Evidence of pathology in early Cenozoic turtles J. HOWARD HUTCHISON1 and FREDRIC L. FRYE2 'Museum of Paleontology, University of California, Berkeley, CA 94720; 'formerly at the Department of Biological Sciences, California State University, Sacramento, 6000 3 Street, Sacramento, CA 95819-2964 Survey of fossil turtle shells and fragments from selected Eocene localities of Wyoming yielded numerous examples of non-developmental pathologic injury. A few of these are clearly the result of predator damage indicated by bite holes and gouge marks. A brief review of the causes of the other injuries, particularly sublaminal pitting of the shell, implicates possible sources from fungal, algal, bryozoan, and bacterial infections, but these remain inconclusive on a case by case basis, pending more detailed study, description, and causes of similar injuries in extant turtles. Comparisons of the relative frequency of pathologies in different turtle paleopopulations suggest that environmen- tal conditions may plav a role in the etiology of these pathologies. PalcoBios 21 (3): 12-19, December 21, 2001 CO 2001 University of California Museum of Paleontology INTRODUCTION The abundance of surficial bone in turtles offers a large and readily preserved canvas upon which to chronicle a record of predator depredations and invertebrate and microbial parasitic activity. Little is published on such evidence, espe- cially in early Cenozoic turtles. We document evidence of such activity on a few turtles from the Eocene (40-50 mil- lion years ago) deposits of Wyoming, where they are among the most common fossils encountered. The principal Eocene predators large enough to inflict scars, punctures or fractures on a mature turtle shell were crocodilians and mammals (carnivores, creodonts, and condylarihs). Some fishes such as Lepisosteus(gar) and Amia (bowfin) reached si/.es large enough to cause scarring and possibly punctures in smaller turtles but probably lacked sufficient force to inflict serious damage to the shell on larger turtles. Evidence of parasitic or predator injury is found com- monly on the shells and shell fragments of fossil turtles (Erickson 1984). Aside from the obvious puncture marks, scrapes and associated infection, the agents of other bone erosive marks are difficult to identify. Similar marks and their agents are not well enough described or known in a suffi- cient variety of modern turtles, particularly tropical species, to provide more than speculation on their cause in the Eocene samples. In Wyoming, the Eocene marked a peak of tropi- cal conditions and turtle diversity for the Cenozoic (Hutchison 1982, 1998). We expect that modern analogues of these erosive agents in Eocene turtle shells exist in the American tropics and subtropics. With the hope of stimulating further observation, par- ticularly on extant turtles, we provide the description of selected specimens that illustrate the variety of parasitic and predator lesions encountered in even a limited fossil sample. Material described below is housed in the University of California Museum of Paleontology' (UCMP). Abbreviations: PL—plastral length, CL—carapace length. EXAMPLES OF PATHOLOGY A selection of specimens described below show the range and nature of the injuries encountered, for each, the nature of the pathology is described and probable cause or causes discussed. Figure 1 UCMP 110605, Baptemys fiartnanii(Cope 1872), Fam- ily Derrnatemydidae, portions of the carapace and plastron; estimated PL-23 cm. Locality V73072, Natrona County, Wyoming, Wind River Formation, D>st Cabin Member, early Eocene (Wasatchian). Pathology': The carapace has seven linear excavations; four of these lesions intersect each other at acute to obtuse angles. The edges of the linear foci appear sharp, but when viewed at lOx magnification, the floor of each is characterized by the presence of rounded reactive new bone. Several addi- tional round to elongate shallow pitting lesions are present in some of the carapacial bones. Fragments of the plastron contain numerous partially healed pits. Cause: The linear excavated lacerations were relatively fresh at the time of the turtle's death. The