Limestone-marl alternations in epeiric sea settings – witnesses of environmental changes, or of rhythmic diagenesis?

Hildegard Westphal 1, Axel Munnecke 2, Florian Böhm 3, Stefan Bornholdt 4

1) Fachbereich Geowissenschaften, Universität Bremen, Marum Building, Leobener Straße, D-28359 Bremen, Germany, [email protected]

2) Institut für Paläontologie, Universität Erlangen, Loewenichstraße 28, D-91054 Erlangen, Germany, [email protected]

3) Leibniz-Institut für Meereswissenschaften, IfM-GEOMAR, Wischhofstraße 1-3, D- 24114 Kiel, Germany, [email protected]

4) Institut für Theoretische Physik, Universität Bremen, Otto-Hahn-Allee, D-28359 Bremen, Germany, [email protected]

For: Dynamics of Epeiric Seas: Sedimentological, Paleontological and Geochemical Perspectives ; edited by Chris Holmden and Brian R. Pratt; Geological Association of Canada Special Volume

“[A test] appears to indicate that many limestones can be produced solely

by rhythmic unmixing of CaCO 3 during diagenesis. Unless such limestones can be clearly distinguished from those that record genuine environmental signals, orbital cycle analysis based on such sequences will give meaningless results“(Hallam, 1986) Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 2

Abstract absolute concentrations). Systematic differences in diagenetically inert parameters can provide Limestone-marl alternations are widespread and unequivocal proof of primary differences. In the typical sediments of epeiric basins and are present in studied limestone-marl alternations, however, such variable abundance throughout the entire parameters do not directly reflect the lithological Phanerozoic. In many cases, their rhythmic rhythm, shedding doubt on limestone-marl appearance is interpreted as a direct response to alternations as direct archives of environmental orbital forcing. However, it is a challenge to change. unequivocally prove a sedimentary origin of the rhythmic intercalation of the two lithologies. This Box model computer simulations visualize difficulty arises from differential diagenesis that possible effects of early diagenetic change acting on alters limestone beds in different ways than limestone-marl alternations, independent of the interlayers (marls), causing a loss of comparability presence or absence of a primary rhythm. The between the lithologies. Differential diagenesis, simulations demonstrate that diagenesis has the between other effects, causes passive enrichment of potential to seriously distort any primary rhythm

Fig. 1: Occurrence of fine-grained calcareous rhythmites (limestone-marl alternations and nodular limestone successions) from literature compilation based on databases including ISI Web of Science, GEOREF and GEOBASE (data compilation available upon request from the authors). The reported occurrences are normalized for the time span of the interval they represent. Also shown: tropical shelf area (TSA) through time (dashed line, after Walker et al., 2002), and ratio of Mg to Ca ions in ocean waters through time (solid line, after Stanley and Hardie, 1998).

the inert non-carbonate fraction in interlayers, where present in the pristine sediment. In particular, calcium carbonate is being dissolved, as well as differential compaction acting mainly on the marl passive dilution in limestone beds, that are cemented interlayers induces distortions of the ratios of the by imported calcium carbonate. Therefore, original frequencies. These simulations emphasize unequivocal information about systematic differences the difficulties in conducting frequency analyses on in the precursor sediments of limestones and carbonate contents of real-world successions. interlayers therefore is preserved only in parameters that are not modified during diagenesis. Such Keywords for index: diagenetically inert parameters include the spectra of anactualism, aragonite, aragonite sea, calcite, calcite organic microfossils (but not their absolute sea, cellular automaton, cementation, compaction, concentration in the bulk sediment) and the ratios of computer simulation, diagenesis, differential diagenetically inert trace elements (again not the Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 3 diagenesis, dissolution, distortion, environmental more distal parts of the basin. (2) A shallow wave archive, environmental conditions, insolubles, base, as usually encountered in epeiric seas, and the limestone-marl alternation, Milankovitch, model, low relief prevent deposition of coarse-grained overprint, palynomorphs, sedimentary record, self- sediments in the distal parts of the basin and favor the organization, stratigraphic record, rhythmites fine-grained sedimentation typical for limestone-marl altenations. (3) Furthermore, the surrounding land provides the input of siliciclastics, of which, in a 1. Introduction – a brief review of limestone- sufficiently large epeiric basin, mostly clay-sized marl alternations material will reach the depositional site where it can form marls. Such an epeiric setting provides all ingredients Limestone-marl alternations are a widespread and for forming potentially sensitive recorders of climate characteristic facies of epeiric sea basins with famous cycles as fine-grained calcareous rhythmites. examples in the Ordovician of N-America and the Therefore, the cyclic patterns of fine-grained western and central High Atlas (Morocco), the calcareous rhythmites are often regarded as indicators Silurian of Northern Europe, the Triassic of the and recorders of orbitally forced climatic cycles (see Carpathians, the Mississippian of Montana, the chapters in de Boer and Smith, 1994 and Einsele and Jurassic and Cretaceous of Central and Western Ricken, 1991). In contrast to shallow-water carbonate Europe, and the Cretaceous of the Western Interior platfoms, the basinal setting allows for continuous Seaway and in Venezuela (e.g., Davaud and records even during sea-level lowstands. Lombard, 1975; Courtinat, 1993; Holmden et al., Nonetheless, the basinal sediments record the 1998; Samtleben et al., 2000; Westphal and environmental changes affecting the shallow-water Munnecke, 2003; Chacrone et al., 2004; Rey et al., platforms in their imported, shallow-water produced 2004; Tomasovytch, 2004; see also several chapters portion. The clay portion also potentially reflects in Einsele et al., 1991). Limestone-marl alternations climatic conditions that influence the weathering of are known from deposits of all Phanerozoic ages, the hinterland. In addition, varying nutrient fluxes even though their abundance varies strongly for the related to variations in weathering patterns are different geologic periods (Fig. 1; see also Westphal potentially recorded in paleontologic parameters such and Munnecke, 2003). The abundance roughly as calcareous nannofossil ratios. Thus, fine-grained follows the oscillations between calcite and aragonite calcareous rhythmites offer a wealth of parameters seas (Sandberg, 1983; Stanley and Hardie 1999) with that turn them into potential archives of climate high abundances during times of calcite seas and variations. lower abundances during times of aragonite seas. In Jurassic and Cretaceous epeiric sea deposits, such One difficulty, however, arises from the fact alternations are particularly widespread. that fine-grained calcareous rhythmites generally are strongly affected by differential diagenesis. Their Limestone-marl alternations are bimodal lithologic character of limestone beds characterized by their conspicuous outcrop intercalated in softer interlayers is largely a product appearance with a pronounced ABAB rhythm of of differential diagenesis, regardless of the presence more weathering-resistant limestone beds and softer or absence of primary sedimentary rhythms (see interbeds (see overview in Einsele and Ricken, 1991). below). This fact is still underestimated in many They can be viewed as part of a continuum of studies that interpret rhythms of various lithological, bimodal micritic alternations ranging from limestone- paleontological, or geochemical parameters in chalk to limestone-shale alternations, and including limestone-marl alternations as direct expressions of lithographic and nodular limestones, and well-bedded orbital climate forcing (cf. Fischer, 1980; Sprenger limestones (Munnecke and Samtleben, 1996; and ten Kate, 1993; Schwarzacher, 2000; Cleaveland Westphal et al., 2000; Munnecke et al., 2001; et al., 2002; Strasser et al. 2005). In this paper we Munnecke and Westphal, 2004). This group of will review and present published and new sedimentary facies is referred to as "fine-grained geochemical data and modeling results to illustrate calcareous rhythmites" in the present manuscript. the problems involved in interpreting fine-grained Large epeiric seas offer favourable conditions calcareous rhythmites as one-to-one recorders of for the generation of fine-grained calcareous climate cycles. rhythmites for several reasons (cf. various chapters in Einsele et al., 1991). (1) Tropical epeiric seas provide marginal areas for the production of shallow-water carbonate deposits that can be winnowed into the Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 4

