Hyperpycnal Flows and Taphonomy of Vertically Embedded Ammonites from the Upper Cretaceous of Antarctica

Eduardo B. Olivero
Laboratorio de Geología Andina CADIC-CONICET, Av. B. Houssay 200; 9410 Ushuaia, Tierra del Fuego, Argentina. E-mail: [email protected]

Introduction
The particular features of ammonite shells make them suitable for theoretical, hydrostatic calculations and Raup (1973) demonstrates that, under ideal conditions, vertically oriented cephalopod shells can only be preserved at very shallow water depths. Raup’s elegant conclusion that pressure-induced filling of intact phragmocones below a water depth of about 10 m will result in the loss of vertical stability was, however, in conflict with field observations. In a recent paper (Olivero, 2007) this conflict was partly solved by showing that transportation of cephalopod shells in turbid water could form a “clay plug” in the siphuncular tube, so that ambient water pressure would not longer be in equilibrium with the gas pressure in the phragmocone. These vertically embedded ammonite shells were transported by gravity flows and the turbulent sediment cloud was forced by differential pressure to enter into the siphuncular tube, originating the clay plug. After flow deceleration, the ammonite shells were frozen in vertical position by deposition of the suspended material. However, as the intact ammonite shells did not preserve the aptychi in the body chamber, the solution of this conflict originated a new taphonomic problem: that the ammonite organisms could not be killed and their shells transported, emptied, and deposited by the same, short sedimentary event. The analysis of this problem is addressed in this study showing that this apparent contradiction could be solved if the ammonite were killed, transported, and deposited by long-lived gravity flows originated by rivers in flood.

Stratigraphic and sedimentologic framework
In the James Ross Basin, Antarctica, the Upper Cretaceous Santa Marta Formation consists of four vertically stacked intergrading facies associations (Facies Associations A to D) that define a regressive sequence about 1 km thick (Scasso et al., 1991). In the lower part of the unit, Facies Association A is dominated by massive or laminated muddy tuffaceous, very fine-grained sandstone. Small fragments of carbonaceous plant material as well as large tree trunks are abundant locally. Articulated inoceramid bivalves as well as heteromorph ammonites (Baculites spp.) preserved commonly in vertical position are dominant in friable muddy sandstones.
Facies Association B consists of a regular alternation of coarsening and thickening upward graded tuffaceous sandy turbidites capped by laminated, carbonized plant fragments. At regular intervals of 4–6 m, highly bioturbated and fossiliferous tuffaceous turbidites are intercalated in the succession. The later beds preserve an abundant and diverse ammonite fauna, with many specimens in vertical position, either dispersed in the bed or concentrated inside and around the body chamber of large pachydiscid shells (shelter preservation).
Facies Association C includes thick, tuffaceous, graded pebbly or coarse-grained sandy turbidites cut erosively by channeled conglomerate and debris flows, both with abundant resedimented fossils and concretions.
At the top of the unit, Facies Association D consists of an alternation of bioturbated, fine-grained, well-sorted, micaceous sandstone, silty fine-grained sandstone, and mudstone, with abundant plant fragments, leaves, and large tree trunks. Small ammonites are associated generally with wood fragments. The dominant bivalves Pterotrigonia sp. and Cucullaea sp. are preserved commonly in their presumed life position. Bioturbated beds are cut erosively by lenticular conglomerate, coquina, and parallel-laminated or cross-bedded sandstone. Vertical ammonites are preserved occasionally within parallel-laminated sandstone.
The vertical stacking of these four facies associations reflects progradation of a deep-water delta system (Scasso et al., 1991). This system includes the following subenvironments arranged in a distal-to-proximal position: prodelta/basin plain (Facies Association A); base-of-slope depositional lobes (Facies Association B); slope-channel complex (Facies Association C); and inner-shelf-delta plain (Facies Association D). In this regressive sequence, the vertical stacking of Facies Associations B and C implies that their respective depositional elements were spatially adjacent, defining a distal (lobe deposits) to proximal (slope-channel complex) trend. Sedimentary structures are difficult to observe in these dominantly friable sedimentites; accordingly, they could not be used for definition of facies tracts. Nonetheless, grain-size, thickness, and bedding geometries vary systematically across these depositional elements and are useful features to characterize facies tracts. The most distal lobe deposits of Facies Association B are typified by thin turbidite beds, which are characterized by broadly parallel bedding, relatively fine-grained sediment caliber, abundant comminuted plant debris, and dense concentrations of vertically embedded ammonites in particular horizons. The proximal channeled conglomerate, debris flows, and thick turbidite beds of Facies Association C are characterized by lenticular bedding geometries with ample evidence of large asymmetric erosive surfaces, abundant coarse-grained sediment caliber, and large plant debris. The asymmetric erosive surfaces are filled with 3-5 m thick fining upward packages that are similar to those described in the literature for the channel-lobe transition zone in the deep sea. The filling of these sharply defined erosive features --typically cut into friable substrates with high relief margins-- suggest rapid deposition during a single sedimentary event. Otherwise, the preservation of vertical, sometimes even negative, slopes would be difficult to model.