discordant orienta- tion, parallel sides, abrupt terminations and smaller plastral pits could have been caused by either microbiological or- ganisms invading scratches or by invertebrates, and appear to have occurred anteniortcm. Other specimens (UCMP V81114/125353, V81104/125354) with this type of mark show that they may be gently to sharply angled. Figure 2 UCMP 126418, unnamed genus C, family? (= large- headed turtle of Hutchison 1980, Emydid C of Hutchison 1998), incomplete shell; estimated CL—22 cm. Locality V81092, Washakie County, Wyoming, Willwood Formation, early Eocene (Wasatchian). Pathology: Several of the peripheral bones contain crateriform lesions characterized by near-half-thickncss os- teolysis. Most often, these foci cross adjacent sutures to contiguous marginal bones. The edges of these osteolytic lesions are developed well above the floor of the defects onto the surrounding normal bone and are consistent with the formation of dense sclerotic reactive bone at the ad- vancing margin of a lesion characterized as chronic active osteomyelitis, an inflammatory condition of bone. 13 PALEOBIOS, VOL. 21, NUMBER 3, DECEMBER 2001 Fig. 1. Baptemys jjarmanii(UCMP V73072/110605), dorsal view of ncurals and right costal hones showing track-type injury. Scale bar = 1 cm. Anterior is towards the bottom. Cause: Undetermined infection. Figure 3 UCMP 126471, unnamed genus C, family? (= large- headed turtle of Hutchison 1980, Emydid C of Hutchison 1998), incomplete shell; estimated PL—18 cm. Locality V81202, Washakie County, Wyoming, Willwood Forma- tion, early Eocene (early Wasatchian). Pathology: The posterior margins of the carapace are characterized by numerous and extensive osteolytic and os- teogenic lesions. Each of these foci exhibits extensive re- modeling. Cause: Severe chronic and multifocal osteomyelitis. Ini- tial cause unknown. Figures 4-5 UCMP 128263 (Fig. 4), plastron; estimated PL-18 cm; from locality V73108. UCMP 136477 (Fig. 5), left hypoplastron; estimated PL—18 cm; from locality V83002. Both specimens arc of Exhmatemys euthneta (Cope 1872), Fig. 2. Unnamed genus C, family ? (UCMP V81092/126418), visceral view of right part of posterior carapace showing protu- berant and cratcriform lesions (arrows) on the peripherals. Scale bar = 1 cm. Anterior is towards the left. Family Bataguridae, and both localities are in Sweetwater County, Wyoming, Wasatch Formation, Cathedral Bluffs Member, middle Eocene (early Bridgerian). Pathology: The ventral and dorsal margins of the scale- covered portions of the plastron contain multiple punctate pitting lesions measuring from less than 1 mm to 3 mm in diameter. Some of their edges are slightly contoured and smooth; suggesting that these were chronic. Other pits do not appear to have been present long enough to have un- dergone repair. Cause: Undetermined. Figure 6 UCMP 136478, Ecbmatemys euthneta, Family Bataguridae, isolated hyoplastron; estimated PL—19 cm. Lo cality V83002, other same as above. Pathology: The ventral (outer) surface contains a curvi- linear or arc-shaped excavation measuring 7 mm x 47 mm (at its widest margins). The lesion extends through the com- pact membranous bone into the subjacent, cancellous, diploc-like bone. When viewed under 10 x magnification, small spicules are visible. The presence of osseous exten- sions into the void suggests that this lesion had been present long enough for early repair to have begun. Cause: The lesion may represent the drag mark of a preda- tor tooth or claw. Figure 7 UCMP 128262, Echmatemys euthneta. Family Bataguridae, plastron, peripherals; PL—18.6 cm. Locality V73108, other data as above. Pathology: The plastron contains three holes with smooth-sided edges that perforate the bone. The two larg- est defects, one of which lies on the ventral midline, are 10.5 mm x 14 mm and 7 mm x 11 mm respectively. Viewed from the dorsal surface, a markedly osteolytic lesion can be HUTCHISON & FKTE-CENOZOIC TURTLE PATHOLOGY 14 Fig. 3. Unnamed genus C, family ? (UCMP V81202/126471), lateral view ot right posterior peripheral region showing exten sive osteolytic and osteogenic lesions. Scale bar = 1 cm. Anterior is towards