Large-sized Late Turonian–Early Coniacian (Late Cretaceous) inoceramid bivalves from Germany: taxonomic issues, temporal framework and palaeoecological implications

A systematic-stratigraphic revision of selected large Late Turonian–Early Coniacian inoceramids of the lamarcki group from Germany showed that both historical specimens of Goldfuss (1836), introduced as Inoceramus annulatus and later becoming the types of Late Turonian I. lamarcki stuemckei Heinz, 1928a and Early Coniacian I. annulatus Goldfuss, 1836, respectively, are of Late Turonian age. I. stuemckei Heinz, 1928a, as emended herein, is well represented by its lectotype. Bulky isolated hinges from Upper Turonian strata, used by Heinz (1932a) to erect Heroceramus hercules, are just fragments of contemporaneous large I. stuemckei; H. hercules is thus invalid. The common Late Turonian I. stuemckei occurs facies-independent from near- to offshore settings and its first common appearance is an excellent marker for the substage base. I. annulatus Goldfuss, 1836 was later applied to very large forms first appearing in the mid-Early Coniacian. Thus, I. annulatus sensu Walaszczyk and Wood (1998) is poorly represented by its medium-sized latest Turonian lectotype, causing some nomenclatorial issues. I. stuemckei (heightmax 500 mm) and I. annulatus (heightmax one metre) are characterized by a stratigraphically abrupt shell enlargement and hinge buttressing across the Middle/Upper Turonian boundary and within the Early Coniacian, respectively. This shift to larger sizes, a common passive defense strategy against predation, is accompanied by widespread evidence for increased Late Turonian predation pressure by marine durophages. We thus speculate that the size increase in Late Turonian–Early Coniacian inoceramid bivalves is part of an escalating arms race between prey and predators, ultimately triggering Late Cretaceous inoceramid gigantism.


Introduction
Very large, thick-shelled inoceramid bivalves with heights > 250 mm and a conspicuous ornament of coarse concentric rugae are a characteristic feature of Upper Turonian and Lower Coniacian strata of Europe, northern America and Asia (e.g., Sowerby 1823; Woods 1912;Tröger 1984;Wood 1998, 2018;Walaszczyk and Cobban 2000). In addition, isolated prisms of inoceramid shells more than 5 mm in length and large, massive calcitic hinges with shell thicknesses of several centimeters were found in abundance, especially in the Upper Turonian. For both inoceramid shells and hinges of large to very large sizes three different names are still in use today: Inoceramus annulatus is used for Lower Coniacian representatives (cf. Walaszczyk and Wood 1998;Walaszczyk and Wood in Niebuhr et al. 1999), Inoceramus lamarcki stuemckei for forms predominantly occurring in the Upper Turonian (cf. Ernst et al. 1983;Wood et al. 1984), and Inoceramus hercules for time-equivalent, isolated massive hinges and shell fragments (cf. Heinz 1928a; Wiese et al. 2004;Tröger and Niebuhr 2014).
The aim of this paper is the evaluation of the taxonomic relationships of the large inoceramid bivalves and hinges found in Upper Turonian to Lower Coniacian strata of Germany, also with respect to the validity and independence of the three taxa (Inoceramus annulatus, I. stuemckei and I. hercules) still used until today and belonging to the heterogenous lamarcki group. Furthermore, their stratigraphic ranges, based on associated ammonite and inoceramid occurrences, are calibrated and the functional morphology of trends in shell size and shell stabilization are discussed in the light of possible evolutionary patterns. The Saxonian Cretaceous is particularly suitable for this approach because of the availability of large Upper Turonian populations of I. stuemckei with sizes up to 450 mm, representing all ontogenetic stages, from both marly offshore (Figs. 5a,7,8,9d) and sandy nearshore facies (Figs. 10,11). I. annulatus (cf. Walaszczyk and Wood 1998), however, has not yet been recorded from Saxony so that occurrences from northern Germany have been investigated (Figs. 2,12,13). However, based on the regional bias on Saxonian faunas and in order to keep the paper short, we will focus on the large Upper Turonian inoceramids and do not detail the taxonomy and its issues with respect to I. annulatus.

Short geological overview
All studied sites are located in northern Germany and are part of the wide warm-temperate epicontinental Boreal shelf sea that covered large parts of western and central Europe north of the Mid-European Island during the Late Cretaceous (e.g., Vejbaek et al. 2010;Fig. 2). Additionally, inoceramid occurrences south of the Mid-European Island from the Danubian Cretaceous Group near Regensburg, Bavaria (site 9 in Fig. 2) have been integrated into the study. The roughly coast-parallel facies zonation on the low-angle, ramp like North German shelf comprises inner shelf sandstones and/or greensands, followed by mid-shelf marls and marllimestone alternations, finally grading into outer shelf white limestones and chalks (Wilmsen et al. 2005;Niebuhr 2006). The Cenomanian to Lower Coniacian succession in the Münsterland, in Lower Saxony and the Subhercynian area (sites 1-6 in Fig. 2) are dominated by calcareous offshore sediments combined in the Plänerkalk Group (Wilmsen and Hiss in Niebuhr et al. 2007). In the Saxonian Cretaceous Basin, siliciclastic input from the eastern part of the Mid-European Island (Bohemian Massif) and the Westsudetic Island was higher and a sandy nearshore facies zone (today forming the Saxonian Switzerland) in the southeast grades into a marly-calcareous offshore facies zone in the northwest (Meißen-Dresden area; sites 7 and 8 in Fig. 2). The strata in Saxony are combined in the Elbtal Group (Voigt and Tröger in Niebuhr et al. 2007;Fig. 2), the marly offshore strata of which are now united in the Meißen-Dresden Subgroup (Wilmsen et al. 2019a). Via a former strait, the Saxonian Cretaceous Basin was directly connected to the Bohemian Cretaceous Basin in the southeast that opened towards the northern margin of the Tethyan Ocean (Fig. 2). For an overview of the current lithostratigraphy in the studied basins, see Hiss et al. (2018).
