Sedimentary Geology 223 (2010) 51–60

Contents lists available at ScienceDirect

Sedimentary Geology

j ourna l homepage: www.e lsev ie r.com/ locate /sedgeo

Diagenesis of silica-rich mound-bedded chalk, the Coniacian ArnagerLimestone, Denmark

H.B. Madsen a,⁎, L. Stemmerik b, F. Surlyk b

a Geological Survey of Denmark and Greenland, Øster Voldgade 10 1350-DK Copenhagen, Denmarkb Department of Geography and Geology, University of Copenhagen, Øster Voldgade 10 1350-DK Copenhagen, Denmark

⁎ Corresponding author.E-mail address: heinebuus@yahoo.com (H.B. Madsen

0037-0738/$ – see front matter © 2009 Elsevier B.V. Aldoi:10.1016/j.sedgeo.2009.10.002

a b s t r a c t

a r t i c l e i n f o

Article history:Received 31 March 2009Received in revised form 14 September 2009Accepted 11 October 2009

Keywords:DiagenesisOpal-CTClinoptiloliteConiacianChalkPorosity

The Coniacian Arnager Limestone Formation is exposed on the Danish island of Bornholm in the Baltic Sea. It iscomposed of mound-bedded siliceous chalk, and X-ray diffraction and scanning electron microscopy indicate acontent of 30–70% insoluble minerals, including authigenic opal-CT, quartz, clinoptilolite, feldspars, calcite,dolomite, and barite. Opal-CT and clinoptilolite are the most common and constitute 16–53% and 2–9%,respectively. The contentof insolubleminerals varies laterallybothwithin themounds and inplanarbeds, and theopal-CT content varies by up to 10% vertically. Themounds consist of twomicrofacies, spiculitic wackestone andbioturbated spiculitic wackestone, containing 10–22% and 7–12% moulds after spicules, respectively.Subsequent to deposition and shallow burial, dissolution of siliceous sponge spicules increased the silica activityof the pore water and initiated precipitation of opal-CT. The opal-CT formed at temperatures around 17°C, theprecipitation lowered the silica activity and the Si/Al ratio of the pore water, resulting in precipitation ofclinoptilolite, feldspar and smectite. Calcite formed synchronouslywith the latest clinoptilolite.Minor amounts ofquartz precipitated in pore water with low silica activity during maximum burial, probably to depths of 200–250 m. The dissolution of sponge spicules and decomposition of the sponge tissue also resulted in the release ofBa2+, Sr2+,Mg2+, Ca2+andCO3

2−, facilitatingprecipitationof barite anddolomite. Precipitationof especially opal-CT reduced the porosity to an average of 40% and cemented the limestone. The study highlights the diageneticpathways of bio-siliceous chalk and the effects on preservation of porosity and permeability.

).

l rights reserved.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

The diagenesis and depositional setting of siliceous chalk differfrom that of the classical pelagic chalk of NW Europe. Precipitation ofauthigenic silicates is common and often reduces the porosity (Malivaand Dickson, 1992; Chaika and Dvorkin, 2000). However, formation ofauthigenic silicates may also result in lithification at shallow burialdepths and hence preserve porosity during deeper burial (Aase et al.,2001), making it difficult to predict the effects of silica diagenesis onreservoir properties.

The diagenesis of sediments rich in biogenic silica includesprecipitation of opal-CT, zeolites, feldspars and smectite, but opal-CTand zeolites are mostly preserved in shallow buried sediments, andrare to absent in more deeply buried successions (Berger and von Rad,1972; Fabricius and Borre, 2007; Gingele and Schulz, 1993; Huggett etal., 2005; Karpoff et al., 2007; Kastner, 1981; williams et al., 1985). Thesilica activity and the Si/Al ratio of the pore water are the maincontrolling factors on diagenesis in silica-rich sediments (Berger andvon Rad, 1972; Abercrombie et al., 1994; Wilkin and Barnes, 1998).The early silica diagenesis is characterized byminerals with high silica

activity such as opal-CT, clinoptilolite and smectite, whereas precip-itation of quartz dominates at lower silica activity and deeper burial(Abercrombie et al., 1994).

The mound-bedded siliceous chalk of the Coniacian ArnagerLimestone Formation on Bornholm, Denmark has only been buried toshallow depths, probably 200–250 m and thus represents an unusualexample of pre-Paleogene chalk where the early diagenetic paragenesisis preserved. The aim of this study is to identify sedimentarymicrofaciesof the formation and to establish the diagenetic processes and theireffects on porosity and permeability. The investigations are based onmapping and sampling of amounded section in a coastal cliff at Arnageron the south coast of Bornholm. The data set allows estimates of thelateral and vertical variabilities in porosity and permeability on themetre to tens of metre scale, horizontally and decimetre to metre scale,vertically.

