Introduction

Burrowed firmground discontinuity surfaces mark depositional breaks (hiatuses) in the Phanerozoic marine sedimentary record. They range from major, interformational surfaces reflecting eustatic sea level oscillations to minor intraformational discontinuities related to periods of increased bottom-current intensity1. The origin of major surfaces was typically related to sea-level lowstands resulting in erosional exhumation of semi-consolidated sediment. This allowed for colonization of exposed firmgrounds by burrowing invertebrates, typically decapod crustaceans, which left distinctive trace fossils of the Glossifungites ichnofacies1,2,3,4. At the later stages of development, the firmground surfaces could often be transformed into fully lithified hardgrounds where bioerosion dominated5,6,7, but these phenomena are not covered by the present work.

Occasionally, heavy bioturbation led to disintegration of the firmground surface and transformed it into a ‘pseudobreccia’ composed of shreds of the older sediment floating in the younger sediment piped-down by burrowers. This phenomenon was first described, from a Santonian phosphatic chalk succession at Taplow in Great Britain, by the renowned ichnologist Bromley5 (his Fig. 18.3A; for details of the section see Jarvis8). A similar process of massive subsurface replacement of an underlying deposit by younger sediment through repetitive burrow excavation and storm infilling was inferred from Holocene settings9.

The burrowed discontinuities are often associated with transgressive lag- or hiatal fossil concentrations3,10,11. These are notoriously difficult to interpret due to their complex anatomy and varied biostratigraphical and taphonomic signatures of fossils10,11. This is of particular importance for units formed during periods of major biotic turnovers, like the Cretaceous–Paleogene (K–Pg) mass extinction. For example, a ‘bonebed’ at the burrowed K–Pg junction in the Edelman Fossil Park of the Rowan University (New Jersey, USA; formerly known as Inversand pit) yields a variety of invertebrate and vertebrate fossils12,13,14,15. These are essential for understanding the K–Pg faunal turnover in North American shallow-marine settings, but their meaning is equivocal due to disputes about origin and age of this bed16,17,18,19.

The burrow-generated pseudobreccia, if present, would certainly contribute to the interpretational problems related to fossil-rich deposits associated with hiatal burrowed surfaces, but this issue has remained unexploited to date. Here, we study a richly fossiliferous hiatal K–Pg boundary interval in Poland to test the influence of the pseudobreccia on the preservation of associated fossil concentrations. Our research includes X-ray computed tomography (CT) imaging of the K–Pg interval to showcase in detail its anatomy and fossil-sediment relationships (see “Material and methods” section).

Geological setting

We have studied a shallow-marine, epicontinental Maastrichtian–Danian boundary succession exposed at the villages Bochotnica and Nasiłów near the town Kazimierz Dolny in the northern part of the Middle Vistula River section, Poland (Fig. 1a; Fig. S1a, b; for details see “Material and methods” section and SI). No formal lithostratigraphical scheme for the study succession has been proposed. Therefore, we have used the traditional, informal lithostratigraphical units, that is, the Kazimierz Opoka, the Greensand, and the Siwak20,21.

Figure 1
figure 1

The K–Pg boundary interval studied. (a) The section with chrono- and lithostratigraphy (informal), and location of the study area in Poland. (b–e) Boundary details in outcrops (b, d—Bochotnica 1, c, e—Nasiłów; the highest position of the limestone shreds marked by white arrow in d). (f) Cross-section of laminated and burrowed infill of a burrow, Nasiłów. (g, h) Scanning Electron Microscope (SEM) images of the soft (g) and hard (h) limestone at the top of Kazimierz Opoka, Bochotnica 1; voids after dissolved spicules in h are largely obliterated by calcitic cement. Figure created using CorelDRAW 2018 (64-Bit) and Adobe Photoshop CS3.

