Foraminifera in the glacial erratic rock Sternberger Gestein from northern Germany

Pleistocene glacial erratics are widely spread in northwestern Europe. They were sheared off from their original source by glacier ice and then transported by ice and meltwater over distances ranging from a few to 1000 kilometers (Smed & Ehlers, 2002). They can be found in large numbers along the Baltic coast and in the hinterland in gravel pits, cuts, and as individual rock pieces on the surface of the soil. Glacial erratics have been the subject of scientific study for over 300 years, aiming to clarify their genesis, origin and the glacial movements (Hesemann, 1975). The interest in glacial erratics is reflected in the existence of the Gesellschaft für Geschiebekunde (Society for the Study of Glacial Erratics) with its ∼300 members, magazines, and local groups (Smed & Ehlers, 2002).

Glacial erratics of magmatic or metamorphic origin are classified and dated based on their lithology and comparisons with potential source areas. As source areas may have similar lithological characteristics, the classification and dating may be difficult (Smed & Ehlers, 2002). Glacial erratics of sedimentary origin are classified lithologically and by facies. Some of them contain fossils which are used to clarify their origin and stratigraphic position (Smed & Ehlers, 2002). This study is part of an effort to depict and analyse the foraminiferal fauna found in glacial erratics of sedimentary origin. Previous studies have been published on Upper Cretaceous chalk marl smears on the Baltic coast and Eocene Heiligenhafener Kieselgestein (Hesemann & Ketelsen, 2017; Hesemann, 2020).

As early as 1711, the Sternberger Gestein (SG) was recognized and first described by Jacob Hieronymus Lochner due to its extraordinarily rich and easily detectable fossil content (Lochner, 1711). It has been the subject of more than 130 publications (Thiede, 2023). The fauna is described as comprising about 620 taxa, of which approximately 350 were molluscs. It thrived from 27 Ma to 25 Ma in the shallow waters of the warm, subtropical paleo North Sea during the Upper Oligocene period. Due to its shell layers and grain-size sorted structures, the SG is interpreted as a tempestite (Suhr & Braasch, 1991; Thiede, 2023). Characteristic are centimeter-thick molluscan shells and shell debris embedded in fine granular, silty sandstone layers (Fig. 1.1; Endler & Herrig, 1995). It is overlain by hundreds of meters of more recent strata. Uplift of upper Permian salt structures in tectonically weak areas has caused the overlying strata to be pushed to the surface. During the Weichselian glaciation in the Pleistocene, uplifted and solidified beds of Upper Oligocene strata were sheared off, broken into pieces, and locally displaced. In the areas of Sternberg, Schwerin, and Krakow, the ice and meltwater deposited many such pieces, which were named SG (Schulz, 2003; Thiede, 2023). The SG is thought to be derived from strata uplifted by salt diapirs in and to the east of the main area of occurrence (Obst et al., 2015). The exact origin is still unknown (Schulz, 2003; Obst et al., 2015). Known ice advances of the Brandenburg, Frankfurt, and Pomeranian phases of the Weichselian, main occurrences of SG and sampling sites are given in Figure 2.

The SG rocks are fine-grained sandstones with sideritic-calcitic binding material. The unweathered parts are light grey. As the weathering progresses, the colour changes from beige to yellow to light to dark brown. Their size commonly ranges from a few to 30 centimeters (Thiede, 2023). They are interpreted as layered deposits of shell beds, washed together by normal and storm tides, within almost fossil-barren sands. Four main lithological types, numbered 1 to 4, are differentiated. The silty lime marl type 1 and the silty, fine sand, conglomeratic type 4 are almost devoid of fossils. The classical sandwich type 3 consists of a rich shell layer sandwiched between two sand layers (Fig. 1.1). Another fossil containing type 2 is conglomeratic (Suhr & Braasch, 1991; Thiede, 2023).