the right. seen. It is roughly rectangular and measures 47 mm x 55 mm. The edges of this partially healed defect are elevated an average of 5 mm above the floor of the subjacent area of osteolysis. The surfaces of the reactive sclerotic bone arc smoothly contoured. Cause: Severe chronic and focal osteomyelitis. Ecbmatemys is an aquatic turtle; the initial wound was probably inflicted by a crocodilian. Infection may have contributed to cause of death. Figure 8 UCMP 128283, Ecbmatemyssp., Family Bataguridae, vir- tually complete shell, somewhat asymmetrically crushed; CL-33 cm, PL-29 cm. Locality V83005, Uinta County, Wyoming, Bridger Formation, Black's Fork Member, Bridger B, middle Eocene (Bridgerian). Pathology: A partially remodeled puncture depression in the carapace to the left of the dorsal midline measures 9 mm x 9 mm. Several other shallow pits are located in the carapace. Two full-thickness, partially remodeled pits arc present on either side of the midline. Cause: The deep dorsal wound was probably caused by a crocodilian bite, as there are no known Eocene mammals with a gape sufficient to inflicting a wound so high on the carapace. Figure 9 UCMP 128284, Hadrianus corsoni (Ixidy 1871), Fam- ily Testudinidae, virtually complete shell of a juvenile; PL— 22 cm. Locality V83006, Uinta County Wyoming, Bridger Formation, Black's Fork Member, Bridger B, middle Fxicene (Bridgerian). Pathology: The left side of the carapace contains an arc of three relatively evenly spaced conical perforations. The Fig. 4. Ecbmatemys euthneta (UCMP V73108/128263), ventral view of hypoplastral area of plastron showing Type I and II pit- ting. Scale bars = 1 cm. Anterior is towards the top. anterior depression measures 10 mm wide x 12 mm long x 6 mm deep (maximum measurements). The central depres- sion is located 39 mm from the nearest edge of the anterior lesion and measures 12 mm wide x 21 mm long x 8 mm deep. The posterior perforation is located 25 mm from the nearest edge of the center hole and measures 11 mm x 16.5 mm long x 6 mm deep. All three lesions exhibit full-thick ness compression fractures and displacement of the keratin shields and compact and cancellous membranous bone ven- trally into the subjacent coelomic cavity. The edges of each of the depressed fractures are smooth and exhibit signifi- cant remodeling without the production of excessive reac- tive or sclerotic bone. When viewed from the ventral (coelomic) surface, the lesions exhibit evidence of osteo- genic repair without evidence of osteomyelitic infection. Cause: The shape and spacing of the depressed fractures of the carapace suggest an unsuccessful attack by a large predator, possibly a crocodilian or mammal. A crocodilian large enough (greater than 3 meters) to have produced such large and widely spaced punctures would probably have suc- ceeded in crushing the shell of this rather lightly built and juvenile turtle. Hadrianus is a terrestrial turtle and even 15 PALEOBIOS, VOL 21, NUMBER 3, DECEMBER 2001 :i r -4 I * Fig. S. Echmatemys euthneta (UCMP V83002/T36477), ven- tral view of left hypoplastron showing extensive Type I and III pitting. Note the serrate femoral stale margin caused by exten- sive pitting on the dorsal side. Seale bar = 1 em. Anterior is to- wards the top. specialized extant mammalian carnivores such as felids arc known to prey on terrestrial (and aquatic) turtles (Emmons 1989). Mammalian predators large enough to have produced such marks include Patriofelis (mteYC&rimc breadth—47 mm). If so, at least two bites are required for the three punctures. The second bite may have overlapped one of the punctures of the first (at the central puncture as preserved in the fos- sil); if so, the intercanine distance would be about 40 mm. DISCUSSION In order to determine which types of pathologies were most common and whether pathologies might reflect as- pects of habitat, we subdivided the observed pathologies (examples given above) into several categories and compiled data on their relative frequency. Table 1 presents a tabula- tion of 137 specimens with evidence of parasitic or preda- tor shell injury (non-developmental) from a survey of two middle