolus 480 Capromys 17 Carollia brevicauda 36, 153, 162, 183 castanea 176, 183 perspicillata 32, 36, 77, 153, 162, 176, 183,224 CaroUiinae 178, 183 Cavia 20, 312 aperea 23, 25, 29, 37, 62 porcellus 20, 23, 50, 69, 80 tschudii 490 Cebalges 255 ff. ^awf// 248 Cebalginae 245 ff. Cebalgoides 255 ff. cebi 247, 254 Cebidae4, 57,231 ff., 258 Cebinae 241 Cebuella pygmaea 3 1 Cebus 232, 238, 240, 256, 258 albifrons3\, 55, 68, 238, 241, 246, 247 apella 18, 21, 23, 29, 31, 36, 40,41,46,57,63,68,74, 238,240,241,246,247, 249, 252 capucinus 55, 57, 68, 248 flavus 23 nigrivittatus 18, 46, 57, 238, 240, 241 Centronycteris maximilliani 36, 1 6 1 Centurio senex 149, 153, 154, 162 Cephalomyidae 108 Cercopithecidae 253 Cerdocyon 455 ff. avius 458, 462 ensenadensis 458, 462 sp. 458 thous 457 ff. Cervidae 60, 473 ff. Cervinae473, 480,481 Cervus elaphus 480 nippon 480 Chaetomys subspinosus 29, 37 Chaetophractus 64 nationi 488 sp. 20 villosus 45, 63, 80 Chalcomys 350, 358 Chelemys 350, 362-364, 394 delfini 363, 489, 494 498 FIELDIANA: ZOOLOGY Chelemys macronyx 351, 359, 363, 489 megalonyx 363, 489 Chinchilla brevicandata 69, 490 lanigera 490 Chinchillidae 407 Chinchillula sahamae 490 Chiroderma 147, 225 trinitatum 184 villosum 148, 162, 184 Chironectes 19 minimus 44 Chiropotes 54, 68, 232, 238 5a/ana5 23, 25, 3 1 , 40, 46, 54- 57, 238, 239 Chiroptera 137 ff., 173 ff., 213 ff. Choeroniscus intermedins 183 minor 183 Choloepus didactylus 40 Cholomys pearsoni 311, 383 Chroeomys 350, 351, 358, 359, 393 Chrotopterus auritus 146, 153, 161, 182 Chrysocyon 455 ff. brachyurus 25, 29, 37, 62, 74, 457, 458, 463 sp. 458 Coendou bicolor 69 insidiosus 29, 37, 63 prehensilis 18, 23, 25, 42, 50 sp. 19 Conepatus 20, 64 chinga 18,29,55,62,65,69, 73, 74, 80, 89, 488 humboldtii 488 rex 488 Cricetidae 3, 283, 347 ff., 402 Ctenomyidae 104, 105, 108, 109, 409 Ctenomys 63, 64, 109,312,381, 409 boliviensis 75 brasiliensis 8 1 fulvus 490 haigii 436 magetlanicus 75, 83, 85, 490 maulinus 405, 490 Ctenomys opimus 490 robustus 490 sp. 405 Cyclopes didactylus 25, 38, 43, 45, 70 Cynopterus sphinx 189 Dactylomyinae 312 Dactylomys 3 1 2 ZJawa 480 Dankomys 351, 363, 389, 390, 395 simpsoni 39 1 sp. 391 Daptomys 486 oyapocki 486 peruviensis 486 venezuelae 486 Dasypodidae 63 Dasyprocta 3 1 2 azarae 62, 75 fuliginosa 25 /epor/na 21, 23, 25, 29, 37, 40, 42, 43, 50, 69 sp. 18 variegata 66, 69 Dasyproctidae 407 Dasypus 19 hybridus 63, 80 novemcinctus 18, 19, 23, 25, 29, 37, 42, 45, 46, 63, 70 sabanicola 46 septemcinctus 23 Deltamys 350, 358, 360, 369, 380, 395 Desmodillus 442, 444 auricularis 443 Desmodontidae 185 Desmodontinae 178 Desmodus 84 rotundus 33, 36, 43, 64, 74, 85, 147, 149, 153, 162, 175, 176, 185,488 youngi 185 Diclidurinae 140 Diclidurus albus 36, 140 Virgo 140, 161 Didelphidae 60, 1 1 2, 1 1 7 ff., 1 25 Didelphis 18 albiventris 20, 21, 23, 60, 80, 129 Didelphis marsupialis 18, 19, 25, 29, 36, 39,40,43,44,69, 125 ff virginiana 129, 