Ted J. Fremd

National Park Service

420 West Main St.

John Day, OR 97845

Xiaoming Wang

Department of Vertebrate Paleontology

American Museum of Natural History

New York, NY 10024

INTRODUCTION

The value and role of many National Park System units in paleontological studies is becoming increasingly evident. In addition to fostering research efforts and interpretation within "outdoor laboratories", these areas contain key depositional sequences for the calibration of biotas throughout North America. This GSA symposium volume offers an opportunity to address an important aspect involved in our renditions of ancient ecosystems, biostratigraphic correlation.

Generally speaking, certain attributes characterize outstanding depositional sequences that are important for biostratigraphic analyses.

1. There are large numbers of very well to excellently preserved specimens available, both in the field and in repositories.

2. The taxonomic diversity of the assemblages is high, sampling a broad spectrum of the ecosystem. Ideally, there are associations of flora and fauna.

3. The fossiliferous strata span a long time interval with few depositional unconformities.

4. The sequence was deposited during momentous processes and events in earth history, such as major climatic changes, adaptive radiations, or mass extinctions.

5. A variety of dateable volcaniclastic marker beds permit ready correlation of the strata both within the depositional reservoir and with those documented from elsewhere.

Outstanding portions of such a record exist in central and eastern Oregon, and include over three hundred and fifty localities within the Clarno Group and the John Day, Mascall, and Rattlesnake Formations. Among early workers, Merriam (1901, 1907) and Chaney (1924) recognized the area as one of a few in the world with such a complete record written on the canyon walls. This is a benefit of proximity to volcanic sources of clastics, from a variety of vents, over a long time. In fact, few Tertiary fossiliferous areas in North America compare to those of the John Day Basin in terms of their importance in documenting long term change.

The accuracy of this "big picture" of biotic change is largely governed by the precision with which the detailed data are collected. At least two critical aspects of the biological assemblages must be addressed. Firstly, the taphonomic biases must be detected and analyzed. Secondly, good stratotypic data must exist for the type specimens; if not, the faunal ranges must be ascertained and a detailed biostratigraphic analysis conducted (see, for example, Rensberger, 1984). This latter problem continues to plague many historically important North American sites in general and the John Day Basin in particular (see Woodburne, 1977).

The John Day Formation itself has had a long and complicated history of investigation, adequately reviewed elsewhere (e.g., Robinson, 1990; Rensberger, 1984). Regrettably, the major collections of vertebrate fossils (UCB, UW, UO, YPM, USNM, and AMNH) from the area lack useful stratigraphic and site data, or the locality information is inaccessible. As a result, confusion has arisen concerning the "John Day fauna", which cannot be viewed as a single fauna, but consists of at least eight different lithosympatric assemblages (sensu Stucky, 1990). The eastern component of the strata has been subdivided into 4 members (Fisher and Rensberger, 1972), recognized largely by color and/or zeolitization features. These subdivisions, while useful for diagenetic analyses, are not as useful for either chronostratigraphic or lithostratigraphic understanding of the strata.

In this short paper, we shall narrow the focus and describe work performed on one taxonomic group, occurring in a restricted interval, in a small geographic area known as "Turtle Cove", of the John Day Formation (sensu Robinson, 1984, 1990). These data are culled from a larger taxonomic work on the John Day canids (Wang and Fremd, in prep).

MATERIALS AND METHODS

Beginning in 1982, field investigators began to retrieve specimens under more systematic conditions (Wagner and Ruben, 1983) than had generally been the case. A system of "cyclic prospecting" begun in 1984 at John Day Fossil Beds National Monument (JODA) has resulted in the salvaging of thousands of specimens that would have been lost to weathering or inadequate data recording (Fremd, 1992a). Work outside the national monument was begun with cooperative agreement with the BLM (Hanson and Allen, 1988, Fremd, 1989, 1992b); this has since grown to include work in the other occurrences of the strata in the state.

Instead of simply duplicating existing and inadequate collections, an effort was made to avoid collecting biases as much as possible. For each specimen encountered, the taxon, element, coordinates (on detailed aerial photography), in situ or float, stratigraphic height, and numerous other features were recorded into the field notes. All collections have been made relative to a number of prominent ash-flow tuffs (see Figure 1.), which permit a much more refined reconstruction of narrow biostratigraphic intervals.

Each item collected was accessioned and catalogued, using the standard DBASEIII+ Automated National Catalog System in use in all national park system units. Additionally analysis was facilitated by interlinking these data supplemented by specimen and locality databases, including information compiled at other major repositories, linked together in IDEALIST, APPROACH, and ECCO in Windows.

As a result of these efforts over 7,000 specimens have been salvaged from strata in the park environs, including a variety of canids, with most specimens pinpointed to within a few meters. One of us (XW) examined specimens of the Canidae (Carnivora, Mammalia) with their preliminary identifications and field numbers, but no stratigraphic information. Tend species, belonging to three subfamilies, can be recognized on morphological grounds. Only after precise taxonomic assignments were made did we plot these into our stratigraphic column, avoiding in such fashion a priori biases and circular reasoning.

Within the newly defined Turtle Cove Member of the John Day Formation, the biostratigraphic distribution of the canids can now be resolved with high precision. Confidence is highest in small canids (the 4 species of Borophaginae) in the lower 2/3 of the Turtle Cove Member, because of the relatively larger samples. Within this new stratigraphic framework, most canid species have more restricted occurrences than have been assumed in the traditional practice of lumping all taxa into a broad "Turtle Cove" unit (Figure 3).

