HARVARD UNIVERSITY Ernst Mayr Library of the Museum of Comparative Zoology LIBRARY DEC 4 2006 .HT- LIBRARY CONTRIBUTIONS IN MARINE MAMMAL PALEONTOLOGY HONORING FRANK C. WHTTMORE, JR. Edited by Annalisa Berta and Thomas A. Demere Incorporating the Proceedings of the Marine Mammal Symposium of the Society of Vertebrate Paleontology 5 1 st Annual Meeting Held at the San Diego Natural History Museum San Diego, California 26 October 1991 f No. 29 1 May 1994 Proceedings of the San Diego Society of Natural History ISSN 1059-8707 MCZ LIBRARY MAY 1 2 1994 HARVARD UNIVERSITY CONTRIBUTIONS IN MARINE MAMMAL PALEONTOLOGY HONORING FRANK C. WHITMORE, JR. Edited by Annalisa Berta and Thomas A. Demere Incorporating the Proceedings of the Marine Mammal Symposium of the Society of Vertebrate Paleontology 51st Annual Meeting Held at the San Diego Natural History Museum San Diego, California 26 October 1991 f No. 29 1 May 1994 Proceedings of the San Diego Society of Natural History ISSN 1059-8707 Contents Preface Annalisa Berta and Thomas A. Demote 1 Tribute to Frank Clifford Whitmore, Jr. Ralph E. Eshelman and Lauck W. Ward 3 I. Marine Mammals: Evolution and Systematics The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematics and Mode of Life Richard H. Tedford, Lawrence G. Barnes, and Clayton E. Ray 1 1 Pinniped Phylogeny Annalisa Berta and Andre R. Wyss 33 Basicranial Evidence for Ursid Affinity of the Oldest Pinnipeds Robert M. Hunt, Jr. and Lawrence G. Barnes 57 The Evolution of Body Size in Phocids: Some Ontogenetic and Phylogenetic Observations Andre R. Wyss 69 Two New Species of Fossil Walruses (Pinnipedia: Odobenidae) from the Upper Pliocene San Diego Formation. California Thomas A. Demere 77 The Family Odobenidae: A Phylogenetic Analysis of Fossil and Living Taxa Thomas A. Demere 99 Phylogenetic Relationships of Platanistoid River Dolphins: Assessing the Significance of Fossil Taxa Sharon L. Messenger 125 Are the Squalodonts Related to the Platanistoids? Christian de Muizon 135 Waipatia maerewhenua, New Genus and New Species (Waipatiidae, New Family), An Archaic Late Oligocene Dolphin (Cetacea: Odontoceti: Platanistoidea) from New Zealand R. Ewan Fordyce 147 A Phylogenetic Analysis of the Sirenia Daryl R Damning 177 A New Middle Miocene Sirenian of the Genus Meta.xytheriam from Baja California and California: Relationships and Paleobiogeographic Implications Francisco J. Aranda-Manteca, Daryl P. Domning, and Lawrence G. Barnes 191 A New Specimen of Behemotops proteits (Order Desmostylia) from the Marine Oligocene of Washington Clayton E. Ray, Daryl P. Domning, and Malcolm C. McKenna 205 II. Marine Mammals: Faunas and Biostratigraphy Neogene Climatic Change and the Emergence of the Modern Whale Fauna of the North Atlantic Ocean Frank C Whitmore, Jr. 223 Miocene Cetaceans of the Chesapeake Group Michael D. Gottfried, David J. Bohaska, and Frank C. Whitmore, Jr. 229 Miocene and Pliocene Marine Mammal Faunas from the Bone Valley Formation of Central Florida Gary S. Morgan 239 PROCEEDINGS of the San Diego Society of Natural History Founded 1S74 Number 29 1 May 1994 Contributions in Marine Mammal Paleontology Honoring Frank C. Whitmore, Jr. Preface On 26 October 1991 a symposium on marine mammal evolution was held at the 51st Annual Meeting of the Society of Vertebrate Paleontology in San Diego, California. It was the first symposium on this topic since the one in 1975 held at the American Institute of Biological Sciences meetings in Corvallis, Oregon. The proceedings of that symposium were published as a special issue of Systematic Zoology (1976, volume 25:301^146) edited by Charles A. Repenning. Ten years after the Corvallis symposium three of the original participants, Lawrence G. Barnes, Daryl P. Domning, and Clayton E. Ray, published an article in Marine Mammal Science (1985, volume 1:15-53) that highlighted notable paleontological discoveries in the intervening years. Since that publication nine years ago, new fosstf discoveries, new techniques for investigating evolutionary relationships, and an increase in the number of interested researchers have contributed important new data. The San Diego symposium was organized to bring together many of these researchers and to present a synthesis of our current knowledge of the evolution, systematica, and biogeography of marine mammals (pinnipeds, cetaceans, and sirenians) and marine mammal faunas. The symposium benefited from an international group of participating scientists representing France, Mexico, New Zealand, Russia, and the United States. Thirteen of the current articles are based on papers presented in San Diego; two additional reports were solicited afterward because of their close relationship to the general theme of the symposium. In the volume, we have arranged the articles into two broad groups, one covering the taxonomy, systematics, and comparative anatomy of specific taxa, the other dealing with the biogeography and biostratigraphy of marine mammal faunas. One of the participants at the symposium, Frank C. Whitmore, Jr.. is widely recognized for his contributions to fossil cetacean systematics and biogeography and for his role as a mentor to many marine mammal paleontologists. Accordingly, we have chosen to recognize Frank's past achievements and continuing research by dedicating this volume to him. The editors gratefully acknowledge the following individuals who, in addition to most contributors, provided critical reviews of manuscripts: J. David Archibald. Jon A. Baskin, Mario A. Cozzuol, Francis H. Fay, John J. Flynn, John E. Heyning, Richard Hulbert, Samuel A. McLeod, James G. Mead, Jeheskel Shoshani, Michael J. Novacek, Donald R. Prothero. and Charles A. Repenning. We are especially indebted to Clayton E. Ray for his advice and encouragement. Philip Unitt, managing editor of the San Diego Society of Natural History's scientific publications, provided skillful editing of the volume. For funding of this publication we also express our appreciation to the National Science Foundation, National Geographic Society. San Diego State University College of Sciences, and the Kellogg fund of the Smithsonian Institution. Annalisa Berta and Thomas A. Demere Tribute to Frank Clifford Whitmore, Jr. Ralph E. Eshelman Department of Paleobiology, National Museum of Natural History, Washington, DC. 20560 Lauck W. Ward Virginia Museum of Natural History, Martinsville, Virginia 24112 With this volume, we pay tribute to our mentor and colleague the "good doctor" Frank Clifford Whitmore. Jr., whose gracious humor, interest in people, and curiosity about the world have meant so much to so many within the field of paleontology and far beyond it. We are pleased to have this opportunity to share our respect and admiration for Dr. Whitmore the scientist, the teacher, and the friend. Frank was born at home in Cambridge. Massachusetts, on No- vember 17. 1915. to Marion Gertrude (Mason) and Frank Clifford Whitmore. a graduate student at Harvard University. When Frank Jr. was two, the growing family began a series of moves occasioned by his father's burgeoning career in organic chemistry: to Houston and the Rice Institute, to Minneapolis and the University of Minne- sota, to Evanston and Northwestern University, to Washington. D. C. and the National Research Council. Frank, his two younger brothers, and younger sister grew up in a lively family, which included their Irish grandmother and. as often as not. a variety of students who rented rooms in their house. Ideas, adventures, and experiments were encouraged. From these grew their later devotion to science, literature, medicine, dance, business, theater, and art. Frank completed his intermediate schooling in Evanston before the family relocated, one last time, to State College. Pennsylvania, and Pennsylvania State College. There his life blossomed with friendship, studies, and sports. His three high-school buddies have remained close friends throughout his life. The reporting he did on the high-school paper and Center Daily Times marks his early work as a writer, one of the important skills of his career. In 1 933, Frank enrolled as an English major at Amherst College in Massachusetts. However, Frank eventually changed his major to something completely different. The change came about in an unusual way. "We had to take a science, and somebody said, 'take geology." Sol took geology and I was kind of bored by it and I got aC. But the guy who'd advised me to take it said. 'I know the first course isn't very good, but you ought to take historical geology. That's really interesting." So I signed up for historical geology and nobody else did. This was with F. B. Loomis. . . . Although 1 was the only person who had registered for the course. [Loomis] agreed to give it. So I had a one-year course in what was not historical geology [but] vertebrate evolution, because that was what Loomis was interested in. and we just sat around three days a week for a year talking about vertebrate evolution. By the end of that year, I decided I wanted to be a vertebrate paleontologist" (Cain 1989). Within his first week at Amherst, Frank was rushed by the fraternity Phi Kappa Psi. At the end of an exhausting day, he sank into a couch and remarked, quoting a character from the Barney Google comic strip. "Moosenose hurt in feet!" One of his new "brothers" immediately nicknamed him Moose, which remains his name to many friends and relatives to this day. In January of 1934, in his freshman year, another fateful event occurred. While taking his date on a sleigh ride with a group of other students, the Smith girl seated on his other side got something in her eye. Frank gallantly helped remove it and life for him was never the same afterwards; he dated Martha Burling Kremers of Niagara Falls. New York, until their graduation and engagement. They were married in 1939, after she'd had a year of business school and he had earned his master's degree at Pennsylvania State University. Although Frank's B. S. cum laude with honorable mention in geology from Amherst centered on vertebrate paleontology, his M. S. in paleontology and stratigraphy dealt instead with invertebrates. Frank M. Swartz was the only resident paleontologist at Penn State; his specialty was ostracodes and so Frank studied ostracodes. That year also provided a good background in sedimentology from Paul Dimitri Kryrine. Frank continued his education at Harvard University, studying under one of the world's foremost vertebrate paleontologists, Alfred Sherwood Romer. Frank was Romer's first student from the geol- ogy department; all previous students had been from biology. Romer referred to himself as a zoologist. Frank recalls. "I had quite a battle with Romer to be allowed to take structural geology be- cause he thought that would be a waste of time. I felt it would be a way to get a job. which it turned out to be. ... I remember as a student, I got to know C. B. Schultz pretty early, and Morris Skinner and Lloyd Tanner and Mylan Stout, the Nebraska folks, and think- ing how strange and rather dull that they were always worrying about stratigraphic correlation. They were talking about the Valen- tine problem, the Marsland Formation, and so on. In a way. they certainly were more geological than we easterners were" (Cain 1989). Thanks to his training at Amherst and Penn State. Frank himself proved to be an exception to this generalization. A particular admonition of Romer's has in turn been heard by Frank's own children and young colleagues, probably because it fit so well his own proclivities: "Learn to write while you're a stu- dent— you can always learn geology later." Frank's first real field experience in vertebrate paleontology came during the summer of 1940. The Harvard field crew headed west toward the Unita Basin of Utah to collect fossil mammals in a used pie truck purchased for $85. "I can remember . . . when we heard that France had fallen, sitting around the campfire. all of us wondering where we'd be a year from then" (Cain 1989). While in the field. Frank became the father of twins. Geoffrey Mason and John Kremers. and their birth kept him out of the army long enough to complete his doctorate. To support his family. Frank served as a teaching fellow and university fellow in paleontology. Harvard was the right place and 1940 the right time to witness an exciting event, the founding meeting of the Society of Vertebrate Paleontology. Frank recalled, "of the starving graduate .students, only those on the premises, like me. could afford to attend. And we couldn't exactly go overboard socially: the cocktail party at the founding meeting cost $1.00. My wife Marty and I took stock of our finances and decided I really should go but that she couldn't afford to (we would have had to get a sitter for our 6-month-old twins)" (Whitmore 1989). Frank's doctoral study, suggested by Romer, was the cranial morphology of three Oligocene artiodactyls. Frank wrote. "It is the purpose of this study to examine in detail the cranial anatomy of some of these extinct genera, because endocranial characteristics are probably nonadaptive. that is. unlikely to be influenced by the In A. Berta and T. A. Demere (eds.) Contributions in Marine Mammal Paleontology Honoring Frank C. Whitmore. Jr. Proc. San Diego Soc. Nat. Hist. 29:3-10, 1994 R. E. Eshelman and L. W. Ward Figure 1 . Frank C. Whitmore. Jr.. while professor al Rhode Island State College. Kingston. Rhode Island (now the University of Rhode Island). ca. 1943. environment, and therefore useful in determining the taxonomic position of groups of animals" (Whitmore 1953:117). For this study, a serial sectioning apparatus was designed and built by F. Russell Olsen of Harvard's Museum of Comparative Zoology. This sectioning technique, perfected for paleobotany with cellulose ac- etate peels, was adapted for vertebrates and described in a paper by Olsen and Whitmore (1944). It was a pioneering achievement in its approach and formed the basis of later work by others, in which details of cranial anatomy such as blood circulation, ear morphol- ogy, and brain conformation have been used in phylogenetic studies of fossil mammals. Frank's first postgraduate job was a teaching position at Rhode Island State College (now the University of Rhode Island) from 1 942 to 1 944. "I was the entire geology department. I got to teach a lot of things I'd never studied before, such as specimen petrology and engineering geology, which I made up myself" (Paleo News 1990). When the Army Specialized Training Program came to the college, he also taught economic and political geography. Prepara- tion for these courses aided him most in the next phase of his career. During this time, Frank Sr. asked his son, "If you could talk to any paleontologist you wanted to, whom would you choose?" Frank named W. B. Scott, and his father offered to pay his way to Princeton for a visit. Frank remembers, "With some qualms, I wrote to Scott and asked if I could come and see him. and he said yes. So I got on the train and went down ... to Princeton, and Dr. Scott met me. He showed me around the lab. I remember he was then about 80. . . . He had me to lunch at the faculty club. . . . Then Scott showed me the Eocene mammals in the Princeton collection. I had spent the previous summer in Utah ... so I was interested. At the end of the day Scott asked me if I would like to collaborate with him in the study of Eocene mammals in the Princeton collection" (Cain 1989). Unfortunately, the war intervened, and Frank was unable to take up the opportunity. His observations about Dr. Scott that day, however, presaged Frank's own support of young people in his field. "I probably never would have done it on my own. . . . But it does give you an idea of what a really decent person Scott was and how he would give up a whole day to an unknown" young instruc- tor (Cain 1989). Frank joined the honorary scientific fraternities. Gamma Alpha and Phi Sigma, and he conducted short field trips while in Rhode Island. But his life was not all science; the war was on. and he had to take his turn up in the college tower as an airplane spotter, mini- mally trained in the silhouettes of domestic and enemy aircraft. In March of 1944. Frank was hired by the U. S. Geological Survey to edit classified reports in the one-year-old Military Geol- ogy Unit. By September of 1945, he had become chief editor, supervising four geologists and 15 typists and draftsmen. Not long after the birth of their first daughter. Katherine Burling, Frank and Marty prepared for his assignment to the Engineer Intel- ligence Division, Southwest Pacific Area, by moving the family to a small house near Niagara Falls. Frank's position as scientific consultant on terrain intelligence took him first to Manila, where he organized the Natural Resources Section of the General Headquarters of the Supreme Commander for the Allied Powers, in preparation for the occupation of Japan. After two months, he relocated to Tokyo, where he served as chief of the Engineering Geology Unit, Natural Resources Section. He supervised the field checking of terrain intelligence reports and consulted with the U. S. Army on foundation conditions, location of construction materials, and selection of airfield and port sites. Frank became a commodity specialist in precious metals, com- piling data on gold and silver production in Japan. As Frank tells it, "Since I was the paleontologist and didn't know much, they looked around for the least harmful thing for me to do. That's how I was put in charge of the precious metals. My job was mainly to hold audience with Japanese gold and silver mine operators and tell them 'no,' they could not mine gold. It was the perfect bureaucrat's job, sitting there all day saying 'no.' I was also in charge of the vaults of Japan, where I saw more money than I will ever see again, with piles of sheet gold one meter on a side" (Paleo News 1990). There were also stacks of platinum crucibles and "buckets of diamonds." The temptations might have been great, except that the military officer from whom Frank took over the vaults was later arrested for trying to re-enter the United States with a pocket full of diamonds. When asked to join the Geological Survey, Frank knew it was likely to mean an overseas assignment, possibly in a war zone. But he could not know that his work, and that of his colleagues, would help directly in saving the lives of Allied troops and hastening the conclusion both of World War II and the Korean War. In 1946, the U. S. Army recognized his work with the highest civilian award the United States bestows, the Medal of Freedom. One would think a tribute of this magnitude would involve a personal presentation if not a public ceremony, but the medal was simply sent to his home with a two-sentence cover letter. One event during Frank's duty in Japan created enough contro- versy that its echoes can still be heard, and so we cannot exclude it from this account. The story of the Peking Man fossils involved Tribute in Frank Clifford Whilmore. Jr. bones of Sinanthropus pekingensis and other relics, which the Japanese had earlier stolen from China. Frank, as part of his duties, was to take custody of these objects, pending their return to the National Geological Survey of China. But there has always been disagreement over the actual location of the objects and the se- quence of events prior to Frank's arrival on the scene. According to one story, in 1941 the specimens were packed in three cases marked "secret" and turned over to U. S. marines who were evacuating Chinwangtao, China, aboard the Dollar Liner President Harrison. The liner ran aground in the Yangtze River near Shanghai on December 8. and the marines were captured. There is documentary evidence that scientists from the Tokyo Imperial University visited Peking in August, 1942, at the request of the Japanese North China Army and took the collection to Tokyo. After the surrender of Japan to the Allied powers, a letter from the Central Liaison Office of Japan alerted the Allies to the collection's existence. The Natural Resources Section was directed to take action to return the speci- mens, and the section chief dispatched Frank to examine the collec- tion. In his memo. Frank stated he could "find no traces of Sinan- thropus" (Lamp and Huang 1990). Many, including Frank, believe the fossils rest at the bottom of the Yangtze River, but he has been accused of keeping secret information on their whereabouts. To this day, the original fossils have never been recovered, but good casts exist and excavations at the Peking Man site have turned up addi- tional specimens. Frank's overseas work did not end in Tokyo. He spent the spring of 1946 assigned to the 24th Corps, U. S. Army, in Korea to survey and map railroads, major highways, landing beaches and ports, including Inchon, which played an important part in the later U. S. invasion. While in Korea, Frank was promoted to chief of the Military Geology Unit. His return to Washington, D. C, did not mean an immediate end to the family's separation. The government had mushroomed dur- ing the war, with little or no increase in housing, and Frank spent months trying to buy or rent a house in the area. Eventually he found one in West Hyattsville. Maryland, and moved the family south from Niagara Falls. Frank and Marty's last child. Susan Hale, was born in 1948, during their ten years in that house. With the end of war-related work, it was assumed that the Military Geology Unit would be shut down, and Frank was looking forward to joining the Paleontology and Stratigraphy Branch and returning to his fossil vertebrates. But the war had demonstrated to the U. S. military how little it knew about foreign geology, and so the unit was transformed into a regular branch of the Geological Survey. Frank stayed on as chief until 1959. His management skills and abilities, in part acquired in Asia, enabled Frank to develop and direct the worldwide activities of the branch, which employed about 120 scientists and support person- nel, with headquarters in Washington and field offices in Tokyo, Heidelberg, and Salzburg. Frank organized interdisciplinary field- mapping programs involving the study of geology, soils, vegeta- tion, hydrology, and topography. Frank's leadership was increasingly recognized and put to use. He chaired numerous groups including the U. S. Geological Survey's Geologic Division Staffing Committee and the committee to com- pile permafrost terms for the first and second editions of the Ameri- can Geological Institute's Glossary- of Geology. He also served as security officer for the Geologic Division between 1948 and 1956. In recognition of the International Geophysical Year in 1958. the Lake Peters Research Station (renamed the G. William Holmes Research Station in 1970) was established in the northeastern part of the Brooks Range of Alaska. Frank was on the team that con- ducted the initial reconnaissance at this offshoot of the Arctic Research Laboratory at Point Barrow and formulated plans for continuing research in the area. Finally, after 15 years of administration, Frank joined the Pale- ontology and Stratigraphy Branch of the Geological Survey as a senior specialist in vertebrate paleontology. He was assigned an office at the National Museum of Natural History. "When I came back to paleontology ... I was for all intents and purposes, fresh out of graduate school, although I'd had my Ph.D. for seventeen years" (Cain 1989). Frank became the informal chief of the survey's vertebrate paleontology staff, which included Charles Repenning at the Menlo Park, California, office and Ed Lewis at the Denver office. He launched a series of diverse investigations. In 1959 and 1960, he collected and studied Miocene and Pleistocene vertebrates from Martha's Vineyard, Massachusetts, as part of the work done by the Engineering Geology Branch. His biostratigraphy of that complexly deformed area helped determine the history of Pleisto- cene deformation on the island. From 1959 through 1965, Frank conducted biostratigraphic studies of Paleozoic and Mesozoic fish and Tertiary mammals from Wyoming and Montana, to aid ongoing geologic mapping there. He was principal investigator for field and laboratory studies of Mio- cene mammals from Panama between 1962 and 1965. This resulted in a biostratigraphic correlation with faunas in Texas and Florida, established that the Miocene mammal fauna of Panama was en- tirely of North American affinity, and helped to define a circum- Caribbean Miocene zoogeographic province and to delineate the southern extent of the North American land mass. These results were published in Science. At about the same time, Frank began collaborating with Bertrand Schultz and Lloyd Tanner of the University of Nebraska on work at Big Bone Lick, Kentucky. This important Pleistocene site is the type locality of Mammut americanum, the American mastodon, and Bootherium botnhifrons. an extinct musk ox. It is also the site where, on Thomas Jefferson's orders, explorers Lewis and Clark collected bones for shipment back to the amateur scien- tist and president of the United States. The team's