THE GIANT TAPIR, TAPIRUSHAYSU, FROM LEISEY SHELL PIT lA AND OTHER FLORIDA IRVINGTONIAN LOCALITIES Richard C. Hulbert, Jr.1 ABSTRACT Twelve new records of the giant North American tapir Tapirus haysi, are reported from Alachua, Invy, Citrus, Polk, and Hillsborough counties, Florida Biochronologic analysis of associated vertebrates indicates T. haysii was limited to the late early-middle Irvingtonian (ca. 1.5-0.6 Ma) in Florida. The sample of T. haysii from the Leisey Shell Pit 1 A site is the most complete yet recovered, and second only to the Port Kennedy Cave population in numbers ofmeasurable teeth. All known cranial features of T. haysii resemble those of T. veroensts, including a low, double-ridged sagittal crest, ontogenetically delayed fusion of the interparietal with adjacent cranial elements massive mandibular ramus, and relatively short diastemi Phylogenetic analysis of New World Tapirus produces two equally parsimonious cladograms. In both, T. haysii and T. veroensis form one monophyletic clade, and the extant Neotropical species T. pinchaque, T. terrestris, and T. bairdii another. They differ with respect to the position of T. simpsoni, a late Miocene North American species. RESUMEN Se reportan 12 nuevos rei:istros del tapir gigante Norte americano Tapirus haysi, provenientes de los condados de Alachua, Levy, Citrus, Polk y Hillborough, Florida. El andlisis biocrono!6gico de v6rtebras asociadas, indica que la ocurrencia de T. haysii en Florida se limita al Irvingloniano temprano y medio (aproximadamente 1.5 - 0.6 Ma atds). La muestra de T. haysii recuperada del dep6sito de conchuelas de Leisey lA es la mas completa registrada hasta ahora y segunda en nOmero de dientes medibles, con respecto a la poblaci6n de la cueva de Puerto Kennedy. Todas las caracteristicas craneales conocidas para T. haysit se asemejan a aquellas de T. veroensis, incluyendo la presencia de una cresta sagital baja de doble borde; el retraso ontog6nico en la fusi6n del interparietal con los elementos craneales adyacentes; una gran rama mandibular y un diastema relativamente corto. El andlisis filogen6tico de las especies del g6nero Tapints del nuevo mundo producen dos cladogramas igualmente parsimoniosos. En ambos cladogramas, T. hays,i y T veroensis conforman un clado monofilatico, mientras que las especies neotropicales actualmente existentes T pinchaque, T. terrestns y T. bairdii conforman otro clado. Ambos cladogramas difieren con respecto a la posici6n de T. 3/mpsom, especie norteamericana del Mioceno tardio. 1 The author is an Assistant P,ofessor of Geology, Georgia Southern University, Statesboro GA 304604149. HULBERT, R.C., JR. 1993. The giant lapir, Tapirus haysii, from Leisey Shell Pit lA and other Florida Irvingtonian localities. Bull. Florida Mus. Nat Hist 37 PL Il(16):515-551. 516 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. M, NO. 16 INTRODUCTION Two Pleistocene species of Tapirus are currently recognized in eastern and central North America: T. veroensis, a well known species similar in size or slightly larger on average than modern New World Tapirus, and T. h«ysii, a poorly known, larger species. As used here, T. veroensis includes T. excelsus Simpson 1945 (Lundelius and Slaughter 1976), and T. haysii includes T. copei Simpson 1945 (Ray and Sanders 1984). Both species have been recorded from Florida (Sellar(Is 1918; Ray 1957, 1964; Ray and Sanders 1984), although many of the earlier reports of T. haysii probably instead represent T. veroensis (e.g. Saber-tooth Cave and Melbourne; Simpson 1945; Ray and Sanders 1984). Ray and Sanders (1984:296) listed only four specimens from Florida that unequivocally represented T. haysii; none was recovered with an associated fauna. Prophetically they noted, "Undoubtedly more specimens of large Pleistocene tapirs [T. haysii] from Florida are already in collections, or will come to light as new faunas, especially of pre- Wisconsin age, are found. The richness of the record in Florida and its southerly latitude make it the most likely source of the material needed for significant improvement of understanding not only of T. haysii but of fossil tapirs generally." Referred specimens of T. haysii are represented in the collections of fossil vertebrates of the Florida Museum of Natural History and the American Museum of Natural History from 12 Florida localities in addition to those listed in Ray and Sanders (1984). Most were collected since 1983. These specimens are important for several reasons. First unlike all previous records from Florida, and most of those from other regions as well (Ray and Sanders 1984:295497), many were found in direct association with diverse suites of other vertebrates. Various geochronologic methods can be used to date these sites and thereby provide for the first time a biochronologic range for T. haysii in Florida. Second, taken together, the Florida sample of T. haysii now constitutes a sufficiently large, geographically (and probably chronologically) restricted sample with which to make valid quantitative comparisons with modern species of Tapirus, with samples of T. veroensis, and with the only previously known population of T. haysii, that of the Port Kennedy Cave site from Pennsylvania (Simpson 1945). Over half of the Florida material derives from a single site, Leisey Shell Pit 14 which thus forms the second largest sample of T. haysii known, exceeded only by the Port Kennedy population in terms of numbers of measurable teeth. Third unlike the Port Kennedy sample, the material from Leisey IA includes several nearly complete, undistorted maxillae and mandibles, well preserved occipital regions of three crania, and a large sample of postcranial elements. In part, the Leisey IA sample fulfills the wishes of previous authors (e. g. Simpson 1945:66; Ray and Sanders 1984:297) for cranial material of T. haysii, without which its phylogenetic affinities and specific status were uncertain. Unfortunately, some of the phylogenetically most critical regions of the skull, including the nasals and HULBERT: THE GIANT TAPIR, TAPIRUS HA YSII FROM LEISEY SHELL Prr 517 frontals, are not present in the Leisey lA sample. Nevertheless, many previously unknown or poorly known cranial features are represented in the new material. The primary purpose of this study is to describe all records of Tapirus haysii from Florida, concentrating on the largest and most complete sample, that of Leisey Shell Pit lA. Description of cranial material receives the most attention, both because it was previously unknown, and because dental and postcranial elements are very conservative in tapirs, and thus of secondary importance in elucidating phylogenetic information (Simpson 1945; Hershkovitz 1954). The biochronologic distribution of species of Tapirus in Florida is discussed, and a tentative phylogeny of New World Tapirus using cladistic methodology is proposed. ACKNOWLEDGEMENTS I thank the many persons who excavated 1-nisey lA in 1984, and Frank A. Garcia Ronald J. Schrader, William Smith, and Lcroy Doll who donated tapir specimens they collected in 1983. Their combined efforts resulted in the fine sample of T. haysii described in this study. The excellent UF collection offossil Tapirus has also benefitted from the donations of Eric Fernandez, L Mastin, G. Heslep, L Roberts Kent Ainslie. and Pierce Brodkorb. Clayton E. Ray provided valuable advise, casts of comparative material, and permission to study USNM fossils. M. Carlton and L Gordon allowed access to the collection of Recent Tapinis skeletal material at the USNM, and loaned comparative specimens. Access to the F:AM collection was provided by Richard H. Tedford. As always. Gary S. Morgan helped numerous times in dealing with the UF collection. Critical reviews of the manuscript by Clayton E. Ray, Ernest IA,ndelius, and S. David Webb helped improve the study. This is University of Florida Contribution to Paleobiology No. 345. ABBREV[ATIONS AMNH - Depailment of Vertebrate Paleontology, American Museum ofNatural History, New York. ANSP - Academy of Natural Sciences, Philadelphia- F:AM - Frick Collection. housed with AMNH collection UF - Florida Museum ofNatural History, University of Florida, Gainesville. UF/FGS - Florida Geological Survey collection offossil vertebrates, housed with UF collection. USNM - National Museum ofNatural History, Smithsonian Institution, Washington, D.C. Vi - upper/lower incisor. C/c - upper/lower canine. P/p - upper/lower premolar (e® P# is an upper fourth premolar) M/m - upper/lower molar (e.g; m2 is slower second molar). D/d - upper/lower deciduous tooth (e.& 42 is a deciduous lower second premolar). MC/MT - metacarpal/metatarsal. x -sample mean 3 - sample standard deviation. n - sample size. CV- sample coefticient of variation. OR - observed range ofa sample. MIN -observed minimum value ofa sample. 518 BULLETIN FLORIDA MUSEUM NATURAL HISrORY VOL. 37, Fr. II, NO. 16 MAX - observed maximum value ofa sample. assoc.-associated. Ma - Mega-anna. millions ofycars before present myr - millions of'years (in duration) L-length AW - anterior width. PW - posterior width. BASALLTH - basilar length ofskull. PALLTH - palate length. SAGCREST- height from basioccipital to top ofskull. 