HMH!HHHIIHWWIM HUME 138 230 “THS Wests \ ..)" . margaw This is to certify that the thesis entitIed METAPODIAL RESTRUCTURING IN MAMMUT AND RECENT ELEPHANTS: EVIDENCE OF DISEASE OR PHYSICAL STRESS? presented by KATHLYN MAI SMITH has men accepted towards fulfillment of the requirements for the MS. degree in Geological Sciences MW Major Professor's Signature Date MSU is an Affirmative Action/Equal Opportunity Institution -.-.-.‘4.-.-.-.-‘-.-.-.-._._.-._l_ - -4 JI-t--A- LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 c:/CIRC/DateDue.p65-p. 15 METAPODIAL RESTRUCTURING IN MIMI/[UT AND RECENT ELEPHANTS: EVIDENCE OF DISEASE OR PHYSICAL STRESS? By Kathlyn Mai Smith A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geological Sciences 2004 ABSTRACT METAPODIAL RESTRUCTURING IN WW AND RECENT ELEPHANTS: EVIDENCE OF DISEASE OR PHYSICAL STRESS? By Kathlyn Mai Smith A variety of pathologies have been noted on Mammut americanum (American mastodon), including subchondral articular surface undermining on metapodials (documented in extinct bison). Metapodial undermining in mastodons has been ascribed to tuberculosis (Mycobacterium tuberculosis), but this diagnosis has not been tested by comparisons with Recent material. The following questions are here addressed by comparisons with Recent proboscideans: (1) To what degree do Recent elephants show undermining? (2) Does undermining vary based on species, age, sex, or whether the animal is wild or captive? (3) Does undermining preferentially affect a specific metapodial? and (4) Can this undermining be firmly linked to tuberculosis in Recent elephants, and be used to interpret the presence of the disease in the American mastodon? To answer these questions, 165 metapodials from 17 skeletal specimens of Recent elephants were examined for the presence of undermining. Of the 165 metapodials studied, 103 (62%) had undermining. The third metapodial most often had undermining (31 of 35; 89%). Undermining differentially affected adult elephants; no juvenile elephants in the study had undermining. Because undermining in Recent elephants affected only adults, and was most often present on the central metapodials, it is likely the result of pressure on the feet from the weight of the animal, and a normal part of Recent proboscidean skeletal anatomy, rather than the result of an infection. ACIG‘IOWLEDGMENTS I would like to thank my advisor, Dr. Michael Gottfried, and my committee members, Dr. J. Alan Holman and Dr. Danita Brandt, for their assistance and guidance throughout this project. Thank you to Dr. Bruce Rothschild (Arthritis Center of Northeast Ohio), for supporting a project based on his original research, and thank you to Dr. Daniel Fisher (University of Michigan), for his many helpful insights, and for lending me the idea for this thesis. This project would not have been possible without the assistance of Collection Managers. Thank you to the Michigan State Museum of Natural History’s Laura Abraczinskas, the Field Museum’s Bill Stanley, and the National Museum of Natural History’s Linda Gordon and Dave Schmidt, who were all extremely generous with their time in order to help me get to the specimens I needed. I would like to acknowledge the College of Natural Science, for awarding me the grant that made my trip to the National Museum of Natural History possible. Finally, I am grateful to my fellow graduate students, my family, and II, who have all supported me throughout this project. iii TABLE OF CONTENTS LIST OF TABLES ........................................................................................................... v LIST OF FIGURES .......................................................................................................... vi KEY TO ABBREVIATIONS .......................................................................................... viii INTRODUCTION ........................................................................................................... 1 The Order Proboscidea ....................................................................................... 2 Extant Proboscideans .......................................................................................... 7 Paleopathology in Proboscideans ....................................................................... 13 Elephant foot skeletal anatomy ........................................................................... 15 Tuberculosis ......................................................................................................... 20 Tuberculosis in elephants .................................................................................... 23 Tuberculosis on bone ........................................................................................... 24 Elephant foot care ............................................................................................... 26 MATERIALS AND METHODS ..................................................................................... 27 RESULTS ........................................................................................................................ 29 Metapodial features ............................................................................................. 29 Articular surface undermining on mastodons ..................................................... 29 Description of elephant specimens ...................................................................... 30 Frequency of undermining on foot bones ............................................................ 38 Preferential occurrences of undermining ............................................................ 44 DISCUSSION .................................................................................................................. 46 Frequency of undermining ................................................................................... 46 Articular surface undermining and lawn-specificity .......................................... 49 Articular surface undermining and sexual dimorphism ...................................... 49 Articular surface undermining and age ............................................................... 50 Articular surface undermining and habitat ......................................................... 50 Articular surface undermining and tuberculosis in Recent proboscideans ........ 52 Articular surface undermining and tuberculosis in fossil proboscideans ........... 53 CONCLUSIONS .............................................................................................................. 54 Future work ......................................................................................................... 55 APPENDD( ...................................................................................................................... 57 REFERENCES ................................................................................................................ 67 iv LIST OF TABLES Number Page 1. Summary of specimens ............................................................................................... 27 2. Summary of foot bones ............................................................................................... 4O LIST OF FIGURES Number Page 1. Comparison of elephant and mastodon feet ................................................................ 3 2. Phylogenetic relationships within Order Proboscidea ................................................. 4 3. Skeletal differences between mammoths and mastodons ........................................... 6 4. Present-day distribution of African elephants ............................................................. 8 5. Present—day distribution of Asian elephants ................................................................ 9 6. African elephant subspecies ........................................................................................ 11 7. Asian elephant subspecies ........................................................................................... 12 8. The elephant manus ..................................................................................................... 16 9. The elephant pes .......................................................................................................... 18 10. Elephant metapodials ................................................................................................. 19 11. (A) Location of distal articular surface on metapodials ............................................ 21 (B) Foot bone dimensions and location of ventral depression .................................. 21 (C) The appearance of articular surface undermining ............................................... 21 (D) The appearance of articular surface lipping ........................................................ 21 (E) Articular surface undermining on a tuberculosis-infected elephant .................... 21 12. (A) “Porous” bones of USNM 588113 ..................................................................... 34 (B) Separated distal articular facets on USNM 588113 ............................................ 35 (C) “Bunching” on USNM 49639 ............................................................................. 35 13. (A) Histogram of metapodials affected by articular surface lipping ......................... 41 (B) Histogram of metapodials affected by articular surface undermining ................ 41 14. (A) Histogram of metacarpals affected by articular surface lipping ......................... 42 (B) Histogram of metacarpfls affected by articular surface undermining ................ 42 15. (A) Histogram of metatarsals affected by articular surface lipping ......................... 43 (B) Histogram of metatarsals affected by articular surface undermining ................ 43 vii L-CAL R/L-MPI (IIJILIVN) BIL-MCI (II,III,IV,V) R/L-MTI (HJIIJVN) RM/LM-PPI (II,III,IV,V) RP/LP-PPI (II,III,IV,V) LM-IPI (II.III.IV.V) KEY TO ABBREVIATIONS left calcaneurn right or left metapodial one (two, three, four, five) right or left metacarpal one (two, three, four, five) right or left metatarsal one (two, three, four, five) first proximal phalanx of the right or left menus (second, third, fourth, fifth) first proximal phalanx of the right or left pes (second, third, fourth, fifth) intermediate phalanx of the left manus (second, third, fourth, fifth) INTRODUCTION One of the most notable members of North America’s Pleistocene fauna is Mammut americanum, the American mastodon. Mastodons are found throughout the continent, but are concentrated in the Great Lakes region (King and Saunders, 1984; Tassy and Shoshani, 1988; Holman, 1995); in Michigan alone mastodon fossils have been recovered from over 250 locations across the state (Abraczinskas, 1993 ), leading Michigan to adopt it as the state fossil in 2002. Paleopathological analyses of mastodons indicate that they suffered from a variety of diseases and injuries (Bricknell, 1987; Fisher, 1984; Rothschild et al., 1994; Rothschild and Helbling, 2001). Pathologies, and non-pathological injuries, attributed to mastodon skeletal elements include arthritis, butchering scars, periodontal disease, and, most recently, tuberculosis. Tuberculosis, as caused by the microbiologic agent Mycobacterium tuberculosis, has, in addition to mastodons, been identified in humans, extinct bison, captive elephants, and in pets and other captive animals in contact with humans (Rothschild et al., 2001; Hoop, 2002; Lomme et al., 1976; Powers and Price, 1967). Rothschild and Helbling (2001) identified tuberculosis in the fossil record on the basis of subchondral articular surface undermining in metapodials (Rothschild and Helbling, 2001), but this has not yet been tested by comparison with Recent material. Despite splitting into separate phyletic lineages 20 million years ago (Tassy, 1996), mastodons and elephants retain similar body plans and skeletal anatomy. The proboscidean skeleton is constructed as a graviportal support system. Adaptations for this body type include columnar limbs, a light skull with a large surface area for muscle attachment, the replacement of a bone marrow cavity with a network of dense, cancellous l bone, and mesaxonic limbs (Shoshani, 1996, and references therein). The difference between mastodon and elephant feet is that mastodons have stockier foot bones, while elephants generally have more slender foot bones (Figure 1). In metapodials, the elements focused on in this study, the ratio of heightzlength is generally less in mastodons than it is in elephants. The number of digits, stance, and general shape of the foot bones for both groups of proboscideans remain the same. The objective of this research is to evaluate subchondral articular surface undermining in mastodon metapodials by comparison with the skeletal anatomy of Recent proboscideans. The following questions will be addressed: (1) To what degree do Recent elephants show the undermining? (2) Does the undermining vary based on species, age, sex, or whether the animals are wild or captive? (3) Does articular surface undermining preferentially affect a specific metapodial? and (4) Can this undermining be firmly linked to tuberculosis in Recent elephants, and thus be used to interpret the presence of the disease in fossil proboscideans, including the American mastodon? The Order Proboscidea Proboscideans are fairly well-represented in the fossil record (Figure 2). Shoshani and Tassy (1996) recognized 8 to 9 families (depending on the inclusion or exclusion of Anthracobune, tentatively considered the oldest Proboscidean), 38 genera, and 162 species. The Order Proboscidea is traditionally accepted to have first appeared in Africa during the Early Eocene with the extinct genus Moeritherium (Fischer and Tassy, 1993; Thewissen and Domning, 1992; Tassy, 1996). Fossil material of Moeritherium shows that it had a variety of proboscidean characteristics, including: the loss of the lower 2 Figure 1. The L-MTIII of MSUVP 1289 (A), an American mastodon, and the L-MTIII of USNM 163318 (B), an African elephant. Note that the mastodon is stouter, and has a wider distal articular surface. Both metatarsals show articular surface undermining. Scale bar = 2 cm. s... . 