2. Geological setting

During the Mesozoic the stable Baltic Shield was separated fromthe subsiding Danish Basin by the block-faulted NW–SE trendingTornquist Zone (Fig. 1) (Liboriussen et al., 1987). The island ofBornholm is a basement horst situated in the zone. During the LateCretaceous, much of the horst became transgressed. The initial Early to

Fig. 1. Structural map of Denmark and southern Sweden and geological map of Bornholm. Modified after Liboriussen et al. (1987) and Gravesen et al. (1982).

52 H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

Middle Cenomanian transgression led to deposition of the ArnagerGreensand Formation in a shallow shelf environment (Christensen,1985; Packer andHart, 1994). The basal part of the overlying ConiacianArnager Limestone Formation is a thin conglomerate composed ofseveral generations of phosphatised and glauconitised pebblesreflecting alternating phases of deposition, non-deposition, erosionandmineralisation (Ravn, 1918). Deposition of the Arnager Limestonetook place in an outer shelf setting (Packer and Hart, 1994).

The Arnager Limestone, 12–20 m thick, is characterized by moundedbedding possibly caused by growth of sponges thickets and baffling ofsediment (Noe-Nygaard and Surlyk, 1985). Siliceous sponges are verycommon and comprise mainly Hexactinellida and rarely Demospongiae(Brückner and Janussen, 2005; Brückner, 2006). The remainder of themacrofauna is dominated by inoceramid bivalves (Tröger and Christen-sen, 1991), whereas other macrofaunal elements, including bivalves,brachiopods, belemnites (Christensen and Schulz, 1997) and ammonites(Kennedy and Christensen, 1991) are very rare. The limestone iscomposedof 34–76% insolubleminerals, primarily opal-CT,which formedduring diagenesis (Noe-Nygaard and Surlyk, 1985; Mogensen, 1994).Flint nodules incorporating siliceous sponges are rare (Ravn, 1918).

Syn-depositional volcanism occurred in southern Sweden, northernGermany and Poland (Norin, 1933, 1934; Dorn and Bräutigam, 1959;Printzlau and Larsen, 1972; Solokowski, 1976; Seibert and Vortisch,1979). On the Arnager Block the Arnager Limestone Formation isoverlain by about 150 m of Santonian Bavnodde Greensand (Ravn,1921; Christensen, 1985, Jensen and Hamann, 1988) and possiblyadditional younger Cretaceous sediments, indicating minimum burialdepth of the Arnager Limestone of 200–250 m. Increased tectonicactivity during Late Cretaceous and Paleogene times caused inversionand uplift of the Arnager Block and erosion of the overlying sediments(Graversen, 2004).

3. Materials and methods

The Arnager Limestone is exposed in a 200 m long coastal cliffprofile west of Arnager harbour (Fig. 1). Mound structures weremapped and 56 samples were taken along the profile. Point countingwas carried out on 20 thin sections to quantify the content. Theinsoluble residue analysiswas based on samples crushed in a porcelainmortar until it passed a 2 mm sieve and calcite was removed using a

buffered acetic acid at pH 4.5 to avoid dissolution of non-calciteminerals. X-ray diffraction (XRD) of the insoluble residue was carriedout on randomly oriented specimens using a Philips 1050 goniometerwith Cu-Κα radiation. Quantification of the minerals was made by theRietveld method using the software Topas 3.

SEM analyses were carried out on small samples of limestone whichwere coatedwith gold in vacuumat 25 kV and 20 mA for 2–3min using aSEM Coating Unit E5000. Photography and qualitative chemical analysiswere made on a PHILIPS XL 40 SEM equipped with a ThermoNoranenergy dispersive X-ray detection system (EDX). Backscatter images ofcarbon-coated polished thin sections were taken at 50× and 1500×magnifications at five places, 6 mm apart and perpendicular to bedding.Quantification of large pores (φlarge, >20 μm) and small pores (φsmall,<5 μm) was based on the backscatter images using the free softwareUTHSCSA Image tool version 3.0, following the method of Røgen et al.(2001). Average φlarge and φsmall values from the image analyses werecorrected by multiplying their relative contribution with the measuredporosity.