130 Dinomyidae 407 Diphylla 32 ecaudata 32, 149, 154, 162, 185 Diplomys 3 1 2 Dipodidae 442 Dipodomys 442, 444 a^///5 443, 444 merriami 443, 444 Dolichotis 20 patagonum 62, 75, 81, 83 Dromiciops australis 1 1 1 ff., 488 Dusicyon 18, 20, 455 If. australis 82, 83, 456 ff avi« 458, 461,467 culpaeus 65, 82, 85, 488 /M/v/pe5 85, 86, 488 ^n5eM5 65, 83, 85, 488 gymnocercus 62, 74, 80 thous 19, 29, 37, 40, 48, 69 Echimyidae 19, 42, 305, 407, 409 Echimys 39, 312 armatus 50 chrysurus 50 sp. 29 Ectophylla alba 189 macconnelli (see also Meso- phylla) 184 £'/ra barbara 25, 29, 37, 38, 40, 47, 55, 62, 68, 89, 448, 452, 453 Eligmodontia 350 puerulus 442, 443, 489 typus 75, 80, 433 ff., 489 Emballonuridae 139, 178, 182 Emballonurinae 139 Enchisthenes hart a (see also Artibeus) 184 Eocardiidae 407 Epidolopinae 103, 105 Eptesicus andinus 185 brasiliensis 32, 185, 213 ff. furinalis 74, 139, 150, 153, 162, 185 fuscus 224 INDICES 499 Eptesicus innoxius 68, 185 lynniin ff. Erethizontidae 407 Eucricetodontinae 392 Eumops auripendulns 68, 139, 152, 162, 186 bonariensis 152, 162 glaucinus 162, 186 perotis 36, 1 86 underwoodi 153, 162 Eumysops 379 Euneomys 283 ff. chinchilloides 83, 283 ff., 490 mordax 285, 287 noei 285 sp. 436 Euphractus sexcinctus 25, 29, 37, 46, 63, 488 Euryzygomatomys spinosus 62 FeUdae 60, 468 Felis catus 126, 127 colocolo 60, 65, 73, 75, 80, 83, 86, 488 concolor 19, 20, 23, 25, 29, 37,43,49,60,65,69,70, 75, 80, 83, 85, 468, 488 geoffroyi 25, 60, 75, 487 guigna 65, 488 jacobita 488 onca 18-21, 23, 25, 29, 37, 43,49,60,69,70,75,80, 468 pardalis 1 8, 25, 29, 37, 43, 49, 60,69 sp. 60 tigrina 23, 29, 49 wiedii 18, 29, 33, 37, 49, 69 yagouaroundi 19, 29, 37, 49, 69 Fonsecalges 255 ff johnjadini 248, 253, 255 saimirii 248, 255 Furipteridae 177, 178, 185 Galea flavidens 75 mmteloides 490 Galenomys 486 garleppi 490 Galictis cuja 62, 65, 74, 80, 448 ff, 488 vittata 25, 43, 48, 447 ff. Geomyidae 108 Geoxus 350, 362-364, 393, 394 michaelseni 363, 494 mldivianus 35 1 , 359, 363, 489 Gerbillinae 442 Gerbillurus 442, 444 paeba 443 Glossophaga commissarisi 139, 147, 161 soricina 32, 36, 68, 153, 161, 176, 179, 183, 213 ff Glossophaginae 147, 178, 183, 224, 225 Graomys griseoflavus 83 Grisonella 448, 451, 452, 454 Groeberia 105, 108 minoprioi 104, 105 Groeberiidae 100 ff Groeberioidea 100, 107, 108 Heterocephalus glaber 108 Heteromyidae 290, 442 Heteromyinae 290 Heteromys 289 ff. anomalus 296 ff desmarestianus 289 ff , gaumeri 289 ff goldmani 299 Hippocamelus 48 1 antisensis 70, 75, 76, 489 bisulcus 21, 65, 83, 85, 473 ff., 489 Histiotiis macrotis 65, 68, 185, 488, 494 montanusSl, 85, 185, 488 velatus 64, 74 Holochilus brasiliensis 80 Hydrochaeris hydrochaeris 23, 37, 43, 50, 62, 69, 80, 88 Hydropotes inermis 480 Hydrurga leptonyx 489 Hylonycteris underwoodi 153 Hypsimys 350, 358-360. 