THE PALEOCENE/EOCENE TRANSITIONON TORNILLO FLAT INBIG BEND NATIONAL PARK, TEXAS
Judith A. Schiebout Louisiana State University Museum of Natural Science Louisiana State University Baton Rouge, Louisiana 70803
Reexamination of the Paleocene/Eocene transition in the most complete applicable section in Big Bend, in light of several recent revisions of taxonomy and of early mammalian biostratigraphy, still leaves some taxa apparently out of synchronization with the picture developed further north, suggesting that ecologic and/or geographic factors are playing a role. Some of the global changes at the Paleocene/Eocene boundary being studied elsewhere, such as a warming trend, appear to be recognizable in the Big Bend region, which has yielded the southernmost large Paleocene mammalian faunas on the North American continent.
Deposition of the Chilicotal Group (Aguja and Tornillo Formations) in Big Bend National Park spans both the Cretaceous/Tertiary and Paleocene/Eocene boundaries. It begins with the first significant influxes of terrestrial sediments as the Cretaceous sea retreated, continues through increasing uplift in the region, and ends with the onset of local volcanism in the Chisos Mountains, which lie today at the heart of the park (Schiebout et al., 1988). The contact of the fluvial Black Peaks and Hannold Hill Members of the Tornillo Formation (Figure 1) is the focus of interest for consideration of the Paleocene/Eocene transition. The longest Tornillo Formation section, the one containing the southernmost major late Paleocene mammal-bearing sites of this continent (Ray's Bonebed and Joe's Bonebed), and the type sections of the Black Peaks and Hannold Hill Members, occurs on western Tornillo Flat in the northern part of Big Bend National Park. Upper case letters are used to distinguish upper teeth from lower throughout. Abbreviations for North American Land Mammal Ages are shown on Figure 1.
THE PALEOCENE/EOCENE BOUNDARY
Current strong interest in the Paleocene/Eocene boundary is evinced by the formation of the IGCP 308 Paleocene/Eocene Stratotype Working Group, working under the auspices of UNESCO. Although the details of boundary stratotypes will be worked out in marine sections, the new plethora of data on events at this time plus developments in understanding of mammal paleogeography, make a close reexamination of data on a good section including the boundary, potentially rewarding. The Paleocene/Eocene boundary is currently considered to lie within magnetochron C24R (Rea et al., 1990, p. 118; Berggren, 1993). Tornillo Flat paleomagnetic data (Rapp et al., 1983) place one vertebrate fossil site, the South Wall Site, in C24R. Among the events possibly associated with the boundary according to Rea et al. (1990) are: global warming produced by an increase in atmospheric CO2 related to tectonism, a drier climate in continental interiors (Wolfe, 1978, 1979); a rapid turnover in mammalian faunas (Gingerich, 1989; Rea et al., 1990); and a lessening of latitudinality with warmer climates in higher latitudes, but a sea level change is not seen.
Until the work of the Stratotype Working Group is complete and, as Berggren (1993) comments, the boundary selected and anchored by a "golden spike", for terrestrial sites it suffices to consider the Paleocene/Eocene boundary to lie at or near the Clarkforkian-Wasatchian Land Mammal Age boundary (Butler et al., 1987; Rea et al., 1990; Gunnell et al., 1993). The North American Land Mammal Ages of Wood have recently been redefined in Woodburne et al. (1987) which defined the base of the Wasatchian at the first appearance of Hyracotherium (West, 1987, p. 85). In 1989, Gingerich defined a new zone at the very base of the Wasatchian (Wa0) based on twenty sites in the Bighorn and Clark's Fork basins of northwest Wyoming, in which he considered the environment to be well drained high floodplains with mature soil development, an environment much less frequently sampled than low floodplains. Gingerich named thirteen new species from Wo faunas, for the most part animals smaller than others of their genera, producing a picture of a distinctive zone, nonetheless not tremendously different taxonomically at the generic level from Wa1 (Gingerich, 1989, p. 90). A small Hyracotherium, H. sandrae, characterizes Wa0 (Gingerich, 1989, p. 58-63).
