Palaeogeography, Palaeoclimatology, Palaeoecology, 63 (1988): 201-223 201 Elsevier Science Publishers B.V., Amsterdam -- Printed in The Netherlands TAPHONOMIC COMPARISON OF PASSIVE AND ACTIVE CONTINENTAL MARGINS: NEOGENE SHELL BEDS OF THE ATLANTIC COASTAL PLAIN AND NORTHERN GULF OF CALIFORNIA SUSAN M. KIDWELL Department of Geophysical Sciences, University of Chicago, 5734 S. Ellis Avenue, Chicago, IL 60637 (U.S.A.) (Received July 21, 1987) Abstract Kidwell, S.M., 1988. Taphonomic comparison of passive and active continental margins: Neogene shell beds of the Atlantic Coastal Plain and Northern Gulf of California. Palaeogeogr., Palaeoclimatol., Palaeoecol., 63:201 223. To test for differences in the nature of the fossil record as a function of terrigenous sedimentation, macroinvertebrate concentrations in a recently rifted continental margin (Pliocene Imperial Formation, southeastern California) are compared with a mature passive margin (Miocene Calvert and Choptank Formations, Maryland coastal plain). Previous work suggested that shell beds are primarily a product of absolute low net sedimentation rather than high skeletal input; the duration and environment of low sedimentation (erosion, omission) determine the quality of paleontologic data through time-averaging and selective preservation of assemblages, and the frequency of low sedimentation episodes determine the stratigraphic density of richly fossiliferous horizons for sampling. Settings of rapid subsidence -- and potentially high average sedimentation rates -- such as active continental margins are thus expected to have fewer and taphonomically less complex skeletal concentrations than slowly subsiding passive margins, and, where steep bathymetric gradients co-occur with high subsidence, skeletal concentrations should be diachronous, patchy in development, and highly variable laterally. These predictions are largely borne out by the shallow marine Imperial Formation, whose average sedimentation rate is several orders of magnitude higher than the Maryland Miocene section. In both tectonic settings, maj or accumulations record relatively prolonged episodes of low terrigenous sedimentation, attesting to the predictive and explanatory power of a sedimentologic model for fossil accumulation. Major complex shell beds in the Imperial Formation (Latrania member) are exceptional deposits, formed under conditions of sediment starvation along the seaward edges of prograding coastal alluvial fans and fan- delta complexes and on bedrock highs; whereas major shell beds in passive margin deposits are stratigraphically condensed records of repeated marine transgressions. Although actualistic experiments suggest that rapid, deep burial is essential to the preservation of skeletal fossils, the stratigraphic record demonstrates that periods of sediment starvation and bypassing -- and thus intermittent or retarded shallow burial of shells -- are fundamental to the formation of many major skeletal concentrations. Introduction ification of the original assemblages to some degree. Because most kinds of paleobiologic Paleontologists collect fossils where they analysis require the comparison of data from find them, both as sparsely dispersed specimens different sampling horizons, the goals of tapho- and from densely packed skeletal concentra- nomic analysis have shifted from the evalu- tions; practically all of these fossiliferous ation of bias within single assemblages to (1) deposits reflect post-morten, taphonomic mod- the identification of taphonomically compar- 0031-0182/88/$03.50 © 1988 Elsevier Science Publishers B.V. 202 able assemblages, regardless of level of bias, mation of southeastern California, part of the and (2) the search for systematic patterns in thickest exposed Plio-Pleistocene section in the distribution of fossiliferous deposits and North America. The Imperial Formation re- their geological and biological controls. Thus cords marine inundation and deltaic prograda- from a paleobiological perspective, tapho- tion in the northern Gulf of California during a nomic analysis reflects a more positive ap- period of tectonic transition, from Miocene proach, stressing what can be done with rifting to Quaternary oblique strike-slip fault- comparable but imperfect and irregularly ing associated with the San Andreas fault spaced paleontological samples, rather than system (Crowell, 1974; Moore and Curray, 1982; focussing on limits imposed by post-mortem Winker and Kidwell, 1986). It thus provides a modification of assemblages, picture of shallow marine sedimentation and Previous analysis of a broad spectrum of skeletal accumulation during rapid subsidence molluscan shell beds from shallow marine on a young divergent margin. Miocene strata of Maryland (Kidwell, 1982, Of