13PIDLTH - diaste,na l=:2* between 13 and Pl alveoli. CPiDLTH - diasuma length between C and Pl alveoli. CANLTH - upper canine alveolar length. CANWTH - ,.:,rwr ...in. alv-lar wiah- MUZWDTH - upper munle' **' * Bi CONDWDTH - breadth of waphl an#la P1M3LTH - length ofupper cheektooth row. P1P4LTH - length of upper premolar scric& MIM)LTH - length ofupper molarseries GRTLTH - greatest length ofmandible. i3p2DLTH - diastema length between B and p2 alveoli. cp2DLTH - diastema length between c and p2 alveoli. SYMPHWDTH - symphysial width across the lower canines. SYMPHLTH - length ofthe mandibular symphysis. p2MDFI11 - madibular depth measured anterior to the p2. mJMDrrH - mandibular depth measured posterior to the m3. canLTH - lower canine alveolar length. canWTH - lower cnnine alveolar width p2m3LTH - length of lower cheektooth series p2,}4LTH - length of lower premolar series. mlm3LTH - length of lower molar series. MATERIALS AND METHODS The vast madority of fossil specimens described herein are housed at the Florida Museum of Natural History, either in the UF or UF/FOS collection. The USNM has a comparatively small collection of Lcisey 1A material, including three specimens of Tapirus that were examined and nicasured. The F:AM collection contains material from one major Irvingtonian site in Florida, McLeod Limerock Mine, Pocket A ne limited sample of Tapirus from McLeod was examined and measured. Tapirus specimens retained by private colle¢torn were not included in this study unless casts we deposited in the UF collection. The localities that produoed the specimens under study are all of Irvingtonian age; detailed descriptions. locality information, and a relative chronology ofthese sites are presented in Webb (1974) and Morgan and Hulbett (this volume). The Apollo R-ch mandible recovered from dredged material provisionally regarded by Ray (1964) as Rancholabrean is much more likely to be Irvingtonian. Unbeknownst to Ray, strata of - 1 ''y widespread is southern Hillsborough County, and tan restricted to the Irvingtonian have been fbund at Apollo Beach (e.g. Pachyarmatherium leiseyi). Comparative material examined included specimens of all four extant species of Tapirus (T. bairdil, 21 skulls, USNM and UF; T. terrestris, 20 skulls, USNM and UF; T. pinchaque, 3 skulls, USNM; T. indicus, 4 skulls, USNM), a large sample of T. veroemis from the Rancholabrean of Florida (including holotype skull UF/FOS 277; numerous palates and maxillae, and mandibles). and referred material of two late Miocene species, T. johnioni and T. simpsoni (F:AM and UF; Yarnell 1980). Museum acronyms used HULBERT: THE GIANT TAPIR, TAPIRUS 84]SU FROM LEISEY SHELL Prr 519 in conjunction with Recent species refer to specimens in mammalogy oollections; those with fossil species refer to specimens in vestebrate paleontology collections Up to 12 cranial and 12 mandibular characters (described in Abbreviation section above) were measured Values greater than 200 mm were taken with anthropometers to the nearest 0.5 mm; those less than 200 mm were taken to the nearest 0.01 mm (but later rounded offlo nearest 0.1 min) with either dial or digital calipers. Only adult individuaIR were measured for cranial and mandibular variables, as judged by the cruption of the P4/p#. The standard three dental measurements for tapir cheekieeth (Simpson 1945), length. anterior width, and posterior width. were taken with calipers (to nearest 0.01 mm). except for the DPI, Pl, dp2,and p2 fof which only length and posterior width were taken Maximum tooth length was measured. umially across the®doloph for uppers and along themidline for lowers. except when lessened by interdental wear. Maximum width was measured at about the enamel-dentine boundary, across the protoloph and metatoph on upper cheekteeth, and the protolophid and hypolophid on lowers. Postcranial elements were measured with a subset of Eisenmann's standard characters used in the study of equids (e* 1986). These generally included greatest length, diaphysial (mid€haft) transverse width and A-P diameter, and width and A-P diameter of the proximal and distal articular surfaces nese data were entered onto a microcomputer and statistically analyzed using the Quattro spreadsheet program Based on examination of the comparative specimens listed above and a review of the literature (especially Sellar€Is 1918; Simpson 1945; Colbal and Hooijer 1953; Hershlcovitz 1934; 1,undelius and Slaughter 1976; and Ray and 6,*4 19841 26 crani•l and dental characters were selected for use in a phylogenetic analysis using the PAUP computer program, version 2.4 (Swoflord 1985). ne BRANCH AND BOUND option of PAUP was used to produce the most parsimonious cladogram(s) possible based on the provided dataset SYSTEMATIC PALEONTOLOGY Order PERISSODACTYLA Owen 1848 Family TAPIRIDAE Burnett 1830 Genus TAPIRUS BrOnnich 1771 Tapirus haysii Leidy 1859 Selected Synopsis of Usage: Tapinu haysil Lcidy 1852a:106 (nomem nu,hlm); Leidy 1852b: 148 (nomen nudum); Leidy 1855:200 (nomen nudum); Lcidy 1859:106 (first valid description); Sellards 1918:59; Simpson 1945:65-66 (regarded as nomen dubium); Ray and Sanders 1984:288-298 (type locality and topotypes described, name regarded as valid). Tapirus copel Simpson 1945:66-69; Ray 1964:61-66; Strain 1966:48-50; Hager 1974:13; Lundelius and Slaughter 1976:227, 237-239; Ray and Sanders 1984:287, 293-296 (regarded as junior synonynn of T. haysiO. Type Specimen.- ANSP 11504, R p4 (Leidy 1859, pl. 17, figs. 7 and 8; Ray and Sanders 1984, fig. 2D). Type Locality.- Benners Estate, Pamlico County, North Carolina (see Ray and Sanders 1984:288-293 for an extensive discussion of the location and geology of this site). 520 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. II, NO. 16 Distribution.- Late Pliocene (late Blancan) through middle Pleistocene (Irvingtonian). Ray and Sanders (1984:293-297, fig. 3) demonstrated a near continuous range for T. hqysii across the southern half of the United States east of the Rocky Mountains, with northernmost records in north-central Nebraska and southeastern Pennsylvania. All records of demonstrative age from Florida are early to middle Irvingtonian, ca. 1.54.6 Ma. Western specimens referred to T. merriami by Jefferson (1989) are tentatively retained in that species pending critical comparison with T. h«yi. Referred Florida Specimens.- Haile 164 Alachua County. UF 91243 R MC m; 91244 L distal MC III; 91245 L MC V; 91246 L distal tibia; 46923 R Mr IV; 91247, 91248 2 R astragali; 91250 2 proximal phalanges, lateral digits. Haile 21 4 Alachua County. UF 62601 L mandible with dp2-ml (cast only); 62602 L mandible with m3. McLeod Limerock Mine, Pocket 4 Levy County. F:AM 37410 L maxilla with DP3-DP4; F:AM 37411 R mandible with dp2-dp3. Crystal River Power Plant Citrus County. UF 17336 proximal phalanx, digit III. Pool Branch, Polk County. UF 11416 L M3. Chicora Mine, Polk County. UF 65575 L scapula. Phosphoria Mine, Polk County. UF 40060 R M2. Tiger Bay Mine, Polk County. UF 22590 L P3. Apollo Beach, Hillsborough County. UF 8225 assoc. R mandible with il-i2, p2-m3 and L mandible with il, p4-m2. Leisey Shell Pit Hillsborough County. Leisey lA: UF 89539 assoc. L maxilla with P2-M) and R&L mandibles with il-m3 (cast only); 82783 assoc. R & L maxillae with DPI-DP4; 84190 assoc. R&L maxillae with Pl-M3; 82782 R maxilla with P4-M3; 87231 R maxilla with Pl-P3, DP4, Ml-M2; 60873 R maxilla with P3-P4; 60874 L maxilla with P2-P), DP4, Ml; 60872 L maxilla with Ml-M3; 65974 L maxilla with P2-P3; 65999 R edentulous maxillary fragment; 86113 assoc. premaxillae with R I143 and L Il, I3; 86363 assoc. premaxillae with R Il- I3 and L I3; 80446 posterior half of skull; 83305 occipital region and R parietal; 83835 occipital region and L squamosal; 85315 L supraoccipital; 80973 assoc. mandibles missing only Ri3 and c; 83580 assoc. mandibles with R&L p2-m3; 87230 assoc. mandibles with R dp2-ml and L dp2-di,3, ml; 87947 L mandibular fragment with dp4; 65971 mandibular symphysis with R il, c, p2 and L c; 81103 R DPI; 82784 L DP4; 86193 R Pl; 80240 L P2; 87948 R Ml; 89533 R M3; 65972, 86777, 86851, 87956, 88117 5 partial upper molars; 82876 L C; 83307, 86890, 81105 3 L Ils; 81104 L I2; 88778 L I3; 65973, 80838 2 R p2s; 87941 R p3; 86940, 86945 2 L p3s; 87935 L ml; 86941, 87937 2 R m2s; 63899, 87942 2 R m3s; 86790, 87468, 87936, 87944 4 partial lower cheekteeth; 86741 R c; 83993 L c; 87951 R il; 80528, 80972, 87946 3 L ils; 86690, 87949 2 R i2s; 87954 L i2; 88250 R i3; 65978, 84362 2 atlases; 65979, 83173 2 axes; 67538 R scapula; HUI.BERT: THE GIANT TAPIR, TAPRUS84YTH FROM LEISEY SHELL Prr 521 65980, 67539 2 L scapulae; 87229 R distal humerus; 87228/87945 R radioulna; 81068 L radioulna· 83721 R magnum; 87955 R scaphoid; 65996 L scaphoid; 125413 R lunar, 81089 L lunar, 87467 R cuneiform; 80447, 87943 2 R unciforms; 81381 L unciform; 85317, 87938 2 R pisiforms; 80576, 81915 2 L pisiforms; 67536, 87958, 89535 3 R MC IIs; 83389, 87959 2 L MC IIs; 87950 R MC m; 81771, 89534 2 L MC Ins; 65990, 89537 2 R MC IV; 83390, 83658, 89538 3 L MC IVs; 84082, 87952, 88777 3 R MC Vs; 82428, 84916 2 L MC Vs; 65975, 80449, 83441 3 L femora; 67537 patella; 65977 R tibia; 65976, 80653, 83579 3 L tibiae; 86866 R fibula; 81633 L fibula; 65986 R calcaneum; 65983-65985, 81235 4 L calcanea; 65982, 86883, 87705, 87960 4 R astragali; 63901, 65981, 67325, 81917, 82403, 87953 6 L astragali; 65997 R navicular, 65998, 85316, 86913 3 L naviculars; 80990 R cuboid; 65992 R ectocuneiform; 65988, 65989 2 R MT IIs; 81311, 82374 2 L Mr IIs; 65987 R Mr II; 81703, 83578 2 L Mr I[Is; 86865 R Mr IV; 63900, 84442, 89536 3 L Mr IVs; 65993, 65994, 80604, 86338, 89532 5 proximal phalanges, digit m; 82375, 83816 2 distal phalanges, digit I[I; 84081, 84169, 87939, 87940, 89528, 89529 6 proximal phalanges, lateral digits; 65996, 81918, 83672, 86852, 88249, 89530, 89531 7 medial phalanges, lateral digits; 82121, 82463, 83996 3 distal phalanges, lateral digits. Ikisey 18: UF 95648 R mandible with m3; 60864 L femur. Leisey 1: UF 115945 RP2; 115946 R maxilla with Ml-M2; 115947 R mandible with m3; 115948 L mandible with m2-m3; 115949 L m2. Idsey 2.: UF 115950 R P3. Leisey 3A: UF 115943 R DP4. Leisey 3B: UF 115944 L P3. Leisey 3: UF 130017 atlas; 142233 L MI' IV; 142234 R astragalus. Revised Diagnosis.