5:333 1 ... . . ._ . astofieemgz i i in 8£§0~W Ila" E: . _ . mas—Ewan .. .. J L IIIIIII I. _ «35:55:: I.— r. n 332...... -- 3 u . _ IIIIII _ u u .u _ u _ u “ 3.8593306 . n _ . _ m u u u .m _ _ . + ...... .. 2:555:32 _ M u l... k . a " oétofiofioe IIIIII . m .W W m . .0. mm m m m H W . o Figure 2. The phylogenetic relationships within Order Proboscidea. Adapted from Shoshani and Tassy (1996) and Thomas et al. (2000) incisor, the loss of the lower canine, the loss of the upper and lower first pre-molar, greatly enlarged, tusk-like second upper and lower incisors, the flattening of the femur, the absence of a saggital crest, and the absence of pneumatization in the crania, among other characters as noted by Tassy (1996), based in part on work by Andrews (1906). The Superfamily Elephantoidea includes the families Mammutidae, Gomphotheridae, Stegodontidae, and Elephantidae. This group shows tooth displacement (Tassy and Shoshani, 1988), in which the cheek teeth erupt behind teeth already in use, and push forward until they become functional; they eventually are replaced themselves and resorbed at the anterior end of dentition. This displacement continues until the third molars are the only teeth present at the end of the full sequence. This study focuses on the families Mammutidae and Elephantidae. Mammutidae emerged in Africa during the Early to Middle Miocene; characteristics of the family include a broad, low cranium, a shortened mandible, and a laterally compressed rostrum. The group reached North America by the late Pliocene with Mammut, the first genus to deveIOp subhypsodont cheek teeth (Saunders, 1996). The Family Elephantidae first appeared in the Late Miocene of Africa (Thomas et al., 2000). Prior to Elephantidae, proboscideans chewed in a grinding and shearing motion; this shifted to horizontal shearing with a fore and aft movement of the jaw in elephantids (Maglio, 1972). Elephantidae includes the iconic Ice Age form, the mammoth (Mammuthus), and the two genera of Recent elephants, Loxodonta and E lephas. There are numerous skeletal differences between mammoths and mastodons. As described by Shoshani (1992), mastodons (Figure 3A) have stockier, heavier bodies; the 5 ‘ 3‘ ., I '5 4". . . a "a. .5: f; ‘3‘ i 'f i . i . I- " z? , s‘ ," . 2| , ‘- I . " W t 2" ti . “ . " } I . B. Jefferson mammoth (Mammuthusjeflersom‘). Figure 3. Skeletal differences between (A) mastodons and (B) mammoths. From Skeels (1962). ' head and shoulder is only slightly above the hindquarters. They have 20-21 thoracic vertebrae, low-domed skulls, straighter tusks, bunodont chewing surfaces, and brachyodont teeth, and they use a grinding surface to chew. By contrast, mammoths (Figure 3B) are more delicately built, with the head and shoulder well above their hindquarters. They have fewer thoracic vertebrae, high-domed skulls, more curvaceous tusks, shorter mandibles, lophodont chewing surfaces, and hypsodont teeth, and they use a grinding motion to chew. Different dental morphologies correspond to the environments inhabited by each taxon. Palynological evidence collected in Michigan for the interstadial intervals of the Wisconsinan glaciation indicate that the environment was boreal forest or forest tundra (Hohnan et al., 1986). This environment was perfect for mastodons, with over 250 specimens found in the state to date (Abraczinskas, 1993). The mastodon adapted well to a range of habitats, and has been termed an ecological generalist. On the contrary, the mammoth was an ecological specialist (Shoshani, 1989), and was better adapted to live in grasslands and treeless tundra-steppe areas (Agenbroad and Mead, 1996). Extant Proboscideans The two surviving genera of Proboscidea are found in areas throughout Africa and Asia (Figures 4 and 5). There are two subspecies of African elephants (Loxodonta africana afiicana and Loxodonta afiicana cyclotis) and three subspecies of Asian elephants (Elephas maximus maximus, Elephas maximus indicus, and Elephas maximus sumatranus). L. a. africana is commonly referred to as the savanna elephant. It is the 7 / wfir I N n F" I an I \w \ \w inr— Eastern Africa 8. Ethiopia __ 9. Kenya 10. Rwanda 39; ,J 13. Tanzania u , . P- . Gambia .193. Western Africa i I v ”o Benin 0 -- - -- 24. Burkina Faso fl 4:- 25. Ghana 26. Guinea Central 12: Z; firmware?“ 2. Central African 29. Liberia Republic 30. Mali 2: 32:0 1. Mauritania . ___ 32 Niger 5. Equatonal Guinea 33. Nigeria 6. Gabon e 34- Senesal 7 2““ 35 Sierra Leone & 36 Togo 8 15 Angola 19 Narmbta ._.. i 1 I I 16 Botswana 20 South Africa 17. Malawi 21. Zambia I500 Irm 18. Mozambique 22- Zimbabwe Figure 4. The present distribution of African elephants (Loxodonta afi'icana). From Douglas-Hamilton and Michelmore (1996) IIIIIJ W 1000km W Churn Bhutan . “N 0 i a: ‘1‘ 1R” i: INDIAN OCEAN Figure 5. The present distribution of Asian elephants (Elephas maximus) elephants. From Sukumar and Santiapillai (1996). largest living land animal, weighing between 8800 and 15,400 pounds, and standing between 9.8 and 13 feet tall (Shoshani, 1992). It has dark skin, hair around the trunk and mouth, triangular ears, and curved, thick tusks (Figure 6A). The forefeet have 4 toenails, and the hind feet have 3 toenails Ramsay and Henry, 2001). The savanna elephant lives in various habitats in Africa, south of the Sahara desert, including open grasslands, forests, deserts, marshes, and lake shores. L. a. cyclotis is commonly referred to as the forest elephant. As noted by Shoshani (1992), it is smaller than the savanna elephant, weighing between 4400 and 9900 pounds, and standing between 6.6 and 9.8 feet tall. It is lighter in color and has less hair; its ears are rounder and smaller, and its tusks are straighter and more slender (Figure 6B). The forefeet have 5 toenails, and the hind feet have 4 toenails (Ramsay and Henry, 2001). The forest elephant lives mainly in equatorial forested regions in central and western Africa, as well as in intermediate zones between forests and grasslands. E. m. maximus is commonly referred to as the Sri Lankan elephant. As described by Shoshani (1992), it weighs between 4400 and 12,100 pounds, and stands 6.5 to 11.5 feet tall. It is the darkest Asian elephant, with large ears, and depigmentation patches on its ears, face, trunk, and belly (Figure 7A). The forefoot has five toenails, and the hind foot has four toenails (Ramsay and Henry, 2001). This subspecies lives only on the island of Sri Lanka, and exists in a variety of habitats, including open grasslands, forests, transitional areas, open savannahs, marshes, and lake shores, from sea level to the mountains. E. m. indicus is the mainland Asian elephant subspecies. Shoshani (1992) notes its weight between 4400 and 11,000 pounds, and height between 6.5 and 11.5 feet. 10 A. The savanna African elephant (L .a. qfricana). B. The forest African elephant (L. a. cyclotis). Figure 6. African elephant (Loxodonta africana) subspecies: (A) savanna elephant (L. a. afi'r'cana); (B) forest elephant (L. a. cyclotis). From Shoshani (1992). A The Sri Lankan Asian elephant (E. m. mwrimus). B. The mainland Asian elephant (E. m. indicm). C. The Sumatran Asian elephant (E. m. sumatranus). Figure 7. Asian elephant (Elephas maximus) subspecies: (A) Sri Lankan elephant (E. m. maximus); (B) mainland elephant (E. m. indicus); (C) Sumatran elephant (E. m. sumatranus). From Shoshani (1992). 12 than the Sumatran elephant; its ears are of variable size (Figure 7B). There are five toenails on the forefoot, and four toenails on the hind foot (Ramsay and Henry, 2001). This elephant lives in 12 mainland countries, from India in the west to Indonesia in the east, and prefers forested areas and transitional zones between forests and grasslands; it can live from sea level to 2000 meters. E. m. sumatranus is the Sumatran elephant. Shoshani (1992) lists its weight as between 4400 and 8800 pounds, and height between 6.5 and 10.5 feet. It is the lightest- colored elephant , and has disproportionately large ears (Figure 7C). Unlike the other Asian subspecies, which have 19 pairs of ribs each, the Sumatran elephant has 20 pairs of ribs. It has five toenails on the forefoot, and four toe nails on the hind foot (Ramsay and Henry, 2001). The Sumatran elephant lives only on the island of Sumatra, mainly in forests and patchy habitats. Paleopathology in proboscideans Paleopathology is the study of ancient diseases. An individual can be affected throughout its lifetime with a variety of diseases, illnesses, and injuries. If these ailments cause bone to restructure, paleopathological interpretations can be made from skeletal remains. Pathologies, and non-pathological injuries, may show up as bony signatures, including lesions, fusings, erosions, fractures and breaks. Previous studies on skeletal elements of fossil proboscideans show that they suffered from a variety of pathologies, including osteoarthritis, spondyloarthropathy, periodontal disease, diffuse idiopathic skeletal hyperostosis, and tuberculosis (Rothschild et al., 1994; Rothschild and Helbling, 2001; Bricknell, 1987). Osteoarthritis is generally observed as a bony overgrowth on the 13 2001; Bricknell, 1987). Osteoarthritis is generally observed as a bony overgrowth on the zygapophyseal joints; severe arthritis may cause grooving of articular surfaces and the changing of bone into an ivory-like mass at the surface. Evidence for spondyloarthropathy includes erosions on the dorsal superior and dorsal inferior borders of the lower thoracic and lumbar vertebrae, giving the vertebral centrum a squared appearance; vertebral fusion can also be a sign of spondyloarthropathy. Periodontal disease causes a ridge and groove where the tooth and gums meet, giving the jaw a lumpy appearance. Diffuse idiopathic skeletal hyperostosis is identified by the ossification of the ligaments between vertebrae, and resembles melting wax. Tuberculosis has been identified in mastodons on the basis of a periosteal reaction on ribs, and subchondral articular surface undermining on foot bones. Evidence of bone restructuring, caused by disease, injury, or everyday activity, can provide information on lifestyle and habitat. For example, two Columbian mammoths (Mammuthus columbi) were discovered with their tusks interlocked (Shultz, 1963). The orientation of the skeletons, in conjunction with the marks on the tusks, suggests that the mammoths were engaging in battles similar to those between bull African elephants (Rothschild and Martin, 1993). Another instance of tusk use can be inferred from wear pattems. African elephants fell trees and break them with their tusks. Living elephants tend to favor one tusk or the other when performing this act, and mastodons likely did the same. When a mastodon skull is recovered with both tusks intact, the tusk on one side is typically shorter or more worn down than the other, suggesting that mastodons were either “right- or left-tusked” (Holman, 1975). Most evidence of disease on bone has been documented on human skeletons, and 14 it is a challenge to identify the disease in non-human vertebrates, as a single disease can affect different vertebrates in distinctly different ways. Thus, it is important to do further testing for the diagnosis of diseases on fossil material based on physical characteristics, in order to assign bone reactions to particular diseases. Rothschild et al. (2001) performed such an analysis on the foot bones of an extinct bison. The pathologies identified on the foot bones were suggestive of tuberculosis. Fragments of DNA from the area affected by pathology were isolated and sequenced. Sequencing of these fragments identified the DNA of a member of the M. tuberculosis complex, confirming the association of the physical character attributed to tuberculosis with the putative infection agent. Elephant foot skeletal anatomy The structure and components of the feet of African and Asian elephants are similar, yet the higher frequency of foot problems in captive Asian elephants suggests an unrecognized biological difference (Ramsay and Henry, 2001). The elephant manus (Figure 8) is semi-digitigrade (Fowler, 2001); it consists of carpals, metacarpals, and phalanges. The elephant carpus comprises two block-like stacks of four bones each The distal four carpals articulate with corresponding metacarpals I through IV, while MCV also articulates with the fourth carpal. In African elephants, the first digit has one phalanx and one sesamoid bone; in Asian elephants, the first digit has two phalanges and one sesamoid bone. Digits two, three, and four each have