Eocene (Bridgerian) rock units. Black's Fork and Twin Buttes Members of the Bridger Formation and Ca- thedral Bluffs Member of the Wasatch Formation, of 2 ' v,i* 5 Fig. 6. Echmatemys euthneta (UCMP V83002/136478), ven- tral view of left hyoplastron showing arcuate gouge mark. Scale bar = 1 cm. Anterior is towards the top. southwestern Wyoming. The sample from which the Bridger Formation specimens is extracted is over ten times larger than that of the Cathedral Bluffs Member and samples a larger spectrum of turtles; although, incidence of injuries is not relatively the same. Five families of turtles ate repre- sented and span a variety of habitats. Hndrianus (Tcs tudinidae) is the only terrestrial turtle in the samples. Echmatemys (Family Bataguridae), Baptemys (Dermatemvdidae) and possibly Chisternon (Baenidae) ap- pear to have preferred quiet water. Baena (Baenidae) ap- pears to have preferred river environments (I Iutchison 1984). For purposes of tabulation and reference, the various in- juries are grouped into six categories, three of which are grouped under the general term of pitting. 1) Pitting Type I: Circular to ovoid pits with flat bot- toms (Figs. 4-5). 2) Pitting Type II: Circular to ovoid pits with rounded bottoms (Figs. 4-5). 3) Pitting Type III: Irregular pits and depressions with relatively discrete margins and local extent (Fig. 5). 4) Track Type: Linear and curvilinear path-like marks of relatively constant width (Fig. 1). 5) Rot: Large areas of irregular depression or patches of dead antemortem lamellar bone (Figs. 2-3). 6) Bites: Compression punctures and tapering scratches (Figs. 6-11). HUTCHISON & FRTH-CY.UOZOIC TURTLE PATHOLOGY 16 Fig. 7. Echmatemys euthneta (UCMP V73108/128262) plas- tron showing lesion development. Above, ventral view of plas- tron showing one smaller perforation (arrow) and two larger perforations (*). Below, the dorsal view showing the hill extent of the central lesion which involves the midline sutures of both hyoplastra and hypoplastic, ami showing the relationship be- tween the larger perforations (*) seen on the ventral surface. Scale bars = 1 cm. Anterior is towards the top. Type I and Type II pitting predominates over all other types and may be different stages of the same pathology. While Echmatemys and Baptemys have the greatest number of injuries, the relative significance of this cannot be ex- tended to the other taxa that were drawn from smaller popu- lations. The higher incidence of track type injury in Baptemys compared to Echmatemys appears to be real and may be a reflection of the probable thinner scales in Baptemys. Baptemys grew to a larger size than Echmatemys, and adults may have been slightly less subject to crocodilian predatioti, although this possible difference is not reflected in the data. Pitting occurs in a wide variety of terrestrial and aquatic turtles. There is a diversity of opinion as well as a lack of much definitive data on the causes of pitting especially as it Fig. 8. Echmatemyssp. (UCMP V83005/128283), dorsal view of the posterior part of the earapaee showing two depression punctures (arrows). Scale bar = 5 em. Anterior is to the top and left. affects bone. Ernst (1971) suggested that the high inci- dence of pitting in females of his sample of Chrysemys picta may be related to calcium and phosphate withdrawal for egg laying. This conjecture is not supported by comparable observations in other taxa (Hulsc 1976b). A variety of mi- crobiologic and metazoan pathogens arc implicated in the etiology of plastral and carapacial pitting in chelonians (Scott 1930, Carpenter 1956, Hunt 1957, 1958, Wallach 1975a, Hulsc 1976b, Frye 1981, Griner 1983, Murphy and Collins 1983, Rothschild and Martin 1993). Most frequently, this pitting is a superficial phenomenon, but once the kerati- nized epithelium is sufficiently penetrated to expose the subjacent bony tissue, severe and widespread osteolytic le- sions may be induced. In aquatic and semiaquatic turtles, pitting is associated with what are usually mutualistic epi- phytic and epizoic organisms, particularly the algae Dermatophyton sp., Basicladia chelonum, B. crassa, and a variety of others (Hunt 