Standard measurements for inoceramid bivalves, taken with a Vernier caliper, comprise height (h) and length (l) of the shell as well as observations on the sculpture (number, size and form of ribs) and the course of the growth axis (for terminology and measured parameters, see Fig. 3). For the maximum height of the shells, the following size classes are used: < 5 cm = small, 5-15 cm = medium-sized, 15-25 cm = large, 25-50 cm = very large, 50-100 cm = huge, and > 100 cm = gigantic (modified according to Tröger and Niebuhr 2014). For the higher rank systematics of the Bivalvia, we refer to Carter et al. (2011).
Locus typicus. Exact place unknown, "grüne und weiße Kreide in Westfalen" according to Goldfuss (1836: p. 114 Description. Inoceramus stuemckei is a large to very large, relatively flat, distinctly ribbed, almost equivalved species with a maximum height and length of 480 mm and 360 mm (FG: 346/1 on Fig. 7), respectively. It is characterized by wide, more-or-less regularly spaced, rounded commarginal concentric rugae ("Anwachsreifen" according to Heinz 1928bHeinz : p. 18, 1928c, running perpendicular to the straight to slightly convex growth axis. The rugae are asymmetric, i.e., the slopes directed to the umbo are shallower compared to the ventrally directed ones. The weakly prosogyre umbo is strongly vaulted and carries small, densely spaced rugae, the distance and height of which constantly increase during ontogeny. Rugae are 15 mm high and separated by inter-rugae spaces up to 40 mm wide on the adult part of the shell in specimens larger than 300 mm. The large, prolate wing has an almost straight hinge line with a broad hinge plate. The general outline is rounded-rectangular. Two basic types can be differentiated among the studied population of Inoceramus stuemckei. Type 1 comprises forms with very strong, rounded to narrowly rounded concentric rugae. Juvenile stages are sturdily convex and thus well differentiated from slightly older I. lamarcki; adult shell parts are only weakly convex. Type 1 forms are characterized rather by box-shaped than circular ventral margins. In outline, they are thus almost symmetrically pentagonal or triangular-oval. Shells, external imprints and favourably preserved internal moulds have regular accretionary rings (growth rings) that also cross the entire wings. All specimens from the sandy nearshore facies in Saxony are members of type 1 and also the marly facies is predominated by this type.
Type 2 forms are very rare and relatively flat (i.e., weakly convex), characterized by only a few, rather weak, shallow, rounded concentric rugae that tend to be strongest in the juvenile and adult growth stages. Shells have regular growth rings while internal moulds are smooth. Their outline is elongated to asymmetric-triangular, resembling in form, but not in sculpture, the lectotype of Goldfuss (1836: pl. 110, fig. 7b). Examples of type 2 forms are BGR: Ma 14106 from the Staffhorst shaft section between 572 and 575 m (Fig. 6) and the specimen FG: 346/1 from the Strehlen Limestone of Saxony (Fig. 7).
The hinges of large specimens (Figs. 8,9), introduced by Heinz (1932a: p. 34) as Heroceramus hercules, are of great thickness, and isolated segments of them are often found separately. The hinges are thickest near the umbo and become thinner towards the posterior end. The ligament pits are numerous, shallow, and several times higher than wide, reaching their maximum height not far from the umbo. Irrespective of the size of the hinge, they are always ~ 2 mm wide, resulting in eight grooves on a length of 15 mm (MMG: SaK 9609; Fig. 9a), 13 grooves on a length of 26 mm (MMG: SaK 9590; Fig. 9d), and 23 grooves on a length of 42 mm (FG: 690-S4/84; Fig. 9b).
The calcitic shell and the massive hinges are only preserved in specimens from calcareous and calcareous marl facies (e.g., Strehlen Limestone of the lower Strehlen Formation in Saxony, Tröger 1984; Hundorf Limestone of the lower Teplice Formation in Czechia, Wiese et al. 2004;Walliser et al. 2020) or calcareous glauconitic sandstones (e.g., Soest Greensand Member of the Duisburg Formation, Münsterland Cretaceous Basin; see MMG: NwK 843, Fig. 9e). The shell thickness increases with distance from the umbo from 0.5 to 3.5 mm. The shell shows fine, regularly spaced, slightly raised growth lines ("Anwachsringreifen", cf . Heinz 1928b: p. 19) that are between ~ 2 and 5 mm wide. In the sandstone facies, these raised growth lines are usually  Heinz, 1928a. BGR: Ma 14105; right valve, 180 mm in height; "Lamarcki-und untere Scaphitenschichten" according to Heinz (1928a); Salder Formation, (4) Lüneburg, Lower Saxony; lower to middle Upper Turonian; Inoceramus Brongniarti according to Heinz (in sched, leg. 1910) ◂ preserved only in external imprints while moulds are characterized by simple rugae in most cases. Only in exceptional cases they are preserved on sandstone moulds (Fig. 10). Specimen MMG: SaK 6622a + b from the sandstone facies of the Saxonian Switzerland shows both preservation types: moulds with simple, regularly spaced, rounded concentric rugae (Fig. 11a2) and external imprints of the same specimens with rugae covered by fine growth lines (Fig. 11a1). This indicates that the occurrence of the raised growth lines in imprints and moulds is related to different preservation forms and not a characteristic feature for specific inoceramid taxa such as the Lower Coniacian species Inoceramus annulatus (see "Discussion" below).