three phalanges; the fifth digit has two phalanges. Digits two through five each have paired sesamoid bones that articulate with the posterior distal articular surface on metacarpals two through five 15 C Figure 8. The African elephant manus; USNM 49489B: (A) right manus anterior view; (B) left manus posterior view (C) right manus lateral view; (D) right manus medial view. Scale bar = 2cm. (Ramsay and Henry, 2001). The elephant pes (Figure 9) is smaller than the elephant manus, and semi- plantigrade (Fowler, 2001); it consists of tarsals, metatarsals, and phalanges. The tarsus comprises seven bones in three rows. Like the forefoot, the four distal tarsal bones articulate with corresponding metatarsals one through four; the fifth metatarsal articulates with the fourth tarsal as well. In Asian elephants, the first digit has one phalanx; in African elephants, the first digit has one sesamoid bone and no phalanx. All other digits have associated paired sesamoid bones located on the distal articular surface of the metatarsals, on the posterior side. In Asian elephants, the second digit has two phalanges; in African elephants, the second digit has three phalanges. In both species, digits three and four have three phalanges each, and digit five has two phalanges (Ramsay and Henry, 2001). The smallest metacarpal is MCI, which is medially positioned, and nearly triangular in shape (Figure 10A). Moving medially to laterally, the next bone is MCI] (Figure 10C), which is approximately twice as long and twice as wide as MCI. MCIII is the largest metacarpd, and centrally positioned (Figure 10E). MCIV (Figure 106) is slightly longer and considerably wider than MCII, and shorter and wider than MCIII. MCV, which is the lateral-most metacarpal, is the most block-like bone (Figure 101). It is slightly shorter than MCIV, and about the same width (Ramsay and Henry, 2001; Smuts and Bezuidenhout, 1993). The metatarsals follow the same size pattern, but they are smaller and stockier than their metacarpal counterparts (Figure 10). Each metapodial widens towards its proximal and distal ends; the distal end is distinctly wider than the proximal end, except for the first metapodial, which has distal l7 Figure 9. The African elephant pes: (A) MR7550 right pes, anterior view; (B) USNM 49849B left pes, posterior view (R-MTI not shown); (C) MR7550 left pes, medial view; (D) MR. 75 50 right pes, lateral view. Scale bar = 2 cm. Figure 10. Elephant metapodials (posterior view). (A) USNM 266911 R-MCI; (B) USNM 49639 L-MTI; (C) USNM 266911 R-MCII; (D) USNM 49639 L- MTII; (E) USNM 266911 R-MCIII; (F) USNM 49639 R—MTIII; (G) USNM 266911 R—MCIV; (H) USNM 49639 L-MTIV; (I) USNM 266911 R-MCV; (J) USNM 49639 L-MTV. Scale bar = 2 cm. and articular ends of about the same width. Elephant metacarpals represent typical features of metapodials in general, in that they are elongate in shape and approximately quadrilateral in cross-sectional shape (Smuts and Bezuidenhout, 1993). There are a number of articular facets on metapodials. The first metapodial has three articular facets, one proximally to articulate with the first carpal, and two distally: one on the anterior that articulates with the first phalanx, and one on the posterior that articulates with the proximal sesamoid bone. The second metapodial has four proximal articular facets (medially for carpals one, two, and three, and laterally for MCIII), and three distally (one for PPII, and two for sesamoid bones). The third metapodial has four facets on the proximal articular surface (medially for MCII, one for the third carpal, one for the fourth carpal, and one laterally for MCIV). Distally, there is one facet for articulation with PPIII, and 2 facets for two sesamoid bones. The fourth metapodial has three articular facets proximally (one for MCIII, one for carpal IV, and one for MCV), and three distally (one for PPIV, and two for sesamoid bones). The fifth metapodial has three proximal articular facets (one for MCIV, one for carpal IV, and one for the ulnar carpal bone), and three distal facets (one for PPV, and two for sesamoid bones) (Smuts and Bezuidenhout, 1993). This study focuses on the distal articular surfaces of the metapodials. The articular facets in questions are those that articulate with sesamoid bones, which are found on the posterior side of the metapodial (Figure 11A). Tuberculosis Tuberculosis is a non-pyogenic form of osteomyelitis; it inflames the bone, but is 20 Figure 11. (A)USNM 266911 R-MTII. Arrow denotes location of distal articular surface. (B) FMNH 49894 R-MCIII: (X) location of width measurement; (Y) location of length measurement. Arrow denotes the location of posterior depression (C) FMNH 60601 R-MTIII. The arrows denote the location of articular surface undermining. (D) USNM 266911 L-MTIV. Arrow denotes “lipping” of the articular surface. (E) USNM 588113 R-MTIV. In life, this individual was infected with tuberculosis. This bone shows articular surface undermining. Scale bar = 2 cm. 21 L m r a r . c‘ ' ‘ s ' T , ‘ a .. . ; 4 4 (r I. .9 ' u .. . , . . ’1‘ . . o ' I . , ' , I ' U U T I . *1, . ', g . - ‘ , . .5‘ .. not pus-forming. It is a bacterial infection, caused by a microbiologic agent, and is transferred through air. The origin of Mj/cobacterium tuberculosis is as yet unknown. In the fossil record it is often difficult to determine which strain of mycobacterium affects the individual; the infecting agent is then termed part of the “Adj/cobacterium tuberculosis complex,” which consists of M tuberculosis, M. bovis, M afiicanum, and M microti (Frothingham et al., 1999). Extinct bison, sheep, musk ox, and other bovids have been diagnosed with skeletal lesions characterized as tubercular, caused by a member of the M tuberculosis complex (Rothschild et al., 2001). Humans (Rothschild and Martin, 1993), captive Asian elephants (Mikota et al., 2000), mastodons (Rothschild and Helbling, 2001), captive oryxes (Lomme et al., 1976), pet birds (Hoop, 2002), and an isolated African elephant (Urbain, 1938) have all been reported as infected with M tuberculosis. ijcobacterium bovis has been reported as a problem in various extant species; white-tailed deer, domestic cattle and African buffaloes are highly susceptible to infection (Michel, 2002). It has been suggested that the human tuberculosis epidemic resulted fi'om the spread of M. tuberculosis complex from domesticated animals to man (Taylor et al., 1999). Mycobacterium tuberculosis, the agent that causes infection in elephants and humans, is suggested to have an origin between 15,000 and 20,000 years before the present (Sreevatsan et al., 1997). The paleopathological character reported as pathognomonic for tuberculosis on bone, articular surface undermining on foot bones, has been noted on humans (Rothschild and Martin, 1993), as well as on extinct longhorn bison (Bison antiquus) (Rothschild et al., 2001), before being identified on mastodons (Rothschild and Helbling, 2001). The bison, dated at 17,870 i 230 years before the 22 present, is the earliest known paleopathological evidence for the occurrence of tuberculosis in North America. Paleopathological evidence of M. tuberculosis was found in human remains from more than 3,700 years ago (Ayvazian, 1993), and in elephants 2,000 years ago (Mikota et al., 2000). Tuberculosis in elephants Tuberculosis can be a debilitating disease for all species, but it is not necessarily fatal. It has a variety of effects on the body, and has different effects on different individuals. Symptoms common to both humans and elephants may include weight loss, anorexia, weakness, coughing, and difficulty breathing (Gutter, 1981; Binkley, 1997; Mikota et al., 2001). Diagnosis of tuberculosis in elephants is most often made post- mortem. Clinical signs of tuberculosis generally do not occur ante-mortem unless the infection is advanced (Francis, 195 8). Elephants brought to veterinarians for treatment of symptoms including weight loss, increased water intake, increased urination, atrophied muscles, depression, and loss of appetite (Gutter, 1981; Saunders, 1983; Binkley, 1997) were found to be infected with M. tuberculosis only after their deaths. M. tuberculosis is common in captive elephants. In North America, 18 of 359 (3.3%) captive elephants surveyed between August 1996 and May 2000 were afflicted with tuberculosis (Mikota et al., 2000). M. tuberculosis was found in 24 captive elephants in California, Illinois, Arkansas, Missouri, Florida, and New Mexico between 1994 and 2001 (Payeur et al., 2002). A study of elephants in North American zoos recorded 8 of 379, or 2.1%, deaths caused by tuberculosis, ranging from prior to 1941 to 2001 (Mikota et al., 2001). Most captive elephants with tuberculosis have been Asian 23 elephants. Since 2001, there has been only one reported case of tuberculosis in an African elephants (Urbain, 1938). In all cases of tuberculosis afflicting elephants, M. tuberculosis has been the responsible agent. The source of tuberculosis for elephants and humans is uncertain. There are no reported cases of tuberculosis in wild elephants (Mikota et al., 2000). It is often difficult to diagnose tuberculosis in elephants while they are alive; without the corpse of an infected wild elephant, it may not be possible to determine incidence of infection. Tuberculosis, as caused by M tuberculosis, seems to be common in domestic animals and wild animals that have close contact with domestic animals. For example, when poultry and game birds contract tuberculosis, the causative agent is typically M avium-intracellularae (Montali et al., 1976); when pet birds contract tuberculosis, the causative agent is typically M genavense or M avium (Hoop et al., 1996). There are, however, reports of pet birds, including a canary (Serinus canaria) and a Blue-fronted Amazon Parrot (Amazona amazona aestiva), becoming infected with M tuberculosis as transmitted from their owners through coughing (Hoop, 2002). Humans also have been known to spread the disease to elephants. There are no recorded instances in which an extant elephant has contracted M tuberculosis without being in contact with a human (Mikota et al., 2000), but it is uncertain whether contact with infected humans is responsible for all the reports of M tuberculosis in elephants. Tuberculosis on bone Rothschild and Martin (1993) reported that evidence of tuberculosis rarely shows 24 up on bone. Tuberculosis can react differently on different bones and on different areas of the same bone. Skeletal evidence for tuberculosis on humans is most often seen on ribs, vertebrae, and metacarpals; signature marks include granular masses, abscesses, necrosis, and subchondral articular surface undermining. The exact mechanism by which M tuberculosis causes bone breakdown is not fully understood Nair et al. (1996) propose three possibilities: (1) bacteria liberate acid and proteases, directly destroying the non-cellular bone components; (2) bacteria initiate cellular processes that stimulate bone degradation; (3) bacteria inhibit the process of bone matrix synthesis, by either increasing osteoclast production or decreasing osteoblast production. In the human spine, M tuberculosis infection decreases the extra-cellular matrix and collapses the vertebrae (Meghji et al., 1997). It is not known whether the bacterial infection causes bone breakdown directly, or if it is an indirect reaction to the introduction of a foreign agent into the cells. Rothschild and Helbling (2001) examined 49 mastodons with foot bones available; 45% showed undermining. This feature was identical to patterns that they had interpreted as pathognomonic for tuberculosis in extinct bison. The undermining was unusual in that it did not preferentially affect mastodons based on age at death, body size, season of death, or location. The authors did not mention whether the data reflect gender differences. Twenty-five percent of examined specimens with both ribs and feet present had articular surface undermining on foot bones in addition to periosteal reaction of the ribs; periosteal rib reaction is a character commonly associated with tuberculosis (Rothschild and Martin, 1993). 25 Elephant foot care One of the most important aspects in managing captive elephants is protecting their feet from harm. Caretakers spend more time on foot problems than on any other aspect of elephant care except feeding and cleaning. As the largest land animal on earth, the foot of the elephant has a substantial amount of weight to brace. Each foot of a 13,200 pound bull African elephant has to support 3300 pounds when stationary; when walking, each foot supports 4400 pounds; when ambling, each foot supports 6600 pounds. Because of the extreme amount of pressure placed on the elephant foot, any damage caused to the foot has serious repercussions for the mobility of the animal (Fowler, 2001). There are several reasons for the frequent occurrence of foot problems in captive elephants, including: lack of exercise, overgrowth of nail and/or sole, improper enclosure surface, too much moisture, insufficient foot grooming, unsanitary enclosures, inherited poor foot structure, malnutrition, and skeletal disorders such as arthritis (Fowler, 2001). Wild elephants maintain their healthy feet by covering large distances daily to eat, bathe, dig, and dust; this exercise strengthens foot muscles and promotes good blood flow to the feet, which most captive elephants lack due to an inactive lifestyle (Roocroft and Oosterhuis, 2001). 