Introduction

In the last few decades, an integrated sedimentological and ichnological analysis has increasingly proven to be a most useful approach for studying sedimentary successions, as the two types of data supplement and enhance each other (Howard 1975; Savrda and Bottjer 1989, 1991; Pemberton et al. 1992a; Pemberton and Wightman 1992; Uchman 1995, 1998; Gingras et al. 1999; MacEachern et al. 1999a; McIlroy 2004; Uchman et al. 2008; Angulo and Buatois 2012). Ichnological data (Bertling et al. 2006, 2022; Seilacher 2007) allow an assessment of the palaeobathymetry, near-bottom salinity, oxygenation and sediment mobility (Seilacher 1967; Frey et al. 1990; MacEachern et al. 1992; Pemberton et al. 1992a, b; Pemberton and MacEachern 1995, 1997, 2007a; Savdra 1992, 1995; Uchman et al. 2008), whilst the corresponding sedimentological evidence allows recognition of the actual sedimentation processes in the palaeoenvironment hosting particular ichnofauna (Howard and Frey 1984; Raychaudhuri and Pemberton 1992; MacEachern and Pemberton 1992, 1994; MacEachern et al. 1999a, 2005; MacEachern and Løseth 2003; Gouramanis et al. 2003; Taylor et al. 2003; Bann et al. 2004; Uchman et al. 2007). Such integrated interdisciplinary studies have revealed, for example, some strong links between the ichnofauna assemblages and dominant style of bioturbation on the one hand and the substrate type and environment hydrodynamic conditions on the other (Howard and Reineck 1981; Howard and Frey 1984; Pemberton and Frey 1984; Vossler and Pemberton 1989; Frey 1990; MacEachern and Pemberton 1992; Pemberton and MacEachern 1995; MacEachern et al. 1999b; Bann and Fielding 2004; Pemberton et al. 2004; Buatois et al. 2008).

The aim of the present study from the Carolinefjellet Formation (Aptian–Albian) in central Spitsbergen, Svalbard (Fig. 1), is to contribute to this line of interdisciplinary research by shedding more light on the sedimentary facies and ichnofauna communities of ancient stormy marine shelves. The archipelago of Svalbard at the western margin of the Barents Sea Shelf, presently a deglaciated high-Arctic polar desert, was an area located around 63–66 °N in the Early Cretaceous (Torsvik et al. 2002, 2008; Worsley 2008; Grundvåg and Olaussen 2017).

Fig. 1
figure 1

A Map of the Svalbard archipelago showing the outcrop of Cretaceous in Spitsbergen (after Hjelle 1993). B The present-day location of Svalbard in the arctic part of North Atlantic province. C Geological map of the Adventdalen area in central Spitsbergen (after Braathen et al. 2012), showing the location of the town of Longyearbyen and the four boreholes (Dh1–Dh4) referred to in the text; the outcrop section studied is a road-cut cliff extending from the town to its airport. D Cretaceous stratigraphy of Spitsbergen (after Mørk et al. 1999; Midtkandal et al. 2008; Grundvåg et al. 2017) and global eustatic curve (after Haq 2014); TL ‒ base-Aptian transgressive lag with global anoxia invasion into Spitsbergen dysoxic estuarine lagoon (cf. Figure 2D)

The Cretaceous period is known for its global greenhouse climate and lack of strong latitudinal temperature gradient (Fischer 1981; Ziegler et al. 1987; Wilkinson and Riding 2007; Huang et al. 2012; Moriya 2011; Scotese et al. 2021), with the emerged Svalbard area in the Barremian hosting both high vegetation and a diversified population of herbivore and carnivore dinosaurs. The Early Cretaceous Svalbard had a generally warm-temperate humid climate (estimated mean annual temperatures of 7–10 °C) that supported an ornithopod dinosaur population, deciduous conifers, ginkgos and local peat accumulation forming coal seams (Heintz 1963; Edwards et al. 1978; Steel and Worsley 1984; Nemec 1992; Hurum et al. 2006). However, the Aptian regional marine transgression in Svalbard area opened the Barents Shelf to the adjacent Arctic Ocean (Steel and Worsley 1984; Blakey 2011), whereby the epicontinental sea with its relatively warm waters became subject to a harsh Arctic wave climate and transient incursions of cold bottom water from spells of polar glacial temperatures (Weissert and Lini 1991; Maher et al. 2004; Kessels et al. 2006; Price and Nunn 2010; Price and Passey 2013; Scotese et al. 2021; Wang et al. 2022). These physical conditions amounted to a specific type of shelf environment, thus far little documented in the global stratigraphic record. The present study reports on the sedimentary deposits and ichnofauna assemblages of this subpolar Early Cretaceous greenhouse shelf.

Regional palaeogeographic setting

Spitsbergen is the main island of the Svalbard archipelago (Fig. 1A, B), exposing Precambrian metamorphic basement overlain by a sedimentary rock succession from the Upper Devonian to Neogene, with a notable local lack of the Upper Cretaceous (Parker 1967; Steel and Worsley 1984; Nøttvedt et al. 1992). The Carolinefjellet Formation constitutes the uppermost part of the Jurassic–Lower Cretaceous Adventdalen Group (Fig. 1D; Mørk et al. 1999), which extends over a large area of Spitsbergen as slivers cropping out at the synclinal fringes of the Central Spitsbergen Palaeogene Basin (Fig. 1A, C). The mud-rich open-marine deposits of the Aptian–Albian Carolinefjellet Formation rest transgressively on the sand-rich, coal-bearing Barremian fluvio-deltaic Helvetiafjellet Formation (Fig. 1D; Nemec et al. 1988; Nemec 1992; Gjelberg and Steel 1995) and are overlain unconformably by the coal-bearing Palaeocene fluvio-deltaic deposits of the lowermost Van Mijenfjorden Group (Steel and Worsley 1984; Nøttvedt et al. 1992; Mørk et al. 1999).

The base of the Carolinefjellet Formation is a transgressive surface marking the onset of marine sublittoral sedimentation over the paralic/littoral deposits of the upper Helvetiafjellet Formation (Fig. 1D; Nemec et al. 1988; Gjelberg and Steel 1995; Mørk et al. 1999). The top is an erosional unconformity, with the preserved thickness of Carolinefjellet Formation ranging in outcrop from 220 m in the NW to nearly 1300 m in the SE (Nagy 1970; Mørk et al. 1999). The erosional truncation increases both northwards and westwards, towards the West Spitsbergen Palaeogene strike-slip orogen (Steel et al. 1985). The maturity of organic matter indicates that possibly up to 1000 m of unknown Upper Cretaceous deposits were removed by erosion in the Svalbard area and the adjacent part of the Barents Sea Shelf (Worsley 2008; Midtkandal et al. 2016; Grundvåg et al. 2019). Extensive Upper Cretaceous deposits, including carbonates, are well documented from offshore wells in the southern part of the Barents Shelf (Nøttvedt et al. 1992; Worsley 2008; Grundvåg et al. 2017) and also westwards in the then adjoining northern Greenland and NE Canadian Arctic (Embry 1991; Worsley 2008). The end-Cretaceous erosion of Svalbard was apparently due to an uplift of the Barents Shelf along its rifted northern and western margins, combined with the transpressional West Spitsbergen orogeny (Steel et al. 1985; Ziegler 1988; Torsvik et al. 2002; Worsley 2008).

The Berriasian to lowest Barremian Rurikfjellet Formation (Fig. 1D) consists of shale-dominated offshore deposits passing upwards into sandy, storm-worked lower shoreface deposits with hummocky stratification (Midtkandal et al. 2008; Grundvåg et al. 2019). In the earliest Barremian, the Svalbard area emerged as part of the volcanically active Lomonosov Ridge that separated the Barents Shelf from the opening Arctic Ocean (Fig. 2A, B). A broad fluvio-deltaic system of the Helvetiafjellet Formation prograded to the SE across the lowland of Spitsbergen (Figs. 1D, 2C; Steel and Worsley 1984; Nemec et al. 1988; Nemec 1992). This system then receded to the NW in the middle to late Barremian due to a stepwise marine transgression (Fig. 1D; Gjelberg and Steel 1995), which turned the southern Spitsbergen area into a dysoxic muddy estuarine embayment (Fig. 2D). The area was eventually drowned and opened to the Arctic Ocean (Fig. 2E, F), with the onset of the open-marine epicontinental environment of the Carolinefjellet Formation (Fig. 1D).

Fig. 2
figure 2

Regional setting. A Palaeogeography of the western part of subarctic region at the beginning of Barremian (~ 130 Ma BP) and in mid-Aptian (~ 115 Ma BP); spherical (orthographic) reconstructions slightly modified from Blakey (2011). BF Hypothetical palaeogeography in Svalbard area (cf. stratigraphy in Fig. 1D), with the background thin black outline of the archipelago’s present-day shoreline (cf. Figure 1A). B The Helvetjafiellet Fm braidplain delta at its maximum extent (early Barremian). C The delta retreat and formation of dysoxic estuarine lagoon (middle to late Barremian). D The lagoon drowning by anoxic waters of OAE1a (Barremian/Aptian transition, base of Carolinefjellet Fm). E The onset of Dalkjegla Mb deposition in transgressively enlarged embayment (earliest Aptian). F The opening of Svalbard seaway and deposition of Carolinefjellet Fm (Aptian‒Albian). Reconstructions inspired by Nemec et al. (1988), Nemec (1992) and Grundvåg et al. (2017, 2019, 2021)

The invasion of anoxia in Svalbard at the Barremian/Aptian transition (Fig. 1D) is linked to the global oceanic anoxia event OAE1a (Midtkandal et al. 2016). Cretaceous worldwide activity of mantle plumes (Johansson et al. 2018; Müller et al. 2019; Scotese et al. 2021) formed several large igneous provinces (LIPs), including the high-Arctic HALIP of the Lomonosov Ridge (Maher 2001; Torsvik et al. 2006; Polteau et al. 2016). The associated lithospheric degassing of CO2 is a widely invoked factor for the OAE1a, with an initially doubled atmospheric CO2 concentration (Ganino and Arndt 2009; Naafs et al. 2016) and possible mass extinctions (Rampino and Self 2015; Scotese et al 2021). The HALIP igneous province, with an area of about 700,000 km2, could alone release an estimated 1000 s GT CO2 and significantly contribute to the global OAE1a (Midtkandal et al. 2016; Polteau et al. 2016). The time lag between the anoxic marine invasion and the ultimate opening of the Svalbard shelf seaway (Fig. 2C, D) is hypothetically estimated as possibly no more than 300 to 400 ka (see next paragraph).