Full size image

The studied succession starts with the upper, but not uppermost Maastrichtian Kazimierz Opoka (Fig. 1a). ‘Opoka’ is a siliceous limestone typified by voids after dissolved sponge spicules and a rigid framework of opal-CT lepispheres22. The top of Kazimierz Opoka is riddled with decapod and ‘worm’ burrows filled with sandy-glauconitic sediment (Fig. 1b–f). The burrowed interval is composed of a hard limestone zone passing upward into a brecciated zone of soft limestone (Fig. 1a–h). The brecciated top of the Kazimierz Opoka has been recognized as a pseudobreccia formed by superposition of successive generations of omission and post-omission burrows at the top of a firmground discontinuity surface20,21,23.

The burrows at the top of Kazimierz Opoka belong predominantly to the ethological category Domichnia24. They run in various directions, appearing in a variety of cross sections on rock surfaces (Fig. 1b–e). Additionally, they are commonly obliterated by subsequent burrow generations which further makes ichnotaxonomic identifications difficult. Despite these problems, the following ichnotaxa have been identified: ‘Thalassinoides saxonicus’ (sensu Kennedy25; see Niebuhr and Wilmsen26 for description of Ophiomorpha saxonica), Th. suevicus, Th. isp., Spongeliomorpha isp. cf. sudolica, Gyrolithes isp., and Chondrites isp. This ichno-assemblage contains both the omission and early post-omission suites of burrows (sensu Bromley5). The pre-omission suite present below the top of the Kazimierz Opoka includes Planolites, Palaeophycus, and the fish-scale lined burrows Lepidenteron. The late post-omission suite in the overlying Greensand is represented by Schaubcylindrichnus isp.21. The burrow outlines reveal no deformation, and faint lamination may be seen in some of their infillings (Fig. 1f). These were therefore open burrows excavated in a semi-cohesive sediment, forming a typical example of the Glossifungites ichnofacies1,2,3,4.

Although the topmost part of the Kazimierz Opoka was commonly referred to as the hardground or the incipient hardground27,28,29, the cross-cuting of successive burrow generations filled with sandy-glauconitic sediment, demonstrates that the cementation of this bed postdated the commencement of the deposition of the overlying glauconitic sand20. Therefore, this is not a hardground in the genetic sense but a firmground discontinuity surface of the interformational type (as defined by Savrda1).

The Kazimierz Opoka is followed, with a break, by a horizon of marly glauconitic sandstone of early, but not earliest Danian age, traditionally referred to as the ‘glauconitic sandstone’ or the Greensand (Fig. 1a, Fig. S2). This fossil-rich deposit has attracted attention for nearly a century, arousing controversies over its extent and age (Fig. S2) due to its gradational upper boundary and mixing of Maastrichtian and Danian fossils20,21,27,28,29,30,31,32. The fossils (often phosphatized) and limestone phosphoclasts form a more or less distinctive tabular concentration, the Phosphorite Layer (Fig. 1a–c), but are also present beneath that level. Both the facies and fossil signatures point to the condensed nature of the glauconitic horizon, which is regarded as the basal, transgressive portion of the Danian sequence21. Higher up, the sandstone passes into fossil-poor Danian gaizes and limestones, the Siwak (Fig. 1a,b; Fig. S2). A partial skeleton of the marine crocodile Thoracosurus was recovered from a lowermost part of Siwak, just above the Phosphorite Layer33.

In terms of the standard dinoflagellate stratigraphy34,35, the soft limestone at the top of Kazimierz Opoka represents the basal part of the Thalassiphora pelagica subzone of the upper Maastrichtian Palynodinium grallator Zone, the Greensand represents the lower, but not lowermost, part of the Senoniasphaera inornata subzone of the lower Danian Danea mutabilis Zone, and the overlying Siwak corresponds to the higher part of the S. inornata subzone (Johan Vellekoop, email communication to M.M., dated 19.10.2023).

In the study area, the burrowed discontinuity surface separates the upper Maastrichtian and lower Danian strata. However, it is traceable across the Lublin Upland as far east as to Lechówka site (Fig. S1a) where it overlies the lowermost Danian strata, including the K–Pg boundary clay with geochemical proxies of the Chixculub impact; this implies the early Danian age of the discontinuity surface studied here21,36,37.