26 MATERIALS AND METHODS The material for this study was provided by the Michigan State University Museum Vertebrate Paleontology collection (MSUVP) and Mammal Research collection (MSUMR), the University of Michigan Museum of Zoology (UMMZ), the Field Museum of Natural History (FMNH), and the National Museum of Natural History (USNM) (Table 1). Fossil material includes American mastodon (Mammut americanum) post- cranial material, which was recovered in Michigan from various Late Pleistocene sediments. Recent material includes 10 Asian elephants (Elephas maximus) and 9 Table1. List of Specimens Specimen # Species Sex Habitat Age Undermining FMNH 53749 L africana ? captive adult present FMNH 60601 E. maximus male captive adult present FMNH 49894 E. maximus male wild adult present USNM 304615 L. africana male wild adult present USNM 494893 L. afn'cana ? ? adult resent USNM 163318 L. africana male wild adult present USNM 270993 L. afn‘cana female captive adult present USNM 588113 L. africana female captive adult present USNM 269391 E. maximus ? wild juvenile absent US NM 49639 E. maximus ? captive adult present USNM 49489 L. efn'cana ? captive adult present USNM 20756* L. afn'cana male captive juvenile C0 USNM 240476“ E. maximus female wild juvenile CO USNM 266911 E. maximus female captive adult present MSUMR no # E. maximus ? captive juvenile absent MSUMR 7550 L. afn‘cana male wild adult present MSUVP 1289 M. amen'canum ? wild adult present UMMZ no # E. maximus female captive adult present UMMZ no data E. maximus (7) ? ? adult present UMMZ 157850 E. maximus female captive adult present *CO = cannot obtain data from this specimen Afiican elephants (Loxodonta afiicana) with associated post-cranial material. elephants specimens include those that died in captivity as well as in the wild. 27 Recent The first part of the project included describing the character defined as pathognomonic for tuberculosis by Rothschild and Helbling (2001): articular surface undermining of foot bones, seen most frequently on metapodials. Mastodon metapodials were examined macroscopically for dimensions, position in the skeleton, description of articular surface undermining, and any other unique features present. Once metapodial features were described in the mastodon, they were then compared to the metapodials of extant elephants, both African and Asian. Each metacarpal and metatarsal was macroscopically examined. Dimensions of the bones were recorded (Figure 11B) (Appendix A). Scaled photographs of each specimen were taken using a 35 mm Minolta Maxxum 5, and some measurements were made from the photographs in cases when it was not convenient to measure on site. The skeletal elements were identified and recorded as positive or negative for articular surface undermining and lipping of the articular surface. The articular surface was examined for the presence of pathologies. This was done to eliminate the possibility that the undermining was caused by spondyloarthropathy, which causes a bone reaction similar to articular surface undermining, but also affects the articular surface (Rothschild and Martin, 1993). Any other unusual characteristics were also noted Histograms were generated showing the percentage of each metapodial affected by undermining to determine if a particular metapodial preferentially shows the feature. The data were also sorted by sex, species, age at death, and whether the elephants were captive or wild, to determine whether the undermining is biased towards any or all of these criteria. 28 RESULTS Metapodial features Based on the examination of Recent elephant foot bones, there appears to be two distinguishing features on metapodials: (1) the undermining of the bone, which is expressed as a depression, or excavation, on the posterior side behind the articular surface (Figure 11C), and (2) the “lipping” of the articular surface, an overgrowth of the articular surface on the posterior side of the bone, that interrupts the smooth contact between the articular surface and the body of the foot bone (Figure 11D). These two characters do not necessarily appear concurrently; lipping of the articular surface may occur without the presence of excavation, though lipping always appears with excavation, as excavation appears to be an extension of this character. Articular surface undermining on mastodons Rothschild and Helbling (2001) identified subchondral articular surface undermining on the metapodials of 22 mastodons, including MSUVP 1289 (Appendix A). This character was noted at the distal articular surface on the posterior facets. MSUVP 1289 is an American mastodon of Late Pleistocene age from Ottawa County, Michigan; it was excavated by the Grand Rapids Public Museum from muck on top of light colored till in June of 1947 (Abraczinskas, 1993). Material found includes tusks, humerus, vertebrae, foot bones, ribs, pelvis, and leg bones. For this study, the foot bones of this specimen were chosen as a standard with which to compare extant elephant bones. Six foot bones were examined fi'om MSUVP 1289, including metacarpals, metatarsals, and phalanges. The four metapodials show articular surface undermining, and no 29 pathologies are apparent on the articular surfaces. Neither phalanx has articular surface undermining, and no pathologies are apparent on the articular surfaces of either bone. Description of elephant specimens The Field Museum of Natural History has three specimens with associated foot bones (Appendix A). FMNH 53749 is an African elephant of unknown sex from the Chicago Zoological Society. It was obtained by the Field Museum in July of 1943. Material studied includes 1 phalanx, 4 metapodials, and one tarsal. Articular surface undermining and lipping of the articular surface is present on RP-PPIV and R-MCIII. Lipping of the articular surface is present on R-MCII, L-MTII, and L-MTIV. There is neither undermining nor lipping on L-CAL. None of the listed bones have articular surfaces apparently affected by pathologies. FMNH 60601 is a male Asian elephant from the Chicago Zoological Society named “Ziggy.“ Ziggy was obtained by the Field Museum in October 1975. Material studied includes 1 phalanx and 12 metapodials. Articular surface undermining and lipping of the articular surface is present on R-MCII, R-MCIII, R-MCIV, R-MCV, L- MCII, L-MCIV, L-MCV, R-MTII, R—M'I'IH, L-MTII, L-MTIII, and L-MCV. Lipping of the articular surface is present on R—MTV and L-MTV. There is no undermining or lipping on LP-PPI. None of the listed bones have articular surfaces apparently affected by pathologies. FMNH 49894 is a male mainland Asian elephant (Elephas maximus indicus) from Hardwar, Uttah Pradesh, India, collected by 1.1. Hauser and obtained by the Field Museum on April 7, 1865. Material studied includes 11 metapodials. Articular surface 30 undermining and lipping of the articular surface is present on R-MCII, R-MCIII, R- MCIV, L-MCII, L-MCIII, R-MTII, R—MTIV, L-MTII, and L-MITV. There is no undermining or lipping on R—MCV and L-MCV. None of the listed bones have articular surfaces apparently affected by pathologies. There is an ovular depression on the posterior side of R-MCIII, near the distal articular surface (Figure 10B). The National Museum of Natural History collections includes 11 individuals with associated foot bones (Appendix A). USNM 304615 is a male African Elephant from Angola, district of Bie Cuando, in the region of Mancuso. It was collected at 17°19’ S, 21°14’ E, 48 miles north-northwest of Mancuso. Nine metapodials were available for study. R-MCIII, L—MCII, L-MCIV, R-MTII, L-MTI, L-MI‘II and L—MTIII show both lipping and excavation. L—MCV shows lipping; R-MCI has no articular surface undermining. None of the listed bones have articular surfaces apparently affected by pathologies. USNM 494893 has no data, but is likely an African elephant, based on its large size and comparatively slender foot bones. Foot bones available for study include 19 metapodials. The articular surfaces for R-MCIII, R-MCIV, L-MCIII, L-MCIV, and L- MTII have been detached from the rest of the bone, so articular surface undermining data is not obtainable from these bones. R-MCI, L-MCI, R—MTIII, R-MTIV, L—MTIII, and L- MTIV show articular surface undermining. There is no undermining on R-MTI. The rest of the listed bones show minor lipping, but no excavation beneath the articular surface. None of the listed bones have articular surfaces apparently affected by pathologies. USNM 163318 is a male African Elephant from Kenya. It was collected on the western slope of Mt. Kenia, at an altitude of 7000 feet. The elephant was added to the 31 collection on August 19, 1909, as part of the Smithsonian Afiican Expedition, and was collected by Theodore Roosevelt. Foot bones available for this specimen include 14 metapodials. R-MCIII, R-MCIV, L-MCIII, L-MCIV, R-MTIII, R-MTIV, L-MTII, and L- MTIII show articular surface undermining and lipping of the articular surface. R-MCV, L-MCV, R-MTII, R-MTV, L-MTIV, and L-MTV show lipping of the articular surface. None of the listed bones have articular surfaces apparently affected by pathologies. USNM 270993 is a female African elephant from the Philadelphia Zoological Gardens. She died on March 12, 1943. Foot bones available for this specimen include 14 metapodials and 1 phalanx. R-MCI, R-MCII, R-MCIII, R-MCIV, L—MCI, L-MCIl, L- MCIII, L-MCIV, R-MTIII, L-MTIII, and L-MTIV articular surface undermining, and lipping of the articular surface. R-MCV, L—MCV, R-‘MTIV, L-MTIV, and LM-PPV show lipping of the articular surface. None of the listed bones have articular surfaces apparently affected by pathologies. USNM 588113 is a female African elephant from an unknown locality who died in captivity at the National Zoological Park. Her remains were added to the collections on August 22, 2000, from an unknown collector. In life, this elephant had foot problems, and post-death examination revealed that she had tuberculosis, though it was not ruled as the cause of death (Linda Gordon, personal communication). Foot bones examined from this specimen are 4 metapodials and 5 phalanges. R-MTI, R-MTII, R-MTIV, RP-PPH, and RP-PPIII show articular surface undermining and lipping of the articular surface (Figure 11E). There is no articular surface undermining or lipping of the articular surface on R-MTV, RP-PPIV, and LM-IPIV. All listed bones have articular surfaces apparently unaffected by pathologies, and are significantly more porous than bones of other 32 elephants examined (Figure 12A). R-MTII shows a separation between the articular facets on the distal portion of the bone (Figure 128). The degree of undermining present, however, is not vastly different than the other specimens. USNM 269391 is a juvenile Asian elephant collected by OS. Huntington. No locality information is provided for this specimen, which was obtained from the Army Medical Museum, US. War Department. Foot bones examined for this specimen include 18 metapodials. Articular surfaces for all listed bones are deformed in some way, as sesamoid bones that were glued to the surface either fell off or were removed, leaving residue behind or taking part of the articular surface away. The following bones from this specimen are missing their distal articular surfaces: R-MCIII, R-MCIV, R-MCV, L- MCH, L-MCIII, L-MCIV, L-MCV, R-MTIII, R-MTIV, L-MTI, L-MTIII, L-MTIV, and L- MTV. Of the bones whose distal articular surfaces were still attached to the body of the bone, there is no articular surface undermining on R-MCII, R-MTII, and L-MTII. There is minor lipping of the articular surface on R-MTI and L-MTII. USNM 49639 is an Asian elephant of unknown sex. The specimen locality is unknown, as is the collector. The individual died on Nov. 6, 1898, at the National Zoological Park. Bones available for this specimen include 20 metapodials. All of the bones have a distinct “bunching” that surrounds the bone directly proximal to the articular surface (Figure 12C); the bunching is porous in nature. This feature makes articular surface undermining difficult to identify. Articular surface undermining and lipping of the articular surface is present on L-MCIV, R-MTI, R-MTIII, R-MTV, L-MTI, and L—MTIII. There is no articular surface undermining on R-MCI, R-MCIV, L-MCI, L- MCII, L-MCV, R-M'I‘IV, and L-MTV. There is lipping of the articular surface on R- 33 Figure 12. (A) USNM 588113 R-MTV (posterior view). Note the increased amormt of pores on this bone. (B) USNM 588113 R-MTIV (distal view). Arrow denotes the location of the separation of articular facets. (C) USNM 49639 L-MTH (anterior view). Arrow denotes the location of “bunching.” Scale bar = 2 cm. MCII, R-MCIII, R-MCV, L-MCIII, R-MTH, L-MTII, L-MTIV. The articular surfaces of all listed bones are apparently unaffected by pathologies. USNM 49849 is an African elephant of unknown sex collected by Dr. S Schuenland in South Afiica, Cape Colony, Addo Bush, near Port Elizabeth. Only the right forelimb and right hind limb are present; the remaining two limbs are on loan. The legs are fully articulated. R-MTI is absent from the articulated foot, and R-MTII is missing its articular surface. Articular surface undermining is present on R-MTIII, R- MTIV, and R-MTV. There is no articular surface undermining on R—MCI. Slight lipping of the articular surface is present on all other bones. None