Controls on Soft Tissue and Cellular Preservation in Late Eocene and Oligocene Vertebrate Fossils from the White River and Arikaree Groups of Nebraska, South Dakota, and Wyoming

Abstract

Previous studies on microtaphonomy have identified multiple types of organic microstructures in fossil vertebrates from a variety of time periods and past environmental settings. This study investigates potential taphonomic, paleoenvironmental, and paleoclimatic controls on soft tissue and cellular preservation in fossil bone. To this end, fifteen vertebrate fossils were studied: eight fossils collected from the Oligocene Sharps Formation of the Arikaree Group in Badlands National Park, South Dakota, and seven fossils from formations in the underlying White River Group, including the Oligocene Brule Formation of Badlands National Park, and the Eocene Chadron Formation of Flagstaff Rim, Wyoming; Toadstool Geologic Park, Nebraska; and Badlands National Park, South Dakota. A portion of each fossil was demineralized to identify any organic microstructures preserved within the fossils. We investigated several factors which may have influenced cellular/soft tissue decay and/or preservation pathways, including taxonomic identity, paleoclimatic conditions, depositional environment, and general diagenetic history (as interpreted through thin section analysis). Soft tissue microstructures were preserved in all fossil samples, and cellular structures morphologically consistent with osteocytes were recovered from 11 of the 15 fossil specimens. Preservation of these microstructures was found to be independent of taxonomy, paleoclimate regime, apatite crystallinity, depositional environment, and general diagenetic history, indicating that biogeochemical reactions operating within microenvironments within skeletal tissues, such as within individual osteocyte lacunae or Haversian canals, may exert stronger controls on soft tissue and biomolecular decay or stabilization than external environmental (or climatic) conditions.

1. Introduction

The dramatic global climate change that occurred at the Eocene–Oligocene boundary has been the focus of extensive paleontological and paleoenvironmental study (e.g., [1,2,3,4]). During the Eocene-Oligocene transition (approximately 34–33 million years ago), global temperatures cooled and terrestrial landscapes became drier, with the vast forests that covered much of the world in the Paleocene and Eocene giving way to grasslands and shrub prairies [3]. This shift in climate and environments led to accompanying shifts in terrestrial faunas, especially in North America [2]. The terrestrial Eocene–Oligocene transition is preserved in the strata of the White River Group, which spans multiple states across the Great Plains and records several million years, from the late Eocene to the Middle Oligocene (Figure 1). Badlands National Park in southwestern South Dakota contains a well-preserved section of the White River Group, with both Eocene and Oligocene strata present and easily accessible within the park boundaries. Previous work on the Badlands has detailed the paleoecology, stratigraphy, and paleopedology of these strata, including the White River Group and the lower Arikaree Group (summarized in [4]). However, the uppermost White River Group (Poleslide Member) and the lowermost Arikaree Group (Sharps Formation) in Badlands National Park have yet to be rigorously studied from lithostratigraphic and paleopedological perspectives (Figure 2 and Figure 3).

Multiple studies have identified soft tissue and cellular structures within vertebrate fossils that are morphologically consistent with those present in extant vertebrates (e.g., [6,7,8,9,10,11,12]). Further studies using immunoassays and Liquid Chromatography–Tandem Mass Spectrometry (LC-MS/MS) have found that soft tissue microstructures preserved in fossils often contain remnants of endogenous biomolecules (e.g., [13,14]). However, most prior studies have focused on Mesozoic or older vertebrate fossils (e.g., [6,7]) with few exceptions (e.g., [9,15,16]), while Cenozoic vertebrates remain relatively unexplored in terms of cellular and soft tissue preservation.