393 Inia geoffrensis 23, 25, 29, 51, 75, 76 Irenomys tarsalis 490 Ischyromyidae 105 Isolobodon 16 Isothrix 3 1 2 Jaculus jaculus 438 Juscelinomys 351, 361, 363, 395, 486 candango 36 1 Kannabateomys 3 1 2 Kerodon rupestris 29, "il Lagenorhynchus cruciger 75 Lagidium 20 penmnum 69 viscacia 69, 490 wolffsohni 490 Lagothrix 243, 256, 258 flavicauda 57, 68, 246 lagothricha 23, 25, 31, 55, 57, 68, 243, 246, 248 Lama glama 70 guanicoe 70, 80, 83, 489 paco5 70 Lasiurus borealis 65, 151, 162, 185,488 a>ieret«64, 185, 488 ega 151, 162 intermedins 152, 162 Lenoxus 350, 363, 394 apicalis 36 1 Leontopithecus rosalia 21, 23, 32, 33, 36, 57, 68 Leporidae 62 Leptonychotes weddelli 489 Leptonycteris sanborni 2 1 3 ff. 500 HELDIANA: ZOOLOGY Lepus capensis 436, 441 magellaniciis 82 Lestoros 1 1 2 Lichonycteris obscura 153, 183 Liomys 297, 299 pictus 299 sped abi lis 299 Lionycteris spurrelli 183 Lissodelphis peroni 75, 83 Lonchophylla handleyi 134, 183 hesperia 183 roZ?M5ra 134, 183 thomasi 183 Lonchorhina aurita 143, 161, 182 Lw/ra 18, 19,48,493 annectens 19 enudris 42, 48 /e//>ifl 65, 69, 83, 85, 488 montana 69 pi at ens is 74, 80 provocax 488 Lutreolina crassicaudata 44, 48, 60, 80 Lycalopex 456, 457, 460 patagonicus 74, 488 Macrophyllum macrophyllum 36, 161, 182 MacrojKxiidae 100, 105 Makalata 3 1 2 Marmosa 19, 64, 486 agricolai 486 andersoni 486 cinerea 36 cracens 486 elegansSS, 113,419,488 handleyi 486 impavida 69 murina 36, 44, 69 noctivaga 69 pusilla 60 scapulata 486 sp. 39, 40 /a/e/ 486 pennanti 448 Mazama 18, 20, 21, 473 ff. americana 19, 25, 29, 37, 38, 40, 42, 50, 60, 70, 474 ff. gouazoubira 18, 29, 37, 50, 60, 70, 75 rw^wfl 19 sp. 20, 50 Megaderma 225 /j^ra 215 spasma 215 Mesomys 3 1 2 Mesophylla macconnelli (see also Ec/o- pM/a) 189 Metachirus nudicaudatus 69 Microbiotheriidae 112, 122 Af/crocav/a australis 75, 83, 436, 490, 494 Micronycteris brachyotis 141, 161 daviesi 182 hirsuta 182 megalotis 142, 153, 154, 161, 182 minuta 182 mce/on 142, 153, 161, 182 schmidtorum 142, 153, 161 Microryzomys 267 Microtragulidae 107 Microtus 404 Microxus 350, 351, 358, 360, 362, 363, 380, 393, 394 bogotensis 360, 394 mimus 357, 360 bennettii 144 cozumelae 144, 153, 154 crenulatum \ 39, 144, 161, 182 koepckeae 182 Mirounga leonina 75, 489 Molossidae 152, 178, 186, 215, 224 Molossops abrasus 135, 186 temminckii 186 Afo/055M5 a/er 32, 64, 68, 153, 162, 186 crassicaudatus 64, 74 mo/o55M5 44, 64, 68, 153, 162, 186, 213 ff. sinaloae 162 Monodelphis 87 americana 23 brevicaudis 60 dimidiata 80, 86, 114 Monodon monoceros 40 Mormoopidae 141, 177, 178, 182,215,224 Mormoops megalophylla 141, 154, 161, 182 Mormopterus kalinowskii 186, 488 Muntiacus muntjak 48 1 A/m5 domesticus 444 AfM5/e/a frenata 18-20, 69 Mustelidae 62, 447, 468 A/>cefe5 rufimanus 68 A/3;o<:a5for acouchy 50 co>'pi« 29, 62, 65, 69, 83, 85, 490 ex/Vw 25, 38, 40, 42, 43 Af3^of/5 albescens 36, 64, 74, 185 atacamensis 488 chiloensis 85, 488 elegans 150, 162 A:eay5/ 162, 185 lucifugus 223 nigricans 36, 74, 176, 185 oxyorM5 185 riparius 185 rwfter 64, 74 simus 185 