of the bones show excavation beneath the articular surface. None of the articular surfaces appear to be affected by pathologies. USNM 20756 is a male African elephant named “Mango.” He died in captivity on April 6, 1882, when he was about 5 years old, and was presented to the museum by Adam Fovepaugh. Evidence of articular surface undermining is unobtainable from this Specimen, as the distal articular surfaces are detached from the body of the foot bone. There is porous bone near the proximal and distal surfaces of the metapodials; the nature of the bone here is similar to that shown in USNM 588113, but to a lesser degree. USNM 240476 is a juvenile Asian elephant collected in Annarn, Vietnam, 20 miles northeast of Vinh, by F .R. Wilsin in 1924, as part of the National Geographic expedition to central Asia. As in USNM 20756, no foot bones have distal articular surfaces fused to the body of the bone. Again, there is porous bone present near the proximal and distal articular surfaces of the metapodials. USNM 266911 is a female Asian elephant who died on August 12, 1937, in the 35 National Zoological Park. Foot bones examined for this specimen include 9 metapodials. All metapodials show bunching near the distal articular surfaces on both the posterior and anterior sides. The metacarpals show a more extreme degree of bunching than do the metatarsals. All metapodials present (R-MCI, R-MCII, R-MCIII, R—MCIV, R-MCV, R-MTII, R—MTIII, R-MTV, and L-MTIV) show articular surface undermining and lipping of the articular surface. No pathologies are apparent on the articular surfaces. R-MCIII has a depression on the posterior side, near the distal articular surface. The mammal collections at Michigan State University hold two recent elephant skeletons with associated foot bones (Appendix A). The first does not have a catalogue number. It is a juvenile Asian elephant, possibly from a circus. The specimen consists of four articulated legs, which included R-MCI, R-MCII, R-MCIII, R-MIV, R—MCV, L- MCI, L-MCII, L—MCIH, L-MCIV, bMCV, R-MTI, R-MTII, R-MTIII, R-MTIV, L-MTI, L-MII, L-MTIII, and L-MTIV. Most of the articular surfaces are detached from the foot bone with which they are associated. For the bones that have their articular surfaces intact, R-MCI, R-MCV, L-MCI and L-MCV show no articular surface undermining. None of the listed foot bones have articular surfaces apparently afi‘ected by pathologies. MSUMR 7550 is a wild male African elephant that was at least 20 years old at death He was collected in Kenya, 35 miles north of Voi by Jens Touborg on February 2, 1962. The specimen is mounted on exhibit at the Michigan State University Museum; nearly all skeletal elements are present. In the feet, the only apparent bone missing is R- MCI. Of the available bones, R-MCII, R-MCII, R-MCIV, L-MCII, L-MCIII, L-MCIV, R- MTII, R—MTIII, R-MTTV, R-MTV, L-MTI, L-MTIII, L-M'I‘IV, and L-MTV have articular surface undermining and lipping of the articular surface. R-MCV, L—MCI, L-MCV, and 36 L-MTII show lipping only. R-MTI has neither feature. None of the listed bones have articular surfaces apparently affected by pathologies. In this specimen, the metatarsals have a greater degree of undermining than do the metacarpals. On the metacarpals, both the R— and L-MCIV have a greater degree of undermining than the other metacarpals. The University of Michigan Museum of Zoology has three elephants with associated foot bones (Appendix A). Two do not have catalogue numbers. The first of these (referred to as UMMZ no number) is “Amber,” an adult female Asian elephant from the Toledo Zoo. Metapodials available for study are R-MCI, L-MCI, R—MCII, L- MCII, L-MCIII, and L-MCIV. All listed bones have articular surface undermining except for R-MCI, which has lipping of the articular surface only. None of the listed bones have articular surfaces apparently affected by pathologies. The proximal articular facets on R- MCII and L-MCII overlap each other, giving the bone a “squished” appearance. R-MCI has separated distal articular facets. The second UMMZ specimen with no number (referred to as UMMZ no data), has no associated data It is an adult, and likely an Asian elephant. Metapodials available for examination are L-MTI, L-MTII, L-MTIII, L-MTTV, and L-MT V. All listed bones have articular surface undermining. L-MTII and L-MTIV have small holes on the distal articular surfaces. The rest of the listed bones have articular surfaces apparently unaffected by pathologies. The metapodials on this individual, along with UMMZ 157850, have the most drastic articular surface undermining of any in this study; L-MTIII and L-MTIV have the most extreme undermining. The bones are extremely porous, and resemble those of USNM 588113. On this specimen, the tarsus also exhibits extreme articular surface undermining. 37 UMMZ 157850 is “Minnie,” an adult female Asian elephant fiom the Brookfield Zoo. Bones available for examination include most of the right hind foot and the left forefoot. Butchering post-death removed parts of the foot; R-MCTV is mostly gone, and R-MCIII and L-MCV are missing most of their distal articular surfaces. All examined bones show articular surface undermining except for R-MTV and R-MTI. The metapodials on this individual have drastic undermining, most notably R-MTIV. The articular surface of L-MCII is slightly chipped Many of the bones are porous, especially R-MTI. The remaining bones have articular surfaces apparently unaffected by pathologies. Frequency of undermining on foot bones Of the 19 extant elephants examined (Table 1), 17 had foot bones with at least some associated distal articular surfaces intact From these 17 elephants, 173 bones were examined. Of these, 7 are phalanges, one is a tarsal, 86 are metacarpals, and 79 are metatarsals. F orty-three percent (3 of 7) of phalanges have articular surface undermining; the tarsal did not have articular surface undermining. Of the 165 metapodials, 141 (85%) have articular lipping of the articular surface (Figure 13A) (Table 2); 103 (62%) are affected by articular surface undermining and lipping of the articular surface (Figure 138) (Table 2). Eighty-three percent (71 of 86) of metacarpals have lipping of the articular surface (Figure 14A) (Table 2); 59% (51 of 86) of metacarpals have articular surface undermining and lipping of the articular surface (Figure 148) (Table 2). Eighty-nine percent (70 of 79) of metatarsals have lipping of the articular surface (Figure 15A) (Table 2). Sixty-six percent (52 of 79) of metatarsals have 38 articular surface undermining and lipping of the articular surface (Figure 15B) (Table 2). A chi-square test applied to the metacarpals affected with articular surface undermining reveals that, at the 0.05 level, the distribution of articular surface undermining on metacarpals is significantly different from a random distribution. Application of the same test to the metatarsals also reveals that, at the 0.05 level, the distribution of articular surface undermining on metatarsals is significantly different from a random distribution (Appendix B). Articular surface undermining shows up most frequently on the third metapodial (31 of 35; 89%) (Table 2) (Figure 13). A chi-square test applied to the metapodial data revealed that at the 0.05 level, the distribution is significantly different from random (Appendix B). Every elephant that has articular surface undermining has both L- and R- MCIII and L- and R-MTIII affected except for two, USNM 49489 and USNM 49639. These two elephants show some lipping on the manus, but no undermining. The hind feet of USNM 49489 show undermining. The R—MTIII is affected; the L-MTIII was not available for study on this specimen. The hind feet of USNM 49639 also show undermining, on both L-MTIII and R-MTIII. The metatarsals were frequently affected to a greater degree than the metacarpals. Articular surface undermining preferentially affects MTIII on the pes (Figure 15), but on the manus MCII, MCIII, and MCIV are nearly equally affected (Figure 14). A chi- square test applied to the MCI], MCIII, and MCIV data reveals that the distribution of undermining on these bones is not significantly different from a random distribution. A chi-square test applied to MTII, MTIH, and MT IV reveals that at the 0.05 level, the data are significantly different from random (Appendix B). 39 Table 2. Summa / of examined foot bones. Affected Bone # Present # with Lipping % with Lipping # with ASU % with ASU Hindlimb CAL 1 0 0 0 O MTI 10 7 70 6 60 MTII 19 17 89.47 1 1 57.89 MT III 18 18 100 17 94.44 MT IV 17 17 100 13 76.47 MTV 15 11 73.33 5 33.33 PF! 1 0 0 O 0 PF" 1 1 100 1 100 PPIII 1 1 100 1 100 PPIV 2 1 50 1 50 PPV 0 0 O 0 0 Forelimb MCI 16 10 62.5 7 43.75 MC" 19 17 89.47 14 73.68 MCIII 17 17 100 14 82.35 MCIV 15 14 93.33 13 86.67 MCV 19 13 68.42 3 15.79 PPI 0 0 O 0 0 PF" 0 0 0 O 0 PPIII 0 0 0 O 0 PPIV 0 O 0 0 0 PPV 1 1 100 0 0 IPIV 1 0 O 0 0 Total 173 145 83.82 106 61 .27 Metapodials any 165 141 85.45 103 62.42 40 Articular Surface Lipping on Metapodials 120 g IMPI g .MPII - MPIII 5 D E DMPN :2 IMPV Mehpodlals A. Percentage of metapodials affected by articular surface lipping. Articular Surface Undermining on Metapodials 100 2 MP1 5 I MPll E :1 MP1" E c: MPN as l MPV Metapodials B. Percentage of metapodials affected by articular surface undermining and lipping of the articular surface. Figure 13. Summary of metapodials affected by (A) lipping of the articular surface and (B) articular surface undermining. 41 Articular Surface Lipping on Metacarpals 120 --._._2.-.._m..-.._. 1.2.2.... h-..“ --2_.._....__.m..- .2” ,3-..-_._i_ai.vw-..‘ l 100 ’; g g a MCI E 8° ’ g I MCII = so - f u MCIII g 40 . E a MCN :1 I MCV 20 - . 0 - i Metacarpalc A. Percentage of metacarpals affected by articular surface lipping. Articular Suface Undermlnlng on Metacarpals MCI I MCII a MCII! El MCN I MCV 96 undermined Metecerpalc .95 Percentage of metacarpals affected by lipping of the articular surface and articular surface undermining. Figure 14. Summary of metacarpals affected by (A) lipping of the articular surface and (B) articular surface undermining. 42 Articular Surface Lipping on Metatarsal: 120 so 3 M11 g. l M'fll .. u MTIII § nwmv :2 I MTV Metatarsal; A Percentage of metatarsals affected by lipping of the articular surface. Articular Surface Undermlning on Metatarsal: 100 -. ~ — - ~ 80 '3 IMTI 5 so - IMTII g uM'nII g 40 - [JMTIV :8 20 - IMW o - w - W. Met-tarsal: B. Percentage of metatarsals affected by lipping of the articular surface and articular surface undermining. Figure 15. Summary of metatarsals affected by (A) lipping of the articular surface and (B) articular surface undemiining. 43 Preferential occurrences of undermining Adult wild elephants (n=4) and adult captive elephants (n=9) show similar degrees of undermining (Table 1). Captive elephants have additional malformations, most notably the “bunching” of the bone around the distal articular surface, on both the anterior and posterior side of the bone, and the increased porosity of the bones. The bunching has an appearance similar to a sock pushed down around the ankle. This character appears in USNM 266911 and USNM 49639. The bunching distorts the bone in the same area that would show articular surface undermining and lipping, if they were present. In the wild, both adult African (n=3) and adult Asian (n=1) elephants have similar degrees of articular surface undermining. Captive adult African (n=2) and adult Asian (n=4) elephants also have similar degrees of undermining (Table 1). Male and female elephants are equally affected by articular surface undermining (Table 1). One hundred percent of the adult male elephants (n=5) exhibit articular surface undermining: three of these are wild African elephants, one is a wild Asian elephant, and one is a captive Asian elephant. One hundred percent of the adult females (n=5) show articular surface undermining: two are captive adult African elephants, and three are captive adult Asian elephants. Of the elephants in this study, four are juveniles. Of these four, one is a captive male African elephant of about 5 years in age; one is a wild female Asian elephant; one is a captive elephant of unknown sex; and one is a wild Asian elephant of unknown sex. The un-firsed epiphyses on the metapodials from the female African and female Asian elephant were not found associated with the associated bone body, or were reattached to 44 the body of the bone post-mortem in a destructive manner. On those two, lipping cannot be identified, and there is no evidence of undermining. Neither of the two with articular surfaces intact show undermining; at best, there is the slightest evidence of articular surface lipping. All of the adults (n=15) included in this study show articular surface undermining or lipping of the articular surface (Table 1). 