Gingerich (1989, p. 89-90) summarized a climate-driven hypothesis for the development and dispersal of the animals whose appearance marks the Wasatchian. He postulated a late Paleocene development of centers of endemism in equatorial areas (northern South America and central America, northern Africa or south Asia) followed by dispersal of a cosmopolitan fauna including modern orders such as the perissodactyls and artiodactyls northward within continents and between continents, across northern routes with the Paleocene/Eocene boundary warming (Gingerich, 1989, p. 89). Sloan (1969) had postulated a Central American source for animals such as rodents and perissodactyls, and Gingerich (1989, p. 90) postulated an African source. Krause and Maas (1990) summarized biogeographic data on the new mammals that mark the late Paleocene and early Eocene in western North America and also suggested a possible origin in Africa or in the Indian subcontinent. New work in China has uncovered the earliest rodents and a possible stem perissodactyl, Radinskya (McKenna et al., 1989). It is possible that China was a stop on the way from an African source (Gingerich, 1989). The pattern of Hyracotherium first occurrences does not unequivocally support this idea. Ting (1993) has named a new very primitive Chinese ceratomorph from an early Wasatchian site, Orientolophus, and suggested that the Mongolian early Eocene Hyracotherium gabunai may be referable to it. Flynn and Novacek (1984) consider the Punta Prieta vertebrate fauna of Baja California, which contains Hyracotherium seekinsi, to be Wasatchian and consider the fauna as a whole to argue against heterochrony for the Wasatchian of the western United States.
BIG BEND BIOSTRATIGRAPHY
Figure 1 summarizes the North American and Big Bend biostratigraphic issues. Columns one to five are extracted from Archibald (1987), 6 through 10 are Tornillo Flat data, and 11 is extracted from Haq et al. and tied to the time scale in column 1. The bases of anomaly 25 and 24 have been used to select a scale for the Tornillo Flat data, and all the rest of its columns hang from 10.
On Tornillo Flat, the stratigraphically lowest occurring Hyracotherium (a maxilla fragment with left upper M1-M3 from the South Wall Site) occurs only eight meters above the base of the Hannold Hill Member as redefined by Schiebout et al. (1987). The Paleocene/Eocene transition falls in a rather small stratigraphic interval. Only 52 and 78 meters, respectively, lower in the well-exposed section lie Joe' Bonebed and Ray's Bonebed and the South Wall Site, fossils have remained very rare. Unfortunately it is in this interval that questions regarding the nature of the transition, particularly those raised by much new data on global changes at the boundary, will have to be answered. Also, no further Hyracotherium material has been forthcoming from the South Wall in over 20 years.
The differences between the Black Peaks and the Hannold Hill Members reflect environmental changes closely coincident with the Paleocene/Eocene transition in Big Bend. The Black Peaks Member is characterized by lenticular sandstones laid down by highly meandering rivers and prominent color banding of overbank mudstones, including readily recognized black layers representing wet floodplain conditions, which have been used as stratigraphic markers on Tornillo Flat (Schiebout, 1974; Schiebout et al., 1988). The Hannold Hill Member is characterized by (Schiebout et al., 1988): a higher proportion of lenticular sandstone bodies, a lower sandstone to mudstone ratio, lack of the distinctive black beds, and an increase of red coloration in overbank mudstones in comparison to gray mudstones. Given that the thickest of the black beds lies at the top of the Black Peaks Member, the lower Hannold Hill may well represent a considerable change in circumstances, probably combining the warming and drying seen worldwide with changes in base level related to minor sea level changes and/or tectonic uplift. Warming and drying would have resulted in vegetational changes as well as changes in soil regimes and additional erosion in source areas. There is no sea level change exactly related to the black mudstone in question, although earlier ones appear approximately correlated with high stands (Figure 1). The black layer in question appears to be forming early in a transgression (Figure 1) or, if the figure of 55.3 Ma is used for the Paleocene/Eocene (Cf3/Wa0) boundary after Kochet et al. (1992), in a high-stand. It must be remembered that the correlation to the sea level curve through the paleomagnetic and age scales also leaves room for some leeway due to shifts in rates of deposition. Aubry (1993) mentions the widespread formation of lignites and carbon rich sediments at the boundary, and the carbon rich black layer certainly fits this pattern. Geochemical work, including a look at the abundant, soil-formed carbonate nodules, is planned to see if some of the geochemical markers of the boundary recognized elsewhere are confirmable here. These abundant nodules greatly affect fossil preservation, hunting, and preparation, in that they cover and fragment bone.