1.1. What is so special about limestone- limestone beds was sourced from calcium carbonate marl alternations? dissolution in the interlayers (“donor” and “receptor” limestones of Bathurst [1971]; see also, e.g. , Ricken Fine-grained calcareous rhythmites show a wide 1986, 1987). Many researchers today favor a model range of morphological variations from limestone- for the origin of fine-grained calcareous rhythmites shale alternations to “classical” limestone-marl where subtle differences in primary carbonate content alternations, to lithographic limestone, and to nodular steer the diagenetic reinforcement that finally leads to limestone. They formed in both, well-oxygenated strong lithologic differences of the diagenetically waters, as indicated by strong bioturbation and mature rhythmite (e.g. , Reboulet and Atrops, 1997; benthic fossils, and under dysoxic or anoxic Schwarzacher, 2000; Hilgen et al., 2003). Other conditions, as indicated by internal lamination and researchers propose a purely diagenetic origin at least the absence of benthic fossils. for certain limestone-marl alternations (e.g., Hallam, 1964; Munnecke and Samtleben, 1996). Despite of the wide range of morphological variations, the two intercalated lithologies are clearly Two models on differential diagenesis of distinguishable in the field because the limestone fine-grained calcareous rhythmites are currently beds are more weathering-resistant than the discussed, that differ fundamentally in their interlayers. The boundaries between the two mechanisms. The now classical model of Ricken lithologies usually are relatively sharp. Within an (1986, 1987) involves pressure dissolution (chemical alternation, the limestone beds generally have higher compaction) of calcite in the interlayers in the deep- carbonate contents than the adjacent softer burial environment as source of the cement for the interlayers, even in successions where on a large limestone beds. In contrast, the model by Munnecke scale the carbonate content in each of the two and Samtleben (1996) is based on dissolution of lithologies varies strongly along the vertical aragonite in the shallow marine burial environment succession. In contrast to other rhythmically sensu Melim et al. (1995, 2002), where aragonite deposited facies such as glacial varves or tidalites, dissolution and reprecipitation as calcite cement fine-grained calcareous rhythmites show indications results from biogeochemically induced gradients in of strong diagenetic alteration that are fundamentally the pore-water (Raiswell, 1988; Munnecke and different for the two intercalated lithologies. The two Samtleben, 1996; Munnecke et al., 1997; Westphal, lithologies have followed two distinctly different 1998; Melim et al., 2002). diagenetic pathways in spite of their common burial Both models have in common that history. This is called differential diagenesis diagenesis takes place outside the reach of fresh (Reinhardt et al., 2000; Westphal et al., 2000), and water influence, and that the soft precursor sediments results in distinctly different characteristics for the of limestones and interlayers were much more alike two lithologies: Limestone beds in such rhythmites in terms of carbonate contents than the diagenetically are largely uncompacted, as indicated by the usually mature limestone beds and marls. A profound undeformed trace fossils and organic-walled difference between both models is the availability of microfossils (e.g., Kent, 1936; Hallam, 1964; carbonate cement. In the model of Munnecke and Henningsmoen, 1974; Jones et al., 1979; Möller and Samtleben (1996) the amount of cement carbonate is Kvingan, 1988; Raiswell, 1988; Westphal and limited by the amount of aragonite in the precursor Munnecke, 1997). This clearly indicates that sediments. When the entire aragonite in the interlayer cementation occurred during early diagenesis, i.e. has been dissolved, early diagenesis comes to a halt. prior to mechanical compaction. Interlayers, in In aragonite-poor sediments, for example, this contrast, are uncemented, resulting in higher necessarily results in a low limestone/interlayer vulnerability to weathering. They are always thickness ratio of the diagenetically mature intensely compacted as indicated by their deformed succession (Munnecke, 1997; Munnecke et al., 2001). trace fossils that in many cases form diagenetic In Ricken’s (1986) model, theoretically the entire pseudolamination (Kent, 1936; Sujkowski, 1958; calcium carbonate present in the precursor sediment Noble and Howells, 1974; Ricken, 1986, Munnecke of the interlayer layer can be dissolved by chemical and Samtleben, 1996). compaction. This differential diagenesis most conclusively The widely applied model of Ricken (1986, is explained by migration of calcium carbonate into 1987) faces the problem that limestone beds are the limestone layers where it forms cement occluding usually uncompacted, which contradicts cementation the initial porosity, whereas no calcium carbonate is in the deep burial diagenetic environment (Kent, imported into the interlayers. It is widely accepted 1936; Hallam, 1964; Henningsmoen, 1974; Jones et today that the calcium carbonate cement in the Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 5 al., 1979; Möller and Kvingan, 1988; Raiswell, 1988; several respects provide more favorable sedimentary Westphal and Munnecke, 1997). Petrographic settings for the deposition of aragonite-bearing observations that have caught differential diagenesis sediment than “aragonite times”, even though sea- “in act” in Neogene sediments from the Bahama water chemistry favors calcite precipitation. “Calcite slopes have revealed that cementation in the times” coincide with times of warmer global climate limestone beds occurs very early and prior to and extensive epeiric seas – both favorable conditions compaction, and at the same time, the interlayers are for carbonate-secreting communities including strongly compacted and show signs of dissolution in aragonitic organisms and thus for the formation of aragonitic components (Westphal, 1998). limestone-marl alternations. These conditions appear Additionally, studies of numerous successions have to have overrun the effect of sea-water chemistry (cf. demonstrated that while all other biotic associations Westphal and Munnecke, 2003; Munnecke and in limestone beds and interlayers might be identical Westphal, 2005). in limestones and interlayers, aragonitic remains only are preserved in limestones as molds or neomorphoses, whereas usually no traces of aragonitic skeletons are preserved in the interlayers 1.2. Environmental signals recorded in (Wepfer, 1926; Kent, 1936; Seibold and Seibold, fine-grained calcareous rhythmites 1953; Walther, 1982; Munnecke and Samtleben, 1996; Reboulet and Atrops, 1997), implying that the Limestone-marl alternations are usually interpreted to absence of originally aragonitic material is caused by directly reflect cyclic palaeoenvironmental signals. postdepositional processes. In contrast, primary The steering mechanisms for differential diagenesis, calcitic fossils down to the size of nanofossils are however, still are a matter of debate. It is tempting to present and usually well preserved both in limestones regard primary differences in sediment composition beds as well as in the interlayers, often with even (e.g., carbonate content, concentration of organic higher abundances in the interlayers compared to matter) as trigger. However, field observations show limestones (Hemleben, 1977; Munnecke and that in some cases primary differences in sediment Samtleben, 1996; Pittet and Mattioli, 2002). These composition do not match lithology. For example, in findings argue against differential diagenesis by Figure 2a, a shell layer passes laterally from calcite pressure dissolution in the deep-burial cemented limestone beds into uncemented environment and for aragonite dissolution in the early interlayers. This indicates that – at least in marine burial diagenesis environment as source of the exceptional cases – the same sediment (here the shell cement in the limestone beds. layer) can be transformed to both a limestone bed as well as an interlayer. On the other hand, some On this basis, the question arises if there is cemented limestone layers consist of different sufficient aragonitic sediment during times of “calcite lithologies, e.g. mudstone and grainstone within a seas” for aragonite-driven differential diagenesis. single bed (Fig. 2b). Here, a specific part of the More strikingly, why are fine-grained calcareous primary sediment is diagenetically “united” within a rhythmites more abundant in times of “calcite seas” single limestone bed by cementation despite of the compared to “aragonite seas” sensu Sandberg (1983) strong primary differences in grain size, carbonate (Fig. 1)? In contrast to the modern interval of an content, permeability etc. between the carbonate sand “aragonite sea”, in times of “calcite seas” non- (now the grainstone) and the carbonate mud (now the skeletal precipitates are thought to be predominantly calcitic. Nevertheless, aragonite-producing organisms mudstone). This cemented limestone is now flourish in “calcite times”, especially in tropical underlain and overlain by carbonate-depleted interlayers (not on the photograph). This indicates, settings, although probably less abundant or less that differential diagenesis in this case did not strictly strongly calcified (Stanley and Hardie, 1998, 1999; follow the strong primary differences in sediment Moñtanez, 2002; Ries, 2005). Therefore, the composition but has “overrun” these differences. relatively low amounts of about 10% initial aragonite (Munnecke et al.; 2001) required for driving Differential diagenesis has profound effects on differential diagenesis in the shallow marine burial the interpretation of calcareous rhythmites: It not environment have potentially been present at all only results in an increase in carbonate content in the times in such tropical settings. The fact that fine- limestone beds, whereas at the same time the grained calcareous rhythmites are much more carbonate content in the interlayers are impoverished. abundant in the rock record of times of “calcite seas” It also accounts for differences in fossil associations than in that of “aragonite seas” nevertheless appears and quantities, in the quantities of non-carbonate contradictory (Fig. 1). However, “calcite times” in constituents, and in petrophysical properties Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 6