The Carolinefjellet Formation

The Carolinefjellet Formation (Harland 1997; Dallmann 1999) consists of dark-grey to blackish mudshales, clayey to silty, intercalated in a variable proportion with thin to thick sandstone sheets. Characteristic features include abundant bioturbation, sporadic shell-debris layers, chamosite ooids, glendonites and siderite concretions (Krajewski and Luks 2003; Mutrux et al. 2008). The basal erosional unconformity (Fig. 1D) is patchily strewn with a transgressive gravelly lag and overlain by a unit of non-bioturbated black mudshales, ~ 15 m thick, correlated to the OAE1a (Midtkandal et al. 2016). The deposition rate of pure black shales is widely estimated as 1‒2 mm ka−1 (compacted values). However, the basal black shales in the present case have numerous thin sandstone interlayers and vary from laminated to massive (Fig. 3; Midtkandal et al. 2016). Their deposition probably involved abundant dysoxic lagoonal mud of the topmost Helvetiafjellet Formation (Nemec et al. 1988; Nemec 1992), resuspended by the transgression, as well as a spontaneous emplacement of massive fluid-mud layers (Baas et al. 2009). When the rate of mud settling and aggradation into a near-bottom layer exceeds the mud-dewatering rate, the fluid mud tends to flow gravitationally according to the local seafloor gradients (Mehta 1991; Kranenburg and Winterwerp 1997). Assuming a conservative mean rate of 4‒5 mm ka−1 (compacted values), the deposition of the anoxic muddy unit in Spitsbergen might have taken 300‒400 ka.

Fig. 3
figure 3

Example log of the Carolinefjellet Fm in central Spitsbergen (well Dh2, Fig. 1C), showing stratigraphic distribution of lithofacies, their associations and the degree of sediment bioturbation

The Carolinefjellet Formation has been divided into five subunits, referred to in their ascending stratigraphic order as the Dalkjegla, Innkjegla, Langstakken, Zillerberget and Schønrockfjellet members (Fig. 2; Parker 1967; Nagy 1970; Dallmann 1999; Grundvåg et al. 2021). These units have conformable boundaries and differ chiefly in their relative proportion of mudshales and sandstones, but also in some other distinctive features. For example (Dallmann 1999; Krajewski and Luks 2003; Maher et al. 2004; Mutrux et al. 2008): The basal part of Dalkjegla Member consists of black shales nearly devoid of burrows. Chamosite ooids occur only in the Dalkjegla Member, 22–30 m above its base. Siderite concretions form several stratigraphic horizons in the Dalkjegla and Innkjegla members. Glendonites appear in the topmost Dalkjegla Member and recur through most of the Innkjegla Member.

Prior to the modern concepts of shelf sedimentation and tempestites, the heterolithic Carolinefjellet Formation was tentatively considered as tidal-flat deposits (Birkenmajer 1966). Researchers presently agree that the formation represents sedimentation in a wave-dominated, open-marine inner to outer shelf environment affected by relative sea-level changes, with minor direct evidence of tidal influences (Nagy 1970; Steel and Worsley 1984; Nøttvedt and Kreisa 1987; Gjelberg and Steel 1995; Dypvik et al. 2002; Maher et al. 2004; Midtkandal 2007; Worsley 2008; Grundvåg et al. 2019, 2021). The Aptian lower part of the Carolinefjellet Formation (Fig. 1D) is generally regarded as a continued record of the punctuated regional marine transgression initiated by deposition of the underlying Glitrefjellet Member of the Helvetiafjellet Formation (Nemec et al. 1988; Gjelberg and Steel 1995; Dypvik et al. 2002). The Albian upper part of the Carolinefjellet Formation (Fig. 1D) is considered to represent a punctuated highstand regression, markedly aggradational (Dallmann 1999; Dypvik et al. 2002; Maher et al. 2004; Midtkandal 2007; Grundvåg et al. 2019, 2021).

However, the exact character of the depositional environment and its palaeogeography are uncertain. The Barremian shoreline of the advancing and retreating braidplain delta of the Helvetiafjellet Formation is widely recognized to have had a NE–SW trend (Fig. 2A; Steel and Worsley 1984; Nemec et al. 1988; Nemec 1992; Gjelberg and Steel 1995; Midtkandal 2007; Worsley 2008; Grundvåg et al. 2021). The location and spatial trend of the Aptian–Albian shoreline in Svalbard is less clear, as the coeval coastal deposits are non-preserved. Maher et al. (2004) suggested an E‒W shoreline trend with a westward alongshore sediment drift. Dallmann (1999) and Dypvik et al. (2002) considered the Carolinefjellet Formation in Spitsbergen to be a prodelta system advancing towards the SW on a wave-dominated shelf, with the shoreline trending NW–SE and located in the middle of Svalbard, perhaps at the northernmost reaches of inner Isfjorden (Fig. 1A). This interpretation would be consistent with the evidence of transported plant debris and the observed slight decrease in the relative proportion of sandstone beds towards the south (Parker 1967; Nagy 1970), obliquely away from the shoreline. The shoreline might well involve some small wave-dominated deltas, but the notion of a prodeltaic environment per se is questionable. First, the fluvial drainage of the Lomonosov island (Fig. 2F) would be sparse, unlikely to produce any major river deltas. Second, wave-dominated deltas generally lack typical river-fed prodelta zone and are instead fronted by wave-worked shorefaces with a storm-controlled offshore transition (Bhattacharya and Walker 1992; Hampson and Howell 2005).

In the first modern sedimentological study of the Carolinefjellet Formation, Nøttvedt and Kreisa (1987) have recognized its sheet-like sandstone beds as tempestites (see also Grundvåg et al. 2021)—deposits characteristic of a storm-dominated lower shoreface to offshore transition zone, rather than diagnostic of a prodelta environment. Nøttvedt and Kreisa (1987) have recognized also frequent localized stacking of tempestites into mounded, isolated or semi-isolated sandbodies, 1–3 m thick, with a clinoformal bed stacking and anisotropic hummocky cross-strata indicating broadly westward (SW to NW range) direction of sediment transport (see also Grundvåg et al. 2021). For a storm-driven transport influenced by Coriolis effect (Walker 1984b), this evidence would be consistent with a shoreline located to the NE (Birkenmajer 1966; Dallmann 1999; Dypvik et al. 2002). However, these mounded sandbodies have been more recently interpreted by Maher et al. (2004) as low-relief barrier-bar complexes formed on a wave-dominated and tidally influenced inner shelf, accreted by a westward alongshore sediment drift and sheltering brackish tidal lagoons with suboxic bottom conditions. This interpretive scenario of a barred shoreline trending E–W in the midst of Spitsbergen intended to explain a presumed lagoonal origin of the chamosite ooids found in the Dalkjegla Member. Although the interpretation might seem consistent with the southward decrease in the proportion of sandstone beds, several other facts render it questionable. First, the mounded sandbodies occur randomly at all levels of the Carolinefjellet Formation and are found also in its southernmost outcrops in Spitsbergen (Nøttvedt and Kreisa 1987), where—nearly 150 km away from the shoreline postulated by Maher et al. (2004)—hardly any sand should have been delivered and no nearshore barrier bars would be expected. Instead, the sedimentary succession there, in Kvalvågen area (Nøttvedt and Kreisa 1987; Århus 1991; Midtkandal 2007), is only slightly poorer in sand and contains similar mounded sandbodies. Second, the mounded sandbodies in the Carolinefjellet Formation show virtually no facies resemblance to barrier bars (cf. Davies et al. 1971; Reading and Collinson 1996) and are compound tempestite stacks (Grundvåg et al. 2021) known from other mid-shelf areas (Midtgaard 1996). Last, but not least, the iron ooids occur in association with offshore-transition tempestites (cf. Price and Nunn 2010), and hence were either derived by storms from erosion of the lagoonal deposits of the uppermost of Helvetiafjellet Formation or formed directly in the offshore transition area (cf. Boyd et al. 1992; Colin et al. 2005; Boyd 2010).

In summary, researchers generally agree that the sedimentological evidence from the Carolinefjellet Formation is incompatible with the earlier Barremian shoreline and that the Svalbard area must have undergone a dramatic palaeogeographic change in the earliest Aptian (Fig. 2), even though the exact pattern of this change is not quite clear. The present study concurs with the palaeogeographic reconstructions postulating that the Barents Sea Shelf in Barremian was sheltered from the opening Arctic Ocean by a volcanically active peninsular land area of the Lomonosov Ridge (Fig. 2A‒C; Ziegler 1988; Brekke et al. 2001; Torsvik et al. 2002; Blakey 2011). The end-Barremian Spitsbergen embayment was sheltered by a prominent sand-barrier complex, ~ 20 m thick (Fig. 2B; Nemec et al. 1988), with the back-barrier lagoonal area hosting dysoxic waters (Nemec 1992). The barrier slightly migrated landwards to the NW, but was drowned by invasion of the OAE1a anoxic water at the Barremian/Aptian transition (Midtkandal et al. 2016). The early Aptian further transgression then opened across Svalbard a wide epicontinental seaway linking the western Barents Sea with the Arctic Ocean (Fig. 2D; Owen 1988; Århus 1991; Nøhr-Hansen 1993; Dypvik et al. 2002; Blakey 2011). The Aptian–Albian local shoreline would then likely assume a NW–SE trend (Fig. 2D; Birkenmajer 1966; Århus 1991; Dypvik et al. 2002), with the island area to the NE being too small to have any extensive fluvial catchments with large deltas. The wave-dominated shorelines of the seaway are unpreserved. The NNW–SSE trend of the Cretaceous outcrop in Spitsbergen, imposed by the Palaeogene foreland syncline (Fig. 1A), would thus be slightly oblique to the Aptian–Albian local shoreline and shelf depositional strike.