Three sites in the Black Peaks Member and three in the Hannold Hill Member exposed on Tornillo Flat are of particular interest regarding the Paleocene/Eocene transition: Ray's Bonebed and Annex, Joe's Bonebed, the New Taeniodont Site, the South Wall, the Exhibit, and TT-Jacks (Figure 1). Screen washing at Ray's and Joe's has yielded small mammals including plesiadapoids which place Ray's in Ti3 and Joe's in Ti5 (Gingerich, 1976). The next stratigraphically higher site, New Taeniodont Site, has yielded the only significant vertebrate find to come from the black mudstone layers, a taeniodont, Psittacotherium (Schoch, 1986). It can be considered a very late occurring member of the genus, if not the latest. Schoch (1986) considered extension of Psittacotherium into Ti5 to be questionable, and the stratigraphic situation suggests a Ti5 or a Clarkforkian age for the specimen, as later recommended in Archibald (1987, p. 28). Schoch (1986) speculated that the rarity of Psittacotherium in well known sites suggested that it was an upland animal. The environment of New Taeniodont Site is where one animal died, and cannot be helpful regarding where the bulk of the population lived. The wet lowland environment it represents, however, would have been widespread on the Gulf coastal plain, so if the animal was an upland dweller, it was far afield.
The South Wall Site has yielded two mammals, a mandible fragment of Hyracotherium and an incisor of Coryphodon (Schiebout et al., 1988), both from float on a sandstone overlying the thickest black mudstone layer in the Black Peaks Member. The Exhibit Site and TT-Jacks have both yielded fragmentary Hyracotherium material referred to H. vasacciense and Coryphodon (Wilson, 1967; Hartnell, 1980). The Exhibit Site received its name from an "in place" display of Wasatchian mammals, especially Coryphodon, and was the first "in place" display of fossil mammals in a U.S. national park (J.A. Wilson, personal comm., 1977). Screening at TT-Jacks has yielded isolated teeth of the rodent Paramys. Korth's (1984) revision of early Tertiary rodents included raising a small Paramys, P. taurus, to species rank. A recent study of late Paleocene and early Eocene rodents in the Clark's Fork Basin in Wyoming has involved considerable study of Paramys taxonomy (Ivy, 1990) and description of a new small species of the genus, P. pycnus, from the Wyoming earliest Wasatchian (Sandcouleean and Graybullian). She also discusses and gives measurements for P. taurus, approximately 15% larger, and ranging from the Clarkforkian to the middle Graybullian (Ivy, 1990, p. 34-37). The TT-Jacks Paramys is small (Figure 2), but there is little correspondence of comparable teeth with the small sample of P. pycnus, so the TT-Jacks material, also a very small sample, will not be reassigned at this time.
The South Wall is best considered to lie in zone Wa1, not in recently defined Wa0 bearing H. sandrae, as the Big Bend oldest Hyracotherium. The upper molars are on the upper boundary of size for H. sandrae and have more complete cingula. The specimen was originally referred to H. angustidens (Schiebout, 1974), which now has been synonymized with H. index (Gingerich, 1989, p. 58). H. sandrae may have ranged neither so far south nor so low on the coastal plain.
In addition to the missing Tiffanian and Wasatchian zones (Figure 1), the Clarkforkian Land Mammal Age and its three included zones cannot be recognized in Big Bend. Paramys and Coryphodon should mark the beginning of the Clarkforkian (Archibald, 1987, p. 64). However, each does not appear until the Wasatchian, with the first Paramys, 64 meters higher in the section than the first Coryphodon. Although the New Taeniodont site could be Clarkforkian, it need not be younger than Tiffanian.