Ammonite taphonomy
Facies Association B has several beds characterized by dense concentrations of intact, vertically embedded ammonite shells, which are dominantly represented by heteromorph, lytoceratid, pachydiscid, kossmaticeratid, and phylloceratid morphotypes. These beds with dense ammonite concentrations mark the base of coarsening and thickening upward sedimentary cycles that are thought to reflect progradation of depositional lobes (Olivero, 2007). The vertical ammonite shells are commonly preserved inside and around the body chamber of large, pachydiscid ammonites (shelter preservation), and in particular cases more than 70 vertically embedded ammonite specimens were recovered in a single case of shelter preservation. Associated with these vertically embedded shells are abundant plant fragments. Several lines of evidence indicate that the vertical orientation of ammonite shells is a primary feature. In particular, primary vertical orientation is indicated by: a) congruent orientation of the inferred buoyancy and gravity centers of the shell; b) congruent orientation of geopetal structures in the phragmocone; c) congruent orientation of associated bivalves, e.g. Pinna sp., in life orientation; d) deformation of vertical ammonite shells congruent with stretching of the shell by vertical loading; and, above all, e) congruent paleocurrents orientation of ammonite shells, other fossil particles, and physical sedimentary structures.
The observation of polished sections cut along the siphuncular plane indicates that the phragmocone of vertical embedded ammonite shells preserve intact chambers filled with drusy calcite. In contrast, the filling material in the body chambers consists of fine-grained sand and silt particles, clayed matrix, and micrite. Interestingly, the filling of the siphuncular tube is also a mixing of clay, silt, and micrite, suggesting that during transportation a clay mixture entered into the siphuncle, avoiding or delaying the waterlogging of the phragmocone. In spite of the complete preservation, none of the sectioned ammonite shells have any evidence of the jaw apparatuses inside the body chamber.

Discussion and conclusions
The absence of reworking, the intact preservation of fragile shells, and the congruent orientation of ammonite shells and physical current structures indicate that the vertically embedded ammonites of the Santa Marta Formation were frozen in that position during single sedimentary events. In particular, the event beds represented by relatively thin turbidites, e.g. the turbidites locate at the base of prograding lobe deposits in Facies Association B, often concentrate large numbers of vertically embedded ammonite inside and around large pachydiscid shells. The common association of abundant small twigs, showing congruent current orientation with ammonite shells, indicates that both the shells and the twigs behave as hydraulically equivalent sedimentary particles during transportation and deposition from the gravity flow.
After death of the organism, the vertical stability of shells floating in the water mass ought to be a transient feature, and thus it is tempting to think that the organisms were killed, and their shells transported and deposited by the same gravity flow. In that case, one would expect to find evidence that the ammonite soft parts were still inside the body chambers, a fact that is generally inferred from the preservation of the hard part (generally the aptychus) of the ammonite jaws. Intact ammonite shells, with their phragmocones filled with drusy calcite and with complete body chambers, have been traditionally searched as perfect targets for recovery of the jaw apparatuses, with the obvious implication that in these cases the soft part of the organism should still be inside the body chamber at the time of burial. In this regards, the ammonites from the Santa Marta Formation are disappointing as none of them show any evidence of the aptychus inside the body chamber, implying that the ammonite soft parts were lost before final burial of the shell.
New experimental results with fresh dead Nautilus animals have demonstrated that waterlogging of the phragmocone can only be accomplished after the mantle tissue detaches from the shell due to decomposition. The detachment process lasts usually more than 1 day and, depending on organism size and water temperature, it could lasts for several days (Wani et al., 2005). As in the ammonites from the Santa Marta Formation aptychi are never found inside the body chamber (implying that soft parts were lost before burial) these results reinforce the assumption that vertically embedded ammonites could not have been killed, transported, and deposited by the same short-lived gravity flow, i.e. the type of flow commonly associated with classical turbidites. On the contrary, it could be assumed that the sequence of involved events --death, tissue decomposition, transportation, and deposition-- was accomplished in different phases. This assumption, however, generates major taphonomic problems and is rejected because it fails to explain how the vertical stability of the shell could have been maintained for relatively long periods, involving the time span from the animal death in the initial phase, to the burial of vertically embedded shells in the final phase.
These inconsistencies, however, could be solved if death of the organisms and final burial of their shells were accomplished during one, long-lived gravity flow. Such a flow would endure enough time as to accommodate the massive killing of organisms, tissue decomposition and lost of soft parts, shell transportation and generation of the clay plug, shell concentration, and rapid sediment deposition that kept the shells in vertical position. Thus, the turbidites of Facies Association B are probably the deposits of long-lived gravity flows generated by rivers in flood. The interpretation of these deposits as hyperpycnites fits well into the delta-influenced settings of the Santa Marta Formation and allows for a better understanding of the dense concentrations of vertically embedded ammonites and plant fragments.

References
Olivero, E.B., 2007. Taphonomy of ammonites from the Santonian-lower Campanian Santa Marta Formation, Antarctica: sedimentological controls on vertically embedded ammonites. Palaios, 22 (6): 586-597.
Raup, D.M., 1973. Depth inferences from vertically imbedded cephalopods. Letahia, 6: 217-226.
Scasso, R.A., Olivero, E.B., and Buatois, L.A., 1991. Lithofacies, biofacies and ichnoassemblages evolution of a shallow submarine volcaniclastic fan‑shelf depositional system (Upper Cretaceous, James Ross Island, Antarctica). Journal of South American Earth Sciences, 4: 239‑260.
Wani, R., Kase, T, Shigeta, Y., and De Ocampo, R., 2005. New look at ammonoid taphonomy, based on field experiments with modern chambered nautilus. Geology, 33 (11): 849-852.

AAPG Search and Discovery Article #90079©2008 AAPG Hedberg Conference, Ushuaia-Patagonia, Argentina