High vegetation in the Svalbard area persisted from Barremian until at least the Palaeocene (Steel and Worsley 1984; Nøttvedt et al. 1992; Worsley 2008) and the Arctic isotope data from fossil molluscs indicate a generally warm (≤ 20 °C) seawater during the Cretaceous (Tarduno et al. 1998). However, the isotope thermometry has also indicated that the polar zone probably experienced transient glacial temperatures (Price and Nunn 2010; Price and Passey 2013; Vickers et al. 2019; see also Weissert and Lini 1991; Kessels et al. 2006). The episodic occurrences of glendonites and preservation of plagioclase grains in a sand-starved sublittoral environment of the Carolinefjellet Formation imply cool (4–7 °C) bottom-water conditions (De Lurio and Frakes 1999; Maher et al. 2004; Price and Nunn 2010). This may indicate invasions of deep polar water onto the Barents Shelf prior to the opening of the North Atlantic and development of its modern oceanic circulation (cf. Ziegler 1988; Doré 1991; Mosar et al. 2002; Torsvik et al. 2002). Invasions of polar water might also explain the predominance of a mixture of non-endemic cold-water dinoflagellates in the Aptian and Albian of the Barents Shelf, similar as in the northern Greenland (Århus 1991; Nøhr-Hansen 1993).

Methods and terminology

The data for the present study were derived from four fully cored boreholes (labelled Dh1 to Dh4, Fig. 1C), drilled near the central Spitsbergen town of Longyearbyen under a UNIS project (Braathen et al. 2012), and from a corresponding road-cut outcrop section, about 2.5 km long, between the town and its airport (Fig. 1C). The well cores show the lower to middle part of the Dalkjegla Member of the Carolinefjellet Formation (Fig. 1D), whereas the outcrop shows the member’s middle to upper part. The cores were studied by detailed sedimentological logging (scale 1:20), with a special focus on biogenic features. Documentation included photographs. The core samples, stored at the UNIS in Longyearbyen, were non-oriented, hence palaeocurrent directions could only be measured in the outcrop section.

The descriptive sedimentological terminology is after Folk (1980), Harms et al. (1982) and Collinson et al. (2006). The term lithofacies refers to the basic types of sedimentary deposits distinguished on the macroscopic basis of their bulk characteristics (Harms et al. 1975; Walker 1984a). The term facies association denotes an assemblage of spatially and genetically related lithofacies, thought to represent a particular zone (sub-environment) of the depositional system. The distinction of shoreface, offshore transition and offshore zones is based on sedimentary facies and pertains to the prevalent depths of fairweather and storm wave bases (Reading and Collinson 1996). The definition of bioturbation structures and trace fossils is after Bertling et al. (2006, 2022). The simplified four-grade scale for sediment bioturbation intensity used in summary figures refers to the bioturbation index (BI) of Taylor and Goldring (1993) as follows: no bioturbation (BI = 0), low (BI = 1), moderate (BI = 2–3) and high bioturbation (BI = 4–6).

Sedimentary facies

The following lithofacies have been recognized as the basic building blocks of the sedimentary succession in the present study:

  • C – basal conglomerate, massive to diffusely planar stratified;

  • Sp – sandstone with planar parallel stratification;

  • Shs – sandstone with hummocky and/or swaley stratification;

  • Sr – sandstone with wave and/or combined-flow ripple cross-lamination;

  • Sm – massive mottled sandstone homogenized by bioturbation;

  • H – heterolithic facies of thinly interlayered sandstone and mudshale;

  • Hm – massive silty to sandy mudstone or muddy sandstone homogenized by bioturbation;

  • Ml – planar parallel laminated silty mudshale;

  • Mm – homogenous (non-laminated) mudshale.

The sedimentary facies are briefly described and genetically interpreted in Table 1, with an example well-core log in Fig. 3 and lithofacies shown in Figs. 4, 5 and 6. For a more detailed depiction of lithofacies, see Grundvåg et al. (2021).

Table 1 Lithofacies of the Dalkjegla Member of the Carolinefjellet Fortmation in the study area
Fig. 4
figure 4

Outcrop examples of lithofacies (cf. Table 1) and their associations. A Broad view of the vertical stacking pattern of lithofacies associations. B Isotropic hummock structure in facies Shs bed. C Composite isotropic hummock in facies Shs bed. D Anisotropic hummock in facies Shs bed. E Large storm bar in facies Shs, formed by lateral clinoformal stacking of hummocks. F Scour trough (arrow), ~ 0.8 m deep and several metres wide, filled with facies Shs sandstones; note the nearly tabular geometry of the enveloping deposits of FA1‒FA3

Fig. 5
figure 5

Well-core examples of lithofacies (cf. Table 1); core diameter 5 cm. A Inversely graded basal conglomerate (C) with large sideritized mudclasts, overlying strongly bioturbated muddy sandstone (Hm) with Schaubcylindrichnus (Sch) burrows. B Sandstones of facies Shs passing upwards into facies Sp. C Composite bed of facies Shs passing upwards into facies Sr. D Thick composite bed of facies Sp

Fig. 6
figure 6

Well-core examples of lithofacies (cf. Table 1); core diameter 5 cm. A Graded beds of facies Sp capped with bioturbated layers of facies H. B Facies C bed with large sideritized mudclasts, overlain by facies Shs and Sr. C Isolated bed of bioturbated facies Sm sandwiched between layers of facies Hm. D Swaley bed of facies Shs underlain by facies Sr and covered by a mudclast conglomerate (C) at the base of facies Sp bed. E Facies Sr rich in tiny iron ooids. F Beds of facies H, with combined-flow ripples and isolated burrows, separated by facies Ml. G Burrowed shelly lag. H Sand injectite intrusion in facies Mm, above the sandstone facies Shs and Sr. I Facies Hm overlain by facies H, showing marked short-term variation in bioturbation intensity. J Sandstone of facies Sr with a flute solemark

The lateral discontinuity of the basal conglomerate layer (lithofacies C), interpreted as a transgressive lag, is probably due to the erosional stripping and lateral reshuffling by sea waves of the coarse sediment from accessible fluvial channel belts of the topmost Helvetiafjellet Formation (cf. Nemec 1992). Sandstone beds range in thickness from 1 to 80 cm and have sharp bases and tops. The thinnest beds consist of lithofacies Sr, with the ripple cross-laminae sets commonly flattened and/or loaded by compaction (Fig. 6F‒I). Thicker beds consist of lithofacies Sp and/or Sr, whereas the thickest ones show mounded segments with lithofacies Shs (Figs. 4B‒E, 5B‒D). The internal structure of hummocks ranges from roughly symmetrical solitary sets of dome-shaped parallel strata to multiple sets separated by convex-up truncations (Fig. 4B), and to sporadic unidirectional cross-strata sets underlain by thin lithofacies Sp and draped by lithofacies Sp and/or Sr (Fig. 4A). Structure asymmetry (anisotropy) indicates transport towards the WNW or NW. Amalgamated composite sandstone beds reach locally 2‒3 m in thickness (Figs. 3, 4). The tops of non-amalgamated bed, overlain by mudshale, show well-preserved oscillatory wave vortex ripples or combined-flow ripples, often with worm trails (Figs. 6F, I, 7B). All these features support Nøttvedt and Kreisa’s (1987) original interpretation of the sandstone beds as tempestites (Table 1; cf. Dott and Bourgeois 1982; Dumas et al. 2005). Bed soles show occasional grooves, small flutes and prod marks indicating flow direction broadly towards the WSW, but often varying by up to 30° on the same bed sole (Birkenmajer 1966). Cross-lamination of combined-flow ripples indicates sand transport in directions ranging between SW and WNW. Asymmetrical wave ripples indicate transport mainly towards the NE or ESE, which may reflect variable wind-modulation of the water oscillatory movement towards the contemporaneous irregular shoreline. Ptygmatic injection dikes, a few centimetres thick and extending above sandstone bed tops (lithofacies Sd; Fig. 6H), are evidence of local post-depositional remobilization of sand by liquefaction, possibly due to seismic events.

Fig. 7
figure 7

Trace fossils in the Dalkjegla Mb of Carolinefjellet Fm. A Strogly bioturbated top of sandstone bed showing ?Aulichnites isp. (Au), Planolites beverleyensis (Pb), P. montanus (Pm) and Planolites isp. (Pl). Loose slab from the outcrop in Longyearbyen. B Gyrochorte comosa (Gy) on wave-rippled top of sandstone bed in Longyearbyen outcrop; the coin (scale) is 2 cm in diameter. C Sandstone with crowded traces ?Helminthopsis isp. (arrow); well Dh2, depth 64–65 m. D Moderately bioturbated facies H, showing vertical cross-section of bedding-oblique, protrusive Rhizocorallium isp. (Rh); well Dh2, depth 123.83–123.88 m. E Vertical cross-section of bedding-parallel spreiten burrow resembling Rhizocorallium (arrow); well Dh1, depth 69.07–69.14 m. F Sandstone with upward-increasing bioturbation intensity, including ?Rhizocorallium isp. (Rh) and funnel-shaped laminated burrow (ss); well Dh1, depth 102–103 m. G Bedding-parallel and vertical Planolites beverleyensis (Pb) at transition from sandstone to mudstone; loose rock slab in Longyearbyen outcrop; scale bar 1 cm. H Facies H showing cross-sections of Planolites isp. (Pl) and poorly visible Phycosiphon incertum (Ps); well Dh1, depth 80–81 m. I Heavily bioturbated muddy sandstone showing vertical cross-section of Phycosiphon ?incertum (Ps), Palaeophycus heberti (Ph) and mud filled small, vermicular, taxonomically undetermined burrows (?Helminthopsis, ?Cosmorhaphe); well Dh2, depth 51–52 m. J Small forms of Palaeophycus heberti (Ph) with differing thicknesses of burrow lining, and small hallowed burrows resembling Phycosiphon incertum (Ps); well Dh1, depth 132–133 m

The sandstones are predominantly well sorted, fine-grained to medium-grained and quartz-rich sublithic to subarkosic arenites (Maher et al. 2004). The rare scattered small pebbles and basal patchy conglomerate layer similarly consist mainly of rounded quartz and chert clasts (Fig. 6D‒E). The textural and mineralogical maturity supports the notion of a wave-worked sediment derived by storms from shoreline zone. However, some of the gravel clasts are merely subrounded. The sand fraction also shows highly varied grain roundness and an admixture of relatively weak mineral grains, such as plagioclase, microcline, chlorite and biotite (Maher et al. 2004). Sediment supply probably involved recycling of older deposits by coastal erosion and local fluvial delivery, with alongshore drift of sediment, its mixing and variable maturation. Sediment provenance included a contemporaneous volcanic source (Nysæther 1966), linked to the HALIP area in the NE part of the Lomonosov Ridge (Ziegler 1988; Tarduno et al. 1998; Maher 2001; Maher et al. 2004; Mutrux et al. 2008).