Tornillo Flat lay much further from the uplands and nearer the sea than many classic localities such as those in the Bighorn Basin, and thus experienced a considerably lower rate of deposition, producing a condensed section. Sloan (1987, p. 187) has summarized rates of deposition for major Paleocene mammal collecting areas, finding a range from 15 bubnofs (meters per million years) for Big Bend to 568 bubnofs for the Hoback Basin in Wyoming. The Bighorn Basin ranged from 200 to 90 bubnofs from west to east (Sloan, 1987). Sloan's calculation for Big Bend used the interval Ti1 to Cf1. A recalculation that avoids an estimate for Cf1, calculating from Ti3 to Wa1, yields a figure of 21 bubnofs, still very low. Rapp et al. (1983) found rates of sedimentation in the Black Peaks Member exposed on Tornillo Flat to be relatively constant, averaging 2 centimeters per thousand years (20 bubnofs). The thickest black mudstone of the Black Peaks section represents a long period of conditions which could well have been inimical to some of the animals of major biostratigraphic significance, followed by a significant environmental change in the switch to characteristic Hannold Hill lithology. The eight meters above the base of the Hannold Hill at which the Hyracotherium was found in float may represent very little time, as the most likely scenario is that the maxilla fragment eroded from the channel sandstone on which it was found and that the sandstone was much more rapidly deposited than the mudstone. More rapid deposition and more upland floodplain conditions rapidly replacing a wet lowland floodplain could very well have been a situation not congenial for types from all zones of the northern intermontane basins where the Clarkforkian and early Wasatchian were defined.
Another possibility is that some types reached Big Bend, but many have done so at considerably different times. It has been estimated that animals with an Asian source dispersing southwards in North America could have arrived in Big Bend later than they do in Wyoming. Rodents, represented by Paramys, have been suggested as an example of this phenomenon (Schiebout et al., 1988) and Coryphodon may also have been affected. The preservation and recovery of fossils in Big Bend is particularly rare, so stratigraphically important types seen in the north may have been here, but will never be found or will be found with future diligent long term search and increasing numbers of washing sites. Twenty years ago, no Big Bend Puercan mammals had been found, although rocks were exposed that seemed likely candidates to yield them. Puercan faunas have now been recovered from several Big Bend sites (Standhardt, 1986). The late first occurrence of Paramys may reflect only that isolated rodent teeth are tiny, and that no screen washing has been done in the stratigraphic interval between Joe's Bonebed and TT-Jacks.
If the Big Bend Eocene is considered to begin with the arrival of a member of the early Wasatchian (Wa1) cosmopolitan fauna, at the first appearance of Hyracotherium, the Eocene appears to be "on time" paleoclimatically and paleomagnetically with the current global picture. It appears that rodents and Coryphodon arrive later in Big Bend than in the very well known faunas of Wyoming, and a lot of zones are "missing". Continued hunting and screening is in order, but it must be remembered that the Big Bend section is condensed and the odds of preservation and recovery are diminished in comparison to the northern intermontane sites. Significant faunal differences between very local upland and lowland areas of fluvial deposition have been documented by Gingerich (1989) in Wyoming, so it is no surprise that the pattern observed in Big Bend is different. Big Bend was an usually wet floodplain environment in comparison to northern intermontane sites immediately pre-Wasatchian, and the Black Peaks-Hannold Hill lithologic transition represents a significant local environmental change coincident to the global Paleocene/Eocene changes.
Support was provided by the National Science Foundation under grant EAR 8216488, the LSU Museum of Geoscience Associates, the LSU Museum of Natural Science, the LSU Department of Geology and Geophysics, and Joe Schiebout. Work was conducted under Antiquities Act Permits granted to the LSU Museum of Geoscience. Conversations with colleagues, including Jill Hartnell, Earl Manning, Steven Rapp, Robert Sloan, Barbara Standhardt, Suyin Ting, and John A. Wilson, have been important. Ruth Hubert and Suyin Ting are thanked for reading the manuscript and providing helpful comments. Cooperation of members of the U.S. National Park Service is much appreciated.
Archibald, J.D., 1987. First North American Land Mammal Ages of the Cenozoic; in Woodburne, M.O., ed., Cenozoic mammals of North America: Berkeley, University of California Press, p. 24-76.
Aubry, M.P., 1993. Paleocene/Eocene boundary events: correlating the marine and terrestrial records: Journal of Vertebrate Paleontology Abstracts of Papers, v. 13.
Berggren, W.A., 1993. NW European and NE Atlantic Paleocene-Eocene boundary interval bio- and sequence stratigraphy and geochronology: Journal of Vertebrate Paleontology Abstracts of Papers, v. 13, p. 26A.