Fig. 2: Sketch showing the transformation of soft sediments to lithified rocks by differential diagenesis. (A) Detail of the Lower Visby Formation (lowermost Wenlock, north of Högklint, Gotland, Sweden) showing that a specific sedimentary layer can be transformed to both limestone and interlayer. (B) Detail of the Upper Visby Formation (lower Wenlock, Hallshuk, Gotland, Sweden). Despite the pronounced primary differences both sediments have been united in one cemented limestone bed. including porosity, permeability, and sonic velocity (e.g., Munnecke and Samtleben, 1996). Secondly, (Bathurst 1987, 1991; Munnecke, 1997; Westphal, because of the different physical behavior of the two 1998; Kenter et al., 2002). Carbonate redistribution lithologies, in many cases they are studied using results in a passive enrichment of insolubles (mainly different methods (e.g., thin sections for the clay minerals and primary calcitic constituents) in limestone beds and sieving for the interlayers) that interlayers where aragonite is dissolved, and a are well suited for each of the two lithologies, but can passive dilution of the insolubles in the limestone introduce artificial differences in the observed faunal beds by cementation. The spectrum of calcareous spectra (e.g. , foraminifers, see Seibold and Seibold, macro- and microfossils in limestone beds and 1953). Thus, the absolute values of many measurable interlayers therefore in many cases are not directly parameters are not directly comparable for the two comparable in terms of primary differences between lithologies, and cannot be directly interpreted in the two lithologies for two reasons. Firstly, the terms of primary, pre-diagenetic, sediment properties. preservation potential in both lithologies is markedly These diagenetic peculiarities result in the different, and delicate fossils are commonly highly controversial discussion on their origin and destroyed by mechanical compaction in interlayers Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 7 interpretation that has a long tradition and is still been analyzed. They may also have been present in ongoing. Are such calcareous ABAB rhythmites the primary sediment parameters that were completely direct result of fluctuating environmental parameters, altered by the diagenetic processes, e.g., differences or are they created by diagenesis, or by a combination in primary porosity or in the content of organic of both (e.g. , Semper, 1917; Seibold, 1952; material. In this case, absence of proof is no proof of Sujkowski, 1958; Hallam, 1964, 1986; Schwarzacher absence. In any case, interpretations with respect to and Fischer, 1982; Ricken, 1986; Munnecke and environmental changes such as orbital forcing should Samtleben, 1996; Westphal et al., 2004b)? It appears not be undertaken until a sedimentary origin of the straightforward to interpret such successions as rhythm is unequivocally proven (Hallam, 1986). This manifestation of changes in input of siliciclastic holds also for frequency analyses, which are no proof sediment or primary carbonate productivity or both, for a primary, environmental cyclicity (see below). in many cases driven by orbital forcing. However, the diagenetic changes related to redistribution of calcium carbonate introduce a severe problem in comparability between the lithologies. Study of 2.1. Paleontological approach diagenetically inert parameters offers the possibility to look behind the veil of diagenesis. Diagenetically inert parameters such as palynomorph assemblages (organic-walled microfossils, e.g. acritarchs, dinoflagellate cysts, etc.) potentially 2. Diagenetically inert parameters as tool for survive alterations imposed by typical early carbonate assessing the record in a calcareous diagenesis, except for strong oxidation (Lind and rhythmite Schiøler, 1994). Systematic differences in the qualitative composition of the assemblages in limestone beds and interlayers would clearly prove As a result of differential diagenesis, the rhythmic changes of environmental conditions. interpretation of many sediment parameters is far Whereas many calcareous microfossils can be from straight-forward. Only parameters that are destroyed by compaction in the interlayers, robust against differential diagenetic change are truly simulating primary differences, palynomorphs are reliable as indicator of primary signals being generally flattened in the interlayers, but usually are represented in the rhythm of the alternation. Study of not destroyed, allowing for comparability of the diagenetically inert parameters is a robust way to assemblages in both limestone beds and interlayers assess the environmental fluctuations recorded in the (Westphal and Munnecke, 1997). Strong oxidation alternating lithologies of fine-grained calcareous that could result in modification or even destruction rhythmites, and to avoid uncertainties imposed by of the palynomorph content can be determined by the diagenetic alterations (Westphal et al., 2000, 2004b). presence/absence of amorphous organic matter Two approaches for proving unequivocal primary (AOM) because this part of the organic residue is signals appear promising: a paleontological and a most vulnerable to oxidation. Palynological bed-by- geochemical one. We here first shortly discuss the bed studies in fine-grained calcareous alternations are paleontological approach before introducing the rare to date because (a) palynologists usually are geochemical approach. We then present new results more interested in long-term trends and therefore of geochemical studies. sample at lower resolution, and (b) palynologists prefer sampling interlayers because the abundance of As mentioned above, there is a long-lasting palynomorphs normally is higher in interlayers discussion in the literature whether or not a compared to limestones (less acid required for diagenetic “unmixing” of a rather homogenous preparation). In the few existing studies, however, sediment is possible. Why is it so difficult to answer changes in the marine palynofacies are not related to this question? To prove an environmental cause of a the cyclical lithology (Bergman, 1987; Brenner, sedimentary rhythm is straight-forward: If limestones 1988; Courtinat, 1993; Westphal et al., 2000, 2004b; and interlayers show systematic differences in the Holstein, 2004). composition (not the quantities) of diagenetically stable constituents, a sedimentary origin of the rhythm is evident (Fig. 3 left arrow). However, if such a systematic difference is not observed (right 2.2. Geochemical approach arrow) it is, in contrast, no proof for a diagenetic origin. In the latter case systematic differences could Early marine burial diagenesis usually affects certain have existed in sediment parameters that have not chemical components of sediment much less than Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 8

Figure 3: Schematic drawing illustrating the “diagenetic dilemma”: Whereas a primary origin of a rhythmic deposit can be proven by systematic differences in the ratio of diagenetically stable components (e.g. differences in the palynomorph communities), it is impossible to prove a diagenetic origin. Even if no differences are observed in such ratios a primary origin of the rhythm is still possible because differences could have existed in either parameters that simply have not been measured, or in parameters that are completely altered by diagenetic processes (e.g. porosity, or content of organic matter).

calcium carbonate components. In particular the carbonate content; and the ratio of the concentrations components that are part of the siliciclastic portion of of such elements has to be studied (e.g.,Ti/Al). the sediment are less mobile than those of calcareous constituents. Many clay minerals and heavy minerals, Clay minerals represent the end product of such as rutile, usually are left unaltered by the early continental weathering, and are ultimately transported diagenetic processes that lead to the dissolution of into sedimentary basins. The climatic conditions and aragonite and reprecipitation of calcite (Bausch, the composition of the rock being weathered have a 1994; Bausch et al. 1994; Deconinck et al., 2003). profound influence on the type of clay being formed This diagenetic inertia offers a tool for assessing (Visser, 1991; Chamley, 1998; Thiry, 2000; Net et environmental conditions such as subaerial al., 2002; Ruffell et al., 2002; John et al., 2003). In a weathering in the provenance area of the siliciclastic setting where alternating environmental conditions portion of mixed successions such as limestone-marl generated a rhythmic succession of limestone beds alternations. Therefore, diagenetically inert elements and marl interlayers, these changing conditions may bound to these minerals (e.g. Ti, Al) potentially be reflected in a unique chemical composition of each preserve information on environmental conditions. lithology. Subsequent early marine burial diagenesis However, the absolute concentrations of these may alter the bulk elemental percentages by adding elements are shifted during diagenesis by dissolution or removing CaCO 3, but will not affect the of calcium carbonate in the interlayers and characteristic element ratios. Plotting two such cementation in the limestones. The values therefore element percentages (e.g. Al 2O3 and TiO 2) in a X/Y need to be normalized independently of calcium chart will show two separate populations of data points corresponding to the two lithologies, each following a different trend line. In particular the ratio Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 9