The earliest canids in the John Day Formation are represented by the simultaneous occurrence of the sister-species "Cormocyon" lemur and "C." latidens in strata equivalent to the upper Brule Formation (Whitneyan) of the northern Great Plains. Both of these small borophagine taxa disappeared just below the Picture Gorge Ignimbrite (PGI, 28.7 m.y. from single-crystal Ar/Ar date), which serves as a prominent marker bed in the regional stratigraphy. Approximately 70 meters above the lowest occurrence of the lemur-latidens pair begins the first appearance of Cormocyon oregonensis, a small, primitive borophagine canid whose ancestry may be traced to the Orellan strata of the Plains (e.g., Hesperocyon pavidus, see Wang, in press; Wang and Tedford, in press). The stratigraphic range of C. oregonensis extends considerably above the PGI, whereas a larger, more derived species, C. copei, is almost entirely confined above the PGI.

Stratigraphic distribution of the median, coyote-sized hesperocyonines can also be tied with their Plains' relatives. The two John Day species of Mesocyon (M. coryphaeus and M. brachyops) can be traced to rocks well below the PGI, close to the earliest occurrence of Cynodesmus in the upper Brule strata of the Plains. Such a contemporary appearance of two genera in the same stage of evolution on either side of the Continental Divide is potentially useful for biostratigraphic correlations, and indicates a pre-Whitneyan divergence time of the Mesocyon-Cynodesmus sister-clades.

CONCLUSION

This work is being linked with several other concurrent taxonomic and stratigraphic investigations, including a contracted project involving highly detailed mapping of the ash-flow sheets, paleosols, and other sediments from the Painted Hills and Clarno Units of JODA (see Bestland et al., this volume). The John Day Basin is assuming an increasingly important role in refining Tertiary biostratigraphy in North America. The results of these efforts are increased resolution of evolutionary processes and events that have long been blurred.

ACKNOWLEDGEMENTS

Former JODA Superintendent Ben Ladd eagerly encouraged collection of these data as part of ongoing resource management activities, and always supported "Paleo First!" at the monument. Hugh Wagner and John Ruben performed the fieldwork in 1982, 1983, and 1984. Data collected by C. Bruce Hanson and Garwood Allen also proved helpful. Seasonal rangers, volunteers, and field workers too numerous to mention apprehended many of the specimens. X. Wang wishes to thank Richard H. Tedford for access to his unpublished manuscripts on borophagine and canine canids, which provided phylogenetic frameworks for the taxa involved. Funding for travel by the junior author was provided by postdoctoral fellowship funds of the American Museum of Natural History.

REFERENCES

Chaney, Ralph W., 1924. Quantitative Studies of the Bridge Creek

Flora: American Journal of Science, Fifth Series, v. 844, 127-144.

Fisher, R.V. and Rensberger, J.M., 1972. Physical stratigraphy of

the John Day Formation, central Oregon: University of California Publications Geological Sciences, v. 101. pp. 1-33.

Fremd, Ted J., 1988. Assemblages of Fossil Vertebrates in

Pre-Ignimbrite Deposits of the Turtle Cove Member, John Day

Formation (Arikareean), from Outcrops within the Sheep

Rock Unit, John Day Fossil Beds National Monument, Oregon: Journal of Vertebrate Paleontology, v. 8(3), 15A.

Fremd, Ted J., 1991. Early Miocene Mammalian Populations from

Turtle Cove, Oregon: Journal of Vertebrate Paleontology, v. 11(3), 29A.

Fremd, Ted J., 1992a. Cooperative Management of Paleontological

Resources in the John Day Basin, Central Oregon: Seventh

Conference on Research and Resource Management, George Wright

Society, 24.

Fremd, Ted J., 1992b. Paleontological Resources Management and

its Practical Applications: Third Conference on Fossils in the NPS, Program with abstracts, 15.

Merriam, John C., 1906. Carnivora from the Tertiary Formations of

the John Day Region: Bulletin of University of California Department of Geology, v. 51. pp. 1-64.

Merriam, John C., 1901. A contribution to the geology of the John

Day Basin [Oreg.]: Bulletin of University of California

Department of Geology, v. 2(9), pp. 269-314.

Rensberger, John M., 1983. Successions of meniscomyine and

allomyine rodents (Aplodontidae)in the Oligo-Miocene John Day

Formation, Oregon: University of California Publications Geological Science, No. 124, xiv-157.

Robinson, P.T., Walker, G.W., and McKee, E.H., 1990.

Eocene (?), Oligocene, and Lower Miocene Rocks of the Blue

Mountains Region; in Walker, George W.,ed. Geology of the Blue Mountains Region of Oregon, Idaho, and Washington: Cenozoic Geology of the Blue Mountains Region: U. S. Geological Survey Professional Paper 1437, p.p. 29-62.

Robinson, P.T., Brem, G.F. and McKee, E.H., 1984. John Day

Formation of Oregon: a distal record of early Cascade

volcanism: Geology (Boulder), v. 12(4), 229-232.

Stucky, Richard K., 1990. Evolution of Land Mammal Diversity in

North America during the Cenozoic: Current Mammalogy, v. 2, 375-432.

Wagner, H. and Ruben, J., 1984. Bone Occurrences in the Turtle

Cove Member of the John Day Formation, Oligocene, Central Oregon: National Park Service CPSU/OSU publication # 84-10.

Wang, X., In press. Phylogenetic systematics of the Hesperocyoninae (Carnivora: Canidae): Bulletin of the American Museum of Natural History.

Wang, X., and Tedford, R.H., In press. The initial diversification of canids in the White River Group, Eocene and Oligocene of North America; in D.R. Prothero and R.J. Emry (eds.), The Terrestrial Eocene-Oligocene Transition in North America: Cambridge University Press.

Woodburne, M.O., 1977. Definition and characterization in

mammalian chronostratigraphy: Journal of Paleontolology, v. 51, No. 2, 220-234.