The environmental, taphonomic, and sedimentologic conditions which facilitate such remarkable preservation in fossil bone remain poorly characterized. Therefore, our goal was to evaluate paleoenvironmental influences on soft tissue preservation by conducting an integrative investigation into the lithology, vertebrate taphonomy, and paleoenvironmental conditions of the Oligocene Sharps Formation, which in turn would allow us to discern the potential controls on soft tissue and cellular preservation within vertebrate fossils from the White River Group and lower Arikaree Group. Results from Oligocene fossils collected in Badlands National Park are compared to those from older Oligocene (Brule Formation) and Eocene (Chadron Formation) fossils from strata in South Dakota, Nebraska, and Wyoming to identify and compare patterns of preservation in fossils from different geologic ages, depositional environments, and paleoclimatic regimes. All fossils used in this study were collected and analyzed as part of an earlier effort to determine the role(s) of rare earth elements (REE) in bone fossilization and the applicability of REE fingerprinting as a method to deter fossil poaching on federal lands (see [17,18,19]). Individual sites were chosen based on geologic age, lithology, and known frequency of poaching activities from personal observations and/or as informed by Badlands National Park staff.

2. Background

2.1. Cellular and Soft Tissue Preservation in Vertebrate Fossils

Over the last decade, there has been an increasing amount of research concerning the context and controls on soft tissue and cellular preservation in fossil vertebrates. Osteocytes and blood vessels have been recovered from a variety of skeletal elements, vertebrate taxa, and depositional environments (marine, freshwater, and terrestrial) from Mesozoic and Cenozoic fossils (e.g., [8,12,15,16]). Bone size also does not appear to be a major controlling factor, as endogenous proteins have been recovered from both large and small skeletal elements (e.g., [20,21]). Researchers have also begun to investigate mineralogical and diagenetic factors involved in soft tissue preservation (e.g., [22]). One study found that iron minerals played an important role in bone cell preservation, forming iron oxide pseudomorphs as a result of early diagenetic reactions [23]. It has also been suggested that soft tissues are more likely to preserve in oxidizing depositional environments [10]—such depositional environments are well-represented in the White River Group, including floodplain, aggradational eolian, stable eolian, lacustrine, forested, and open, lightly vegetated environments [4,24]. In contrast, soft tissues in fossil specimens have been recovered from reducing depositional environments [11,25], possibly stemming from rapid burial in iron-rich sediments.

Following death and significant decay of ultrastructural soft tissues (i.e., muscles and adipose tissues), vertebrate bones recrystallize to become carbonate hydroxy-fluorapatite. During the process of fossilization, the organic components of bone tissue (i.e., osteocytes, blood vessels, fibrous matrix) and the biomolecules comprising them may be damaged or entirely lost as a result of diagenesis. Bioapatite has a very low crystallinity (small crystallite size) during life, but during fossilization its crystallinity increases [26,27]. The extent of recrystallization and homogeneity of hydroxyapatite crystal size in fossil bone is determined by the Crystallinity Index, which can be used to investigate diagenetic effects on fossil bone [28]. An inverse relationship between the degree of crystallinity of a fossil bone and the amount of organic material preserved within it has been noted [28]. Similarly, a correlation between fossil bone apatite crystallinity and trace element uptake was noted [29], with the suggestion that fast recrystallization rates (as inferred from trace element concentration profiles) may facilitate long-term retention of endogenous microstructures and their constituent biomolecules.

2.2. Sedimentology of the White River Group

The lowermost units of the White River Group (Chamberlain Pass and Chadron Formations) primarily consist of extensive fluvial channel sandstones and floodplain paleosols altered by the humid late Eocene climate and expressed as colorful bands of strata in parts of the White River Badlands [30,31]. In contrast, the Oligocene Brule Formation is composed of brown and tan mudstones gradually giving way to tan siltstones higher up-section [4], indicating a transition from subhumid to subarid conditions (Figure 2). Volcanic materials sourced from Nevada and Utah were transported east and are found throughout the White River Group [24]. Lower in the section, volcanic material was mixed with alluvial sediments derived from the Black Hills to form the Chamberlain Pass Formation, up through the Scenic Member of the Brule Formation. Higher in the section, volcanic materials were deposited as dryland loess, as commonly seen in the thick siltstone layers of the upper Poleslide Member of the Brule Formation [4,24].