- Significantly larger than T. johnsoni, T. simpsoni, T. veroensis, T. bairdii, T. terrestris, or T. pinchaque, with only minor overlap Of any) in cranial and dental dimensions. On average, larger than but with considerable overlap in OR of cranial and dental dimensions of T. indicus. Smaller than T. augustus and, on average, T. merriami. Lower toothrow length (p2-m3) 153 to 167 mm. Development of sagittal crest as in T. johnsoni, T. veroensis, and T. pinchaque; low, double-ridged in subadults; not like either the sagittal table of T. bairdii nor the high crest of T. terrestris. Large, subtriangular interparietal remains unfused relatively late in ontogeny, until at least the full eruption of M3. Lambdoidal crests strongly developed, flair outward. Posterodorsal process of premaxilla terminates in an acute point. Upper premolars relatively "molarized," with well developed protolophs on P2 and P3, and a variably developed transverse loph on the Pl. Description.- The following description of cranial features of T. haysii from Leisey lA follows the same sequence as the comprehensive summary presented by Ray and Sanders (1984:303-308) that compared T. veroensis with the four extant species of Tapirus. 522 BULLEMN FLORIDA MUSEUM NATURAL HISrORY VOL. 37, Fr. II, NO. 16 Both available premaxillae separated from their respective maxillae along the sutures. On the right side of the more complete specimen~ UF 86113, it is evident that the posterior process of the premaxilla ends in an acute point, although the posteriormost portion is broken off. In dorsal view, the premaxillary-maxillary suture veers slightly laterally, which would have allowed minor exposure of the maxilla medial to the premaxilla. This arrangement is most similar to that observed in the majority of specimens of T. veroensis and T. terrestris. In T. indicus the amount of exposed maxilla is greater; in T. pinchaque (and some T. terrestris and T. veroensis) very little or none of the maxilla is exposed; and in T. bairdii the premaxilla terminates more bluntly. The greatest length of the premaxilla is about 105 mm (this assumes about 5 mm broken off UF 86113). In T. veroensis, premaxillary length is 89 to 95 mm (n = 4). The dorsal margins on all available maxillae are broken. The most complete specimen in this regard is the juvenile UF 82783. Portions of the maxilla dorsal to the infraorbital foramen are preserved, and show the beginnings of the smoothly rounded dorsomedial border evident on maxillae of all other Tapirus except T. bairdii. T. bairdii has a strong dorsally directed flange on the maxilla in this region. Since this flange forms very early in ontogeny in T. bairdii (Ray and Sanders 1984), the young age of UF 82783 does not contribute to its absence. Although conclusive evidence is lacking, T. hagsii probably lacked the ossified anterior mesethmoid cartilage of T. bairdii, as it is associated with the development of the dorsal maxillary flange. No portions of the posterodorsally ascending process of the maxilla. the lacrimals, nasals, or frontals are preserved in the available Leisey lA sample, or any other sample of this species. The development of the sagittal crest in T. haysii is shown in two specimens, UF 80446 and 83305, both of which include the occipital region and at least one complete parietal (Fig. lA). In both the cmnial suturcs between the parietals, interparietal, frontals, and supraoccipital had not fused, implying that the age at death for both individuals was at most young adulL USNM 14219, a female T. bairdii with an erupted but only slightly worn M3, retains similarly open cranial sutures, as does the holotype of T. veroensis, UF/FGS 277. UF 83305 was found in close proximity with a number of other specimens, which together probably represent a single individual. These include the palate UF 84190, premaxillae (UF 86113), atlas (UF 84362), axis (UF 83173), and mandibles (UF 83580). The M3 of UF 81490 is slightly worn and represents the same wear stage as the M3 of USNM 14219. UF 83305 therefore very likely represents the young-adult stage of development of the sagittal crest in T. hqysii. The sagittal crest on the anterior portion of the parietal begins the wide divergence that would have been carried further on the frontals. The low, double-ridged, narrow crest is formed solely on the parietals. Both UF 80446 and 83305 have very similar sagittal prests to those of young adults of T. johnsoni (Yarnell 1980), T. veroensis (Lundelius and Slaughter 1976), and T. pinchaque (Hershkovitz 1954; Ray and Sanders 1984). At HULBERT: THE GIANT TAPIR, TAPIRUS HAYSH FROM LEISEY SHELL Prr 523 B ..P7.2- r flr' Cr 3 - . r.9/0 6& a i:i Figure 1. Dorsal (A), ventral (B),and posterior (C) views ofUF 80446, partial skull of Tapirus hays,i from Leisey Shell Pit 14 Hillsborough County, Florida Length of scale bar = 10 cm. BULLEI'IN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. II, NO. 16 524 A B - Figure 2. Labial (A) and lingual (B) views of UF 80973, left mandible of Tapirus haysii from Leisey Shell Pit 1 4 Hillsborough County, Florida. See Figure 58 for an occlusal view of the cheekteeth of this specimen Lzngth of scale bar = 10 cm. its narrowest point, the crest on UF 80446 is about 12 mm across, that of UF 83305 about 10 mm. The latter is probably the older individual of the two. This type of sagittal crest (with minor differences) is found on all Tapirus except T. bairdii and T. terrestris. T. terrestris develops a very pronounced sagittal ridge, which is evident even injuvenile individuals (Hershkovitz 1954). T. bairdii retains a broad, low sagittal table throughout ontogeny, a modification of the juvenile condition in other species. Following Simpson (1945:56), most authors have stated that T. indicus has a sagittal table similar to T. bairdii. This is true only in juveniles, but adults develop a double crest across the parietals as in T. veroensis (e.g. USNM HULBERT: THE GIANT TAPIR, TAPIRUSHA YS// FROM LEISEY SHELL Prr 525 155410; Colbert and Hooijer 1953). The dorsal surface of the supraoccipital is flat and smooth, not expanded and rugose as in T. indicus and T. augustus. UF 80446 and 83305 both ret=ined large, unfused interparietal bones. Although not preserved, their shape and size can be discerned from the vacancy left in the cranium (Fig. lA). The interparietal is approximately the shape of an isosceles triangle, about 36 mm long and 26 mm wide on UF 80446,25 mm long and 29 mm wide on UF 83305. On a third, and apparently older, individual (UF 83835) the interparietal had fused with the supraoccipital and the suture between them is not visible. These specimens suggest that T. haysii, like T. veroensis, retained an independent interparietal into early adulthood, but that it fused with the surrounding cranial elements in older adults. The lambdoidal crests are strongly developed, more so than T. veroensis, and project both laterally and posteriorly (Fig. 14 10. In this they resemble most individuals of T. veroensis, T. pinchaque, and T. indicus, and differ from T. bairdii and T. terrestris in which there is much less of a lateral component (Simpson 1945). The strong development of these crests is related to the increased size of the species (Colbert and Hooijer 1953). The anterodorsal process of the maxilla that projects medial to the premaxilla would have been narrowly visible in lateral view. It thus resembles most individuals of T. veroensis and T. terrestris in this region. The ventral premaxillary-maxillary suture in lateral view descends anteroventrally to just anterior to the canine, as in all other New World Tapirus. In T. indicus and T. augustus, the suture descends towards the middle of the canine (Ray and Sanders 1984). The postglenoid and mastoid processes converge very slightly, but the external auditory meatus is open ventrally on the left side of UF 80446 (Fig. 1B). This is the only specimen where both processes are well preserved. Their convergence is less than in T. veroensis (e.g. UF/FGS 277), and is most similar to T. indicus among living tapirs in this regard. Quantitative comparisons of cranial, mandibular, and dental characters of T. haysii with other species are presented in Tables 1-3. Greatest length of the mandible is the best indicator of body size known by several individuals in T. h®sii (n = 4, including one individual from Port Kennedy). For this linear character, T. haysii is about 13% greater than T. bairdii or T. veroensis, and 20% greater than T. pinchaque or T. terrestris. Among living tapirs only T. indicus displays considerable overlap with T. haysii in size related dimensions (Tables 1- 2). Simpson (1945:59) erred when he stated that the teeth of T. indicus were smaller than those of T. veroensis. The "giant" T. hoysii is dwarfed by the enormous Asiatic Pleistocene T. augustus, which has a mandibular length 30% greater than that of T. haysii (445-450 mm, Colbert and Hooijer 1953) Relative muzzle width of T. haysii and T. veroensis appear to fall within the "normal" range for Topirus when adjusted for size, but diastema length is on average lower (Fig. 2; Table 2). T. bairdii has a relatively long muzzle and small teeth. Relative to 526 B U LLE TIN FLO R ID A M U S E U M N A H IS rO R Y V O L. 37, F r. II, N O . 