Remarks. In contrast to former authors (e.g., Tröger 1967;Keller 1982) we regard Inoceramus stuemckei Heinz as an individual species, morphologically sufficiently distinct from Inoceramus lamarcki Parkinson (also Walaszczyk and Cobban 2000: p. 46 regard Heinz's taxon as a species). Parkinson's species is generally smaller than 150 mm in height (i.e., medium-sized), and the shells are more convex with strongly domed umbo and dorso-ventrally elongated, characterized by weaker sculpture of shallow, oval rugae. Furthermore, both taxa are stratigraphically largely discrete, I. lamarcki characterizing the (late) Middle Turonian and I. stuemckei, after its appearance in the late Middle Turonian, is dominating in the early to mid-Late Turonian. Nevertheless, both species are closely allied.
The species Heroceramus hercules was introduced by Heinz (1932a: p. 34) for large, massive calcitic inoceramid hinges from the "obere Scaphitenschichten" (upper Upper Turonian) in the Lüneburg section, northern Germany (Heinz 1932b). Additional findings of such largescale h inges are known, e.g., from the mid-Upper Turonian Streh len Limestone of the lower Strehlen Formation, Saxony (Figs. 9a-d;Geinitz 1839Geinitz , 1873Tröger 1984;Tröger and Niebuhr 2014) and the time-equivalent Hudcov (Hundorf) Limestone of the lower Teplice Formation, Czechia (e.g., Wiese et al. 2004). Tröger (1984: pl. 2) designated both an isolated hinge and a large composite mould from the Strehlen Limestone with parts of the shell preserved as neotypes for Inoceramus (Heroceramus) hercules (Figs. 7, 9b). However, we consider I. (H.) hercules Heinz, 1932a as a younger subjective synonym of Inoceramus stuemckei Heinz, 1928a. As can be shown by conjugate occurrences [e.g., specimen MMG: Sak 8417 from the mid-Upper Turonian Strehlen Limestone at Weinböhla (Fig. 8) in which the hinge is in contact with the composite mould, or the specimen illustrated by Tröger and Christensen (1991: pl. 3, fig. 4)], the massive hinges and hinge plates belong to shells of I. stuemckei and are not remains of another large species.
In Tröger and Wolf (1960) "Inoceramus cuvieri var. annulata Goldfuss" and "Inoceramus latus Man. var. stümckei Heinz" from the marls of the upper Räcknitz Formation a few meters below the Strehlen Limestone are described as synonyms. It should be noted that there is in fact some taxonomic confusion concerning the separation of the two species and that we regard the hitherto accepted lectotype of Inoceramus annulatus [Goldfuss 1836, pl. 110, fig 7a) from the "mittlere und obere Scaphitenschichten" (Heinz , 1928a; middle to upper Upper Turonian) as a possible representative of I. lusatiae Andert, 1911 (see "Discussion" below). Also, the medium-sized Inoceramus annulatus Goldfuss part Heinz 1928of Bräutigam (1962 fig. 4), from his lower vancouverensis Zone at Langelsheim, Harz foreland near Goslar, ca. 30 m above the first appearance of the zonal index Inoceramus perplexus and ca. 20 m below the FO of Mytiloides striatoconcentricus (Bräutigam 1962; app. 7), is questionable. This level equals to the I. perplexus/I. stuemckei/I. inaequivalvis Zone, far below the first appearance of Lower Coniacian I. annulatus in the sense of Walaszczyk and Wood (1998), and we thus include it in I. stuemckei. The specimen of Wolanski (1932: p. 29, pl. 5, fig. 1), however, described as Boehmiceramus sp./Inoceramus (Boehmiceramus) sp. from Upper Turonian limestones of Wusterhusen-Gustebin, Vorpommern [not "Mucronatensenon", see Deecke (1908)], and labeled by von Hagenow as Inoceramus annulatus Goldf., is a typical I. stuemckei due to the presence of wide, regularly spaced, round-topped concentric rugae covered by conspicuous, regularly spaced, raised growth lines which are not developed in the "Senonian" subfamily Boehmiceraminae of Heinz (1932b: p. 18).
The large, allegedly Coniacian species Inoceramus sachsi Bodylevsky, 1958 (p. 79, pl. 30, fig. 1; pl. 32, figs. 1a, 6; pl. 33, fig. 1) is similar to I. stuemckei and, according to the stratigraphic revision of its stratum typicum in Taymyr (western Sibiria), of Late Turonian age (Walaszczyk and Cobban 2000: p. 50), too, such as the North American records reported therein. I. sachsi may well turn out to as a younger subjective synonym of Heinz's species but a revision of Bodylevsky's taxon is beyond the scope of this paper.