Carbonate clasts are rare, including fragments of pelecypod shells, fine-grained contemporaneous hardground limestones and eroded intraformational concretions of microspar calcite or siderite (Maher et al. 2004). In addition to resedimented iron ooids in the lower part the formation, some coalified plant detritus and scattered petrified fragments of driftwood are sporadically found (Mutrux et al. 2008). Early-diagenetic glauconite occurs and increases in abundance upwards in the succession (Maher et al. 2004), which may indicate a decreasing bulk rate of sediment accumulation. Small concretions of bacterial framboidal pyrite tend to be associated with clay pellets and animal burrows. Ball-shaped and strata-bound sideritic ironstone concretions in mudshale beds indicate an early subsurface cementation of uncompacted sediment by non-ferroan calcite in the upper part of sulphite-reduction zone, enriched in framboidal pyrite at the local sites of enhanced bacterial decomposition of organic matter (Krajewski and Luks 2003).

Trace fossils

Trace fossils are common in the studied succession, slightly more abundant in wells Dh1 and Dh2 than in wells Dh3 and Dh4 (Fig. 1C, 3). Only the basal conglomerate (lithofacies C), the overlying black mudshale unit (lithofacies Ml/Mm) and the majority of thick sandstone beds (lithofacies Sp and Shs) are nearly devoid of bioturbation structures (Figs. 6B, D and 8B). The distinct basal unit of lithofacies Ml/Mm shows merely sporadic structures reminiscent of thread-like burrow fills, manifested on the core surface as small spots of lighter-shade very fine-grained sand, 1‒2 mm in diameter. In contrast, all units of lithofacies Sm are highly bioturbated (BI = 5–6; Fig. 7C).

Fig. 8
figure 8

Trace fossils in the Dalkjegla Mb of Carolinefjellet Fm. A Schaubcylindrichnus freyi (Sf, arrows), variably oriented, in lithofacies H; well Dh3, depth 140–141 m. B Schaubcylindrichnus coronus (arrows) in oblique section in lithofacies Sm; well Dh1, depth 124–125 m. C Palaeophycus ?sulcatus (Pu), P. tubularis (Pt) and Thalassinoides ?suevicus (Th) on mudstone bedding surface; loose rock slab in Longyearbyen outcrop. D Strongly bioturbated lithofacies Sm showing mud-lined, variably oriented burrows interpreted as Palaeophycus isp. (Pa); well Dh1, depth 124–125 m. E Ichnotraces ?Rosselia isp. (R) and Palaeophycus isp. (Pa) in lithofacies Sp; well Dh1, depth 120–121 m. F Mud-lined Palaeophycus isp. (arrow) in sandstone lithofacies; well Dh1, depth 124–125 m. G Ichnotraces Arenicolites isp. (Ar), Diplocraterion isp. A (Dp) and Palaeophycus tubularis (Pt) in lithofacies H; loose rock slab in Longyearbyen outcrop; scale bar 1 cm. H Ichnotraces ?Chondrites targionii (Ch) and Siphonichnus isp. (Si) in lithofacies H; well Dh2, depth 20–21 m. I Lithofacies Hm with ?Phycosiphon isp. (arrow); well Dh1, depth 95–96 m. J Lithofacies Hm with ?Phycosiphon isp. (arrows); well Dh2, depth 81–82 m. K Clusters of sand-filled burrows (arrows) in lithofacies H, interpreted as vertical section of spiral burrows resembling Gyrolithes; well Dh1, depth 117–118 m

The BI varies from 0 to 5 on a bed thickness scale and on the scale of bed packages several metres thick. Burrows are virtually absent in the lower parts of sandstone beds thicker than 5 cm and are increasingly more common in their top parts, at the contact with the overlying mudshale (Fig. 6A, C), which means ‘lam-scram’ sequences sensu Bromley (1990). Burrows are also relatively rare or absent in the heterolithic deposits of lithofacies H and in sideritized mudshales (Fig. 6B, F). On a larger stratigraphic scale, burrows are most abundant in the interval from 30 to 55 m above the formation base, whereas the higher part of the Dalkjegla Member shows marked fluctuations in bioturbation intensity (Fig. 3).

The burrows indicate a range of seafloor animal ethological activities, mainly feeding, dwelling, grazing, resting and crawling. Most bioturbation structures visible in non-slabbed core samples are small portions of burrow systems of unclear taxonomic affiliation. Many of them were produced by sediment scrambling, often multiple, or by scrambling of a soupy-state sediment substrate, which renders them taxonomically non-classifiable. In general, an exact taxonomic classification of trace fossils at ichnospecies level was seldom possible, but 38 ichnotaxa were inferred in the sedimentary succession (Figs. 7, 8, 9, 10 and 11), with three undetermined and 14 uncertain (labelled with question marks). A systematic ichnological description of the trace fossils is given in Table 2.

Fig. 9
figure 9

Trace fossils in the Dalkjegla Mb of Carolinefjellet Fm. A Hallowed burrows resembling large Phycosiphon incertum (arrows) in strongly bioturbated silty upper part of lithofacies Sp graded bed; the dark triangular structure in overlying Sp bed seems to represent a dwelling bivalve burrow (domichnion); well Dh1, depth 94–95 m. B Strongly bioturbated bed of lithofacies Sm with distinct Asterosoma isp. (arrows); well Dh1, depth 79–80 m. C Moderate bioturbation with Asterosoma (As) and ?Planolites isp. (Pl) in lithofacies H; well Dh1, depth 30–31 m. D Strongly bioturbated lithofacies Hm with Asterosoma isp. (As), Planolites beverleyensis (Pl) and structures resembling Teichichnus isp. (?Te); well Dh1, depth 111–112 m. E Asterosoma isp. (As) in lithofacies Sm; well Dh1, depth 80–81 m. F Ophiomorpha cf. rudis (O), Phycosiphon incertum (Ps) and undetermined burrows in lithofacies H; well Dh1, depth 30–31 m. G Sandstone bedding plane showing Ophiomorpha cf. rudis (O); loose rock slab in Longyearbyen outcrop. H Sandstone bed sole showing Thalassinoides suevicus (Th); loose rock slab in Longyearbyen outcrop. I Strongly bioturbated mud and sparsely bioturbated sand in lithofacies H with ?Phycodes isp. (?Py) Palaeophycus tubularis (Pt) and Thalassinoides isp. (?Th); well Dh3, depth 79–80 m. J Thickly mud-mantled sand-filled tubes interpreted as Phoebichnus isp. (Po) and Palaeopycus heberti (Ph); well Dh1, depth 107–108 m. K Structure resembling lower part of Diplocraterion isp. (?D); well Dh1, depth 129–130 m

Fig. 10
figure 10

Trace fossils in the Dalkjegla Mb of Carolinefjellet Fm. A Moderately bioturbated lithofacies H with ?Astreosoma isp. (?As), ?Cylindrichnus isp. (?Cy), Phoebichnus isp. (Po), Phycosiphon (Ps) and Siphonichnus isp. (Si); scale 1 cm, well Dh3, depth ~ 66.5 m. B Weakly bioturbated lithofacies H with Phoebichnus isp. (Po) and Siphonichnus isp. (Si); well Dh3, depth 72.0–72.1 m. C Arenicolites isp. (arrows) in moderately bioturbated lithofacies H; well Dh3, depth 95–96 m. D Strongly bioturbated lithofacies H with ?Rosselia isp. (?Ro) and ?Asterosoma isp. (?As); well Dh1, depth 108–109 m. E Moderately bioturbated lithofacies H with Conichnus aff. papillatus (arrow); well Dh1, depth 130–131 m. F Strongly bioturbated lithofacies Sm with presumed Conichnus isp. (arrow); well Dh1, depth 114–115 m

Fig. 11
figure 11

Trace fossils in the Dalkjegla Mb of Carolinefjellet Fm. A Moderately bioturbated sandstone with sideritized structure resembling Conichnus isp. (arrow); well Dh1, depth 106–107 m. B Strongly bioturbated lithofacies Hm with ?Cylindrichnus concentricus aff. (Cy); well Dh2, depth 106–107 m. C ?Cylindrichnus concentricus aff. (Cy) in moderately bioturbated lithofacies H; well Dh1, depth 129–130 m. D Vertical section of ?Lingulichnus isp. (Li) in lithofacies D; well Dh2, depth 22–23 m. E Vertical section of ?Lingulichnus isp. (Li) extending from mudstone interlayer into sandstone; well Dh1, depth 102–103 m. F Vertical section of lithofacies Sm with irregularly aligned Skolithos isp. (arrows); well Dh1, depth 123–124 m. G Vertical section of Teichichnus isp. (Te) in lithofacies Hm; well Dh3, depth 68.7–68.9 m. H Meniscoid laminated structures (?equilibrichnia) in lithofacies H: a shallow cup-shaped structure with sharp irregular margins (A) and a deeper, post-shaped laminated structure with frayed muddy margins (B); well Dh1, depth 113–114 m. I Sand-dominated lithofacies H with strongly bioturbated (B) and unbioturbated (U) parts, showing vertical sections of cup-shaped trace fossil with differing sizes and pattern (CS), a funnel-shaped fugichnion or equilibrichnion (f), a trace fossil resembling Cylindrichnus isp. (A), and a bilobate repichnion considered to be Gyrochorte isp. (G); well Dh1, depth 113–114 m

Table 2 Taxonomy of bioturbation structures in the Dalkjegla Mb of the Carolinefjellet Fm in the study area, with ichnogenera ordered alphabetically

The occurrence and stratigraphic distribution of particular ichnotaxa and trace-fossil assemblages is generally related to lithofacies. For example, the ichnotaxa found in the thick sandstone beds of lithofacies Sp, Shs and Sr are different from those in lithofacies Sm or Hm. These differences entail variation in ichnofabric. An ichnofabric dominated by Phycosiphon-like burrows (Phycosiphon incertum, ?Phycosiphon isp., Helminthopsis isp.) is characteristic of lithofacies Sm (Figs. 7C, I, J, 8I, J) and locally Mm/Ml, and of the bioturbated divisions of lithofacies H (Fig. 7H). An ichnofabric dominated by Asterosoma, locally accompanied by Teichichnus, Thalassinoides and plug-shaped equilibrium structures (cf. Bromley and Uchman 2003), characterizes the top parts of the thick sandstone beds of lithofacies Sp, Shs and Sr, subsequently covered with the mudshales of lithofacies M (Fig. 9B, D, F). Bedding surfaces in the outcrop show also ?Aulichnites isp., Gyrochorte comosa, Palaeophycus sulcatus, ?Phycodes isp. and Thalassinoides suevicus (Fig. 7G, H). A package of sandstone lithofacies 20–26 m above the base of the formation in well cores (Fig. 3) shows an ichnofabric dominated by Skolithos, Palaeophycus isp. and Arenicolithes burrows (Figs. 10C, 11F).