A first account of eight foraminiferal taxa in the SG, along with illustrations of five of them, was provided in 1846 by Boll (Boll & Brückner, 1846). In 1849, Karsten reported 34 taxa without illustrations from Sternberg in the collections of the University of Rostock (Karsten, 1849). In his extensive studies of the Tertiary in Germany, Reuss reported 146 taxa for the Upper Oligocene. Of these, he found 27 taxa in the SG (Reuss, 1865). Drawings of almost all 146 taxa were published by Reuss in different works (Reuss, 1856, 1865). With few exceptions, the illustrations show only one view. An example is given in Figure 3. In 1914, Beutler reported 86 foraminiferal taxa for the SG, based on material from the collections of the University of Rostock (Beutler, 1914). For illustrations, he referred to the works of Reuss (1856, 1865), Bornemann (1855), d’Orbigny (1846) and Roemer (1839). Since Beutler's publication in 1914, no new reports on foraminifera in the SG have been made.

During the Oligocene, northern Germany was covered by the North Sea and was located on the periphery of the North Sea basin. Broadly summarized, three main depth-related biofacies were observed: an inner sublittoral, an outer sublittoral-upper bathyal, and the deep-water Rhabdammina biofacies (King, 1983). Based on the analysis of benthic foraminiferal assemblages in 250 cores, the Oligocene of Western Mecklenburg in Northern Germany was divided into seven horizons, as shown in Table 1 (Müller, 2000). Absolute age data were not provided. The horizons showed varying thicknesses between cores. For example, the Asterigerinoides guerichi (Franke, 1912) horizon ranged from 3 m to 50 m. The base of the Chattian was identified by the abrupt and frequent occurrence of A. guerichi (Müller, 2000). This biohorizon was recognized as a suitable stratigraphic correlation marker in shallow waters throughout the entire North Sea Basin (Vandenberghe et al., 2012). The Global Boundary Stratotype Section and Point (GSSP) of the Chattian was erected in 2016 in a bathyal section (Coccioni et al., 2018), which is not applicable in a shelf environment. In the regional Chattian, four horizons were recognized (Müller, 2000). The A. guerichi horizon is dominated by the species that gives it its name, which abruptly disappears at the top of the horizon. The overlying P. oblonga horizon is characterized by frequent occurrences of Astacolus gladius, Nodosaria intermittens, Elphidium subnodosum, and P. oblonga. These taxa decrease in abundance in the overlying Almaena osnabrugensis (Münster, 1838) horizon, where the taxon that gives it its name appears first and often. The overlying Saracenaria magma Spiegler, 1974 horizon lacks the taxa A. osnabrugensis, E. subnodosum, and P. oblonga, while S. magma first appears. For the top of the Chattian and base of the Aquitanian, Müller reports a gradual decrease in the abundance of foraminifera and species diversity which is linked to marine regression. He recognizes an Oligocene-Miocene interval rather than a distinct boundary (Müller, 2000).

Twenty-one rock pieces from five gravel pits each weighing between 0.1 and 1.2 kg were collected and analysed. They were chosen to represent differences in SG pieces with respect to occurrence, lithology, grain-size composition, and weathering. A list is provided in  Appendix A. Five pieces are illustrated in Figure 1. Fifteen pieces were found in the gravel pit of Kobrow, two pieces each in the pits of Consrade and Pinnow Ausbau, and one piece each in the pits of Plate and Zarrentin. The coordinates of the sampling sites are provided in  Appendix A. Besides Zarrentin, the pits are located within the primary area of SG occurrence, as depicted in the map (Fig. 2). Eighteen samples belong to the classical sandwich type three lithologically, and one sample belongs to the conglomeratic type two. One piece belongs to type one, and another piece belongs to type four. The analysed parts of fifteen pieces contain a wide range of grain sizes and are referred to as “full”. Four pieces referred to as “coarse” contain 95% or more particles that are bigger than 1 mm. Two pieces addressed as “fine” only contain particles smaller than 1 mm. The colour reflects the weathering of the analysed pieces. A light beige and beige colour indicates little to no weathering, while a yellowish and brown colour indicates moderate to heavy weathering (Endler & Herrig, 1995). The pieces were soaked in water, deep-frozen, and thawed twenty-five times or more until at least 30 g were disaggregated. The disaggregated fractions were boiled for 30 minutes in a 10% solution of sodium carbonate (Na₂CO₃) in water and left to stand for a day.