all active margin records available for 1985, 1986a, b; Kidwell and Jablonski, 1983; study, the Imperial Formation was particularly Kidwell and Aigner, 1985) indicated that epi- attractive because it was reported (Menden- sodes of low net sedimentation governed the hall, 1910; Vaughan, 1917; Hanna, 1926; Wood- formation of skeletal concentrations in that ring, 1932; Woodard, 1963) to contain skeletal mature passive margin setting. Periods of concentrations similar in physical scale and erosion, sediment starvation, and bypassing faunal diversity to major shell deposits in the determined the stratigraphic distribution of Maryland Miocene. The sedimentologic model almost all skeletal concentrations, minor and would predict few major shell beds formed major in scale, and strongly influenced the through prolonged episodes of low net sedi- quality of shell preservation, time scales of mentation in active margins because of rapid skeletal accumulation (and thus time-averag- burial of bedding surfaces; the great thickness ing of assemblages), and patterns of benthic of the Imperial Formation suggested that colonization through taphonomic feedback, sediment supply probably had been sufficient to If patterns of sedimentation generally gov- keep up with rapid subsidence. Large-scale ern the formation of skeletal concentrations, skeletal concentrations should thus be prim- then it should be possible to predict the arily biogenic, the product of episodes or distribution of fossiliferous deposits and their environments of high skeletal input rather qualities, both as repositories of paleontologi- than low terrigenous sedimentation. Unless cal information as settings for biotic inter- the Imperial shell beds were biogenic in origin, actions. Moreover, the nature of the fossil they presented a possible contradiction to the record shoiuld differ quantitatively if not sedimentologic model. qualitatively among tectonic settings charac- terized by different patterns of sedimentation Californiaand Maryland Neogene settings and subsidence, as well as among different compared depositional systems and paleolatitudes. If, on the other hand, skeletal concentrations are The non-deltaic shallow marine facies of the primarily the product of high skeletal input Pliocene Imperial Formation (Latrania mem- and preservation potential and are influenced ber) and the Miocene Calvert (Plum Point only to a small degree by sedimentation rates, Member) and Choptank Formations are advan- then patterns of skeletal accumulation should tageous units for comparison (Fig.l). Both the vary less predictably through the record, active and passive continental margin records To test the role of sedimentation in dictating are characterized by terrigenous clastics de- skeletal accumulations, I examined biostrati- posited in a range of low-energy, intertidal to nomic patterns in the Pliocene Imperial For- shallow subtidal marine environments out of 203 Fig.1. Outcrop belts (black) of Pliocene Imperial Formation in southeastern California and of Miocene Calvert and Choptank Formations in the Maryland-Virginia coastal plain. A. CM= Coyote Mountain study area on northeast side of Elsinore Fault; FCV= Fish Creek Vallecito area; stippled pattern indicates active spreading centers; arrows indicate flow direction of northern distributaries of the Colorado River delta; NW-SE trending faults are part of San Andreas system. B. C = Calvert Cliffs; P= cliff exposures along Potomac River. the direct influence of deltas. Both contain biostratigraphic control [Melillo, 1982; diverse mollusk-echinoid-coral faunas with Kidwell, 1984; Schreiber, 1984; ~7.5 or 9.0 m.y. many extant congeners, duration for exposed section (Langhian to mid- The two records contrast strongly, however, Tortonian) based on scales of Berggren et al., in features related to tectonic setting. 1985 and Haq et al., 1987]. (1) Net rates of sediment accumulation were These time-averaged values of course say more than an order of magnitude higher along little about differences in short-term sedimen- the edges of the Pliocene Salton Trough. tation rates or steadiness of sediment accumu- Johnson et al. (1983) estimated post-compac- lation. However, the lack of significant uncon- tion accumulation rates of 4.0-5.5 mm/yr for formities in the Imperial section (Winker, 1987) the approximately 1 km thick marine deltaic and presence of numerous breaks in the part of the Imperial section from the Fish Maryland Miocene section (Kidwell, 1984; Creek-Vallecito area (Fig.lA) toward the cen- 1986c) suggest major differences in the com- ter of the basin, as calibrated by magneto- pleteness of the two records. stratigraphy (4.0-4.3 Ma). These deltaic strata (2) Initial dips (and thus the paleobathyme- overlie and intertongue