The studied succession as a whole bears a mixed Cruziana–Skolithos ichnofacies (cf. Seilacher 1964, 1967, 2007; Bromley 1990; MacEachern et al. 2007a). The trace fossil suite in the above-mentioned sandstone package in well profiles represents the Skolithos ichnofacies, with the highly burrowed (scrambled) sediment indicating a distal variety of this ichnofacies. Trace fossils in the remaining part of the succession represent the Cruziana ichnofacies, which varies between the following two suites:

  • A proximal to archetypal variety of the Cruziana ichnofacies (MacEachern et al. 2007a, c) dominated by Asterosoma isp., Cylindrichnus isp., ?Phoebichnus isp. and Teichichnus isp., often accompanied by Schaubcylindrichnus coronus, Rosselia isp., Palaeophycus heberti, Thalassinoides ?div. isp., Rhizocorallium (?Taenidium), Diplocraterion div. isp., Conichnus div. isp., plug-shaped equilibrium structures and undetermined cup-shaped burrows (Table 2).

  • A distal variety of the Cruziana ichnofacies (MacEachern et al. 2007a, c) dominated by Phycosiphon-like trace fossils (Phycosiphon incertum, ?Phycosiphon isp., ?Helminthopsis isp.) with some Planolites isp., Diplocraterion div. isp., plug-shaped equilibrium structures, ?Lingulichnus isp., Siphonichnus isp. and rarely some other ichnotaxa.

The first suite occurs in the top parts of medium to thick sandstone beds overlain by mudshale and in some packages of the heterolithic deposits of lithofacies H. The second suite occurs in lithofacies Sm and in some beds of lithofacies H.

Notably, both the Skolithos and Cruziana ichnofacies here show quite specific composition in comparison to that commonly reported from similar sublittoral deposits in other regions (cf. Buatois and Angriman 1991; MacEachern and Pemberton 1992). The Skolithos ichnofacies differs from its ‘classic variety’ by nearly lacking such trace fossils as Ophiomorpha, Bergaueria, Taenidium and Macaronichnus (cf. MacEachern et al. 2007a, 2012; Buatois and Mangano 2011), whereas the Cruziana ichnofacies lacks its most characteristic arthropod tracks and Zoophycos burrows (cf. Buatois and Mangano 2011). Moreover, the facies-crossing ichnogenus Chondrites, so commonly reported from such deposits, has not been evidenced in the present case. Its absence is particularly remarkable in the lower part of the succession, in the blackish to brownish dark-grey mudshales of lithofacies Mm overlying the non-bioturbated basal lithofacies Ml, where other trace fossils begin to appear, such as Phycosiphon-like burrows and Schaubcylindrichnus freyi.

Lithofacies associations

The lithofacies (Table 1) have been recognized to form four main associations, labelled FA1 to FA4, which range from sand-rich to muddy and are vertically alternating with one another in the stratigraphic succession (Fig. 3). The extensive outcrop section of gently-inclined deposits and the two pairs of closely-spaced wells (Fig. 1C) show also considerable lateral changes over the distances of a few hundred metres, with some lithofacies pinching out and with one lithofacies association passing laterally into another.

The lithofacies associations and their ichnofauna assemblages are described and interpreted below. The lithofacies associations are thought to represent different inner to outer shelf zones and for simplicity are given interpretive environmental labels, but their descriptions are separated from interpretations in the text. The stratigraphic alternation of lithofacies associations (Fig. 3) is attributed to the lateral shifting of shelf zones in response to relative sea-level fluctuations accompanied by morphodynamic changes of the shoreface–shelf profile (cf. Hampson 2000; Hampson and Storms 2003; Grundvåg et al. 2021).

FA1: Lower shoreface deposits

The deposits of lithofacies association FA1 occur several times in the stratigraphic succession (Fig. 3), forming packages up to 6 m thick, enveloped by the mud-richer association FA2 and passing laterally into the latter. Their boundaries are mainly gradational, marked by a rapid upward change in the thickness and relative proportion of sandstone and mudstone beds, but the basal contact of FA1 in some cases is sharp and recognizably erosional.

Description

This lithofacies assemblage (Figs. 3, 4) consists mainly of lithofacies Sp and Shs, with drapes or intercalations of lithofacies Sr and subordinate thin interbeds of lithofacies Sm, Hm and/or Mm, rarely Ml. Lithofacies Shs volumetrically dominates, as the sandstone beds with hummocky stratification are the thickest. Beds of sandstone lithofacies tend to be amalgamated by erosion, whereby the interbeds of muddy lithofacies are commonly truncated, discontinuous or virtually removed. Mudclasts, up to 7 cm in length, occur scattered along internal erosional surfaces.

Sandstone beds in the outcrop are generally extensive sheets, but their geometry varies from tabular to irregularly mounded or lenticular (Fig. 3; see also Grundvåg et al. 2021). Some beds show a uniform thickness over lateral distances of several tens of metres, before thinning markedly or pinching out within a few metres. Other sandstone beds, particularly the thickest ones, show irregular mounds in the form of simple low-amplitude domes (isotropic to anisotropic hummocks) or compound, broader and thicker (≤ 1.5 m) domes with a clinoformal lateral stacking of the successive units of lithofacies Shs and subordinate Sr, mainly towards the W or WSW (Fig. 3). The amalgamated clinoforms are convex-upwards to sigmoidal, inclined at 5–20°. Most of the sand mounds have erosional bases with an irregular low relief (< 10 cm), but reaching a relief of up to 50 cm where forming a sharp lower boundary of FA1 package (Fig. 3). There are also solitary channel-like scours (Fig. 3F), up to 1.5 m deep and a few tens of metres wide, trending towards the W or WSW and filled with vertically accreted packages of lithofacies H or obliquely inclined (10‒15°) packages of lithofacies Sp and Sr (see Birkenmajer 1966, Figs. 34).

Animal burrows in lithofacies association FA1 are relatively uncommon and limited mainly to infrequent bioturbated horizons, generally corresponding to the mudshale-covered top parts of sandstone beds or their amalgamated packages. The degree of bioturbation thus varies from none to high (BI = 0 to 6) on the thickness scale of a bed or a bed package. The following ichnotaxa have been recognized: Arenicolithes isp., Asterosoma isp., Paleophycus tubularis, Palaeophycus isp., ?Phoebichnus isp., Planolites isp., Rosselia isp., Schaubcylindrichnus coronus, Skolithos isp. and Teichichnus rectus, with accompanying ?Helminthopsis-like traces and sporadic Ophiomorpha cf. rudis and Diplocraterion isp. (Table 2).

The FA1 packages are generally hosting a proximal variety of the Cruziana ichnofacies (cf. MacEachern et al. 2007a), dominated by either Asterosoma or undefined scrambled ichnofabric overprinted by other taxa as deeper tiers. Only one sandstone package, about 1 m thick and located 20–26 m above the base of Carolinefjellet Formation in the wells (Fig. 12), hosts a distal variety of the Skolithos ichnofacies (cf. MacEachern et al. 2007a) dominated by suspension-feeder traces and comprising Skolithos isp., Arenicolithes isp., Palaeophycus tubularis, Palaeophycus isp. and sporadic Ophiomorpha and Diplocraterion (Figs. 10C, 11F).

Fig. 12
figure 12

Well-log correlation panel showing the interpretive depositional architecture of the Dalkjegla Mb of Carolinefjellet Fm in the study area (Fig. 1C) and the vertical distribution of its lithofacies and ichnofacies associations. Legend as in Fig. 3. The panel shows the lower part of the Carolinefjellet Fm and its basal contact with the Helvetiafjellet Fm. Note the stratigraphic repetition due to tectonic thrusting in logs Dh1 and Dh2. The right-hand continuation of the figure shows the topmost part of the succession in wells Dh1 and Dh2, lacking in wells Dh3 and Dh4; note again the slight stratigraphic repetition due to thrusting

Interpretation

The assemblage of sandstone lithofacies Sp, Shs and Sr indicates the action of littoral waves with variable orbital velocities, punctuated by storm combined-flow events (Table 1). The clinoformal stacking of amalgamated sandstone units suggests broad subaqueous sand bars formed by series of high frequency storms, with the geostrophic currents deviated from shoreline by the Coriolis effect (Walker 1984b) and with the initial hummocks instigating subsequent deposition in their hydraulic shadow. Midtgaard (1996) described similar sandbodies, accreted seawards at ~ 70° away from shoreline, from the Lower Cretaceous mid-shelf deposits in western Greenland. The channel features, limited to FA1, were originally attributed to tidal currents (Birkenmajer 1966), but this lithofacies association represents sedimentation dominated by waves and storm-generated currents. These features are likely storm-current elongate scours (cf. Bentley et al. 2002, Fig. 19) or bypass troughs formed by storm-boosted rip currents (cf. Gruszczyński et al. 1993; Stone et al. 1995; Mathers and Zalasiewicz 1996), filled in directly by the storm and/or subsequent fairweather sedimentation.

The scarcity of burrows in sandstone beds implies seafloor conditions of high and quasi-perennial sediment mobility, and hence deposition mainly above the fairweather wave base. The numerous internal truncation and amalgamation surfaces imply that a considerable part of fairweather sand (lithofacies Sp and Sr) was probably removed by erosive storms and transferred seawards. The interbeds of muddy facies and isolated bioturbation horizons represent episodes when the wave base stayed above the seafloor. The evidence as a whole indicates deposition between the maximum and mean depth of fairweather wave base, which means a lower shoreface zone (Clifton 1981; Brenchley 1985; Tillman 1985; Reading and Collinson 1996; Hampson 2000).