2.3. Stratigraphic and Paleopedological Context of the Sharps Formation

The Sharps Formation overlies the Upper Poleslide Member of the Brule Formation (Figure 3), which is the highest (currently) recognized unit within the White River Group, with a distinctive white layer (commonly referred to as the Rockyford Ash) separating the two rock units. While the Sharps Formation and Rockyford Ash are currently regarded as the lowest units of the Arikaree Group, the Sharps Formation and Upper Poleslide Member of the Brule Formation appear to be lithologically identical (Figure 2), with both units containing tan and brown, silty sandstone and sandy siltstone beds [4]. The similarity between the uppermost White River Group siltstone beds and the Sharps Formation led earlier researchers to conclude that the two are equivalent [32]. This distinctive white layer in the Cedar Pass area of Badlands National Park is not an ash bed, but instead a mixture of weathered volcanic materials and epiclastic debris that does not correlate with the true Rockyford Ash at the base of the Sharps Formation on Sheep Mountain Table near Rockyford, South Dakota [4]. We follow the nomenclature of [4] and refer to this distinctive bed as the Cedar Pass white layer and use it as a stratigraphic reference point throughout this study. Formal revision of the stratigraphic nomenclature of the Brule vs. Sharps Formations has yet to be published, and as such we refer to the strata above the Cedar Pass white layer as the Sharps Formation, although they are identical to strata within the underlying Poleslide Member of the Brule Formation.

In the Poleslide Member, massive siltstone beds are interpreted as volcaniclastic loess deposits, whereas sandstones represent stream deposits; the strata of the Sharps Formation are interpreted to have very similar origins based on their similar lithologies and overall geomorphic expression [4,30]. In terms of paleopedology, a shift is observed from forest soils (alfisols) at the base of the Brule Formation (which includes both the Scenic Member and the Poleslide Member) to grassland soils (mollisols), which formed within with fluvial, lacustrine, and eolian environments, each trading prominence throughout the Poleslide Member [4,30].

3. Methods

3.1. Field Collection

Vertebrate fossil specimens of large, postcranial bones (including tortoise carapace fragments) were collected in situ from each major depositional environment within the Sharps Formation (channel sandstones, stable floodplain environments, and aggradational eolian environments) in the Cedar Pass area of Badlands National Park, generating a total of eight specimens (Figure 4 and Figure 5). The stratigraphic position and preservational attributes (e.g., fracturing, color) of specimens were noted in order to place each specimen into temporal context and to interpret its taphonomic history. Seven fossil bones previously collected from channel and floodplain deposits within the underlying White River Group (Chadron and Brule formations) were also analyzed for comparison with the new fossils from the Sharps Formation (Figure 3).

3.2. Demineralization and Soft Tissue Analysis

Fragments of bone material were demineralized following established protocols [11]. Roughly 0.5 cm³ fragments of each specimen were demineralized in 0.22 µm filtered 0.5 M Ethylenediaminetetraacetic acid (EDTA) at pH 8 in polystyrene cell culture plates. Demineralization occurred over 2–6 weeks, with the EDTA solution replaced approximately every 48 h. Demineralization products were then drawn from the culture plates with sterile pipettes, placed on glass slides, and capped by a glass cover slip. Any isolated soft tissue or cellular structures were identified via standard (transmitted light) optical microscopy, with identifications based on morphologic consistency with counterpart cells/tissues from extant vertebrates (e.g., [11,15]). As negative controls, sediment samples from the four collection localities in the Sharps Formation (Rockyford Ash, lower siltstone bed, and upper channel sandstone) were also demineralized and examined following the same protocols used with the fossil specimens. Images of demineralization products (both from the fossils and sediment controls) were captured using a Nikon Eclipse microscope equipped with a Nikon DSFi-3 camera and NIS Elements software (Nikon, Tokyo, Japan).

Patterns of soft tissue and cellular preservation were compared between individual vertebrate fossils and cross-referenced with associated sedimentological data from strata lower in the White River Group [18,33,34,35,36,37,38]. This permitted a comprehensive analysis of the transition from late Eocene paleoenvironments (humid and forested fluvial channels and floodplains) to those of the early and middle Oligocene (open, eolian dominated savannah conditions) with respect to any potential influence of each paleoclimatic/paleoenvironmental regime on soft tissue and cellular preservation.