Palaeogeographic occurrence. Central and East Europe (very common), Middle East (Iran) to (Central) Asia, northern America. The following occurrences in Europe can be specified: England. Yorkshire to Kent = lower Upper Chalk.

Discussion
As outlined in the taxonomic part, there is only one very large-sized inoceramid species in Upper Turonian of northern Germany, i.e., Inoceramus stuemckei Heinz, 1928a. It is followed by another very large to huge Early Coniacian species, i.e., Inoceramus annulatus sensu Walaszczyk and Wood (1998). Their stratigraphic distribution and possible evolutionary ties as well as the functional morphology of shell traits and inferred modes of life are discussed below.

Stratigraphic ranges of selected species of the lamarcki group
The lamarcki group comprises a rather heterogeneous complex of about 50 species and subspecies (many of which probably synonymous) with poorly understood phylogentic relationships that appeared in the Middle Turonian and ranged into the basal Upper Coniacian (Walaszczyk and Cobban 2000: p. 50). Taxa of the group occur all over the northern hemisphere and show considerable morphological plasticity. Herein, we focus on large to huge Upper Turonian to Lower Coniacian representatives from selected sections in northern Germany.
There are three sections in Lower Saxony where both Inoceramus stuemckei and I. annulatus were found (Fig. 12): the former quarry at Lüneburg, study area of Rudolf Heinz, which is inaccessible since a long time (site 4 in Fig. 2), the Staffhorst shaft section, from which the rich assemblage of inoceramid bivalves is housed in the BGR collection (site 2 in Fig. 2), and the famous quarry section in Salzgitter-Salder, GSSP for the base of the Coniacian (Walaszczyk et al. 2021, site 5 in Fig. 2).
The range of Inoceramus stuemckei in northern Germany is fairly well constrained, comprising the basal Upper Turonian up to the first appearance of M. scupini in the upper Upper Turonian. An overlap of small-to mediumsized Inoceramus lamarcki (< 150 mm in height) and large  Heinz, 1928a. BGR: Ma 14106; type 2, most parts of the shell preserved; note the wide, regularly spaced, round-topped concentric rugae covered by fine, regularly spaced, raised growth lines ("Anwachsringreifen"), 240 mm in height; Salder Formation, (2) 572-575 m depth in the Staffhorst shaft, Lower Saxony; I. inaequivalvis/I. stuemckei/I. perplexus Zone, lower Upper Turonian ◂ to very large Inoceramus stuemckei (commonly > 250 mm in height) in the Middle-Upper Turonian boundary interval is proposed by Walaszczyk (1988, 1992, pers. commun. May 2021), but not seen in our material. In the Staffhorst shaft section, the lowermost Upper Turonian I. perplexus/I. stuemckei/I. inaequivalvis Zone is differentiated in a lower part without, and an upper part with Inoceramus perplexus (Walaszczyk and Wood in Niebuhr et al. 1999;see Wiese et al. 2020 for an alternative Upper Turonian inoceramid biostratigraphy based on interval zones only); however, very large Inoceramus stuemckei, such as specimen BGR: Ma 14106 (Fig. 6), are a characteristic feature of the lower Upper Turonian I. perplexus/I. stuemckei/I. inaequivalvis Zone (Ernst et al. 1983;Wood et al. 1984;Walaszczyk and Wood in Niebuhr et al. 1999;Kaplan 2011). Also in Saxony, thin-shelled I. perplexus, limited to the offshore marl and Pläner facies, are associated with large I. stuemckei. Thus, the appearance of common large I. stuemckei in Sandstone a 3 of the Postelwitz Formation in the nearshore zone (Saxonian Switzerland; cf. Wilmsen andNiebuhr 2009, 2014a) is considered as simultaneous, i.e., marking the base of the Upper Turonian in the shallow-water sandstone facies where the thin-shelled index inoceramid I. perplexus is missing. This is also supported by the fact that below this level, also in the sandstone facies of the Saxonian Switzerland, the lamarcki group is only represented by small-to mediumsized Inoceramus lamarcki ssp. which usually do not exceed a height of 150 mm, being a feature of the upper Middle Turonian of central Europe (Walaszczyk and Wood in Niebuhr et al. 1999). The upper limit of the stratigraphic range of I. stuemckei is demonstrated in the Staffhorst shaft section where the first appearance of M. scupini is associated with the last occurrence of large Inoceramus ex gr. lamarcki and I. ex gr. inaequivalvis (Walaszczyk and Wood in Niebuhr et al. 1999: p. 188 In the Lüneburg section, the range of Inoceramus annulatus Goldfuss, 1836according to Heinz [(1928a, i.e., Goldfuss' specimen 7a (1836: pl. 110, fig. 7a)], is identical with the range of Mytiloides scupini (= Inoceramus stillei of Heinz 1928a: pl. 3; Inoceramus stillei var. scupini n. var. of Heinz 1930). This interval is in turn characterized by small-to medium-sized representatives of the lamarcki group, e.g., Inoceramus lusatiae, in central European sections (Walaszczyk and Wood 1998). Very large and huge inoceramids > 250 mm in height are missing in this interval. This suggests that Heinz ( , 1928a by no means saw in the small Goldfuss specimen 7a (< 85 mm in height) a representative of a huge inoceramid species with "shells exceeded three feet [914 mm] in length and twenty inches [508 mm] in width" (Mantell 1822: p. 213; see Fig. 13), often noted from the "schloenbachi beds" (Lower Coniacian) and the Lower-Middle Coniacian transitional beds (Walaszczyk 1992), e.g., Inoceramus annulatus in the sense of Walaszczyk and   fig. 5)]. However, this identification was not confirmed by Andert (1934: p. 128), Tröger (1967: p. 73) and Walaszczyk (1996: p. 369). Moreover, Walaszczyk and Wood (1998) are of the opinion that Inoceramus lusatiae and I. annulatus, both belonging to the lamarcki group, are extremely difficult to distinguish and do not include the original of Andert (1911: pl. 8, fig. 5) in their taxonomic concept of the small-to medium-sized species I. lusatiae. However, I. stuemckei, I. lusatiae, the Goldfuss specimen 7a (1836: pl. 110, fig. 7a), and the Early Coniacian I. annulatus exhibit very similar lamarcki-like ornaments with nearly concentric rugae, covered by regularly spaced, raised growth lines (Anwachsringreifen, cf. Heinz 1928b: p. 19, 23;1928c: p. 75). Furthermore, the fact that I. annulatus of Heinz ( , 1928a and I. lusatiae have an overlap in the topmost Turonian Mytiloides scupini Zone suggests that Goldfuss' specimen 7a (1836: pl. 110, fig. 7a) may rather correspond to Andert's species concept of I. lusatiae. This, in turn, would indicate that I. annulatus of Heinz ( , 1928a) is a younger synonym of Inoceramus lusatiae Andert, 1911 and the name Inoceramus annulatus Goldfuss, 1836 is a nomen nudum which does not fulfil the criteria of the ICZN (Ride et al. 1999).