16 Table 1. Univariate statiatics for cranial measurements of Tapirus. BASAL PAL SAG P1M3 PIP# M1M3 BPlD CPlD CAN CAN MUZ COND LTH LTH CREST LTH In[ LTH LTH LTH LTH WTH WDTH WDTH Tapirus bairdii n 18 21 20 18 20 19 21 21 20 20 21 18 379.39 223.43 120.34 138.17 74.94 65.59 74.84 53.89 11.38 9.87 50.49 83.26 11.45 8.27 5.25 5.27 3.30 2.56 5.00 4.50 1.63 1.18 2.96 4.18 MIN 361.0 207.5 110.0 126.0 67.2 61.1 65.8 43.5 7.2 7.6 44.6 77.1 MAX 409.00 239.0 132.7 148.4 80.9 70.8 84.0 60.7 14.6 11.9 55.1 91.9 3.02 3.70 4.36 3.82 4.40 3.90 6.68 8.35 14.30 11.94 5.87 5.02 Tapirus terrestris n 16 18 16 18 18 18 18 17 18 18 16 16 350.59 190.77 122.78 134.75 71.83 65.37 63.31 47.92 8.74 7.20 42.97 71.40 s 15.83 8.72 5.99 4.93 3.42 2.48 4.27 4.35 2.46 1.96 3.37 3.88 MIN 328.0 172.5 115.5 125.3 66.0 61.0 56.3 40.7 0.0 0.0 37.0 64.7 MAX 384.0 211.0 132.2 147.4 80.8 71.4 73.1 57.4 11.0 8.8 50.0 79.3 (17 4 .52 4 .57 4 .88 3 .65 4 .77 3 .80 6 .75 9 .07 28 . 14 27 .22 7 . 85 5 .44 Tapirus pinchaque n 333232333333 347.50 200.33 97.12 141.01 74.12 67.67 67.18 48.21 10.28 8.35 43.55 76.27 7.47 3.51 5.06 0.83 3.85 1.00 7.89 6.38 0.35 0.58 4.08 2.55 MIN 342.0 197.0 91.3 140.4 69.8 67.0 58.2 40.9 9.9 7.7 40.6 74.5 MAX 356.0 204.0 100.7 141.6 77.3 68.4 72.8 52.8 10.5 8.7 48.2 79.2 (17 2. 15 1 .75 5 .21 0 .59 5 .20 1 .47 11 .74 13 .23 3 .37 6 .96 9 .37 3 .34 Table 1 continued. BASAL PAL SAG PIM3 P1P4 M1M3 DPID CPiD CAN CAN MUZ COND LTH LTH CREST LTH LTH LTH LTH LTH LTH WTH WDTH WDTH Tapirus indicus n 333333333333 403.33 222.00 115.58 163.97 89.28 78.05 71.35 49.85 10.30 8.78 50.32 89.23 4.37 6.08 8.07 11.06 5.11 6.52 8.23 8.43 1.80 1.07 0.69 4.02 MIN 398.5 218.0 106.5 151.8 84.5 70.6 62.0 40.1 8.4 7.7 49.5 86.9 MAX 407.0 229.0 122.0 173.3 94.7 82.4 77.5 55.1 12.0 9.8 50.9 93.9 (N 1 .08 2.74 6 .98 6 .75 5 .73 8 .35 11 .54 16 .91 17 .51 12 .24 1 . 37 4 .50 H U LB E R T: TH E G IA N T TA P IR , TAPIR U S H AYS# FR O M LE IS E Y S H E LL PrT 527 Tapina veroensis n 242598564432 349.50 184.72 106.03 145.77 77.97 71.08 59.53 44.04 7.80 7.85 46.36 77.66 17.68 6.32 1.23 8.22 4.24 4.20 5.86 5.42 0.93 0.42 0.94 3.08 MIN 337.0 179.9 105.2 139.5 71.0 65.5 50.9 35.0 6.8 7.3 45.7 75.5 MAX 362.0 194.0 106.9 159.0 84.2 77.9 65.6 50.1 9.0 8.2 47.4 79.8 (N 5 .06 3 .42 1 . 16 5 .64 5 .44 5 .90 9 . 85 12.30 11 .98 5 .31 2 .02 3 .97 Tapirus haysii n 003114002223 -- 115.23 166.58 87.69 79.83 13.86 13.29 50.25 83.50 -- 3.20 - - 1.63 0.13 1.07 2.19 1.03 MIN -- 112.0 166.6 87.7 77.5 13.8 12.5 48.7 82.5 MAX -- 118.4 166.6 87.7 81.3 14.0 14.1 51.8 84.5 CV - 2.78 2.04 0.92 8.09 4.36 1.23 528 BUI I.RI'IN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. II, NO. 16 Table 2. Univariate statistics for mandibular measurements of Tapirus. GRT Bp2D cp2D p2m3 p2p4 mlm3 SYMPH SYMPH p2M in3M can can ith Ith Ith Ith ith Ith wdth Ith dpth dpth Ith wth Tapirus bairdii n 20 21 21 17 18 19 21 21 21 20 21 21 304.23 81.44 65.30 128.33 61.93 66.86 50.95 92.45 47.59 55.25 16.78 14.33 s 9.80 4.74 4.72 4.86 2.63 2.35 3.43 5.06 2.20 3.14 1.61 1.48 MIN 289.0 71.7 53.9 119.7 58.3 61.7 47.6 80.1 41.7 51.3 13.4 11.6 MAX 323.0 92.6 73.1 138.5 67.9 70.9 61.4 100.9 51.3 64.6 19.6 17.1 (17 3 .22 5 . 82 7 .23 3 .79 4 .25 3 .51 6 .73 5 .47 4 .63 5 .68 9 .59 10 .35 Tapirus terrestris . X m 14 14 15 14 14 15 13 13 15 15 15 15 283.64 68.32 54.73 126.60 59.87 67.38 44.06 89.77 44.75 52.21 14.85 11.71 13.66 4.73 4.64 5.51 2.97 2.77 4.34 5.84 4.94 3.56 2.01 1.10 MIN 266.5 62.5 48.2 118.5 55.9 63.0 36.6 80.8 37.6 46.9 11.4 10.2 MAX 315.0 80.2 65.9 141.1 67.1 74.8 53.8 98.2 54.6 57.4 17.3 13.3 crq 4 .82 6 .92 8 .49 4 .35 4 .96 4. 11 9 . 86 6 .50 11 .03 6 . 82 13 .57 9 .37 Tapirus pinchaque n 333121333233 287.83 72.12 56.83 132.85 63.48 68.70 45.22 83.54 43.15 50.06 15.75 12.77 s 1 .71 4 .97 6 .08 - 1 .60 - 2 .96 5 .32 1 .59 0 .65 1 .25 0 .31 MIN 281.5 66.5 49.9 132.9 62.4 68.7 42.5 77.4 41.4 49.6 14.5 12.5 MAX 296.5 75.8 61.4 132.9 64.6 68.7 48.4 87.2 44.4 50.5 17.0 13.1 (77 2.70 6 .89 10.69 2.52 6 .54 6 .37 3 .68 1 . 30 7.94 2 .46 Table 2 continued. GRT Op2D cp2D p2m3 p2p4 mlm3 SYMPH SYMPH p2M naM can can Ith Ith 1th Ith Ith ith wdth Ith dpth dpth ith wth Tapints indicus 333333333233 336.83 77.48 59.26 151.86 73.58 79.05 52.97 102.38 54.14 60.22 17.68 13.44 3.88 4.56 5.58 12.63 4.05 7.58 2.47 2.51 2.46 2.98 0.90 1.32 PE# *.~ 332.5 72.6 53.3 137.9 69.5 70.4 50.8 99.6 51.6 58.1 17.0 12.6 340.0 81.6 64.3 162.5 77.6 84.5 55.7 104.5 56.5 62.3 18.7 15.0 H U I-B E R T: TH E G IA N T TA P IA , TAPIRUS HAYSH FR O M LEISEY S H E LL P rr 529 1.15 5.89 9.42 8.32 5.50 9.58 4.66 2.45 4.54 4.96 5.06 9.80 Tapirus veroensis n 133225322333 X 308.00 59.58 46.76 134.38 65.97 72.11 44.53 81.02 45.96 61.15 11.83 11.00 - 5.48 3,09 0.37 0.10 2.66 3.47 1.97 5.46 3.52 0.54 0.99MIN 308.0 56.3 44.9 134.1 65.9 69.4 41.1 79.6 42.1 58.5 11.4 10.0MAX 308.0 65.9 50.3 134.7 66.0 75.2 48.0 82.4 49.8 65.1 12.4 12.0(N 9 .20 6 .61 0.28 0. 15 3 .69 7.79 2.44 11 . 88 5 . 75 4 .56 8 . 96 Tapirus haysii n 454556346544 X 343.63 74.31 61.20 158.14 73.87 82.81 52.89 93.22 53.78 71.63 13.59 12.60 S 13.54 3.68 5.40 6.13 2.72 4.83 1.60 8.92 4.48 4.15 0.68 0.49 MIN 326.0 70.5 55.7 153.4 71.7 76.7 51.1 82.8 45.6 67.3 12.8 12.3MAX 359.0 79.9 68.4 166.1 78.5 89.6 54.2 102.4 58.2 78.2 14.2 13.3 3.94 4.95 8.83 3.88 3.68 5.83 3.02 9.57 8.32 5.79 5.01 3.93 530 BULLErIN FLORIDA MUSEUM NATURAL HISrORY VOL. 37, Fr. II, NO. 16 modern Tapirus, both T. haysii and T. veroensis have smaller lower canines and deeper mandibles below the m3 (Table 2; Fig. 2). Measured characters of the cheekteeth show only minor differences between the referred Florida and Port Kennedy populations of T. hoysii (Table 3). Only four of 37 parameters (P2AW, P4L, MlL, and p4L) are significantly different based on univariate t-tests (0.05*ZO.01). The Florida T. haysii population is significantly larger on average than that of T. veroensis for 36 of the same 37 parameters, although there is overlap in almost all variates (Table 3). The T. veroensis sample was derived from numerous localities, of which only a few (e.g. Vero, Arredondo, Hornsby Springs, Rock Springs) were collected in situ with an associated Rancholabrean fauna. The majority instead came from the beds of rivers, including the Ichetucknee, Sante Fe, Aucilla, Wacissg Waccasassa, and Oklawaha. The preponderance of vertebrates produced by these are Rancholabrean, but the presence of early Pleistocene specimens is possible. A similar sample was measured by Lundelius and Slaughter (1976; it included many of the same UF specimens), who also found more substantial overlap in tooth dimensions between T. veroensis and T. haysii (their T. copei) than did Simpson (1945). Their identifications were questioned by Ray and Sanders (1984), who queried why large individuals assigned to T. veroensis did not represent T. haysii. In the case of UF 8225, the Apollo Beach mandible, Ray and Sanders were undoubtedly correct in assigning this specimen to T. haysii (its absence probably accounts for the much larger observed maximum values for T. veroensis lower cheekteeth reported by Lundelius and Slaughter [1976, table 3] than those listed here in Table 3). Other large-sized, putative T. veroensis individuals are less easily reassigned. In many cases the longest tooth for a particular locus is not the widest, many having widths that fall below the OR of T. haysii, and their apparent great length is because they are unworn and have not suffered interdental wear. Thus to reassign species identification to T. haysii on the basis of single measurements would cause substantial decreases in the means and observed minima in other characters, and this would in turn lead to greater apparent overlap between the two species. The only morphological criterion of the dentition that has been used to separate the two species other than size is the more highly molarized anterior premolars of T. haysii (Simpson 1945; Ray and Sanders 1984). A quantitative measure of this is the ratio P2AW/P2PW, which is greater (on average) in T. haysii than T. veroensis. If the larger individuals referred to T. veroensis by Lundelius and Slaughter (1976) and this study are actually Rancholabrean T. haysii, then large P25 (those with lengths greater than 20.6 mm, the MIN for T. haysii) should have relatively high AW/PW ratios. The five P25 with lengths greater than 20.6 mm assigned to T. veroensis in fact have low AW/PW ratios (87-90, x = 89) which are not different from those of smaller P25. Without morphological evidence, it seems best to retain these individuals in T. veroensis until unquestionable Rancholabrean T. haysii is recovered from Florida, HULBERT: THE GIANT TAP]R, TAPIRUS HA YSUFROM LEISEY SHELL PIT 531 Table 3. Univariate statistics for measurements on cheekteeth of fossil species of Tapirus. The Florida sample of T. h«ysii is a composite from a number of carly Iningtonian localities, although about three- fburths are from a single site. Lcisey Shell