«Paleoenvironments and Paleoecology of the Vertebrate Fauna from a Late Cretaceous Marine Bonebed, Canada Stephen L. CUMBAA, Charlie J. UNDERWOOD and ...»
Mesozoic Fishes 5 – Global Diversity and Evolution, G. Arratia, H.-P. Schultze & M. V. H. Wilson (eds.): pp. 509-524, 5 figs., 1 tab.
© 2013 by Verlag Dr. Friedrich Pfeil, München, Germany – ISBN 978-3-89937-159-8
Paleoenvironments and Paleoecology
of the Vertebrate Fauna
from a Late Cretaceous Marine Bonebed, Canada
Stephen L. CUMBAA, Charlie J. UNDERWOOD
and Claudia J. SCHRÖDER-ADAMS
Bonebeds – concentrations of bioclastic debris of vertebrates in geological strata – can accumulate under a variety of conditions. They are common in marine deposits of the Late Cretaceous Western Interior Seaway of North America, and are often characterized as lag deposits. In general, these deposits represent unknown periods of accumulation and contain a mélange of taxa, possibly transported from a variety of habitats. As such, the contents of these marine bonebeds are often considered less useful for studies of paleoecology and paleoenvironments than are fossils recovered from rock units that represent continuous sedimentary deposition within one or more contiguous paleoenvironments. We examined the fossil contents of one particularly rich marine bonebed of “middle” Cenomanian age to determine if useful conclusions can be drawn with respect to the habitats and probable interactions of its pre-depositional fauna. This bonebed occurs as discontinuous lenses in shales of the upper part of the Belle Fourche Member of the Ashville Formation in the Pasquia Hills of Saskatchewan, Canada.
Acid preparation of these lenses revealed an assemblage containing: 20 chondrichthyan taxa: a chimaeriform, hybodontiforms, diverse lamniforms and rare rajiforms; 15 actinopterygian taxa: a caturid, pycnodonts, an aspidorhynchid, a pachycormid, a plethodid, ichthyodectids, pachyrhizodontids, an albulid, a putative salmoniform, enchodontids, and an acanthomorph; and nine tetrapods including turtles, pliosaurs and elasmosaurs, four marine bird taxa, a terrestrial bird, and a lizard. Evaluation of probable habitats of the fish fossils reveals many pelagic forms, but a sparse nectobenthic fauna. Only one or two genera of those identified are considered to be euryhaline, and there are no obligate freshwater forms. The overwhelming majority of taxa identified are fully marine. Inferred feeding strategies for fish taxa include several durophagous forms including Ptychodus, a pycnodont and an albulid, with most other taxa, including the most abundant shark species (lamniforms) and osteichthyans (Enchodus) appearing to have been active predators with piercing dentitions. There are also predators/scavengers with cutting dentitions (anacoracid sharks) and a probable planktivore (Cretomanta). Interpretation of the taphonomy and faunal content indicates that the bonebed accumulated near shore over a period up to tens of thousands of years and was largely composed of bioclastic detritus from shallow, contiguous habitats, with some input from a deeper, anoxic shelf assemblage.
The Western Interior Seaway (WIS) was an epicontinental sea that covered the middle of North America during the last 35 million years of the Cretaceous, linking the proto-Atlantic/Tethys seas to the south with the paleo-Arctic Ocean to the north (Fig. 1). Vertebrates were diverse in the seaway by the “mid”Cenomanian, with more than 70 taxa known from fossil localities in Canada and the United States (RUSSELL 1988, CUMBAA et al. 2010).
Most marine vertebrate remains of this age in North America occur in bonebeds, typically transgressive lag deposits, which are relatively common in the “mid”-Late Cenomanian. However, in sediment-starved portions of the basin, such as the eastern margin, bonebeds often occur as the result of winnowing during basin-wide sea-level lowstands and/or storm deposits (SCHRÖDER-ADAMS et al. 2001). The eastern margin of the seaway received dramatically less siliciclastic sedimentation than the Rocky Mountain foredeep on the western side of the basin (MCNEIL & CALDWELL 1981, CHRISTOPHER et al. 2006), a physical factor that facilitated the concentration of biogenic components in the sediment. The bonebed discussed here, from the Pasquia Hills of Saskatchewan, is thought to have been formed by winnowing on a shallow seafloor, probably near shore, above storm wave base and under tidal influence (SCHRÖDER-ADAMS et al. 2001, CUMBAA et al. 2006).