between Al 2O3/ TiO 2 appears suitable for detecting “Weltenburg”) do not reflect a primary sedimentary primary differences in the sediment, because the cyclicity (Fig. 4B). The data in each of the titanium concentration shows a wide range for successions plot on a single regression line with high different clay minerals. Other elements are less linear correlation coefficients (average r2 = 0.98). suitable for this approach, e.g., elements such This means that even though limestones are potassium that can adhere to clay minerals, show a characterized by lower contents of Al 2O3 and TiO 2 less stable behavior during early diagenesis. In the compared to the interlayers, both lithologies are literature, this rather simple test for a primary, non- indistinguishable in terms of the ratio of these diagenetic signal in a fine-grained calcareous diagenetically stable constituents. This uniformity of rhythmite, however, is rarely applied, and systematic Al/Ti ratios in limestones and interlayers was geochemical studies focusing on the ratios of confirmed by an additional statistical test. Assuming diagenetically stable constituents are rare (e.g. that Al/Ti ratios are equal in limestones and Bausch, 1992, 2001, 2004; Bausch et al., 1994; interlayers and that the observed variability is mainly Biernacka et al., 2005; Niebuhr, 2005). due to random variations, the Al/Ti ratios are expected to be normally distributed. As shown in the histogram insets in Figure 4B this appears to be the 3. Geochemical analyses case; in contrast to the clearly bimodal distribution of the "Trubi" Formation (Fig. 4A). The probabilities that the measured Al/Ti frequency distributions Several successions of fine-grained calcareous conform to a normal distribution are high in all cases, rhythmites were studied with this geochemical except for the "Trubi" Formation (Fig. 4). Only in approach, based on own data or on published data this latter case we can reject the hypothesis of a sets, by analyzing the ratios between Al 2O3 and TiO 2 normal Al/Ti distribution (goodness-of-fit !2 test; concentration. In order to investigate the reliability of Davis, 2002). Clearly, this is no proof that the this method we compare “classical” calcareous rhythmic successions summarized in Figure 4B rhythmites with a section that clearly reflects an represent entirely diagenetic rhythms. It only shows environmental rhythm (“Trubi” Formation): that a primary origin of these rhythms is not proven (1) The “Trubi” Formation; lower Pliocene; Punta di by the measured parameters (according to Fig. 3, Maiata, Sicily. In outcrop, the “Trubi” Formation right arrow). shows grey-white-beige-white quadripartite colour cycles. The grey layers consist of bioturbated marl with an average carbonate content of 67%, whereas 4. Computer simulations the beige coloured layers represent indistinctly laminated marls with an average carbonate content of 61%. The whitish layers, separating the two coloured marl types, have slightly elevated carbonate contents (average 72%; own unpublished data; cf. Nijenhuis et 4.1. Simulations as a tool to assess post- al., 1999). A primary, environmental origin of the depositional distortions alternating lithologies is evident from the differences in bioturbation and lamination between the grey and To gain a better understanding of the dynamics and beige marls. Additionally, the two different marl complex effects of differential diagenesis, we created types are clearly distinguished in the geochemical a computer model that implements in a schematic data that show a bimodal distribution in their Al/Ti way the diagenesis model of aragonite dissolution ratios (Fig. 4A). The white layers are less well and calcite reprecipitation described above. The defined by Al/Ti ratios. They may represent a variety crucial questions to be answered by the model of the marls that was diagenetically enriched in approach are: What had diagenetically overprinted calcium carbonate. Both, sedimentological and successions looked like prior to diagenesis? Is a geochemical data unequivocally prove a primary, rhythmic variation required in the original non-diagenetic origin of the cyclicity in the “Trubi” composition of the sedimentary succession, or can Formation. differential diagenesis generate a cyclic succession (2) “Classical” limestone-marl alternations of from homogeneous sediment? Can differential Permian, Jurassic, and Cretaceous age. Without the diagenesis transform random variations in the pristine development of two different populations, the sedimentary composition into a cyclic succession?

Al 2O3/TiO 2 ratios of five "classical" limestone-marl On the other hand, if there were a cyclic succession alternations and one Plattenkalk succession (Fig. 4B of pristine sediments, how would differential Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 10

Fig. 4: Al 2O3/TiO 2 data from different sections. (A) Data from “Trubi” Formation (S-Sicily) plot on two separate trend lines, indicating a bimodal chemical composition of the non-carbonate fraction, and, thereby varying environmental conditions.. (B) Data from several calcareous rhythmic successions (own and published data) with a linear correlation. Frequency distributions of the Al/Ti ratios are shown in a histogram for each section. A normal distribution is expected if the Al/Ti ratios vary randomly about a single mean value. More complex variations, producing a multimodal distribution, are expected for non-random (e.g. rhythmic) fluctuations. The probability (p) that a measured Al/Ti frequency distribution conforms to a normal distribution was calculated with a goodness-of-fit !2 test (Davis, 2002). The p values allow only for the “Trubi” Marls to reject the hypothesis of a normal Al/Ti distribution. diagenesis alter the original cyclic signal? We here compaction) to answer these questions. This employ a two-dimensional computer simulation of approach allows us to study lateral as well as vertical the relevant aspects of differential diagenesis processes. (aragonite dissolution, calcite reprecipitation; Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 11

4.2. Approach and method state, “aragonitic”, represents a pristine sediment consisting of aragonite, calcite and clay minerals; (2) The computer simulation is based on a cellular The second state, “non-aragonitic”, represents either automaton. Space is divided into cells on a a pristine, originally aragonite-free sediment, or a rectangular lattice, where each cell models a diagenetically altered, aragonite-depleted sediment macroscopic volume of the sediment such that the consisting of calcite and clay minerals; and (3) the whole field can represent a geologic succession with third state, “cemented”, represents sediment in which vertical and lateral dimensions. The interaction calcite cement is added to the pristine (aragonitic or between the cells implements the aragonite non-aragonitic) sediment. In the pristine sediment, dissolution and calcite reprecipitation dynamics of “aragonitic” and “non-aragonitic” cells are randomly the rhythmic diagenesis model by Munnecke and distributed in each new layer. For implementing a Samtleben (1996) described above. It thus simulates time-series signal, the percentage of “aragonitic” cells the diagenetic redistribution of calcium carbonate in the pristine layers is varied randomly or during early diagenesis in the shallow marine burial systematically. E.g., to illustrate the effects of environment. The computer model originally has rhythmic diagenesis on a cyclic sedimentary been designed to produce output in the time domain, succession we use the 65°N summer insolation curve i.e., for constant sedimentation rates and without of Berger and Loutre (1991) to which we couple the compaction (Böhm et al., 2003; Westphal et al., percentage of “aragonite” cells in a layer. The 2004a). Here we present a modified version where insolation curve includes the influences of compaction has been implemented. A brief precession, obliquity, and eccentricity, i.e., the basic description of the computer model is given below; for “Milankovitch parameters”. It represents a complex a detailed description see Böhm et al. (2003) and cyclic signal that has variance in a range of different Westphal et al. (2004a). frequency bands of geologically meaningful wavelengths. Similar to standard stratigraphic simulations, the model is a cellular automaton model setup The model diagenetic environment is layered consisting of a rectangular matrix of cells. Here a into several zones (Fig. 5): (i) The uppermost layer is version is employed, where each cell can assume a non-reactive zone below the sediment-water only discrete states. In our case the model requires interface. (ii) The zone underneath is a cementation three possible states of a cell (Fig. 5): (1) The first zone, where calcite cement is precipitated. (iii) Further below, a non-reactive zone may be present