Pit lA likewise the T. veroens,3 sample is a composite from Rancholabrean localities from throughout Florida. The Port Kennedy sample of T. haysii is taken from Simpson (1945). Results of Mests between the two samples of T. hays#. and between the two Florida samples are reported to the right ofthe sdatistics for T. veroensis and the Port Kennedy T. hays#. Simpson (1945) did not report the standard deviation for upper ckeektooth measurements; Mests were done assuming equal variances between the two samples of T. haysi, for these teeth. "NS" indicates that the means are not significantly different from one another (p 2 0.05); "*: a significant difference with 0.05>pk 0.01; "**". a signilicant difference with 0.01 >p 2 0.001; and "***", a significant difference withp < 0.001. n x , MIN MAX CV TAPIRUSHAYSH(FLORIDA) L 4 21.73 1.55 20.3 23.3 7.13 Pl PW 4 20.21 1.46 18.9 22.0 7.21 L 7 22.26 1.33 20.6 24.3 6.00 P2 AW 8 24.10 1.26 22.0 23.4 5.22 PW 8 26.47 1.51 23.8 28.0 5.69 L 8 22.84 1.20 21.6 24.5 5.25 P3 AW 8 27.19 1.20 25.7 29.0 4.40 PW 9 26.94 1.37 25.4 29.3 5.07 L 5 23.39 0.96 22.6 25.0 4.12 P4 AW 5 29.27 1.73 28.0 31.9 5.90 PW 5 28.46 1.68 27.2 30.8 5.89 L 10 24.94 0.76 23.9 25.9 3.06 Ml AW 9 29.14 1.20 26.9 30.7 4.11 PW 10 26.19 1.26 24.5 28.5 4.83 L 9 27.79 0.91 26.3 29.2 3.26 M2 AW 9 32.67 1.42 30.9 34.5 4.34 PW 8 28.98 1.16 27.4 30.5 4.01 L 5 27.94 0.58 27.1 28.5 2.08 M3 AW 3 32.60 0.81 31.5 33.6 2.49 PW 5 26.83 1.36 23.1 28.9 3.06 L 6 27.22 1.27 25.6 28.8 4.68 p2 PW 7 17.71 1.33 15.4 19.2 7.53 L 9 23.71 0.93 22.5 25.2 3.92 p3 AW 9 17.42 0.44 16.8 17.9 2.51 PW 9 19.35 1.09 18.1 21.5 5.63 532 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. Ii, NO. 16 Table 3 continued. n x s MIN MAX CV t df sig L 5 23.59 0.89 22.5 25.0 3.76 p4 AW 5 20.43 0.27 20.0 20.6 1.34 PW 5 21.35 0.72 20.2 21.9 3.37 L 8 25.71 1.61 23.0 27.6 6.25 ml AW 8 21.13 0.99 19.9 22.3 4.70 PW 8 19.61 0.78 18.3 20.7 399 L 9 28.30 0.95 27.0 30.4 3.37 m2 AW 22.17 0.35 21.8 22.7 1.59 PW 9 21.20 0.94 19.5 22.7 4.41 L 11 30.36 1.60 28.7 33.5 5.28 n,3 AW 10 22.21 0.88 21.2 24.2 3.99 PW 11 19.93 1.37 17.3 22.0 6.88 TAPIRUSHAYS#(PORT KENNEDY) L 3 23.80 22.4 24.9 -1.702 5 NS Pl PW 3 20.40 19.6 21.5 -0.167 5 NS L 5 22.60 21.9 24.0 4.435 10 NS P2 AW 4 23.90 25.5 26.5 -2.279 10 * PW 4 27.60 27.4 27.9 -1.198 10 NS L 5 23.70 22.7 24.3 4.240 11 NS P3 AW 4 28.60 27.0 29.5 -1.876 10 NS PW 4 28.10 26.1 29.0 -1.379 11 NS L 4 23.30 24.1 26.4 -2.916 7 * P4 AW 4 30.60 29.9 31.8 -1.134 7 NS PW 4 29.30 28.4 30.1 -0.738 7 NS L 5 26.10 25.8 26.4 -2.729 13 * Ml AW 5 30.30 28.9 31.1 -1.709 12 NS PW 5 27.30 25.8 27.9 -1.570 13 NS L 5 28.10 27.3 29.7 4.609 12 NS M2 AW 4 32.60 31.3 34.9 0.085 11 NS PW 4 29.70 28.0 31.5 4.992 10 NS L 5 28.20 26.8 29.2 -0.709 8 NS M3 AW 5 32.30 31.0 34.1 0.585 8 NS PW 5 27.80 26.5 29.0 -1.128 8 NS L 6 25.90 0.70 24.7 27.0 2.70 2.218 10 NS p2 PW 6 16.37 0.67 15.5 17.6 4.04 1.890 11 NS HULBERT: THE GIANT TAPm, TAPIRUS HAYSH FROM LEISEY SHELL Prr 533 Table 3 continued. n x , MIN MAX CV f df sig L 9 24.22 0.67 23.0 25.1 2.77 -1.342 16 NS p3 AW 9 17.01 0.73 16.1 18.0 4.29 1.446 16 NS PW 9 19.19 0.68 17.8 20.2 3.54 0.382 16 NS L 6 24.60 0.26 24.1 24.9 1.06 -2.688 9 * p4 AW 8 20.41 1.13 18.3 21.7 5.54 0.031 11 NS PW 8 21.28 1.11 19.5 22.8 5.22 0.125 11 NS L 6 25.33 1.10 23.4 27.0 4.31 0.232 12 NS ml AW 7 20.83 0.97 19.8 22.9 4.66 0.636 13 NS PW 7 19.56 0.58 18.7 20.8 2.97 0.149 13 NS L 10 28.91 1.06 27.4 30.8 3.67 -1.318 17 NS n12 AW 9 21.56 1.09 20.5 24.1 5.06 1.518 15 NS PW 10 20.17 1.26 18.3 22.8 6.25 2.005 17 NS L 6 30.78 0.58 30.1 31.5 1.88 -0.606 15 NS n13 AW 6 22.70 0.69 21.7 23.8 3.04 -1.168 14 NS PW 5 19.84 0.31 19.2 20.5 2.57 0.141 14 NS TAPJRUS EROENSIS (FLORIDA) L 17 18.75 1.33 16.4 20.6 7.12 3.920 19 *** Pl PW 17 16.40 1.73 13.7 18.8 10.67 4.015 19 *** L 15 19.83 1.02 18.4 21.7 5.14 4.719 20 *** n AW 15 21.22 0.91 19.5 23.1 4.27 6.352 21 *** PW 15 23.59 1.02 21.9 25.8 4.31 5.458 21 *** L 14 20.47 0.94 19.4 22.6 4.60 5.140 20 *** n AW 14 24.88 1.01 22.7 26.5 4.05 4.832 20 *** PW 14 24.90 1.07 22.8 26.7 4.32 4.007 21 *** L 13 21.37 1.12 19.6 23.0 5.25 3.545 16 ** P4 AW 13 26.83 1.26 24.1 28.7 4.71 3.325 16 ** PW 13 26.46 1.04 24.7 28.2 3.93 3.093 16 ** L 14 22.70 1.45 20.4 25.5 6.39 4.432 22 *** Ml AW 13 26.24 1.32 24.6 28.5 5.04 5.239 20 *** PW 14 23.91 1.14 22.4 26.1 4.77 4.625 22 *** L 18 25.14 1.21 23.1 27.5 4.83 5.737 25 *** M2 AW 17 29.07 1.31 26.9 31.2 4.49 6.496 24 *** PW 18 23.85 1.08 24.2 28.3 4.16 6.673 24 *** 534 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. 11, NO. 16 Table 3 continued. n x : MIN MAX CF f df sig L 14 24.84 1.03 22.7 26.3 4.15 6.304 17 *** M3 AW 14 28.89 1.35 26.7 31.5 4.68 5.720 17 *** PW 14 24.39 1.00 22.5 26.2 4.10 4.284 17 *** L 9 23.86 0.91 22.7 25.1 3.82 5.968 13 *** p2 PW 9 14.73 1.34 12.6 16.7 9.10 4.416 14 *** L 8 21.61 1.53 20.4 24.6 7.07 3.471 15 ** p3 AW 8 15.81 0.91 14.8 16.8 5.75 4.730 15 *** PW 8 17.98 1.29 15.9 19.8 7.18 2.379 15 * L 11 22.20 0.94 20.8 23.7 4.25 2.768 14 * p4 AW 11 19.18 1.09 17.6 21.0 5.68 2.485 14 * PW 11 20.29 1.35 18.0 22.9 6.64 1.642 14 NS L 15 22.88 1.23 21.0 25.8 5.35 4.735 21 *** ml AW 14 18.99 1.17 17.3 20.8 6.16 4.392 20 *** PW 15 17.69 1.22 15.9 20.6 6.89 4.024 21 *** L 14 25.42 1.09 23.7 27.6 4.27 6.482 21 *** ml AW 13 20.54 1.19 18.8 22.3 5.80 3.752 19 ** PW 13 19.59 1.22 18.2 21.8 6.21 3.339 20 ** L 14 26.63 1.31 24.4 29.8 4.92 6.415 23 *** m3 AW 14 20.18 0.97 18.5 21.5 4.79 5.233 22 *** PW 14 18.44 1.12 16.7 20.0 6.10 2.996 23 ** The only qualitative characters commonly attributed to T. haysii dentitions that differentiate it from T. veroensis are the relatively molarized Pl and P2. As discussed above, the latter can be analyzed by comparing the AW relative to the PW, and also by the development of the protoloph on the P2. On the Pl, molarization is judged by relative width, strength of the inner cusp, development of a transverse loph, and presence of an anterolingual cingulum (Simpson 1945; Hershkovitz 1954). In all of these characters, the Leisey lA sample is intermediate between lhat observed in the Port Kennedy sample of T. haysii, and that in T. veroensis or T. tenestris (Figs. 3,4C). As in all samples of Tapirus, there is considerable intraspecific variation in these characters, which should temper their widespread use as systematic indicators. For example, Simpson (1945) reported a range for the ratio of P2AW to P2PW of 92 to 95 (n = 4) for T. haysii, and 87 to 92 (n = 7) for T. veroensis. In the Leisey lA sample it ranges from 90 to 96 (n = 6), but in four of the six it ranges between 90 and 92, below the OR of the Port HULBERT: THE GIANT TAPIR, TAPIRUS H,1 YSH FROM LEISEY SHELL Prr 535 A C Figure 3. Occlusal views of upper cheekteeth of Tapirus hbys,i from Leisey Shell Pit 14 Hillsborough County, Florida. (A) UF 84190, left Pl-M3,(B) UF 89539, left P2-M3; (C) UF 84190, detail of left Pl- P2. Length of scale bar for A and Bis 50 mm; for C it is 20 mm. BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. n, NO. 16 536 A B - Figure 4. Occlusal views of upper cheekteeth of Tapirus haysii from Ltisey Shell Pit LA, Hillsborough County, Florida. (A) UF 82783, left DPI-DP3; (B) UF 87231, right Pl-P2. Scale bar is 20 mm Kennedy population. This difference results from a significantly shorter P2AW in the Leisey lA sample, while the P2L and P2PW are not significantly different from the Port Kennedy population (Table 3; Figs. 3,413). Similarly, the protoloph of the P2 and the transverse loph of the P 1 in the Leisey sample vary between the relatively weakly developed states observed in T. veroensis, T. terrestris and T. bairdii, and the well developed states of Port Kennedy T. haysii (Figs. 34; Simpson 1945, fig. 10A). While Simpson (1945) and subsequent authors have noted this intergradation caused by intraspecific variation (e.g. Hershkovitz 1954 regarding the Pl of T. pinchaque), Simpson especially appears to have underemphasized the extent of the morphological overlap in these characters. The relative "primitiveness" of the modal states of the Leisey T. haysii may be due to their greater antiquity (by about 0.5-1.0 myr) than the Port Kennedy population. The lower cheekteeth in the Leisey lA sample exhibit a single morphological character not observed in T. veroensis, the presence of well developed metastylids HULBERT: THE GIANTTAPIR, TAPIRUSHAKS# FROMLEISEY SHELLPrr 537 on the dp2-dp4 (Fig. 5,6). In T. veroensis, metastylids are very weakly developed or absent on the deciduous premolars (n = 13). Weak metastylids are also the observed character state in the referred Love Site and McGehee Farm populations of T. simpsoni (Fig. 6C, 6D, 6F). As the sample of dps from Leisey lA is quite small and represents only two individuals, the consistency of this character in uncertain. The metastyli(is are relatively weak on the dp2 and dp3 ofF:AM 37411 (McLeod), suggesting that this character too is subject to intraspecific variation. Excluding this character, size, and slightly more molarized Pls and P25, the dentitions of Florida T. haysii are practically indistinguishable from those of T. veroensis and Nestropical Tapirus. Postcranial elements of a large tapir were also recovered from Leisey lA (Table 4). There is no indication from their statistical distribution that more than one species is represented. Relative to cranial dimensions and the proportions of the postcranial elements of other species (especially T. indicus), the limbs of T. haysii seem relatively gracile. The limbs are characteristicaly tapirid, with no peculiar features. The skeletal element with the highest relative representation at Leisey lA is the astragalus. The minimum number of individuals (MNI) of T. hqysii recovered at Lcisey lA is six, based on right astragali. A greater MNI of 9 is obtained from maxillae, if it is assumed that unmatched right and left maxillae represent different individuals. A limited number of postcranial elements of T. haysii are known from other localities, most notably Haile 164 from which no cranial material was recovered. The size of the elements from Haile 16A supports their identification as T. hqysii (Table 4), as does its Imingtonian age. Discussion.