field and lab studies, from 1962 through 1970. of the late Pleistocene mammals and stratigraphy of Kentucky contrib- uted to the geomorphic history and paleoclimatology of the Ohio valley. After five summers of field work, their results helped con- vince the state of Kentucky to create Big Bone Lick State Park, ensuring preservation of the site. For his efforts, Frank was anointed an Honorable Kentucky Colonel by the state. Meanwhile, Frank was also being exposed to fossil marine mammals, thanks to his close association with Remington Kellogg, world-renowned expert, who worked in the Paleobiology Depart- ment of the museum. It was Dr. Kellogg who made the west side of the Chesapeake Bay a permanent fixture in Frank's life as Frank increasingly helped the elder paleontologist and, after his death in 1 969, took over some of his work. During the late 1960s and early 1970s. Frank was principal investigator for the Calvert Cliffs Paleontology Project on Chesa- peake Bay. This project entailed detailed interdisciplinary paleo- ecological and stratigraphic studies during excavation for the Calvert Cliffs nuclear power plant. Funding for this work came from the Ford Foundation and National Geographic Society. The Calvert Cliffs Paleontology Project opened the door to Frank's association with the National Geographic Society. In 1971, he was asked to join the prestigious Committee for Research and Exploration, serving as Dr. Kellogg's replacement. The committee grants millions of dollars each year for research projects throughout the world, some of which are later described in the society's maga- zine. In 1 972, Frank returned to Alaska, this time to Amchitka Island, where he collected fossils of the historically extinct Steller's sea cow (Hydrodamalis gigas). He and others worked on the rate and R. E. Eshelman and L. W. Ward Figure 2. Frank C. Whitmore, Jr., 1965, holding skull of fossil musk ox collected in 1 807 by William Clark (of Lewis and Clark fame) and exhibited in the White House in the same year. mode of Pleistocene uplift of the island, as indicated by beach deposits, which were critical to the prediction of the effects of nuclear testing. Work on Oligocene whales from South Carolina resulted in two publications in 1975 and 1976. Since 1976, Frank has been princi- pal investigator on the study of Paleocene vertebrates from Saudi Arabia. Paleoecologic studies of this estuarine fauna established the geographical position of part of the ancient Tethys Sea, and contrib- uted to the delineation of lime deposits needed for cement manufac- ture. Like that of the other research scientists at the Geological Survey, part of Frank's job involved handling "examination and report" (E & R) requests. Some were submitted by colleagues in other disciplines whose investigations turned up pieces of what might be bone. Others came in via USGS public-relations people from citizens who wanted to know something about their backyard digs or vacation treasures. One E & R stands out above all others, both for its unusual nature and because it landed on the desk of a man whose knowledge of historical time is as acute as his understanding of geologic time. This was "The Case of the Papal Proboscidean." Sylvio Bedini, then deputy director of the National Museum of History and Technology (now the National Museum of American History) asked Frank to identify, from photographs, some bones dug up during the air-conditioning of the papal apartments in the Vatican. Everyone was puzzled when Frank identified at least one bone as that of an elephant. Further research revealed that in 1514, when Pope Leo X had a stranglehold on the spice trade to the Far Fast. King Emmanuel the Great of Portugal wanted a share of the action. To get on the good side of the pope. Emmanuel presented him with a young elephant. No elephant had been seen in Rome since the time of Hannibal, and it proved to be a great curiosity — especially as it had been trained to genuflect whenever the pope appeared. It also held water in its trunk and squirted designated victims on the command of its trainer. One day. the elephant's keepers decided they would gild the elephant from head to toe as a surprise for the pope. The unexpected surprise was that the gilding killed the elephant. The devastated pope directed the papal painter, who happened to be Raphael, to paint a life-size mural of the elephant; Raphael felt this was beneath him and ordered an appren- tice to complete the mural on the palace wall. The elephant was subsequently buried beneath the painting. The mural is now gone but the bones remain (Whitmore 1978). This and many others stories have been shared gleefully over lunch at the museum, whose collegiality is perfectly suited to Frank's temperament. For years, he took a bag lunch to the office of his longtime friend Harry Ladd. As the other members of the lunch mess would appear, they would array themselves around the desk with their sandwiches and the talk would begin. Here, one might say, is an example of the cross-fertilization that takes place in a great museum — scholars engaged in the exchange of ideas. Ideas certainly were exchanged — on politics, reminiscences of fieldwork, stories of questionable taste, the Washington Redskins; innumer- able small bets were made, usually on sports or politics, and duly recorded on Harry's desk calendar (Whitmore and Tracey 1984). A story Frank tells on himself concerns a day. decades past, when he was collecting a fossil whale along the shores of Chesa- peake Bay. Typically, there were a few helpers along that day — amateur collectors and aspiring young scientists. In the course of the long day, one of these youths said to Frank — surely out of the greatest respect — "You must be one of the last of the old-time collectors!" And so he is. Perhaps it is Frank's memory of W. B. Scott and that day at Princeton, or perhaps it is simply Frank's openness and interest in people that cause him to be so generous with his time and his knowledge. He frequently gives tours behind the scenes at the museum to school and college groups, out-of-town visitors, and amateur collectors. In the 1950s, Washington-area school children watched him on the early WETA television science series Time for Science. And school children nationwide now see him on one segment of PBS's 26-segment story The Voyage of the Mimi when the young protagonist visits Frank at the museum to learn about whales and paleontologists. Frank retired from the United States Geological Survey in 1984, but he continues his work as a research associate and curator emeritus of the Smithsonian Institution in his old office there. His current studies include the taxonomy and description of fossil Plio- cene whales and terrestrial mammals from the Lee Creek phosphate mine at Aurora. North Carolina, and description of Miocene marine mammals from the Pisco Formation of Peru. Many of his papers and publications have been deposited in the Smithsonian Institution archives. While Frank came late to his life's research, his colleagues and employers valued highly the management and leadership he brought with him to the museum. His ability to listen, to draw others out. and to mediate discussion of touchy subjects has often been tested. Frank was appointed chair of the joint U. S. Geological Survey/ Smithsonian Institution committees for the design of new labs and for decisions regarding the paleontology collections. In 1971, he was general chair of the Geological Society of America meetings in Washington, with 4300 people attending. For the American Asso- ciation for the Advancement of Science and the Mexican Science Council, he chaired a symposium on land connections between North and South America. Tribute to Frank Clifford Whitmore, Jr. < Figure 3. Frank C. Whitmore, Jr.. studying fossil whales at the Univer- sity of Otago. Dunedin, New Zealand. 1988. The list of institutions on whose panels and committees he has served is long and varied. It includes the Department of Defense, American Geological Institute, National Research Council, and the Center for the Study of the First Americans. He has provided scientific guidance to the Schoelkopf Geological Museum in Niagara Falls, New York, and to exhibit specialists at the National Museum of Natural History and the state of South Carolina. In 1979, he served as general chair of the International Centen- nial Symposium of the USGS on "Resources for the 21st Century." Brought together were some 500 scientists, corporate executives, and government officials from 48 countries. Frank spent five years planning the agenda, booking speakers, and editing and rewriting many of the foreign papers to make them publishable in English. Professional societies have provided Frank with a continuous thread throughout his varied career. And they have all benefited by his membership. He's at home both at small, convivial monthly meetings and at huge national conventions. His wit at the dinner table is as valuable to many as his chairing of committees and symposia. Frank was a founding member of the Society of Vertebrate Paleontology in 1940: he later served on its executive board and is an honorary life member. He's been a member of the Paleontologi- cal Society of America since 1942. In 1944. he joined the Paleonto- logical Society of Washington, later serving as vice-president and president. That same year, he joined the Geological Society of Washington, becoming councillor, secretary, first vice-president, and then president. In 1945, perhaps to show that he couldn't live without meetings wherever he might be stationed, he helped found the Geological Society of the Philippines. He became a fellow of the Geological Society of America in 1947, and has served on its Penrose Medal Committee and as its Penrose citationist. Three years later, he became a fellow of the American Association for the Advancement of Science, eventually serving as section secretary and chairman, councillor, and chair of the Newcomb Cleveland Prize Committee. Although not of the highest scientific significance, one mem- bership gave Frank a wider forum for his well-known humorous talents. In 1944. he joined other young U. S. Geological Survey types as a member of the Pick and Hammer Club. Begun in 1 894 as the Association of Aspiring Assistants, it was originally an offshoot of the Geological Society of Washington. By the mid-twentieth century, this platform for talks by unestablished geologists was best known for its annual show, which spoofs the stuffed shirts of the geologic bureaucracy. Frank participated fully in writing songs and dialogue, acting in minor and major roles, singing lustily, and dancing, among other things, the academic gavotte in cap and gown. In 1967. he was invited to be a guest on the television program To Tell the Truth. He was, in fact, the "true" vertebrate paleontolo- gist, but got himself in hot water by misunderstanding one of the questions. Instead of answering "dinosaur" to the query. "What was the largest carnivore in the world?" he replied, "bear." Conse- quently, the panel was led astray and guessed the wrong contestant. Immediately following the show. Kitty Carlisle lit into him, "Imag- ine, a bear! How ridiculous!" And when Frank returned to his office at the museum, there in his chair sat a huge femur of a meat-eating dinosaur. In 1981, Frank was traveling in China when he was awarded the Meritorious Service Award by the Interior Department. This got him out of shaking hands, on stage, with then-secretary James G. Watt, not one of his heroes. When it was Frank's time to receive his 40-year service scroll and pin, a representative came to his office, stated something to the effect that he knew Frank wouldn't want a lot of pomp and ceremony, discovered he didn't have the award in his pocket, and left saying it would be sent in the mail. There is a type of honor frequently bestowed by his colleagues that has been harder to receive. Because of his friendship, respect for biography, and skill as a writer, over the years Frank has been asked to prepare memorials to several fellow scientists. As a result, his bibliography includes memorials to Alfred Romer, Remington Kellogg. Harry Ladd, Willy Postel. Louis Ray, John Huddle, and Charley Johnson. They are graceful tributes to men whose work and character he admired. In addition to his own writing. Frank has reviewed many books and papers. One stands out from all the rest, and we quote it here as an example of the good doctor's breadth of knowledge, attention to detail, and mellow humor. From his review of the sixth edition of "Eoornis pterovelox gobiensis" Whitmore stated, "Far ahead of his time. Fotheringham thoughtfully melded together every aspect of the natural history of his subject: the occurrence of its fossil ances- tors, references to it in Egyptian hieroglyphics, its appearance in Cro-Magnon cave paintings, and at the other extreme of scientific inquiry, the most painstaking physiologic studies of wing-beat fre- quency and of the pH cycle of the bird's beak fluid, observed under the most difficult conditions over an entire year. The author's discovery that each dumbbell-shaped egg contains a male and a female, and his analysis of the part played by parthenogenesis in the evolution of the genus, makes the mind boggle" (Whitmore 1967). But a truer indication of a person's worldliness and erudition R E. Eshelman and L. W. Ward may be the letter to the editor, of which Frank has written many. Here we quote from a joint response by Whitmore and Hotton to a 1972 article in Smithsonian Magazine entitled "Fantastic Animals Proved Tall Timber of our Mythology." "The taxonomy of the sidehill gouger {Membriinequales declivitous) is more complicated than Carson suspects. There are subspecies, not yet formally named but recognized by local inhabitants, and there is more than a scintilla of evidence that the morphology and habits of these taxa can be correlated. Among many examples we can cite is the sidehill dodger, which inhabits the Driftless Area of Wisconsin; the dextrosinistral limb ratio approaches unity although the metapodials on the down- hill side are noticeably stouter. The sidehill gouger is the common Pennsylvanian species, but it has been speculated (Hotton, in lilt. 1972) that all of the different varieties spring from the little-known strike-runner (Crestophilus ambiguus) in which the right front and the left rear limbs are short and the left front and right rear are long, or vice versa. These animals ran the long Tennessee ridges in late Pleistocene time but became extinct during the Pleistocene epoch, perhaps because of trauma resulting from attempts to run ridges of newly formed glacial cirques" (Hotton and Whitmore 1972). A professional life as busy as Frank*s might leave the impres- sion that he has little activity outside his career, but his nonscientific interests are many. When their family was young, he and Marty took their children on many day trips, to Rock Creek Park and the zoo, on picnics, to Civil War battlefields, and the like (mysteriously, after he began working at the museum, Saturday trips to the Smithsonian decreased markedly). Drives to the country, and longer trips to visit relatives, invariably involved a stop along the road so he could give a brief lesson on an especially noteworthy outcrop. Despite the financial strains of a growing family, the Whitmores made sure to attend an occasional play and took the children along at a young age. This meant Saturday matinees at the National Theater, where dramas and musicals tried out on the way to Broad- way, and a brand-new local group. Arena Stage, put on productions at the Hippodrome and the Old Vat. The Whitmores have had season tickets to Arena Stage since the 1960s. Certain aspects of Frank's work have intersected with his and Marty's other interests, especially their love of travel. Field trips to Martha's Vineyard took them back to the New England of their student years. Frank's membership on the Committee for Research and Exploration means their joining the National Geographic Society's triennial tours of current research projects; hence their trips to South America. Africa, Europe, Asia, and Australia. At home, alongside mementoes of their long life and many travels together, the Whitmores display tokens of affection from friends and family. Half of the pictures, statuettes, and trinkets depict the moose, Frank's long-time alter ego. The other half are whales, in honor of the marine-mammal paleontologist. In addition. he has accumulated a veritable wardrobe of whale neckties, in colors suitable for every occasion — from Christmas to St. Patrick's Day to a meeting at the National Geographic Society. Frank continues to serve as vice-chair of the National Geo- graphic Society's Committee for Research and Exploration. He attends to his mail and his research at the museum. When in town. Frank and Marty still frequent the Tuesday Lunch Group of mu- seum denizens, at the Beeffeeders on Tenth Street. They continue to travel extensively, in part to keep up with the four children, now scattered across the country. There are five grandsons, three step- grandchildren, and great-grands may soon be on the way. Many of us owe some measure of our professional progress to this man who took lime to listen, teach, and put up with the ignorance of budding scientists and up-and-coming collectors. While working with Frank in the late 1960s, we used to refer to him as "the good doc," a nickname we still use with reverence and respect for a man who means that and more. To us. he will always be "the good doc." ACKNOWLEDGMENTS We are first and foremost thankful to Martha Kremers Whitmore. who was generous in her assistance. When Ralph Eshelman called Marty at home (knowing Frank was safely down- town in his office) to tell her of this tribute, he told her he had something secret to discuss concerning her husband. In typical Marty fashion she responded in a happy, friendly voice. '•How- mysterious and exciting! What is it?" She immediately offered a file she has kept on Frank from early in their marriage, including newspaper clippings, programs, personal notes, letters, photo- graphs, certificates, and more. Without this material, our tribute to Frank would be woefully incomplete. But just as revealing were the little notes she has frequently written him. supporting his work and commending his achievements. They illustrate her admiration for, and devotion to. her husband. For this love he is most fortunate. Many of Frank's friends and co-workers added to this tribute. Everyone approached was enthusiastic and supportive, making our job all the easier. It is obvious from these contacts that Frank has many friends who wish him well. Among them are Edwin Snider. Joshua Tracey. Ellis Yochelson, Barry Bishop, Tom Dutro. Richard and Mary Ellen Williams, Pamela Henson, Clayton Ray, David Bohaska, Nicholas Hotton, Warren Blow, and Joseph Cain. LITERATURE CITED Cain. J. 1989. Oral history interview of Frank C. Whitmore. Jr.. 8 August 1989. Smithsonian Institution Archives. Hotton. N.. and Whitmore. F. C, Jr. 1972. Letters to the editor: Fantastic Animals. Smithsonian Magazine 3 (7): 13. Lamp. J., and Weiwen. H. 1990. The story of Peking Man: From archaeology to mystery. Foreign Language Press. Beijing. China. and Oxford University Press, Hong Kong. Olsen. F. R.. and Whitmore. F. C Jr. 1 944. Machine for serial sectioning of fossils. Journal of Paleontology 18: 210-215. Paleo News. 1990. Department of Paleobiology newsletter. 1(1): 5. Whitmore, F. C, Jr. 1953. Cranial morphology of some Oligocene Artiodactyla. United States Geological Survey Professional Paper 243-H: 117-159. . 1967. Review: "Eoornis pterovelox gobiensis" by Augustus G. Fotheringham. Journal of Paleontology 41: 1302-1303. — . 1978. The papal proboscidean. The Cross Section 4(3): 12. 1989. Letter to John A. Wilson. 17 March 1989. Smithsonian Institution Archives — , and J. I. Tracey, Jr. 1984. Memorial to Harry Stephen Ladd. 1899-1982: Geological Society of America Memorials 14: 1-7. SELECTED BIBLIOGRAPHY OF FRANK C. WHITMORE. JR. 1938. (with F. B. Phleger. Jr.). Two young merycoidodonts. American Journal of Science, 5th ser., 36 (215): 377-388. 1942. Endocranial anatomy of some Oligocene Artiodactyla. Geological Society of America Bulletin 53: 1842-1843. 1944. (with F. R. Olsen). Machine for serial sectioning of fossils. Journal of Paleontology 18: 210-215. 1946. Planned peacetime work in military geology. Geological Society of America Bulletin 57: 1242. 1947. Cranial morphology of some early Tertiary Artiodactyla. Harvard University, Summary of Theses 1943-1945, pp. 198-200. 1948. Military geology. Professional Geographer 7: 7-16. — . Digest on problems of military geology in the event of a national emergency. Research and Development Board. National Military Establishment, Digest Series 10. 1040. Review: "Geology and Paleontology (Fiat Review of German Science, 1939-1946)." Science 109 (2834): 425-426. 1951. Translation: Kleinschmidt. A. 1951. Uber ein skelet und eine Rekonstruktion des ausseren Habitus der Reisenseekuh, Rhytina gigas Zimmerman 1780. Zoologischer Anzeiger 146, heft 9-10' 292-314. 1953. Cranial morphology of some Oligocene Artiodactyla. United States Geological Survey Professional Paper 243-H: 1 17-159. Tribute to Frank Clifford Whitmore, Jr. . Currenl Japanese studies of Desmostylus. Society of Vertebrate Paleontology News Bulletin 37: 9. 1956. (with F. M. Swart/). Ostracoda of the Silurian Decker and Manlius limestones in New Jersey and eastern New York. Journal of Paleon- tology 30: 1029-1091. 1957. (with R. Kellogg). Marine mammals — annotated bibliography. Pp. 1223-1226 in J. W. Hedgpeth (ed.). Treatise of marine ecology and paleoecology: Ecology. Geological Society of America Mem- oir 67. vol. I — . (with R. Kellogg). Mammals — annotated bibliography. Pp. 1021-1024 in H. S. Ladd (ed.). Treatise on marine ecology and paleoecology: Paleoecology. Geological Society of America Memoir 67. vol. 2. 1958. Further information on archeological and fossil material from Choukoutien. China. Asian Perspectives 2(1): 54-56. — . Geologic writing for the nongeologist. Geological Society of America Bulletin 69: 1662. 1960. Fossil mammals from Ishigaki-shima. Ryukyu-retto. United States Geological Survey Professional Paper 400-B. art. 171: B372- B374. 1961. Edited: Discoveries in Ancient Life, by V. Brown. Science Materi- als Center. New York. 1962. Paleontology, evolution, findings on 50-foot fossil whale. United States Geological Survey Professional Paper 450-A: 74. — . Review: "Mestonakhozhdeniya Tretichnykh Nazemnykh Mlekopitayushchikh na Territorii SSSR." by A. A. Borisiak and E. I. Behaeva (Tertiary terrestrial mammalian localities in the territory of the USSR). International Geology Review 4: 863-864. 1963. (with C. B. Schultz, L. G. Tanner. L. L. Ray. and E. C. Crawford). Paleontologic investigations at Big Bone Lick State Park, Ken- tucky: A preliminary report. Science 142 (3596): 1 167-1 169. 1965. (with R. H. Stewart). Miocene mammals and Central American seaways. Science 148 (3667): 180-185. — . Presentation of the Paleontological Society Medal to G. Arthur Cooper. Journal of Paleontology 39: 520-522. 1966. (with 1. L. Ray. C. B. Schultz, L. G. Tanner, and E. C. Crawford). Kentucky. Pp. 53-63 in Guidebook for Field Conference G, Great Lakes-Ohio River Valley: International Association for Quaternary Research. 7th Congress, U.S.A., 1965. Nebraska Academy of Sci- ence, Lincoln, Nebraska. — . Review: "The Age of Reptiles," by E. H. Colbert. Geotimes 10 (6): 32. — . Review: "Dinosaur Hunt," by G. O. Whitaker. Geotimes 10(6): 36-37. — (withC. B. Schultz and L. G. Tanner). Pleistocene mammals and stratigraphy of Big Bone Lick State Park. Kentucky. Geological Society of America Special Paper 87: 262-263. 1967. (with H. L. Foster). Pamhera atrox (Mammalia: Felidae) from central Alaska. Journal of Paleontology 4 1 : 247-25 1 . . (with C. B. Schultz, L. G. Tanner, L. L. Ray. and E. C. Crawford). Big Bone Lick, Kentucky: A pictorial story of the paleontological excavations at this famous fossil locality from 1962 to 1966. University of Nebraska State Museum. Museum Notes 33. — . (with K. O. Emery, H. B. S. Cooke, and D. J. P. Swift). Elephant teeth from the Atlantic continental shelf. Science 1 56(378 1 ): 1 477- 1480. — . Elephants under the sea. Science Digest 61 (4): 15-16. — . Edited: "The Changing Earth." by J. Viorst. Bantam. New York. — . Presentation of the Paleontological Society Medal to Alfred S. Romer. Journal of Paleontology 41: 817-819. — . Review: "Marsh's Dinosaurs." by J. H. Ostrom and J. S. Mcin- tosh. Geotimes 12(4): 34-36. — . Review: "Fossil Lake. Oregon: Its Geology and Fossil Faunas." by I. S.Allison. Journal of Paleontology 41: 814-815. — . Review: "A Review of the Macropodid Genus Slhenurus," by R. H. Tedtord. Journal of Paleontology 41: 815. — . Review: "Eoornis pterovelox gobiensis," by A. G. Fotheringham. Journal of Paleontology 41: 1302-1303. — . Review : "The Fossil Macropodidae from Lake Menmdee. New South Wales." by R. H.Tedford. Journal of Paleontology 41: 1303- 1 304. 