Results

Pseudobreccia anatomy

For detailed examination of the anatomy and fossil-sediment relationships in the K–Pg interval, we have investigated a 19 × 18 × 15 cm rock sample (Fig. 2a–e) from the outcrop at Bochotnica 1 (Fig. S1b). The slab encompasses the pseudobreccia zone and the overlying Phosphorite Layer. An irregular ‘garland’ and shreds of soft limestone are preserved in a sandstone matrix (Fig. 2a, b). The slab contains phosphoclasts and abundant fossils, mainly disarticulated valves of the scallop Dhondtichlamys acuteplicata (Fig. 2a–e). A belemnite rostrum, probably a subadult individual of Belemnella kazimiroviensis, crosses the limestone–sandstone interface (Fig. 2a). The polished side surface of the sample reveals a thoroughly bioturbated fabric, with better visible last-generation burrows (Fig. 2b). Scallop shells commonly protrude from the soft limestone into sandstone, often bearing soft limestone remnants on their surfaces (Fig. 2b, d, e). Many scallop valves floating in sandstone display soft limestone vestiges as well (Fig. 2c). Fossil orientation is random, except for stacking of the convex-up scallop valves in the upper part of the sample (Fig. 2c).

Figure 2
figure 2

Anatomy of the pseudobreccia in the slab from Bochotnica 1. (a, b) Gross views, the top indicated by arrows; dashed line in b indicates a postomission burrow. (c–e) Close-up views of selected fragments to show fossil-sediment relationships. Figure created using CorelDRAW 2018 (64-Bit) and Adobe Photoshop CS3.

Full size image

The X-ray computed tomography imaging of the sample has revealed the fossil-sediment relationships in three-dimensional (3D) view (Fig. 3a–i). In 3D images, the ‘garland’ turns out to be a cross section of a soft limestone protrusion rooted in the underlying substrate of the same lithology, similar to the bosses or pinnacles visible in outcrops (Fig. 1e). Irregular pieces of soft limestone in a sandstone matrix (Fig. 3b) correspond to the soft limestone ‘spots’ discernible in 2D views (Figs. 1d, 3a). The upper portion of the rock sample is devoid of soft limestone shreds, containing only fossils and phosphoclasts of the Phosphorite Layer (Fig. 3b).

Figure 3
figure 3

Three-dimensional pseudobreccia anatomy and sediment-fossil relationships in the slab presented in Fig. 2. (a, d, g) Natural views of the slab. (b, c, e, f, h, i) X-ray CT imaging of the slab (soft limestone signal removed in c, f, i). Abbreviations explained in a. Figure created using CorelDRAW 2018 (64-Bit) and Adobe Photoshop CS3.

Full size image

Fossil-sediment relationships

The molds after dissolved aragonitic shells are not recorded in our CT images, which therefore provide data only on the relationships between calcitic fossils and the surrounding sediments (Fig. 3b, c, e, f, h, i). These fossils are all identified as late Maastrichtian taxa; the Danian biota preserved in the Greensand are dominated by casts of minute, originally aragonitic bivalves and gastropods27,38.

The CT images demonstrate that protrusion of shells from the limestone into sandstone was a common phenomenon. When the soft limestone signal is removed from some images (Fig. 3c, f, i), the fossil concentration appears bipartite, with the lower level formed by fossils associated with the pseudobreccia zone, and the upper level composed of phosphates and fossils belonging to the Phosphorite Layer (compare Fig. 3b, c).

Hand-specimens further highlight the relationships between fossils, the pseudobreccia, and burrows (Fig. 4a, f–o). The pseudobreccia fossils are very well preseved, as exemplified by: (1) the original color-pattern preserved on the scallop shell (Fig. 4h), (2) superbly preserved echinoid test (Fig. 4j), (3) growth rings on brachiopod shells (Fig. 4m), and (4) delicate bryozoan etchings preserved on the otherwise pristine surface of the belemnite rostrum (Fig. 4o). There are no preservational differences between the sandstone- and limestone embedded parts of individual specimens. In contrast to the pseudobreccia assemblage, the Maastrichtian specimens from the Phosphorite Layer are typically worn, bored, fragmented and phosphatized (Fig. 4b–e).