Porosity and gas permeability were measured in the core labo-ratory at theGeological Survey ofDenmark andGreenland (GEUS). Theporosity was determined by using a Helium porosimeter on clean anddried plugs of known dimensions oriented parallel to bedding and bysubtracting grain volume from themeasured bulk volume. Boyle's Lawwas employed to determine the grain volume. The gas permeabilitywas measured on a steady state instrument with confining pressureof 400 psi and by flowing nitrogen gas through plugs at differentialpressures between 0 and 1 bar.

4. Sedimentology

The base of the Arnager Limestone Formation is exposed in theeastern part of the coastal cliff west of Arnager harbour (Fig. 3). Theformation dips 5° to the W, thus exposing stratigraphically youngerlimestone going from the east to west along the cliff.

Noe-Nygaard and Surlyk (1985) demonstrated that the stronglyjointed limestone in wet condition reveals markedly moundedbedding (Fig. 2). In the basal part of the formation, the mounds orhummocks are 20–30 m long and around 1 m high (Fig. 3). They areasymmetric with the eastern slope generally dipping 10–12° and thewestern slope dipping 6–8°. The crests of successive beds in some

Fig. 2. a) Mound structure at 75 m on the profile shown in Fig. 3. b) Interpreted bed tracing defining a one metre high and several metres long mound downlapped and onlapped byyounger mound beds.

53H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

mounds appear to be slightly displaced toward the west. The gentlywestward dipping beds are usually the thickest. Minor erosion isrecognized on the eastern side of some mounds, e.g. at 35 m (Fig. 3).The oldest mounds downlap onto the basal conglomerate, withyounger mounds downlapping or onlapping the older mounds. Inthe upper part of the formation internal bedding is difficult to depictbecause of poor exposure but bedding seems to be more planar withlittle lateral thickness variation. However, a mound structure at least15 m long and 0.5 m high is recognized at 170 m (Fig. 3).

The formation consists of fine-grained siliceous chalk. It is lightgreyish with dark grey areas (Fig. 4). The transition between the lightand dark grey chalk is commonly abrupt and irregular and crosscutsboth Thalassinoides and lenses of siliceous sponge spicules, and thedark grey areas are thus a post-depositional phenomenon. Imprints ofsiliceous sponges are very common on bedding planes, occasionallywith recognizable morphology. Also moulds after sponge spicules androotlets are common. The sediment contains 30–70% insolubleminerals dominated by opal-CT, quartz, clinoptilolite, orthoclase,microcline, anorthite and mica (Fig. 5, Table 1). Subordinate dolomite,smectite, pyrite and barite have been detected in the SEM. Theinsoluble fraction includes 16–53% opal-CT, 2–7% quartz, 2–9%clinoptilolite, and 4–10% mica of the bulk limestone (Table 1). Thefeldspars are anorthite, microcline and orthoclase, constituting 0–3%.The content of insoluble minerals varies laterally both within themounds and in the planar beds, and the opal-CT content varies by up to10% vertically. Two different microfacies, spiculitic wackestone andbioturbated spiculitic wackestone are distinguished, in addition to thebasal conglomerate (Fig. 6).

4.1. Spiculitic wackestone

In this microfacies moulds after spicules constitute 10–22% by area(Table 2). Themoulds occur on bedding planes in amixture of μm-sizedopal-CT and carbonate, forming a spiculitic wackestone (Fig. 6a–b and

d–e). This microfacies also contains macrofossil fragments, glauconite,detrital mica, quartz and feldspar grains, occasionally concentrated insubhorizontal lenses (Table 2). Preferred orientation of skeletal grainsand moulds after sponge spicules is roughly parallel to the lenses(Fig. 6a–b). The skeletal grains comprise foraminifers with subordinatefragments of brachiopods, inoceramid bivalves andbryozoans (Fig. 6a–band d–e) (Noe-Nygaard and Surlyk, 1985). Glauconite and detritalgrains of quartz and feldspar are angular to sub-angular and constitute0.4–2.0% and 1.2–4.8%, respectively.

The matrix of the spiculitic wackestone consists primarily of fine-grained opal-CT and carbonate (Fig. 6d–e). The carbonate comprisesmainly coccoliths and crystallites from coccoliths, 2–8 μm in size, andshell detritus up to 30 μm long. In some samples the carbonatefraction constitutes most of the matrix, whereas opal-CT dominates inothers. The spiculitic wackestone has porosities of 36–42% with anaverage of 39%, and permeabilities of 4–20 mD with an average of12 mD (Table 1).