1468. Review: "The Bering Land Bridge," edited by D. M. Hopkins American Scientist 56: I50A-152A. — . Memorial to Albert Williams Postel. Geological Society of America. Proceedings for 1966. pp. 341-346. 1%9. Adaptive radiation of Cetacea. United States Geological Survey Professional Paper 650-A: Alt) — . (with C. B. Schultz. L. G. Tanner, and L. L. Ray). Geologic and faunal evidence of the Quaternary deposits at Big Bone Lick. Kentucky. Geological Society of America. South Central Section Meeting. Lawrence. Kansas. Abstracts with Programs for 1969. pt. 2, pp. 24-25. — . Review: "Fossil Vertebrates of Southern California," by T Downs. Journal of Paleontology 43: 1307-1308. . Review: "Notes and Comments on Vertebrate Paleontology," by Alfred S. Romer. Geotimes 14 (9): 33, 36. -. Ecologic and stratigraphic implications of some Cenozoic ver- tebrates in the Atlantic coastal plain. Geological Society of America. Abstracts with Programs for 1969. pt. 7, p. 236. 1970. (with A. E. Sanders). Extinct porpoises collected near Charleston. South Carolina. Marine Newsletter I (6): 1. — . Cenozoic of the U. S.. New toothed whale from the Yorktown Formation in Virginia. United States Geological Survey Profes- sional Paper 700-A: A 145. 1 97 1 . Calvert Cliffs project. Science 173(3993): 192-193. . Vertebrate biofacies and paleoenvironmental history of Maryland Miocene. Guidebook 3. Maryland Geological Survey, pp. 31-36. — . (with R. E. Gemant and T G. Gibson). Description of selected Miocene exposures. Guidebook 3. Maryland Geological Survey, pp. 49-58. 1972. (with L. M. Gard. Jr., and G. E. Lewis). Steller's sea cow in Pleistocene interglacial beach deposits on Amchitka, Aleutian Is- lands. Geological Society of America Bulletin 83: 867-870. — . (with W. I. Finch and J. D. Sims). Stratigraphy, morphology, and paleoecology of a fossil peccary herd from western Kentucky. United States Geological Survey Professional Paper 790. — . Remington Kellogg ( 1892-1969). American Philosophical So- ciety Yearbook 1972. pp. 205-210. 1973. (with C. E. Ray). Paleontology. Pp. 67-68 in T Simkin, W. G. Reeder. and C. MacFarland (eds.). Galapagos Science: 1972 Status and Needs. Smithsonian Institution, Washington. D. C. 1974. Collections of South African mammal-like reptiles in the United States: A correction. Journal of Paleontology 48: 607. — . Memorial to Remington Kellogg, 1892-1969. Geological Soci- ety of America Memorials 4: 117-129. — . (with J. Knapp). Remington Kellogg. October 5. 1892-May 8. 1969. Biographical Memoirs, National Academy of Sciences 46: 159-189. — . (with A. E. Sanders). The Cetacea 30 million years ago. Ameri- can Zoologist 15: 812. . A thousand nights' entertainment. The Geological Society of Washington. 1893-1975. Geotimes 20 (7): 14-15. . Presentation of the Penrose Medal to Maurice Ewing. citation by Frank C. Whitmore. Jr. Geological Society of America Bulletin 86: 1161-1168. 1976. (with A. E. Sanders). Review of the Oligocene Cetacea. System- atic Zoology 25: 304-320. 1977. Memorial to Alfred Sherwood Romer. 1894-1973. Geological Society of America Memorials 5: 1-10. — . Review: "Origin and Evolution of the Elephantidae." by V J. Maglio. Journal of Mammalogy 58: 457—159. — . (with S. H. Whitmore). Review: "Dinosaurs," by N. Sullivan. AAAS Science Books and Films 13(2): 94-95. — . (with L. M. Gard. Jr.) Steller's sea cow (Hydrodamalis giga.s) of late Pleistocene age from Amchitka. Alaska. United States Geo- logical Survey Professional Paper 1036. . The papal proboscidean. The Cross Section 9(3): 12. — . (with S. H. Whitmore). Review: "What Really Happened to the Dinosaurs'?" by D. Cohen. AAAS Science Books and Films 14(3): 181. — . Review: "Athlon: Essays on Palaeontology in Honour of Loris Shano Russell." edited by C. S. Churcher. Journal of Paleontology 52: 1401-1403. 10 R. E. Eshelman and L. W. Ward 1979. Alexander Wetmore. 1886-1978. Society of Vertebrate Paleontol- ogy News Bulletin 1 16: 64-65. . Review: "The Ecology of Fossils: Illustrated Guide," edited by W. S. McKerrow. AAAS Science Books and Films 15(3): 147. . (with C. T. Madden. 1. M. Nagvi. D. L. Schmidt. W. Langston, — . Review: "A New Look at the Dinosaurs." by A. Charig. AAAS Science Books and Films 19 (2): 78. Hemphillian vertebrate fauna from Mobile County. Alabama. Jr.. and R. C. Woodl. Paleocene vertebrates from coastal deposits in the Harrat Hadan area. At Taif region. Kingdom of Saudi Arabia. United States Geological Survey. Saudi Arabian Mission. Project Report 269, Open-file Report OF-80-227. 1980. Memorial to Louis Lamy Ray. 1909-1975. Geological Society of America Memorials 10: 1-8. . Review: "Dinosaurs and People: Fossils. Facts, and Fantasies." by L. Pringle. AAAS Science Books and Films 15 (4): 226. — Review: "Mesozoic Mammals: The First Two-thirds of Mamma- lian History." edited by J. A. Lillegraven, Z. Kielan-Jaworowska, and W. A. Clemens. AAAS Science Books and Films 16(2): 75. . Resources for the 21st Century: Summary and conclusions of the International Centennial Symposium of the United States Geo- logical Survey. United States Geological Survey Circular 857. 1982. Review: "Dinosaurs in Your Backyard," by W. Manetti. AAAS Science Books and Films 18 (2): 76. — . (with M. E. Williams). Edited: Resources for the Twenty-first Century — Proceedings of the International Centennial Symposium of the United States Geological Survey. United States Geological Survey Professional Paper 1 193. — . The International Centennial Symposium: Background and objectives. P. 2 in F C. Whitmore, Jr.. and M. E. Williams (eds.). Resources for the Twenty-first Century — Proceedings of the Inter- national Centennial Symposium of the United States Geological Survey. United States Geological Survey Professional Paper 1193. . Remains of Delphimdae from Sahabi Formation. Garyounis Scientific Bulletin. University of Garyounis, Libya. Special Issue 4: 27-28. . (with C. T. Madden. K. W. Glennie. R. Dhem. D. L. Schmidt, R. J. Ferfaglia, and P. J. Whybrow). Stegotatra belodont (Proboscidea. Gomphothenidae) from the Miocene of Abu Dhabi. United States Geological Survey, Saudi Arabian Project Report, Jiddah, pp. 1-22. 1983. (with J. E. Repetski). Memorial to John Warfield Huddle, 1907- 1975. Geological Society of America Memorials 13: 1-7. — . (with C. T. Madden and D. L. Schmidt). Masritherium (Artiodactyla, Anthracotheriidae) from Wadi Sabya, southwestern Saudi Arabia: An earliest Miocene age for continental rift-valley volcanic deposits of the Red Sea margin. United States Geological Survey Open-file Report OF-83-61 . — . Review: "First Look at Dinosaurs." by M. E. Selsam and J. Hunt. AAAS Science Books and Films 18 (5): 273-274. . Remington Kellogg, 1892-1969. Pp. 15-24 in C. E. Ray (ed.). Geology and Paleontology of the Lee Creek Mine, North Carolina, vol. 1. Smithsonian Contributions to Paleobiology 53. . (with C. T. Madden). Tertiary vertebrate faunas of Arabian Peninsula. Geological Society of America, Abstracts with Pro- grams 15 (6),: 633. Pp. 71-72 in W. C. Isphording and G. C. Flowers (eds). Differen- tiation of unfossiliferous clastic sediments: Solutions from the southern portion of the Alabama-Mississippi coastal plain. Tulane Studies in Geology and Paleontology 17(3). 1984. Cetaceans from the Calvert and Eastover Formations. Pamunkey River. Virginia. Pp. 227-231 in L. W. Ward and K. Krafft (eds.). Stratigraphy and paleontology of the outcropping Tertiary beds in the Pamunkey River region, central Virginia coastal plain. Guide- book for Atlantic Coastal Plain Geological Association 1984 Field Trip. Atlantic Coastal Plain Geological Association. . Land mammals from the Calvert Formation. Pamunkey River, Virginia. Pp. 236-239 in L. W. Ward and K. Krafft (eds.). Stratigra- phy and paleontology of the outcropping Tertiary beds in the Pamunkey River region, central Virginia coastal plain. Guidebook for Atlantic Coastal Plain Geological Association 1984 Field Trip. Atlantic Coastal Plain Geological Association. . (with J. I. Tracey, Jr.). Memorial to Harry Stephen Ladd, 1899- 1982. Geological Society of America Memorials 14: 1-7. — . Review: "American Science in the Age of Jefferson," by J. C. Green. Earth Science History 3: 188-190. 1986. (with G. V Morejohn and H. T Mullins). Fossil beaked whales — Mesoplodon longirostris dredged from the ocean bottom. National Geographic Research 2(1): 47-56. . Whale worries confirmed. Geotimes 31 (4): 2. . Review: "General Features of the Paleobiological Evolution of Cetacea," by G. A. Mchedlidze. Quarterly Review of Biology 61: 249-250. 1987. Cetacea from the Sahabi Formation. Libya. Pp. 145-152 in N. T. Boaz et al. (eds.). Neogene Paleontology and Geology of Sahabi. Liss, New York, New York. — . Review: "Mammal Evolution: An Illustrated Guide." by R. J. G. Savage. AAAS Science Books and Films 22 (4): 232-233. — . (with R. N. Oldale and J. R. Grimes). Elephant teeth from the western Gulf of Maine, and their implications. National Geographic Research 3 (4): 439-146. . A delphinoid ear bone from the Dam Formation (Miocene) of Saudi Arabia. Pp. 447—450 in P. J. Wybrow (ed.). Miocene geology and paleontology of Ad Dabtiyah, Saudi Arabia: Bulletin of the British Museum (Natural History) (Geology) 41: 447-450. . (with R. N. Oldale and J. R. Grimes). Late-Wisconsinan el- ephant teeth from the western Gulf of Maine. Geological Society of America, Abstracts with Programs, p. 794. -. Edited: "Fossil Cetacea of the Caucasus," by G. A. Mchedlidze. Translation published by Smithsonian Institution Press. Washing- ton. D. C. 1990. Review: "Frozen Fauna of the Mammoth Steppe: The Story of Blue Babe." by R. D. Guthrie. AAAS Science Books and Films 26 (2): 111). The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematics and Mode of Life Richard H. Tedford Department of Vertebrate Paleontology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024 Lawrence G. Barnes Vertebrate Paleontology Section, Natural History Museum of Los Angeles County, WO Exposition Boulevard, Los Angeles, California 90007 Clayton E. Ray Department of Paleobiology, United States National Museum of Natural History, Smithsonian Institution. Washington. D.C. 20560 ABSTRACT. — Species of the large extinct early Miocene carnivoran Kolponomos Stirton. 1960. are known from a few fossils found in marine rocks along the northeastern margin of the Pacific Ocean in Oregon and Washington, U.S.A. These animals are notable for their massive skulls with markedly deflected rostra and broad, crushing cheek teeth like those of a sea otter. Originally based on an incompletely preserved snout from the marine lower Miocene Clallam Formation at Clallam Bay. Clallam County. Washington, and questionably assigned by Stirton to the Procyonidae, the taxon has until recently remained enigmatic and not certainly assigned to any particular carnivoran family. Additional specimens from the type locality, including a nearly complete cranium with some teeth, provide new data on the cranial morphology of the species. Another specimen. consisting of a nearly complete cranium, mandible with dentition, and some postcranial bones, from the lower Miocene Nye Mudstone on the Oregon Coast, represents a new species. Kolponomos newportensis. The new material demonstrates that Kolponomos is an ursoid most closely related to members of the paraphyletic family Amphicynodontidae. Similar phylogenetic roots have been postulated for the pinnipeds as a whole, and cladistic analysis implies a sister-taxon relationship of Kolponomos with the Pinnipedimorpha. The few postcranial bones available demonstrate that Kolponomos was amphibious but not a strong swimmer. Kolponomos was probably littoral in distribution, all specimens having been discovered in nearshore marine rocks. The crushing cheek teeth would have been suited to a diet of hard-shelled marine invertebrates. The anteriorly directed eyes and narrow snout indicate that Kolponomos could view objects directly in front of its head, of benefit to an animal that would selectively eat epifaunal marine invertebrates. The elongated upper canine and third incisor teeth clustered in thickened bone at the anterior end of the down-turned snout and the posteriorly retracted nasal opening are adaptations that would allow the animal to pry organisms from rocks while keeping its nostrils away from the substrate. Large paroccipital and mastoid processes indicate strong neck muscles that could provide powerful downward movements of the head. These features indicate that Kolponomos probably fed on marine invertebrates living on rocky substrates, prying them off with the incisors and canines, crushing their shells, and extracting the soft parts, as do sea otters. Kolponomos represents an unique aquatic adaptation for marine carnivorans, whose mode of living and ecological niche are approached only by modem sea otters. INTRODUCTION all known specimens of Kolponomos, to redescribe and rediagnose The early Miocene carnivoran genus Kolponomos Stirton, 1960, K. clallamensis. to describe a new species from Oregon, to corn- is known from a few fossils found in marine rocks along the ment on the relationships and taxonomy of the genus, and to discuss northeastern margin of the Pacific Ocean in Oregon and Washing- implications for its functional morphology and behavior, ton. Kolponomos was originally based on an incompletely pre- served snout from the marine lower Miocene Clallam Formation at Clallam Bay, Clallam County. Washington. The relationships and morphology of the type species of the genus. K. clallamensis All specimens were in hard concretionary sandstone matrix. Stirton. 1960, have remained problematic, and for many years the Those from Washington were prepared by use of pneumatic chisels animal was not assigned with certainty to any particular carnivoran and formic acid; those from Oregon were prepared by mechanical family. Stirton (I960) questionably assigned it to the Procyonidae, and air abrasive methods. and was followed by Piveteau ( 1961 ), Romer ( 1966). and Thenius Geologic ages cited herein are modified according to the re- (1969). Carroll (1988) classified the taxon as Carnivora, incertae vised radiometric scale of Dalrymple (1979) and the correlations sedis, and Ray (in Barnes et al. 1985:43) regarded it as an proposed by Armentroutet al. ( 1983). The acronyms for institutions enaliarctine pinniped. are as follows: AMNH, American Museum of Natural History, New Additional specimens, including a nearly complete cranium York, New York; BM(NH). British Museum (Natural History), with some teeth, have now been collected from the Clallam Forma- London, England; LACM. Natural History Museum of Los Angeles tion near the type locality of Kolponomos clallamensis at Clallam County. Los Angeles, California; UCMP. University of California Bay. These provide additional data on the cranial morphology of the Museum of Paleontology, Berkeley, California; USNM. National type species. Barnes et al. (1985) announced the discovery of a Museum of Natural History, Smithsonian Institution. Washington, nearly complete cranium and mandible with some postcranial bones D.C. of Kolponomos from the lower Miocene Nye Mudstone on the Casts of the crania have been placed in AMNH, USNM, UCMP, Oregon coast. This specimen represents a new species of LACM, and the University of Nebraska State Museum. Measure - Kolponomos. The purpose of this study is to describe and illustrate ments, in millimeters, of the crania, dentitions, and mandible of the In A. Berta and T. A. Demere (eds.) Contributions in Marine Mammal Paleontology Honoring Frank C. Whitmore, Jr. Proc. San Diego Soc. Nat. Hist. 29:1 1-32, 1994 METHODS AND MATERIALS 12 R. H. Tedford. L. G. Barnes, and C. 1 Raj species of Kolponomos have been provided in Tables 1 and 2. Cranial restorations of Kolponomos clallamensis are based on all available specimens. SYSTEMATICS Class Mammalia Linnaeus. 1758 Order Carnivora Bowdich, 1821 Suborder Caniformia Kretzoi, 1943 lnfraorder Arctoidea Flower. 1869 Parvorder Ursida Tedford, 1 976 Superfamily Ursoidea (Gray). 1825 Family Amphicynodontidae (Simpson). 1945 Included genera. — This paraphyletic family includes Amphi- cynodon Filhol. 1882. Pachycynodon Schlosser. 1887 (including Paracynodon Schlosser, 1899). Allocyon Merriam, 1930. and Kolponomos Stirton, 1960. Kolponomos Stirton. 1960 Kolponomos Stirton, 1960:346. Kolponomus Carroll. 1988:635. 672. Emended diagnosis of genus. — Stirton's diagnosis was based solely on the holotype of K. clallamensis and amounted to a de- scription of that specimen. New material allows a refinement of that diagnosis. The characters noted are all derived with respect to other Carnivora: I3 with large root; I, vestigial or missing; cheek teeth with strongly inflated principal cusps; P1"3 and P, , with anterior and posterior cingular cusps, P4 also with prominent posterolingual cingular cusp; P4 molariform with large protoconc; M' with large conules. lingual cingulum only between principal cusps; M: mark- edly smaller than and lying posterolingual to M1 with postero- lingually placed metaconule; M, quadrate in occlusal outline; M, triangular with reduced talonid; M, absent. Facial region of skull markedly flexed downward relative to basicranial plane; muzzle deep; nasal retracted to above P1 or P:, its sutural contact with frontal wide, slightly wider anteriorly than posteriorly; long process of premaxilla nearly meeting correspond- ing processes from frontal along nasal suture; palate highly vaulted anteriorly; infraorbital foramen greatly enlarged and opening into shallow fossa in maxilla; infraorbital canal very short; orbit facing forward and relatively small; zygomatic arch widely flaring with strong postorbital process and variably developed masseteric pro- cess; postorbital process of frontal lacking; lacrimal foramen small, variably present; sphenopalatine foramen large and closely associ- ated with posterior palatine foramen; optic foramen small, nearly same size as ethmoid foramen; anterior process of alisphenoid forming strut bracing palate against braincase; mastoid process hypertrophied into long column extending laterally and ventrally; posterior carotid foramen well anterior to posterior lacerate fora- men; foramina for venous occipital sinus in foramen magnum; lambdoidal crest strongly extended posteriorly on either side of midline. Type species. — By original designation. Kolponomos clalla- mensis Stirton, I960. Included species. — Kolponomos clallamensis Stirton, I960, late Early Miocene, Washington; Kolponomos newportensis, new spe- cies. Early Miocene. Oregon. Kolponomos clallamensis Stirton, 1960 Figures 1-7, 13-14 Kolponomos clallamensis Stirton, 1960:347, figs. 1— t. Diagnosis of species. — A species of Kolponomos differing from K newportensis, new species, by the following derived features: cranium with anterior part of palate more highly vaulted; infraorbital foramen larger, approximately twice the diameter, open- ing into prominent fossa; large masseteric process on ventral sur- face of zygomatic arch at jugal/maxillary suture: maxilla rather than jugal forming anterior rim of orbit; paroccipital process larger, oriented vertically rather than posteroventrally; basioccipital nar- rower, especially posteriorly between posterior lacerate foramina; narrow, prominent vertical crest on occiput dorsal to foramen mag- num lacking; zygomatic arch more strongly arched dorsally. In addition, K. clallamensis is distinguished by the following primi- tive features: rostrum narrower, with anterolateral margin of snout around canine and I' not flaring laterally, nearly vertical; hamular process of pterygoid straighter, smaller, not extending so far ven- trally: mastoid process shorter, straighter. and oriented vertically rather than being twisted and projecting anteroventrally beneath the external auditory meatus; intercondylar notch present and deep. Holotype. — UCMP 50056. anterior part of cranium with roots of left I3 and both M"s, collected in 1957 by Mrs. Betty Willison. Type locality.— UCMP V5761, 250 yards east of Slip Point lighthouse near section line. NW 1/4 NE~ 1/4. Sec. 21, T. 32N. R. 12W. and LACM 5933, Slip Point, Clallam Bay, Clallam County, Washington. Referred specimens. — LACM 1 3 1 148. from the type locality, a nearly complete cranium with parts of the left I2 and I' and right and left P: and M1, collected by Albin Zukofsky, II, February 1988; LACM 123547. from Merrick's Bay. Clallam Co., Washington, a fragment of tooth (M,7) collected by William Buchanan. May 1983. Formation and age. — Extensive marine sedimentary deposits are exposed on the north side of the Olympic Peninsula in Washing- ton. The lower Miocene Clallam Formation and underlying Eocene and Oligocene deposits are exposed in wave-cut cliffs and terraces and in man-made excavations for more than 70 miles (112 km) along the south shore of the Strait of Juan de Fuca. The stratigraphy and invertebrate paleontology of this thick sedimentary sequence are well known (e.g., Addicott 1976a,b.c; Armentrout et al. 1983: Feldmanetal. 1991; Rau 1964;Snavely 1983; Snavely et al. 1980; Tabor and Cady 1978). but only a very few fossil vertebrates have been recorded [Stirton I960 (carnivore); Olson 1980 (bird): Domning et al. 1986. and Ray et al. 1994, this volume (desmo- stylian); Barnes 1987, 1989 (whale)]. Specimens of Kolponomos clallamensis are from the Clallam Formation, deposited during the Pillarian Molluscan Stage (Addicott 1976a; Moore and Addicott 1987) of the early Miocene. A measured section at Slip Point (Addicott 1976b: tig. 4) shows that the rocks exposed there belong to the lower part of the Clallam Formation. The lower part of the Pillarian Molluscan Stage containing K. clallamensis includes an interval of time correlative with the late Arikareean North Ameri- can Mammal Age. Skull.— The nearly complete referred skull (LACM 131148) has parts of the left i:and I3 and both P:'s and M''s, but is missing the right zygoma and the anterolateral margin of the right premax- illa and maxilla, much of the ascending ramus of the right maxilla, the right nasal, the dorsal surface of the interorbital region, the sagittal crest, much of the roof of the braincase on the right side, and the lambdoidal crest (Figs. 1-4). Structures of the right orbit are fully exposed. The Early Miocene Littoral Cirsoid Carnivoran Kolponomos: Systematics and Mode of Life 13 The skull of Kolponomos clallamensis is roughly triangular in dorsal aspect, with a broad occipital region and a narrow snout (Figs. 5-7). The dorsal profile is arched, and the zygomatic arches are prominent. At its anterior extremity, the rostrum is thick and ventrally deflected. In anterior view, the narial opening is elongated dorsoventrally and tapered ventrally. A considerable thickness of the premaxillae anterior to the narial opening separates the incisors from the anterior margin of the naris. There is a rather prominent vertically oriented premaxillary protuberance. On either side of it. the anterior surface of the premaxilla slopes abruptly antero- ventrally. On either side in this area, the distal part of the root for I3 forms a bulge in the premaxilla. Otherwise, the bone surface in this area is depressed. The vertical premaxillary eminence extends posterodorsally and is continuous with the relatively narrow and sharp margins of the naris. Immediately anterior to the naris, the bone surface is rugose and punctured by many small foramina. The maxilla— premaxilla suture is fused and obliterated anteri- orly on both the holotype and the referred cranium, but on both specimens the sutures bounding the premaxilla lateral to the naris and the nasal bone are discernable. The ascending process of the premaxilla extends posteriorly to about mid-length on the nasal. It does not meet the anterior process of the frontal as in living bears but stops approximately 2 to 3 mm from the frontal. Along the lateral margin of the naris. the premaxilla forms a nearly vertical lateral surface. It is almost horizontal adjacent to the nasal bone, however. The lateral surface of the snout is dorsoventrally high and flat. Between the canine and the zygomatic arch it is concave. There is a nasolabials fossa on the dorsal part of the ascending ramus of the maxilla immediately anterior to the orbital margin. This fossa is broad and shallow, and is bordered anterodorsally by a slight protuberance, dorsally by a faint horizontal ridge, and poste- riorly by the orbital margin, which has a vertically elongated antorbital process. The infraorbital foramen has a large anterior opening. 