Inoceramus annulatus in the sense of Walaszczyk and Wood (1998) first appears some way-up in the Lower Coniacian (Fig. 12), maybe as a phylogenetic successor of Inoceramus lusatiae, its most closely resembling species (Walaszczyk and Wood 1998). I. lusatiae disappears nearly simultaneously with the first appearance of I. annulatus, however, at different stratigraphic levels at certain central European sections, e.g., the C. c. crassus Zone at Staffhorst (Walaszczyk and Wood 2018; site 2 in Figs. 2, 12; see also Fig. 13) and the base of the C. w. hannovrensis Zone of Słupia Nadbrzeżna, Poland (Walaszczyk and Wood 1998). Furthermore, a transitional form of I. lusatiae/annulatus was found at the latter section (Walaszczyk and Wood 1998: pl. 10, fig. 5) which equals the small Goldfuss specimen (1836: pl. 110, fig. 7a) very well. Also, in the Salzgitter-Salder section, I. lusatiae has its highest occurrence at the base of the C. w. hannovrensis Zone (163 m depth according to Wood and Ernst 1998: figs. 55, 56) while I. annulatus first appears 12 m above, respectively (175 m depth; Walaszczyk et al. 2021). This strongly suggests that I. annulatus should be regarded as a successor of I. lusatiae, and that I. lusatiae from the C. w. hannovrensis and C. c. inconstans zones at Staffhorst equals to small-to medium-sized specimens of I. annulatus of Salzgitter-Salder and Słupia Nadbrzeżna, Poland. I. annulatus ranges to the top of the Lower Coniacian, possibly also into the basal Middle Coniacian (Walaszczyk and Wood 1998: p. 424).
In a nutshell, I. annulatus Goldfuss, 1836sensu Heinz (1926, 1928a and I. annulatus sensu Walaszczyk and Wood (1998) are almost certainly different species with different stratigraphic ranges-so far with an identical lectotype of latest Turonian age (see Fig. 1). The current paper does not provide a final solution for the taxonomic issues arising from this conclusion but we suggest that the name I. annulatus sensu Walaszczyk and Wood (1998) should be retained for the very large to huge Early Coniacian forms in order to reflect the taxonomic usage of the last decades. Medium-sized I. annulatus sensu Heinz ( , 1928a from the latest Turonian may be conspecific to contemporaneous I. lusatiae Andert, 1911 but further systematic studies are needed in this respect in order to  Heinz of Tröger (1981, pl . 1, fig. 4, 4a), Inoceramus sp. (n. sp.?) ex gr. Inoceramus lamarcki Parkinson of Tröger (1984: pl. 1, fig. 3, 3a) and Inoceramus hercules (Heinz, 1932a, b) of Tröger and Niebuhr (2014, fig. 4a). b FG: 346/2; massive hinge with 23 preserved grooves on a length of 42 mm, lower and upper side; original and neotype of Inoceramus (Heroceramus) hercules Heinz of Tröger (1984: pl . 1, fig. 4, 4a). c MMG: SaK 9640; c1, upper side of a hinge encrusted by small oysters; note growth lines of the outer shell, c2, large prisms of the hinge up to 14 mm thick and small prisms of the shell up to 2 mm thick visible in the cross-section. d MMG: SaK 9590; small hinge with 13 preserved grooves on a length of 25 mm; original of Inoceramus Brongniarti Sow. of Geinitz (1873: pl. II.11, fig . 7). e MMG: NwK 843; fragments of thick-shelled inoceramid valves partly in contact from the "grüne Kreide von Westphalen" of Goldfuss (1836); Soest Greensand Member of the Duisburg Formation, (1) Anröchte-Klieve, Münsterland Cretaceous Basin; M. labiatoidiformis/M. striatoconcentricus Zone, mid-Upper Turonian satisfactorily solve the complex nomenclatorial dilemma revolving around I. annulatus according to the ICZN (Ride et al. 1999).