- Based on tooth size (which forms the only mutual basis for comparison), there is little doubt that the Florida and Port Kennedy samples of large tapirs represent the same species. The species name T. haysii seems the most appropriate for these populations (Ray and Sanders 1984). Although the holotype p4 from North Carolina falls above the OR of the Florida sample for length and AW (it also is longer than any of the Port Kennedy p45), the size of most of the referred topotypes described by Ray and Sanders (1984) are within the OR of the reRrred samples for all three measured variates. The differences in Pl and P2 morphology between the Florida and Port Kennedy samples, if not purely an artifact of sampling (the sample sizes are still relatively small, especially for the Pl), most likely represent chronoclinal variation. A similar wide degree of intraspecific variation is observed for these characters in other species. The Leisey Shell Pit and other Florida Irvingtonian records of Tapirus confirm the existence of a tapir significantly larger on average than the typical Rancholabrean T. veroensis. They do little in the way of demonstrating significant morphological differences between T. veroensis and T. haysii, as all cranial features now known for T. haysii resemble those of T. veroensis. Arguably, with its relatively primitive upper premolars, the Leisey sample blurs what little 538 BULLETIN FLORIDA MUSEUM NATURAL HISrORY VOL. 37, Fr. II, NO. 16 Table 4. M 1 elements of Tapirus hays,7 from Florida. See Referred Specimens li,ung for exact 1-lities. M..,rements after r (1986) HUMERUS UF 87229 MIDSHAFT TRW 31.4 MIDSHAFr A-P L 40.3 DIST ART W 62.4 DIST MEDIAL A-P L 68.0 RADIUS UF 81068 UF 87228 x s CV ORT LENGTH - 235.0 235.0 MIDSHAFT TRW 28.9 30.1 29.3 0.85 2.88 MIDSHAFT A-P L 24.1 21.3 22.7 1.98 8.72 PROX ART W 58.6 62.9 60.8 3.04 5.01 PROX ART A-P L - 33.9 33.9 DIST ART W 66.1 66.4 66.3 0.21 0.32 DIST A-P L 35.7 33.2 34.5 1.77 5.13 METACARPAL II UF 83389 UF 87958 UF 87959 UF 89535 x 5 CV GRT LENGTH 114.1 109.5 109.3 110.0 110.7 2.27 2.05 MIDSHAFT TRW 19.4 21.5 22.5 - 21.1 1.58 7.49 MIDSHAFT A-P L 11.1 11.2 13.5 - 11.9 1.36 11.38 PROXARTW 22.5 23.1 - 25.6 23.7 1.64 6.93 PROX ART A-P L 17.5 20.5 18.8 - 18.9 1.50 7.95 DIST ART W 21.7 21.0 22.1 - 21.6 0.56 2.58 DIST A-P L 24.4 25.0 27.2 - 25.5 1.47 3.77 METACARPAL In UF 81771 UF 87950 UF 89534 UF 91243 UF 91244 x S CV GRT LENGTH 132.9 128.9 133.8 125.5 - 130.3 3.83 2.94 MIDSHAFT TRW 28.8 30.5 29.7 31.0 28.9 30.0 0.96 3.21 HULBERT: THE GIANT TAPIR, TAPEUS HAYS#FROM LEISEY SHELL Prr 539 Table 4 Continued. METACARPAL III Continued UF 81771 UF 87950 UF 89534 UF 91243 UF 91244 x , CV MIDSHAFr A-P L 13.9 14.2 15.4 16.5 15.8 15.0 1.19 7.94 PROXARTW 34.2 34.8 35.7 35.2 - 35.0 0.63 1.81 PROX ART A-P L 27.0 26.3 - 26.7 - 26.7 0.25 0.94 DIST ART W 31.6 32.5 33.2 33.4 31.6 32.7 0.81 2.49 DIST A-P L 25.4 24.9 24.7 24.8 25.6 24.9 0.31 1.25 METACARPAL IV UF 65990 UF 83390 UF 89537 UF 89538 x icv GRT LENGTH - - 107.0 107.7 107.3 0.49 0.46 MIDSHAFT TRW 20.9 - 19.9 22.7 21.2 1.42 6.70 MIDSHAFT A-P L 12.8 - 12.4 13.1 12.8 0.35 2.75 PROX ART W 22.1 23.2 22.2 - 22.5 0.61 2.70 PROX ART A-P L 24.4 25.1 - 24.6 24.7 0.36 146 DIST ART W - - 20.2 - 20.2 DIST A-P L - - 23.0 - 23.0 METACARPAL V UF 84082 UF 84916 UF 87952 UF 88777 UF 91245 x 1 CV GRT LENGTH 83.7 86.7 88.4 86.5 78.3 86.3 1.95 2.25 MIDSHAFT TRW 15.8 18.3 17.9 18.0 19.2 17.5 1.15 6.55 MIDSHAFT A-P L 10.2 10.1 11.0 10.2 10.6 10.4 0.42 4.04 PROX ART W 12.8 13.5 12.4 13.7 12.6 13.1 0.61 4.62 PROX ART A-P L 25.4 - 21.7 23.3 23.5 23.5 1.86 7.91 DIST ART W 18.4 18.3 19.0 18.3 17.1 18.5 0.34 1.82 DIST A-P L 20.7 20.6 21.1 - 20.3 20.8 0.26 1.27 FEMUR UF 60864 UF 80449 UF 65975 UF 83441 x , CV GRT LENGTH - 342.0 337.0 320.5 333.2 11.25 3.38 MEDIAL LENGTH 330.5 - 316.0 301.0 315.8 14.75 4.67 MIDSHAFr TRW 38.6 34.9 37.9 37.1 37.1 1.60 4.32 MIDSHAFT A-P L 39.9 36.9 39.3 33.6 37.9 2.02 5.33 PROXW 105.2 - 109.8 106.8 107.3 2.34 2.18 A-P DIAM HEAD 46.2 - 46.9 44.8 46.0 1.07 2.33 540 BULLErIN FLORIDA MUSEUM NATURAL HISrORY VOL. 37, Fr. II, NO. 16 Table 4 Continued FEMUR Continued UF 60864 UF 80449 UF 65975 UF 83441 x CV DIST ART W - 82.3 77.3 78.5 79.4 2.61 3.29 DIST MED A-P L 95.0 105.3 96.0 96.2 98.1 4.81 4.90 TIBIA UF 65976 UF 65977 UF 80653 UF 83579 UF 91246 x a CV ORT LENGTH 260.0 259.0 254.0 248.0 - 255.3 5.50 2.15 MIDSHAFr TRW 28.9 29.6 28.7 29.0 30.8 29.1 0.39 1.33 MIDSHAFT A-P L 34.0 33.0 31.4 31.2 34.8 32.4 1.34 4.13 PROXARTW 81.7 84.4 81.4 79.7 - 81.8 1.94 2.38 PROX ART A-P L - - 66.0 67.0 - 66.5 0.71 1.06 DIST ART W 43.4 42.8 41.5 41.4 46.3 42.3 0.98 2.33 DIST A-P L 34.9 34.5 37.4 36.3 36.6 35.8 1.33 3.72 ASTRAGALUS UF 63901 UF 63981 UF 65982 UF 67325 UF 81917 UF 86883 UF 87953 MEDIAL LENGTH 55.2 54.3 54.6 50.5 49.9 50.6 52.4 LATERAL LENGTH 60.7 62.9 62.1 39.7 60.4 60.2 60.3 MEDIAL L TROCHLEA 44.2 43.2 43.5 41.8 41.6 41.4 43.1 ORT WIDTH 50.1 53.9 37.2 50.9 49.8 53.4 51.1 DIST ART W 44.6 47.7 49.4 45.3 44.9 45.1 44.4 DIST ART A-P L 29.9 - - 27.4 30.2 29.4 29.2 ASTRAGALUS Continued UF 87960 UF 91247 UF 91248 x : CF MEI)IAL LENGTH 53.4 51.5 58.8 53.1 2.73 5.14 LATERAL , LENGTH 59.8 60.9 - 60.8 1.07 1.75 MEDIAL L TROCHLEA 42.5 42.8 48.0 43.2 1.90 4.40 GRT WIDTH 52.0 33.1 56.5 53.0 2.64 4.99 DIST ART W 45.1 47.6 48.6 46.3 1.85 4.00 DIST ART A-P L 29.8 27.9 29.8 29.2 1.01 3.47 HULBERT: THE GIANT TAPIR, TAPIRUS HAYSH FROM LEISEY SHELL Prr 541 Table 4 Continued CALCANEUM UF 65983 UF 65984 UF 65986 UF 81235 x CV GRT LENGTH 114.3 - - 108.9 111.6 3.82 3.42 GRT TRW 53.3 49.2 53.7 51.9 32.0 2.04 3.91 GRT A-P L 46.1 42.6 43.3 45.9 44.5 1.79 4.02 METATARSAL II UF 65988 UF 65989 UF 82374 UF 81311 UF 89536 x 5 CV ORT LENGTH 113.5 109.9 112.3 110.5 114.9 111.5 1.65 1.48 MIDSHAFT TRW 22.2 19.0 20.9 18.9 21.2 20.3 1.59 7.87 MIDSHAFT A-P L 14.5 13.4 14.1 13.5 14.3 13.9 0.32 3.74 PROXARTW 16.4 15.8 17.6 19.0 17.1 17.2 1.41 8.22 PROX ART A-P L 14.8 - 17.9 16.2 18.5 16.3 1.55 9.52 DIST ART W 22.8 19.3 20.7 20.8 21.3 20.9 1.44 6.89 DIST A-P L 25.8 26.0 28.0 26.1 28.4 26.5 1.02 3.87 METATARSAL III UF 65987 UF 81703 UF 83578 X 3 CV ORT LENGTH 130.0 130.0 125.8 128.6 2.42 1.89 MmSHAFT TRW 28.2 29.3 26.9 28.1 1.20 4.27 MIDSHAFT A-P L 14.9 14.8 14.0 14.6 0.49 3.39 PROXARTW 32.0 30.3 28.9 30.4 1.55 5.11 PROX ART A-P L 25.2 25.1 25.3 25.2 0.10 0.40 DIST ART W 32.2 30.7 31.1 31.3 0.78 2.48 DIST A-P L 25.1 25.2 24.3 24.9 0.49 1.98 METATARSAL IV UF 46923 UF 63900 UF 84442 UF 86865 UF 142233 x 1CV GRT LENGTH 107.4 111.4 110.1 - 117.6 111.6 4.32 3.87 MIDSHAFT TRW 19.8 18.8 18.4 19.6 22.5 19.8 1.60 8.09 MIDSHAFT A-P L 13.9 14.7 12.7 13.4 15.3 14.0 1.03 7.35 PROXARTW 20.3 19.8 18.9 18.6 20.7 19.7 0.90 4.56 PROX ART A-P L 23.1 - 20.6 22.2 - 22.0 1.27 5.76 DIST ART W 20.8 20.1 20.3 - 21.3 20.6 0.34 2.61 DIST A-P L 27.6 27.6 26.3 - 30.5 28.0 1.78 6.34 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. II, NO. 16 542 A B .. , 44 - ..0 '. 104 4 -- Figure 5. Occlusal views of lower cheekteeth of Tapirus haym from kisey Shell Pit 14 Hillsborough County, Florida. (A) UF 83580, left p2-013, (B) UF 80973, left p2-m3. Scale bar is 20 mm. HULBERT: THE GIANT TAPIR, TAPIRUS HAYSH FROM LEISEY SHELL Prr 543 A B 0173,20 file//0,4 ---T"% /3/<./$A C Figure 6. Occlusal views oflower deciduous cheekteeth of Tapirus from Florida. (A) UF 87230, right dp2- dp3 of T. hayi, Leisey Shell Pit lA (early Irvingtonian); (B) UF 11117, right dp2-dp3 of T. simpsoni, McGehee Farm Site, Aachua County (early Hemphillian), (C) UF 47928, right dp2-di,3 of T. veroensis, Aucilla River 14 Jefferson-Taylor county line (late Rancholabrean); (D) UF 11117, right dp4. (E) UF 87230, right dp4. Scale bar is 10 mm. morphological distinction was thought to exist between the two species. If the two different size populations cannot be shown to have been sympatric, and if the absence of significant morphological differences continues, then the hypothesis that T. haysii and T. veroensis represent a single chronoclinal lineage is difficult to falsify, as was perceived by Ray (1964:65). Late Pleistocene (Rancholabrean) records of large tapirs in eastern North America (e.g. in Kurt6n and Anderson 1980) have not proven to be T. haysii, and instead represent larger than average- sized individuals of T. veroensis. Some late Pleistocene records of unequivocally T. haysii-sized individuals are known from the southwestern United States, e.g. the Lehner Mammoth Site (Lance 1959; Ray and Sanders 1984), but these have been 544 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. U, NO. 16 referred to the western species T. merriami by Jefferson (1989). There rem2ing much to be learned about the enigmatic giant Nonh American tapir. BIOCHRONOLOGY OF FLORIDA TAPIRS Since the publication