3.3. Thin Section Preparation and Microscopy

Following standard techniques (e.g., [39]), a fragment of each fossil was cut using a Buehler Isomet diamond trim saw, mounted on a glass slide with epoxy, polished with fine (600 µm) silicon carbide grit, and imaged in both cross-polarized and transmitted light microscopy. Features of interest included Haversian canals in the cortical bone based on comparisons to earlier work [40], as well as secondary mineral infills in cavities, which were interpreted based on crystal morphologies to infer associated paleoclimatic and diagenetic conditions documented in previous studies (e.g., [41,42]).

3.4. Mineralogy and Crystallinity Analysis

Mineralogy and crystallinity of fossil vertebrate specimens were determined by X-ray diffraction (XRD). Preparation of the fossils for XRD analysis followed existing techniques [29]. Specifically, a fragment of each fossil was cleaned with hand tools to remove excess sediment before being ground into a powder with a mortar and pestle. Each fossil powder was then filtered through a sieve (63 µm mesh) to acquire grains of uniform size. The fossil powders were analyzed using a Bruker D8 Advance powder X-ray diffraction system, fitted with a rotating stage and a copper Kα radiation tube set to 40 kV and 40 mA, scanning a 2θ range of 0° to 70° in a continuous mode. Each fossil was then assigned a hydroxyapatite Crystallinity Index (CI) value [28].

4. Results

4.1. Demineralization Products and Bone Apatite Crystallinity

All demineralized fossils yielded microstructures matching the morphology of vertebrate blood vessels and fibrous proteinaceous bone matrix, and all but four of the specimens preserved structures consistent in morphology with vertebrate osteocytes (

Table 1. Summary of soft tissue and cellular microstructures recovered from demineralization of each fossil, including crystallinity index (CI) values for each specimen [28]. Abbreviation: indet.: indeterminate.

Specimen	Taxon	Age	Element	Lithology	Osteocytes	Blood Vessels	Fibrous Matrix	CI Value
F19-50	Tortoise	Oligocene	Shell	Channel Sandstone	Absent	Uncommon	Rare	0.42
F19-49	Mammal	Oligocene	Indet. postcranial	Channel Sandstone	Rare	Frequent	Rare	0.42
F19-28	Mammal	Oligocene	Indet. Rib	Channel Sandstone	Rare	Uncommon	Rare	0.42
F19-27	Tortoise	Oligocene	Shell	Channel Sandstone	Abundant	Uncommon	Rare	0.43
F19-24	Tortoise	Oligocene	Shell	Volcanic Ash	Abundant	Frequent	Rare	0.50
F19-19	Carnivore	Oligocene	Ulna	Eolian Siltstone	Absent	Uncommon	Rare	0.46
F19-18	Oreodont	Oligocene	Rib	Eolian Siltstone	Rare	Uncommon	Rare	0.51
F19-10	Mammal	Oligocene	Indet. rib	Eolian Siltstone	Abundant	Abundant	Rare	0.40
F19-9	Tortoise	Oligocene	Shell	Volcanic Ash	Rare	Rare	Rare	0.43
F19-8	Tortoise	Oligocene	Shell	Volcanic Ash	Uncommon	Uncommon	Frequent	0.53
F19-7	Tortoise	Oligocene	Shell	Eolian Siltstone	Rare	Abundant	Absent	0.48
F19-6	Tortoise	Oligocene	Shell	Volcanic Ash	Abundant	Uncommon	Rare	0.60
F19-3	Tortoise	Oligocene	Shell	Eolian Siltstone	Absent	Abundant	Absent	0.52
F19-1	Mammal	Oligocene	Indet. limb Bone	Eolian Siltstone	Absent	Absent	Absent	0.53
F101	Tortoise	Oligocene	Shell	Eolian Siltstone	Abundant	Frequent	Abundant	0.36
F47	Tortoise	Oligocene	Shell	Floodplain Mudstone	Abundant	Frequent	Frequent	0.28
F12-70	Tortoise	Eocene	Shell	Floodplain Mudstone	Abundant	Uncommon	Uncommon	0.44
F15	Mammal	Eocene	Indet. Limb	Floodplain Mudstone	Abundant	Uncommon	Rare	0.32
FR09-09	Mammal	Eocene	Indet. Rib	Floodplain Mudstone	Rare	Rare	Uncommon	0.46
F08-87	Tortoise	Eocene	Shell	Channel Sandstone	Absent	Frequent	Rare	0.44
F08-80	Brontothere	Eocene	Rib	Floodplain Mudstone	Abundant	Abundant	Frequent	0.61
F08-77	Oreodont	Eocene	Stylopodial	Floodplain Mudstone	Absent	Uncommon	Rare	0.65
F08-10	Brontothere	Eocene	Rib	Floodplain Mudstone	Frequent	Abundant	Uncommon	0.43
F08-09	Brontothere	Eocene	Femur	Floodplain Mudstone	Abundant	Abundant	Frequent	0.47
F08-08	Brontothere	Eocene	Indet. Metapodial	Floodplain Mudstone	Uncommon	Abundant	Frequent	0.37
F08-915-R	Brontothere	Eocene	Scapula	Floodplain Mudstone	Rare	Abundant	Uncommon	0.47