Palaeoecological implications
When evaluating the morphological trends in the above considered late Middle Turonian to Early Coniacian representatives of the lamarcki group, an overall maximum size increase becomes evident, accompanied by a thickening of the shell and buttressing the hinge areas (Fig. 14). The relatively abrupt increase in shell size from medium-sized Inoceramus lamarcki to very large Inoceramus stuemckei at the transition from the Middle to the Late Turonian may have been fostered by a global eustatic sea-level fall at this time (sequence boundary SB Tu 3 of Janetschke et al. 2015), resulting in widespread shallowing of the shelf seas and triggering speciation in high-stress situations (cf. Kauffman 1978a). This speciation event seems to follow the evolutionary mode of punctuated equilibria sensu Eldridge and Gould (1972) and Gould and Eldredge (1977) and have been documented for other Upper Cretaceous inoceramids (e.g., Kauffman 1977). Corresponding evolutionary events often occurred during high-stress situations affecting whole communities during short time intervals such as rapid marine regression (Kauffman 1978a) when habitat segregation favored allopatric speciation. In the henceforth shoaling early Late Turonian seas, the newly evolved largesized and thickened shells were probably beneficial in two ways: on the one hand, a thick and strongly ribbed shell may provide increased mechanical protection and sea-floor adhesion in agitated shallow waters. However, this interpretation alone is certainly not the main reason because the same inoceramid morphs also occur in contemporaneous deeper water facies, i.e., under low-energy conditions. On the other hand, size increase is, together with sculptural features such as ornamentation and thick shells, a common passive defense strategy among shelled organisms in order to reach a size refugium in which the prey becomes unassailable for a predator (e.g., Klompmaker et al. 2019). This is especially important for immobile epifaunal organisms such as inoceramid bivalves that simply cannot escape from attacks. Most likely the development of the massive hinges in I. stuemckei also aimed at improved breakup-protection against predator attempts to open the valves, converting the live shells into a "security container" (Coen 1985 pointed to the important role of the bivalve hinge in withstanding shear forces). Thus, large, thick and strongly ribbed shells with buttressed hinges in the first place provided improved shelter against shell-breaking marine predators such as ptychodontoid sharks and marine reptiles (e.g., mosasaurid lizards) that are much more common in shallow waters (cf. Ward 1996;Kelley and Hansen 2001) and increased in importance during the Mesozoic marine revolution, especially in the (Late) Cretaceous (Vermeij 1977). We thus propose that predation pressure increased during the Late Turonian shallowing phase, triggering the size increase and the development of high-strength shells in some representatives of the lamarcki group (Fig. 14), providing additional evidence for the escalation hypothesis of Vermeij (1987,1994). In support of this view, Ekrt et al. (2001) and Wiese et al. (2004) provided firm evidence for the abundance of marine reptiles and sharks (i.e., potential predators) from the lower Teplice Formation (mid-Upper Turonian) of the Bohemian Cretaceous Basin; also, the contemporaneous Strehlen Limestone of Saxony yielded numerous corresponding vertebrate remains (Licht et al. 2016;Sachs et al. 2016Sachs et al. , 2017, especially massive shark vertebrae and teeth, including specimens that can be attributed to the durophagous taxon Ptychodus (Fig. 15a). Moreover, a relatively abrupt increase in predation intensity on inoceramids by vertebrates was proposed for the Late Turonian of the US Western Interior Seaway by Harries and Ozanne (1998); they quoted an increasing abundance of Ptychodus (shark) teeth and inferred predation marks on up to 40% of the inoceramids in certain levels of the Upper Turonian. Ptychodontid sharks attained considerable sizes during the Late Cretaceous (Jambura and Kriwet 2020) and are known to have preyed upon inoceramids (Kauffman 1978b). Furthermore, durophagous adaptions such as molariform teeth on the fingers of the chelae, heterochely (e.g., enlarged crusher claws and/or specialized cutting claws) and calcified mandibles had appeared among multiple clades of the Decapoda by the Late Cretaceous (Schweitzer and Feldmann 2010). Especially nephropid and erymid lobsters were widely distributed during the Cretaceous and had claws that could be used for crushing or capturing prey (Feldmann and Schweitzer 2014); large erymids are not uncommon in the Saxo-Bohemian Cretaceous (see Geinitz 1875;Fritsch and Kafka 1887;Gründel 1974), e.g., Enoploclytia leachii (Mantell) which has also been found in the Strehlen Limestone (Fig. 15c), along with abundant shark remains including ptychodontoid teeth and vertebrae (Fig. 15a, b). Harries and Ozanne (1998: p. 382) also brought two further decapod groups into play who attained durophagy during the Late Cretaceous, corresponding to the interval where evidence for inoceramid predation increases: stomatopod and brachyuran crustaceans. Experiments on crushing crab predation showed that the shells of the Blue Mussel Mytilus edulis, a well-suited model organism for the typical shell bauplan  Heinz, 1928a. FG: 690-871/96; left valve with fine growth lines, parts of the wing and imprint of the hinge are preserved, 190 mm in height; Sandstone a 3 of the Postelwitz Formation, (8) former quarries of Bad Schandau-Postelwitz, Saxony; I. inaequivalvis/I. stuemckei/I. perplexus Zone, lower Upper Turonian ◂ of many Cenomanian-Middle Turonian inoceramids, works well with its thin shell when staying small, but an increase in size is structurally disadvantageous (Nagel-Myers et al. 2019). This observation may be the reason that the size increase of Late Turonian inoceramids is not seen in mytiliform taxa and only certain groups started the arms race with the diversifying durophagous predators, also applying shell-thickening and sculpture formation. Consequently, the numerous isolated massive hinges and thick shell fragments found in Upper Turonian strata, originally giving rise for the erection of I. hercules, might also be seen as evidence of (successful) predation. However, such fragments may be indistinguishable from shell material resulting from postmortem shell breakdown by physical processes (Schweitzer and Feldmann 2010). After a gap in the fossil record during the late Late Turonian and earliest Coniacian when large-to huge-sized inoceramids largely disappeared (interval with I. lusatiae in Figs. 12, 14), I. annulatus readopted the trend of maximum size increase in the lamarcki group during the Early Coniacian (Fig. 14). Even if I. annulatus may be seen as an phylogenetic successor of I. lusatiae (to which it has the closest morphological relationship; cf. Walaszczyk and Wood 1998), the exact evolutionary relationships in this interval remain elusive; further detailed stratigraphic and systematic studies are needed.