Deposition below the mean fairweather wave base is consistent with the occurrence of Cruziana ichnofacies and with the bioturbation horizons indicating distinct time-windows for benthic colonisation (Pollard et al. 1993; Goldring et al. 2004, 2007). The solitary sandstone package with Skolithos ichnofacies (Fig. 12) indicates a highly mobile sandy substrate and low sedimentation rate (cf. MacEachern et al. 2007a; MacEachern and Bann 2008), which would suggest deposition around the mean fairweather wave base in a middle shoreface zone and imply the maximum fall of relative sea level recorded within the sedimentary succession. This episode might either signify the greatest shoreline advance or represent brief opportunistic colonization of lower shoreface by Skolithos ichnofauna following an exceptionally high delivery of sand by a series of strong storms (e.g., Vossler and Pemberton 1988, 1989; Pemberton et al. 1992b, 2012; Pemberton and MacEachern 1997). In either case, the sparsity of bioturbation in this sandstone package would reflect the short duration of inter-storm benthic colonisation windows (MacEachern and Pemberton 1992; Dashtgard et al. 2012; Pemberton et al. 2012).

FA2: Proximal offshore-transition deposits

Lithofacies association FA2 shows a comparable thickness proportion of alternating sandy and muddy deposits, forming packages about 1 to 4 m thick. They underlie and overlie lithofacies association FA1 or occur in isolation from the latter (Fig. 3). The outcrop section also shows FA1 passing laterally into FA2 by the lateral thinning or pinch-out of sandstone beds and an increase in the relative proportion of muddy deposits.

Description

This lithofacies assemblage (Figs. 3, 4) consists of the heterolithic deposits of lithofacies H and mudshale lithofacies Mm interspersed with discrete sheet-like sandstone beds showing various combinations of lithofacies Sp, Shs and Sr (Table 1; Fig. 5B‒D). Subordinate are beds of lithofacies Sm or intervals of Hm. The muddy and sandy beds are up to 50 cm thick, but mainly 10–20 cm. Sandstone beds have mainly flat bases, sharp to slightly erosional, whereas their tops in the outcrop section show broad undulations on a lateral scale of several to a few tens of metres. The undulations render some of the thinner beds discontinuous, split into irregular broad lenses (probably 3D patches). Erosional amalgamation of sandstone beds is uncommon and only local.

Bioturbation in lithofacies association FA2 is more common than in FA1 and ranges from isolated to clustered burrows, but occurs mainly in the muddy facies and reaches its highest intensity in distinct horizons of unknown lateral extent (observation from well cores). The trace fossils include Asterosoma isp., Paleophycus tubularis, Palaeophycus isp., ?Cylindrichnus concentricus, Teichichnus rectus, Thalassinoides isp., Rosselia isp., Conichnus div. isp., Planolites isp., Arencolites isp., ?Helminthopsis isp., ?Phycosiphon incertum and various plug-shaped equilibrium structures, with sporadic Diplocraterion ?div. isp., Rhizocorallium isp. (?Taenidium isp.), ?Lingulichnus isp., Siphonichnus isp., ?Phoebichnus isp., ?Phycodes isp. and ?Chondrites targionii (Table 2). In addition, Planolites montanus, P. beverleyensis, Palaeophycus sulcatus, Gyrochorte comosa and ?Aulichnites isp. were found on sandstone bedding surfaces in the outcrop (Fig. 9A, B, G).

The bioturbation intensity in FA2 varies from none to high (BI = 0 to 6). Most common are stratigraphic intervals with no or little bioturbation (BI = 1), hosting isolated burrows (Figs. 6A, E, F, H, I, 7D). Similarly varied is ichnofabric, dominated by horizontal burrows. The rare strongly bioturbated horizons show a scrambled fabric intersected by Asterosoma and/or Teichichnus, rarely other ichnotaxa. The less bioturbated deposits show a full range of isolated burrows, rarely intersecting one another.

Interpretation

Compared to FA1, this lithofacies association contains thinner sandstone beds and a higher proportion of muddy facies, which suggests a more distal depositional zone relative to palaeoshoreline. The discrete sandstone sheets with wave-generated and combined-flow structures indicate sand emplacement by brief events, considered storms. The thicker beds of muddy facies and greater abundance of bioturbation indicate considerably longer periods of sand-starved seafloor conditions below the fairweather wave base. Such conditions characterize the lowermost shoreface to upper offshore-transition zone, at a water depth range from around the maximum fairweather wave base to the mean storm wave base, where sand is delivered and spread mainly by storms (Howard and Reineck 1981; Kreisa 1981; Reading and Collinson 1996; Hampson 2000). The lenticularity of sandstone beds on a lateral scale of up to a few hundred metres (Nøttvedt and Kreisa 1987; Grundvåg et al. 2021) probably reflects a patchy style of the spatial distribution of sand by storms (cf. Bentley et al. 2002; Keen et al. 2004).

The ichnofauna assemblage in FA2 indicates environmental conditions hospitable to a wide range of benthic animals, but also reflects the ecological stress imposed by episodic sand emplacement (cf. Pemberton et al. 1992b; Pemberton and MacEachern 1997; MacEachern et al. 2007b; Bann et al. 2008; MacEachern and Bann 2008). The predominance of horizonal burrows indicates a deposit-feeding benthic fauna (cf. Pemberton et al. 2012), which implies relatively long periods of low-energy bottom water conditions. The scarcity of suspension-feeding fauna indicates a quick burial of the episodically emplaced sand layers by ubiquitous mud, which disfavoured seafloor colonization by organisms preferring a sandy substrate (Pemberton et al. 2012). The event sedimentation and variable time-windows for benthic colonisation may explain the varying bioturbation degree of the deposits (Pemberton et al. 1992b, 2012).

FA3: Distal offshore-transition deposits

Lithofacies association FA3 contains a much higher thickness proportion of muddy lithofacies, with sandstone beds generally thinner and finer-grained. These mud-dominated deposits form packages up to ~ 20 m thick, both underlying and overlying lithofacies association FA2 (Fig. 3). The contacts of these two associations are conformable and transitional, marked by changes in the relative proportion of sandy and muddy lithofacies. In the outcrop section, FA2 is also passing laterally into FA3 within a few hundred metres.

Description

The assemblage FA3 (Fig. 3) consists mainly of lithofacies Mm, Ml and Hm, interspersed with thin sandstone sheets of lithofacies Sr (Table 1, Fig. 6F, I). The sandstone beds are very fine- to fine-grained and locally up to 20 cm thick, but are mainly thinner than 5 cm and commonly discontinuous, composed of isolated lenses with a lateral extent of 10 cm to several metres. The boundaries of sandstone beds are sharp, but the bases are seldom recognizably erosional and the tops show well-preserved ripple forms. Ripple crests are trending mainly SE, and the ripple cross-laminae sets in some beds show mud drapes, occasionally rich in carbonaceous plant detritus.

FA3 in its lowest stratigraphic occurrence (Fig. 12), particularly in its basal part, is distinctly less burrowed and shows a much lower diversity of trace fossils than in the higher occurrences. Moderate to high bioturbation occurs in thickness intervals of 1 m to a few metres, particularly in lithofacies Hm. There are isolated horizons of intense burrowing as well as random solitary burrows. High bioturbation (BI = 5–6) prevails in the FA3 package at 40–55 m above the base of the formation.

Trace fossils include Helminthopsis isp. and some other Phycosiphon-like ichnotaxa, Asterosoma isp., Planolites isp., Thalassinoides isp., Teichichnus rectus, Palaeophycus isp., Siphonichnus isp., ?Cylindrichnus concentricus and Rosselia isp., accompanied by sporadic Diplocraterion ?div. isp., ?Lingulichnus isp. and undetermined cup-shaped, funnel-shaped and fringed U-shaped burrows (Table 2). Ichnofabric is dominated by Phycosiphon-like burrows, locally with Asterosoma, Planolites, Thalassinoides and Teichichnus. Asterosoma is characteristic of the sand-richer parts of FA3 packages. The lowermost stratigraphic package of FA3 (Fig. 12), particularly at its transition from the underlying FA4, shows only ?Phycosiphon isp., Schaubcylindrichnus freyi and some sand-filled thin pipes reminiscent of Planolites. Overall, the FA3 deposits bear a trace-fossil assemblage of the Cruziana ichnofacies, mainly its distal variety (cf. Savrda et al. 2001; MacEachern et al. 2007a).

Interpretation

The differences in lithofacies and ichnofauna and the stratigraphic relationship between lithofacies association FA2 and mud-dominated FA3 indicate that the latter is a more distal seaward equivalent of the former. The isolated thin sandstone beds are thought to be distal tempestites, deposited at water depths where only the strongest storm-generated currents were delivering fine sand and where the sand was fully reworked by storm waves with mainly low near-bottom orbital velocities (Komar and Miller 1975; Clifton 1981; Clifton and Dingler 1984). Ripple mud drapes indicate brief rhythmic detachments of the wave base from the seafloor, which may reflect pulsating storms or the impact of tides on the wave base during storms lasting for several days. The pinch-and-swell geometry of thin sandstone beds, on the scale of ripples or small groups of ripples, indicates sand-starved seafloor conditions. The lenticularity of thicker beds, on a lateral scale of several tens to hundreds of metres (Nøttvedt and Kreisa 1987), reflects the inherently patchy pattern of spatial sand distribution by storms (cf. Bentley et al. 2002; Keen et al. 2004).

The evidence as a whole indicates deposition in the distal offshore-transition zone, below the mean storm-wave base (Howard and Reineck 1981; Reading and Collinson 1996; Hampson 2000). This interpretation is consistent with a distal Cruziana ichnofacies, dominated by traces of deposit-feeding, dwellings and grazing fauna. The general lack of suspension-feeders reflects the sparsity and infrequent availability of sandy substrate.

The poor assemblage of ichnofauna in the stratigraphically lowest package of FA3, at its transition from the underlying FA4, apparently reflects a benthic ecological change from an unhospitable (anoxic?) muddy environment to a more hospitable environment with episodic delivery of sand and improved seafloor oxygenation. A similar ichnological case was reported, for example, by Savrda et al. (2001) from the Pleistocene shelf of New Jersey.

FA4: Offshore deposits

Lithofacies association FA4 occurs only in the basal part of the succession (Figs. 3, 12), where these muddy deposits form a unit ranging in thickness from ~ 2 m in wells Dh1/Dh2 to 5 m in wells Dh3/Dh4 (Figs. 1C, 3) and to nearly 15 m in the southern outcrops more distal from the inferred palaeoshoreline (Midtkandal et al. 2016). This muddy unit is underlain by a solitary erosional conglomeratic sheet (wells Dh1 to Dh3) or locally by a correlative erosional bypass surface (well Dh4). The conglomeratic layer, although laterally discontinuous, is a regionally widespread feature (Birkenmajer 1966) considered in mapping to mark the base of the Carolinefjellet Formation (Parker 1967; Nagy 1970; Mørk et al. 1999). Similarly widespread is the basal mudshale unit, which apparently extends beyond the coastal outcrops in SE Spitsbergen (Nemec et al. 1988; Århus 1991).