Figure 4
figure 4

The preservation of late Maastrichtian fossils in the studied K–Pg interval. (a) Stratigraphy: 1, 2—informal lithostratigraphy (1—traditional, 2—proposed), 3—chronostratigraphy, 4—section log; PLa—Phosphorite Layer assemblage, Pa—pseudobreccia assemblage. (b) Phosphatic mold of the scallop Dhondtichlamys acuteplicata. (c) Phosphatized and burrowed infill of the oyster Pycnodonte vesicularis. (d) Glauconitized shell of the brachiopod Neoliothyrina obesa. (e) Worn and bored fragment of a belemnite. (f) Cast of D. acuteplicata with glauconite-filled burrow. (g) Shell of P. vesicularis with burrowed infilling. (h) Valve of D. acuteplicata. (i) Brachiopod in a soft limestone shred. (j) Echinoid Pleurosalenia bonissenti. (k) Isolated valve of the bivalve Limatula sp. in a soft limestone shred. (l) External cast of the bivalve Mutiella sp. with a sandstone-filled burrow. (m) Brachiopods Carneithyris sp. protruding from the hard limestone into the interior of a burrow filled with glauconitic sandstone. (n) Belemnite Belemnella kazimiroviensis straddling the limestone–sandstone interface (the same specimen in Figs. 2 and 3). (o) Bryozoan etching on the belemnite surface. Specimens in (b)—(e) belong to Phosphorite Layer assemblage; specimens (f)—(o) to the Pseudobreccia assemblage. For provenance and repository of the specimens see Material and methods. Abbreviations: K. Op.—Kazimierz Opoka, S.—Siwak, b—burrows, cbor—clionaid sponge borings, gsb—glauconitic sandstone-filled burrowes, betch—bryozoan etchings, hl—hard limestone, sl—soft limestone, gs—glauconitic sandstone. Figure created using CorelDRAW 2018 (64-Bit) and Adobe Photoshop CS3.

Full size image

Discussion

Based on Machalski et al.21and the present observations, the succession of events recorded at the top of the Kazimierz Opoka may be reconstructed as follows: (1) erosional removal of uppermost Maastrichtian deposits during a sea level drop in the early Danian; (2) sedimentary omission and colonization of the exhumed carbonate-siliceous ooze by burrowers; (3) commencement of sedimentation of glauconitic sand (initially intermittent as indicated by laminated burrow infillings, Fig. 1f); (4) multi-event erosion leading to the concentration of fossils and phosphoclasts of various derivation into the Phosphorite Layer. Persistent bioturbation through the omission and postomission stages led to the formation of pseudobreccia at the top of the Kazimierz Opoka and vertical translocation of many components of the Phosphorite Layer.

Soft limestone shreds extend up to the level of Phosphorite Layer (Fig. 1d), indicating the actual position of the originally discrete discontinuity surface. The glauconitic sand beneath this level is a secondary deposit that was transported down by burrowing animals, which gradually replaced the Maastrichtian ooze in the way described by Bromley5, and Tedesco and Wanless9. The pseudobreccia zone was assigned to the ‘glauconitic sandstone’ or Greensand by several workers27,29,30,39,40, but in our view it actually represents the top of the Kazimierz Opoka. This paves the way for interpretation of the Phosphorite Layer, traditionally regarded as the ‘upper Greensand’, as a basal lag or conglomerate of the Danian transgression. Due to the hybrid nature of the Greensand, its distinction as a separate unit is misleading and we propose to abandon it (Fig. 4a).