Myrmecophaga 19 tridactyla 19, 23, 29, 36,43, 44,60 Myrmecophagidae 60 A^o^wa 19, 448 narica 448 «a5wa 19, 21, 23, 25, 29, 37, 43, 47, 62, 66, 68, 74, 448^50 Nasuella olivacea 447 ff. Natalidae 150 Natalus stramineus 150, 153, 162 INDICES 501 Necrolestidae 106, 107 Necrolestoidea 107 Necromys conifer 38 1 Nectomys squamipes 25, 40, 381 Neotomys ebriosus 490 Nesophontes 17 Noctilio albiventris 32, 68, 74, 1 33, 1 82 leporinus 32, 36, 64, 68, 74, 140, 153, 161, 182 Noctilionidae 140, 178, 182 Notiomys 350, 362-364, 380, 393, 394 edwardsi 362, 436 Nyctereutes 460, 462, 466, 47 1 Nyctinomops laticaudatus 153, 162, 186 macrotis 186 Octodon 402, 407, 487 bridgesi 402, 403, 490 degus 65, 69, 75, 85, 402, 403, 413 ff., 490 lunatus 402, 403, 490 Octodontidae 108, 401 ff., 407, 415 Octodontoidea 108 Octodontomys 404 gliroides 75, 404, 405, 490 Octomys 404 barrerae 404 mimax 404 Odocoileinae 479-48 1 Odocoileus 18, 473 ff. hemionus 476, 477, 480 virginianus 18, 19, 25, 42, 51, 474 ff. Oligoryzomys 26 1 ff. Onychomys torridus 443 Oryzomyini 364, 392 Oryzomysiex ff., 487 angouya 62 chacoensis 26 1 ff. chaparensis 262, 267, 274 delicatus Idl ff. destructor 264, 271, 273 flavescens 80, 262 ff. fornesi 262 ff. galapagoensis 84, 86 longicaudatus 69, 82, 83, 85, 261 ff., 419, 489 Oryzomys mattogrossae 266, 267, 273 megacephalus 62 melanostoma 69 microtis 26 1 ff. nigripes 63, 262 ff. stolzmanni 262, 264, 27 1 utiaritensis 266, 273, 274 Otaria byronia 489 flavescens 20, 69, 75 Otocyon 458, 460, 470 Oxymycterus 350, 351, 356, 358, 360 ff., 393-395 akodontius 36 1 angularis 36 1 delator 361 hispidus 361 iheringi 361 />ica^ 36 1 nasutus 36 1 paramensis 36 1 , 394 platensis 36 1 roberti 36 1 rw/us 37, 62, 80, 357, 361 Palaeothentinae 104 Parabderites bicrispatus 104 Parabderitini 104 Paramyidae Patagonia 99 ff. peregrina 99 ff. Patagoniidae 99 ff. Patagonioidea 99 ff. Peramelidae 113, 114 Perognathus 434, 444 /a//ax 443 flavus 443 longimembris 443, 444 Peromyscus 442 eremicus 443, 444 maniculatus 434, 443, 444 yucat aniens 297 Peropteryx kappleri 161 macrotis 36, 161, 182 Petauridae 1 1 3 Philander opossum 36, 39, 40, 43, 44, 69 Phylloderma stenops 161 Phyllostomidae 141, 178, 189, 213,215,223,224 Phyllostominae 141, 178, 182, 224, 225 Phyllostomus discolor 68, 145, 153, 161, 179, 182 elongatus6S, 182, 213 ff. hastatus 33, 36, 68, 88, 182 latifolius 2\3 ff. obscurus 36 stenops 145, 161, 182 Phyllotini 364 Phyllotis darwini 69, 85, 86, 413 ff., 436, 443, 444, 489 gerbi litis 442 magister 489 osgoodi 489 xanthopygus 83, 489 Pithanotomys 375 Pithecia 46, 232, 240, 243, 256, 258 hirsuta 246, 249 monachus 23, 25, 26, 31, 57, 249 pithecia 3 1 , 40, 42, 46, 57, 88, 239, 243 Pithecinae 239 Plagiodontia 16 Platalina genovensium 183 Platyceros dama 476, 477, 480 Platyrrhini 63, 258 Pliolestes 377 Podoxymys 350, 361, 363, 393, 394, 486 roraimae 361 Polydolopidae 103-105 Pontoporia blainvillei 75 Potoroinae 105 y7avM5 