Butler, R.F., Krause, D.W., and Gingerich, P.D., 1987. Magnetic polarity stratigraphy and biostratigraphy of middle-late Paleocene continental deposits of south-central Montana: Journal of Geology, v. 95, p. 647-657.
Flynn, J.J. and Novacek, M.J., 1984. Early Eocene vertebrates from Baja California: evidence for intracontinental age correlations: Science, v. 224, p. 151-153.
Gingerich, P.D., 1989. New earliest Wasatchian mammalian fauna from the Eocene of northwestern Wyoming: composition and diversity in a rarely sampled high-floodplain assemblage: University of Michigan Papers on Paleontology, no. 28, 97 p.
Gingerich, P.D., 1976. Cranial anatomy and evolution of early Tertiary Plesiadapidae (Mammalia, Primates): University of Michigan Papers in Paleontology, v. 15, p. 1-141.
Gunnell, G.F., Bartels, W.S., and Gingerich, P.D., 1993. Paleocene-Eocene boundary in continental North America: biostratigraphy and geochronology, northern Bighorn Basin, Wyoming: in Lucas, S.G. and Zidek, J., eds., Vertebrate Paleontology in New Mexico: New Mexico Museum of Natural History and Science Bulletin 2, p. 137-144. Haq, B.U., Hardenbol, J., Vail, P.R., Wright, R.C., Stover, L.E., Baum, G., Loutit, T., Gombos, A., Davies, T., Jan Du Chene, R., Pflum, C., Romine, K., and Posamentier, H., 1987. Cenozoic global cycle chart: Version 3.1A.
Hartnell, J.A., 1980. The vertebrate paleontology, depositional environment, and sandstone provenance of early Eocene rocks on Tornillo Flat, Big Bend National Park, Brewster Country, Texas [M.S. Thesis]: Austin, University of Texas, 174 p.
Ivy, L.D., 1990. Systematics of late Paleocene and early Eocene Rodentia (Mammalia) from the Clark's Fork Basin, Wyoming: Contributions from the Museum of Paleontology, University of Michigan, v. 28(2), p. 21-71.
Koch, P.L., Zachos, J.C., and Gingerich, P.D., 1992. Correlation between isotope records in marine and continental carbon reservoirs near the Paleocene/Eocene boundary: Nature, v. 358, p. 319-322.
Korth, W.W., 1984. Earliest Tertiary evolution and radiation of rodents in North America: Bulletin of Carnegie Museum of Natural History, no. 24, 71 p.
Krause, D.L. and Maas, M.C., 1990. The biogeographic origins of late Paleocene - early Eocene mammalian immigrants to the Western Interior of North America; in Bown, T.M. and Rose, K.D., eds., Dawn of the age of mammals in the northern part of the Rocky Mountain interior: Geological Society of America Special Paper 243, p. 71-105. Lucas, S.G. and Williamson, T.E., 1993. Late Cretaceous to early Eocene vertebrate stratigraphy and biochronology of the San Juan Basin, New Mexico; in Lucas, S.G. and Zidek, J., eds., Vertebrate paleontology in New Mexico: New Mexico Museum of Natural History and Science Bulletin 2, p. 93-104.
McKenna, M.C., Chow, M., Ting, S., and Luo, Z., 1989. Radinskya yupingae, a perissodactyl-like mammal from the late Paleocene of China; in Prothero, D.R. and Schoch, R.M., eds., The evolution of perissodactyls: New York, Clarendon Press, p. 24-36.
Rapp, S.D., MacFadden, B.J., and Schiebout, J.A., 1983. Magnetic polarity stratigraphy of the early Tertiary Black Peaks Formation, Big Bend National Park, Texas: Journal of Geology, v. 91, p. 555-572.
Rea, D.K., Zachos, J.C., Owen, R.M., and Gingerich, P.D., 1990. Global change at the Paleocene/Eocene boundary: climatic and evolutionary consequences of tectonic events: Paleogeography, Paleoclimatology, Paleoecology, v. 79, p. 117-128.
Schiebout, J.A., 1974. Vertebrate paleontology and paleoecology of Paleocene Black Peaks Formation, Big Bend National Park, Texas: Texas Memorial Museum Bulletin 24, 87 p.