The disaggregated fractions were washed through 125-µm sieves, and the retained residues were air-dried under a cover. The dry residues were split using a micro splitter. About 15% of the samples were utilized to select and capture specimens for an illustrated list of taxa. These splits were excluded from the count. Optical images were captured using a Keyence VHX 900F, while scanning electron micrographs were obtained with the VEGA3-TESCAN at Senckenberg am Meer. The remaining splits were dispersed on a tray and fully counted until at least 100 specimens per sample were found. The number of 100 was considered sufficient for analysing a tempestite that only had a broad relationship to the original live assemblage. About 5% of the specimens were broken and only counted if more than half of the test was intact.

The assignment to taxa was undertaken using the illustrated list of taxa. The list of taxa is given in  Appendix B and the occurrences of the species in each sample are detailed in  Appendix C. Ten of the illustrated taxa were so rare that they were not found in the counts. Dashes are used to indicate individual sample and taxa data with zero totals. The identification of taxa was based on Staesche & Hiltermann (1940), Bettenstaedt et al. (1962), Loeblich & Tappan (1987), Müller (2000), and Ellis & Messina (2020). The nomenclature was updated to Hayward et al. (2023), except for the genus Palmula and the species Astacolus gladius. The assignment to size categories (small, mid, and large) was based on adult specimens of a taxon, rather than on individual specimens.

To determine and test the affinities between the assemblages, we applied cluster analysis (CA) and analysis of similarities (ANOSIM) using the PAST (PAlaeontological STatistics) data analysis package version 4.13 (Hammer, 2023), based on the specimen counts at the species level for the nineteen pieces with fossil content. We performed CA using hierarchical cluster analysis in Q-mode with Ward’s method algorithm (Ward, 1963; Hammer, 2023). We performed ANOSIM on the eighteen pieces in clusters 1 to 3 using the one-way ANOSIM with the Bray-Curtis similarity index (Hammer, 2023).

Fisher's alpha (α), Pielou's equitability (J), and Shannon's homogeneity (H; Fisher et al., 1943; Shannon, 1948; Pielou, 1966) were calculated as faunal indices. Fisher's alpha index (α) is a measure of diversity that describes the relationship between the number of species and the number of specimens by assuming that species abundance follows a log distribution (Murray, 2006). Pielou's equitability (J), a measure of evenness, describes how specimens are distributed among species and ranges between 0 and 1. The Shannon H index takes into account both the number of species and the evenness of their distribution. For one specimen it starts at 0 and increases as the number of species and the evenness of the assemblage rise. The raw materials, disaggregated fractions, and picked specimens are deposited at the Foraminifera.eu lab, Waterloostr. 24, 22769 Hamburg, Germany.

A total of over 3,000 specimens were extracted from twenty-one pieces of SG. Nineteen pieces of the lithological types two and three, regardless of grain-size composition and weathering grade, were found to contain a minimum of one hundred foraminifera. The SG pieces of type one (KOB_14) and four (KOB_15) were barren of fossils, as reported by collectors.

A total of 82 foraminiferal taxa were found and are listed alphabetically by order together with the naming authority in  Appendix B. For the count, 1,978 foraminifera were identified to the species level.  Appendix C provides complete assemblage data. Nine agglutinated specimens could not be identified. Sixty-nine taxa are illustrated in Figures 4–10. The illustrated taxa represent 97% of all specimens counted ( Appendix C). Sixteen of these taxa were identified as “sp.” The preservation state of the specimens was generally good to very good with only a few exceptions. There are no significant signs of dissolution. Fragmented small taxa and unidentifiable chamber pieces were observed rarely. Delicate structures of the surface and aperture were either minimally damaged or well intact, as illustrated, for example, in Figures 8.3c and 8.4. Agglutinated foraminifera tend to disintegrate easily. Few small fragments and no crushed specimens were found.

The fauna is characterized overall by large specimens, with 98% exceeding the size of 500 µm and 52% exceeding the size of 1,000 µm. Even in the small-grained samples KOB_08 and KOB_11, the proportion of specimens smaller than 500 µm is low, at 4% and 11% respectively ( Appendix D).