18, 25, 37, 47, 55, 68, 74 Prepidolopidae 103 Primates 245 ff. Priodontes giganteus 45, 46 maximus 29, 37, 42, 63 Proargyrolagus bolivianus 106 Procebalges 255, 256, 258 pitheciae 249 302 HELDIANA: ZOOLOGY Procyon cancrivoms 25, 29, 37, 40, 47, 62, 74, 448^50 Procyonidae 62, 447, 468 Proechimys 39, 69, 305 ff. amphichoricus 308 arabupu 308, 313, 325, 330 arescens 308 biomensis 308 boliviensis 309 brevicauda 309, 322, 330, 333, 338 burrus 3 1 1 calidior 3 1 1 canicollis 311, 326, 337 canicollis-group 305 fF., 311 fr., 315-317, 326-329, 331, 334 ff., 337, 340 centralis 3 1 1 cherriei 307 chiriquinm 3 1 1 chrysaeolus 309, 317, 326, 337, 339, 344 colombianus 3 1 1 CMv/m 309, 322, 327, 333, 339, 340 ff. CMv/>r/-group 305 ff., 309, 311, 314, 317 ff., 322, 323, 328, 329, 331, 333 ff, 339, 340 ff. decumanus 312, 326, 337 decumanus-group 305 ff., 308, 312, 315 ff., 326 ff., 328, 329, 331, 334 ff, 337, 341 elassopus 309, 338 goeldii 308, 338 goeldii-gToup 305 ff., 308 ff., 314-316, 323, 324, 328, 329, 331, 332, 334 ff., 338, 341 ff. goldmani 3 1 1 gorgonae 3 1 1 guairae 310, 317, 326, 337, 339, 344 gularis 309, 338 guyannensis 307, 330, 336 guyannensis-group 305 ff., 307 ff, 312 ff., 324 ff., 328, 329, 331, 334 ff., 338, 339, 342 hendeei 309, 338 hilda 308 hoplomyoides 310, 317, 339, 344 hyleae 308 Proechimys ignotus 3 1 1 kermiti 308 leioprimna 308 leucomystax 309 liminalis 308 longicaudatus 64, 309, 333, 338 longicandatus-group 305 ff., 309, 310, 314, 316 ff., 322,323, 328,329, 331, 333 ff., 338, 342 ff magdalenae 310, 317, 339, 345 mincae 309, 326, 337, 339, 345 myosuros 37 nesiotes 308 nigrofulvus 309, 338 ochraceouslXQ, 326, 339, 345 oco«ne///311,320, 329, 339, 344 oris 307, 325, 338 pachita 308 panamensis 311 -^ poliopus 310, 339, 345 quadruplicatus 308, 327, 329 rattinus 308 ribeiroi 309 riparum 308 roZ?em- 307, 3 1 3, 325, 336, 338 ro5a 3 1 1 rubellus 3 1 1 securus 309 semispinosus 311, 320, 324, 327, 332, 338, 339, 343 ff semispinosus-group 305 ff., 310ff, 314, 319ff., 323, 324,328, 329,331, 332, 334 ff., 338 ff., 343 ff. simonsi 309, 325, 330, 336, 338 5/moAJ5/-group 305 ff., 309, 313ff., 314, 323ff., 328, 329, 331, 334 ff., 337 ff, 344 sp. 29 steerei 308, 338 trinitatus 309, 317, 339, 345 trinitatus-gronp 305 ff., 309 ff., 314, 315, 317, 325 ff., 328, 329, 331, 334 ff., 337, 339, 344 ff. Proechimys urichi 309, 339, 345 vacillator 307 villacauda 309 warreni 307 Promops centralis 186 nasutus 32, 68 Protocyon 455 ff. orcesi 458, 462 scagliarum 457, 458, 462 troglodytes 458, 462, 463 Pseudalopex 455 ff. culpaeus 457 ff. griseus 457, 458, 460, 468 gymnocercus 457, 458, 460, 461 penianus 458, 460, 461 sechurae 457, 458, 460, 461 vetulus 457, 458, 460, 461 Psoroptidae 245 ff. Pteronotus davyi 139, 141, 154, 161, 182 gymnonotus 182 parnellii 153, 161, 182, 213 ff. personatus 141, 161, 182 Pteronura 48 brasiliensis 23, 25, 29, 37, 48, 62 Pteropodidae 189 Pudu 473 ff, 493 puda 65, 474, 