Description

This lithofacies association consists chiefly of the blackish grey, parallel-laminated clayey to silty mudshales of lithofacies Ml (Fig. 6F) interspersed towards the top with sporadic thin (< 1 cm) and mainly discontinuous, pinch-and-swell sheets of very fine-grained sandstone lithofacies Sp and/or Sr. Ripple cross-laminae sets in lithofacies S4 occasionally show carbonaceous mud drapes. The underlying polymict conglomeratic sheet consists of poorly sorted lithofacies C, generally no more than 10 cm thick, alternating laterally between massive and faintly planar stratified (Table 1, Figs. 3, 5A, 12). The conglomerate layer contains intraformational mudclasts and has an uneven erosional base and a flat top locally covered with a thin discontinuous layer of lithofacies Sr.

Lithofacies association FA4, although the most muddy (Table 1), generally lacks bioturbation. Sporadic tiny sand spots (~ 1 mm in diameter) occur on the non-slabbed core surfaces, but it is unclear if these are some sand-filled very thin burrowing pipes or rather floating tiny aggregates (intraclasts) of very fine-grained sand.

Interpretation

The unit of muddy lithofacies Ml intercalated with sporadic very thin sheets of finest-grained sand indicates deposition in an offshore zone, where only some of the strongest storms would occasionally spread sparse sand (Howard and Reineck 1981; Reading and Collinson 1996; Hampson 2000). Non-bioturbated black mudshales may indicate either a prolonged episode of sand-starved anoxic seafloor conditions or a high-rate deposition of ubiquitous mud rich in organic carbon. The widespread occurrence of these muddy deposits directly above the gravel-lain erosional surface indicates an abrupt ultimate drowning of the retreating fluvio-deltaic system of the Helvetiafjellet Formation (Fig. 2C). Therefore, the discontinuous basal conglomerate sheet has been interpreted in the regional stratigraphy to be a transgressive lag, with the sand and gravel fraction derived by erosional reworking of the fluvio-deltaic substrate (Nemec et al. 1988; Gjelberg and Steel 1995).

The basal muddy unit (Fig. 12), recording the global episode OAE1a (Midtkandal et al. 2016, Fig. 3), was deposited in environmental continuity with the increasingly anoxic Svalbard lagoonal embayment of the topmost Helvetiafjellet Formation (Fig. 2C, D). The retreating broad Barremian delta of the Helvetiafjellet Formation (Gjelberg and Steel 1995) accumulated abundant organic-rich black mud (TOC 2‒5%) in its interdistributary bays and back-barrier lagoon (Fig. 2C; Nemec et al. 1988; Nemec 1992). The abrupt marine invasion probably resuspended the latest of these deposits and spread anoxia unhospitable to benthic fauna in the resulting Svalbard embayment (Fig. 2D), while increasing its bulk water energy, sparse sand delivery and aeration. The mud deposition at this stage may have involved resedimentation of bottom fluid mud, flowing in accordance to the local seafloor gradients (Allison et al. 2000; Traykovski et al. 2000; Sheremet et al. 2005; Ichaso and Dalrymple 2009). Large volumes of mobile seafloor mud might thus be gravitationally redeposited within the Svalbard embayment avoiding bioturbation (Mehta 1991; Trowbridge and Kineke 1994). The volumetric concentration of clay aggregates in fluid mud may reach 95% (Wells 1989), which allows the flow to carry silt particles and possibly tiny clumps of fine sand grains in its rigid-plug zone (Baas et al. 2009). Thin rhythmic sets of graded silty to clayey mud layers (Fig. 6F) may be due to a pulsating downward flux of sediment settling from storm-generated suspension plumes, caused by the depletion of their unstable density gradient by deposition-driven convection (Kerr 1991; Nemec 1995).

Discussion

Shelf sedimentation conditions

The present study concurs with the regional reconstructions postulating Barremian major palaeogeographic changes in Svalbard (Figs. 1D, 2), with the Aptian–Albian sourcing shoreline in the midst of Svalbard and trending approximately NW–SE (Birkenmajer 1966; Århus 1991; Dypvik et al. 2002; Blakey 2011). The opposite shoreline of the Svalbard Aptian–Albian seaway was in NE Greenland (Fig. 2D), but its relative regional position had been changed by the Palaeogene opening of the North Atlantic and Arctic Ocean.

The present study is limited to the Dalkjegla Member of the Carolinefjellet Formation (Fig. 1D), but a similar range of lithofacies characterizes its higher stratigraphic members (Grundvåg et al. 2017, 2019, 2021). The sand-rich lithofacies association FA1 of the Dalkjegla Member is considered as deposited within the maximum reaches of a fairweather wave base, but the majority of the isolated or amalgamated sandstone beds in FA1 and all the sandstone sheets in lithofacies associations FA2 to FA4 are tempestites—as originally postulated by Nøttvedt and Kreisa (1987). The large-scale lateral discontinuity of tempestites apparently reflects the inherently patchy style of sand distribution by storms on an inner shelf (cf. Bentley et al. 2002; Keen et al. 2004). The mounded geometry of thick tempestite beds, mainly composite (Fig. 4B‒D), is thought to reflect large hummocks and their common stacking into some kind of storm bars by vertical and/or lateral accretion, the latter directed variably seawards or landwards (Fig. 4E, F; Jelby et al. 2020; Grundvåg et al. 2021).

The abundance and considerable thicknesses of both solitary and amalgamated sandstone tempestites in the study area, several tens of kilometres away from the inferred northeastern palaeoshoreline, imply high-frequency severe storms with a magnitude commonly exceeding that of the world’s largest modern storms. For comparison, the largest modern hurricane storms in the Gulf of Mexico deposited relatively thin and very fine-grained sand layers on the shelf, only locally reaching a thickness of 10‒20 cm (Snedden et al. 1988; Bentley et al. 2002; Keen et al. 2004, 2006; Allison et al. 2005). The single sandstone tempestites in the present case are up 100 cm thick and their amalgamated stacks are up to 350 cm (cf. Grundvåg et al. 2021). Sandstone tempestites of similar thickness have also been reported from other stormy Cretaceous shelves (Burgeois 1980; Duke 1985; Århus 1991; Arnott 1993b; Midtgaard 1996), which seems to support the notion that a major global climatic change triggers extreme weather phenomena (Easterling et al. 2000; Meehl et al. 2000; Rosenzweig et al. 2001; Francis and Vavrus 2012). This notion has important stratigraphic palaeoenvironmental implications in the context of the Earth history of cyclic changes between icehouse and greenhouse climatic conditions (Scotese et al. 2021). It also serves as a predictive warning for the impending modern climate change, as shown by the recent increase in the frequency and magnitude of Atlantic hurricanes and Pacific typhoons (Johnson et al. 2018, 2020).

As pointed out by Leckie and Krystinik (1989), the vast majority of reported ancient shelves show storm-generated compensation currents flowing orthogonally away from the shoreline, rather than being strongly deviated subparallel to the shore by the Coriolis effect and turned into classic theoretical geostrophic currents (Walker 1984b). The Early Cretaceous shelf of Svalbard and NE Greenland (Midtgaard 1996), with the storm-driven currents directed at ~ 70° away from the inferred shoreline, would be among the relatively few ancient cases where typical geostrophic currents have been recognized (cf. Duke 1985). Could the Svalbard shelf be somehow unique? Its high latitude and palaeogeographic setting must be taken into account. First, the horizontal deflection of storm compensation currents by the Coriolis effect is greatest near the pole and smallest at the equator. Second, this subarctic shelf was within a relatively narrow seaway, where the axial tidal currents—enhanced by confinement—could impose their action on the storm compensation flows.

It is by no means certain that the iron ooids found in the Dalkjegla Member are of lagoonal origin, as suggested by Mutrux et al. (2008), even though a dysoxic brackish lagoon formed in Spitsbergen during the Barremian marine transgression (Fig. 2C; Nemec 1992; cf. Boyd 2010). The Svalbard area in the Early Cretaceous was under the influence of the HALIP volcanic activity (Maher et al. 2004). Iron ooids could have formed by exhalative hydrothermal fluids rising to the seafloor (Heikoop et al. 1996) or by the alteration of volcanic ash (Sturesson et al. 1999, 2000), with possible mineralogical changes by diagenetic reactions. The iron ooids might thus not derive from the non-evidenced, hypothetical multiple coastal lagoons postulated by Mutrux et al. (2008), and could form in the early Aptian nearshore or even offshore-transition zone affected by frequent storm wave action (Collin et al. 2005).

The stratigraphic organization of the Carolinefjellet Formation into five members (Fig. 1D) may reflect 3rd-order eurybatic sea level changes (sensu Haq 2014). The Carolinefjellet Formation has an estimated time span of about 20 Ma. Although its mud-rich Innkjegla and Zillerberget members are somewhat thicker than the sand-richer Dalkjegla, Langstakken and Schønrockfjellet members (Fig. 1D), these latter members abound in the internal surfaces of erosion and seaward sediment bypass. The five members might thus have had a roughly comparable time span of about 4 Ma.

The stratigraphic alternation of lithofacies associations FA1‒FA4 within the Dalkjegla Member (Figs. 3, 4A, 12) may then be due to eurybatic 4th-order sea level changes, with a mean time span of ~ 0.45 Ma (cf. Plint 1991; Hampson and Storms 2003; Haq 2014), driven astronomically and affecting the shelf wave climate. Seismic and interpretive correlation sections (Grundvåg et al. 2017, 2021) show large-scale morphodynamic adjustments of the inner shelf profile with a repetitive low-angle (< 1°) clinoformal stacking of deposits, which may reflect eurybatic 4th-order sea level changes, expectedly accompanied by changes in the sediment supply and wave climate (Clifton 1981; Hampson 2000; Hampson and Storms 2003). The vertical changes of lithofacies associations (Figs. 4A, 12) may thus reflect a steepening of the shoreface–shelf profile during its early progradation and a maintenance of uniformly gentle profile during the later progradation (Swift et al. 1987; Hampson and Storms 2003). The seasonal peaks of stormy wave climate would keep the progradation in check.