Of the four benthic Chattian marker species in Western Mecklenburg (Müller, 2000), only Palmula oblonga was present. The other three species, Almaena osnabrugensis, Asterigerinoides guerichi, and Saracenaria magma, were not found. Four abundant species in the SG Elphidium subnodosum, Palmula oblonga, Nodosaria intermittens, and Lenticulina osnabrugensis were reported by Müller (2000) as biostratigraphically relevant (Table 1). Collectively, they account for 33% of the total counts ( Appendix D).

The assemblages comprised benthic foraminifera, with the exception of one planktonic specimen of the species Ciperoella anguliofficinalis. The abundance of species was unbalanced, with the five most abundant species accounting for more than half of the individuals. Furthermore, the 11 most abundant species accounted for three quarters of the total population (Fig. 11, Appendices C, D). The four most abundant species occurred in 80% to 100% of the nineteen pieces with foraminiferal content (Fig. 11). Elphidium subnodosum was found in all pieces with fossil content and accounted for 11% of the total. Cribrostomoides jeffreysii and N. intermittens were present in 17 specimens, with total shares of 16% and 9%, respectively. Palmula oblonga occurred in 16 pieces, while Cyclammina placenta and Heterolepa dutemplei were found in 15 pieces. Their shares were 10%, 7% and 7%, respectively ( Appendix D).

Three main clusters, consisting of 18 SG pieces and one individual piece, were identified from the cluster analysis (Fig. 12). High dissimilarities between the three clusters are indicated in ANOSIM with R = 0.8439 (Fig. 13A) and paired R values ranging from 0.72 to 0.91 (Fig. 13B). The Bonferroni corrected p-values of 0.0021 for clusters 1 and 3, and of 0.0036 for 1 and 2 indicate a very strong dissimilarity, while 2 and 3 show a less strong dissimilarity with a p-value of 0.0834 (Fig. 13C). The faunal indices Fisher's alpha (α), Pielou's equitability (J), and Shannon (H) are given in  Appendix E.

The following considerations refer to the 19 pieces with fossil content.

The assemblages are considered to be thanatocoenoses. They only partially reflect the original biocoenoses. In tempestites, sediments, including foraminiferal tests, are transported from their origin and deposited, sorted by size and weight. The tests can be damaged to the point of being completely destroyed. These storm deposits were fossilized over millions of years and formed the source rock of the SG. Parts of this rock were sheared off, broken into pieces, and dispersed by glaciers and meltwater (Schulz, 2003; Thiede, 2023). Each piece investigated contained a random fraction of the layered deposits, resulting in a mixture of particle sizes.

The cluster analysis shows that particle size is the primary factor determining the affinity between the pieces. Analysis of similarities indicates that the three clusters show significantly distinct faunal compositions. Cluster 3 comprises four coarse pieces containing 95% of particles larger than 1 mm (KOB_01, KOB_02, KOB_05, and PIN_02). Accordingly, they have the highest proportions of large specimens, measuring 1 mm and above, namely 80% or more. Cluster 2 comprises four pieces containing a wide range of particle sizes and relatively high shares of large specimens: 65% (KOB_03), 63% (PIN_01), 61% (KOB_04), and 48% (CON_02). The share of large specimens overlaps with those of the ten pieces in cluster 1 (KOB_06 to KOB_13, CON_01, and ZAR_01). Their proportion of large specimens is less than 50%. Seven of them have shares below 35%. The piece PLA_01 stands out from all the clusters. Some specimens do not fall into the size category of their taxon. Due to the rarity of such specimens, which are classified as large or medium in size, the potential impact on the proportions is viewed as negligible.

Diversity indices also demonstrate a relationship to particle size ( Appendix E), although not as strongly as observed in the cluster analysis. In terms of diversity, three out of the four coarse pieces (KOB_01, KOB_02, and PIN_02) have significantly lower values (α < 4) compared to the other pieces, except for KOB_06. Apart from the outstanding piece PLA_01, which has the significantly lowest J value of 0.59, the evenness is lowest in the four coarse pieces.

No other significant correlations, for example with the degree of weathering, sample site, or facies, were found. Most likely, the number of pieces available for study was too small.