45 DISCUSSION Frequency of undermining The lipping of the articular surface is present on 85% (141 of 165) of all metapodials examined in this study. Sixty-two percent (103 of 165) of these metapodials show, in addition, the character of articular surface undermining; twenty-three percent (38 of 164) show lipping of the articular surface only. The remaining bones (24 of 165; 15%) show neither undermining nor lipping. Rothschild and Helbling (2001) reported that articular surface undermining indiscriminately affected mastodons in their study regardless of age at death, body size, season of death, or location. For Recent elephants, this does not appear to be true. Articular surface undermining is present on at least one metapodial from every adult elephant in this study. The juvenile elephants show no articular surface undermining; if there is any restructuring on the foot bones, it is a slight lipping of the articular surface without undermining involved. The occurrence of articular surface undermining of Recent elephants is biased towards adults, suggesting a strong relation to ontogeny. Articular surface undermining preferentially affects the third metapodial, which is the longest and most centrally positioned (Figure 12). Assuming a weight of 15,000 pounds for the largest African elephant, with a slipper area (sole) of 254 square inches (Fowler, 2001), the pressure on the elephant foot can reach nearly 15 pounds per square inch. Most of this weight is centered on the third metapodial, due to the elephant’s mesaxonic limb structure. Marsh (1884) devised a classification system for ungulates based on functional limb symmetry. This classification, as described by Klaits (1972), can be applied to proboscideans, which are mesaxonic because the axis of the limb 46 can be applied to proboscideans, which are mesaxonic because the axis of the limb passes through the longest, centralized digit: the third. As the foot contacts the ground, torsion of the leg is produced, causing the first row of carpals or tarsals to apply pressure to the second row (Cope, 1887). The combination of immense weight, columnar legs, and overall limb structure of the proboscideans that show the highest degree of undermining on their third metapodials suggests that the undermining is actually restructuring of bone related to pressure. Furthermore, articular surface undermining does not occur on juvenile individuals, although some juveniles do show lipping of the articular surface. Lipping is suggested to be the first stage in restructuring the foot bone, eventually culminating in undermining. Even a young elephant carries a large amount of weight on its skeleton, but seemingly not enough to cause undermining. As the elephant grows larger, more pressure is applied to the feet, and the bones continue to restructure. Articular surface undermining occurs with a higher frequency on MTIII than on MCHI (Figures 14 and 15). The sembdigitigrade stance of the forelimbs may cause more stress to the foot bones than the semi-plantigrade stance of the hind limbs, causing the metatarsals to be restructured more often than metacarpals. The hind feet almost always have a greater role in propulsion than do the fore feet, though in the symmetrical elephant gait, the intervals between contact of the fore and hind limbs is about equal (Hildebrand, 1976). When an elephant ambles, the forward swing of the hind leg has slightly more kinetic energy than the foreleg. The difference is kinetic energy is due to the small difference in mass between the bind and fore leg; the foreleg of the elephant, at approximately 250 kg, is slightly lighter than the hind leg, at 268 kg (Hildebrand and Hurley, 1985). Increased mass on the hind leg results in greater 47 of undermining on MTIV and MTV, as compared with MCIV and MCV, suggests that the presence and degree of undermining is an effect of how the weight of an elephant is supported by its skeletal anatomy. Articular surface undermining preferentially affects MTIII on the pes (Figure 15 ), but on the manus MCII] and MCIV are nearly equally affected (Figure 14). The burden of support on the mesaxonic manus, then, may not be primarily MCHI. Because MCIII and MCIV are affected by articular surface undermining to approximately the same degree, it appears that the combination of MCHI and MCIV provides the primary means of support. The rhinoceros manus, also mesaxonic, divides the burden of support between the second and third digits when the manus lifts, and the third and fourth digit when the digit manus lands (Klaits, 1972). The elephant manus, then, may still have a plane of structural symmetry through the middle digit, but this digit apparently acts in concert with the fourth digit. The second metapodials in the manus and pes are also often affected by undermining, though not as frequently as the third and fourth metapodials (Figures 14 and 15). MCII is more frequently affected with undermining than MTH. These results suggest that pressure on the manus is more evenly distributed by the three centralized digits (MCII, MCIII, and MCIV). In the pes, however, the pressure from the limb is transmitted most heavily to the MTHI (Figure 15), as there is a wider gap between the percentage of affected MTIII and the percentages of affected surrounding metatarsals than there is between the percentage of affected MCII] and the percentages of affected surrounding metacarpals (Figure 14 and 15). 48 Articular surface undermining and fawn—specificity Articular surface undermining is not species-specific. In captivity, Asian elephants appear to suffer more foot problems than African elephants (Fowler, 2001). It might be suggested, then, that Asian elephants would have a higher degree of articular surface undermining, but this is not true. In fact, all of the adult African (n=8) and Asian (n= 6) elephants in this study, both captive and wild, show articular surface undermining (Table 1). These data support a structural origin for undermining. Both Asian and African elephants, related at the family level, are graviportal animals with mesaxonic limbs. The metapodials of Afiican and Asian elephants are very similar in shape; their articulated feet, too, are nearly identical. The primary difference between the limbs of these two proboscidean genera is size; African elephants are generally larger. Articular surface undermining and sexual dimorphism Sexual dimorphism is present in elephants; the males are larger than the females within the same species. More pressure would be applied to the feet on a larger elephant than on the feet of a smaller elephant, so the larger, heavier males might be expected to show a higher frequency of articular surface undermining. However, for elephants whose sexes are known, the frequency of undermining between the sexes is the same. One hundred percent of adult males (n=5, Table l) and adult females (n=5, Table 1) exhibit articular surface undermining. 49 Articular surface undermining and age Four elephants in this study are juveniles. Of these four, one is a captive male African elephant of about 5 years in age (USNM 20756), one is a wild female Asian elephant (U SNM 240476), one is a captive elephant of unknown sex (MSUMR no ti), and one is a wild Asian elephant of unknown sex (U SNM 269391). The female African and female Asian elephant do not have intact articular surfaces (Table 1). On these two specimens, lipping cannot be identified, and there is no evidence of undermining. Neither of the two juveniles with articular surfaces intact show undermining; at best, there is the slightest evidence of articular surface lipping. In contrast, 100% of adult elephants included in the study (n=15) have at least one metapodial with articular surface undermining or lipping of the articular surface. The juveniles in this study were too young to have fused epiphyses, which is why the articular surfaces were often absent or not strongly attached to the body of the bone. Until the individual becomes an adult, articular surface undermining might not have a chance to completely restructure the bone. Also, as a juvenile, the individual is apparently not large enough for the bone to begin to restructure. Articular surface undermining, then, is likely only found on larger, mature individuals. Articular surface undermining and habitat The potential relationship of articular surface undermining to habitat of the individual cannot be assessed using the data available in this study. Several of the elephant skeletons had no accompanying location data; of the 6 with location data, 2 are savanna African elephants from Kenya, 1 is a savanna African elephant from Angola, 1 is 50 a savanna African elephant from South Africa, 1 is a mainland Asian elephant from Uttah Pradesh, and 1 is a mainland Asian elephant from Vietnam. All show articular surface undermining except for one juvenile mainland Asian elephant, from which articular surface undermining data could not be obtained. More data are needed to establish whether a relationship exists between articular surface undermining and habitat. The presence of articular surface undermining is not related to whether the individual is enclosed in a captive habitat (Table 1). The adult captive elephants (n=9) in this study show severe metapodial undermining. After years in captivity, this is expected for an elephant (Fowler, 2001). Adult wild elephants (n=4), however, also show articular surface undermining. The nature and degree of undermining on the metapodials of the captive and wild elephants is comparable, suggesting that the occurrence of such a malformation is caused by the weight of the animal rather than by an outside factor; it is likely a feature of elephant skeletal anatomy, rather than a result of years in captivity. It is possible that captivity increases the likelihood that undermining would be present at an earlier age, or in smaller individuals, but the sample size is not large enough to support or refute this claim. To explore this idea further, the data must include wild adult females and more juveniles. A character that is likely a result of years in captivity is the “bunching” appearance of the bone around the distal articular surface, on both the posterior and anterior side (Figure 12C), found in this study exclusively in captive individuals. The bunching distorts the bone in the same area that would show articular surface undermining and lipping, if present. Bunching, like undermining, could be a result of the pressure from the weight of the animal on the metapodials. It may be evidence of a stress 51 fracture, as a similar feature on a ceratopsian phalanx was described as such (Rothschild, 1988). This feature appears on the anterior side of metapodials from several wild individuals in this study, including the mastodon. The extreme development of this feature on captive individuals in this study, however, may relate to the increased sedentary lifestyle of the captive elephant, and is likely an abnormal part of proboscidean anatomy. The argument for sexual dimorphism would be clarified by comparing elephants from equivalent situations. All adult females from which data were available in this study were captive. Captive elephants have more stress placed upon their feet due to lack of exercise. Captivity may compensate for smaller body size, causing undermining on the metapodials of females in this study. This idea should be revisited with data including wild female elephants, for comparison. Articular surface undermining and tuberculosis in Recent proboscideans Rothschild and Helbling (2001) suggested a causal connection between articular surface undermining and tuberculosis in fossil proboscideans. This relationship is not supported for Recent elephants. USNM 58813 is a female captive African elephant that suffered from tuberculosis. This specimen showed undermining; the undermining, however, was not different in appearance from the other adult elephants with articular surface undermining. Captive elephants are already at high risk for foot problems, due to lack of exercise combined with the hard substrate of most captive elephant habitats (Sadler, 2001). Tuberculosis may have weakened USNM 588113, and made it more difficult for her to get the proper exercise, but given the ubiquity of articular surface 52 undermining among non-tuberculin elephants, undermining was likely not a direct result of the tuberculosis. Articular surface undermining and tuberculosis in fossil proboscideans Rothschild et al. (2001) defined articular surface undermining as the pathognomonic character of tuberculosis in extinct bison skeletal material. Although bison and proboscideans are both graviportal animals, the bison foot does not function in the same way as the proboscidean foot. The elephant manus rotates upon a medial axis when the elephant takes a step. The bison has paraxonic limbs, like most members of the Order Artiodactyla (Klaits, 1972), and is not, then, primarily supported by a single metapodial (or three central metapodials) as is the proboscidean foot; rather, the axis of symmetry on the bison foot is between the two middle digits. In addition, the bison has unguligrade feet, as opposed to the semi-digitigrade and semi-plantigrade stances of the elephant fore- and hind feet, respectively. Because of these anatomical differences, the skeletal elements in the feet of the two animals likely react to stresses differently; therefore, inferred effects of tuberculosis on bison metapodials is not the best model for identifying the potential effects of tuberculosis on proboscidean metapodials. 