). These microstructures were recovered from all fossils regardless of geologic age, skeletal element, host lithology, and source taxon. Of the four fossils that did not yield osteocytes, two were tortoises and two were mammals, two were Oligocene in age while two were from late Eocene strata, two were from fluvial channel sandstones, and one each from a floodplain mudstone and eolian siltstone. Fragments of extracellular matrix were identified by their fibrous texture (Figure 6A–C), while blood vessel fragments were recognized by their tubular, branching forms and hollowness (Figure 6D–G), and osteocytes were identifiable either individually as small (generally no more than 20 µm in length), prolate structures with thin, branching extensions consistent in form with cellular filopodia (Figure 6H,I), or in groups clustered together within partially demineralized bone matrix (Figure 6J–L). Demineralization of associated sediment samples (Figure 7) yielded small masses of argillaceous material and silt-size quartz grains but no identifiable organic microstructures.

Apatite crystallinity was also examined to assess the relationship between diagenetic recrystallization of bioapatite and extent of soft tissue preservation (Table 1). CI values were found to vary from 0.28 to 0.65, with no clear pattern in relation to the recovery of any of the three microstructures examined (Table 1). There is no significant difference between Eocene and Oligocene fossils. Although the sample size was small, there was no significant difference in CI values between fossils with and without osteocytes. However, calcite was absent in most fossils in which osteocytes were also absent.

4.2. Bone and Secondary Mineral Micromorphology and Diagenetic History

Thin sections of many of the fossil specimens revealed well-preserved microstructural preservation of extensive Haversian canal systems, especially in the external cortex (Figure 8A–C), with primary osteons, secondary osteons, and individual osteocyte lacunae each still clearly visible. Several specimens contained elongate or circular features representing resorption cavities (e.g., Figure 8D–G). Virtually all such cavities were infilled with calcite, which was primarily sparry in form (Figure 8D–F). Specimen F19-06 contained iron oxide as well, usually clustered around the edges of cavities, with sparry calcite forming a second-generation infilling cement that occupied the rest of the space (Figure 8E,F). In this specimen, iron oxide tended to most often be found on one side of larger cavities, presumably representing “spirit-level” secondary mineral deposits formed on the quarry-down side within these larger spaces (cf., [43]; Figure 8F), but in smaller cavities it was typically found to coat the entire perimeter (Figure 8E). Some fossils (F47, F101, and F19-28) possess thin micritic calcite linings on the walls of cavities which are then infilled by sparry calcite. Specimens F19-19 and F19-50 generally exhibit poorly preserved bone texture, with fine histologic structures (i.e., osteons, lines of arrested growth) within the cortex largely altered via microbial attack/bioerosion and apparent recrystallization (cf., [44]; Figure 8G–I).

5. Discussion

5.1. Trends in Soft Tissue Preservation

While all fossil specimens examined were found to retain organic microstructures (vessels and fibrous matrix fragments), considerable variation was encountered in terms of the abundance and quality of osteocyte preservation. Fossils with poorly preserved histology in the thin section (e.g., Figure 8H,I) did not yield any identifiable osteocytes after demineralization, indicating that the apparent pattern in osteocyte preservation in Table 1 corresponds to the extent of diagenetic alteration. Intriguingly, this finding parallels a recent trend identified in fossil bones preserved in shallow marine environments [25], indicating that there are commonalities to bone decay among disparate depositional environments which, in turn, impart similar outcomes in terms of cellular and soft tissue preservation.