The common occurrence of Maastrichtian fossils which intersect the limestone-sandstone interfaces, or float in sandstone with attached limestone remnants (Figs. 2, 3 and 4), requires an explanation. The bioturbated fabric and close connection of the fossils with burrows (e.g., Fig. 2d, 4f, g, l) suggest a peculiar process of fossil preparation by burrowers during the formation of the pseudobreccia. Successive generations of burrowers avoided hard objects when excavating the Maastrichtian ooze, leaving them exposed in burrows, which were subsequently filled with sand. Repetition of this process led to a gradual subsurface replacement of the host sediment of these fossils by new sediment. The ‘bio-preparation’ must have been a delicate process as we have not detected bioglyphs on the pseudobreccia fossils (scratches are occasionally present on the burrow walls below, forming a delicate pattern similar to that of Spongeliomorpha sudolica41,42). We have failed to find reports on similar processes, except for the paper by Kennedy25 (p. 147), who noted that calcitic fossils occasionally protruded into burrows in some Cretaceous hardgrounds. In recent settings, the burrowing activity typically leads to a patchy concentration of shells in the sediment43.

The ichno-taphonomic process inferred above explains the co-occurrence of worn and excellently preserved Maastrichtian fossils in the Greensand. Several workers have drawn attention to records of belemnite rostra in excellent preservation from this horizon next to specimens that were heavily damaged by erosion or chemical corrosion28,30,39,40. The occurrence of intact Maastrichtian fossils was even interpreted as evidence for the late Maastrichtian age of the glauconitic sandstone30,39. Our view is that there are two distinctive fossil assemblages in the interval studied, reflecting different taphonomic pathways. These are: (1) the lower assemblage linked to the pseudobreccia zone, termed here the Pseudobreccia assemblage (Pa in Fig. 4a; Fig. 4f–o), composed of well-preserved Maastrichtian fossils, partially or entirely freed from their host sediment by burrowers, (2) the upper assemblage linked to the Phosphorite Layer, the Phosphorite Layer assemblage (PLa in Fig. 4a; Fig. 4b–e), composed of derived Maastrichtian and broadly indigenous Danian fossils. The pseudobreccia is well resolvable in the sections studied on account of the clear contrast between the limestone and sandstone. However, if the contrasts were weaker or non-existent, the correct identification of the pseudobreccia and its fossil assemblage would be difficult or impossible (compare Fig. 3b, e, h with Fig. 3c, f, i).

Implications

A phenomenon referred here to as the burrow-generated pseudobreccia is not uncommon in the Phanerozoic marine sedimentary record, occurring often in association with fossil concentrations3,44. Amongst the K–Pg boundary sites, the burrow-generated pseudobreccia was explicitly identified in Alabama45 and implicitly at a site in New Jersey by Landman et al.46 (their Fig. 1). However, the pseudobreccia escaped attention of researchers of the Edelman Fossil Park succession, although it is visible at the burrowed top of the Maastrichtian Navesink Formation in field photos provided in Wiest et al.18 (their Fig. 3) and Voegele et al.19 (their suppl. Figure 1), which in consequence has been erroneusly interpreted as the lower part of the overlying Hornerstown Formation by these authors (Fig. 5). This observational failure significantly contributed, in our view, to the confusion surrounding the fossil concentration at this site.

Figure 5
figure 5

Proposed stratigraphical correlation and genetic interpretation of studied K–Pg interval at Nasiłów and Bochotnica in Poland (left) with the corresponding interval exposed in the Edelman Fossil Park, New Jersey, USA (right, based on Wiest et al.18, their Fig. 3 and Voegele et al.19, their suppl. Figure 1). Abbreviations: 1—traditional lithostratigraphy (informal in Poland, formal in the New Jersey section); 2—proposed new lithostratigraphy; 3—proposed chronostratigraphy; 4—section log; Pa—Pseudobreccia assemblage; PLa—Phosphorite Layer assemblage; MFLa—Main Fossiliferous Layer assemblage; OLa—Oyster Layer assemblage; a—burrows; b—soft sediment shreds; c—invertebrate fossils; d—isolated vertebrate fossils; e—articulated vertebrate fossils; f—phosphoclasts and phosphatized fossils. Figure created using CorelDRAW 2018 (64-Bit).

