AN (ORELLAN) SIRENIAN FROM THE BUCATUNNA FORMATION OF ALABAMA

KENNETH WHETSTONE and LARRY D. MARTIN DEPARTMENT OF SYSTEMATICS AND ECOLOGY AND MUSEUM OF NATU UNIVERSITY OF KANSAS, LAWRENCE, KANSAS RAL HISTORY

INTRODUCTION were recovered from the top of the Buca­ Tertiary sirenians from the New World tunna Formation at its upper contact with have been recently reviewed by Reinhart the Chickasawhay Limestone. (1976) and Savage (1976). The earliest New The Bucatunna clay in Alabama and Mis­ World sirenians are known from the presence sissippi is predominantly composed of dark of pachyostose ribs in strata of the carbonaceous clays and. silts. l t is locally southeastern United States and from an sandy and bentonitic (Murray, 1961) and is Eocene skull and mandible from Jamaica. about 2 5 feet ( 7.62 m) thick at the Lone Eocene sirenians have been reported from Star Quarry (Glawe, 1967). The Bucatunna the Gosport Formation (Bridgerian) of is middle Oligocene (Orellan) in age and cor­ Alabama (Siler, 1964; Russell, 1955) and the relates with rocks of age in Europe Avon Park Limestone (Uintan) of Florida (Cooke et al., 1943). Although the Bucatun­ (Reinhart, 1976 ). The middle Eocene Pro­ na is sometimes grouped with the underlying rastomus sirenoides is based on a single Byram Marl and Glendon Limestone as the specimen from Jamaica. Several occurrences Byram Formation, we have followed Murray ( 1961) in considering these units to be sepa­ of Oligocene sirenians have previously been reported in the New World. Two pach y­ rate formations. Fossil vertebrates have not previously ostosc ribs were described from Chiapas, been described from Oligocene strata of Mexico, by Mullerried ( 19 32) and Maldo­ Alabama, although Cooke ( 1926, p. 2 91 J nado ( 19 53). A skull from the middle Oligo­ noted the occurrence of otoliths in the cene of Puerto Rico was described as Cari­ Byram Marl. The sircnian specimens from bosiren turneri by Reinhart (1 959), and the Lone Star Quarry were associated with Trelles (1936) noted the occurrence of siren­ crocodilian scutcs (GSA-V -108 5), spines and ians in the Oligocene deposits of Cuba. Arata dental plates of myliobatid rdys (GSA V and Jackson (1965) reported sirenians from 1083, I 084), sharks teeth of the genera the Oligocene Red Bluff and Chickasawhay Odontaspis, Galeocerdo and lo;urus (GSA­ Formations in Mississippi. The age of the V-1089), and vertebrae and otoliths of Mexican records is uncertain, and Savage several teleost fishes. These specimens arc (1976) refers to these as "Eocene/Oligo­ catalogued into the collections of the Geo­ cene". The reference by Reinhart (1976, p. logical Survey of Alabama (GSA) and the 236) to an Oligocene Halitherium from University of Kansas (KUVP). Florida is an error; elsewhere in this descrip­ tion these specimens are referred to the Mio­ cene. The specimens described herein, there­ DESCRIPTION fore, represent the first Oligocene sirenians The specimens reported here include part reported from Alabama. of a left ramus (KUVP 43130), part of d In 197 3 a field party from the University neural arch (GSA-V-1086), and several ribs of Alabama led by D. E. Jones collected the ~KUVP 43131, GSA-V-1090). Thev are from first of the sirenian remains from the Lone a small sircnian with very short pachyostosc Star Cement Quarry on the Tombigbee River ribs. The mandible appears to be relatively 2.2 miles (3.5 km) northeast of St. Stephens, short and moderately deep. The mandibular Alabama. Subsequent collections by R. C. canal is large and exits posterior to the last Price and K. N. Whetstone produced several molar just below the anterior edge of· the additional specimens including the figured ascending ramus. The portion of the as­ portion of a right mandible. The specimens cending ram us preserved is straight but is not