4.2. Bioturbated spiculitic wackestone

This microfacies contains well defined trace fossils (Thalassinoides)and in addition differs from the spiculitic wackestone by the randomorientation of fossil fragments andmoulds after sponge spicules, and themore random distribution of glauconite and detrital grains, probablycaused by bioturbation (Fig. 6c and f). The moulds after spiculesconstitute 7–12% by area, which is lower than in the spiculiticwackestone (Table 2). The fossil content is similar to that of thespiculitic wackestone. The content of glauconite, detrital quartz andfeldspars is slightly lower than in the spiculiticwackestone, constituting0.3–0.9% and 1.0–3.3%, respectively. The matrix in the bioturbatedspiculitic wackestone is similar to the spiculitic wackestone. However,the abundance of pores appears to be lower (Fig. 6f). This facies hasporosities in the range of 32–41% with an average of 38% andpermeabilities of 5–13 mD with an average of 10 mD.

Fig. 3. Tracing of mounded bedding in the 200 m long coastal cliff west of Arnager.

54 H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

5. Authigenic minerals

The presence of opal-CT, siliceous heulandite, feldspars and mixedlayered smectite–illite was noted by Noe-Nygaard and Surlyk (1985)and Mogensen (1994). The authigenic minerals identified in thisstudy include opal-CT, quartz, clinoptilolite, orthoclase, microcline,anorthite, smectite, barite, calcite and dolomite. The structure ofclinoptilolite is very similar to siliceous heulandite and can easily bemistaken for this mineral.

Opal-CT occurs as pore-filling cement both in thematrix, inmouldsafter spicules, and in foraminifers and calcispheres (Noe-Nygaard and

Surlyk, 1985). It is also commonon the surface of detrital quartz grains.Opal-CT occurs as lepispheres, microspherical aggregates that consistof bladed opal-CT crystals composed of nano-sized crystallites(Fig. 7a). The lepispheres occur in different stages where the smallestand least complex commonly sit on larger and more complexlepispheres, up to 8 μm in diameter. The blades appear to be randomlyoriented, but a 60° angle between blades is common, most obvious inthe small, embryonic lepispheres with few blades (Fig. 7a). Thelepispheres are either spherical, which is the most commonmorphol-ogy, oval, or form sheets of opal-CT blades on the walls of pores. Opal-CT lines the pores andwas the first diagenetic phase to precipitate. The

Fig. 4. Polished and oil wetted slab of Arnager Limestone. Burrows (Bur) and lenses ofmoulds after spicules (Ms) are common in the limestone. Light to dark grey shadingswith occasionally irregular boundaries (Ib) are not primary structures but of diageneticorigin.

55H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

Opal-CT includes silica with both tridymitic and cristobalitic structure,in the XRD spectres, and tridymite dominates over cristobalite in theopal-CT structure. The tridymite is characterized by peaks at 4.32,4.09–4.11, 3.83 and 2.50 Å,whereas the cristobalite is characterized bypeaks at 4.03 and 2.48 Åwith less intensity than tridymite (Fig. 5). TheEDX analyses indicate that Si, O and traces of Al are incorporated in theopal-CT structure with a Si/Al ratio of 66.8.

Clinoptilolite is common and forms almost perfect crystals withsmooth surfaces, up to 10 μm long and 3 μmwide (Fig. 7b). The crystalshave a pseudotetragonal symmetry but the top of the prisms indicates amonoclinic habit. Growth-stops on parts of the (100), (010) and (001)

Fig. 5. XRD spectrum of sample Arn 24 containing opal-CT, clinoptilolite (Cl), quartz (Q),orthoclase (Or) and glauconite (Gl). Note that the tridymite (T) structure dominatesover cristobalite (Cr) in opal-CT.

crystal surfaces are common, giving some of the crystals a staircase-likeappearance. However, no complete interruptions of growth seem tohaveoccurredduringprecipitation. The clinoptilolite crystals commonlysit on opal-CT, occasionally with a next generation of opal-CT lepi-spheres on top. Rarely are they observed on calcite crystals that grew onopal-CT (Fig. 7c). Late opal-CT has not been observed on eitherclinoptilolite or calcite crystals. Qualitative EDX analyses indicate thatthe clinoptilolite contains Na, K, Ca, Al, Si and O with an average Si/Alratio of 4.9. XRD analyses show characteristic peaks at 8.92, 7.89, 6.76,5.10, 3.97, 297 and 2.79 Å,where the latter occasionally consists of threepeaks (Fig. 5). The intensities and distribution of peaks are similar toclinoptilolite 07-1870 in the Powder Diffraction File (PDF), with thechemical formula (Na,K,Ca)6Al6Si30O72·18H2O.