1958), one or more Phycomycctes (Hulsc 1976a, Wallach 1975b) and the bryozoan Plumatella (Dixon 1960). Ernst (1971) and Hulsc (1976b) stated that epiphytic- growth on the carapace may be causative agents of pitting, but only Hunt (1958) presented direct evidence under arti- ficial conditions of at least one potential means of pitting along the sulci of the dorsal shields. He documented a set of conditions in an aquarium in which the repeated desicca- tion due to basking and re-establishment of new algae lead to a curling-up of the vertebral and pleural shield laminae at their margins culminating in the loss of entire laminae. This cycle is repeated until the epidermal laminae eroded and destruction of the Malpighian layer and underlying bone occurs followed by death at these extremes. Behavior, water quality, air temperature, and humidity may all be factors in 17 PA LEO BIOS, VOL. 21, NUMBER 3, DECEMBER 2001 stabilizing this process at certain stages. The manifestation of this process on the bones was not illustrated or described by Hunt, but from the description, we would expect deep- ) --*< Fig. 9. Hadrianus corsoni (UCMP V83006/128284), left side of a juvenile carapace with three bite compression pits in ventral (top; anterior is to the left) and dorsal (bottom; anterior is to the right) view. In ventral view, one of the 3 compressions is obscured bv matrix. Scale bar = 2 cm. ening, broadening and increasing irregularity of the dorsal sulci on the bones. Fungal (mycotic) origin of the pitting was asserted but not demonstrated as a major cause of necrosis of the shell of some turtles by Hunt (1957, 1958) and Rcickcnbach-Kinke and Elkcn (1965). Jacobson (1980), Grincr (1983) and Frye (1977, 1981) provide direct evidence of fungal infection, although only Fusarium oxysporum and Candida albicans are identified. Reickenback-Kinke and Elkcn ( 1965 ) and Frye (1981) also mentioned Mucorsp. and Dermatophyton sp. as a cause of pitting, but the detailed manifestation of these infections on the bones of the shell are not described or illustrated. Bacteria also appear to be a major cause of pitting (Scott 1930, Wallach 1975a) but only Beneckea chitinovora is posi- tively identified (Wallach 1975a). Murphy and Collins (1983) implicated Mycobacterium as a source of plastral pitting but their cited source for this (Rhodin and Anver 1977) does not record any shell infection or damage. The illustrations of Wallach ( 1975a] of Bcnckca damage in Trionyx resemble in outline those of our Type I-III pits, but detailed mor- phology of the pits on the bones was not provided. The fossil trionychids in our samples are virtually free of similar marks. Mehrtens (1967) notes that "Australian chelids develop sublaminal lesions when kept in water which registers a pH value of 7.6 or more." He figures (p. 17) a live Chelodina lonjjicollis with Type I and Type II pitting within the con- fines of the laminal margins. This may indicate favorable environmental conditions for some microbiological or in- vertebrate agent. Frye (1981) notes that the incidence of fungal infections may be directly related to the pH of the water because growth of most fungi is inhibited at a pH of Table 1. Distribution of shell injury types from two middle Eocene formations in Wyoming. N is total number of specimens show ing some form of injury. Numbers in parentheses are percent of specimens examined within a taxon that show injury. See text tor descriptions of categories. One specimen may exhibit multiple injury types. Taxon -----Pitting — Tracks Roi Bites N Type I Type II Type III Wasatch Formation Echmatemys euthneta 58 1(2) 54(93) - - - 3(5)" Baptemys wyomingensis 3 2(67) 1(33) — — 1(33) — Bridgcr Formation Hadrianus corsoni 3 - 1(33) 1(33) - - 1(33) Echmatemys sp. 40 6(15) 26(65) 8(20) 2(5) 1(2.5) 7(17.5) Baptemys wyomingensis 26 17(65) 15(58) 7(27) 11(42) 6(23) 3(12) Chistemon undatum 5 2(40) 4(80) 3(60) 2(40) — — Bacna arenosa 2 — 1(50) - 1(50) — — TOTALS 137 28(20) 102(74) 19(14) 16(12) 8(6) 14(9) * Includes one scratch mark. HUTCHISON &¦ FRTE-CEK07.01C TURTLE PATHOLOGY 18 less than 6.5. Whether this is the case in the Australian population requires additional