Faunal interpretation based on individual SG pieces is not considered reasonable. Affinities between single pieces are related to secondary sedimentation processes, as described in the previous paragraph. Transport, sorted sedimentation, and random fractionation by glacial-fluvial forces have obscured the original live faunas. Overall, however, the assemblages in the pieces show strong similarities.

The share of small specimens with sizes below 500 µm is low, with only KOB_11 having a percentage above 10% (11%). Nine pieces do not contain any small specimens. It is concluded that this largely reflects the original fauna. Small sediment particles are abundant in the ‘full’ pieces and are also present in the coarse pieces. The good to excellent condition of the specimens and the rarity of fragmentation suggest relatively short transport and that loss of small, fragile specimens is unlikely.

The six taxa, Cribrostomoides jeffreysii, Cyclammina placenta, Elphidium subnodosum, Heterolepa dutemplei, Nodosaria intermittens, and Palmula oblonga, were found to be the characteristic and important taxa. They were present in 79% of the pieces, with E. subnodosum present in all pieces and C. jeffreysii and N. intermittens present in 89% ( Appendix C). The cumulative abundance of these six taxa across all pieces was 60% and varied from 30% to 83% ( Appendix E).

Except for one specimen, planktonic taxa are missing. From what has been said about small specimens, it is concluded that this reflects the original fauna. A low or zero planktonic/benthic ratio in modern oceans is indicative of shallow waters that are less than 50 m deep (Gibson, 1989; Van der Zwaan et al., 1990; Bellier et al., 2010). The presence and importance of E. subnodosum support the interpretation of the SG as containing an inner shelf assemblage from depths less than 50 m. The genus Elphidium is reported from modern oceans at depths of 0–50 m in temperate to warm waters (Murray, 2006). For Cribrostomoides and Heterolepa, a wider depth range is reported in modern oceans, including depths of 0–50 m (Murray, 2006). For Heterolepa dutemplei, an outer neritic to upper bathyal distribution has been reported for the Neogene (Van Morkhoven et al., 1986). No data are given for Nodosaria and Palmula. Of the characteristic taxa, Cyclammina is an outlier with a reported range of depths from 100 m to the abyss (Murray, 2006). In the Oligocene and regionally, its preferences may have been different. Using current observations to interpret the past has its limitations, anyway. Individual taxa may have changed their habitat preferences over time. Changing habitat conditions drive evolutionary processes. However, this seems to be the best available approach and is widely used (King, 1983; Spiegler, 1986; Gibson, 1989; Bellier et al., 2010; Murray & Alve, 2011).

The assemblages are characterised by a predominance of hyaline species over agglutinated and porcelaneous species. With the exception of KOB_09, their percentages range from 48% to 88% with an average of 66% ( Appendix D). In summary, the assemblages show the characteristics of the inner sublittoral biofacies of King (1983): low numbers of planktonic specimens and predominantly benthic hyaline taxa.

As mentioned in the section on regional biostratigraphy, the Oligocene of Western Mecklenburg is divided into seven horizons (Table 1; Müller, 2000). The assemblages of the investigated SG pieces belong to the Chattian Palmula oblonga horizon. The presence of the important species Elphidium subnodosum in all pieces together with the absence of Asterigerinoides guerichi and Almaena osnabrugensis is indicative of this horizon (Table 1). Müller (2000) reported that A. guerichi was abundant in the underlying horizon and A. osnabrugensis abundant in the overlying horizon. Both were absent in the Palmula oblonga horizon. The presence of the important species P. oblonga in 79% of the pieces supports this interpretation, as it is reported to be abundant only in the Palmula oblonga horizon. The important Nodosaria intermittens, which is common in the Palmula oblonga horizon and parts of the overlying horizon, is present in 89% of the samples.

The seven horizons are only of regional significance. The assemblages in western Mecklenburg show similarities with those of the Lower Rhine (Ellermann, 1960), the Kasseler Meeressande in northern Hesse (Kümmerle, 1963), and the Doberg near Bünde (Grossheide & Trunkó, 1965). In detail, however, there are differences in the composition of the assemblages and the range of species. From a global perspective, the SG assemblages are of Chattian age. The sudden appearance of A. guerichi and in large numbers is reported throughout the North Sea basin and is interpreted as the initial part of the Chattian (Ellermann, 1960; King, 1983). The Palmula oblonga horizon lies regionally above this Chattian horizon and below the first Aquitanian horizon or Upper Chattian boundary (Müller, 2000). The only planktonic species found, Ciperoella anguliofficinalis, has the highest reported abundance in the Chattian, but its occurrence ranges from the Upper Eocene to the Lower Miocene (Olsson et al., 2018). In the absence of planktonic marker species, the age range cannot be further restricted.