18 to 21 mm high by 10 to 11 mm wide. The foramen opens into a broad fossa, strongly emarginated ventrally. and extending anteriorly nearly to the P2. The nasal bones are elongated, nearly parallel-sided, with rounded posterior borders. Where they join posteriorly they are not separated by the frontals. They slope anteroventrally and are nearly flat transversely. On the holotype they are approximately 74 mm long; on the referred cranium, although incomplete, the left nasal is 69 mm long as preserved. The nasals are narrowest just anterior to their mid-length and are wider both anteriorly and posteriorly. Their anterior margin is thick and slightly up-turned, especially near the sagittal line. They expand laterally at the anterior margin. The posterior part of the nasals is at the highest part of the cranium, which has a smooth, domelike profile. The low supraor- bital ridges of the frontals extend posterodorsomedially from just lateral to the posterior ends of the nasals. Posteriorly the interorbital constriction tapers uniformly toward the braincase, to its narrowest point in the intertemporal region. The top surface of the interorbital region on the referred cranium is weathered away, but on the holo- type cranium a low crest extends posteromedially from each su- praorbital process toward the sagittal plane. These crests lap anteri- orly onto the frontals, and where they merge posteriorly on the mid- line they form a slight V-shaped sulcus. Posteriorly from this point there is a low, broad sagittal ridge on the holotype, but the cranium is missing posterior to the intertemporal region. The mid-sagittal re- gion posterior to this area is broken on the referred cranium also. The braincase is elongated and tapers anteriorly toward the interorbital constriction. Its dorsal surface slopes laterally toward the temporal fossa and flares posterodorsolaterally to the nuchal crest, which is, however, almost entirely lost on the referred cranium. The surface of the bone is slightly undulating but not strongly rugose or pitted. The frontal-parietal suture ascends the anterolateral wall of the braincase from the back of the orbital region, approaches the midline, then bends posteriorly over the dorsal surface of the brain- case to extend posteromedially toward the midline. The squamosal fossa, forming the floor of the temporal fossa over the squamosal, is broad and shallow, does not slope anteriorly, and extends posteriorly into a broad sulcus in the lateral part of the nuchal crest. The occipital shield is in the shape of a broad isosceles triangle, with the apical part broken away. A broad median crest extends dorsally from the foramen magnum toward the nuchal crest. This crest is flanked by a pair of broad fossae. In turn, each fossa is flanked laterally by a broad eminence, preserved on the left side and in part on the right of the referred cranium, that extends dorso- lateral^ to merge with the posterior side of the nuchal crest. From this point, the nuchal crest becomes narrower ventrolaterally and curls posteriorly in the area posterior to the temporal fossa. Dorsal to each condyle is a prominent transversely oriented fossa that is continuous with a large fossa lying dorsal to the paroccipital pro- cess and below the nuchal crest. The paroccipital process curves posteroventrally almost as far ventrally as do the occipital condyles. It has a vertical posterior swelling that extends dorsally into the lateral fossa. The foramen magnum is wide and compressed dorso- ventrally. Its dorsal margin is a broad arch, only slightly peaked medially. The occipital condyles are relatively small, canted dorso- lateral^", with sharp edges around the lateral and ventral margins of the articular facets. The medial side of each is slightly excavated and has a small condyloid foramen. The condyles are separated ventrally by a broad U-shaped intercondylar notch. Their articular surfaces are not continuous ventrally but separated by a fossa that is a continuation of the intercondylar notch and aligned antero- posteriorly with the median ridge on the basioccipital. The condyles are positioned relatively ventrally in relationship to the basi- cranium. At the anterior margin of each condyle, immediately posterior to the hypoglossal foramen, is a recess in the margin of the articular surface. The orbit is of rather small diameter, measuring approximately 33 mm transversely in the referred cranium. At the anterior margin of the orbit, immediately ventral to the antorbital process, is a small (3 mm diameter) lacrimal foramen. The anterior part of the orbit has a convex medial wall that protrudes into the orbit in the area posterior to the infraorbital foramen. Immediately posterior to this area is a large (9 mm diameter), round sphenopalatine foramen that is imme- diately dorsal to and very close to the slightly smaller (6 mm) orbital aperture of the posterior palatine foramen. The tract for the optic nerve is elongated and relatively deep, leading posteroventrally from the ethmoidal to the optic foramen and into the orbital fissure; these foramina are approximately equal in size. The anterior lacerate foramen ( orbital fissure) and foramen rotundum lie in a large fossa as in pinnipeds, although a partition of bone separates them. As in other caniform carnivorans, the anterior aperture of the alisphenoid canal opens into the ventral wall of the anterior lacerate foramen and does not have a separate opening into the orbit. The zygomatic arch is stout and curves uniformly outward from the orbit. The zygomatic process of the squamosal increases only slightly in thickness posteriorly, and the posterior extremity of the jugal diminishes equally in diameter as it extends posteriorly. The zygomatic arch is not straight in its middle but curves both laterally and dorsally. At its anterior end. it is massive, forming thick dorsal and ventral margins around the infraorbital foramen. The ventral root of the zygomatic arch is very stout and descends vertically to form a buttress dorsal to the M'. This buttress forms a vertical component to the anterior end of the zygomatic arch that continues dorsally through the dorsal margin of the infraorbital foramen. 14 R. H. Tedford, L. G. Barnes, and C. E. Ray B t \ \ 5cm - 'W "■m** Figure 1 . Lateral views of the crania of species of Kolponomos. A, K. clallamensis Stirton, 1 460. referred, LACM 1311 48, left side; B, K. clallamensis, holotype, UCMP 50056, left and right sides; C, K. newportensis n. sp„ holotype, USNM 2 1 5070, right side reversed to left for comparison. All specimens to same scale. From this point, the zygomatic arch flares posterolateral^ to form the lateral margin of the orbit. The jugal bone extends anteromedially over the maxilla, forming a partially mortised joint. The jugal does not form the anterior rim of the orbit. Ventral to the orbit, the maxilla flares where it meets the jugal, and together they form a prominent masseteric process. This process projects ventrolateral^, and a crest extends posteriorly from it along the ventrolateral border of the jugal. The postorbital process of the jugal is stout, broad anteroposteriorly, and its apex is located anteri- orly. The anterior extremity of the zygomatic process of the squa- mosal abuts the posterior side of the postorbital process and is dorsoventrally expanded and slightly up-turned. From this point the zygomatic process curves uniformly posteroventrally to the glenoid fossa. At the glenoid fossa, the zygomatic process curves medially to form the dorsal surface of the glenoid fossa. The glenoid fossa is elongated transversely, narrow anteropos- teriorly. and has a smoothly curved articular surface. In contrast to that of typical Ursinae. the glenoid fossa is situated in a plane dorsal to the basioccipital plane. In ursines. the glenoid fossa is ventral to the plane of the basioccipital. As is typical of the Ursidae. the postglenoid process is well developed medially, projecting antero- ventrally to the glenoid fossa, and diminishes laterally. The postglenoid process is thin anteroposteriorly and does not form an anteroposteriorly thickened buttress as in the Ursinae. There is a low preglenoid process laterally. On the dorsal surface of the zygomatic process of the squamosal. The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematics and Mode of Life 15 Figure 2. Outline drawings of restored crania of Kolponomos species viewed from the left and oriented so that the hasicranial plane is horizontal. A. K. clallamensis Stirton. 1 960. referred. LACM 131 148; B. /C. clallamensis. holotype; UCMP 50056; C. K. newponensis n. sp„ holotype, USNM 2 1 5070 with tooth row restored. All drawings to same scale. Symbols for anatomical features: ac, alisphenoid canal (posterior aperture!; earn, external acoustic meatus; fio. infraorbital foramen; fl, lacrimal foramen; fla. anterior lacerate foramen; fo, foramen ovale; fop, optic foramen; Fr, frontal; fr, foramen rotundum; fs, sphenopalatine foramen; Ju. jugal; mp, mastoid process; Mx. maxilla; Na, nasal; nf. nasolabialis fossa; Pa. parietal; Pmx. premaxilla; pp. paroccipital ( = jugular) process; Sq. squamosal. 16 R. H. Tedford, L. G. Barnes, and C. E Ray V I B Figure 3. Ventral views of the crania of species of Kolponomos. A, K. clallamensis Stirton. 1 960. referred. LACM 1 3 1 1 48; B. A', clallamensis, holotype. UCMP 50056; C. K. newportensis n. sp.. holotype, USNM 21 5070. AM specimens to same scale. where the zygomatic arch meets the squamosal fossa, there is a prominent tuberosity. This tuberosity also is at the anterolateral edge of the shelf dorsal to the external auditory meatus. This shelf slopes posteroventrally and expands dorsoventrally as it merges with the base of the mastoid process. The palate is elongated and concave both anteroposteriorly and transversely. On either side of the midline the palate is subplanar, essentially flat transversely and gently arched anteroposteriorly. from the incisive foramina to the palatal notch. Anteriorly and laterally the palate descends abruptly to the inner margin of the dental arcade. The anterior end of the palate is deflected ventrally. so that the alveolar margins of the canines and incisors are posi- tioned more ventrally than those of the cheek teeth. The septum separating the incisive foramina is continuous pos- teriorly with a slight raised ridge extending more than 30 mm posteriorly along the midline suture of the palate. These foramina are large and reniform. On either side of the palate, at the posterolat- eral corner just medial to the M: alveoli, are the posterior palatine foramina, closely associated with the maxillo-palatine suture. These foramina are variable in size and number. All are large on the holotype. On the referred cranium, the anterior foramina are both of intermediate size, while the posterior foramina are of different 17 5cm occ ten Figure 4. Outline drawings of restored crania of Kolponomos species, viewed ventrally. A, K. clallamensis Stirton, 1960. referred, LACM 131148; B./C. clallamensis, holotype, UCMP 50056; C, K. newportensis n. sp., holotype, USNM 215070, with tooth row restored. All drawings to same scale. Symbols for anatomical features: Bo, basioccipital; Bs, basisphenoid; cc. carotid canal; earn, external acoustic meatus; fh. hypoglossal foramen; fi. incisive foramen ( = palatine fissure); flp. posterior lacerate foramen; fpal, palatine foramen; fsm. stylomastoid foramen; hf, tympanohyal pit (= hyoid fossa); mp, mastoid process; Mx, maxilla; occ, occipital condyle; Pal, palatine; pp. paroccipital (= jugular) process; Ps, presphenoid; Pt, pterygoid; tec, ectotympanic; ten, entotympanic. sizes, intermediate on the left and small on the right. On both crania the anterior foramen is continuous with a prominent antero- posteriorly elongated sulcus that extends anteriorly to a point where it disappears medial to the P\ Posterior to the palatine foramina and the M2 the posterolateral palatal margin is formed by a narrow vertically oriented crest pierced by a foramen. This crest is continu- ous posteriorly with the sharply keeled ventral border of the ptery- goid hamulus. On the referred cranium, the hamulus is very thin transversely and concave laterally, as is typical of the Ursinae, but is bent sharply ventrally. The narrow posterior process of the ptery- goid hamulus extends posteriorly. The main part of the pterygoid- palatine strut ascends posteriorly, to join the basicranium around the posterior aperture of the alisphenoid canal. The lateral surface of the strut is rounded and convex and continues onto the ventrolat- 18 R. H. Tedford. L. G. Barnes, and C. E. Ray r: w ■*» B The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematica and Mode of Life 19 - B B 5cm 5 cm Figure 6. Anterior views of the crania of Kolponomos species. A, K. clallamensis Stirton, 1960, referred, LACM 131148; B, K. clallamensis, holotype, UCMP 50056; C, K. newportensis n. sp., holotype, USNM 215070. All specimens to same scale. eral surface of the braincase wall. From the pterygoid hamulus, a fine ridge extends posteriorly along the medial side of the posterior aperture of the alisphenoid canal and the foramen ovale and contin- ues into the auditory tube of the bulla. The internal narial opening is highly arched and wide. The palatal notch is broadly rounded in the referred specimen, has a slightly acute apex in the holotype, and extends anteriorly almost to a point between the centers of the M:'s. On either side of the internal narial opening, the palatine-pterygoid struts sweep medi- ally to form a sharp, underhanging border. The roof of the internal narial opening ascends anteriorly and in its anterior part has a medial keel formed by the vomer and presphenoid. The pre- sphenoid-basisphenoid suture is transversely oriented and is not Figure 7. Posterior views of the crania of Kolponomos species. A, K. clallamensis Stirton, 1960. referred, LACM 131148; B, K. clallamensis, holotype, UCMP 50056; C, K. newportensis n. sp., holotype, USNM 215070. All specimens to same scale. coossified. The basisphenoid is nearly flat where it forms the roof of the internal naris between the pterygoid hamulae. It expands posteriorly and at its lateral edge, dorsal to the pterygoid hamulus, bears a groove that extends posterolaterally into the median lacerate foramen. The median lacerate foramen is elongated antero- posteriorly and is situated at the anteromedial corner of the bulla. The basioccipital-basisphenoid suture is fused, and its precise location is not visible. The basioccipital has a median crest that widens posteriorly and spreads toward each condyle. On either side are a curved fossa and a rugosity that mark the insertion of the rectus capitis ventralis muscles. Posterolateral to the fossa, between the condyle and the bulla, is the hypoglossal foramen, which is transversely oval and approximately 3 mm in diameter. The tympanic bulla is small and has a rugose ventral surface. It is fused laterally to the squamosal and the base of the mastoid Figure 5. Dorsal views of the crania of Kolponomos species. A, K. clallamensis Stirton, 1960, referred, LACM 131 148; B. K. clallamensis. holotype, UCMP 50056; C, K. newportensis n. sp., holotype. USNM 215070. All specimens to same scale. 20 R. H. Tedford. L. G. Barnes, and C. E. Ray process. Posteromedially it is separated from the basioccipital by a sulcus. Anterolaterally it expands posleroventral to the medial part of the glenoid fossa and is broadly appressed to the posteromedial part of the postglenoid process. An oblique crest on the ventral surface of the bulla that extends from the stylomastoid foramen to the anteromedial margin appears to mark the junction between the entotympanic and the ectotympanic. If this is true, the entotympanic contribution to the tympanic bulla is approximately equal to that of the ectotympanic. There appears to be a small postglenoid foramen located in a fissure where the medial part of the postglenoid process is over- lapped by the bulla. The ventral surface of the bulla is retracted posteriorly at the anteromedial corner ventral to the median lacerate foramen. The posterior lacerate foramen lies at the posteromedial corner of the bulla and is semicircular and positioned obliquely. The external auditory meatus is round, approximately 4 mm in diameter, and recessed far beneath a wide shelf formed by the squamosal. The mastoid process is very long, extending outward from the cranium variably 38 and 44 mm on either side, measured from the notch where it joins the paroccipital process. It projects anteroventrolaterally from the basicranium. The process is basi- cally three-sided; one flattened surface medial, one anterior, and one posterior. The medial surface is concave in contrast to the other two, which are slightly convex. At its distal end, the mastoid process is compressed anteroposteriorly. The concave medial surface expands proximally toward the basicranium and is confluent with a large recess surrounding the hyoid fossa. This same recess extends onto the anterolateral surface of the paroccipital process, which is at this place deeply excavated. The paroccipital process projects posteroventrolaterally from the basicranium and is com- pressed transversely. Its anteroventral margin is a crest that extends anteromedially toward the bulla, reaching the posterior side of the bulla between the stylomastoid foramen and the posterior lacerate foramen. As in the Ursinae. the hyoid fossa is separated from the poste- rior lacerate foramen by a ridge of bone. The hyoid fossa sits within a deep recess. In ursines, the hyoid fossa is widely separated from the external auditory meatus by a wide expanse of the bulla. In Kolponomos clallamensis, however, the hyoid fossa is very close to the external auditory meatus. Also, in K. clallamensis the stylomastoid foramen lies midway between the hyoid fossa and the external auditory meatus, whereas in the Ursinae the stylomastoid foramen is widely separated from the external acoustic meatus and is within the recess that houses the hyoid fossa. Dentition. — The upper dentition consists of l'~\ canine, P1 4, and M1 2. The actual teeth present in the referred cranium are the roots of the left I2"-' and the complete left and right P:'s and M''s. On the left side, all alveolar margins are preserved, and it is that side that forms the basis for the following description. The incisors and canines are clustered, without significant di- astemata, in the thickened and downturned anterior end of the snout. The incisors are aligned transversely anterior to the canines. I1 and I2 are small and have transversely compressed roots. I1 is smaller than I2, and both teeth are implanted essentially vertically in the palate. The I^'s are much larger, being approximately four times the diameter of I2 at the alveolar rim. Unlike the medial incisors, the P's are procumbent and deeply rooted in the premaxilla between the canine and the incisive foramen. The left I1 measures 18.2 mm anteroposteriorly and 12.4 mm transversely at the alveolar rim. A diastema of 4 mm separates the alveolus for the left I1 from that of the upper canine. The canine alveolus is oval and measures 20 mm anteroposteriorly and 17 mm transversely at the alveolar rim. The bulge in the lateral side of the rostrum indicates that the root for the canine is extremely long and extends nearly as far into the rostrum as the lateral edge of the nasal bone. The root is procumbent. The cheek-tooth row curves laterally from the canine posteri- orly to M'. then M2 is positioned more medially. P1 has a single root that is round in cross-section, procumbent, and closely appressed to the posterior side of the canine alveolus. The alveolus indicates that the root was tapered, approximately 7 mm in diameter at the alveo- lar margin, and extended for approximately 1 5 mm into the maxilla. P2 is a robust tooth, with two roots and a large smooth crown. The tooth is oriented obliquely to the axis of the cheek-tooth row. the anterior root medial to the axis. The posterior root is on the axis of the tooth row and is approximately twice the diameter of the anterior root. The tooth tilts medially into the oral cavity. The crown of this tooth has a flat apical wear facet on the principal cusp. A smooth cingulum borders the lingual side of the crown from the anterior to the posterior border of the tooth. The posterior part of the crown is formed into a talon lying between the principal cusp and the posterior end of the cingulum. P1 had roots aligned in the axis of the cheek-tooth row. Judged by the size of the alveoli, the two roots of this tooth were more nearly equal in size than those of P2, the posterior one being only slightly larger in diameter than the anterior one. The maxilla projects ventrally, forming a crest of bone between the two roots of this tooth. P4 was a large tooth, nearly as large as M1. and had three roots. Of these, the medial (protocone) root is the largest, being more than twice the depth and diameter of either of the lateral (paracone and metacone) roots. Between P3 and P4, the lateral margin of the maxilla begins a strong lateral bend, so that P4 is oriented obliquely to the cheek-tooth row. A diastema of approximately 3 mm sepa- rates the anterior (paracone) root of P4 from the posterior root of P\ M1 is a massive tooth. It appears to have been approximately 30% larger than P4 in surface area. It is greatly expanded trans- versely and is triangular in occlusal view. It has five mammiform cusps of nearly equal sizes. The lateral two cusps are the paracone and metacone; the most medial cusp forms the apex of the triangle and is the protocone; intermediate cusps are the para- and metaconules. The surface of each M' shows extensive wear that breaks through the enamel to expose the inner dentinal core. There is a labial cingulum between the para- and metacones, but other cingula are lacking. The M2 alveoli indicate that this tooth was approximately half the size of the M1. M2 is positioned lingually opposite the talon of M'. Like both P4 and M1, M2 had three roots. Of these, the anterolateral (paracone) root was broadly joined with the medial (protocone) root to form a transversely oriented bilobate root. The posterior root is the metacone root and is rotated so that it is actually the most medial root of the tooth. Kolponomos newportensis, new species Figures 1-12 Kolponomos clallamensis Stilton, 1960. Barnes etal.. 1985:43, figs. 9a. b. Diagnosis. — A species of Kolponomos differing from K. clalla- mensis in the following derived cranial features: broad muzzle flaring laterally above canines and incisors; mastoid process twisted clockwise, extending forward beneath external auditory meatus as far as postglenoid process; intercondylar notch lacking so that the articular surfaces of the occipital condyles are continuous ventrally. In addition. K. newportensis is distinguished from K. clallamensis by the following primitive features: infraorbital foramen smaller and lacking marked excavation of maxilla anterior to it; prominent masseteric process of maxilla lacking and masseteric process on jugal reduced; jugal forming anterior rim of orbit; paroccipital process smaller and less downwardly pointing; paroccipital process lacking hyoid fossa. The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematic* and Mode of Lite 21 Holotype. — USNM 215070, cranium lacking parts of dorsal surface with only right and left P: in situ; mandible lacking tips of coronoid processes and lacking incisors and right P,; isolated asso- ciated right I3 and C. right P1. left P1, left P\ right P4. left M'. and right and left M:. Recovered from the same concretion were the axis, third cervical vertebra, a broken proximal lumbar vertebra, a sternebra, a proximal part of an anterior rib, a thyrohyal lacking the proximal end, a complete '.'ceratohyal, a metapodial lacking the distal end and half of the proximal end, a median phalanx, and unidentifiable bone fragments. The original half of the concretion containing the occipital part of the skull and most postcranial elements was collected by Douglas R. Emlong, October 1969 (Emlong field no. 603). On 26 January 1976 Emlong found the remainder of the concretion, containing the balance of the skull, mandible, and isolated teeth (Emlong field no. E76-B). recognizing it as associated and pertaining to Kolponomos. Type locality. — A concretion found on the beach at low tide line, approximately 300 yards (274 m) south of the mouth of Big Creek and 100 yards (91 m) seaward of the sea cliff just north of Newport, Lincoln County. Oregon. Formation and age. — Lower part of the Nye Mudstone, repre- senting the early Pillarian Molluscan Stage ( Addicott 1976a; Moore and Addicott 1987). correlative with the Late Arikareean Land Mammal Age. early Miocene. Etymology. — Named for the town near the type locality to record the occurrence of the type species in a manner similar to that for the genoholotype. Skull.