The most surprising finding in this study was the great prevalence of soft tissue and cellular preservation within the specimens, particularly of vessels and fibrous matrix fragments (Table 1). Recovery of these microstructures across all taxa, depositional environments, skeletal elements, and geologic ages supports the interpretation that organic microstructure preservation may be independent of these factors (e.g., [9,12,15,16,45]. However, the fact that endogenous microstructures were recovered from all fossils, regardless of taxonomy, host lithology and depositional environment, remains unexpected given conventional wisdom and historical hypotheses. In contrast, recent research suggests that the type of fossil bone (compact vs. cancellous) preserves different apatite crystallinity values that may correlate to the type of soft tissue preserved, that taxonomy shows a correlation with the type of soft tissue that is preserved, and that biomechanical function shows no correlation with soft tissue preservation [46,47].

Regardless, when placed in context with other recent studies (e.g., [12,25]), our results suggest that preservation of soft tissue and cellular microstructures may be more common in fossil bones than previously thought.

5.2. Bone Histology and Post-Depositional Conditions

Of the three fossil specimens that exhibited poorly preserved bone histology (F08-77, F19-19, and F19-50), all were from different depositional environments and ages. The only shared trait was their lack of identifiable osteocytes in the thin section. Hence, the poor histological and cellular preservation observed in these three fossils is more likely a result of individual diagenetic alteration rather than shared climatic or environmental conditions at the time of fossilization.

In all fossils with calcite infills (F47, F101, F19-06, F19-18, F19-28, and F19-50), equant and blocky morphologies were dominant (e.g., Figure 8D–F). The prevalence of sparry, diagenetic calcite is notable, especially in the Oligocene fossils collected from eolian siltstones (e.g., F19-18), since calcite of this form is indicative of saturated or phreatic diagenetic conditions [41,42]. The sparry calcite cavity infills are likely the result of post-depositional phreatic saturation since their depositional environments have been interpreted as dry [4,30,33]. The Eocene fossils, in contrast, either lacked any clear cavity infill or exhibited micrite instead of more crystalline, sparry calcite (e.g., F08-87). Micritic calcite is indicative of vadose conditions [41,42]; the presence of micrite lining the cavity walls around the centrally located sparry calcite in some fossils (F47, F101, and F19-28) implies a relatively dry, unsaturated setting followed by phreatic saturation after burial. The presence of iron oxide in F19-06 suggests oxidizing conditions, though this too may be reflective of diagenetic redox conditions instead of those of the initial depositional environment. The apparent lack of iron in the cells, vessels, and fibrous matrix fragments contrasts with the findings of previous studies (e.g., [7,9,11]), which identified a connection between iron-rich sediments and endogenous microstructure preservation. Thus, cellular and soft tissue microstructure preservation may not always be reliant on diagenetic iron mineralization, as some authors have suggested (e.g., [23]).

6. Conclusions

This study is among the first to examine the influences of a broad range of factors on soft tissue and cellular preservation in fossil bones, including geologic age, depositional environment, and paleoclimatic conditions. Neither taxonomic identity, host lithology, nor paleoclimatic regime (e.g., warm and humid Eocene vs. cooler and drier Oligocene) were found to influence soft tissue or cellular recovery. These findings, in combination with the observed prevalence and morphologic integrity of the tissue and cellular structures documented in this study, suggest that these endogenous microstructures may not be as rare in fossil bones as historically thought.

Future studies on this topic are clearly warranted, with special attention toward diagenetic factors, such as metamorphic history, chemical alteration, and trace element uptake. Incorporating additional, complementary methods of analysis such as cathodoluminescence and scanning electron microscopy in conjunction with energy dispersive X-ray spectroscopy could also be used to corroborate and refine observations and interpretations from thin-section petrography and optical microscopy, thereby providing a more complete picture of post-depositional conditions and diagenetic history. Multidisciplinary studies of fossil specimens from diverse depositional environments offer the most informative path forward to discerning the relative importance of large-scale/external (i.e., climatic, geologic) versus small-scale/internal (i.e., redox, pH) conditions as controls on soft tissue and biomolecular preservation in the fossil record.