of the first definite records of T haysii from Florida (Simpson 1945; Ray 1964), much has been learned about the pre-Rancholabrean Pleistocene history of the state (Webb 1974; Morgan and Hulbert this volume). Of the 12 new records of T. haysii reported in this study, the following eight were collected in direct association with a suite of mammalian tan that in combination are now recognized as diagnostic of either the late early or middle Iningtonian in Florida: Leisey 14 Leisey 34 Leisey 3B, Haile 164 Haile 214 McLeod Limerock Mine, Crystal River Power Plant, and Pool Branch. These mammals indude Sigmodon libitinus, Canis edwardii, Smilodon gracilis, Mammuthus hayi, and Pachyannatherium leiseyi (Webb et al. 1989; Morgan and Hulbert this volume). They typically lack Blancan holdovers such as Chasmaporthetes, Sigmodon medius, and Capromeryx arizonensis, as well as tan charactedstic of the late Irvingtonian such as Didelphis and Smilodon populator. All but one of the new records of T. h«ysii, plus all previous records of the species from Florida, are late early Irvingtonian (early Pleistocene). The youngest record of T. haysii from Florida is McLeod, middle Irvingtonian, ca. 0.6 Ma (Morgan and Hulbert this volume). This distribution confirms and amplifies previous suggestions that T. haysii is especially characteristic of (or limited to?) the Blancan and Irvingtonian (Hibbard and Dalquest 1966: Ray and Sanders 1984). Present evidence suggests that T. h«ysii was not present in Florida during the late Blancan or earliest Irvingtonian (ca. 2.5 to 1.6 Ma). Faunas of that interval either lack Tapirus altogether (e.g. M,casphalt [= APAC] Shell Pit, St. Petersburg Times Site; Morgan and Ridgway 1987), or contain a relatively small, presently undescribed species of Tapirus (e.g. Inglis 14 De Soto Shell Pit, Haile 7C, Haile 15A). As T. haysii is known from the late Blancan of Texas and Colorado (Strain 1966; Hager 1974; Ray and Sanders 1984) and T. merriami the Blancan of California (Jefferson 1989), it seems most likely that three species of Tapirus were present but not sympatric in North America at th:,t time. The late Irvingtonian in Florida is presently represented by a single fauna Coleman 2A. Two specimens, a distal tibia (UF 125414) and an MC II (UF 125415), record Topirus at Coleman. They represent Tapirus veroensis and not T. haysii on the basis of size. This is the first report of a tapir from the Coleman 2A 1.f., and theoldest record of T. veroensis in Florida. The number of known Rancholabrean local faunas in Florida is quite large, probably numbering over a hundred. Tapirus is a common constituent of the Florida late Pleistocene megafauna. All tapir specimens directly associated with HULBERT: THE GIANT TAPm, TAPIRUS 84YS# FROM LEISEY SHELL Prr 545 Rancholabrean mammals in Florida are T. veroensis. The biochronologic range of T. veroensis within the Rancholabrean in Florida is difficult to establish because mnny faunas have not been or c=nnot be more precisely dated. Early Rancholabrean records of T. veroensis include Williston (Holman 1959:18) and Bradenton 51st Street (UF 2361, 3275, 4762). T. veroensis persisted into the latest Pleistocene in Florida based on records associated with Pateoindian artifacts (Aucilla River). PHYLOGENY OF NEW WORLD TAPIRUS The earliest records of Tapirus in North America are late middle Miocene (early Clarendonian, ca. 11 Ma) from the Ash Hollow Formation of Nebraska (Schultz et al. 1975; Yarnell 1980). The genus first appeared in the Oligocene of Europe (Radinsky 1965; Guerin and Eisenmann 1982), so its first occurrence in North America reflects an immigration event, presumably from Asia. Schultz et al. (1975) described two Miocene species from Nebraska, ?Tapirus johnsoni (Clarendonian) and ?Tapirus simpsoni (early Hemphillian), but only provisionally assigned them to the extant genus. On the basis of more complete cranial material, Yamell (1980) demonstrated that both could be referred to Tapirus, and extended the range of T. simpsoni to the late Clarendonian (Love Site). The only early Pliocene sample of Tapirus is derived from the Palmetto Fauna of central Florida and has yet to be studied. Tapirus is first recorded in South America during the Uquian (late Pliocene), when it was a member of the first major exchange of terrestrial vertebrates between North and South America following the formation of the Panamanian Isthmus (Marshall et al. 1984). Tapirus n. sp. (Florida), T. memami (California), and T. haysii (Texas, Colorado) are known from similar- aged late Blancan faunas in North America (Strain 1966; Hager 1974; Jefferson 1989; Hulbert in prep.). The phylogenetic relationships among extant Neotropical Tapirus and fossil North American species have long remained enigmatic (Simpson 1945; Hershkovitz 1954). In order to elucidate these relationships, 26 cranial and dental characters were selected for cladistic analysis (Table 5). Most of these characters have been widely used in the literature to describe fossil and Recent Tapirus. Character states for each of the seven species under study (T. johnsoni, T. simpsoni, T. haysii, T. veroensis, T. pinchaque, T. bairdii, and T. terrestris) were determined by observation of pertinent skeletal material from the USNM, F:AM, and UF collections. T. johnsoni, the oldest species, was used as the outgroup to determine character state polarities. PAUP analysis of the data (Table 6) produces two equally parsimonious cladograms (Fig. 7). Even excluding characters with single autapomorphies, each cladogram has a high overall consistency index of 0.75. 546 BULLEr[N FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. Il, NO. 16 Table 5. Characters and character states used in the phylogenetic analysis ofNew World Tapirus. 1. Height of sagittal crest; 0 =low, 1 -high. 2. Adult width of sagittal crest; 0 = narrow. 1 - broad. 3. Dorsal surface offrontals, 0 - relatively broad, 1 = relatively narrow. 4. Frontal inflation; 0 = none or slight, 1 = moderate to very inflated. 5. Interparietal; 0 = fused relatively early in ontogeny. 1 - fused late in ontogeny. 6. Antcromedial process offrontal; 0 = extends between nasals. 1 = does nat extend between nasals. 7. Nasal leng#,4 0 - long, 1 = short 8. Depth ofspiral groove on nasal; 0 = shallow, 1 = deep. 9. Spiral groove on namal; 0 = extensive, close to midline ofskull. 1 = not extensive or close to midline. 10. Depth ofascending groove for nasal dive,liculum; 0 - shallow, 1 = deep. 11. Lamhdoidal crests; 0= flair both posteriorly and laterally, 1 = directed posteriorly with little or no outward 8air. 12. Dorsal contour ofskull; 0 = relatively straight 1 = nasals notably slept down from frontals. 13. Posteriorprocessofpremaxilla; 0 =long. terminates inacute point; 1 -shon,bluntly terminated. 14. Anteromedial process of maxilla; 0 =very slender. little ornolateralexposure, 1 -slender, moderate lateral eigosure, 2 = broad, well exposed latcrally. 15. Premaxillary-maxillary suture; 0 =descends anterior to canine in lateral view, 1 = descends to middle ofcanine. 16. Dorsal maxillary flange; 0 = absent, 1 = present 17. Relative width ofparoccipilal process, 0 = narrow and slender, 1 = broad and massive. 18. Anterior lacrimal process(es); 0 - nk-*or weak, 1 =welldeveloped. 19. Lacrimal depression; O =absent, l= present 20. Relative diaste,na length; 0 = short, 1 = moderate, 2 = long 21. Pl shape; 0 -sho,1,broad, 1 =elongate 22. Metastylids on 42*4; O =weak, often absent on dp) and dp#, 1 - well developed on all three deciduous premolan 23. Pl molarization; 0 = poor Oow AW/PW ratio). 1 = moderate. 2 = well molarized. 24. Relative mandibular ramus depth; 0 = slender, 1 = massive. 25. Crown height ofcheekteeth; 0 = short-crow·ned teeth, 1 - higher crowned teeth. 26. Cheektooth width; 0 = teeth relatively narrow, 1 = teeth broad Both cladograms in Figure 7 hypothesize that T. haysii and T. veroensis share a closer common ancestor than either does with any of the Neotropical species. Similarly, the three extant species form a monophyletic group united by five synapomorphies (node 4, Fig. 7). T. terrestris and T. bairdii are closest sister tan. The difference between the two cladograms is in the position of the late Miocene species T. simpsoni. In one, T. simpsoni is the sister taxon to the five Pleistocene and Recent species (Fig. 7A). The latter group (node 2, Fig. 7A) is hypothesized to share two derived character states in the dentition, wider cheekteeth and a more molarized P2 (the latter is found in the primitive state in T. pinchaque; this is regarded as a reversal in Fig. 7A). Figure 7A more closely corresponds with the HULBERT: THE GIANT TAPIR, TAPRUS HA YS# FROM LEISEY SHELL Prr 547 Table 6. Character states for seven New World species of Tapirus. Character numbers (1-26) and codes for states refer to those listed in Table 5. A "?" denotes an unknown or missing value for a character state. Character 1 2 3 4 3 6 7 8 9 10 11 12 13 T. johnsoni 0000 ? 00000000 T. aimpsont 00 ?? 000 ? 1 ? 000 T. haysii 00 ?? 1 ????? 0 ? 0 T.veroen:13 0 0 0 1 1 0 ? 0 0 0 0 0 0 T.pinchaque 0 0 0 1 0 1 0 0 1 1000 T. ferrestris 1010001011110 T. bairdil 0110000111111 Character 14 15 16 17 18 19 20 21 22 23 24 25 26 T.johnsoni ?? 70000000000 T. simpsoni 10017 ? 