— The skull of USNM 215070 lacks most of the dorsal surface and is toothless except for both P:'s. which are crushed into their alveoli and forward into those for P1. The concretion enclosing the specimen was subspherical and aproximately 27 cm in diameter. Prior to its consolidation, all upper teeth except left and right P: had fallen out. In the course of gross preparation these were found in a tight cluster beneath the palate and between the horizontal rami of the mandible. Also prior to consolidation, the mandible had slipped out of articulation and moved forward and upward forcibly, coming to rest in a symmetrical undershot false occlusal position, undoubt- edly causing the anterior displacement of left and right P2 and severely crushing and comminuting the thin alveolar walls of left and right M: and M1 and, to a lesser extent, the alveolar margins of the upper premolars. The apices of the coronoid processes were removed by abrasion of the smooth surface of the concretion as was much of the dorsal surface of the skull. The tightly appressed mandible was painstakingly separated from the skull and the tightly clustered isolated teeth were extracted by Gladwyn B. Sullivan in 1976 in the course of gross preparation of the specimen. This specimen represents an old individual as judged by its heavily worn teeth and advanced cranial fusion. Despite the latter, it is possible to make out many sutures, particularly in the orbital region. A striking major feature of the skull, in common with all re- mains of the genus, is the flexure of the facial part of the skull relative to the basicranial plane (Fig. 2). Measured as the angle between the palate and basisphenoid, the flexure is approximately 155° in USNM 215070. The widely flaring zygomatic arches, the forward-oriented orbits, and the great hypertrophy of the mastoid processes are also distinctive features of the remarkable skulls of Kolponomos, In USNM 2 1 5070 the muzzle is nearly as broad at the carnassial as the palate. The large anteroposteriorly elongated incisive fo- ramina lie in the trough of the strongly arched palate with distinct grooves extending anteriorly from them nearly to the incisor al- veoli. The interforamen septum forms a low S-curve anteriorly. Posteriorly the vaulted palate has strong anteroposteriorly oriented depressions along either side of the flattened medial part of the palate, becoming progressively deeper posteriorly and leading into the anterior palatine foramina at the maxillary-palatine suture adja- cent to the anterior root of M:. Behind that a series of pits extends the posterior palatine groove to a foramen that penetrates the thin rim of the palatine portion of the palate. The pterygoid hamuli are arcuate in palatal view and terminate in dorsoventrally flattened processes. Sutures with bones surround- ing the pterygoid are too coossified for the precise outline of this element to be determined. The anterior or pterygoid process of the alisphenoid is defined by its suture with the palatine; with the palatine it forms a strong strut bracing the back of the palate against the braincase. Ventral to this strut a depression for the origin of the pterygoid muscle extends downward toward the hamular process. The posterior end of the large alisphenoid canal penetrates the base of the strut. This opening is closely followed by the foramen ovale, which lies in a common pit with the opening of the canal on the left side of USNM 215070, but on the right a groove joins these orifices as in most arctoids. The dorsal and posterior sutures of the alisphenoid with adjacent bones are closed. The basisphenoid and basioccipital bones are strongly coossified and the sutures between them are eliminated. They form a trapezoidal figure with its base lying across the rectus capitis insertions just anterior to the posterior lacerate foramina. Thus the basioccipital is broadest across the rectus insertions where the winglike lateral processes of this bone overlap the medial edge of the petrosal and presumably floor the large tract for the inferior petrosal sinus. The knoblike processes for the rectus are situated bilaterally at the posterolateral corners of ovoid shallow depres- sions for muscle insertion that presumably extend forward onto the basisphenoid and medially to a low crest at the midline. The exoccipital is solidly fused with surrounding elements, except for its irregular contact with the posterior end of the bulla (caudal entotympanic). This bone contains the hypoglossal foramen, which is situated posteromedial to the posterior lacerate foramen as well as the posterior rim of the latter opening. Presumably the exoccipital also forms the medial wall and spine of the paroccipital process. The occipital condyles protrude slightly posterior to the nuchal crest. There is no intercondylar notch as the condyles are conjoined ventrally, uniting their articular surfaces. A shelflike extension of the floor of the foramen magnum extends posteromedially beyond these articular surfaces. Inside the foramen magnum the paired posterior openings of the hypoglossal foramina can be seen on its floor. An additional pair of foramina lying on the lateral wall of the foramen magnum at the level of the dorsal part of the condyles presumably accommodated venous drainage for sinuses within the occiput. The auditory region is very small and nestled deeply within the ventrally projecting elements surrounding it, particularly the greatly hypertrophied mastoid process. The bulla is uninflated and exten- sively coossified with surrounding elements; nevertheless, most of its outline as well as its composition can be determined from bilaterally symmetrical suture traces and rugose coossification tracts. These observations indicate that the ectotympanic lacks an ossified meatal tube; its anterior limb spreads over the postero- medial surface of the postglenoid. forming the posterior wall of the slitlike postglenoid foramen. Anteromedially the ectotympanic overlaps the entotympanic and coossifies laterally with the alisphenoid behind the foramen ovale. A styloid process of the ectotympanic lies beneath the opening for the eustachian tube. The posterior limb of the ectotympanic is fused to the base of the mastoid process. Behind this union, the stylomastoid foramen emerges from the mastoid. The facial canal is thus separated from the large pit for the tympanohyal that opens above a prominent hyoid process of the entotympanic at the posterolateral corner of the bulla. There is no large hyoid fossa excavated into the anterior wall of the paroccipital process as in K. clallamensis. 22 R. H. Tedford, L. G. Barnes, and C. E. Ray The entotympanic is irregularly exposed along the medial edge of the bulla because of variable overlap of the ectotympanic; poste- riorly the caudal element is sutured to a process from the exoccipital and posterolateral^ to a process from the mastoid. There is a large posterior opening of the carotid canal well anterior to the posterior lacerate foramen. This opening is formed medially by the basioc- cipital wing and laterally by the entotympanic, but anteriorly the arterial tube is nearly completely surrounded by the entotympanic. The anterior carotid foramen lies medial to the styloid process of the ectotympanic and opens into a groove in the basisphenoid anterior to the median lacerate foramen. From this groove the artery must loop posteriorly to enter the median lacerate foramen and/or the presumed channel in the basioccipital that accommodates the inferior petrosal sinus. The large mastoid process takes the form of an anteroposteriorly flattened column bending outward and downward from its base and curving forward at its tip to pass nearly under the postglenoid process (Fig. 7C). Its components are totally coossified. but the nuchal crest extends along the lateral surface of the process, thus marking the position of the mastoid-squamosal suture and indicat- ing that the process is composed about equally of the two elements. The process terminates in a raised ovoid area that lies within the suture. This element may represent the secondary ossification cen- ter (epiphysis) frequently observed at the tip of the mastoid process in adult ursids. A ridge arises from the posteroproximal surface of the mastoid process and passes upward and posteriorly to join the paroccipital process. The paroccipital process curves postero- ventrally and terminates in a sharp point. The supraoccipital bones are solidly coossified with surround- ing elements. They are concave and highly rugose beneath the nuchal crest; a thin ridge is present sagittally. At their lateral ex- tremities a shallow pit is present dorsal to the base of the paroccipital processes. The lateral processes of the nuchal crest extend behind the inion and beyond the occipital condyles; conse- quently, the crest has a broad inflection at the midline. The parietal and squamosal bones are coossified. but their junction is probably marked by the bilaterally symmetrical collapse of the braincase wall under lithostatic load. Breakage dorsally has removed most of the sagittal crest of the parietal, but the remaining evidence indi- cates the presence of at least a low crest. The squamosal apparently makes a significant contribution to the anterior part of the mastoid process. Its glenoid fossa forms a cylindrical articulation nearly at right angles to the basicranial axis. Prominent recurved postglenoid processes are present, deepest medially, and the anterolateral part of the articular surface is bordered by a low preglenoid process. Most of the squamosal-jugal suture is visible; the squamosal contribution to the zygomatic arch seems to taper out at the base of the jugal postorbital process. The anterior ends of the frontals have been removed by erosion, exposing a natural section of the narial cavity. Frontal-parietal sutures are coossified and not traceable, but they may have crossed the midline at about the point where the parasagittal crests appear to diverge anteriorly. The frontal sinuses seem to have extended back- ward over the braincase to about this point. Beneath these sinuses the braincase is sharply constricted anteriorly, probably indicating the greatly constricted form of the olfactory lobes of the brain as shown by the holotype of K. clallamensis. Sutures in the orbital wall can be partially seen and the relative position of foramina and bones can be determined. In general the arrangement is like that described by Stirton for the holotype of the genotypic species. There is a large common opening for the anterior orifice of the alisphenoid canal, the foramen rotundum, and the anterior lacerate foramen as in pinnipeds. Anterodorsal to this opening the slitlike small optic foramen opens into a short groove. Dorsal to the Table 1. Measurements (in mm) of crania of Kolponomos clallamensis and A', newportensis, new species. Total (condylobasal) length Postpalatal length (palatal notch to basion) Basion to anterior edge of zygomatic root Length C alveolus to M: alveolus Width of rostrum across canines Width of skull across alveoli of M1 Width of skull at infraorbital foramen Width of skull across antorbital process Width of greatest intertemporal constriction Width of braincase, anterior edge of glenoid fossa Zygomatic width Auditory width Mastoid width Paroccipital width Greatest width of occipital condyle Greatest width, anterior nares Greatest height, anterior nares Greatest width of foramen magnum Greatest height of foramen magnum Transverse diameter of infraorbital foramen Height of infraorbital foramen Kolponomos Kolponomos newportensis clallamensis USNM UCMP LACM 2 1 5070 50056" 131 148'' 253.1 258.4 107.3 — 107.0 155.2 — 153.7 114.2 ca. 108 117.7 75.2 — (71) ca. no ca. 98 119.6 78.3 75.2 74.3 74.9 80.9 (82) 46.2 46.8 45.4 72.3 — 82.3 179.5 — (178) 137.0 — 141.2 182.0 — 183.7 116.5 — 123.0 61.5 — 74.0 — 37.0 38.5 — 36.3 — 26.8 — 33.3 18.3 — 15.0 11.1 11.6 13.0 11.3 20.0 21.6 "For additional measurements see Stirton ( 1960:355). Bilateral measurements of the referred cranium of K. clallamensis are made on the left side. Parentheses indicate estimated measurements made hy doubling a half width. The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematics and Mode of Life 23 Table 2. Measurements of cheek teeth and mandible of Kolponomos. Where available, measurements of the left side en- tered first. K. newportensis K. clallamensis (USNM 215070) (LACM 131148) Upper teeth (length x width) I' 12.8 x 10.7 — C 15.4 x 13.9 — P1 7.3 x 11.5 — P: 16.2x — 17.2 x 12.5 15.7 x 10.3 P' ca. 16 x 12 — P4 18.5x25.5 — M1 22.3 x 28.9 21.9x31.4 M: 16.8 x 16.4 17.8 x 17.1 — Lower teeth (length x width) C 17.2 x 13.1 16.5 x 13.5 — P, — x ca. 9 7.6x9.9 — P2 13.7 x 10.8 — 14.0 x 10.5 p, 16.1 x 11.2 16.2 x ll.l — P4 18.0 x 14.0 18.7 x 13.6 — M, length 27.3: 27.5 — M, width trigonid 24.0; 24.5 — M, width talonid — ; 22.7 — M, 21.5x22.3 — 21.3x — Mandible Length of horizontal ramus. condyle to tip of rostrum 201.5 — Length of base of coronoid process 55.5; 55.7 — Depth of mandible beneath P, 59.5; 59.8 — Depth of mandible beneath anterior part of M, 43.8; 44.2 — Depth of mandible beneath posterior part of M? 40.0; 42.2 — Width of mandible atP, 18.8; 20.5 — Width of mandible at M, 21.8; 24.2 — Length of symphysis 53.5 — optic foramen lies the equal-sized ethmoid foramen. The orbitosphenoid bone containing these foramina seems to pass anterodorsally along the orbital wall to a greater extent than in K. clallamensis. Ventral to these foramina, the sphenopalatine and poste- rior palatine foramina are closely associated, the former on the orbitosphenoid-palatine suture and the latter penetrating the adjacent palatine bone. The sphenopalatine foramen is about twice the size of the posterior palatine foramen; neither is as large as its counterpart in K. clallamensis. Indistinct sutures suggest that the palatine is attenu- ated between the frontal and maxilla and fails to reach the anterior end of the orbital fossa. There is a very small lacrimal foramen but no trace of the limits of the corresponding lacrimal bone. The jugal forms the ventral part of the orbit and appears to extend nearly to the lacrimal foramen (in contrast to K. clallamensis). With the maxilla it forms the roof of the short infraorbital canal. There is a strong postorbital process and a low, elongated, rugose masseteric process formed by the jugal. The dorsal part of the maxilla is eroded away so that only the lateral and palatal parts of this bone are preserved. The oval infraorbital canal penetrates the zygomatic process of the maxilla. This canal is large (see Table 1) and short (11.0 mm left), as is characteristic of Kolponomos and pinnipeds. It opens onto the face into a shallow depression (infraorbital fossa ) that is more extensively developed in K. clallamensis. There appears to have been a shallow fossa for the levator labii marked by a dorsal preorbital fossa as in A'. clallamensis. The suture with the premaxilla is thoroughly coossified. The rostrum anterior to the canines is laterally expanded and bears low rises over the roots of the canine and I3. A natural oblique section through the narial cavity exposes the dorsal part of this region from the level of the canines to just behind the orbits, showing that the ethmoturbinals are placed laterally and are excluded from the narial opening as in other arctoids. The dorsal frontal sinus is also evident and must extend backward close to the frontoparietal suture. Upper dentition. — Most of the upper cheek teeth of USNM 215070 (Fig. 8). with the exception of the in situ P2's, were found grouped together in the matrix between the rami of the associated mandible. When prepared and fitted into their respective alveoli the following teeth were recovered: right l\ C, right P1, left P1, left P\ right P4, left M', and right and left M:. All except I' and C are highly worn and so furnish little information about their crowns' mor- phology. The alveoli for I'~: indicate that these teeth were much smaller than I3. I3 is procumbent and has a long cylindrical root and a relatively short crown with a recurved tip. There is no cingulum or carina. A facet for the lower canine is present, worn through the enamel at the posterolateral base of the crown. The tip has been fractured and a transverse groove has been cut across the anterior face of the crown, both probably representing damage during use of this tooth as a lever. The upper canine is broken across the root so that only about half the total length of the tooth remains. It is also procumbent; its crown is an attenuated cone, slightly recurved at the tip, and has no cingulum or carina. A wear facet for the lower canine extends from the tip to within 2 mm of the base of the crown. The tip is also worn apically. and a short transverse groove cuts the anterior face near the tip. The P1 is closely appressed to the canine. Its crown is oriented transverse to the axis of the tooth row. The crown has an ovoid outline and a posterior cingulum. Heavy wear has truncated the apex to a medially sloping wear surface cut to the crown's base. Its single root is anteroposteriorly flattened and bends posteriorly to accommodate the roots of C and P2. The P: has two roots; its crown is an elongated oval in occlusal outline with cingular shelves ante- rior and posterior to the stout principal cusp. Wear has formed a medially truncated surface across the principal cusp that has cut nearly to the crown's base. P3 is only slightly larger than P2 and two- rooted, the more anterior root passing inside the posterior root of P2 so that the tooth has an orientation oblique to the tooth row. The crown is ovoid in occlusal outline; wear has removed the principal cusp and anterior cingulum. but the posterior cingular shelf re- mains. This cingulum bears a low cuspule laterally and is bounded medially by a well-developed facet of interdental wear. P4 bears three roots, the lateral pair about the size of those of the anterior premolars, although the most posterior is slightly smaller than the anterior. A large medial root is nearly symmetrically placed with the lateral roots, but the whole tooth has an oblique orientation to align with the anterior surface of M', thus removing the embra- sure pit found in most carnivores where carnassial shear is impor- tant. At the advanced stage of wear shown by the P4 only an encircling band of thin enamel remains. The enamel is broken on the posterior side by abrasion between P4 and the adjacent M1. A remnant of the metastyle remains on the posterolateral part of the crown marked by a notch that may represent the base of the carnas- sial notch. Dentine-filled pulp cavities on the gently concave worn crown indicate the presence of a principal external cusp or coa- lesced cusps forming an anteroposteriorly elongated structure and large internal cusp ("protocone") supported by the strong internal 24 R. H. Tedford. L G Bames. and C. E. Rav 5cm Figure 8. Upper cheek teeth of holotype of Kolponomos newportensis n. sp„ USNM 215070. Right side with left P:. P3, and M1"2 reversed to restore the complete cheek-tooth series. root. Anterolateral to the latter a filled pulp cavity indicates the presence of another smaller cusp linked to the "protocone," prob- ably indicating that the anterior cingulum bore a cusp analogous to the paraconule of tribosphenic molars. Thus the upper carnassial of USNM 215070 has been molarized, its sectorial nature changed to function with the molars as part of the masticatory battery. The first upper molar has three roots. The labial roots support- ing the paracone and metacone are anteroposteriorly compressed structures; the lingual root supporting the protocone is a short faceted cone. Filled pulp cavities indicate a pattern of cusps similar to the M1 of A-, clallamensis. A short section of the labial cingulum bridges the indentation between the paracone and metacone and. as in K. clallamensis, indicates that the paracone was larger than the metacone. Thin enamel rims the concave wear surface of this tooth, broken only at the junction with P4. The M: is triangular in form and least worn on the left side. It is positioned lingually opposite the M1 talon, in contrast to a labial position opposite the trigon as in most carnivorans. It bears two short stout roots; the anterior is anteropos- teriorly flattened, the posterior triangular. The worn crown shows four inflated cusps that have coalesced, separated only by thin grooves similar to the condition in the less worn M' of K. clallamensis. These are interpreted as follows: the most labial is the paracone with the closely allied metacone immediately postero- lingual to it; the large protocone occupies the anterolingual border. Figure 9. Mandible of Kolponomos newportensis n. sp., holotype, USNM 215070. A, occlusal view; B, left side. At same scale as figures of crania. the labial projection of its wear facet representing the paraconule; the cusp at the posterolingual corner of the tooth is the metaconule. The enamel covering of these cusps is remarkably thin, as revealed by apical wear. An interdental wear facet with M1 occurs on the anterior face of the tooth. Mandible. — The nearly complete mandible of 215070 lacks only the tips of both coronoid processes (Figs. 9, 10). The rami are thoroughly ankylosed at the symphysis; the junction is raised exter- nally, creating a symphyseal boss ventrally. The horizontal ramus is deepest at this boss and shallowest posteriorly. Anteroventrally paired foramina lie on either side of the symphyseal suture about at mid-depth. Laterally there are three mental foramina; the largest is the most ventral, lying beneath P, at the posterior end of a shallow depression. A second foramen lies anterodorsal to the first and beneath P, on the right side or posteriorly beneath the anterior root of P4 on the left side. The third and most posterior foramen lies at mid-depth of the horizontal ramus beneath the posterior root of P4. The rami are markedly rugose beneath the molars; the right side shows bone resorption around the protoconid root ofM,, and on the left side there is a pit in the lateral surface adjacent to the hypoconid root of the same tooth. The masseteric fossa has a deep anteroposte- riorly elongated pit in its deepest recess. The masseteric crest does The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematic* and Mode of Life 25 Figure 1 0. Outline drawing of the holotype of Kolponomos newportensis n. sp., USNM 215070, viewed from the left. Upper dentition restored from isolated teeth found with the type. not extend to the anterior end of the fossa but arises just above the base of the angular process and passes to the articular condyle. In harmony with the form of the glenoid fossa, the articular condyle is cylindrical, pointed laterally, and deepest medially; a pit for inser- tion of the external pterygoid muscle occurs at the anteromedial base of the condyle. The angular process is relatively small and markedly inflected medially. Its dorsal surface has a pit and ridge, and the medial surface is rugose, all for insertion of parts of the internal pterygoid muscle. There is a large mandibular foramen that lies below the level of the tooth row and the condyle and just above the level of the dorsal surface of the angular process, about midway along the base of the ascending ramus. Lower dentition. — The central incisors seem to have been lost in life in USNM 215070 but it is not certain that I, was in fact present. The position of this tooth is occupied by spongy bone. The alveolus for the right I2 is filled with spongy bone but the alveolus for the left 1, is discernable. The roots of both I,'s are present. The canine has a long root and short crown. It is fully preserved only on the left side; the right canine appears to have had its apex broken away in life; the broken surface is polished and a secondary wear facet was established on the medial side of the broken tip. Occlu- sion with the upper canine has worn the tip and posterolateral side of each lower canine. The premolar row diverges slightly posteriorly along an axis that lies wholly inside the axis of the molar row. The left P, is badly broken; the right P, was recovered from the surrounding matrix. Like its counterpart in the maxilla, the P, is anteroposteriorly compressed to fit between the base of the canine and the overhang- ing anterior margin of P:. Its crown is heavily worn apically on a posteriorly slanting surface. There is a posterior cingulum pitted by breakage. The P, and P, are similar in form; the P, is larger. The apically worn crowns show a robust principal cusp and anterior and posterior cingular cusps connected by a lingual cingulum. There is no labial cingulum. The P4 departs from this form in that a strong posterolingual cusp is also present; the anterior cingular cusp and / -*%a_ F+^r B i i Figure 11. Cervical vertebrae of holotype of Kolponomos newportensis n. sp., USNM 215070. A, axis, anterior and left lateral views, lateral view reversed from right side; B, third cervical, anterior and left lateral views. posterolabial cingular shelf combine to give this tooth a molariform appearance. The M, is offset laterally so that only the anterolingual face of its paraconid overhangs the posterolabial cingulum of P4; its crown is only a little longer than wide and worn nearly flat. On the left side the entoconid region is worn away, but on the right the crown is complete with its encircling thin enamel. A remnant of the enamel in the right talonid basin is preserved at this advanced stage of wear. Filled pulp cavities indicate the full tribosphenic complement of cusps; the protoconid, metaconid, and hypoconid were subequal in size, the paraconid and entoconid smaller. There are indentations internally at the carnassial notch and externally between the talonid and trigonid. There is no trace of a cingulum. M, is wider than long, triangular in occlusal outline, and widest