After extracting over 3,000 specimens from 21 investigated SG pieces, a total of 82 foraminiferal taxa were found. Out of these, 69 taxa were illustrated, which accounted for 97% of the 1,978 counted specimens.

Particle size was found to be the primary factor determining the affinity between the pieces, as revealed by cluster analysis. This is in accordance with previous studies interpreting the SG as containing a tempestite-built thanatocoenosis. Further obscuration of the native fauna in single SG pieces occurred as a result of abrasion and random dispersal of the source layer caused by glacier ice and meltwaters during the Pleistocene.

Detailed interpretation of individual SG pieces in terms of fauna and biostratigraphy was not considered reasonable. Overall, however, the assemblages in the pieces show strong similarities. The fauna is characterized by large specimens, with 98% exceeding the size of 500 µm and 52% exceeding the size of 1,000 µm. The species abundance was unbalanced, with the six most abundant species accounting for over 60% of the individuals. These were Cribrostomoides jeffreysii, Cyclammina placenta, Elphidium subnodosum, Heterolepa dutemplei, Nodosaria intermittens, and Palmula oblonga. Each one was present in at least 79% of the pieces.

The SG was found to have originated from the inner sublittoral zone at water depths below 50 m, in accordance with King (1983). Indicative are the scarcity to absence of planktonic taxa, the presence of the important genus Elphidium in all pieces, and the predominance of hyaline taxa.

The SG is found to belong to the Palmula oblonga horizon of Chattian age, which was regionally recognized in undisturbed strata of Western Mecklenburg through borings (Müller, 2000). Indicative are the frequent occurrences of E. subnodosum, N. intermittens, and P. oblonga, alongside the complete absence of Asterigerinoides guerichi and Almaena osnabrugensis.

This paper is dedicated to the memory of our dear co-worker Karina Thiede (formerly Karina Endler), who passed away while this paper was being written. The authors are deeply grateful to Karina and Nils Thiede, who provided most of the pieces and assisted in finding additional pieces, along with Stefan Polkowsky. We thank Brian Ottway and Leon Hoffman for their careful review of the manuscript, helpful comments, and constructive suggestions. We are also very grateful to Fatematuz Zohora Nishi, an anonymous reviewer, and the associate editor for their constructive comments. Conceptualization, methodology, investigation, formal analysis, data curation, writing: Michael Hesemann. Imaging and identification of taxa, imaging of pieces: Michael Hesemann, Dieter Ketelsen. Both authors have read and agreed to the published version of the manuscript. This research received no external funding. The authors declare no conflict of interest. The Appendices can be found linked to the online version of this article.

Appendix A. List of samples with sites, geographic coordinates (WGS-84 in decimal degrees), facies type, and collector. Facies types: 1 = silty lime marl; 2 = fossil rich, conglomeratic; 3 = fossil rich, sandwich layered; 4 = silty, fine sand, conglomeratic. Dimensions are given as length, width, and height in centimeters. Collector: KNT = Karina and Nils Thiede, MH = Michael Hesemann, SP = Stefan Polkowsky.

Appendix B. List of species found in the Sternberger Gestein by order.

Appendix C. Abundances per species. Cumulative share of illustrated and important species, presence of important species in pieces. Dashes indicate zero values. Size: <500 µm = small, 500–1000 µm = mid, >1000 µm = large. Wall material: agglutinated = A, porcelaneous = P, hyaline = H.

Appendix D. Aggregated abundances per size, wall material and for important species. Dashes indicate zero values. Number of species. Size: <500 µm = small, 500–1000 µm = mid, >1000 µm = large. Wall material: agglutinated = A, porcelaneous = P, hyaline = H.

Appendix E. Diversity data and indices.