Full size image

The Edelman Fossil Park succession reveals striking similarities to the K–Pg interval studied here (Fig. 5), including the presence of pseudobreccia and bipartite nature of the fossil concentration, recently documented by Voegele et al.19. More specifically, we propose that the lower fossil assemblage at the Edelman Fossil Park site, that is the Oyster Layer assemblage19 should be interpreted as a genetical counterpart of the late Maastrichtian Pseudobreccia assemblage (Pa) from the Middle Vistula River section (Fig. 5). Consequently, the upper fossil assemblage at the Edelman Fossil Park, that is the Main Fossiliferous Layer (MFL) as defined by Voegele et al.19, is best interpreted as the basal Danian transgressive lag or conglomerate, in analogy with the Phosphorite Layer assemblage (PLa) from Poland. The MFL and Pa assemblages share a broad array of taphonomic pathways, being composed of fossils with varied taphonomic and age signatures12,13,16. Even more specifically, in both cases the articulated remains of post-Cretaceous marine vertebrates (Fig. 5) seem to occur not within the transgressive lag itself but just at its top16,33. Obviously, our interpretations require future testing by field-work at the Edelman Fossil Park site along the lines suggested by the present study. If correct, they may bring new insights into our understanding of the marine fossil record across the K–Pg boundary in North America.

Conclusion

Our study demonstrates that the recognition of pseudobreccias is a prerequisite for understanding the associated hiatal fossil concentrations, allowing us to reconsider the studied K–Pg interval in Poland. It also documents a hitherto unreported ichno-taphonomic process of subsurface preparation of fossils by burrowers during the pseudobreccia formation. Future recognition of similar features in genetically-related units elsewhere should prove valuable for their interpretation, with particular significance for intervals formed during major turnovers in the history of life, like the end-Cretaceous mass extinction.

Material and methods

Localities

The material studied was collected at three outcrops located near the town Kazimierz Dolny (Fig. S1b). Locality Nasiłów is located on the left bank of the Vistula River, opposite of Kazimierz Dolny. This is a large abandoned quarry (51° 20′ 38.126″ N, 21° 57′ 34.749″ E) on the Vistula escarpment east of the village Nasiłów. The other locality, Bochotnica 1, is located on the right bank of the Vistula, north of Kazimierz Dolny. This is an abandoned rural quarry behind the old watermill (51° 20′ 18.953″ N, 22° 0′ 13.785″ E) in the eastern part of the village Bochotnica. Part of the material comes from another, now unaccessible rural quarry on the Vistula escarpment in the northern part of the village, marked here as Bochotnica 2 (Fig. S1b). See Machalski et al.21 for detailed description of Nasiłów and Bochotnica 1 localities.

Provenance and repository of material

Our materials consist of lithological samples and paleontological specimens from Bochotnica 1 and 2 and Nasiłów outcrops (see above) housed at the Institute of Paleobiology, Polish Academy of Sciences (IP PAS, collection acronym ZPAL). A single echinoid specimen (Fig. 4j) comes from the Museum of the Earth, PAS (collection acronym MZ); both institutions in Warsaw, Poland. The provenance and repository of samples and specimens illustrated in the present paper is as follows: Fig. 1f—Nasiłów, ZPAL. 82/1; Fig. 2 and Fig. 3—Bochotnica 1, ZPAL V.82/11; Fig. 4b—Nasiłów, ZPAL V.82/8; Fig. 4c—Nasiłów, ZPAL L.18/118, Fig. 4d—Bochotnica 2, ZPAL V.82/7; Fig. 4e—Nasiłów, ZPAL V.82/9, Fig. 4f—Bochotnica 2, ZPAL V.82/5, Fig. 4g—Nasiłów, ZPAL L.18/115; Fig. 4h—Nasiłów, ZPAL V.82/10; Fig. 4i—Bochotnica 2, ZPAL V.82/2; Fig. 4j—Bochotnica, unspecified outcrop, MZ VIII Ee 836; Fig. 4k—Bochotnica 2, ZPAL V.82/6; Fig. 4l—Nasiłów, ZPAL V.82/3; Fig. 4m—Nasiłów, ZPAL V.82/4; Fig. 4n, o—Bochotnica 1, ZPAL V.82/11.