across the trigonid. Wear has removed the anterolingual comer of the right M, but not the left, which has lost marginal enamel along the anterior part of the tooth. Filled pulp cavities indicate the presence of three trigonid cusps and the hypoconid, which is repre- sented by encircling enamel on the right and has not worn into the pulp cavity on the left. The pulp cavities seem best interpreted in comparison with the trigonid of M,: a large protoconid on the anterolabial corner, a small paraconid directly lingual to it. and a large metaconid on the median lingual margin. Pulp cavities of the latter two cusps are connected. There is no M3. Postcranial .skeleton. — An axis and third cervical vertebra are available and all structures are preserved on one side or the other of these bones. The axis (Fig. 1 1 ) has a short centrum and high neural spine as in pinnipeds and lutrine mustelids. The general form of the bone most resembles that seen in phoeids or Enhydra, although the neural spine is larger overall than in the latter. The odontoid process points anterodorsally as in terrestrial carnivores and in contrast to most pinnipeds, and there is a shallow groove on its dorsal surface for the transverse ligament of the atlas, a feature usually missing in pinnipeds. A broad ridge continues the odontoid process into the spinal canal. The atlantoaxial articulations are joined beneath the odontoid process as in ursids, not separated by a notch as in pinni- peds. The centrum (less the odontoid process) is wider than long. Its 26 R. H. Tedford. L. G. Barnes, and C. E. Ray It I? ' /•'-' B r 44 Figure 12. Metapodial and phalanx of holotype of Kolponomos newpor- tensis n. sp.. USNM 215070. A. metapodial. dorsal and ventral (right) views; B, phalanx, dorsal and ventral ( right) views. posterior articulation is dorsoventrally flattened. A marked ventral keel leads from the joined atlantoaxial articulation to the posterior articular epiphysis where there is a thickening of the keel into a low process. Enhydra and phocids show a similar structure. The robust transverse process sweeps in an arc backward beyond the centrum. as in pinnipeds. The vertebrarterial canal pierces the base of the transverse process. It is of large caliber as in pinnipeds and lutrine mustelids. The neural arch has a narrow base corresponding to the short centrum. The postzygapophyses project as flanges laterally; their articulations slant upward posteriorly and laterally. The high bladelike neural spine hooks posteriorly beyond the neural arch of the third cervical when in articulation. It is strongly inclined anteri- orly, ending in a process for the rectus capitis that lies above the base of the odontoid. The neural spine is thin, wider only at the rectus capitis origin. Its form most approximates that of phocids rather than that of otariids or ursids, in which the spine is more robust and has a marked posterior process. The postzygapophyses lack dorsal processes for insertion of the axial muscles, but the base of the arch beneath the overhanging neural spine is excavated for the more medial elements of this muscle system. The centrum of the third cervical vertebra is about as long as wide, and flattened dorsoventrally (Fig. 11) and keeled with a posterior enlargement, as in pinnipeds. The prominent transverse processes sweep posterolateral^ and terminate in twin processes. They seem to lack an anterior spine as in phocids and in contrast to other carnivores. Large vertebrarterial canals pierce the bases of the transverse processes. The neural arch is low, and the neural canal is correspondingly flattened dorsoventrally, as in pinnipeds. Pre- and postzygapophyses are stout; the latter have low processes for the axial musculature on their dorsal surfaces. This vertebra lacks a neural spine as in phocids, Enhydra. and ursids. Elements of the foot of A", newportensis are represented only by a much eroded metapodial and a nearly complete median phalanx (Fig. 12). The proportions of the metapodial suggest that it may be a third metatarsal, but without the proximal end this identification is uncer- tain. The shaft of this bone is markedly flattened dorsoventrally. as in pinnipeds. Enough remains of the distal end to indicate that the articulation was hemispherical and had a well-developed ventral keel, as in terrestrial carnivores. The median phalanx is also markedly flattened but not elongated as in pinnipeds or Enhydra. Its distal end is trochleated. and the proximal articulation indicates that the distal end of the proximal phalanx was also trochleated. The proximal end has strong lateral processes for flexor tendons and a dorsomedial process for extensors, implying powerful movement of the digits. DISCUSSION Relationships Between the Species of Kolponomos Although we have only one nearly complete cranium of each species, the characters cited in the diagnoses are similar to those that distinguish other nominal species of arctoid carnivores. More- over, the two specimens of A", clallamensis are similar in important particulars that separate them from the specimen here described as K. newportensis. lessening the possibility that the differences be- tween the specimens from Washington and Oregon are due to sexual dimorphism or individual variation. Nor is there evidence that the individuals differ in ways usually associated with sexual dimorphism in arctoid carnivores (gross size, size of canines, and development of muscular processes of the skull). In some ways the cranium of Kolponomos newportensis seems the more primitive of the two, having some characters more like those seen in other arctoids. It has a less highly arched palate, a shorter paroccipital process, a smaller hyoid fossa, and a smaller infraorbital foramen. Also, the jugal rims the anteroventral part of the orbit. Some of its other diagnostic characters, however, such as the broad snout, lack of an intercondylar notch, and the more extremely developed mastoid process, are derived relative to K. clallamensis and other arctoids. K. clallamensis appears to have had a more specialized feeding mode, a more modified rostrum, and greater innervation to the fleshy lips. Relationships of Kolponomos Among the Arctoid Carnivora In 1960 Stirton compared Kolponomos clallamensis exten- sively with Allocyon loganensis Merriam. 1930, from the mid- Arikareean (Oligocene) John Day Formation at Logan Butte, Crook County, Oregon, and concluded that A. loganensis was "the carni- vore most closely related to Kolponomos." Among the 29 features that he delineated, the following similarities seem most informative cladistically; presence of a nasolabials fossa dorsoanterior to the orbit, short infraorbital canal and large infraorbital foramen with infraorbital fossa, and lack of a postorbital process. With the addi- tional material of Kolponomos now available the following can be added: the basioccipital is wide posteriorly, the mastoid process is hypertrophied. and there is a depression anterior to the median lacerate foramen for the first loop of the internal carotid. The latter feature is correlated in Allocyon with a deeply excavated lateral margin of the basioccipital for reception of the carotid artery and inferior petrosal sinus, typical of ursids and amphicyonids. Allocyon (Figs. 13, 14) and Kolponomos are similar in the following dental features that appear to represent synapomorphies: The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematics and Mode of Lite 27 P4 is triangular in outline with a protocone nearly the size of (Allodesmus) or larger than {Kolponomos) the paracone; M: has a posteriorly expanded "heel" (the metaconule and posterior cingu- lum). and the M, talonid is as wide and long as the trigonid. All of these features seem to support the close phyletic relation- ship between A llocyon and Kolponomos. In many features Allocyon is more primitive than Kolponomos, especially those that can be interpreted as adaptations to molluscivory in the latter. Since the relationships of Allocyon have not been made explicit heretofore, we explore the evidence here as a means of placing Allocyon and thus Kolponomos within the Carnivora. The synapomorphies listed above uniting Allocyon and Kolponomos also support their relationship with basal members of the arctoid clade, especially the amphicyonids and ursids. Particu- larly important is the presence of the "ursid loop" in which the internal carotid artery is nested in the inferior petrosal sinus (Hunt 1977). This system has a clear bony signature in the deep marginal invagination of the basioccipital. Although this feature cannot be completely seen in Kolponomos. it can in Allocyon, and it is entirely comparable to the structure in amphicyonids and ursids. A position closer to ursids is supported dentally by loss of M\ enlargement of the protocone of P4, and development of a "heel" in M2 by enlarge- ment of the metaconule and associated posterior cingulum. Within the Ursidae significant autapomorphies unite the sub- families Hemicyoninae and Ursinae and exclude Allocyon and Kolponomos, whose relationships lie near or within the most basal ursoid group, usually referred to as the Amphicynodontinae (Simpson 1945). Amphicynodon and Pachycynodon are the most completely known taxa (Cirot and de Bonis 1992; Cirot 1992) included in this group. The latter, and larger, form resembles Allocyon particularly in its posteriorly extended palate (Fig. 14), but it resembles both Allocyon and Kolponomos in having a short infraorbital canal, fossa nasolabialis, enlarged infraorbital foramen, and a similarly reduced postorbital process. Our conclusions about the relationships of Allocyon and Kolponomos with the primitive ursoids are similar to those postu- lated for the most primitive pinnipedimorph, Enaliarctos, by Flynn et al. (1988). Berta (1991). and Hunt and Barnes (1994, this vol- ume). In cranial morphology Kolponomos and Allocyon resemble the pinnipedimorphs in having a fossa for the origin of the nasolabialis muscle, a short infraorbital canal, a large infraorbital foramen opening into a fossa, a small optic foramen, and in lacking a postorbital process. Furthermore, in Kolponomos the lacrimal is small, fusing early to adjacent bones, and in A", clallamensis the maxilla forms the anterodorsal orbital rim. In Kolponomos the foramen rotundum has a common opening with the anterior lacerate foramen, the postglenoid foramen is vestigial, the posterior carotid foramen is clearly anterior to the posterior lacerate foramen, and M, is absent. Some of these features and others noted in the vertebral column may represent trends parallel to those seen in pinnipeds, especially with respect to aquatic adaptation (e.g., emphasis on the internal jugular drainage of the cranium and thus loss of the postglenoid exit), but the sum total suggests that Allocyon and Kolponomos represent early offshoots of the same stock that yielded enaliarctine pinnipedimorphs and that both have their roots in the earliest differentiation within the Superfamily Ursoidea. A more explicit hypothesis of the relationships of Kolponomos to other ursoids and pinnipedimorphs can be made by using the TABLE 3. Distribution of cranial and dental features discussed in the text (0, primitive state; 1, derived state). Taxon" Derived state AMP MUS URS AMC PAC ALL KOL ENA 1. Basioccipital excavated laterally 1 0 1 1 1 1 1 2. Shallow suprameatal fossa 0 1 1 9 0 0 0 3. M3 absent 0 1 1 1 1 1 1 4. Basioccipital wide posteriorly 0 0 1 1 1 1 1 5. P4 large protocone 0 0 1 1 1 1 1 6. M: with "heel" 0 0 1 1 1 1 1 7. M'~2 loss parastyle 0 0 1 1 1 1 0 8. M'~: loss paraconule 0 0 1 1 7 0 0 9. M1 lingual metaconule 0 0 0 1 7 0 0 10. M, reduced paraconid 0 0 1 1 7 0 •> 1 1. M, size metaconid = protoconid 0 0 1 1 7 1 9 1 2. Infraorbital canal short 0 0 0 1 1 1 1 1 1 3. Infraorbital foramen large 0 0 0 1 1 1 1 14. P4 short metastyle 0 0 0 1 1 1 1 15. Palate posteriorly extended 0 0 0 0 1 0 1 16. M, talonid as wide as trigonid 0 0 0 0 1 1 0 17. M, metaconid large 0 0 0 0 1 1 0 18. Nasolabialis fossa present 0 0 0 0 0 1 1 19. Infraorbital fossa present 0 0 0 0 0 1 1 20. Mastoid process large 0 0 0 0 0 1 0 21. Postorbital process absent 0 0 0 0 0 1 1 22. Alisphenoid "strut" present 0 0 0 0 0 0 1 1 23. Postglenoid foramen vestigial 0 0 0 0 (1 0 1 1 24. M2 lingual to M' 0 0 0 0 0 0 1 1 25. Anterior lacerate foramen and foramen rotundum in common fossa 0 0 0 0 (1 0 1 1 26. M, absent 0 1 0 0 0 0 1 1 "AMP, Amphicyonidae; MUS, Mustelida; URS, Ursinae and Hemicyoninae; AMC, Amphicynodon: PAC, Pachycyno- don; ALL, Allocyon: KOL, Kolponomos: ENA, Enaliarctos. 28 R. H. Tedford. L. G. Barnes, and C. E. Ray Figure 13. Comparative outline drawings of left side of crania. A. Kolponomos clallamensis Stirton, 1960, referred, LACM 131 148: B, Allocyon loganensis Merriam. 1930, holotype. UCMP 24106, from Merriam (1930: fig. 1); C, Pachycynodon boriei (Filhol, 1877), holotype, from Filhol (1877: fig. 59). sister taxon, the Mustelida (Procyonidae + Mustelidae), and a basal arctoid group, the Amphicyonidae. as outgroups. For this purpose we scored 26 of the binary characters discussed above among eight taxa (Table 3). The taxa are the Amphicyonidae (represented by Daphoenodon), Mustelida (represented by the archaic forms Mustelictis, Amphictis, and Plesictis; Schmidt-Kittler 1981). Ursidae (including the hemicyonine Cephalogale and the ursine Ursus). Amphicynodon (mostly A. typicus, BM(NH) M749I). Pachycynodon boriei (Filhol 1877:pl. 58-60, as Cynodictis gryei: Cirot \992), Allocyon (Merriam 1930, UCMP24106). Kolponomos (both species), and Enaliarctos (mostly E. meulsi, but also data from other species described by Berta 1991). The branch-and- bound algorithm of PAUP(2.4.1 ) found a single most parsimonious tree (Fig. 15) with a branch length of 36, a consistency index of 0.72, and a retention index of 0.73. Further explanation of the characters used as synapomorphies are as follows: 1. Basioccipital deeply excavated laterally. — As Hunt (1977) has shown in living ursids. the large inferior petrosal sinus contain- ing the intracranial loop of the internal carotid artery lies in a deep excavation in the lateral margin of the basioccipital that extends to the posterior lacerate foramen. A morphologically identical struc- ture occurs in the amphicyonids (including the daphoenines), im- plying a similar vascular pathway and a synapomorphy for the Arctoidea. Amphicynodontids and pinnipedimorphs [Enaliarctos; Hunt and Barnes 1994, this volume) retain this feature. Derived pinnipeds and members of the Mustelida lack it. 2. Shallow suprameatal fossa present. — All arctoids above the Amphicyonidae show a suprameatal fossa that may be later trans- formed into a deep pit in the squamosal dorsal to the external auditory meatus or may exist as shallow structures floored by the auditory tube and obliterated in ontogeny by fusion with the meatus (Schmidt-Kittler 1981). A small shallow fossa excavated dorso- The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematic* and Mode of Life 29 B Figure 14. Comparative outline drawings of ventral side of crania. A, Kolponomos clallamensis Stirton, 1960. referred. LACM 131148; B. Allocyon loganensis Merriam. 1930, holotype, UCMP 24106. from Merriam (1930: fig. 3); C, Pachycynodon boriei (Filhol, 1877), holotype. from Filhol (1877: fig. 60). posteriorly into the squamosal contribution to the mastoid process seems to be the most primitive state of this feature. In ursids and amphicynodontids this structure is shallow primitively and is lost in derived taxa rather than being obliterated by growth of the tubular external auditory meatus. 3. M1, absent. — A feature uniting all arctoids above the Amphicyonidae. 4. Basioccipital wide posteriorly. — The greater width of the basioccipital across the posterior lacerate foramen versus its width at the basisphenoid suture is a derived feature of ursids and higher arctoids. This feature was cited by Wyss (1987) as a pinniped synapomorphy. later modified by Berta (1991: table 7, no. 44) to indicate the short and wide basioccipital that describes the condi- tion of pinnipeds above the Otariidae. 5. Fourth upper premolar with large protocone. — All ursoids have an upper carnassial with a broad protocone that is usually shelflike because of incorporation of the lingual cingulum. Further enlargement of this cusp in Kolponomos relative to the labial cusps is an autapomorphy that serves to "molarize" the carnassial. 6. Second upper molar with "heel." — All ursoids have an M: that differs from the tribosphenic form of that of other arctoids by the posterior shelf or "heel" formed by a well-developed posterior cingulum, often incorporating the metaconule. This appears to be the case in Kolponomos and probably Allocyon. Although the M: of Enaliarctos is very reduced it seems to include a shelflike heel behind the trigon (as in E. emlongi; Berta 1991 ) and so is coded as derived in this feature. 7. Reduction and loss of parastyle on M'~:. — Early parastyle reduction and loss is a feature of ursines and hemicyonines. Amphicynodontids lost this cusp later in phylogeny as similarly hypocarnivorous forms (e.g., Pachycynodon) arose. 8. Reduction and loss of paraconule on M'2. — Full loss of the paraconule took place at different times in the ursine, hemicyonine, and amphicynodontid lineages, but reduction in the size of this cusp characterizes the early members of all groups. Curiously, Kolponomos retains this cusp, inflated like all the molar cusps, to form the broad grinding surface as in another molluscivore. Enhydra. 9. Lingual position of M1 metaconule. — The ursid metaconule is strongly connected to the longitudinally elongated protocone in M1 (and M:). It has lost its primitive connection with the metacone and occupies a more lingual position on the crown (Cirot and de Bonis 1992). Amphicynodon retains a primitively labial position of the M1 metaconule but this cusp is large and well connected to the protocone by a crista. Kolponomos also retains a primitive tribos- phenic form of M1 but all cusps are inflated and lack connecting cristae. 10. Reduction of paraconid of A/,. — Modification of M-, in ursoids involves enlargement of the talonid relative to the trigonid. Reduction and loss of the paraconid accompanies attainment of subequal size of the metaconid and protoconid and their assumption of a more transverse relationship. Again. Kolponomos appears to retain a paraconid in its large M, trigonid. 30 R. H. Tedford, L. G. Barnes, and C. E. Ray P.O. URSIDA S.F URSOIDEA PINNIPEDI- MORPHA F AMPHICYNODONTIDAE 7 16 22,23,24.25,26 7. 20 8,19,20,21 2? 15, 16,17 12,13.14 4,5,6,7,8,10,11 Figure 15. Phyletic relationships of taxa discussed in the text. For character numbers see text and Table 3. Asterisks indicate reversal to primitive state. 1 1 . Size of metaconid equal to protoconid on A/-,. — The reduc- tion of the trigonid relative to the talonid in the M, of ursids and amphicynodontids involves enlargement of the metaconid to the same size as the protoconid. 12. Infraorbital canal short. — The distance from the anterior edge of the orbit to the opening of the canal is unusually abbrevi- ated in the amphicynodontid ursoids and pinnipedimorphs. 13. Infraorbital foramen large. — Amphicynodontid ursoids and pinnipedimorphs also have a very large anterior opening of the short infraorbital canal. 14. Metastyle ofP4 short. — In a trend toward hypocamivory the amphicynodontids shorten the carnassial blade by reduction of the metastyle. 1 5. Palate posteriorly extended. — The palate is prolonged in the midline by posterior extension of the palatine bones so that the internal nares lie at a considerable distance behind the tooth row. This condition is derived in the Ursoidea but has an independent distribution that implies some homoplasy (i.e.. it is present in the Ursinae and some amphicynodontids but not in Kolponomos or pinnipedimorphs generally!. 1 6. M f talonid as wide as or wider than trigonid. — Correspond- ing to modification of the M' talon (character 9), amphicynodontids enlarge the talonid of the lower carnassial to accommodate the longitudinal protocone and associated metaconule of M1. 17. M j metaconid large. — Modification of the M, trigonid more for crushing in derived amphicynodontids involved enlargement of the metaconid relative to surrounding cusps. In volume it comes to match the paraconid and is nearly as high in the unworn crown as the protoconid. 18. Nasolabialis fossa present. — A prominent fossa just dorsoanterior to the orbital rim. presumably for the nasolabialis muscle, is a derived condition in Allocyon, Kolponomos, and primi- tive pinnipedimorphs. From the perspective of pinniped evolution Berta ( 1 99 1 ) coded the loss of this feature in pinnipeds as derived, i.e., a reversal to the primitive arctoid state. 19. Infraorbital fossa present. — This broad depression lies just anterior to the opening of the infraorbital canal onto the face. It is not correlated with the large foramen of amphicynodontids but characterizes Allocyon, Kolponomos. and pinnipedimorphs. 20. Mastoid process hypertrophied. — This is a synapomorphy for Allocyon and Kolponomos. although the latter has greatly elon- gated the process ventrally. The massive backward-pointing paroccipital process in Allocyon is an autapomorphy for that genus. 2 1 . Postorbital process lacking. — In contrast to other arctoids, in Allocyon and Kolponomos the postorbital processes of the frontals are lacking. Low supraorbital ridges at the anterior ends of the parasagittal crests represent the position of the processes in these genera and in primitive pinnipedimorphs (Berta 1991 ). 22. Alisphenoid "strut" present. — A reinforced region extends from the palatine process of the alisphenoid dorsoanteriorly to a correspondingly reinforced pterygoid process of the palatine. These elements combine to form a strut bracing the posterior part of the palate against the braincase. Such a structure is present in Kolponomos and pinnipedimorphs. The Early Miocene Littoral Ursoid Carnivoran Kolponomos: Systematica and Mode of Life 31 23. Postglenoid foramen vestigial. — Reduction of this opening is correlated with greater emphasis on the internal jugular system as the main venous drainage of the braincase. Although the posterior lacerate foramen in Kolponomos is not conspicuously enlarged, the postglenoid foramen is very reduced as in pinnipedimorphs. 24. M2 lingual to M' . — A peculiar feature of the dentition of pinnipedimorphs (and Potamotherium) that retain M2 is the lingual position of this tooth adjacent to the talon of M' rather than labial as in most carnivores (M. Wolsan, pers. comm.). Kolponomos shows this feature. 25. Foramen rotundum and anterior lacerate foramen lie in a common fossa. — Berta (1991) discussed the distribution of this pinnipedimorph synapomorphy. It also clearly occurs in Kolponomos. but other amphicynodontids have the primitive state in which these foramina are separated by a bony lamina visible in lateral view of the skull. 26. My absent. — As in pinnipedimorphs this tooth is absent in Kolponomos. Figure 15 summarizes the distribution of these synapomorphies, indicating the paraphyly of the Amphicynodontidae when the Pinnipedimorpha (sensu Berta 1991 ) are placed within this group as the sister taxon of Kolponomos. Some of the characters thought to be synapomorphies of the Pinnipedimorpha by Berta ( 1991 ) actu- ally have a wider distribution (e.g.. characters 13. 21. 