476, 480, 489 Rattus rattus 126, 127 Reithrodon 82 auritus 62, 381,436,441 chinchilloides 85 physodes 80, 83, 490 Rhinophylla alethina 134 fischeri 183 pumilio 183 Rhipidomys leucodactylus 69 Rhogeessa tumida 153, 154, 162 Rhyncholestes raphanurus 111 ff., 488 R hynchonycteris naso 32, 36, 140, 153, 154, 161, 182 INDICES 503 Saccopteryx bilineata 140, 153, 161, 182 leptura 139, 161, 182 Saguinus 256 bicolor 3 1 fuscicollis 31, 57, 68 labiatus 23, 25, 26, 57 midas 25, 38, 40, 42, 46, 47, 57, 68, 89 mystax 3 1 , 68 nigricollis3l,6S, 248 oedipus 31, 55-57, 247 Saimiri 232, 256, 258 boliviensiseS, 73, 74, 232, 249 oerstedi 201, 232 sciureus 40-42, 46, 57, 232, 233, 246, 248, 249 ustus 23, 25, 26, 232 Saimirinae 233 Saimirioptes hershkovitzi 249 fF. paradoxus 249, 252 Scapteromys sp. 383 tumidus 80 Schizopodalges 256-258 lagothricola 248, 255 Sciurillus pusillus 39 Sciurus aestuans 18, 23, 29, 37, 39, 42, 50, 69 granatensis 55, 56 igniventris 25 pyrrhinus 69 spadiceus 25, 69, 75 stramineus 69 Scolomys 486 Sigmodontinae 347 ff., 413 Solenodon 17 Sotalia 54 fluviatilis 25 Spalacopus 405, 409 cyanus 65, 85, 402 ff., 490 Sparassocynus 379 Speo//io5 21,455 ff. pacivorus 458, 463 venaticus 25, 40, 458, 463 Stenodermatinae 147, 178, 184, 189, 213,223,225 Strepsirrhini 253 Sturnira 147 aratathomasi 134 fe/V/e/w 183 Sturnira bogotensis 177 erythromos 6S, 183 lilium 64, 153, 154, 162, 183, 213 ff ludovici 183 magna 183 /7a/7a 183 oporophilum 68 f/Wae 183 Stumirinae 178, 183 scrofa 42 5>'/v/7a^5 brasiliensis 18, 19, 21, 23, 29, 37, 62, 70 Jloridanus 19 sp. 18 Tadarida brasiliensis 68, 74, 80, 85, 186, 213 ff., 488 laticaudata 64 Tamandua 19, 20 tetradactyla 18, 20, 23, 25, 29, 36,43,45,60, 61, 70 Tapirus pinchaque 19, 70 terrestris 19-21, 23, 29, 37, 40, 41, 43, 50, 59, 70 villosus 7 1 Taterillus 442 pygargus 443 Tayassu 59 pecan 19, 37, 38, 42, 50, 59, 70 tajacu 19, 23, 25, 37, 42, 50, 59, 70 Thalpomys 350, 356, 358 Thaptomys 350, 351, 358 Theriodictis 455 ff. p/a/e«5/5 458, 461,462 tarijensis 458, 461 Thrinacodus 3 1 2 Thylatheridium 379 Thyroptera 32 tricolor 31, 162, 185 Thyropteridae 177, 178, 185 Tolypeutes 64 matacus 63, 80 tricinctus 23, 29 Tomopeas 185 Tonatia bidens 32, 44, 143, 161, 182, 213 ff. brasiliense 144 carrikeri 182 evof/5l44, 148, 153, 154, 161 w/rtM/fl 139, 144, 154, 161 nicaraguae 144 sylvicola 74, 144, 182, 213 ff. Trachops cirrhosus 32, 139, 145, 153, 161, 182, 213 ff. Tremarctos ornatus 19, 20, 68, 74 Trichechus inunguis 25, 29, 51, 55 manatus 18, 37, 41-43, 54 Tylomyinae 392 Urocyon 455 ff. cinereoargenteus 457-459 progressus 458, 459 Uroderma 147, 187 ff. bilobatum 147, 153, 154, 162, 184, 189, 202 ff. magnirostrum 184, 189, 207, 208 Vampyressa 147 bidens 184 melissa 184 /7M5///a 139, 148, 153, 154, 162, 184, 189 Vampyrodes caraccioli 148, 154, 162, 184 Vampyrops 147 brachycephalus 184 £/or5a//5 176, 184 /!c//er/ 147, 184 infuscus 184 lineatus 64, 1 84 vittatus 184 