Material and methods
The vertebrate fossil material reported here was discovered along the Bainbridge River in the Pasquia Hills of Saskatchewan, Canada, during field work led by the senior author, under a Saskatchewan Heritage Palaeontological Resource Investigation permit. The bulk of the material was collected by the authors in 2006, and is catalogued in the Earth Sciences collections of the Royal Saskatchewan Museum, Regina.
The intact bonebed lenses were brought back to the laboratories of the Canadian Museum of Nature, and were cleaned and then immersed in deep trays of water for several days, at which time a meter was used to record baseline pH. The trays were then drained, and the lenses re-immersed in buffered 5-10 percent acetic acid.
After repeated acid changes over many weeks, the disaggregated material was removed, rinsed, and placed in regularly changed tap water until the pH matched the baseline. The specimens were wet sieved and separated by size for ease of sorting using a standard stacking graduated mesh series to 0.5 mm, retaining the fines that passed through the last screen. One large lens from the locality had been exposed to the elements for several years; it had disaggregated naturally, and was collected in place. The dried matrix was sorted under magnification, using a binocular microscope. All residue 1 mm or larger was sorted. After cursory examination of several samples of the finer sieved fractions under a binocular microscope yielded no new taxa, we picked through only a representative sample.
Some larger lenses (~60 cm by 20 cm) of the Bainbridge River bonebed were found in situ a few metres apart, within the same bedding plane and filling the troughs or rippled surfaces (CUMBAA & BRYANT 2001). The individual lenses taper at each end and show layering, with deposition of coarse bioclastic material sandwiching thin mudrock interbeds (PHILLIPS 2008). Ripple marks on the upper surface of some of the bonebed lenses indicate wave influence, and mud drapes at the bottom of others show tidal influence during bed formation (SCHRÖDER-ADAMS et al. 2001). Large calcite-lined vugs encasing bentonitic clay are common within the lenses, most likely indicating chunks of bentonite having been ripped up from the seafloor during a storm, ultimately becoming part of the lenses as bioclast-studded mud balls. The calcitic matrix includes a small percentage of inoceramid-derived prismatic calcite, often as single prisms but occasionally in bundles (SCHRÖDER-ADAMS et al. 2001). Rarely, a larger portion of a single inoceramid valve is incorporated into a lens. The bonebed lenses are characterized by obvious vertebrate fossils, primarily teeth, coprolites and disarticulated bones, which are highly concentrated, vary considerably in size, and are for the most part randomly oriented (Fig. 3).
Bonebed taphonomy and depositional setting Although bonebeds have been extensively examined, the vast majority of studies have focused on bone concentrations in non-marine palaeoenvironments. A taphonomic study was conducted of the bonebed discussed here and others of its age in the Manitoba Escarpment of eastern Saskatchewan and western Manitoba (PHILLIPS 2008). This bonebed formed in thin layers and lenses, which contain more than 55 % phosphatic bioclasts (PHILLIPS 2008); an earlier study characterized samples as containing 48-70 % bone and phosphatic nodules, many of the latter clearly being coprolites (SCHRÖDER-ADAMS et al. 2001).
There was little or no terrigenous sedimentation along most of the eastern shore of the Seaway during the “mid” Cenomanian, resulting in condensed sections. Observations along the entire Manitoba Escarpment demonstrate that lands bordering the eastern margin of the WIS were flat to low-lying, with little gradient to rivers and streams emptying into the sea (MCNEIL & CALDWELL 1981). The lack of sediment from terrestrial erosion is confirmed by detailed analysis of the bonebed; argillaceous material and siliciclastic grains are rare to absent, with the phosphatic clasts suspended in a matrix of sparry calcite cement. Less than three percent inoceramid-derived prismatic calcite is observable in point count analysis, but is readily seen in SEM microphotographs (PHILLIPS 2008).