23, 25. and 26) within the Ursoidea or have ursoid precursor states (character 4). Synapomorphies specifically linking the Pinnipedimorpha with Allocyon and Kolponomos (characters 18, 19. and 21-26) indicate that these terrestrial and amphibious arctoids. although dentally specialized for hypocarnivory. approximate the stem group for the pinnipedimorphs. A classification consonant with the phyletic relationship pos- tulated for the carnivorans discussed above is indicated in Figure 15. The Pinnipedimorpha were not ranked by Berta (1991) and are so indicated on Figure 15 as an unranked taxon within the parvorder Ursida of Tedford (1976). Since the traditional "suborder" Pinnipedia is subsumed in the Pinnipedimorpha it too must remain unranked in the present attempt to construct a taxonomy that expresses the phylogenetic conclusions of this paper. Speculations About the Mode of Life of Kolponomos The few postcranial bones presently known indicate that Kolponomos was not fully aquatic, at least in the sense that the pinnipeds are. Its foot bones clearly indicate retention of significant ability for terrestrial locomotion and an amphibious existence. It was probably littoral in distribution. All known specimens are from near-shore, shallow-water marine deposits that contain abundant fossil mollusks, including large mussels and giant pectinids. The broad, sea-otterlike crushing cheek teeth would have been ideally suited to a diet of hard-shelled marine invertebrates. The teeth are well worn, indicating that the diet included very hard-shelled ani- mals, possibly mussels, limpets, abalone. pectinids. and echinoids. Coupled with accidentally ingested abrasive sediment, these would account for the heavily worn condition of the cheek teeth. The orbits are directed anteriorly rather than laterally as in living bears, suggesting that Kolponomos probably could view objects directly in front of its head. This would be of benefit to an animal selectively eating rock-dwelling benthic or attached (sessile) marine inverte- brates. The infraorbital foramen is large, quite so in K. clallamensis. and the mental foramina on the lateral side of the dentary of K. newportensis are also large. These probably indicate enhanced tactile sensitivity of the lips and muzzle. Kolponomos might also have had a large upper lip approaching that in living walruses, and this would be concomitant with the depth of the premaxilla between the incisors and the narial opening. Walruses have very sensitive lips and tactile vibrissas, that apparently aid in distinguishing prey items when visibility is poor (Fay 1982). Kolponomos might also have had highly developed tactile vibrissae. The palate is flexed downward relative to the basicranial plane; the occipital condyles face ventrally and are positioned posteroventrally relative to the basicranium, suggesting that the head was carried downward in relation to the vertebral column. The upper canine and incisor teeth are large and clustered in thickened bone at the extreme anterior end of the downturned snout. The nasal opening is retracted posteriorly, an adaptation that would keep the nostrils away from the substrate. Large paroccipital and mastoid processes indicate powerful neck muscles that could have provided strong downward movements of the skull. All these adaptations suggest that Kolponomos fed on epifaunal marine invertebrates living on rocky substrates. Kolponomos probably obtained its food by levering tightly clinging animals off the substrate and twisting and prying with its head. The robust median phalanx also suggests that the digits were capable of powerful movement and these too may have been used to procure food. This method of feeding is somewhat different from that of living sea otters. Sea otters actively swim in shallow to moderate depths and obtain bottom-dwelling animals, largely by pulling and prying them off rocks with their forelimbs. Sea otters have rela- tively long digits and strong forelimb musculature. They have large, flat, crushing cheek teeth, but they do not have enlarged mastoid and paroccipital processes. This correlates with the fact that they do not pull their food off rocks by using their heads. Kolponomos also is unlike the living walrus, which belongs to a pinniped group that later in the Tertiary became very diverse, successful, and widespread along the coasts of the northern hemi- sphere. Modern walruses occupy shorelines, at least part of the time, when they haul out at specific locations along the shore. When feeding, however, they are offshore, diving pinnipeds. They typi- cally dive only to shallow or moderate depths, where they exploit food resources for the most part different from those of other pinnipeds, mostly benthic shelled and nonshelled invertebrates. They do not crush mollusk shells by chewing (Fay 1982) but rather use the tongue in a pistonlike method to suck the soft parts out of gaping bivalve shells. They also use the tongue as a piston to direct a jet of water from the mouth onto the substrate in a method of hydraulic mining of infaunal prey. Walruses do not chew up shells of their prey and they do not swallow shells or broken shells, so, although the general category of food of walruses is the same as that proposed for Kolponomos. the locating of the food and manner of gathering and eating it is apparently different. The specialized dusignathine otariid Gomphotaria pugnax may be a relatively close functional counterpart to Kolponomos. This large pinniped is known from upper Miocene rocks of the Califor- nia coast. Like Kolponomos. Gomphotaria had elongated upper as well as lower canines, even more fully developed as tusks. Also like Kolponomos, Gomphotaria had large cheek teeth, which although not expanded transversely were broken and worn during life from feeding on resistant prey. Barnes and Raschke ( 1991 ) proposed that Gomphotaria was a shallow-water or littoral pinniped that pried its food off rocks, the food presumably being shelled mollusks as we have postulated for Kolponomos. and that rather than sucking its food into the mouth it crushed the animals with the cheek teeth. Kolponomos appears to be an ursid variation on the sea otter adaptation. On the west coast of North America, middle and upper Miocene horizons bearing fossil marine vertebrates have been ex- tensively prospected, much more extensively than have the lower Miocene deposits. Nothing related to Kolponomos has as yet been found in these younger deposits. Kolponomos might very well be the end of its lineage. 32 R. H. Tedford, L. G. Barnes, and C. E. Ray ACKNOWLEDGMENTS We thank Albin Zukofsky. II, for his donation of the important skull of Kolponomos clallamensis from Clallam Bay. We thank Robert L. (Fritz) Clark for preparing and casting this new skull. The photographs of the specimen were made by LACM staff photogra- pher Donald Meyer. We thank Donald E. Savage and J. Howard Hutchison for making the holotype of Kolponomos clallamensis available for our work. We thank Jim Goedert for making observa- tions relating to the collecting sites of the holotype and referred specimens of Kolponomos clallamensis. for making notes on the stratigraphy, and for analyzing the associated mollusks to make inferences about the paleoecology. The holotype of Kolponomos newportensis was collected by the late Douglas Emlong in two fragments of one concretion, found on separate occasions more than six years apart. Emlong recognized that the second, major part pertained to the first, minor one, even though the latter had no intelligible bone showing and had not been seen by him for several years. Assembly of the broken concretion containing the type of K. newportensis and gross preparation, including separation of the jammed mandible and skull and loose teeth, was done by Gladwyn B. (Tut) Sullivan. Final preparation of the type specimens of K. clallamensis and K. newportensis was skillfully done by Edward Pedersen of the American Museum of Natural History. Photographs of these specimens and line drawings were carefully prepared by Chester Tarka and Lorraine Meeker of the American Museum. Xiaoming Wang helped with the PAUP analysis of phylogeny. LITERATURE CITED Addicott, W.O. 1976a. Neogene molluscan stages of Oregon and Wash- ington. Pp. 95-115 in A. E. Fritsche, H. Ter Best, Jr.. and W. W. Womardt (eds.). The Neogene Symposium. Society of Economic Paleontologists and Mineralogists (Annual Meeting. Pacific Sec- tion). San Francisco. California. . 1976b. New molluscan assemblages from the Upper Member of the Twin River Formation, western Washington: Significance in Neogene chronostratigraphy. United States Geological Survey Journal of Research 4:437-447. -. 1976c. Molluscan paleontology of the lower Miocene Clallam Formation, northwestern Washington. United States Geological Survey Professional Paper 976. Armentrout, J. M., D. A. Hull. J. D. Beaulieu, and W. W. Rau (eds.). 1983. Correlation of Cenozoic stratigraphic units of western Or- egon and Washington. Oregon Department of Geology and Mineral Industries Oil and Gas Investigations 7. Portland. Oregon. Bames, L. G. 1987. Aetiocetus and Chonecetus, primitive Oligocene toothed mysticetes and the origin of baleen whales. Journal of Vertebrate Paleontology 7(3) supplement: 10A. . 1989. Aetiocetus and Chonecetus (Mammalia: Cetacea); primi- tive Oligocene toothed mysticetes and the origin of baleen whales. Abstracts of Posters and Papers, Fifth International Theriological Congress. Rome. Italy, 22-29 August 1989, 1 :479. , and R. E. Raschke. 1991. Gomphotaria pugnax, a new genus and species of late Miocene dusignathine otariid pinniped (Mammalia: Carnivora) from California. Natural History Museum of Los Angeles County Contributions in Science 426. , D. P. Domning. and C. E. Ray 1985. Status of studies on fossil marine mammals. Marine Mammal Science 1:15-53. Berta, A. 1991. New Enaliarctos* (Pinnipedimorphai from the Oligocene and Miocene of Oregon and the role of "enaliarctids" in pinniped phylogeny. Smithsonian Contributions to Paleobiology 69. Carroll. R. L. 1988. Vertebrate Paleontology and Evolution, W. H. Freeman, New York, New York. Cirot, E. 1992. Etude phylogenetique dequelques genres d'Arctoideade 1'Oligocene eurasiatique. Comparaison des donnees morpho- logiques et moleculaires. Ph.D. Dissertation. Faculte des Sciences Fondamentales et Appliquees de Poitiers. Poitiers, France. — , and L. de Bonis. 1992. Revision du genre Amphicynodon, carnivore de 1'Oligocene. Palaeontographica. Pt. A. 220: 103-130. Dalrymple, B. 1979. Critical tables for conversion of K-Ar ages from old to new constants. Geology 7:558-560. Domning. D. P., C. E. Ray. and M. C. McKenna. 1986. Two new Oligocene desmostylians and a discussion of tethytherian system- atics. Smithsonian Contributions to Paleobiology 59. Fay, F H. 1982. Ecology and behavior of the Pacific walrus. Odobenus rosmarus divergens Illiger. North American Fauna 74. Feldmann. R. M., A. B. Tucker, and R. E. Berglund. 1991. Fossil crustaceans, paleobathymetry and decapod crustaceans. Washing- ton. National Geographic Research and Exploration 7:352-363. Flynn, J. J., N. A. Neff. and R. H. Tedford. 1988. Phylogeny of the Carnivora. Pp. 73-116 in M. J. Benton (ed.). The Phylogeny and Classification of the Tetrapods, vol. 2. Systematics Association Special Volume 35B. Clarendon Press. Oxford, England. Hunt. R. M.. Jr. 1977. Basicranial anatomy of Cynelos Jourdan (Mammalia: Carnivora). an Aquitanian amphicyonid from the Allier Basin, France. Journal of Paleontology 5 1 :826-843. . and L. G. Barnes. 1994. Basicranial evidence for ursid affinity of the oldest pinnipeds, in A. Berta and T. A. Demere (eds.) Contri- butions in marine mammal paleontology honoring Frank C. Whitmore. Jr. Proceedings of the San Diego Society of Natural History 29:57-68. Memam. C. W. 1930. Allocyon, a new canid genus from the John Day beds of Oregon. University of California Publications. Bulletin of the Department of Geological Sciences 19:229-244. Moore. E. J., and W. O. Addicott. 1987. The Miocene Pillarian and Newportian (Molluscan) stages of Washington and Oregon and their usefulness in correlations from Alaska to California. United States Geological Survey Bulletin 1664A. Olson. S. L. 1980. A new genus of penguin-like pelecaniform bird from the Oligocene of Washington (Pelecaniformes: Plotopteridae). Natural History Museum of Los Angeles County Contributions in Science 382. Piveteau, J. 1961. Mammiferes: Origine reptilienne, evolution. Traitede Paleontologie 6:1-1 138. Rau, W. W. 1964. Foraminifera from the northern Olympic Peninsula, Washington. United States Geological Survey Professional Paper 374-G. Ray, C. E., D. P. Domning. and M. C. McKenna. 1994. A new specimen of Behemotops proteus (order Desmostylia) from the marine Oli- gocene of Washington, in A. Berta and T. A. Demere (eds.) Contri- butions in marine mammal paleontology honoring Frank C. Whitmore. Jr. Proceedings of the San Diego Society of Natural History 29:207-224. Romer, A. S. 1966. Vertebrate Paleontology. University of Chicago Press. Chicago, Illinois. Schmidt-Kittler. N. 1981. Zur stammesgeschicte der marderverwandten Raubtiergruppen (Musteloidea. Carnivora). Eclogae Geologicae Helvetiae 74:753-801. Simpson, G. G. 1945. The principles of classification and a classifica- tion of mammals. Bulletin of the American Museum of Natural History 85: 1-350. Snavely, P. D.. Jr. 1983. Peripheral rocks: Tertiary geology of the north- western part of the Olympic Peninsula. Washington. Geological Association of Canada. Victoria. British Columbia. Field Trip Guidebook 12:6-31. — , A. R. Niem. N. S. MacLeod. J. E. Pearl, and W. W. Rau. 1980. Makah Formation — a deep-marginal-basin sequence of Late Eo- cene and Oligocene age in the northwestern Olympic Peninsula. Washington. United States Geological Survey Professional Paper 1162-B. Stirton. R. A. 1960. A marine carnivore from the Clallam Miocene Formation, Washington. Its correlation with nonmarine faunas. University of California Publications in Geological Sciences 36:345-368. Tabor, R. W., and W. M. Cady. 1978. Geologic map of the Olympic Peninsula. Washington. United States Geological Survey Miscella- neous Investigations Series Map 1-994. Tedford. R. H. 1977. Relationships of pinnipeds to other carnivores (Mammalia). Systematic Zoology 25 (for 1976):363-374. Thenius. F. 1969. Stammesgeschichte der Siiugethiere (Einschliesslich der Hominiden). Handbuch der Zoologie. Band 8. Teil 2. Vol. 1:1- 368. Wyss. A. R., 1987. The walrus auditory region and the monophyly of pinnipeds. American Museum Novitates 2871. Pinniped Phylogeny Annalisa Berta Department of Biology, San Diego State University, San Diego. California 92182 Andre R. Wyss Department of Geological Sciences, University of California. Santa Barbara, California 93106 ABSTRACT. — In our view, presentations inferring pinniped cJiphyly provide inadequate evidence of "otarioid" monophyly and inadequate evidence that phocids are related to some nonpinniped group. The integrated assessment of higher-level pinniped relationships presented here, based on cranial, postcranial, and soft-anatomical characters from most living and adequately known fossil pinnipeds, supports pinniped monophyly. We scored more than 150 character transformations on a generic-level character matrix and used a computer parsimony algorithm (PAUP) to construct a maximally parsimonious phylogenetic hypothesis for the group. Its major outlines are as follows: (Enaliarctos (Pteronarctos (Otariidae (Odobenidae (Pinnarctidion, Desmatophoca, Allodesmus (Phocidae))))). Internally, the data are highly consistent. Convergence is much less pervasive than generally assumed, with reversals being the dominant pattern of homoplasy. INTRODUCTION Few mammalian examples more forcefully illustrate the impact of phylogenetic systematic methods on notions of a particular group's evolutionary history than does the case of the pinnipeds. Recent cladistic studies have addressed questions of relationships within the otariids (fur seals and sea lions) (Berta and Demere 1986) and phocids (true seals) (Muizon 1982a; Wyss 1988b). odobenid (walrus) affinities (Wyss 1987). and the placement of certain archaic fossil taxa (Wyss 1987; Berta et al. 1989; Berta 1991). The net result of these efforts is a concept of pinniped relationships drastically different from what was generally accepted until relatively recently. We outline here the novel aspects of these recent proposals and present their methodological basis. In addition to providing new information on the distribution of morphological characters, we combine and revise the data sets of these previous studies in an attempt synthesize what we regard to be the currently best supported hypothesis of pinniped relationships. We present the evidence for this hypothesis in the form of a taxon-character matrix in hopes that it may serve as a starting point for future phylogenetic analyses of pinnipeds. If this matrix generates debate about charac- ter coding, or discussion over the in- or exclusion of certain charac- ters in the analysis, or if it inspires the examination and description of additional characters that either support or refute the relation- ships we favor, in short, if it evolves, it will have served its purpose. HISTORICAL CONSIDERATIONS All recent workers agree that pinnipeds are members of a carnivoran subclade. the Arctoidea, that includes among terrestrial lineages mustelids, ursids. and procyonids (Tedford 1976). Contro- versy about relationships among the major groups of pinnipeds centers on the relationship of phocids to the rest of the Arctoidea. This disagreement reduces to two fundamental questions of mono- phyly. Before we consider these (to eliminate possible ambiguity) we must define our usage of "monophyly." We use the term (sensu Hennig 1966) to denote a group of taxa derived from a common ancestor and including all of the descendants of that common ancestor. Evidence for monophyly of a particular group consists of the shared possession of evolutionary novelties (synapomorphies) by its members. The two central questions concerning the phylogenetics of "fin-footed" arctoids are ( 1 ) is the group as a whole descended from an exclusive common ancestor? (i.e., are pinnipeds monophyletic?) and (2) do pinnipeds excluding phocids have a common ancestor not also shared by phocids? (i.e.. are the Otarioidea, defined as the Otariidae and Odobenidae plus certain extinct forms, monophyletic?) (Fig. 1 ). The question of single versus multiple origin! s) dates from Mivart's ( 1885) suggestion that the group's origin was likely com- pound, with sea lions and walruses derived from ursids and true seals derived from mustelids, otters in particular (Fig. IB). Al- though this view was dismissed by several workers over the next half century (e.g., Weber 1904; Gregory 1910), it was not dis- counted altogether by others (e.g., Kellogg 1922; Howell 1929; Simpson 1945). Thereafter, the notion of multiple pinniped origins regained wide support in the morphological and paleontological literature (McLaren 1960; Tedford 1976; Muizon 1982a,b), a shift influenced particularly by the detailed descriptions of the fossil taxa Potamotherium (Savage 1957) and Enaliarctos (Mitchell and Tedford 1973). More recently, one of us argued, on the basis of anatomical criteria, in favor of a return to the single-origin interpre- tation (Wyss 1987) (Fig. 1A), a conclusion consistent with but independent of certain biomolecular and cytologic results (Fay et al. 1967; Arnason 1986; de Jong 1982, 1986). Several subsequent studies (Flynnetal. 1988; Berta etal. 1989; Wyss 1988, 1989) have yielded additional evidence supporting this conclusion, but it con- tinues to engender debate (Wozencraft 1989; Repenning 1990; Bonner 1990). The second question concerns the phylogenetic validity of the Otarioidea. Since their recognition as distinct groups of mammals, otariids and odobenids have nearly universally been regarded as being more closely related to each other than either is to some third taxon. The observation that the walrus is in many respects more nearly intermediate between otariids and phocids than had been previously appreciated (Fay et al. 1967) signaled an important break from this view. The argument that odobenids are related more closely related to phocids than to otariids took this suggestion one logical step further (Wyss 1987). This proposed phocid-odobenid linkage opened the broader question of where this pair should be placed relative to the other arctoids. Either a special link between mustelids and phocids (plus now odobenids) could continue to be recognized (rendering the Otarioidea polyphyletic), or the associa- tion of phocids and odobenids could be recognized within the context of a monophyletic Pinnipedia (rendering the Otarioidea paraphyletic). Thus the questions of otarioid and pinniped mono- phyly are to some degree interwoven, yet if care is taken both may be evaluated with considerable independence. To two phylogenetic questions, are pinnipeds monophyletic and are otarioids monophyletic, there are four alternative pairs of re- sponses. All except one of these have recent historical precedent: affirmative, affirmative (no recent proponents); affirmative, nega- tive (Fig. 1A) (Wyss 1987; Berta et al. 1989; Flynn et al. 1988; Wyss and Flynn 1993); negative, negative (Mitchell and Tedford In A. Berta and T. A. Demere (eds. i Contributions in Marine Mammal Paleontology Honoring Frank C. Whitmore. Jr. Proc. San Diego Soc. Nat. Hist 29:33-56, 1994 34 A. BertaandA. R. Wyss Terrestrial Arctoids Enaliarctos Otariidae Desmatophoca Odobenidae Allodesmus Phocidae . Otarioidea Mustelidae Potamotherium Figure 1. Two competing hypotheses regarding phylogenetic relation- ships among pinnipeds. A, Current yiew of pinniped monophyly proposes common ancestry for all pinnipeds from a terrestrial arctoid and supports a sister-group relationship between walruses, seals, and their extinct relatives (Wyss 1987; Berta et al. 1989; Flynn et al. 1988). B. Alternative view of otarioid monophyly proposes independent origins of otarioid and phocid lineages from different terrestrial artcoid groups and supports a sister-group relationship between walruses, sea lions, and their supposed fossil relatives (Barnes 1989. and others). 1973: 205, 278); negative, affirmative (Fig. IB) (Barnes 1989; Wozencraft 1989; Repenning et al. 1979). We begin with the otarioid question because in some senses it is less complex and easier to address in isolation. We emphasize again that to test the hypothesis of otarioid monophyly apart from the question of pin- niped monophyly we must restrict our attention to those derived features characteristic of otarioids not also occuring in phocids. OTARIOID MONOPHYLY Were pinnipeds demonstrably polyphyletic both the otarioids and phocids could be diagnosed with characters each shares but acquired independently. But because we view otarioid monophyly as a hypothesis in legitimate need of testing we regard claimed otarioid synapomorphies occurring also in phocids as questionable, at least for the initial part of the analysis. Barnes (1989: fig. 9) presented the most recent, most detailed, and, so far as we are aware, only nominally cladistic diagnosis of the Otarioidea (= his Otariidae). He listed on a branching diagram 20 characters diagnostic of the Otarioidea. Following the procedure of Wible ( 1 99 1 ), we evaluated these 20 features, grouping them as follows: ( 1 ) characters for which the reported derived state occurs in relevant outgroups (i.e.. nonpin- niped Arctoidea) and are therefore primitive and phylogenetically uninformative, (2) characters for which the derived state occurs also in phocids and thus are of uncertain value. (3) characters dubiously described, and (4) characters for which descriptions and distribution are correct but for which we offer comment. Characters are labeled with letters corresponding to the order in which they were listed by Barnes ( 1989: fig. 9) at node 1, the basal node of his branching diagram. Barnes' otarioid "synapomorphies" occurring also in relevant outgroups. — The derived state of the following characters occurs elsewhere among the Arctoidea and thus represents a level of generality broader than supposed by Barnes ( 1989). (a) Neck lengthened. Wyss (1987: 11) discussed previously problems associated with this character. Even if it did characterize otarioids primitively (among odobenids, it doesn"t characterize at least Odobenus, the only odobenid for which this character can currently be scored), it is not a derived feature among the Carnivora (Bisaillonet al. 1976). (e) Foramen ovale and posterior opening of alisphenoid canal joined in an elongated recess. The arrangement in ursids is identical (Davis 1964) and almost certainly represents the ancestral condi- tion of the Arctoidea. (h) Embayment formed in lateral edge of basioccipital for loop of median branch of internal carotid artery. This character is well known in ursids (Hunt 1974) and amphicyonids (Hunt 1977) and is unquestionably derived at a level broader than the Otarioidea. Although this embayment is absent in all living pinnipeds, in cer- tain phocids (e.g., Monachus) sharp crests on the dorsal surface of the basioccipital may represent an osseous vestige of it. (j) Basal whorl of cochlea directed posteriorly. As discussed by Wyss (1987), this condition characterizes all therian mammals except phocids, which are uniquely specialized in having a more transversely oriented basal whorl. (1) Auditory ossicles not enlarged. This feature is obviously primitive; the derived condition (ossicles enlarged) originated three times among otarioids according to Barnes' scheme and again (presumably independently) in phocids (see below under Assessing the Pattern of Homoplasy). (m) Entotympanic restricted to medial part of bulla around carotid canal. This condition corresponds to the type "A" bulla of Hunt ( 1974), which characterizes ursids, some amphicyonids, some mustelids, and perhaps arctoids ancestrally. It is not uniquely diag- nostic of otarioids and is therefore not relevant to the question of otarioid monophyly. (n) Internal acoustic meatus round. A round internal auditory meatus is widespread among terrestrial carnivorans and is unques- tionably primitive at the level suggested here. The partially or completely divided condition seen in certain "otarioids" and phocids is derived (see discussion of character 24 in analysis be- low). The meatus is not round in odobenids, Pinnarctidion, Desmatophoca. or Allodesmus. (p) Bony tentorium in braincase closely appressed to dorsal surface of eminence containing semicircular canals and floccular fossa. As discussed by Wyss (1987: 24), this is the typical carnivoran condition and is undoubtedly primitive. Because the bony