Schiebout, J.A., Rigsby, C.A., Rapp, S.D., Hartnell, J.A., and Standhardt, B.R., 1987. Stratigraphy of the Cretaceous/Tertiary and Paleocene/Eocene transition rocks of Big Bend National Park, Texas: Journal of Geology, v. 95, p. 359-371.
Schoch, R.M., 1986. Systematics, functional morphology, and macroevolution of the extinct mammalian order Taeniodonta: Peabody Museum of Natural History Bulletin, no. 42307.
Sloan, R.E., 1987. Paleocene and latest Cretaceous mammal ages, biozones, magnetozones, rates of sedimentation, and evolution: Geological Society of America Special Paper, 209 p.
Sloan, R.E., 1969. Cretaceous and Paleocene terrestrial communities of western North America: Proceedings of the North American Paleontological Convention, part E, p. 427-453.
Standhardt, B.R., 1986. Vertebrate paleontology of the Cretaceous/ Tertiary transition of Big Bend National Park, Texas [Ph.D. Dissertation]: Baton Rouge, Louisiana State Universtiy, 299 p.
Ting, S., 1993. A preliminary report on an early Eocene mammalian fauna from Hengdong, Hunan Province, China: Kaupia, Darmstadter Beitrage zur Naturgeschichte. Heft3, p. 201-207.
West, R.M., 1987. Eocene (Wasatchian through Duchesnean) biochronology of North America: in Woodburne, M.O., ed., Cenozoic mammals of North America: Berkeley, University of California Press, p. 77-117. Wilson, J.A., 1967. Early Tertiary mammals; in Maxwell, R.A., Lonsdale, J.T., Hazzard, R.T., and Wilson, J.A., Geology of Big Bend National Park, Brewster County, Texas: University of Texas Bureau of Economic Geology, v. 671, p. 157-169.
Wolfe, J.A., 1989. North American Eocene vegetation, and its climatic implications: Transactions of the American Geophysical Union (EOS), v. 70, p. 375.
Wolfe, J.A., 1978. A paleobotanical interpretation of Tertiary climates in the northern hemisphere: American Scientist, v. 66, p.694-703.
Wood, H.E., Chaney, R.W., Clark, J., Colbert, E.H., Jepsen, G.L., Reeside, J.B., and Stock, C., 1941. Nomenclature and correlation of the North American continental Tertiary: Geological Society of America Bulletin 52, 48 p.
Woodburne, M.O., ed., 1987. Cenozoic mammals of North America: Berkeley, University of California Press, 336 p.
Figure 1. Big Bend biostratigraphy in global context. Where vertical zigzags are shown, a column has been condensed vertically, and where horizontal zigzags are shown, the column has been terminated (although further data is available). Columns 1-6 are modified from Woodburne (1987) and relate to all of North America: Column 1 - geochronology in millions of years; Column 2 - magnetic anomalies; Column 3 - North American Land Mammal Ages; Column 4 - zones or subzones; Column 5 - selected first appearances for North America; Columns 6-8 refer to Tornillo Flat, Big Bend, Texas and are modified from Schiebout et al. (1988). Column 9 was derived from Rapp et al. (1983). The bases of anomaly 26 and 24 have been used to select a scale for the Tornillo Flat data and to align it with that from Woodburne (1987). All of the Tornillo Flat columns are aligned with column 9. Column 6 - selected first appearances for Big Bend; Column 7 - fossil sites with height of Member in meters; Column 8 - partial stratigraphic section of Tornillo Flat in Big Bend, showing distinctive black mudstone layers; Column 9 - magnetic anomalies measured on Tornillo Flat; Column 10- stratigraphic nomenclature in Big Bend; Column 11 is extracted from Haq et al. (1987), which is global in scale. It was aligned with column 1 through a "time in millions of years" scale on the original. Column 11 - short term eustatic curve, labeled for associated systems tracts as follows: high stand (HS), low stand wedge (LSW), shelf-margin wedge (SMW), transgressive deposits (TR).
Table 1. Size comparisons of Big Bend Hyracotherium from the South Wall Site and Paramys from TT-Jacks with comparable animals from the northern intermontane basins. Figures on Big Bend Hyracotherium from Schiebout (1974), on H. sandrae from Gingerich (1989), on Big Bend Paramys from Hartnell (1980) and on P. taurus from Ivy (1990).