Authigenic anorthite, orthoclase and microcline most commonlyoccur in the matrix but are also observed in pores (Fig. 7d and e). Inthematrix the feldspars have a tabular habit and are up to 20 μmwideand 30 μm long, commonly showing perfect cleavage. Impressionsafter fossil fragments are occasionally observed on the surfaces offeldspar grains, indicating that they grew in the matrix (Fig. 7d). Thepore-filling feldspars form perfect euhedral crystals, up to 6 μm wideand 10 μm long. They are composed of K, Al, Si and O according to theEDX analyses, and are either orthoclase or microcline. The Si/Al ratioof the pore-filling alkali feldspar is 4.8 compared to 2.6 for the matrixfeldspar. The pore-filling alkali feldspar sits on opal-CT with a latergeneration of opal-CT on the surfaces (Fig. 7e). The orthoclase hascharacteristic peaks at 3.25, 3.32 and 3.77 Å and the microcline ischaracterized by three peaks around 3.25 Å and peaks at 3.29, 3.46and 3.70 Å (Fig. 5). Not all peaks are distinguished because of overlapwith peaks from other minerals.

Trace amounts of clay, probably smectite are indicated by a broadpeak around 11 Å (Fig. 5), and delicate and well preserved clayflakes have been observed between opal-CT lepispheres (Fig. 7f).EDX analyses indicate that they are composed of Fe, Mg, Ca, K, Al, Siand O.

Authigenic quartz occurs in small amounts in the form of anhedralmasses on opal-CT lepispheres and detrital quartz grains (Fig. 7g andh). Granular nano-sized quartz crystals are occasionally present onthe lepispheres, blurring and partly erasing the delicate structure ofthe opal-CT blades. Anhedral to subhedral masses of authigenic quartzare associated with opal-CT lepispheres on detrital quartz grains(Fig. 7h), and quartz also replaces the delicate blades in some opal-CTlepispheres. No authigenic minerals are observed to post-date theauthigenic quartz.

Few dolomite crystals, up to 30 μm in size, with perfect rhombohe-dral shape, are observed in the SEM (Fig. 7i). The surface of the crystalscommonly has imprints from coccoliths and coccolith fragments areoccasionally embedded in the crystals. The dolomite occurs only in thematrix. Barite is found in a few samples as up to 15 μm long wedgeshape blades in voids together with opal-CT. The barite sits on theopal-CT and may fill the central parts of pores (Fig. 6b). EDX analysesindicate that the barite contains minor Sr.

The relative timing of precipitation of the authigenic minerals canbe established based on the SEM analyses. Opal-CT was the firstdiagenetic mineral to form followed by clinoptilolite and feldspars.Calcite started to precipitate at the end of clinoptilolite precipitation.The baritefills the centre of voids and sits on opal-CT and clinoptilolite,indicating a late formation. Authigenic quartz postdates opal-CT andpossibly all other minerals. The timing of dolomite and smectite is notwell established, but the smectite probably precipitated synchronous-ly with the opal-CT.

6. Discussion

The boundaries between light and dark grey chalk crosscut tracefossils, lenses of detrital minerals and fossil fragments, indicating thatit is a result of diagenesis. The SEM images indicate that silica,

Table 1Mineralogy, permeability and porosity.

Sample Microfacies Mineral (vol. %) Permeability Porosity

Quartz Opal-CT Clinoptilolite Orthoclase Albite Microcline Glauconite Apatite mD %