field observations or experi- mental data to resolve. Terrestrial turtles have been observed with actively ulcer- ated plastral and carapacial pitting lesions associated with saprophytic feeding of soil invertebrates. This form of kera- tin and superficial bony lesion appears to be most com- monly associated with turtles and tortoises during periods of torpor or hibernation (Frye 1981). Species of the terres- trial emydid Terrapene may have scries of discrete sublaniinal pitting aligned, but not coincident, with the pleural sulci; however, the agent is unknown (Carpenter 1956, Legler 1960). Tooth marks and bite holes in the shell of modern turtles result from both crocodilian and mammal predation. Such large wounds in the shell may provide sites for opportunis- tic parasitism by insects such as the flesh-fly, Sarcopkajja cristudinis (Jackson et al. 1969). Whether these would leave diagnostic traces of their own on the bones is unknown. The data presented in Table 1 show that evidence of para- sitic and/or microbiologic activity is common on fossil turtle shells. Both geographic (site) and host incidence differ for the various types of lesions. The high incidence of pitting in the Cathedral Bluffs sample, mostly from the same local level, suggests that some environmental conditions encour- aged the agents of such marks. This is even more apparent in light of the fact that the Bridger sample yielding Echnatemys'm fable 1 is an order of magnitude larger than that from the Cathedral Bluffs sample. The Cathedral Bluffs unit represents an evaporitic phase of the Eocene Green River lake (Bradley 1926). The presence of both limy de- posits and red beds, indicating aerial exposure and oxida- tion in the local area, suggest an alkaline pond. These types of sediments are rarer or absent in the Bridger Formation. The Cathedral Bluffs Echmatemys are also smaller than gen- erally encountered, further suggesting a marginal habitat. There are interesting parallels between the pitting and habi- tat in Australian chelids mentioned by Mehrtens (1967) and those of the Cathedral Bluffs sample. Although trionychids are attacked by bone erosive bacte- ria (Scott 1930, Wallach 1975a), the fossil specimens that we sampled were notably free from injuries. The natural pitted sculpturing of the bones may also mask minor pit- ting. In the Eocene, diverse crocodilians such as Borealosuchus, Leidyosuchus, Crocodylus, and AUqgnathosuchus were prob- ably the principal predators of adult aquatic turtles. In the Eocene the geographic range of some of these crocodilians extended from coast to coast and north above the Arctic Circle; thus there was no refuge from large aquatic preda- tors as there is today in the middle latitudes. Shallow-water and smaller turtles may also have suffered significant mam- malian predation from such carnivores as Miacis and Uintacyon, creodonts like Patriofelis, and omnivorous condvlarths such as Pbcnacodus. The bite holes we noted were not immediately fatal but once the bone marrow spaces within the membranous bones of the chelonian shell are exposed and invaded by either a predator's attack or by the slower agency of algae, fungi, bacteria or invertebrates, deep, chronic and often aggressive bone-erosive infections may result as well as more fulminating and fatal generalized sep- ticemia. Aside from the bite marks, none of the pathologic pit- ting on the shells of fossil turtles is ascribed to a specific agent or narrow class of agents. Detailed description of the nature and agents of pathologic pitting in the bones of ex- tant turtles may provide diagnostic data for recognition of the etiologic agents in the fossils. Better understanding of the physical environments and their correlation to specific types of pitting may offer insights for paleoecological re- constructions. ACKNOWLEDGMENTS The Annie M. Alexander Endowment to the University of California Museum of Paleontology provided research support for Hutchison. We thank James W. Westgate and Bruce Rothschild for reviewing the paper, and James F. Parham for photography and editorial assistance. This is UCMP contribution number 1753. LITERATURE CITED Bradley, W.H. 