53 CONCLUSIONS This study of 165 metapodials belonging to 17 individual elephants revealed articular surface undermining and lipping of the articular surface in 62% (103 of 165) of the metapodials; 88% (15 of 17) of the elephants in this study had at least one metapodial with articular surface undermining and lipping of the articular surface. All affected elephants are adults, both males and females; no juveniles with intact articular surfaces (n=2) show undermining. Articular surface undermining is widespread in Recent adult elephants; it is not sex-specific, habitat-specific, or species-specific. The third metapodial is most commonly affected by articular surface undermining (Figure 13); this increased frequency may be related to the fact that this metapodial receives the most pressure from the 10,000-plus pound animal with mesaxonic limbs. Articular surface undermining shows up more frequently on MTIII than on MCIII. Its presence on the MTIII of one specimen (U SNM 49639) and absence on the MCIII of the same individual suggests that the pressure applied to the hind foot may be greater than that applied to the fore foot. This differential occurrence of articular surface undermining is likely related to the semi-plantigrade stance of the pes, compared to the semi-digitigrade stance of the manus. In a comparison of all elephants, the MTIII (pes) and MCIV (manus) most frequently show undermining. The three central metapodials provide the main support for the graviportal animal, especially in the elephant manus; MCIV and MCII are more often affected by undermining than MT IV and WE, and MTIII is affected by undermining more often than MCIII (Appendix, Table B). On the forefoot, MCIII shares the brunt of the pressure with MCIV, to a greater degree than MTIII shares the pressure with MTIV on the hind foot. The same 54 distribution is observed in the mesaxonic limbs of the rhinoceros (Klaits, 1972), in which the plane of functional symmetry is through the third digit, but the third digit shares the burden of weight with the fourth and second digit, respectively, when the manus lands and when the manus raises. When the rhinoceros stands, the burden of support is shared by the third and fourth digit. For the Recent elephants in this study, MCH is also more often affected than MTH. This suggests that, on the manus, the pressure is more evenly distributed between the three central digits. On the pes, the difference between the percentage of bones with undermining between MT 111 and the immediately surrounding digits is greater than the difference in percentage of bones with undermining between MCIII and the immediately surrounding digits, suggesting that the MT 111 takes on a disproportionate amount of pressure. In conclusion, articular surface undermining is widely deve10ped in Recent adult elephants, and its presence is most reliably predicted by the age of the individual. The frequency of articular surface undermining is highest on foot bones that bear the greatest stress, suggesting a structural explanation for the observed frequency and distribution of the feature. Future work The results in this study suggest several avenues for future work: (1) Increase the sample size to further elaborate on the patterns that have been suggested in this thesis. (2) Pursue the possible link between habitat and the presence or degree of articular surface undermining. In Recent elephants, the slippers are softer for forest elephants, and rougher for desert elephants (Roocroft and Oosterhuis, 2001); it is possible that the 55 hardness of the substrate is related to the presence or degree of undermining in individuals. In captivity, most elephants reside in habitats with concrete or asphalt floors (Roocrofi and Oosterhuis, 2001); these substrates exert additional pressure onto the foot of the individual (Sadler, 2001). Although there is no discemable difference in the appearance of undermining observed in captive and wild elephants examined for this study, it would be interesting to test this idea with elephants from a wider range of habitats. (3) Apply the methods in this study to metapodials of other extinct proboscideans. It would be most interesting to compare the results of this study with a similar study done on Mammuthus, considering the phylogenetic relationship between mammoths and elephants. (4) Conduct a microbiological test of the bone in the area where undermining is present (as performed by Rothschild et al. [2001], on bison metacarpals) to clarify where M tuberculosis is present. 56 APPENDIX 57 Appendix A. Summary of examined foot bones ASU = articular surface undenniang NU = no underminifingpresent AAS = affected articular surface U = underminingpresent 1 CO = cannot obtain data from this specimen L = lippirg present (no undermining)__ Y = yes lM = indeterminate measurement N = no Specimen # Bone ASU AAS Length (ch3 Width (cm) LIW FMNH 53749 L-CAL NU N 17.3 8 2.1625 FMNH 53749 R-MCIII U N 14.3 4.3 33255814 FMNH 53749 L-MCI U N 8 3.5 2.2857143 FMNH 53749 L-MT ll L N 7.3 3 2.4333333 FMNH 53749 L-MTlll L N 9 3.5 2.5714286 FMNH 53749 RP-PPIV L N 8.5 3.5 2.4285714 FMNH 60601 R-MClll U N 23.2 7.4 3.1351351 FMNH 60601 R-MCN U N 22.5 8 2.8125 FMNH 60601 R-MCV U N 14.5 5.5 2.6363636 FMNH 60601 L-MClI U N M 7 IM FMNH 60601 L-MCIV U N 22.5 8 2.8125 FMNH 60601 L-MCV U N 14.3 6 2.3833333 FMNH 60601 R-MTlll U N 16.3 6 2.7166667 FMNH 60601 R-MTV L N 11 7 1.5714286 FMNH 60601 L-MTII U N IM 6 lM FMNH 60601 L-MTIII U N 16.5 6 2.75 FMNH 60601 L-MTV L N 11 7 1.5714286 FMNH 60601 LP-PPI NU N 9 6 1.5 FMNH 49894 R-MCII U N 22 9 2.4444444 FMNH 49894 R-MCIII U N 23.5 8.5 2.7647059 FMNH 49894 R-MCIV U N 21.3 8.5 2.5058824 FMNH 49894 R-MCV NU N 15.3 9 1.7 FMNH 49894 L-MCII U N 21 7.75 2.7096774 FMNH 49894 L-MClll U N 23.3 9 2.5888889 FMNH 49894 L-MCV NU N 16 8 2 FMNH 49894 R—MTll U N 14 6.3 2.2222222 FMNH 49894 R-MTIV U N 22 8.3 2.6506024 FMNH 49894 L-MTll U N 14.5 6 2.4166667 FMNH 49894 L-MTIV U N 14 6.3 2.2222222 USNM 304615 R-MCl NU N 12.8 5 2.56 USNM 304615 R-MCIII U N 25.4 7.4 3.4324324 USNM 304615 L-MCll U N 23.8 6.8 3.5 USNM 304615 L-MClV U N 23.3 6.7 3.4776119 USNM 304615 L-MCV L N 21.1 6.5 3.2461538 USNM 304615 R-MTII U N 15.1 7 2.1571429 USNM 304615 L-MTl U N 8.5 3 2.8333333 USNM 304615 L-MTll U N 16.6 5.5 3.0181818 USNM 304615 L-MTlll U N 18.8 6.5 2.8923077 USNM 494898 R-MCI U N 8.4 4 2.1 58 USNM 494898 R-MCII L N 17 5.5 3.0909091 USNM 494898 R-MClll 00 N 17.8 5.4 3.2962963 USNM 494898 R-MCIV CO N 16.5 5.5 3 USNM 494898 R-MCV L N 12.6 6 2.1 USNM 494898 L-MCI U N 8.5 3.5 2.4285714 USNM 494898 L—MCII L N 15.5 5 3.1 USNM 494898 L-MCIII 00 N 18.7 5.8 3.2241379 USNM 494898 L-MCIV CO N 15.5 5.9 2.6271186 USNM 494898 L-MCV L N 14.4 6.1 2.3606557 USNM 494898 R-MTI NU N 5 2.5 2 USNM 494898 R—MTIl L N 9.5 3.7 2.5675676 USNM 494898 R-MTlll U N 12.5 5 2.5 USNM 494898 R-MT IV U N 10.6 4.6 2.3043478 USNM 494898 R—MTV L N 7.6 5 1.52 USNM 494898 L-MTII CO N 9.5 3.6 2.6388889 USNM 494898 L-MTIII U N 13.2 4.5 2.9333333 USNM 494898 L-MTN U N 12 4.9 2.4489796 USNM 494898 L-MTV L N 7.4 6.2 1.1935484 USNM 163318 R—MCIIl U N 19 6.3 3.015873 USNM 163318 R-MCN U N 17 5.5 3.0909091 USNM 163318 R-MCV L N 16.2 6.8 2. 3823529 USNM 163318 L-MCIII U N 18.6 7 2.6571429 USNM 163318 L-MCIV U N 16.5 6.1 2.704918 USNM 163318 L-MCV L N 13.1 6 2.1833333 USNM 163318 R-MTII L N 11.5 3.5 3.2857143 USNM 163318 R-MTIII U N 13.5 6 2.25 USNM 163318 R-MTIV U N 13.4 5.4 2.4814815 USNM 163318 R-MTV L N 7.7 6 1.2833333 USNM 163318 L—MTII U N 11 4.4 2.5 USNM 163318 L-MTIII U N 13.5 5.1 2.6470588 USNM 163318 L-MTIV L N 12.2 4 3.05 USNM 163318 L—MTV L N 8.7 6.3 1.3809524 USNM 270993 R-MCI U N 7 3 2.3333333 USNM 270993 R-MCll U N 14.5 4.3 3.372093 USNM 270993 R-MClll U N 15.2 4 3.8 USNM 270993 R—MC IV U N 14.5 4.4 3.2954545 USNM 270993 R-MCV L N 13 5 2.6 USNM 270993 L-MCI U N 6.8 3.4 2 USNM 270993 L-MCII U N 3.4 4.2 0.8095238 USNM 270993 L-MCIII U N 15.3 4.4 3.4772727 USNM 270993 L-MCIV U N 13.2 4.2 3.1428571 USNM 270993 L-MCV L N 12.5 3.5 3.5714286 USNM 270993 R-MTlll U N 10.5 3.3 3.1818182 USNM 270993 R—MT IV L N 9.2 3.5 2.6285714 USNM 270993 L-MTIII U N 10.5 3.2 3.28125 USNM 270993 L—MT IV L N 9.5 3.5 2.7142857 USNM 270993 LM-PPV L N 6 4 1.5 USNM 588113 R-MTI U N 6.5 3 2.1666667 USNM 588113 R-MTII U N 10 4 2.5 USNM 588113 R-MTN U N 1 1.5 4.8 2.3958333 59 USNM 588113 R-MTV NU N 7.2 4.7 1.5319149 USNM 5881 13 RP-PPII U N 6.5 3.8 1.7105263 USNM 588113 RP-PPIII U N 6.7 3.8 1.7631579 USNM 588113 RP-PPIV NU N 4.5 3 1.5 USNM 588113 LM-IPIV NU N 4.5 2.5 1.8 USNM 269391 R-MCII NU Y 10 4 2.5 USNM 269391 R-MCIII CO Y 10.2 4.2 2.4285714 USNM 269391 R-MCIV CO Y 9.3 4.3 2.1627907 USNM 269391 R-MCV CO Y 9 4.3 2.0930233 USNM 269391 L-MCII CO Y 9.4 3.5 2.6857143 USNM 269391 L-MCIII CO CO 9.5 3 3.1666667 USNM 269391 L—MCIV CO Y 10.2 4.3 2.372093 USNM 269391 L-MCV CO Y 7.7 3.9 1.974359 USNM 269391 R-MTI L Y 2.8 1 2.8 USNM 269391 R-MTII NU Y 5.8 2.6 2.2307692 USNM 269391 R-MTIII CO CO 7 3 2.3333333 USNM 269391 R-MTN CO CO 6.5 3.3 1.969697 USNM 269391 R-MTV CO CO 4.5 2.5 1.8 USNM 269391 L-MT I CO Y 3.2 1 3.2 USNM 269391 L-MTII L Y 6 2.6 2.3076923 USNM 269391 L-MTIII CO Y 7.9 3.8 2.0789474 USNM 269391 L-MT IV CO CO 6.3 2.9 2.1724138 USNM 269391 L-MTV CO Y 5 3 1.6666667 USNM 49639 R—MCI NU N 9 4 2.25 USNM 49639 R-MCII L N 15.5 5.5 2.8181818 USNM 49639 R-MCIII L N 16.9 6 2.8166667 USNM 49639 R-MCIV NU N 15 5 3 USNM 49639 R-MCV L N 14.3 5.5 2.6 USNM 49639 L-MCI NU N 9.5 3.9 2.4358974 USNM 49639 L—MCII NU N 15.2 5.2 2.9230769 USNM 49639 L-MCIII L N 16.3 5.8 2.8103448 USNM 49639 L-MCIV U N 15 5 3 USNM 49639 L-MCV NU N 12.5 5.5 2.2727273 USNM 49639 R-MT I U N 5 2.5 2 USNM 49639 R-MTII L N 9.5 4.2 2.2619048 USNM 49639 R-MTIII U N 1 1.6 5.3 2.1886792 USNM 49639 R—MT IV U N 10 4.2 2.3809524 USNM 49639 R-MTV NU N 7 5 1.4 USNM 49639 L-MTI U N 5.2 2.5 2.08 USNM 49639 L—MTII L N 9.5 4 2.375 USNM 49639 L-MTIII U N 12 5.7 2.1052632 USNM 49639 L-MTIV L Y 10.7 4.5 2.3777778 USNM 49639 L-MTV NU N 7.2 5.5 1.3090909 USNM 49489 R-MCI NU N 6.8 3 2.2666667 USNM 49489 R-MCII L N 12.8 5 2.56 USNM 49489 R-MCIII L N 15 5.5 2.7272727 USNM 49489 R-MCIV L N 13 5.8 2.2413793 USNM 49489 R-MCV U N 11.3 5.5 2.0545455 USNM 49489 R-MTII CO Y 8 4 2 USNM 49489 R-MTIII U N 10.5 4.5 2.3333333 USNM 49489 R-MT IV U N 9.2 4.1 2.2439024 USNM 49489 R-MTV U N 5.7 4.7 1.212766 USNM 20756 R~MCI| CO Y 6.5 4 1.625 USNM 20756 R-MClll CO Y 8.5 4.2 2.0238095 USNM 20756 R-MCN CO Y 7.2 4.4 1.6363636 USNM 20756 R-MCV C0 Y 5.5 2.8 1.9642857 USNM 20756 L-MCll CO Y 7.5 3.6 2.0833333 USNM 20756 L-MCIll CO Y 8 4.5 1.7777778 USNM 20756 LMCIV CO Y 7 4.3 1.627907 USNM 20756 L-MCV CO Y 5.5 2.7 2.037037 USNM 20756 R-MTll CO Y 4.7 3 1.5666667 USNM 20756 R-MTlll C0 Y 6.2 3.5 1.7714286 USNM 20756 R-MT IV CO Y 5.3 3.3 1.6060606 USNM 20756 R-MTV CO Y 3.5 2.2 1.5909091 USNM 20756 L-MTII CO Y 4.7 3 1.5666667 USNM 20756 L-MTlll CO Y 6 3.7 1.6216216 USNM 20756 L-MTIV CO Y 5.1 4 1.275 USNM 20756 L-M‘N CO Y 3.3 2.1 1.5714286 USNM 266911 R-MCI U N 9.2 4 2.3 USNM 266911 R-MCll U N 14.5 4.9 2.9591837 USNM 266911 R-MCIII U N 17.3 6 2.8833333 USNM 266911 R-MCIV U N 15.5 5 3.1 USNM 266911 R-MCV U N 13.9 5 2.78 USNM 266911 R-MTII U N 10.5 3.7 2.8378378 USNM 266911 R-MTlll U N 11.5 4.5 2.5555556 USNM 266911 R-MTV U N 6.7 5 1.34 USNM 266911 L-MT IV U N 10.9 4 2.725 MSUMR no data R-MCI NU N 7 2.6 2.6923077 MSUMR no data R-MCV NU N 7.2 3.5 2.0571429 MSUMR no data LoMCl NU N 7.7 2.7 2.8518519 MSUMR no data L-MCV NU N 7.4 4.1 1.804878 MSUVP 1289 R-MCll U N 12.7 7.6 1.6710526 MSUVP 1289 L-MCI U N 8.2 4.9 1.6734694 MSUVP 1289 L-MCll U N 12 8.3 1.4457831 MSUVP 1289 RMTN U N 10.8 6.6 1.6363636 MSUVP 1289 RM-PPIV NU N 12.7 7.6 1.6710526 MSUVP 1289 LM-PPlll NU N 7.2 7.1 1.0140845 MSUMR 7550 R-MCll U N 18.8 6.5 2.8923077 MSUMR 7550 R-MCIII U N 20.1 7.7 2.6103896 MSUMR 7550 R-MCIV U N 18.3 8 2.2875 MSUMR 7550 R-MCV L N 15.1 6.6 2.2878788 MSUMR 7550 L-MCl L N 11.6 4.6 2.5217391 MSUMR 7550 L-MCll U N 17.5 9.5 1.8421053 MSUMR 7550 L-MCIII U N 19.9 7.8 2.5512821 MSUMR 7550 L-MCIV U N 18.2 7.6 2.3947368 MSUMR 7550 L—MCV L N 14.7 6.9 2.1304348 MSUMR 7550 R-MTl N N 7.2 2.9 2.4827586 MSUMR 7550 R-MTII U N 11.1 5 2.22 MSUMR 7550 R-MTlll U N 13.3 6 2.2166667 MSUMR 7550 R-MT N U N 13.1 6.6 1.9848485 61 MSUMR 7550 R-MTV U N 8.4 5.8 1.4482759 MSUMR 7550 L-MTI U N 7 2.8 2.5 MSUMR 7550 L-MTII L N 11.5 5 2.3 MSUMR 7550 L-MTlll U N 14.1 6.2 2.2741935 MSUMR 7550 L-MT IV U N 13.7 6.6 2.0757576 MSUMR 7550 L—MTV U N 8.3 5.8 1.4310345 UMMZ no # R-MCI L N 10 3.6 2.7777778 UMMZ no # L-MCI U N 9.4 3.7 2.5405405 UMMZ no # R-MCII U N 15.3 5.6 2.7321429 UMMZ no # L-MCII U N 15.4 15.5 0.9935484 UMMZ no # L-MClll U N 15.9 6 2.65 UMMZ no # L-MClV U N 15.3 6.2 2.4677419 UMMZ no data L-MTI U N 5.1 2.8 1.8214286 UMMZ no data L-MTII U Y 10 4.9 2.0408163 UMMZ no data L-MTIII U N 12 5 2.4 UMMZ no data L-MTlV U N 10.9 4.9 2.2244898 UMMZ no data L-MTV U N 7.5 5.4 1.3888889 UMMZ 157850 R-MClll U CO 15.4 5.8 2.6551724 UMMZ 157850 L-MCI U N 8.6 3.2 2.6875 UMMZ 157850 L-MCII U Y 14.7 5.4 2.7222222 UMMZ 157850 L—MCV U CO 12.3 4.9 2.5102041 UMMZ 157850 R-MTI N N 5.5 2.6 2.1 153846 UMMZ 157850 R-MTII U N 9.8 4.4 2.2272727 UMMZ 157850 R-MTlll U N 10.5 5 2.1 UMMZ 157850 R-MTIV U N 9.4 4.8 1 .9583333 UMMZ 157850 R—MTV N N 6.8 6.3 1.0793651 62 Appendix B. Chi-square tests for significance. ASU = articular surface undermining I df = degrees of freedom I o = observed I e = expected I Metapodials: Metapodial ASU No ASU Total a (Yes) eLNo) I 13 13 26 16.2303 9.769697 II 25 13 38 23.72121 14.27879 Ill 31 4 35 21.84848 13.15152 IV 26 6 32 19.97576 12.02424 V 8 26 34 21.22424 12.77576 Total 103 62 165 Metapodial Undermining o e o-e (o-e)“2/e I Y 13 16.2303 -3.23030303 0.642924 I N 13 9.769697 3.23030303 1.068084 II Y 25 23.72121 1.27878788 0.068938 II N 13 14.27879 -1.27878788 0.114526 Ill Y 31 21.84848 9.15151515 3.833228 III N 4 13.15152 915151515 