Fig. 5: (A) Simplified diagenesis model for the computer implementation. Sediment is progressively buried and moves through stable diagenetic zones where aragonite dissolution and calcite cementation take place. Time step I: layer B is cemented by calcite cement derived from dissolution of aragonite in layer A. Time step II: The cemented layer B enters the aragonite dissolution zone, but the aragonite present in layer B is “sealed” by cement, so little to no dissolution takes place; therefore layer C is not being cemented. Time step III: The uncemented layer C enters the aragonite dissolution zone, aragonite is dissolved and reprecipitated as calcite cement in layer D. This model demonstrates the possibility of self-organized diagenetic patterns. (B) Cellular automata model for diagenesis simulations. With each time step, cells representing the sediment move one layer downward. Three possible states of cells are: 1 “aragonitic”: consisting of aragonite, calcite, and terrigenous material; 2 “cemented”: calcite cement is added; 3 “non-aragonitic”: consisting of calcite and terrigenous material, either originally aragonite-free, or diagenetically aragonite-depleted. As a layer passes through the aragonite dissolution zone, aragonite is removed from state 1 cells. In this simulation run, for each cell where aragonite is dissolved, two cells in the cementation zone are cemented. An insolation signal is implemented by varying the percentage of “aragonitic” cells (after Böhm et al., 2003 and Westphal et al., 2004). Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 12 that is situated between the cementation zone and (iv) the underlying aragonite dissolution zone. (v) Finally, a non-reactive “historical zone” follows where no further shallow marine-burial diagenesis takes place. In the model, the thickness of the individual zones can be varied. In the dissolution zone, “aragonitic” cells are transformed to “non-aragonitic” cells. In the cementation zone, “aragonitic” and “non-aragonitic” cells are transformed to “cemented” cells. The number of the latter is limited by the number of cells that are dissolved in the dissolution zone. Layers of dominantly “cemented” cells form diagenetically mature limestone beds, whereas layers of dominantly “non-aragonitic” cells form diagenetically mature marl interlayers. For the simulations shown here, two cells are cemented per dissolved aragonite cell. In all sequential operations, cells are accessed in random order to avoid artificial anisotropies. The simulation output is a sediment column with cemented and aragonite-free cells (Figure 6 gives an example of a simulated succession). Earlier simulations with the computer program used here (Böhm et al., 2003; Westphal et al., 2004a) did not consider further distortion of the primary signal by differential compaction. Here we extend the model with the ability to simulate differential compaction of the sediment column. Differential compaction generates uncertainties in the depth-to- time transformation (Hinnov, 2000). In fine-grained calcareous rhythmites cemented limestone beds usually are not compacted whereas interbeds always are compacted (see above). This early compaction is a combination of loss of primary porosity and volume loss by aragonite dissolution. A section measured in a rock succession, i.e., in the depth domain, therefore cannot be directly fed into frequency analysis, i.e., into the time domain. Differential compaction in a fine-grained calcareous rhythmite distorts any primary signal. Decompaction calculations that distinguish between the two intercalated lithologies are required to original cellular automaton model is a succession of layers that, as sedimentation rates are held constant, represent equal time of sediment accumulation. The compaction algorithm uses as Fig. 6: Example of simulated diagenetically mature fine-grained input the average aragonite content of each individual calcareous rhythmites. The modeled sequence includes a layer, measured before diagenetic alteration, and the Milankovitch signal (65°N summer insolation spectrum; average cement content of the same layer, measured Berger and Loutre, 1991) that is distorted by the diagenetic after diagenetic alteration. The thickness of each overprint. Left: Milankovitch signal fed into the simulation. Right: sedimentary column of cemented “limestone beds” layer is then reduced according to these two variables and uncemented “interbeds”. Conspicuous intervals with as follows. Layers with cement content above a decoupling of neighboring cell domains and subsequent specified threshold value, Ct (40% in the standard reorganization of well-defined layering are indicated on the calculation, 70% in the enhanced-compaction right. (Parameters used for simulation: diagenetic cycle period: 50 ka; distance between cementation and dissolution experiment), are not compacted. The thickness of all layer: 25 layers; simulation time step: 1 ka (1 layer per ka); layers with less cement is reduced by 20% to account simulated time: 3 Ma (3003 layers.)). for mechanical compaction of primary pores. Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 13

Additional compaction of these layers results from such layers reduces compaction and is accounted for volume loss due to the dissolution of aragonite. In our by subtracting the cement portion from the aragonite experiments we define that aragonite dissolution portion. The compacted layer thickness (CLT) is leads to a compaction of 75% of the aragonitic therefore calculated as portion of each layer in addition to the 20% mechanical compaction. The remaining 5% account for non-dissolvable residuals including primary CLT = 1 - k * (0.2 + 0.75 * (ara - cem)) calcite and non-carbonate materials. As a result, the maximal compaction of an uncemented aragonitic layer is 95%, which would correspond to a parting where the initial layer thickness is 1, k is a between limestone beds. Slight early cementation of scaling factor, ara and cem are the aragonite and

Fig. 7: Power spectra of simulated rhythmic sequences (bold lines) overprinted by different diagenetic cycle frequencies compared to primary signal (65°N summer insolation spectrum; Berger and Loutre, 1991) (fine lines). The original frequency spectra are modified in four different ways: (A) Amplification: Minor peak in the insolation spectrum (11 ka) is amplified by resonance with the diagenetic cyclicity. (Note: diagenetic cement spectrum shows harmonics due to the rectangular shape of the diagenetic cycles.) (B) Emergence: spectral peaks (114 ka and 12 ka) not present in the input signal emerge. (C) Suppression: precession frequencies are suppressed. (D) Displacement: obliquity cycle period (41 ka) displaced towards shorter wavelength. MEM=maximum entropy method. (From Westphal et al., 2004a) Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 14 cement content in vol%, respectively, and ara > cem . series of cement content values for frequency For ara < cem the last term is set to zero. The scaling analysis (which corresponds to equal sampling factor k was varied between 0 (no compaction) and 1 distance in real-world successions), the sequences (full compaction) to study the impact of increasing need to be resampled. Frequency spectra of the compaction intensity on the recorded signal. resulting sequences were analyzed with the maximum entropy method (MEM) using the SSA-MTM toolkit The cement content of a compacted layer has of Ghil et al. (2002). to be recalculated, because this value represents a volume-based percentage in our model and therefore changes with compaction: 4.3. Results of earlier simulations without compaction cem c = cem i / CLT In simulations with the diagenetic cellular automaton model, differential diagenesis generates laterally where subscripts c and i refer to compacted and extended layers of fine-grained calcareous rhythmites initial layers, respectively. only where a primary sedimentary signal The compaction procedure results in a synchronizes the oscillating cells to form continuous sequence of layers of non-uniform thickness and beds. Otherwise, randomly distributed nodules form varying cement content. To generate an equal-spaced due to the limited horizontal coupling of the

Fig. 8: Effects of varying compaction on the frequency spectra of modeled fine-grained calcareous rhythmites. Simulated sediment column is given in Figure 6, uncompacted frequency spectrum is shown in Figure 7C. Compaction further disturbs the spectrum. Ten spectra are plotted in each panel with compaction intensity (k) increasing from top to bottom. Corresponding peaks are connected with grey bars. With increasing compaction spectral peaks are shifted towards higher frequencies. (A) Standard compaction scenario with 40% specific