Mode of life
Inoceramid bivalves were epibenthic, predominantly sessile organisms that often dominated Late Cretaceous offshore level-bottom communities (e.g., Kauffman et al. 2007;Wilmsen 2008;Wilmsen et al. 2021), but they are also found in nearshore sands, suggesting that they had comparatively wide tolerances for physico-chemical environmental factors at genus and species level (e.g., Harries et al. 1996). From the perspective of functional morphology, the flat, disc-like shells of I. stuemckei and I. annulatus provide large resting surfaces that are, in conjunction with the thick shell, most likely an adaption to a recumbent lifestyle (see Kauffman et al. 2007 for Platyceramus and Wilmsen et al. 2021 for disc-like Gnesioceramus crippsi). An upright, semi-infaunal mode of life with the dorsal part buried and the ventral part directed into the water column is less likely given the fact that in the sandy nearshore facies of Saxony I. stuemckei has always been found in horizontal position, whereas contemporaneous pinnid bivalves often occurred in their vertical semi-infaunal life position (e.g., Wilmsen and Niebuhr 2014b: figs. 13c, 13d).
Where preserved, the regular growth rings of both I. stuemckei and I. annulatus show a rather equal width between 2-5 mm across the major part of the shell (or external imprint) of individual specimens (see for example Figs. 6 and 11b). The equally broad increments suggest constantly recurring growth cycles with no or little ontogenetic variation. Environmental variables in fact significantly influence modern bivalve shell formation with temperature and food availability exerting the main controls on growth rate (Schöne et al. 2005;Walliser et al. 2018). Regular growth rings of similar dimension (2 mm) in Late Cretaceous inoceramids from Demerara Rise have been assigned to short-term cycles of food availability in deep-oceanic bottom waters, driven by seasonal/annual productivity changes in surface waters (Jiménez Berrocoso et al. 2008). Furthermore, Walliser et al. (2020) suggest that growth bands in Late Turonian Inoceramus hercules (= I. stuemckei) were produced annually and shell growth was nearly continuous throughout the year. Accepting this, a half-to one-meter-high inoceramid bivalve would have attained a maximum age of 250-500 years. However, even if some modern bivalves may reach ages of > 200 year under more temperate conditions and with slower growth rates (e.g., Arctica islandica; Thompson et al. 1980), such an old age (and consequently slow growth) for inoceramids seems very unlikely, especially when considering that ontogenetic size increase as fast as possible was of utmost importance in the predator-littered Late Cretaceous marine environment in order to reach the size refugium, following the leitmotif: grow fast or die young. Furthermore, intra-annual growth increments have been observed by Schöne et al. (2005) and we speculate that the incremental growth is reflecting in fact sub-annual growth cycles; monthly neap-spring tidal cycles, known to potentially be recorded in bivalve shells (e.g., Pannella and MacClintock 1968;Richardson et al. 1980), may be an option. Their effects are most pronounced in inter-to shallow subtidal environments but spring-neap modulation of coastal water circulation may result in subtidal flows that are known to be capable of rapidly flushing buoyant material out of the coastal zone and into the sea at neap tides (Ribeiro et al. 2004), exerting a strong control on food availability for and, thus, incremental shell growth of shelf biota. In this scenario, the possible life span of the large inoceramids would be reduced to 20-40 years, but a final statement on the origin of the inoceramid growth rings is currently not possible.  Inoceramus stuemckei Heinz, 1928a. MMG: SaK 6620a + b; hypo-and epirelief of a shell bed with at least five juvenile specimens, specimens I and II are 110 mm in height; a1 (left), imprints of specimens I and II show nicely preserved "Anwachsringreifen" of the inner shell and the completely preserved wing (specimen I), a2 (right), on the composite moulds of specimens I and II only the strong rugae are visible; Sandstone a 3 of the Postelwitz Formation, (8) former quarry Reinhardtsdorf 1 in Reinhardtsdorf-Kleingießhübel, Saxony; I. inaequivalvis/I. stuemckei/I. perplexus Zone, lower Upper Turonian

Conclusions