Preparation of macrosamples

Lithological samples and paleontological specimens were prepared, both in the field and laboratory, using standard mechanical methods: hammer, chisel, needles, and a vibrotool (Paleotools, ME-9100).

Specimen photography

The specimens were photographed using a Canon EOS 450 D camera with lenses: Canon Macro Lens EF 100 mm 1:2.8, Canon Compact-Macro Lens EF 50 mm 1:2.5 and Macro Conversion Lens Raynox DCR-150.

Scanning electron miscroscope (SEM)

For SEM studies, rock samples from the top of the Kazimierz Opoka were placed on stubs with double-sided adhesive tape and sputter-coated with a conductive carbon film. Analyses were conducted at the Institute of Paleobiology PAS, using a Philips XL20 scanning electron microscope. The instrument was operated at an acceleration voltage of 25 kV, a beam current of 98–103 nA and a spot diameter of 3.5 μm.

Computed tomography (CT) procedures

In order to obtain the 3D imaging of the K–Pg contact zone (Figs. 2 and 3), a representative rock slab from Bochotnica 1 outcrop was collected and scanned using the GE PHOENIX v|tom|x s system (General Electric Sensing and Inspection Technologies/Phoenix X-ray, Wunstorf, Germany) at the Computed Microtomography Laboratory of the University of Silesia in Katowice, Poland. The system consists of three key components: an X-ray source, a turntable and an X-ray detector panel (2024 × 2024 pixels). The X-ray source produces a cone-shaped X-ray beam by bombarding a metal target with electrons, generated by passing a high-energy electric current through a tungsten filament. The cone is focused onto a sample, which is mounted on the turntable and rotated through 360°. During the rotation the detector panel collects a series of two-dimensional projections (radiographs) usually at the intervals between 0.1° to 0.05°. The projections measure the amount of X-ray energy transmitted by the sample. In order to produce a radiograph with high contrast and brightness, the X-ray beam must adequately penetrate the sample without over-exposing the panel. Hence, the energy of the X-rays (determined by the voltage and current) has to be carefully selected by the CT user. Penetration of the beam (and projection contrast) is largely determined by the voltage, whilst the number of X-rays (and projection brightness) is mainly determined by the current. The rock-sample from Bochotnica 1 locality was scanned with an X ray beam set at 200 kV and 350 µA.

After scanning, a computer has been used to line up and centre the projections. Each row of pixels in the detector panel will become a slice, so a cross-section of the examined sample is created from each line of pixels in the radiograph using a method called the back-projection. Essentially, the projections for each slice are converted to digital profiles then smeared across each other to create digital images of the corresponding CT slices. Each slice is made up of voxels, i.e., three-dimensional pixels. A CT reconstruction (or volume) is thus essentially a matrix. Each voxel is assigned a CT number and grey value. Thus a CT scan is akin to a 3D extension of a greyscale digital photograph, but based on X-ray transmission. The size of the voxels is determined by the magnification of the projected sample image on the panel. The turntable can be moved to varying distances between the X-ray source and detector until the projection of the studied sample fills the panel. Therefore, the best magnification is the one where the voxel size is 1/2000th of the width or height of the sample. For the examined rock-sample, voxel size was 0.246 mm.

The projections from the X-ray scanner have been reconstructed, followed by manual segmentation of key elements (soft limestone, phosphoclasts and fossils) into distinctive colors and preparation of the final 3-D visualization of the studied interval in the Drishti program (Ajay Limaye, ANU Vizlab). The shots were made from eight different angles, with particular elements of the examined sample visible in various configurations. All these screenshots were created on a black background (.jpg and .bmp extensions) and white (.bmp only). In total, over 110 files were created during the present study.

Figure processing

All figures in this paper were prepared using CorelDraw 18 and Adobe Photoshop CS3.