threshold value C t (for explanation of this value see text). (B): In the enhanced compaction scenario (C t = 70%) an additional peak occurs at a periodicity of 19-20 ka at high compaction intensities (k>0.4). High-frequency peaks (periods <11 ka) show no consistent trends with increasing compaction. Spectra are calculated in the linear depth domain. Peaks are labeled with the corresponding time domain periodicities. Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 15 diagenetic behavior of neighboring cells (Böhm et al., 4.4. New simulations: the effect of 2003). With typical diffusion coefficients and differential compaction compactional flow velocities for fine-grained sediments, diffusive transport prevails only at In addition to the distortions of the primary distances of less than a few meters (Berner 1980). At sedimentary signals by dissolution and cementation, larger distances transport will be dominated by the vertical compactional flow. This limits the horizontal differential compaction further distorts the spectral range of chemical interactions to scales that are much characteristics of fine-grained calcareous rhythmites. smaller than those typical for the laterally extensive This compaction affects individual limestone and beds of fine-grained calcareous rhythmites. interlayers with different intensities depending on the primary aragonite content and the degree of An external signal, implemented in the precompactional cementation. Therefore the model by variations in the aragonite portion, distortions are not removed by simple linear synchronizes the diagenetic bedding; even a random decompaction algorithms. In addition, bio- or or low frequency external signal imposes sufficient magnetostratigraphic resolution is usually not control to synchronize the high-frequency couplets. sufficient to resolve differential compaction effects, Temporal decoupling of neighboring cell domains is so they cannot be accounted for in time-depth shown in Figure 6 (middle and upper part of the models. column). Phase offsets between neighboring cell domains are caused by small random disturbances, A straightforward effect of the decreased thickness of either when the primary signal becomes too weak or interlayers due to compaction is that the spectral when the diagenetic cycle and the primary signal peaks are shifted towards shorter periodicities with have different frequencies and get out of phase. increasing compaction intensity (Fig. 8). Higher Synchronization by a strong primary signal, in phase frequencies are generally more affected than lower with the diagenetic oscillator, can reorganize the cell ones, leading to an increasing distortion of the domains to form laterally extended homogenous frequency ratios with increasing compaction. This is layers (Fig. 6). particularly relevant as these ratios are often used by Thus, according to our simulations and sedimentologists to identify Milankovitch cyclicities general considerations, a primary sedimentary signal in sedimentary sequences. The most prominent is a necessary prerequisite for generating laterally example for such frequency ratios is the 1:5 bundling extensive fine-grained calcareous rhythmites. A of the eccentricity/precession relationship that is purely self-organized origin of laterally extensive often used as a spectral Milankovitch fingerprint beds is not possible with the currently available (Schwarzacher, 2000). Finally, with very strong diagenetic concepts. On the other hand, our compaction, spurious spectral peaks may emerge, simulations show that differential diagenesis is able further obscuring the signal (Fig. 8). In summary, our to distort the primary sedimentary signal in a number simulations show that a mature limestone-marl of ways. A primary (regular, random, or alternation may constitute a very poor representation Milankovitch) signal undergoes dramatic changes in of its original sedimentary precursor sequence, character during diagenesis. The frequency of the especially with respect to its frequency signal not only may be shifted (Böhm et al., 2003), characteristics. but in addition new frequencies may emerge, other frequencies may be suppressed, and some frequencies may be amplified while others are reduced (Fig. 7; Westphal et al., 2004a). These distortions of the original frequencies have serious implications for the 5. Conclusions interpretation of diagenetically mature fine-grained calcareous rhythmites. Carbonate contents and lithology (limestone beds versus interlayers) cannot In this article we reviewed calcareous directly be interpreted in terms of frequencies of rhythmites (including the “classical” limestone-marl original environmental signals without independent alternations) and their diagenesis, both, from the proof (paleontological, geochemical) of the primary perspective of geological models as well as with the origin of the intercalated lithologies. aid of a computer simulation that makes visible the spatio-temporal dynamics of the diagenesis model. Fine-grained calcareous rhythmites have undergone differential diagenesis. Cemented, uncompacted limestone beds are intercalated in uncemented, compacted interlayers. Despite the distinct cyclicity Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 16 seen in outcrop or core, however, it is surprisingly understood for reliably interpreting fine-grained difficult to determine the presence of an rhythmic carbonate successions. environmental trigger forcing or underlying the lithologic intercalation. Differential diagenesis modifies many sediment parameters in different ways Acknowledgements in limestone beds and interlayers, destroying comparability. The investigation of diagenetically inert parameters including ratios of trace elements This study was supported by the German Science proves an important tool for approaching this Foundation DFG (Fr-1134/4 und We-2492/1) and by question, because these ratios are not altered by a HWP grant of the University of Erlangen/State of differential diagenesis. Clearly, searching for Bavaria to HW. We gratefully acknowledge critical unequivocal fingerprints of primary signals is comments by Paul Myrow and Mario Coniglio. The essential for interpreting such successions, e.g. , in geochemical data of the “Trubi” Formation samples terms of orbital forcing and chronostratigraphy. were produced as part of Martin Wolf’s Master’s Frequency analyses are only reliable for successions thesis. where a sedimentary origin is clearly established, and where the differential compaction of limestone beds and interlayers can be accounted for. Surprisingly, none of the investigated alternations shows unequivocal evidence for environmental changes References between limestones and interlayers.

The simulation results emphasize the possible Bathurst, R.G.C., 1971, Carbonate Sediments and dramatic effects of differential diagenesis, i.e., Their Diagenesis: Amsterdam, Elsevier, aragonite dissolution in interlayers and cementation Developments in Sedimentology, v. 12, 620 p. in limestone beds, in overprinting and even generating lithologic alternations. Diagenetic Bathurst, R.G.C., 1987, Diagenetically enhanced redistribution of calcium carbonate and differential bedding in argillaceous platform limestones: compaction has the potential to severely distort any stratified cementation and selective primary signals recorded in the sedimentary column. compaction: Sedimentology v. 34, p. The processes of differential diagenesis potentially 749-778. result in amplification, displacement, suppression, Bathurst, R.G.C., 1991, Pressure-dissolution and and even in emergence of new cycle frequencies. limestone bedding: the influence of stratified Cyclic external input is not needed for the formation cementation, in Einsele, G., Ricken, W. and of cyclically alternating limestone beds and Seilacher, A. eds., Cycles and Events in interlayers. Stochastic fluctuations in environmental Stratigraphy, Springer Verlag, Berlin, p. 450- parameters appear sufficient for the production of 463. laterally extensive beds over distances much larger than nodule size. If external periodic signals are Bausch, W.M., 1992, Geochemische Kennzeichnung recorded in the pristine sediment, these signals may des Profils Weltenburg: Erlanger Beiträge be severely distorted in the diagenetically mature Petrographie Mineralogie, v. 2, p. 1-14. succession. The simulations imply that even though Bausch, W.M., 1994, Geochemische Analyse von primary, external fluctuations are required for the Karbonatgesteinssequenzen: Abhandlungen formation of cyclically intercalated limestone beds der Geologischen Bundesanstalt, v. 50, p. 25- and interlayers, the original frequency spectrum of 29. such primary signals may not survive differential diagenesis. Compaction additionally modifies the Bausch, W.M., 2001, Mineralogie und Geochemie frequency spectrum of cycles. The thoroughly des Profils „Frankenschotter“ im dynamic processes of diagenesis thus strongly Treuchtlinger Marmor (und die Kalk-Mergel- influence the distribution of the interlayered Hyperbel): Erlanger Beiträge Petrographie lithologies in fine-grained calcareous rhythmites in a Mineralogie, v. 11, p. 9-45. non-trivial way where new spectral peaks may Bausch, W.M., 2004, Geochemie des Profils emerge that do not carry any information on primary, Maxberg/Solnhofen: Erlanger Beiträge environmental, frequencies. Diagenetic effects are far Petrographie Mineralogie, v. 14, p. 35-51. more than a “bother” that can be neglected – they have to be critically examined and ideally should be Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 17