Based on the systematic and stratigraphic revision of selected large inoceramids of the lamarcki group from Upper Turonian and Lower Coniacian strata of Germany, the following taxonomic, biostratigraphic, evolutionary and pa laeoecological conclusions can be drawn: Fig. 12 The three known localities in northern Germany where both very large Inoceramus stuemckei and huge Inoceramus annulatus were found in-situ (for locations see Fig. 2); note occurrences compared to the FOs and LOs of additional inoceramid bivalves. (2) Staffhorst shaft section (see Walaszczyk and Wood in Niebuhr et al. 1999;Walaszczyk and Wood 2018); (4) Lüneburg section (see Heinz 1928a); (5) Salzgitter-Salder section (see Wood and Ernst 1998;Walaszczyk and Wood in Niebuhr et al. 1999;Walaszczyk et al. 2021); Cww = C. waltersdorfensis waltersdorfensis, Ccc = C. crassus crassus, Cdd = C. deformis deformis ◂ Fig. 13 Inoceramus annulatus sensu Walaszczyk and Wood (1998). BGR: Ma 14107; double-valved specimen in butterfly position with most of the shell preserved, more than 450 mm in height, with nicely preserved "Anwachsringreifen" on the composite mould, but not on the outer surface of the shell; Erwitte Formation, (2) 443.50-447 m depth in the Staffhorst shaft, Lower Saxony; Cremnoceramus crassus crassus Zone, Lower Coniacian • In his Petrefacta Germaniae, Goldfuss (1836) illustrated two inoceramid specimens as Inoceramus annulatus which were differentiated by Heinz ( , 1928a into "Inoceramus latus Mant., var. stümckei n. var." (i.e., the specimen on pl. 110, fig. 7b in Goldfuss 1836) and "Inoceramus annulatus Goldf., part. Heinz" (i.e., the specimen on pl. 110, fig. 7a Heinz, 1928a as emended herein; the name is thus no longer valid. • Inoceramus annulatus Goldfuss, 1836 remains taxonomically problematic. Despite being based on an only medium-sized lectotype from the latest Turonian, the name was later applied to very large to huge forms characterized by a triangular-rounded outline, more widely and regularly spaced concentric rugae, an upright form and a much larger maximum size, first appearing in the mid-Early Coniacian. Thus, I. annulatus in the sense of Walaszczyk and Wood (1998) is poorly represented by its lectotype that we regard being very close the latest Turonian-earliest Coniacian species I. lusatiae Andert, 1911, a medium-sized representative of the lamarcki group. The complex nomenclatorial issues revolving around I. annulatus cannot be solved herein and further systematic studies are needed.
• I. annulatus sensu Walaszczyk and Wood (1998) is comparably rare and appears in the mid-Lower Coniacian Cremnoceramus waltersdorfensis hannovrensis Zone, ranging into the basal Middle Coniacian. • Valves of both I. stuemckei Heinz, 1928a andI. annulatus sensu Walaszczyk and are ornamented with regularly spaced, round-topped rugae, covered by fine, regularly spaced and slightly raised growth lines ("Anwachsringreifen" of Heinz 1928b). The presence/ absence of this feature in imprints and moulds is related to different preservation forms and cannot be used as a taxonomic criterion. • Inoceramus stuemckei Heinz, 1928a and I. annulatus sensu Walaszczyk and Wood (1998) are characterized by a stratigraphically abrupt enlargement of their shells and buttressing of the hinges across the Middle/Upper Turonian boundary interval and within the Early Conia- SaK 10537; large vertebra assigned to Ptychodus (cf. Schlesiger 2021). c MMG: SaK 10196; right claw of the durophagous crustacean Enoploclytia leachii (Mantell) cian, respectively. Up to 500 mm were reached in Late Turonian I. stuemckei, and Early Coniacian I. annulatus attain shell heights up to one meter. • The evolutionarily fairly abrupt shift to large/very large inoceramid shells in the Late Turonian is accompanied by widespread evidence for increased predation by durophagous organism such as sharks, mosasaurs and crustaceans. Size increase is a common passive defense strategy among (shelled) organisms in order to reach a size refugium in which the prey becomes unassailable for a predator. We thus speculate that the considerable size increase in Late Turonian to Early Coniacian inoceramid bivalves is part of an escalating arms race between prey and predators, ultimately triggering Late Cretaceous inoceramid gigantism. • We infer a recumbent lifestyle for the very large to huge, flat and disc-like I. stuemckei Heinz, 1928a and I. annulatus sensu Walaszczyk and Wood (1998) and rapid ontogenetic shell growth, reflected by the regularly spaced growth lines that most likely document subannual growth cycles.
Acknowledgements Constructive reviews by J. Todes (Chicago) and I. Walaszczyk (Warsaw) are gratefully acknowledged. Furthermore, I. Walaszczyk is thanked for further discussion in the aftermath of the review process. We also thank R. Winkler and M. Röthel (Senckenberg Naturhistorische Sammlungen Dresden) for digital photography of the MMG specimens and collection work, respectively. D. Erler (Senckenberg Naturhistorische Sammlungen Dresden) helped to obtain important literature. B. Gaitzsch and I. Kogan helped with the photographic documentation of the specimens housed in the geological collection of the TU Bergakademie Freiberg. C. Heunisch and A. Bornemann (Geozentrum Hannover) supported the collection work in the BGR in various ways. Our final thanks go to M. Amler (Köln) und M. Reich (Munich) for the editorial handling.
Funding Open Access funding enabled and organized by Projekt DEAL.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.