tentorium varies widely among the Carnivora (Nojima 1990: table 2) and is difficult to identify, we excluded it from our analysis. Barnes' otarioid "synapomorphies" that occur also in phocids. — The derived state of the following characters occurs also in phocids, a group not included in Barnes" ( 1989) analysis. (b) Proximal limb elements shortened. Phocids have previously been recognized as possessing short proximal limb bones (Weber 1904; Howell 1929), and this character has been identified at a Pinniped Phylogeny 35 more general level, the Pinnipedimorpha, comprising all pinnipeds plus Enaliarctos (Berta et al. 1989). (c) Maxilla forms part of wall of orbit. Wyss ( 19S7) reported that the derived state in whieh the maxilla makes a significant contribution to the medial orbital wall and forms the anterior orbital rim also occurs in phocids. (d) Foramen rotundum and anterior opening of alisphenoid canal combined into one large orbital fissure. Barnes' diagram fails to indicate that Pinnarctidion and Desmatophoca are exceptions. Phocids also share this derived condition (see discussion of charac- ter 19. Appendix 1 ). (f) Sphenopalatine foramen enlarged. This derived state also occurs in phocids (see character 12, Appendix 1). (g) Petrosal isolated from surrounding cranial bones. Repenning ( 1972) discussed this feature as occurring in phocids also. We have not analyzed it because of the difficulty of quantifying it. We observe only subtle differences in this feature among pinnipeds and terrestrial carnivorans. (o) Posterior lacerate foramen enlarged, not expanded trans- versely. The posterior lacerate foramen is enlarged in all phocids as well. In some, however, it is also expanded transversely, but this is apparently a secondary transformation. The condition likely primi- tive for phocids (e.g.. that seen in Monachus) is indistinguishable from the supposed "otarioid" condition. (q) Postglenoid foramen reduced. Phocids are also character- ized by having reduced or lost the postglenoid foramen (see charac- ter 40, Appendix 1 ). (r| Entepicondylar foramen lost from humerus. An entepi- condylar foramen is variably present among phocids. From a previ- ous phylogenetic study of the group (Wyss 19881 and our present analysis, presence of this foramen in phocines and most early fossil "monachines" is probably secondary for the group. (s) Olecranon process of ulna enlarged. As illustrated by Howell ( 1929: fig. 10). phocids possess a condition of the olecranon pro- cess similar to that seen in otariids and odobenids. Barnes' description dubious. — (k) Head lost from incus. The loss of a head on the incus presupposes that a head was once present, which to us seems highly unlikely. By comparison with the outgroups identified here, the head on the incus is a phocid autapomorphy. absent in all other carnivorans. Barnes' (1989) descriptions require modification. — (i) Mastoid process large and cubic. The size and shape of the mastoid process in nonphocid pinnipeds is not significantly changed over the condi- tion in ursids. Wyss (1987) critiqued the use of this feature at greater length. (t) Aquatic propulsion by fore- and hindlimbs, principally the forelimbs. Living pinnipeds swim in two different ways. Otariids generate propulsion principally by use of the forelimbs, whereas phocids and odobenids use principally the hindlimbs (English 1976; Gordon 1981, 1983). It has been argued that the ancestor of pinni- peds (or even the ancestor of "otarioids," if this group should prove monophyletic) likely generated propulsion by using all four limbs, as Enaliarctos probably did (Berta et al. 1989). This argument applies equally to the ancestor of phocids even if they are not related to other pinnipeds. That some distant ancestor of phocids was a four-limb swimmer is indicated by the phocids' forelimbs' being highly modified (used in steering) despite their propelling themselves by the hindlimbs. If phocids had evolved hindlimb swimming directly from a terrestrial ancestor, the forelimbs should not be as highly transformed as they are. In summary, Barnes' analysis does little to bolster the case for otarioid monophyly: indeed, it fails to reveal a single persuasive synapomorphy for the group. We recognize that a proposed otarioid synapomorphy is not automatically invalidated by its appearance in phocids. Plausibly, the Phocidae and "Otarioidea" could be diag- nosed with some of the same (convergently acquired) characters, provided that additional characters demonstrated a phylogenetic separation between the two groups. Historically, the assumed linkage between phocids and mustelids provided this separation, but, as discussed below, recent reviews of characters previously cited in support of this pairing call it into question. The weakness of the evidence supporting the relationship of one pinniped subgroup (phocids) to a terrestrial arctoid lineage (mustelids) to the exclusion of other pinnipeds (a requisite for the acceptance of convergence between otarioids and phocids) leads us to dismiss at least initially apomorphies occurring in both "otarioids" and phocids as necessarily indicating "otarioid" monophyly. Certain "otarioid" synapomorphies might represent convergences: however, we fail to see the logic of accepting this claim in the absence of phylogenetic evidence substantiating a linkage of phocids to some terrestrial lineage of arctoids. Citing Repenning and Tedford ( 1977), Wozencraft (1989: 516) argued that there are "many" synapomorphies supporting a walrus- otariid clade. yet he did not list a single shared derived feature in support of this contention. Of the 1 1 features listed as diagnostic of otarioids in the earlier study, all are primitive or of otherwise unclear phylogenetic significance (Wyss 1987). Thus neither Barnes' nor Wozencraft's analyses identify synapomorphies cor- roborating otarioid monophyly. PINNIPED MONOPHYLY Before addressing the question of pinniped monophyly. we first examine the recent arguments in favor of pinniped diphyly. Accep- tance of pinniped diphyly requires that two criteria be satisfied: evidence of otarioid monophyly and evidence that phocids are related to some nonpinniped terrestrial group. We concluded in the previous section that otarioid monophyly was not well founded. With respect to the second question. Wyss (1987) reviewed the characters used by Tedford (1976) and Muizon (1982a) to unite phocids and mustelids, concluding that no strong case could be made for a mustelid-phocid pairing. Wozencraft ( 1989) argued in favor of a mustelid-phocid link but did not discuss the synapo- morphies supporting nodes on his maximally parsimonious trees. To consider all possible pinniped-terrestrial arctoid pairings we include as outgroups the Ursidae, Mustehdae, Procyonidae, and extinct Amphicyonidae. The monophyly of these groups is gener- ally accepted (Flynn et al. 1988). Principal references for these taxa are as follows: Amphicyonidae, Hunt (1974), Hough (1948); Ursidae, Davis (1964), Beaumont (1965); Mustelidae. Savage (1957), Schmidt-Kittler (1981): Procyonidae, Baskin (1982). Wozencraft and Decker ( 1 99 1 ). METHODS AND MATERIALS Our assessment of relationships among pinnipeds relies upon outgroup comparison. Flynn et al. ( 1988) reviewed the relationship of pinnipeds to other arctoids. proposing two principal hypotheses: pinnipeds as the sister group of ursids and pinnipeds as part of a polytomy with other arctoid families. Berta ( 1991 ) used the Ursidae and the Amphicyonidae as the first and second outgroups to pinnipedimorphs on the basis of their retaining the excavated bassioccipital and presumed loop of the internal carotid artery, a synapomorphy (see Hunt and Barnes 1994. this volume). It is worth mentioning that no extant pinnipeds have the internal carotid loop, and the excavated basioccipital. most extreme in Enaliarctos. is presumably lost. Fortunately, strong postcranial evidence that Enaliarctos is related to pinnipeds supports the presumed loss of this feature at the level of the Pinnipedia. Proponents of both di- and 36 A. Berta and A. R \V\ss monophyly must accept this loss. Thus this feature can in no way be judged to favor a monophyletic Otarioidea. Four synapomorphies link ursids and pinnipedimorphs: (1) shelflike anteromedially placed I*4 protocone. (2) narrow M1 with longitudinally elongated protocone (Flvnn et al. 1988). (3) knoblike acromion process of scapula, and (4) robust olecranon process on ulna (Berta 19911. Wyss and Flynn (1993) used similar evidence to support a sister- group relationship between the Ursoidea (defined as the common ancestor of ursids and amphicyonids plus all of its descendants ) and the Pinnipedia. In addition to living representatives of the three pinniped fami- lies, we include, as terminal taxa, their extinct relatives and indicate their degree of completeness (Table 1). With two exceptions the monophyly of these taxa is generally accepted. On the basis of comparative anatomical evidence Wyss (1988) questioned the monophyly of "Monachus" (indicated by quotes). Berta (1991) recognized Enaliarctos as a metataxon [term formulated by Gauthier (1986); see also Gauthier et al. (1988) and Donoghue ( 1985)] since there is no unambiguous evidence supporting either its monophyly or paraphyly. Initially we included all fossil taxa and in later runs of the data selectively removed them to determine their effect on the tree. PAUP ANALYSIS We scored 143 skeletal character transformations on a taxon- ■ character matrix (Table 2). Of these characters, 73 were craniodental (64 binary and 11 multistate), 52 were postcranial (48 binary and 4 multistate), and 15 were soft anatomical. Some 160 character transformations were possible. We subjected the data to Swofford's ( 1991 ) computer algorithm PAUP, version 3.0s, using the heuristic search option. In all runs multistate characters were entered as unordered. In the initial PAUP run eight characters, 8, 13, 37, 47, 63, 74, 82, and 138. were excluded since they could not be unambiguously polarized. Our initial PAUP analysis considering all fossil taxa resulted in over 100 most parsimonious trees. Principal differences among the 100 trees were in the position of poorly known odobenids including Alachtherium, Dusignathus, Pliopedia, and Pontolis. Later analy- ses excluded these taxa, including only those at least 53% complete. Table I. Completeness of fossil taxa studied as a percentage of the number of characters scored (Appendix 1 ). Percentage Complete Taxon Cranial Postcranial Enaliarctos 93 78 Pteronarctos 94 0 Thalassoleon 93 71 Imagotaria 91 60 Aivukus 77 33 Desmatophoca 90 0 Allodesmus 86 93 Pinnarctidion 53 0 Piscophoca 77 58 At rophoca 60 73 Homiphot a 90 45 Comphotaria 63 45 Alachtherium 16 0 Pontolis 23 0 Pliopedia 0 24 Prorosmarus 4 0 Dusignathus 26 0 We obtained the same result. 100+ most parsimonious trees. Each cladogram had a branch length (BL) of 239, a consistency index (CI) of 0.640, and a rescaled consistency (RC) index of 0.554. The RC excludes autapomorphies from the analysis as well as totally homoplastic characters (see Wiley, et al. 1991 ). A strict-consensus tree is presented in Fig. 2. DISCUSSION Pinniped monophyly is supported by diverse anatomical data. We discuss below the major groupings shown in Fig. 2. The various characters are numbered as in Appendix I . Diagnostic characters for the nodes and terminal taxa in Fig. 2 are listed in Appendix 2. Pinnnipedimorpha Berta et al. (1989) proposed the name Pinnipedimorpha as a term for the monophyletic group including Enaliarctos and the Pinnipediformes. Postcranial (Berta and Ray 1990) and cranial features (Berta 1991 ) have been used to diagnose the group (Table 2, Figs. 3-5 ). We recognize 1 8 unequivocal characters and 6 equivo- cal characters diagnosing the Pinnipedimorpha. Among unequivo- cal synapomorphies are 10 craniodental features: (11) infraorbital foramen large (Fig. 3), (15) anterior palatine foramina anterior of maxillary-palatine suture, (25) round window large with round window fossula developed, (27) basal whorl of scala tympani en- larged, (40) postglenoid foramen vestigial or absent, (43) jugular foramen enlarged, (48) processus gracilis and anterior lamina of malleus reduced. (66) M1 : reduced in size relative to premolars. (67) M1"2 cingulum reduced or absent, and (72) M, metaconid reduced or absent. Unequivocal postcranial synapomorphies of pinnipedimorphs include structural details of the flippers such as (87) greater and lesser humeral tuberosities enlarged (Fig. 4), (88) deltopectoral crest strongly developed (Fig. 4). (90) humerus short and robust. (92) olecranon fossa shallow. (98) digit I on the manus emphasized (Fig. 4). ( 105) digits I and V on the pes emphasized (Fig. 5). ( 1 10) ilium short, and (118) and femoral condyles strongly inclined medially. An additional 16 equivocal synapomorphies might be diagnos- tic of this clade but are subject to equally parsimonious alternative distributions. Seven of these characters, not preserved in Enaliarctos, we assigned to the less inclusive level of the Pinnipedia: (31) cochlear aqueduct large, (49) middle ear cavity and external auditory meatus with distensible cavernous tissue, (54) deciduous teeth fewer. (60) number of lower incisors reduced. (101) metacarpal I longer than metacarpal II, (103) foreflipper claws short, and (104) manus, digit V, intermediate phalanx strongly reduced. The oldest known pinnipedimorph, Enaliarctos, described on the basis of crania and isolated teeth (Mitchell and Tedford 1977; Barnes 1979), is now known from a nearly complete skeleton collected from the late Oligocene or early Miocene Pyramid Hill Sandstone Member of the Jewett Sand in central California (Berta et al. 1989; Berta and Ray 1990). Other crania and associated lower jaws and postcranial elements referred to this taxon are described from deposits of similar age in coastal Oregon (Berta 1991 ). Pinnipediformes The name Pinnipediformes encompasses the ancestor of Pteronarctos and all its other descendants, the Pinnipedia. This group can be diagnosed on the basis of 14 characters, two of which are unequivocal: (14) embrasure pit on palate between P4 and M1 shallow to absent and (24) mastoid process close to paroccipital process, the two connected by a high continuous ridge (state I of multistate character). Pinniped Phytogeny 37 TABLE 2. Distribution among pinniped taxa and relevant outgroups of characters scored. Numbers in heading correspond to numbers of characters in Appendix 1 . 0. Ancestral state; 1 , 2. and 3, derived states; '.'. character missing or not preserved. 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 1 Amphicyonidae 0 0 0 0 0 0 0 9 0 0 0 0 0 0 0 0 0 0 9 0 0 0 0 0 2 Mustelidae 0 0 0 0 0 0 1 0 0 0 0 0/1 0 0 0 0 0 9 0/1 0 0 0 0 3 Procyonidae 0 0 0 0 0 0 1 0 0 0 0 0/1 0/1 0 0 0 0 ? 0 0 0 0 0 4 Ursidae 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 Enaliarctos 0 0 0 0 0 0 0 9 0 0 0 0 1 0 0 0 0 0 0 6 Pteronarctos 0 0 0/1 0 0 0 1 0 0 0 1 0 0 0 0 0 7 Arctocephalus 0 0 1 0 1 1 1 1 0 1 3 0 0 0 0 0 8 Callorhinus 0 0 1 0 1 1 1 1 0 1 3 0 0 0 0 0 9 Otariinae 0 0 1 0 1 1 1 1 0 1 3 0 0 0 0 0 1 0 Thalassoleon 0 0 1 0 1 1 1 1 0 1 3 0 0 0 0 0 1 1 Aivukus 0 0 0 0 1 9 1 0 0 9 1 2 0 0 0 0 0 1 2 Alachitherium 9 9 9 9 9 ? 9 9 9 9 ? 9 9 ? ? 9 9 9 9 9 9 9 9 1 3 Dusignathus 9 ? 9 9 9 ? 9 9 ? 9 9 9 9 ? 9 9 ? 9 9 9 9 0 9 1 4 Gomphotana ? 9 0 9 9 9 1 1 1 0 1 1 0 0 9 0 0 0 0 9 1 5 Imagotaria 0 0 1 0 0 1 1 1 0 0 1 1 2 0 1 0 0 0 0 1 1 6 Odobenus 1 0 2 0 0 1 1 1 1 0 1 1 2 0 1 0 0 0 0 1 1 7 Pliopedia 9 9 9 9 9 ? 9 9 9 9 9 9 9 9 ? 9 9 9 9 ? 9 1 8 Pontolis 9 9 9 9 ? ? 9 ? 9 ? ? 9 ? 9 ? 9 9 0 0 0 1 1 9 Allodesmus 1 1 0 0 2 1 1 0/1 1 0 1 0 1 0 0 0 2 2 0 Desmatophoca 1 1 0 0 2 1 1 0 1 0 1 0 1 0 0 0 2 2 1 Pinnarctidion 9 9 9 9 2 0 ? ? ? 0 1 0 0 0 0 0 2 2 2 Acrophoca 9 9 ? ? 9 9 9 ? 9 0 2 0 9 ? 2 2 3 Cystophora 1 0 0 0 1 1 1 1 9 2 1 1 1 2 24 Erignathus 1 0 0 0 1 1 1 1 1 2 0 1 1 2 2 S Homiphoca 1 0 1 0 1 1 1 1 1 2 0 9 0 2 26 Lobodontini 1 1 0 0 1 1 1 1 1 2 0 1 0 2 27 Mirounga 1 1 0 0 1 1 1 1 1 2 0 1 0 2 2 8 "Monachus" 1 0/1 0 0 1 1 1 1 0 2 0 1 0 2 29 Phocini 1 0/ 1 0 0 1 1 1 1 0 2 1 1 1 2 30 Piscophoca 1 9 0 0 1 1 9 1 0 2 0 9 9 2 25 26 27 2 8 2 9 3 0 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 3 9 4 0 4 1 4 2 4 3 4 4 45 4 6 4 7 4 8 1 Amphicyonidae ? 9 ? 9 0 0 9 ? 0 0 0 0 9 0 0 0 0 0 0 9 9 9 ? 0 2 Mustelidae 0 9 ? 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 Procyonidae 0 9 ? 0 0 9 0 0 0 0 0 0 0/1 0 0 0 0 0 0 0 0 0 0 0 4 Ursidae 0 0 0 0 0 0 0 0 0 0 0 0 0/1 0 0 0 0 0 0 0 0 0 0 5 Enaliarctos 0 0 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 6 Pteronarctos 0 0 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 7 Arctocephalus 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 Callorhinus 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 Otariinae 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 Thalassoleon 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ? ? 1 1 Aivukus ? 9 0 0 9 9 0 0 0 1 0 0 0 1 0 0 9 9 1 2 Alachitherium 9 9 9 0 ? ? 9 9 9 9 ? 9 9 9 9 9 9 0 ? ? ? 1 3 Dusignathus 9 1 9 9 9 ? 9 9 0 9 ? 9 9 9 9 9 9 9 9 9 9 1 4 Gomphotaria 9 9 9 0 ? 9 9 0 9 0 0 0 0 ? 9 9 0 9 ? 9 1 5 Imagotaria 1 1 1 0 0 9 9 0 0 0 1 0 0 0 0 0 ? 1 1 6 Odobenus 1 1 1 0 0 1 1 0 0 0 1 0 0 0 0 0 1 1 7 Pliopedia 9 9 9 9 9 9 ? 9 ? 9 9 9 ? 9 9 ? 9 9 9 ? 9 1 8 Pontolis 9 1 ? 0 9 9 0 0 0 1 0 9 0 0 0 9 9 1 9 Allodesmus 1 1 1 0 9 9 0 0 1 9 2 0 0 1 9 20 Desmatophoca 1 2 1 0 9 ? 0 0 1 9 1 0 0 1 9 2 1 Pinnarctidion ? 2 9 0 9 9 0 0 1 ? 1 0 0 1 ? 9 2 2 Acrophoca 9 2 9 9 ? 9 9 9 1 2 9 2 0 0 9 9 2 3 Cystophora 1 ? 1 1 0 2 1 2 2 1 0 24 Engnathus 1 2 1 1 0 2 0 2 1 0 25 Homiphoca 1 2 1 9 1 2 9 2 0 0 26 Lobodontini 1 2 1 1 1 2 0 2 0 0 27 Mirounga 1 2 1 1 0 2 0 2 0 0 28 "Monachus" 1 2 1 0/1 1 0 2 0 2 0 0 29 Phocini 1 2 1 9 1 0 2 1 2 0/1 2 1 0 30 Piscophoca 9 9 9 ? 9 9 9 9 1 2 9 2 1 0 0 38 A. Berta and A. R. Wvss Table 2 (continued). 49 50 5 1 5 2 5 3 54 55 5 6 57 5 8 59 60 6 1 6 2 6 3 6 4 6 5 6 6 6 7 6 8 69 70 7 1 7 2 1 Amphicyonidae 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 Mustelidae 0 0 0 0 0 0 0 0 1 0 7 0 0 0 1 0 0 0 0 0/ 1 0 0 0 0 3 Procyonidae 0 0 0 0 0 0 0 0 7 0 7 0 0 0 1 0 0 0 0 0 0 0 0 0 4 Ursidae 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 Enaliarctos 7 1 0 0 0 ? 0 0 0 0 0 7 0 0 0 0 1 0 0 1 0 6 Pteronarctos 7 0/1 0 0 0 7 0 0 0 0 0 7 0 0 0 1 1 0 1 0 0 7 Arctocephalus 1 0 0 0 0 1 0 0 0 0 0 3 2 0 0 0 8 Callorhinus 1 0 0 0 0 1 0 0 0 0 0 3 2 0 1 0 9 Otaninae 1 0 0 0 0 1 0 0 0 1 0 3 2 0 0/1 0 1 0 Thalassoleon 7 0 0 0 0 7 0 0 0 0 0 0 2 1 0 0 0 1 1 Aivukus ? 0 7 7 7 7 1 0 7 1 1 3 7 1 0 0 1 2 Alachitherium 9 7 1 0 7 7 7 7 7 7 7 1 7 7 7 7 7 7 0 0 1 3 Dusignathus 7 7 1 0 0 7 7 7 7 7 7 0 3 2 7 1 0 0 1 4 Gomphotaria 7 0 1 0 0 7 1 0 7 7 7 1 7 3 1 7 7 1 0 7 7 7 1 5 Imagotaria 7 0 ? 0 0 7 0 0 7 1 0 0 2 1 1 0 0 0 1 6 Odobenus 1 0 1 0 0 1 1 0 7 7 7 1 3 2 1 1 0 0 1 7 Pliopedia 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 1 8 Pontolis 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 1 9 Allodesmus 7 0 1 1 7 0 0 7 0 7 1 0 3 2 1 0 0 0 20 Desmatophoca 7 0 1 1 7 0 0 7 0 7 0 2 1 /2 1 0 0 1 2 1 Pinnarctidion 7 0 7 7 7 7 0 7 0 7 7 7 0 0 0 1 7 7 7 7 7 7 2 2 Acrophoca ? 0 1 1 7 7 7 7 7 1 7 0 0 2 2 1 7 0 0 2 3 Cystophora 2 0 ? 0 1 1 0 1 7 0 0 3 2 1 0 0 24 Erignathus 2 0 7 0 1 0 1 7 0 0 3 2 1 0 0 25 Homiphoca ? 0 1 7 7 1 1 7 7 0 0 2 2 1 0 0 26 Lobodontini 2 0 1 0 1 1 1 1 0 0 3 2 1 0 0 27 Mirounga 2 0 1 0 1 1 1 1 0 0 3 2 1 0 0 28 "Monachus" 1 /2 0 1 0 1 1 1 1 0 0 3 2 1 0 0 29 Phocini 2 0 7 0 1 0 0 1 0 0 0 3 2 1 0 0 30 Piscophoca ? 0 1 1 7 1 1 7 1 0 0 2 1 1 0 0 7 7 7 3 7 4 7 5 76 7 7 7 8 7 9 80 8 1 8 2 8 3 8 4 8 5 8 6 8 7 8 8 8 9 9 0 9 1 | 9 2 9 3 9 4 I 9 5 9 6 1 Amphicyonidae 0 0 0 ? ? 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 Mustelidae 0 0 7 ? 0 0 0 0 0/1 0 0 0 0 0 0 0 0 0 0/1 0 0 0 0/1 3 Procyonidae 0 0 ? ? 0 ? ? 0/1 0 7 ? 0 0 0 0 0 0 0 0 ? 0 0 0 4 Ursidae 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0/1 0 0 0 0 0 5 Enaliarctos 0 0 ? ? 0 0 1 0 0 0 0 1 ? 1 0 1 1 1 0 0 0 1 6 Pteronarctos 0 0 7 ? 0 7 7 7 ? 7 ? ? 7 ? 1 ? ? ? ? ? ? 7 ? 7 Arctocephalus 0 0 0 0 0 0 1 0 0 0 1 1 1 1 1 1 0 1 1 1 8 Callorhinus 0 0 0 0 0 0 1 0 0 0 1 1 1 1 1 1 0 1 1 1 9 Otaninae 0 0 ? 0 0 0 1 0 0 0 1 1 1 1 1 1 0 1 1 1 1 0 Thalassoleon 0 7 ? ? 7 ? 1 0 0 0 1 1 1 1 1 1 0 1 1 1 1 1 Aivukus 0 7 ? ? 7 7 7 7 ? 7 7 ? 1 1 1 1 1 1 1 1 2 1 2 Alachitherium 0 ? ? 7 7 ? ? ? ? ? ? ? ? 7 7 ? 7 7 ? ? 7 7 1 3 Dusignathus 1 ? ? 7 7 ? ? 7 7 ? ? ? ? 7 ? ? ? ? 7 ? 7 ? 1 4 Gomphotaria 0 ? ? ? ? ? ? 0 0 0 1 7 ? 1 1 1 1 1 ? 1 1 ? 1 5 Imagotaria 0 7 ? 0 7 7 7 0 0 0 1 0 1 1 1 1 1 1 1 1 1 2 1 6 Odobenus 0 1 1 1 1 1 0/1 0 0 0 1 0 1 1 1 1 1 1 1 1 1 2 1 7 Pliopedia 7 7 7 ? ? ? 7 7 7 7 ? ? ? 7 1 1 1 1 1 1 1 1 1 2 1 8 Pontolis 7 7 7 7 ? ? ? ? 7 ? 7 ? 7 7 7 ? ? 7 ? ? ? ? ? 7 1 9 Allodesmus 0 0 0 0 0 1 1 0 0 0 1 ? 1 1 1 1 1 1 0 1 1 1 20 Desmatophoca 0/1 ? 7 ? ? ? ? ? ? 7 ? 7 ? ? 7 ? 7 ? ? ? ? 7 2 1 Pinnarctidion 7 7 0 ? ? ? ? ? 7 ? ? 7 7 ? ? ? ? ? 7 ? 7 ? ? 7 2 2 Acrophoca 1 1 1 1 1 ? 0 ? ? 1 0 1 1 1 1 1 1 0 1 1 2 2 3 Cystophora 1 1 0 1 1 1 1 1 0 0 0 1 2 0 1 0 1 0 1 1 1 2 4 Erignathus 1 1 0 1 1 1 1 1 0 0 0 1 2 0 1 0 1 1 1 1 1 25 Homiphoca ? ? 7 7 ? ? 7 7 7 ? ? ? 1 1 ? 1 0 1 7 1 1 2 26 Lobodontini 1 1 1 1 1 1 0 2 1 0 1 1 1 1 1 1 0 1 1 1 27 Mirounga 1 1 1 1 1 0 0 1 1 0 1 1 1 1 1 1 0 1 1 2 2 8 "Monachus" 1 1 7 1 1 0 0 1 1 0 1 1 1 1 1 1 0 1 1 2 29 Phocini 1 1 0 1 1 1 1 0 0 0 1 2 0 1 0 1 1 1 1 1 30 Piscophoca 1 1 1 ? 7 1 7 1 1 0 1 1 1 1 1 1 0 1 1 2 Pinniped Phylogeny 39 Tablh 2 (continued). 97 9 8 9 9 1 00 1 0 1 1 02 1 03 1 0 4Jl 0 5 1 0 6 1 07 1 0 8 1 09 1 1 0 1 1 1 1 1 2 1 1 3 1 1 4 1 1 5 1 1 6 1 1 7 1 1 8 1 1 9 1 20 1 Amphicyonidae 9 0 0 0 9 7 9 ? 0 9 0 0 0 0 0 0 0 0 9 0 0 0 0 7 2 Mustelidae 0 0 0 0 0 7 0 ? 0 0 0 0 0 0 0 0 0 0 9 9 0 0 0 ? 3 Procyonidae 0 0 0 0 0 7 0 ? 0 0 0 0 0 0 0 0 0 9 0 0 ? ? ? 0 4 Ursidae 0 0 0 0 0 ? 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 Enaliarctos 0 1 0 0 7 ? ? 7 0 7 0 0 0 0 0 0 0 0 0 1 0 1 6 Pteronarctos ? ? 7 ? ? 7 ? 7 ? 7 7 ? ? 7 ? ? ? ? ? ? 7 7 9 7 Arctocephalus 0 0 1 1 1 1 7 0 0 0 0 0/1 0 1 0 1 1 1 8 Callorhinus 0 0 1 1 1 1 0 0 0 0 1 0 1 0 1 1 1 9 Otariinae 0 0 1 1 1 1 0 0 0 0 0 0 1 0 1 1 1 1 0 Thalassoleon 0 0 1 1 ? 7 ? 7 0 0 0 0 1 0 1 0 1 1 7 1 1 Aivukus ? 2 1 7 7 7 ? 7 ? 7 7 ? 0 7 ? 1 ? ? 7 ? 7 1 2 Alachitherium ? ? 7 7 7 7 ? 7 7 7 7 ? ? ? 9 ? 7 ? 7 9 ? 7 1 3 Dusignalhus ? ? 7 7 ? 7 ? ? 7 7 ? 7 ? ? ? 7 ? 7 ? 9 9 1 4 Gomphotana ? ? 1 7 7 7 7 ? ? ? 1 9 7 0 0 7 1 7 1 1 1 1 5 Imagotaria 1 1 1 7 ? ? 7 ? ? ? 1 ? 7 0 7 7 1 0 1 1 0 1 6 Odobenus 0 2 1 1 1 1 0 1 1 0 0 1 1 1 1 1 1 1 7 Pliopedia 7 1 1 7 ? ? ? 7 ? ? 7 7 ? ? 7 ? 7 7 ? 7 7 1 8 Pontolis 7 ? 7 7 9 ? 9 7 7 ? 7 ? ? ? 9 ? 9 ? ? 7 1 9 Allodesmus 0 0 1 1 1 7 0 ? 1 0 0 1 0 1 1 1 1 20 Desmatophoca 7 ? 7 7 7 ? ? 7 ? -? ? ? 7 ? 9 9 9 9 2 1 Pinnarctidion 7 ? ? 7 ? 7 ? ? ? ? ? ? ? ? ? 7 ? 7 2 2 Acrophoca 7 0 1 7 ? 7 1 ? 1 0 1 0 1 1 7 23 Cystophora 1 0 0 0 0 0 0 1 0 0 0 1 0 1 1 0 24 Erignathus 1 0 0 0 0 0 0 0 0 0 0 1 0 1 1 0 2 5 Homtphoca 7 7 7 7 7 ? ? ? ? ? 0 1 0/1 0 1 1 7 2 6 Lobodontini 0 0 1 1 0/1 1 1 1 1 1 0 1 0 1 1 0 2 7 Mirounga 0 0 1 1 0 1 1 1 1 1 0 1 0 1 1 0 28 "Monachus" 0 0 1 1 0 1 0/1 1 1 1 0 1 0/1 1 1 0 29 Phocini 1 0 0 0 0 0 0 0 0 0 1 1 0 1 1 0 30 Piscophoca 7 0 1 ? ? ? 7 ? ? 1 0 1 0 7 ? ? ? 0 7 1 2 1 1 22 1 23 1 2 4h 2 5 1 26 1 27 1 2 8Jl 29I1 3 0 1 3 1 1 3 2I1 3 3 1 34 1 3 5|l 3 6|l 3 7|l 3 8 1 39 1 40 1 4 1 1 42 1 43 1 Amphicyonidae 0 0 0 0 0 0 0 7 7 9 7 7 7 9 9 7 7 7 9 7 7 9 7 2 Mustelidae 0 0 0 0 0 0 0 0 0 0 0 7 ? ? 0 7 7 0 0 7 0 0 0 3 Procyonidae ? 0 0 0 0 0 7 0 0 0 ? ? 9 7 0 7 7 7 0 7 0 0 0 4 Ursidae 0 0 0 0 0 0 0 0 0 0 9 ? 0 7 0 ? 0 7 0 7 0 0 0 5 Enaliarctos 0 0 7 7 0 0 7 ? ? 9 9 ? 0 9 ? 7 7 7 7 7 7 7 7 6 Pteronarctos 7 ? ? ? ? ? ? 7 ? ? ? 7 9 ? 7 7 7 7 7 7 7 9 9 7 Arctocephalus 0 1 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 8 Callorhinus 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 9 Otariinae 0 1 1 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 1 1 0 Thalassoleon 0 0/1 0 7 0 0 ? ? ? ? ? 0 ? 7 7 7 7 7 7 7 7 7 9 1 1 Aivukus ? ? ? 7 7 ? ? ? ? ? 7 ? ? ? 7 7 7 ? 7 7 7 7 9 1 2 Alachitherium ? ? ? 7 ? ? 7 7 ? ? ? ? 7 ? ? 7 ? 7 ? 7 7 7 9 1 3 Dusignathus ? ? ? ? 7 7 ? ? 7 ? 7 7 7 7 7 ? 7 ? 7 7 9 7 9 1 4 Gomphotaria ? 0 ? 1 1 1 7 7 ? 1 7 7 ? 9 ? 7 9 9 7 9 7 7 9 1 5 Imagotaria 0 ? 0 1 1 1 7 7 7 7 7 7 7 ? ? 7 7 7 7 7 7 7 7 1 6 Odobenus 0 0 0 1 1 1 1 2 0 1 0 ? 0 1 0 1 1 1 0 0 1 7 Pliopedia ? ? ? ? ? 7 ? ? ? ? ? 9 ? ? 7 7 ? 7 7 9 7 7 7 1 8 Pontolis 7 ? 7 ? ? ? ? ? ? ? 7 7 ? ? 7 ? 7 ? 7 7 7 7 1 9 Allodesmus 0 0 0 1 0 1 ? ? 7 7 7 7 ? ? 7 9 7 9 7 7 7 20 Desmatophoca ? ? 7 ? ? ? ? ? ? 7 7 ? ? 7 7 ? 9 7 7 7 7 2 1 Pinnarctidion ? ? 7 ? 9 ? 7 ? ? ? ? 7 ? ? 7 7 7 7 7 7 9 22 Acrophoca 0 1 0 1 0 2 ? 7 7 ? 7 ? 7 7 ? 7 7 7 7 9 2 3 Cystophora 1 1 0 1 0 2 1 1 1 1 1 0 0 1 1 1 1 1 1 0 24 Erignathus 1 1 0 1 0 2 1 1 1 1 0 0 0 1 0 1 1 1 1 0 2 S Homiphoca 1 1 ? ? 7 ? ? ? ? ? ? 7 7 7 7 7 7 7 7 9 26 Lobodontini ? 1 0 1 0 2 1 1 1 1 0/1 0 0 1 1 1 1 1 1 0 27 Mirounga 0 1 0 1 0 2 1 0/1 2 0 1 1 1 0 1 1 1 1 1 1 0 28 "Monachus" 0 0/1 0 1 0 2 1 2 0 1 0/1 1 0 1 0 1 1 1 0/1 0 29 Phocini 1 1 0 1 0 2 1 1 1 1 1 0 0 1 1 1 1 1 1 0 30 Piscophoca ? ? ? 1 0 2 ? ? ? ? ? ? ? ? ? ? ? ? ? 9 40 A. Berta and A. R. Wyss 3 O 4-» 3 o V) O c o 10 O M