Bausch, W.M., Viohl, G., Bernier, P., Barale, G., Interior Basin, USA: Marine Bourseau, J.-P., Buffetaut, E., Gaillard, C., Micropaleontolgy, v. 21, p. 249-257. Gall, J.C. and Wenz, S., 1994, Eichstätt and Davaud, E. and Lombard, A., 1975, Statistical Cerin: Geochemical comparison and definition approach to the problem of alternating beds of of two different Plattenkalk types: Geobios, limestone and marl (Upper Oxfordian of the M.S., v. 16, p. 107-125. French Jura): Eclogae geol. Helv., v 68(3), p. Berger, A. and Loutre, M.F., 1991, Insolation values 491-509. for the climate of the last 10 million of years: Davis J.C., 2002, Statistics and Data Analysis in Quaternary Science Reviews, v. 10, p. Geology. 3rd edition, 638 p., John Wiley & 297–317. Sons, New York. Bergman, C. F., 1987, Silurian paulitinid jawed de Boer, P.L. and Smith, D.G., eds., 1994, Orbital polychaetes from Gotland: PhD Dissertation Forcing and Cyclic Sequences: Special University Lund (Sweden), Lund, 205 pp. Publication of the International Association of Berner, R.A., 1980, Early diagenesis. A theoretical Sedimentologists, v. 19, 559 p. approach. Princeton, Princeton University Deconinck, J.F., Hesselbo, S.P., Debuisser, N., Press, 241p. Averbuch, O., Baudin, F. and Bessa, J., 2003, Biernacka, J., Borysiuk, K. and Raczynski, P., 2005, Environmental controls on clay mineralogy of The Zechstein (Ca1) limestone-marl an Early Jurassic mudrock (Blue Lias alternations from the North-Sudetic Basin, Formation, southern England: International Journal of Earth Sciences, v. 92(2), p. 255- Poland: depositional or diagenetic rhythms?: 266. Geological Quarterly, v. 49(1), p. 1-14. Einsele, G. and Ricken, W. 1991, Limestone-marl Böhm, F., Westphal, H. and Bornholdt, S., 2003, alternations - an overview, in Einsele, G., Required but disguised: Environmental signals Ricken, W. and Seilacher, A. eds., Cycles and in limestone-marl alternations: Events in Stratigraphy, Springer Verlag, P a l a e o g e o g r a p h y , Palaeoclimatology, Berlin, p. 23-47. Palaeoecology, v. 189, p. 161-178. Einsele, G., Ricken, W. and Seilacher, A. (eds.), Brenner, W., 1988, Dinoflagellaten aus dem Unteren 1991, Cycles and Events in Stratigraphy, Malm (Oberer Jura) von Süddeutschland; Springer Verlag, Berlin, 955 pp. Morphologie, Ökologie, Stratigraphie: T ü b i n g e r Mikropaläontologische Fischer, A.G., 1980, Gilbert-bedding rhythms and Mitteilungen, v. 6, p. 1-157. geochronology: Geological Society of America Special Paper, v. 183, p. 93-104. Chacrone, C., Hamoumi, N. and Attou, A., 2004, Climatic and tectonic control of Ordovician Ghil, M., Allen, R.M., Dettinger, M.D., Ide, K., sedimentation in the western and central High Kondrashov, D., Mann, M.E., Robertson, A., Atlas (Morocco): Journal of African Earth Saunders, A., Tian, Y., Varadi, F. and Yiou, Sciences, v. 39, p. 329-336. P., 2002, Advanced spectral methods for climatic time series: Reviews of Geophysics, Chamley, H., 1998, Clay mineral sedimentation in v. 40, 1003, doi:10.1029/2000RG000092. the Ocean, in Paquet, H., and Clauer, N., eds., Soils and sediments (mineralogy and Hallam, A., 1964, Origin of the limestone-shale geochemistry): Berlin, Springer-Verlag, p. rythm in the Blue Lias of England: a 269–302. composite theory: Journal of Geology, v. 72, p. 157-169. Cleaveland, L.C., Jensen, J., Goese, S., Bice, D.M. and Montanari, A., 2002, Cyclostratigraphic Hallam, A., 1986, Origin of minor limestone-shale analysis of pelagic carbonates at Monte dei cycles: Climatically induced or diagenetic?: Corvi (Ancona, Italy) and astronomical Geology, v. 14, p. 609-612. correlation of the Serravallian-Tortonian boundary: Geology, v. 30(10), p. 931-934. Hemleben, C., 1977, Autochthone und allochthone Sedimentanteile in den Solnhofener Courtinat, B., 1993, The significance of palynofacies Plattenkalken: Neues Jahrbuch für Geologie fluctiations in the Greenhorn Formation und Paläontologie, Monatshefte, v. 5, p. 257- (Cenomanian-Turonian) of the Western 271. Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 18

Henningsmoen, G., 1974, A comment. Origin of to identify diagenetic environments: Geology, limestone nodules in the Lower Palaeozoic of v. 23(8), p. 755-758. the Oslo Region: Norsk Geologisk Tidsskrift, Melim, L.A, Westphal, H., Swart, P.K., Eberli, G.P. v. 54, p. 401-412. and Munnecke, A., 2002, Questioning Hilgen, F.J., Abdul Aziz, H., Krijgsman, W., Raffi, I. carbonate diagenetic paradigms: evidence and Turco, E., 2003, Integrated stratigraphy from the Neogene of the Bahamas: Marine and astronomical tuning of the Serravallian Geology, v. 185, p. 27-53. and lower Tortonian at Monte dei Corvi Möller, N.K. and Kvingan, K., 1988, The genesis of (Middle-Upper Miocene, northern Italy): nodular limestones in the Ordovician and P a l a e o g e o g r a p h y , Palaeoclimatology, Silurian of the Oslo Region: Sedimentology, Palaeoecology, v. 199, p. 229-264. v. 35, p. 405-420. Hinnov, L.A., 2000, New perspectives on orbitally Moñtanez, I.P., 2002, Biological skeletal carbonate forced stratigraphy: Annual Review Earth records changes in major-ion chemistry of Planetary Sciences, v. 28, p. 419-475. paleo-oceans: PNAS, v. 99(25), p. 15852- Holmden, C., Creaser, R.A., Muehlenbachs, K., 15854. Leslie, S.A. and Bergström, S.M., 1998, Munnecke, A., 1997, Bildung mikritischer Kalke im Isotopic evidence for geochemical decoupling Silur auf Gotland: Courier Forschungsinstitut between ancient epeiric seas and bordering Senckenberg, v. 198, p. 1-71. oceans: implications for secular curves: Geology, v. 26(6), p. 567-570. Munnecke, A. and Samtleben, C., 1996, The formation of micritic limestones and the Holstein, B., 2004, Palynologische Untersuchungen development of limestone-marl alternations in der Kössener Schichten (Rhät, Alpine the Silurian of Gotland, Sweden: Facies, v. 34, Obertrias): Jahrbuch der Geologischen p. 159-176. Bundesanstalt, v. 144, p. 261-365. Munnecke, A. and Westphal, H., 2004, Shallow- John, C.M., Mutti, M. and Adatte, T., 2003, Mixed water aragonite recorded in bundles of carbonate-siliciclastic record on the North limestone-marl alternations – the Upper African margin (Malta) - coupling of Jurassic of SW Germany: Sedimentary weathering processes and mid Miocene Geology, v. 164(3/4), p. 191-202. climate: GSA Bulletin, v. 115(2), p. 217-229. Munnecke, A. and Westphal, H., 2005, Variations in Jones, B., Oldershaw, A.E. and Narbonne, G.M., primary aragonite, calcite, and clay in fine- 1979, Nature and origin of rubbly limestone in grained calcareous rhythmites of Cambrian to the Upper Silurian Read Bay Formation of Jurassic age – an environmental archive?: Arctic Canada: Sedimentary Geology, v. 24, p. Facies, v. 51, p. 611-626. 227-252. Munnecke, A., Westphal, H., Reijmer, J.J.G. and Kent, P.E., 1936, The formation of the hydraulic Samtleben, C., 1997, Microspar development limestones of the Lower Lias: Geological during early marine burial diagenesis: a Magazine, v. 73, p. 476-478. comparison of Pliocene carbonates from the Kenter, J.A.M., Anselmetti, F.S. , Kramer, P., Bahamas with Silurian limestones from Gotland Westphal, H. and Vandamme, M.G.M., 2002, (Sweden): Sedimentology, v. 44(6), p. 977-990. Acoustic properties of "young" carbonate Munnecke, A., Westphal, H., Elrick, M. and Reijmer, rocks, ODP Leg 166 and Holes Clino and J.J.G., 2001, The mineralogical composition of Unda, Western Great Bahama Bank: Journal precursor sediments of calcareous rhythmites - a of Sedimentary Research, v.72, p. 129-137. new approach: International Journal of Earth Lind, I. and Schiøler, P., 1994, Dinoflagellate Sciences, v. 90, p. 795-812. concentration as an independent reference for Net, L.I., Alonso, M.S. and Limarino, C.O., 2002, monitoring mineral mobilization in stylolites: Source rock and environmental control on clay Sedimentary Geology, v. 92, p. 53–65. mineral associations, Lower Section of Paganzo Melim, L A., Swart P K. and Maliva, R.G., 1995, Group (Carboniferous), Northwest Argentina: Meteoric-like fabrics forming in marine Sedimentary Geology, v. 152, 183-199. waters: Implications for the use of petrography Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 19