161 162 Tulane Studies in Geology and Paleontology Vol. 14

complete enough to rule out the possibility zoic vertebrates from the Gulf Coastal Plain - that it may have inclined forward as in the I: Tulane Stud. Geol., vol. 3, p. 175-177. , Trichecus. The four roots pre­ COOKE, C. W., 1926, The Cenozoic formations in Geology of Alabama (G . I. Adams, C. Butts, L. served on the ramus are probably from two W. Stephenson and W. Cooke, ed.): Geol. Surv. molars. The anterior tooth has two broad, Alabama Special Report 14, p. 251-297. flat roots, while the posterior tooth has a COOKE, C. W., J. A. GARDNER and W. P. flat anterior root and a round posterior root WOODRING, 1943, Correlation of the Ceno­ zoic form at ions of the Atlantic and Gulf Coast­ indicating that this is the last molar. No al Plain and the Caribbean region: Geol. Soc. alveolae occur anterior to the roots pre­ Amer., Bull., vol. 54, p. 1713-1722. served on the fragment, and it would appear G LAWE, L. N., 1967, Geologic section exposed in that only two molars were present. Lone Star Cement Company Quarry in Geology of the Coastal Plain of Alabama (D. E. Jones, Very few systematic inferences can be ed .) . Geol. Soc. Am er., Guidebook, p. J 10-113. made from such fragmentary material. How­ MALDONADO-KOERDELL, MANUEL, 1953, ever, the apparent tooth re duction would Segundo hallazgo de sirenidos fosiles en Mexi­ suggest that it should not be referred to the co : Ciencia (Mexico), vol. 13, p. 146-148 . . Trichechidae. It is also not from a as MULLERRIED, F. K. G., 1932, Primer hallazgo de un sirenido fosil en Ia republica Mexicana: Uni­ it possesses rooted teeth and short massive versidad Instituto Biologica Anales, vol. 3, p. ribs. It might be related to Hy drodamalis as 71-73. this genus is edentulous with ve ry massive MURRAY, G. E. , 1961, Geology of the Atlantic dense ribs (Simpson, 1932). Of the fossil and Gulf Coastal provinces of North America. taxa, does not show the New York, Harper Bro. 692 p. REINHART, R. H., 1959, A review of the tooth reduction that appe ars to be present in and desmostyla: Univ. Calif., Publ. Geol. Sci., the Alabama fos sil. At the present time the vol. 36, 146 p. best assignment that we can make for this REINHART, R. H., 1976, Fossil sirenians and mat erial is , genus undetermined. desmostylids from Florida and elsewhere: Florida State Museum, Bull., vol. 20, p. If the tooth reduction suggested here should 187-300. prove to be real, then a new generic taxon RUSSELL, R. J. (ed.), 1955, Guides to South­ w ould be required. eastern Geology: Geological Soc. Amer., p. 1-592. ACKNOWLEDGEMENTS SAVAGE, R. J. G., 1976, Review of early sirenia: Systematic Zoology, vol. 25, p. 344-351. We thank S. W. Shannon of the Geologi­ SILER, W. S., 1964, Middle Eocene sirenian m cal Survey of Alabama for the loan of Alabama: Jour. Paleontology, vol. 38, p. material under his c;:are. Jessie a Harrison read 1108-1109. SIMPSON, G. G., 1932, Fossil sirenia of Florida an earlier version ~f the manuscript and and the evolution of the sirenia: American D awn Adams prepared the illustrations. Museum Nat. Hist., Bull., vol. 59, p. 419-503. TRELLES, DUELO, L., 1936, Restos fosilizados REFERENCES CITED de un manati extinguido del periodo Oligocene Inferior: Sociedad Cubana de Historia Natural ARATA, A. A. , and C. G . JACKSON, 1965, Ceno- 'Felipe Poey' Havana, Mem., vol. 9, p. 269-270.

December 28, 1978

PLATE 1 Oligoce~e s1remans from the Bucatunna Formation of Alabama. 1. right mandible (KUVP 43130) m lateral (la.) , dorsal (lb.), and medial view (lc.); 2. and 3. ribs (KUVP 43131 and GSA V-1090 respectively) . Scale is 2 em. No.4 Oligocene Sirenian from Alabama 163

2em

---·-...

l I \ \ \ \ \ \ \

\ la \ \ I \ I I \ \ ' ...... - .. -

lb

lc

3

PLATE 1