spectrum 43, 146, 153, 161, 182 Vespertilionidae 150, 178, 185, 215, 224 Vespertilioninae 150 Vicugna vicugna 65, 66, 70, 489 Vizcacia vizcacia 62, 65, 81 S04 HELDIANA: ZOOLOGY Vombatus ur sinus 1 14 Vulpes 457, 458, 460, 470 Wiedomys pyrrhorhinos 29, 33, 37 Xylomys 290 Zaedyus pichiy 63, 80, 488 Zygodontomys 350, 351, 354 brevicauda 354 microtinus 355 thomasi 355 Subject Index Acinar cells 2 1 3 ff. Age variation 129, 131, 266, 293,312,448 Bacular morphology 3 1 2 IF., 447 ff. Behavior 187 ff., 402 ff., 442 Biogeographic history 108, 392, 407 Body size 485 ff. C-band analysis 361, 477 Censusing413 ff., 440 Chilean mammals 485 ff. Chronicles, Neotropical Region 14-21 Coevolutionary patterns 245 ff. Cohort definition 420 Collecting bias 176, 485 ff. Cranial morphology 121, 165, 167ff., 267ff., 320 ff., 347 ff, 457 ff Demography 413 ff., 433 ff. Dentition 1 00 ff., 1 2 1 , 1 65, 1 68, 267 ff., 330 ff., 347 ff., 402, 454, 457 ff. Desert rodents 443 Diet 125 ff., 175, 213 ff., 362, 364, 402 ff., 427, 441 Digestive system, morphology 347 ff. Discovery, Neotropical Region 14-21 Distribution, geographic 133 ff.. 137 ff., 163 ff., 173 ff., 191 ff., 231 ff., 261 ff., 274, 283 ff., 292, 305ff., 347ff.,401 ff, 455 ff. Ectoparasites 164, 165, 245 ff., 359 Endangered status 485 ff. Evolutionary diversification 99 ff., 347 ff, 455 ff. Exploration and description. Neotropical Region 21 ff. External features 1 17, 164, 165, 167 ff., 267 ff., 457 ff. Extinction 468 Family nov. 99 ff. Faunal origins, dispersal 3, 87 ff., 347 ff., 402, 464 Faunal representation 485 ff. Foraging habits 137 ff. Fossil record 457 G-band analysis 356, 361, 477 Genital morphology 122, 350 Genus nov. 99 ff. Geographic variation 1 64, 2 1 6, 287, 289 ff. Glands, male accessory 358, 364 Habitat 1 33 ff., 137 ff., 165, 167, 189, 402 ff. Hershkovitz, Philip 1 ff. biographical sketch 1 ff. bibliography 4 ff. Immigration 425 Incisive foramen 321 ff. Infraorbital foramen 328 Karyotypic analysis 1 1 1 ff., 296, 347 ff., 473 ff. Karyotypic evolution 473 ff. Key to identification 1 70 Life history 433 ff. Life zone 485 ff. Mandibular morphology 100 ff., 347 ff. Mesopterygoid fossa 328 ff. Metachromism 3 Morphological adaptation 108, 362, 434 Morphological variation 293 Neotropical mammalogy, his- tory 1 1 ff. Nongeographic variation 216, 266, 293 NOR analysis 477 Osgood, W. H. 486 ff. Phylogenetic relationships 245 ff, 457 Physiological adaptation 434 Population regulation 4 1 3 ff. Population survivorship 423 Postcranial morphology 122, 448, 457 ff. INDICES 505 Precipitation 175, 426, 442 Predation 201 Recruitment 423 Reproduction 122, 173 ff., 419, 433 ff. Roosting behavior 187 ff. Salivary glands 2 1 3 ff. Secretory granules 2 1 3 ff Sex-chromosome mosaicism 211 ff Sex ratio 419, 438 Sexual variation 128, 216, 266, 293, 438 Species nov. 99 ff, 1 64, 1 67, 249 ff , 377, 379, 383 Steppe community 433 ff Superfamily nov. 99 ff Systematic