Niebuhr, B., 2005, Geochemistry and time-series Ricken, W., 1987, The carbonate compaction law: a analyses of orbitally forced Upper Cretaceous new tool: Sedimentology, v. 34, p. 1-14. marl–limestone rhythmites (Lehrte West Ries, J.B., 2005, Aragonite production in calcite seas: Syncline, northern Germany): Geological effect of seawater Mg/Ca ratio and the Magazine, v. 142, p. 31-55. calcification and growth of the calcareous alga Nijenhuis, I.A., Lourens, L.J. and de Lange, G.J., Penicillus capitatus: Paleobiology, v. 31(3), p. 1999, Diatomites at Lido Rossello: 445-458. Geochemistry and implications for formation Ruffell, A.H., Price, G.D., Mutterlose, J., Kessels, K., models of the "Trubi" marls, in Nijenhuis, Baraboshkin, E. and Gröcke, D.R., 2002, I.A., Geochemistry of eastern Mediterranean Palaeoclimate indicators (clay minerals, sedimentary cycles: on the origin of Miocene calcareous nannofossils, stable isotopes) to Pleistocene sapropels, laminites and compared from two successions in the late diatomites: Geologica Ulraiectina, v. 167, p. Jurassic of the Volga Basin (SE Russia): 56-73. Geological Journal, v. 37, p. 17-33. Noble, J.P.A. and Howells, K.D.M., 1974, Early Samtleben, C., Munnecke, A. and Bickert, T., 2000, marine lithification of the nodular limestones Development of facies and C/O-isotopes in in the Silurian of New Brunswick: transects through the Ludlow of Gotland: Sedimentology, v. 21, p. 597-609. Evidence for global and local influences on a shallow-marine environment: Facies, v. 43, p. 1- Pittet, B. and Mattioli, E., 2002: The carbonate signal 38. and calcareous nannofossil distribution in an Upper Jurassic section (Balingen-Tieringen, Sandberg, P. A., 1983, An oscillating trend in Late Oxfordian, southern Germany): Phanerozoic non-skeletal carbonate P a l a e o g e o g r a p h y , Palaeoclimatology, mineralogy: Nature, v. 305, p. 19-22. Palaeoecology, v. 179, p. 71-96. Schwarzacher, W., 2000, Repetitions and cycles in Raiswell, R., 1988, Chemical model for the origin of stratigraphy: Earth Science Reviews, v. 50, p. minor limestone-shale cycles by anaerobic 51-75. methane oxidation: Geology, v. 16, p. 641- Schwarzacher, W. and Fischer, A.G., 1982, 644. Limestone-shale bedding and pertubations of Reboulet, S. and Atrops, F., 1997, Quantitative the earth's orbit, in Einsele, G. and Seilacher, variations in the Valanginian ammonite fauna A., eds., Cyclic and Event Stratification: of the Vocontian Basin (southeastern France) Springer, Berlin, p. 72-95. within limestone-marl cycles and within Seibold, E., 1952, Chemische Untersuchungen zur parasequence sets: Palaeogeography, Bankung im unteren Malm Schwabens: Neues Palaeoclimatology, Palaeoecology, v. 135, p. Jahrbuch für Geologie und Paläontologie - 145-155. Abhandlungen, v. 95, p. 337-370. Reinhardt, E.G., Cavazza, W., Patterson, R.T. and Seibold, E. and Seibold, I., 1953. Foraminiferenfauna Blenkinsop, J., 2000, Differential diagenesis und Kalkgehalt eines Profils im gebankten of sedimentary components and the unteren Malm Schwabens. Neues Jahrbuch für implication for strontium isotope analysis of Geologie und Paläontologie – Abhandlungen, carbonate rocks: Chemical Geology, v. 164, p. v. 98, p. 28-86. 331-343. Semper, M., 1917, Schichtung und Bankung: Rey, O., Simo, J.A., and Lorente, M.A., 2004, A Geologische Rundschau, v. 7, p. 53-56. record of long- and short-term environmental and climatic change during OAE3: La Luna Sprenger, A. and ten Kate, W.G., 1993, Orbital Formation, Late Cretaceous (Santonian–early forcing of calcilutite-marl cycles in southeast Campanian), Venezuela: Sedimentary Spain and an estimate for the duration of the Geology, v. 170, p. 85-105. Berriasian stage: Geological Society of America Bulletin, v. 105, p. 807-818. Ricken, W., 1986, Diagenetic Bedding: A Model for Limestone-Marl Alternations: Lecture Notes Stanley, M. and Hardie, L.A., 1998, Secular in Earth Sciences, v. 6, Berlin, Springer oscillations in the carbonate mineralogy of Verlag, 210 p. reef-building and sediment-producing Westphal et al. “ Limestone-marl alternations in epeiric sea settings ” 20

organisms driven by tectonically forced shifts Walther, M., 1982, A contribution to the origin of in seawater chemistry. Palaeogeography limestone-shale sequences, in Einsele, G. and Palaeoclimatology Palaeoecology, v. 144, p. Seilacher, A.: Cyclic and Event Stratification, 3-19 p. 113-120, Springer, Berlin. Stanley, S.M. and Hardie, L.A., 1999, Wepfer, E., 1926, Die Auslaugungsdiagenese, ihre Hypercalcification: Paleontology links plate Wirkung auf Gestein und Fossilinhalt: Neues tectonics and geochemistry to sedimentology: Jahrbuch für Mineralogie Beil.-Bd., v. GSA Today, v. 9(2), p. 1-7. LIV(B), p. 17-94. Strasser, A., Aurell, M., Bádenas, B., Meléndez, G. Westphal, H., 1998, Carbonate Platform Slopes - A and Tomás, S., 2005, From platform to basin Record of Changing Conditions: Lecture to swell: orbital control on sedimentary Notes in Earth Sciences, v. 75: Berlin, sequences in the Oxfordian, Spain: Terra Heidelberg, New York, Springer Verlag, 179 Nova, v. 17(5), p. 407-413. p. Sujkowski, Z.L., 1958, Diagenesis: Bulletin of the Westphal, H. and Munnecke, A., 1997, Mechanical American Association of Petroleum compaction versus early cementation in fine- Geologists, v. 42, p. 2692-2717. grained limestones: differentiation by the preservation of organic microfossils: Thiry, M., 2000, Palaeoclimatic interpretation of clay Sedimentary Geology, v. 112, p. 33-42. minerals in marine deposits: an outlook from the continental origin: Earth Science Reviews, Westphal, H., and Munnecke, A., 2003, Limestone- v. 49, p. 201-221. marl alternations - a warm-water phenomenon? Geology, v. 31, p. 263-266. Tomasovych, A., 2004, Effect of extrinsic factors on biofabric and brachiopod alteration in a Westphal, H., Head, M.J. and Munnecke, A., 2000, shallow intraplatform carbonate setting (Upper Differential diagenesis of rhythmic limestone Triassic, west Carpathians): Palaios, v. 19(4), alternations supported by palynological p. 349-371. evidence: Journal of Sedimentary Research, v. 70, p. 715-725. Visser, J.P., 1991, Clay mineral stratigraphy of Miocene to Recent marine sediments in the Westphal, H., Böhm, F. and Bornholdt, S., 2004a, central Mediterranean: Geologica Ultraiectina, Orbital frequencies in the sedimentary record: v. 75, p. 1-243. distorted by diagenesis?: Facies, v. 50, p. 3-11. Walker, L.J., Wilkinson, B.H. and Ivany, L.C. 2002, Westphal, H., Munnecke, A., Pross, J. and Herrle, J., Continental drift and Phanerozoic carbonate 2004b, Multiproxy approach to understanding accumulation in shallow-shelf and deep- the origin of Cretaceous pelagic limestone- marine settings: The Journal of Geology, v. marl alternations (DSDP Site 391, Blake- 110, p. 75-87. Bahama Basin): Sedimentology, v. 51, p. 109- 126.