Element shape would seem to be the main predictor of good preservation within the bonebed. Chondrichthyan elements, almost exclusively teeth, exhibit very little wear and breakage attributable to postdepositional processes. Osteichthyan remains include elements from the entire body, but only teeth and vertebrae are well preserved. Flat, blade-like and brittle elements such as the bones of the skull and the opercular series are almost always broken, usually beyond identifiability. Scales are present, but rarely so, and are almost always broken. Long and cylindrical elements such as teeth, on the other hand, fare very well, but are only occasionally preserved in place in tooth-bearing elements. The spines and processes of vertebrae are generally missing, but the spindle or barrel-shaped centra are very well preserved. Bird teeth and bones representing the entire body, with the exception of the relatively flat and thin sternum, are present and exhibit relatively little abrasion (SANCHEZ 2010). Some large tetrapod elements, such as those of plesiosaurs, can show considerable abrasion, while others, such as small flipper elements, are very well preserved. Overall, the breakage, abrasion and disarticulation of the vertebrate fossils indicate a significant level of transport and/or winnowing, but comparatively less so than in many other bonebeds in the region (PHILLIPS 2008).
Coprolites, almost certainly of piscine origin due to their size, number, and similarity in form and content to intestinal molds (cololites) that we have observed in articulated fossil fishes in localities of Early Turonian age within the region, are a significant component of the bonebed. It is likely that most of the rounded calcium phosphate nodules in the bonebed are of coprolitic origin. These range in size from about 3-20 mm in greatest dimension, and many have obvious fish vertebrae and scale inclusions. The sheer number of what appear to be fully formed and undistorted coprolites, distributed throughout all lenses, indicates the accumulation of a continuous rain of biotic debris on and into a soft bottom, where the coprolites stayed long enough to harden. This suggests depositional conditions in relatively quiet water with little initial disturbance or transport, and lack of a significant detritivore fauna. This is indicative of a relatively inhospitable initial depositional environment, such as anoxic bottom waters postulated for these levels in the description of the Ashville Formation (MCNEIL & CALDWELL 1981) and for the layer containing the bonebed lenses at the Bainbridge locality (SCHRÖDER-ADAMS et al. 2001).
Many mechanisms could result in the accumulation of biotic debris, including deposition in place, storm deposits, erosional deposits, or a time-averaged accumulation as the result of a transgressive lag, with incoming waters “rolling up” bones and teeth exposed for possibly considerable and certainly variable periods of time after initial deposition (WALKER & BAMBACH 1971, KIDWELL & BOSENCE 1991, ROGERS & KIDWELL 2000). Deposition in place (“within-habitat” of KIDWELL & BOSENCE 1991) can be eliminated as a mechanism by the evidence of the coprolites, and probably by the large size of the faunal community of differing habitats. Storm deposition – a single event – also appears unlikely due to the high concentration of almost exclusively vertebrate remains and the virtually complete lack of articulated specimens. ROGERS & KIDWELL (2000) examined Late Cretaceous (Campanian) vertebrate deposits, including those from the transgressive advance of the Bearpaw Sea, the last major expansion of the WIS.
They found that vertebrate remains seem to ‘behave’ differently from shell deposits, and postulated shoreface on shoreface erosional deposits – erosion of older shoreline deposits by wave and current action – to explain the accumulation of vertebrate remains. Although those deposits from the western edge of the WIS represent a similar assemblage to that of the eastern deposit discussed here, detailed stratigraphic study of the Bainbridge locality (SCHRÖDER-ADAMS et al. 2001) reveals no sign of the kind of erosion described by ROGERS & KIDWELL (2000).
Fauna and their potential paleoenvironments
The taphonomy of the bonebed suggests variation in the bottom conditions under which the bioclasts were deposited, including depth and degree of transport. In a hypothetical transect from a shelf-like, quiet and anoxic environment (SCHRÖDER-ADAMS et al. 2001, SIMONS et al. 2003) to a more active and oxygenated shoreline, more than one potential paleoenvironment and a variety of habitats are likely to have been
10 mm B Fig. 3.
Bainbridge River bonebed. A, thin-section through a bonebed lens. Note random element concentration and orientation in the upper two-thirds of the thin-section versus a stronger alignment of elongated elements and slightly coarser, more compacted matrix in the lower third, hinting at variable uni-directional current action throughout the deposition of the bonebed. B, Partially acid-prepared surface of a bonebed lens from the locality.
Representative specimens include: 1, one of several coprolites and oblong to rounded calcium phosphate pebbles;
2, osteichthyan vertebrae; 3, Enchodus palatine; and 4, Archaeolamna tooth. Teeth of sharks and bony fishes are the most common identifiable elements represented, followed by osteichthyan vertebrae and coprolites.
present, eventually contributing fauna to the bonebed. Table 1 lists the vertebrates identified from the bonebed, the relative abundance of each taxon, and its presumed ‘usual’ habitat or paleoenvironment(s).