The late Aptian–Albian net eustatic sea-level rise did not exceed 50 m (Fig. 1D; Haq 2014), and hence the succession of open-marine shelf deposits with a preserved compacted thickness of at least 1200 m implies considerable syndepositional tectonic subsidence. Pulses of subsidence may have reduced the record of eustatic sea level falls to an apparent normal regression (cf. Jervey 1988; Messina et al. 2007), with a gradual upward change from FA2 to FA1. A sharp erosional base of FA1 indicates cases of a forced regression non-compensated by subsidence and representing a bypass hiatus. Transgressive systems tracts FA1 → FA2 → FA3(→ FA4) imply depositional and highly aggradational transgressions, which can be attributed to the active tectonic subsidence with a high rate of sediment supply by coastal erosion and alongshore drift. The vertical changes of lithofacies associations may thus have important implications for the palaeoshoreline trajectory, reflecting an interplay of eustatic sea-level changes, tectonic subsidence and sediment supply (Hampson 2000; Hampson and Storms 2003; Haq 2014; Grundvåg et al. 2021).

Shelf ichnofauna ecology

Except for two minor stratigraphic intervals, the studied shelf succession shows a trace fossil assemblage of the Cruziana ichnofacies in its proximal to distal expression. However, it should be noted that the succession spans a relatively narrow bathymetric range and a limited range of shelf subenvironments from lower shoreface to offshore zone. The two exceptions are the non-bioturbated black mudshales in the basal part of the succession and the occurrence of a distal expression of the Skolithos ichnofacies in a thin package of amalgamated sandstone beds in the lower mid-part of the succession (Fig. 12). The black shales formed in anoxic conditions linked to the OAE1a flooding (Midtkandal et al. 2016), whereas the nature of the sandstone package with Skolithos ichnofacies is less clear. Considering its small thickness and isolated occurrence at the turnabout level of the succession regressive to transgressive trend (Fig. 12), this unit may represent a brief encroachment of middle shoreface conditions at the regression culmination.

The intensity of bioturbation varies in the succession profiles, but similar lithofacies associations show a comparable intensity throughout the stratigraphic succession (Figs. 3, 12), which implies an ichnofauna ecology controlled by the seafloor hydraulic regime and oxygenation, and hence mainly by the wave climate and eurybatic sea-level changes. The forcing mechanism could thus be climate combined with eustasy (Fig. 1D) and regional tectonism. The intensity of bioturbation was decreasing during periods of both elevated and lowered sea level. A rise of sea level would reduce the impact of storms on the seafloor, resulting in poorer sand supply and benthic water aeration, while allowing polar water incursions. A fall of sea level would increase the impact of storms and benthic aeration, while creating unstable, mobile sandy substrate. The optimal seafloor conditions for a markedly increased benthic fertility are recorded by the most borrowed stratigraphic interval 40–55 m above the formation base, dominated by FA3 deposits (Figs. 3, 12). This interval represents specific seafloor conditions attributed to a marine transgression in harsh subpolar wave climate, with increased bottom oxygenation and decreased frequency of sand incursions from the receding shoreline.

Some lateral variation in the bioturbation intensity was due to the seafloor uneven depositional morphology (cf. Bentley et al. 2002; Hampson and Storms 2003; Keen et al. 2004; Grundvåg et al. 2021). Storm-accreted elevated areas had a better-aerated and steadily sandy substrate, while the least aerated local depressions with an unsteady heterogeneous substrate were less hospitable to benthic fauna. The lack of bioturbation in the muddy FA4 (Figs. 3, 12) reflects probably a sparsity of seaward nutrient supply, which contrasts with the generally fertile prodelta environments (Díaz et al. 1996; Ayranci et al. 2014) and supports the notion of a storm-dominated, non-deltaic inner shelf (cf. Hamblin and Walker 1979; Hampson 2000; Hampson and Storms 2003; Grundvåg et al. 2021). Small deltas undoubtedly existed, but their role was probably limited to the feeding of a wave-dominated shoreline subject to alongshore sediment drift.

The stratigraphic succession shows a trace-fossil assemblage of the Cruziana ichnofacies in its distal to proximal expression. The documented variation in its species composition and bioturbation intensity is limited to a relatively narrow bathymetric range and a narrow environmental systems tract spanning lower shoreface to offshore zone. The only recognizable exceptions are the non-bioturbated black shales of FA4 in the basal part of the succession and the thin correlative sandy package with a distal expression of the Skolithos ichnofacies ~ 20 m above the succession base (Figs. 3, 12). The shale unit is attributed to deposition in a dysoxic lagoon transformed into anoxic embayment by the early Aptian OAE1a (Fig. 2C, D; Midtkandal et al. 2016), and the sandstone unit to a brief encroachment of the sandy mid-shoreface zone at the culmination of the shelf regressive trend (Fig. 12).

Noteworthy in the Cruziana ichnofacies here is the lack of arthropod trackways, its denominative component, which may suggest an absence of epibenthic arthropods in the sedimentary environment or be simply an artefact of observations made mainly on drilling cores, where the recognition of epichnial traces and horizontal burrows is difficult. More certain is the lack of such ichnotaxa as Ophiomorpha, Chondrites and Zoophycos, which may indicate seafloor ecological conditions unhospitable to certain benthic animals. Despite the global greenhouse climate, the Svalbard shelf area was subject to the invasions of cold polar water (Wilkinson and Riding 2007; Herle et al. 2015; Vickers et al. 2019). As noted by Cadée (2001), the present-day Arctic and temperate coasts lack Callianassids and crabs, which are producers of Ophiomorpha, Thalassinoides and Spongeliomorpha traces. According to Dworschak (2000), the Thalassinideans today are unknown from nearshore deposits at latitudes higher than 70 ºN and 50 ºS. These observations had led Goldring et al. (2007) to a tentative suggestion that the occurrence of pellet-lined burrows, such as Ophiomorpha, in the shoreface, lower foreshore and estuarine sandy deposits is restricted to warm, tropical or subtropical environments. The lack of Ophiomorpha, Chondrites and Zoophycos in the present case may thus be due to the specific conditions of a highly unsteady and episodically polar-influenced seafloor. Notably, abundant Ophiomorpha are known from the more stable Santonian–Campanian environment of western Greenland (Pedersen and Bromley 2006) and the Palaeocene environment of Spitsbergen (Svinth 2013). Likewise, both Chondrites and Zoophycos are known from the late Palaeozoic shallow glaciomarine deposits (Netto et al. 2012).

The black to dark grey shales with a high TOC content imply a low-oxygen sediment pore waters, which should generally be suitable for at least the occurrence of Chondrites. The lack of Chondrites, Ophiomorpha and Zoophycos in the present regional case was thus probably due to the benthic anoxia combined with unstable and insufficient water salinity in the Spitsbergen lagoon transformed into embayment (Fig. 2C, D) surrounded by land-draining river outlets.

Apart from the basal unit of FA4 (Fig. 12), the distribution and intensity of bioturbation in the sedimentary succession was apparently controlled mainly by the time span of colonisation windows and the magnitude of sand-emplacing storm events. The colonisation windows were expectedly a function of the storm frequency and sea level, whereas the thickness of tempestites depended on the storm magnitude. Powerful storms eroded the substrate and covered it with a sediment layer too thick to be burrowed in continuity with the underlying pre-storm deposit. The deposition of the basal black shales in Spitsbergen area (Fig. 2C, D) could similarly be affected by rapid en masse emplacement of fluid-mud layers.

The ichnofauna assemblages in the studied succession are generally consistent with their hosting lithofacies association and hence vary with the sedimentary environment conditions. The environmental changes in the study area commenced with a transgressive invasion of anoxic offshore environment and were followed by fluctuations between offshore transition and lower shoreface subenvironments, with a brief encroachment of sandy mid-shoreface conditions at the turnaround of the shelf general regressive to transgressive trend (Fig. 12). The short-term environmental fluctuations and related ecological changes on the Svalbard shelf were probably a result of the interplay of 4th-order eustatic cycles (Haq 2014) and the active rifting at the Barents Shelf margins (Ziegler 1988; Maher 2001; Torsvik et al. 2002; Worsley 2008).

Conclusions

The lithofacies associations and ichnofauna assemblages in the studied part of the Carolinefjellet Formation, Spitsbergen, record important in loco changes of environment conditions at the western margin of the Barents Shelf, controlled by both the Cretaceous global greenhouse climate and the shelf local opening to a harsh wave climate of the Arctic Ocean. The Aptian shelf subenvironments ranged from lower shoreface to offshore zone, with a brief encroachment of mid-shoreface conditions at the culmination of regressive aggradational stratigraphic trend. Sedimentological data indicate high-frequency storms, often exceeding the magnitude of largest modern hurricane events. While the individual lithofacies reflect short-term (annual?) weather changes in the shelf environment conditions, their varying associations record morphodynamic changes in the shelf topography driven by 4th-order eurybatic sea-level fluctuations. The formation’s five lithostratigraphic members, alternatingly sand- and mud-dominated, probably represent 3rd-order eurybatic sea-level changes.

The deposits are rich in ichnofauna, but the bioturbation degree varies between none (BI = 0) and total (BI = 6). Free of burrows are the basal transgressive conglomerate layer and the overlying package of black mudshales, the majority of thick sandstone beds, and some units of recurring heterolithic deposits. Pervasive total bioturbation prevails in the mudshales and heterolithic deposits located at 40–55 m above the formation base, which apparently recorded the shelf optimal fertility conditions for benthic fauna: a modest sedimentation rate with high nutrient supply and water circulation allowing for a stable and moderately dysoxic benthic environment.

Thirty-eight ichnotaxa have been recognized, mainly on core surface, with fourteen of them uncertain. The trace fossils include vertical, inclined and horizontal burrows, mostly fondichnia and domichnia, produced by mobile benthic organisms. The Aptian succession as a whole shows trace fossils assemblages of the Cruziana ichnofacies in its proximal to distal expression, with ichnotaxa impoverishment indicating bathymetry generally shallower than that of the archetypal Cruziana ichnofacies. Exceptions are a solitary sandstone package with the Skolithos ichnofacies in its distal expression and the basal non-bioturbated unit of black mudshales. The basal mudshales represent an end-Barremian dysoxic estuarine Spitsbergen lagoon invaded by the anoxic waters of the global event OAE1a. The sandy deposits with the Skolithos ichnofacies were formed close to the maximum fairweather wave base at the mid-Aptian culmination of a eurybatic shelf regressive trend.