relationships 105, 122, 163 ff, 191 ff, 213ff, 252, 261 ff, 283 ff, 297, 305ff, 347fr., 401 ff, 451, 454, 455 ff Temporal ridge 327 ff Tent construction 187 ff. Thorn scrub community 4 1 3 ff Trophic habits 108, 173 ff ,467, 485 ff. Ultrastructure analysis, com- parative 213 ff 506 FIELDIANA: ZOOLOGY Other Fieldiana: Zoology Publications in Mammalogy The Bals ol' Iran: Sv»;tpmatics. Distrihulicn. Frolom !v. V'M^i -i, i DoRlaso \^) illus., 22 tables. riiDUcation khi/, 5>^il.uu 1980. 579 pages, 164 illus.. 58 tables. Publication 13U9, $31.00 A Multivariate Study ol the l-amil> Molossidac (.Mammalia, Chiroplcra): Morphology, Ecology, Evo- lution. B\ Patricia Waring Freeman. 1981. 173 pages, 25 illus., 12 tables. Publication 1316, $13.25 Taxonomy and Evolution of the Sinica Group of Macaques: 2. Species and Subspecies Accounts of the |ndi;in RnnriiM \Ta(;i(iiii^ Xfin-nrn rnil'inlii R\' \:\c\ FoodiMi I OS 1 S"* n;K'f»eer (Cervidae). Part I. Pudus, Genus Pudu Gray. By Philip Hershkovitz. 1982. 86 pages, 37 illus.. 10 tables. Publication 1330, $11.25 The Effectiveness of Methods of Shape Analysis. By Cliff A. Lemen. 1983. 17 pages, 5 illus., 6 tables. Publication 1343, $3.00 laxonomy and Evolution of the Sinica Group of Macaques: 4. Species Account of Macaca thibetana. W\ \acV Foodcn IQS"? ""O n;iPrs '^ illns 1 !;ihlcs. Publication 1345, $3.50 On the Phyletic Weight of Mensural Cranial Characters in Chipmunks and Their Allies (Rodentia: Sciuridae). By Bn - '"^ "attcrson. 1983. 24 pages. 4 illus.. 7 tables. Publication 1348, $3.50 Annotated Checklist of Bird and Mammal Species of Cocha Cashu Biological Station, Manu National Park, Peru. By John W. Terborgh, John W. Fitzpatrick, and Louise Emmons. 1984. 29 pages, 2 illus.. 3 tables. Publication 1352, $3.75 Systematics of Mice of the Subgenus Akodon (Rodentia: Cricetidae) in Southern South America, with the Description of a New Specie*- R^ u^,.,,^ i^ p-.it, MtrMi vTiiK.n m r;-.iiw,i,A ->nH w.ihx r Ki-,.nc 1984. 16 pages, 6 illus.. 1 tabic Publication 1355, $4.00 A Preliminary laxonomic Review oi mc >ouih American Bearded Saki Monkeys Genus Chiwpotes (Cebidac, Platyrrhini), with the Description of a New Subspecies. By Philip Hershkovitz. 1985. 46 pages, 14 illus., 9 tables. Publication 1363, $7.00 I : momy and Evolution of the Sinica Group of Macaques: >f Natural Hislor\. Rv Jack I ooden. 1986. 22 pages, 2 illus., 4 tables. Publitaiion i, ' ' Order by publication number anc .,. „ ,., ,^. „ ..^. . ,^-.. , ,...^. ,,.^,,,. . Illinois residents add current lax. All foreign orders are payable in U.S. dollar-checks drawn on any I'.S. i^ank ■ ' ibsidiary of any foreign bank. Address all requests to: FIELD MUSEUM OF NATURAL HISTORY Library' - Publications Division Roosevelt Road at Lake Shore Drive Chicago, Illinois 60605-249X Field Museum of Natural History Rooscvcll Road at Lake Shore Drive Chicago, Illinois 60605-2496 Telephone: (312) 922-9410