0000010 T. haysli 1001??0012111 T. veroensis 1000010001111 T. pmchaque 00001011?0011 T. terrestris 0000101001011 T. bairdii 2010002001011 known fossil record of the group, and with their biogeographic relationships. It suggests that the closest ancestor (node 2) shared by North and South American Tapirus was of early Pliocene age, unfortunately a time when the fossil record is very poor. The alternate phylogenetic hypothesis (Fig. 7B) has T. simpsoni as the sister taxon to the three Neotropical species only, and T. veroensis + T. haysii is the sister group of these four species. Only a single known apomorphy unites T. simpsoni and the exlant species at node 3, a reduction in the development of the nasal diverticulum scroll. This arrangement requires no reversal to attain a primitive P2 in T. pinchaque, but rather parallel development of a molarized P2 in the groups T. terrestris + T. bairdii and T. haysii + T. veroensis. This arrangement implies the existence of an as yet unknown common ancestor of T. veroensis and T. haysii in the late Miocene and early Pliocene. Which of the two competing hypotheses more closely reflects the true phylogenetic history of the group is difficult to determine at present, but each can be tested with a more complete fossil record, or the inclusion of more taxa in the analysis. The states of two important cranial characters, the development of an 548 BULLETIN FLORIDA MUSEUM NATURAL HMTORY VOL 37, Pr. II, NO. 16 40 \5 'd 06 61> 7 8 9 10 11 5 4 2 1 4/ 9~$9 ~~0.-1106.t . f. 97 8 10 11 2 5 4 3 1 Figure 7. Two equally most parsimonious cladograms produced by the PAUP program for seven New World species of Tapinis. T. johnsoni was used as an outgroup in the analysis. Numbers refer to nodes mentioned in text HULBERT: THE GIANT TAPIR, TAPIRUS HAYSH FROM LEISEY SHELL Prr 549 anterior lacrimal process and the depth of the ascending diverticular channel, are presently unknown in T. simpsoni. If either were shown to be synapomorphic with the derived sinte observed in the Neotropical tapirs, then the cladogram in Fig, 78 would become more parsimonious than the other. This analysis is obviously very tentative, as it includes neither any South American fossil species (which are very poorly known), nor Asiatic species. It assumes that T. indicus is more closely related to other Asiatic and European species, and that the few similarities it shares with Neotropical Tapirus represent homoplasy. Only a comprehensive study of all fossil and Recent Tapirus could substantiate that assumption and possibly eliminate one (or both) of the two alternative phylogenies proposed here. CONCLUSIONS Three Plio-Pleistocene tapirs are known from Florida. A large species, referred to Tapirus haysii Leidy, is known from at least 14 late early to middle Irvingtonian (ca. 1.5-0.6 Ma) localities. The sample from Leisey Shell Pit lA is the largest and most complete of these, and includes the first described cranial elements of the species. It demonstrates that T. hcfys, i shares many cranial character states with the late Pleistocene T. veroensis, although a majority of these are interpreted to be plesiomorphic for New World Tapirus. The Pl and P2 of the Leisey sample are, on average, less "molarized" than those of the best known population of T. haysii, from the Port Kennedy site, Pennsylvania. The difference is interpreted as the result of chronoclinal variation, and the two samples are otherwise quite similar. All previous records of T. haysii from the Rancholabrean ofFlorida are either instances of misinterpretation of the age of the specimen (e. g. Ray 1964), or are relatively large individuals of T. veroensis. T. haysii and T. veroensis are very closely related, and a hypothesis that they represent a single, polytypic species is not easily falsified, as their chronologic and geographic distributions never seem to intersect (at least with identifiable specimens). T. h®sii has at least two apomorphies that imply they are distinct species, well developed metastylids on deciduous premolars, and more "molarized" anterior upper premolars. There is, however, intraspecific variation in both of these, and only small sample sizes are known for the former. The late Pliocene species of Tapirus from Florida is slightly smaller than average-sized individuals of T. veroensis, but falls within the observed ranges of the latter for dental dimensions. It is thus much smaller than the giant Tapirus haysii, and apparently represents a new species. It occurs only in faunas that are older than those producing Tapirus haysii (Inglis 1 4 Haile 7C, Haile 15A; late Blancan-very early Irvingtonian), so there is no evidence the two were sympatric. The absence of T. haysii from the earliest Irvingtonian and late Blancan in Florida is interesting, as it is known from similar-aged sites in Texas and Colorado. T. 550 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, Fr. II, NO. 16 veroensis ranges from the late Irvingtonian (Coleman 2A) to the latest Rancholabrean. LITERATURE CITED Colbert E. H., and D. A Hooijer. 1953. Pleist=ne mammals from the limestone fissures of Szechwan, China. Bull. Amer. Mus. Nat. Hist 102(1):1-134. .V. 1986. Comparative osteology of modem andfossil horsca, half-asses,andasses. Pp. 67-116 in R.H. Meadow and H.P. Uerpman, eds Equids in the ancient wodd. Dr. I.udwig Reichert Verlag, Wiesba(len. Guerin, C., and V. r . 1982. Repartition stratigraphique des tapirs (Mammalia, Perissodactyla) dam le Neogene et le Quaternaire d'Europe occidentale. P. 298 in 90 Reunion des Sciences de la Terre. Paris Hager, M. W. 1974. Late Pliocene and Pleistocene history of the Donnelly Ranch vertebrate site, southeastern Colorado. Univ. Wyoming Contrib. Gmt. Spec. Pap. 2:142. Hershkovitz, P. 1954. Mammals of northern Columbia, preliminazy report no. 7: Tapirs (genus Tapirus), with a systematic review of American specia Proc. U. S. Nat. Mus. 103:465-496. Hibbard, C W., and W.W. Dalquest 1966. Fossils from the Seymour Formation of Knox and Baylor counties, Texas, and their bearing on the late Kansan climate ofthat region. Contrib. Mus. Paleon„ Univ. Michigan 21(1):1-66. Holman, J. A 1959. Birds and mammals from the Pleistocene of Williston, Florida. Bull. Florida State Mus„ Biol. Sci. 5(1):1-24. Jefferson, G. T. 1989. Late Cenozoic tapirs (Mammatia: Perissodactyla) of western North America. Contrib. Sci. NaL Hist. Mus. Ins Angeles Co. 406:1-21. Kurth B., and E. Anderson. 1980. Pleistocene mammals of Notth America. Columbia Univ. Press, New York. 442 p Lance, J. F. 1959. Faunal remains Som the L.chner mammoth site. Amer. Antiq. 25:35-39. Lcidy, J. 1852£ [Reference to afossil tooth of a tapir.] Proc. Acad. NaL Sci. Philadelphia 6: 106. . 1852b. [Remarks on TapinuHays#.1 Proc. Acad. Nat. Sci. Philadelphia 6:148. 1855. Notice of some fossil bones discovered by Mr. Francis A Lincke, in the banks of the Ohio River, Indiana. Proc. Acad. NaL Sci. Philadelphia 7:199-201. 1859. Descriptions of vertebrate lossils. Pp 99-122 in F.S. Holmes, ed. Post-Pleiocene fossils of South-Carolina. Russell and Jones, Charleston. I.undelius, E. L, and B. H. Slaughter. 1976. Notes on American Pleistocene tapirs. Pp. 226-243 in C.S. Churcher. ed. Athlon: Essa), in Paleobiology in Honour of 1//is Shano Russell. Royal Ontario Museum, Toronto. Marshall, L G., A Berta, R Homtetter, R. Pascual, O. A Rei; M. Bombin, and A Mones. 1984. Mammals and stratigraphy: geochronology of the continental mammal-bearing Quaternazy of South America. Palaeovertebrata Man Extr. 1984. 76 p. Morgan, G. S.. and B. Ridgway. 1987. Late Pliocene (late Blancan) vertebrates from the St. Petersburg Times Site, Pinellas County, Florida, with a brief review of Florida Blaman faunas. Pap. Florida Paleon. 1:1-22. Radinsky, L B. 1965. Evolution of the tapiroid skeleton from Heprodon to Tap,rus. Bull. Mus. Comp. Zool. 13*3):69-106. Ray, C. E. 1957. A list bibliography, and index ofthe fosail vertebrates of Florida Florida Gmt. Surv. Spec. Pub. 3:1-175. .1964. Tapinis copei in the Pleistocene of Florida. Quart J. Florida Acad. Sci. 27(1):59-66. -, and AE. Sanders. 1984. Pleistocene tapirs in the eastern United States. Pp. 283-315 in H.H. Genoways and M.R. Da,~on, eds. Contributions in Quaternary Vertebrate Paleontology: a Volume in Memorial to John E. Guilday. Carnegie Mus. NaL Hist, Spec. Pub. 8. Schultz, C. B., L D. Martin, and R. G. Corner. 1975. Middle and Late Cenozoic tapirs from Nebraska Bull. Univ. Nebraska State Mus. 10(1): 1-21. HULBERT: THE GIANT TAPIR, TAPIRUS 84 YSU FROM LEISEY SHELL PIT 551 Sellards, E. H. 1918. The skull of a Pleistocene tapir including description of a new species and a note on the associated fauna and flora. Aim Rept. Florida Gcol. Surv. 8:127-130. Simpson. 0.0.1945. Notes on Pleisto©ene and Recent tapirs. Bull. Amer. Mus. NaL Hist 8«2):33-82. Strain, W. S. 1966. Blancan mammalian fauns and Pleistocene formations, Hudspeth County. Texas. Bull. Texas Ment. Mus. 10:1-55. Swofford, D. L 1985. PAUP, phylogentic anal#gis using parsimony, version 2.4. Illinois Nat. Hist. Surv., Cl 'v . 81 P. Yarnell, K. 11 1980. Srd..„dics of late Miocene Tapiridae (Man,matia, Perissodactyla) from Florida and Nebrs,nic. M.S. thaqi< Univ. Flofida, Gainesville. Webb, S. D. 1974. Chronology of Florida Pleistoome mammals. Pp. 5-31 in S. D. Webb, ed. Pleisto©ene mammals of Florida. Univ. Presses Florida, Gainesville. . G. S. Morgan, R. C. Hulbert. Jr., D. S. Jones, B. J. MacFadden, and P. A Mueller. 1989. Geochronology of a rich early Pleistocene vertebrate fauna, 1£isey Shell Pit Tampa Bay, Florida. QuaL Res. 32(1):96-110.