1926. Shore phases of the Green River Formation in northern Sweetwater County, Wyoming. U.S. Geological Survey Professional Paper 140-D: 121-131. Carpenter, C.C. 1956. Carapacial pits in the three-toed box turtle, Terrapene Carolina triunjjuis (Chelonia; Emydidae). Southwest- ern Naturalist 1:83-86. Dixon, J.R. 1960. Epizoophytic algae on some turtles of Texas and Mexico. Texas Journal of Science 12:36-38. Emmons, I..H. 1989. Jaguar predation on ehelonians. Journal of Herpetolqgy 23:311 -314. Erickson, B.R. 1984. Chelonivorous habits of the Paleocene croco- dile Lcidyosucbus formidabilis. Science Publications of the Sci- ence Museum oj' Minnesota, new series, 5(4): 1-9. Ernst, C.H. 1971. Observation on the painted turtle, Chrysemys picta. Journal of' Herpctoltiqy 5:216-220. Frye, F.L. 1977. Bacterial and fungal diseases of captive reptiles. pp. 787-791 in R.E. Kirk (ed.). Current Veterinary Therapy VI. W.B. Saunders Co., Inc., Philadelphia. Frye, F.L. 1981. Biomedical and surgical aspects of captive reptile husbandry. Veterinary Medicine 1'ubl. Co., Inc., Edwardsville, Kansas. 456 pp. Griner, I..A. 1983. Pathology of zoo animals. Zoological Society of San Diego. 608 pp. Hulse, A.C. 1976a. Carapacial and plastral flora and fauna of the Sonoran mud turtle, Kinosternon sonoriense (Reptilia, Testudines, Kinosternidae). Journal of HerpetoUnjy 10:45-48. Hulse, A.C. 1976b. Growth and morphometries of Kinosternon sonoriense (Reptilia, Testudines, Kinosternidae). Journal of Herpetoloay 10:45-48. 19 PALEOBIOS, VOL 21, NUMBER 3, DECEMBER 2001 Hunt, T.J. 1957. Notes on diseases and mortality in testudines. Hcrpetologica 13:19-23. Hunt, T.J. 1958. Influence of environment on necrosis of turtle shells. Herpetolojjica 14:45-46. Hutchison, J.H. 1980. Turtle stratigraphy of the Willwood For- mation, Wyoming preliminary results. Univeristy of Michigan Museum of Paleontology Papers on Paleontology 24:115-118. Hutchison, J.H. 1982. Turtle, crocodilian, and champsosaur di- versity changes in the Cenozoic of the north central region of western United States. Palacogeographv, Palaeoclimatology., Palaeoecology 37:147-164. Hutchison, J.H. 1984. Determinate growth in the Bacnidac (Testudines): Taxonomic, ecologic and stratigraphic signifi- cance. Journal of Vertebrate Paleontology 3:148 -151. Hutchison, J.H. 1998. Turtles across the Paleocene/Eocene Ep- och boundary in west-central North America, pp. 401-408 in M.-P. Aubry, S. Lucas, and W.A. Berggren (eds.). Late Pale- ocene-Early Eocene Climate and Biotic Events in the Marine and Terrestrial Records. Columbia University Press, New York. Jackson, C.G., Jr., M.M. Jackson, and J.D. Davis. 1969. Cutane- ous myinis in the three-toed box turtle, 'Perrapenc Carolina triunguis. Bulletin of the Wildlife Disease Association 5:114. Jacobson, E.R. 1980. Mycotic diseases of reptiles. Society for the Study of Amphibians and Reptiles, Contributions to Hcrpctol- o^yl:235-241. Ixgler, J.M. 1960. Natural history of the ornate box turtle, Tcrrapenc omata ornata Agassi/,. University of Kansas Mu- seum of Natural History Publications 11:527-669. Mchrtens, J.M. 1967. Chelids of Australia. International Turtle and Tortoise Society Journal 1(6): 14-21. Murphy, J.B., and J.T. Collins. 1983. A review of the diseases and treatments of captive turtles. AMS Publications, Lawrence, Kansas. 56 pp. Rekhenbach-Kinke, H.H., and E. Elkan. 1965. The principal dis- eases of lower vertebrates. Academic Press, New York. 256 pp. Rhodin, A.G.J., and M.R. Anvcr. 1977. Mycobacteriosus in turtles: Cutaneous and hapatosphenic involvement in a Phrynops hilari. Journal of Wildlife Diseases 13:180-183. Rothschild, B.M., and L. Martin. 1992. Paleopathology: Disease in the fossil record. CRC Press, Inc., Boca Raton, Florida. 400 pp. Scott, H.H. 1930. Tuberculosis in man and lower animals. Medi- cal Research Council (Great Britain), Special Reports Series 149:1-270. Wallach, J.D. 1975a. The pathogenesis and etiology of ulcerative shell disease in turtles. Journal of Zoo Animal Medicine 6:11 - 13. Wallach, J.D. 1975b. Ulcerative shell disease in turtles: Identifi- cation, prophylaxis and treatment. International Zoo Yearbook 17:170-171.