6.368105 IV Y 26 19.97576 6.02424242 1.816777 IV N 6 12.02424 602424242 3.018194 V Y 8 21.22424 -1 3.2242424 8.239662 V N 26 12.77576 13.2242424 13.68847 df = (M )*(c~1) = 4 Chi-square 38.85891 63 Metacarpals Metacarpal ASU No ASU Total 6 (Yes) e (No) I 7 9 16 9.488372 6.511628 ll 14 5 19 11.26744 7.732558 III 14 3 17 10.0814 6.918605 IV 13 2 15 8.895349 6.104651 V 3 16 19 11.26744 7.732558 Total 51 35 86 Metacerpal Undemining o e o—e (o-e)“2/e I Y 7 9.488372 -2.48837209 0.652588 I N 9 6.511628 2.48837209 0.950914 II Y 14 11.26744 2.73255814 0.662695 II N 5 7.732558 273255814 0.965641 lll Y 14 10.0814 3.91860465 1.523149 III N 3 6.918605 -3.91860465 2.219445 N Y 13 8.895349 4.10465116 1.894042 IV N 2 6.104651 410465116 2.759889 V Y 3 11.26744 826744186 6.066203 V N 16 7.732558 8.26744186 8.839325 df =(r-1)*(c—1) = 4 Chi-square 26.53389 Metatarsals Metatarsal ASU No ASU Total 9 (yes) a (no) I 6 4 10 6.582278 3.417722 II 11 8 19 12.50633 6.493671 III 17 1 18 11.8481 6.151899 IV 13 4 17 2.797468 5.810127 V 5 10 15 9.873418 5.126582 Total 52 27 79 Metatarsal Underminin o e o-e (o—e)"2le I Y 6 6.582278 058227848 0.051509 I N 4 3.417722 0.58227848 0.099203 II Y 11 12.50633 -1.50632911 0.18143 II N 8 6.493671 1.50632911 0.349421 III Y 17 11.8481 5.15189873 2.240195 III N 1 6.151899 515189873 4.31445 IV Y 13 2.797468 10.2025316 37.20923 IV N 4 5.810127 -1.81012658 0.563939 V Y 5 9.873418 487341772 2.405469 V N 10 5.126582 4.87341772 4.632755 df = (r-1)*(c-1) = 4 Chi-square 52.04761 Metacarpals H, III, and IV Metacarpal ASU No ASU Total e (yes) 9 (no) II 14 5 19 15.27451 3.72549 III 14 3 17 13.66667 3.333333 IV 13 2 15 12.05882 2.941176 Total 41 10 51 Metacarpal Unden'nini_ng o e o-e (o-e)"ZIe ll Y 14 15.27451 -1 .2745098 0.106345 II N 5 3.72549 1.2745098 0.436017 III Y 14 13.66667 0.33333333 0.00813 III N 3 3.333333 033333333 0.033333 IV Y 13 12.05882 0.94117647 0.073458 IV N 2 2.941176 -0.94117647 0.301176 df = (3-1)*(2—1)= 2 Chi-square 0.95846 65 Metatarsals H, II], and IV Metatarsal ASU No ASU Total e (yes) 9 (no) II 11 8 19 14.42593 4.574074 III 17 1 18 13.66667 4.333333 N 13 4 17 12.90741 4.092593 Total 41 13 54 Metatarsal Underminin o e o—e (o-e)"2le II Y 11 14.42593 -3.42592593 0.813602 II N 8 4.574074 342592593 2.565977 III Y 17 13.66667 3.33333333 0.813008 III N 1 4.333333 -3.33333333 2.564103 N Y 13 12.90741 0.09259259 0.000664 N N 4 4.092593 009259259 0.002095 df=(3.1)"(2-1)= 2 Chi-square 6.759449 66 REFERENCES 67 Abraczinskas, L.M. 1993. Pleistocene Proboscidean sites in Michigan: New records and an update on published sites. Michigan Academician 25:443-490. Agenbroad, D. and Mead, J .I. 1996. Distribution and palaeoecology of central and western North American Mammuthus. In Shoshani, J ., and Tassy, P. (eds). The Proboscidea: Evolution and palaeoecology of elephants and their relatives. Oxford University Press, Oxford, 472 pp. Andrews, CW. 1906. A descriptive catalogue of the Tertiary Vertebrata of the Fayum, Egypt. London, British Museum, 319 pp. Ayvazian, LP. 1993. History of tuberculosis. In Reichman, LB. and Hershfield, E.S. (eds). Tuberculosis: A comprehensive international approach. New York, Marcel Dekker, 752 pp. Binkley, M. 1997. Tuberculosis in captive elephants. Proceedings of the American Association of Zoo Veterinarians. Seattle, WA, p. 116-119. Bricknell, I. 1987 . Palaeopathology of Pleistocene proboscideans in Britain. Modern Geology 11:295-309. Cope, ED. 1887. The Perissodactyla. American Naturalist21(11): 985-1007. Douglas-Hamilton, I. And Michelmore, F. 1996. Loxodonta afi'icana: range and distribution, past and present. Shoshani, J ., and Tassy, P. (eds). The Proboscidea: Evolution and palaeoecology of elephants and their relatives. Oxford University Press, Oxford, 472 pp. Fisher, DC. 1984. Taphonomic analysis of late Pleistocene mastodon occurrences: evidence of butchery by North American Paleo-Indians Paleobiology 10:338- 358. Fischer, MS. and Pascal Tassy. 1993. The interrelation between Proboscidea, Sirenia, Hyracoidea, and Mesaxonia: the morphological evidence. In Szalay, F.S., Novacek, M.J., and McKenna, M.C. (eds). Mammal phylogeny: placentals, Springer-Verlage, New York, 1-321 pp. Fowler, ME. 2001. An overview of foot conditions in Asian and African elephants. In Csuti, B., Sargent, E.L., and Bechert, U.S. (eds). The elephant’s foot. Iowa State University Press, Ames, 163 pp. Francis, J. 1958. Pathogenesis and Pathology. In Tuberculosis in animals and man: a study in comparative pathology. Cassell, London, 11-43. Frothingharn, R., Strickland, P.L., Bretzel, G., Ramasqarny, S., Musser, J .M., and 68 Williams, D.L. 1999. Phenotypic and genotypic characterization of Mycobacterium africanum Isolates from West Africa. Journal of Clinical Microbiology 37(6): 1921-1926. Gutter, A. 1981. Mycobacterium tuberculosis in an Asian elephant. Proceedings of the American Association of Zoo Veterinarians. Seattle, WA., p. 105-106. Hildebrand, M. and Hurley, J .P. 1985. Energy of the oscillating legs of a fast-moving cheetah, pronghom, jackrabbit, and elephant. Journal of Morphology 184223-31. Hildebrand, M. 1976. Analysis of tetrapod gaits: general considerations and symmetrical gaits. In R.M. Herman et al. (eds). Neural Control of Locomotion. Advances in Behavioral Biology, 18. New York, Plenum Press. Holman, J .A. 1975. Michigan’s Fossil Vertebrates. Stone Printing Company, Lansing, Michigan, 54 pp. Holman, J .A. 1995. Ancient life of the Great Lakes Basin: Precambrian to Pleistocene. University of Michigan Press, Ann Arbor, Michigan, 287 pp. Holman, J .A., Fisher, DC, and Kapp, RD. 1986. Recent discoveries of fossil vertebrates in the Lower Peninsula of Michigan. Michigan Academician 28(3):431-463. Hoop, R.K. 2002. Mycobacterium tuberculosis infection in a canary (Serinus canaria L.) and a Blue-Fronted Amazon Parrot (Amazona amazona aestiva), Avian Diseases 46:502-504. Hoop, R.K., Bettger, EC, and Pfyffer, GE. 1996. Etiological agents of mycobacterioses in pet birds between 1986 and 1995. Journal of Clinical Microbiology 34:991-992. King, J .E. and Saunders, J .J . 1984. Environmental Insularity and the Extinction of the American mastodont. In Martin, RS. and Klein, R.G (eds). Quaternary Extinctions. University of Arizona Press, Tuscon, 315-339. Klaits, BC. 1972. The moving mesaxonic menus: a comparison of tapirs and rhinoceroses. Mammalia 36(1): 126-145. Lomme, J .R., et al. 1976. Mycobacterium tuberculosis infection in two East African oryxes. Journal of the American Veterinary Medical Association 169(9):912- 914. Maglio, V.J. 1972. Evolution of Mastication in the Elephantidae. Evolution 26:63 8- 658. 69 Marsh, CC. 1884. Dinocerata, a monograph of an extinct order of gigantic mammals. US. Geological Survey Monography 10 (28): 1-237. Meghji, S., White, P.A., Nair, S.P., Reddi, K., Heron, K., Henderson, 8, Zalianai, A., Fossati, G., Mascagni, P., Hunt, J .F., Roberts, MM. and Coates, A.R.M. 1997. Mycobacterium tuberculosis Chaperonin 10 stimulates bone resorption: a potential contributory factor in Pott’s Disease. Journal of Experimental Medicine 186(8):]241-1246. Michel, AL. 2002. Implications of tuberculosis in African wildlife and livestock Annals of the New York Academy of Sciences 969:251-255. Mikota, S.K., Larsen, RS, and Montali, RJ. 2000. Tuberculosis in elephants in North America. Zoo Biology 19:393-403. Mikota, S.K., Peddie, L., Peddie, J ., Isaza, R., Dunker, R., West, G., Lindsay, W., Larsen, R.S., Salman, M.D., Chatterjee, D., Payeur, J ., Whipple, D., Thoen, C., Davis, D.S., Sedgwick, C., Montali, R.J., Ziccardi, M., and Maslow, J. 2001. Epidemiology and diagnosis of Mycobacterium tuberculosis in captive Asian elephants (Elephas maximus). Journal of Zoo and Wildlife Medicine 32(1): 1-16. Montali, R.J., Bush, M., Thoen, CO, and Smith, E. 1976. Tuberculosis in captive exotic birds. Journal of the American Veterinary Medical Association 1692920- 927. Nair, S.P., Meghji, S., Wilson, M., Reddi, K., White, P., and Henderson, B. 1996. Bacterially induced bone destruction: mechanisms and misconceptions. Infection and Immunity 64: 2371-2380. Payeur, J .B., J .L. Jamagin, J .G. Marquardt, and D.L. Whipple. 2002. Mycobacterial isolations in captive elephants in the United States. Annals of the New York Academy of Sciences 969:256—258. Powers, RD. and Price, RA 1967. Human tuberculosis in a rhinoceros. Journal of the American Veterinary Medical Association 151(7):890-892. Ramsay, EC. and Henry, R.W. 2001. Anatomy of the elephant foot. In Csuti, B., Sargent, E.L., and Bechert, U.S. (eds). The elephant’s foot. Iowa State University Press, Ames, 163 pp. Richman, L.K., RJ. Montali, R.C. Cambre, J. Lehnhardt, M. Kennedy, S. Karria, and L. Potgieter. 1996. Endothelial inclusion body disease: a newly recognized fatal herpes-like infection in Asian elephants. Proceedings of the American Association of Zoo Veterinarians 1996:483-485. 70 Roocrofi, A. and Oosterhuis, J. 2001. Foot care for captive elephants. In Csuti, B., Sargent, E.L., and Bechert, U.S. (eds). The elephant’s foot. Iowa State University Press, Ames, 163 pp. Rothschild, BM. 1988. Stress fracture in a ceratopsian phalanx. Journal of Paleontology 62(2):302-303. Rothschild, B. and Helbling, M. 2001. Documentation of hyperdisease in the Late Pleistocene: validation of an early 20th century hypothesis. Journal of Vertebrate Paleontology, Abstracts of Papers 21(3):94A. Rothschild, BM. and Martin, L. (eds). 1993. Paleopathology: disease in the fossil record. CRC Press, London, England, 361 pp. Rothschild, B.M., Martin, L.D., Lee, G. Bercovier, H., Bar-Gar, G.K., Greenblatt, C., Donoghue, H., Spigelman, M., and Brittain, D. 2001. Mycobacterium tuberculosis complex DNA from an extinct bison dated 17,000 years before the present Clinical Infectious Diseases 33:305-11. Rothschild, B.M., Xiaoming, W., and Shoshani, J. 1994. Spondyloarthropathy in proboscideans. Journal of Zoo and Wildlife Medicine 25:360-366. Sadler, WC. 2001. The role of nutrition and its possible impact on elephant foot care. In Csuti, B., Sargent, E.L., and Bechert, U.S. (eds). The elephant’s foot. Iowa State University Press, Ames, 163 pp. Saunders, G. 1983. Pulmonary Mycobacteriwn tuberculosis infection in a circus elephant. Journal of the American Veterinary Medical Association 183(11):]311- 1312. Saunders, J .J . 1996. North American Mammutidae. In Shoshani, J ., and Tassy, P. (eds). The Proboscidea: Evolution and palaeoecology of elephants and their relatives. Oxford University Press, Oxford, 472 pp. Schultz, CB. 1963. Fossil hunting yesterday and today. Museum Notes 22:18. Shoshani, J. 1989. A report on the Shelton mastodon site and a discussion of the numbers of mastodons and mammoths in Michigan. Michigan Academician 21(2):]15-132 Shoshani, J. 1992. Comparing the living elephants. In Shoshani, J. (ed.). Elephants: Majestic creatures of the wild Rodale Press, Inc., Pennsylvania, 240 pp. Shoshani, J. 1996. Skeletal and other basic anatomical features of elephants. In Shoshani, J ., and Tassy, P. (eds). The Proboscidea: Evolution and palaeoecology 71 of elephants and their relatives. Oxford University Press, Oxford, 472 pp. Shoshani, J. and Tassy, P. 1996. Summary, conclusions, and a glimpse into the future. In Shoshani, J ., and Tassy, P. (eds). The Proboscidea: Evolution and palaeoecology of elephants and their relatives. Oxford University Press, Oxford, 472 pp. Skeels, MA 1962. The mastodons and mammoths of Michigan. Papers of the Michigan Academy of Science, Arts, and Letters 12103-33. Smuts, M.M.S. and Bezuidenhout, A.J. 1993. Osteology of the thoracic limb of the African elephant (Loxodonta afiicana). Onderstepoort Journal of Vet. Res. 60: 1- 14. Sreevatsan, 8, Pan, X., Stockbauer, K.E., Connell, N.D., Kreiswirth, B.N., Whittam, TS, and Musser, J.M. 1997. Restricted structural gene polymorphism in the Mycobacterium tuberculosis complex indicates evolutionarily recent global dissemination. Proceedings of the National Academy of Sciences 94:9869-9874. Sukumar, R. and Santiapillai, C. 1996. Elephas maximus: status and distribution In Shoshani, J ., and Tassy, P. (eds). The Proboscidea: Evolution and palaeoecology of elephants and their relatives. Oxford University Press, Oxford, 472 pp. Tassy, P. 1996. Who is who among the Proboscidea? In Shoshani, J ., and Tassy, P. (eds). The Proboscidea: Evolution and palaeoecology of elephants and their relatives. Oxford University Press, Oxford, 472 pp. Tassy, P. and Shoshani, J. 1988. The Tethytheria: elephants and their relatives. In Benton, M]. (ed). The phylogeny and classification of the tetrapods, volume 2: Mammals. Clarendon Press, Oxford, 344 pp. Taylor, G.M., Goya], M., Legge, A.J., Shaw, R.J., and Young, D. 1999. Genotypic analysis of Mycobacterium tuberculosis from medieval human remains. Microbiology 145:899-904. Thewissen, J GM. and Domning, DP. 1992. The role of phenacodentids in the origin of the modern orders of ungulate mammals. Journal of Vertebrate Paleontology 12(4):494-504. Thomas, M.G., Hagelberg, E., Jones, H.B., Yang, Z., and Lister, AM. 2000. Molecular and morphological evidence on the phylogeny of Elephantidae. Proceedings of the Royal Society of London (Series B) 267 :2493-2450. Urbain, A. 193 8. L’infection tuberculeuse spontanee chez les mammiferes sauvages en 72 captivite. American Institute Pasteur 61 :705-730. 73 S E R A R m L Y H B F E V m U E T A VI S N A m N m M llllll 1293 02504 4136 ill lllll Hill I 1 Ill 1 3