Arn 1 6.8 29.4 7.4 0.8 0.8 – 5.4 – 13.1 43.7Arn 2 7.0 25.1 8.4 – 0.3 2.0 7.0 – 13.2 43.5Arn 3 3.3 37.9 5.0 – – 1.3 5.4 – 6.6 41.1Arn 4 SW 4.0 26.1 5.2 – – 1.6 4.5 – 13.4 37.7Arn 5 BSW 2.0 37.3 4.2 0.2 – 1.3 6.2 – 9.0 37.1Arn 6 1.9 25.9 4.5 0.1 0.4 – 5.5 – 19.5 38.6Arn 7 5.0 52.8 4.2 – 0.6 1.7 7.9 – 6.4 37.7Arn 8 2.2 33.8 6.1 – – 1.3 9.9 – 9.4 42.7Arn 9 4.2 32.7 7.5 – 0.6 2.5 6.0 – 17.3 39.6Arn 10 2.6 35.8 4.1 1.3 0.8 – 7.0 – 5.6 35.4Arn 11 3.0 16.1 6.2 – – 2.4 5.6 – 10.6 41.6Arn 12 2.8 43.2 3.3 – 0.8 1.2 7.9 – 5.6 36.3Arn 13 3.2 19.3 6.8 0.2 – 1.1 6.6 – 9.8 41.0Arn 14 2.5 15.9 5.3 1.3 – 1.8 9.8 – 8.3 40.7Arn 15 3.5 13.2 6.2 – 0.7 2.0 4.4 – 2.0 38.3Arn 16 2.8 31.5 4.2 – 0.7 0.9 7.3 – 5.8 40.7Arn 17 3.0 45.4 3.8 – 0.5 0.9 5.6 – 10.8 41.0Arn 18 2.9 27.1 8.5 0.9 – 1.7 8.9 – 7.2 45.3Arn 20 SW 4.1 29.4 5.7 0.3 – 2.1 6.3 – 10.9 41.5Arn 21 SW 2.2 34.2 3.6 0.5 0.8 1.0 6.7 – 13.1 36.9Arn 22 BSW 2.7 43.5 3.6 – – 1.7 7.2 – 13.1 36.8Arn 23 SW 2.4 33.0 5.0 – 0.5 1.2 5.3 – 4.3 41.9Arn 24 BSW 2.3 43.5 2.9 0.5 – 1.3 6.7 – 11.7 41.1Arn 25 BSW 3.9 29.8 6.7 – 0.7 1.9 6.5 – 11.2 41.3Arn 26 BSW 4.2 44.7 3.8 – – 2.0 6.4 – 9.5 38.3Arn 27 SW 2.4 48.2 2.5 – 0.7 1.4 7.1 – 8.2 36.2Arn 28 4.7 22.8 7.0 – – 1.9 5.1 – 4.8 40.0Arn 29 2.5 39.8 2.6 – 0.3 1.2 6.8 – 5.5 34.0Arn 30 2.8 31.2 3.8 – 0.5 1.2 5.1 – 6.4 36.4Arn 31 3.3 27.4 5.7 1.0 0.7 – 6.9 – 12.2 42.6Arn 32 2.9 39.5 5.5 – – 1.1 7.2 – 15.4 43.6Arn 33 3.3 37.0 8.7 0.3 – 1.8 7.6 – 16.8 44.4Arn 34 2.1 40.5 3.4 – – 0.7 7.9 – 9.4 37.3Arn 35 3.3 52.6 3.5 – 1.0 1.2 9.7 – 18.9 37.7Arn 36 3.2 39.8 5.9 0.2 – – 6.6 – 17.5 40.4Arn 37 SW 2.2 27.0 5.4 – 0.4 1.5 7.4 – 13.7 37.9Arn 38 SW 2.9 24.2 6.1 0.1 0.6 1.5 6.7 – 20.0 39.8Arn 39 SW 3.3 20.8 6.9 0.1 0.4 2.4 6.3 – 12.8 41.9Arn 40 BSW 3.0 28.0 5.6 0.2 0.5 0.9 6.4 – 4.5 32.1Arn 41 SW 2.5 28.9 4.7 1.5 0.4 – 6.2 – 14.5 38.6Arn 42 SW 3.5 33.1 5.3 – – 1.2 7.3 – 9.4 37.0Arn 43 SW 6.4 16.2 7.9 0.1 – 2.8 5.3 – 7.3 41.7Arn 44 3.2 22.5 5.3 – 0.6 – 5.0 – 9.0 39.6Arn 45 2.7 18.2 5.8 – 0.2 1.4 5.1 – 12.3 39.5Arn 46 3.2 29.4 4.0 – 0.4 1.5 4.7 – 7.3 31.9Arn 47 3.9 16.0 7.4 – – 2.1 5.5 – 5.8 39.1Arn 48 3.5 19.7 6.2 – – 1.0 3.9 – 6.4 38.9Arn 49 3.5 18.7 7.3 0.1 – 1.3 7.2 – 7.8 41.2Arn 5OA BC 4.5 31.9 7.5 1.0 – – 8.4 – 1.7 38.2Arn 50B BC 14.7 44.4 2.7 – – 1.2 8.3 7.4 1.3 29.4Arn 51 2.6 33.6 3.5 – – 0.8 5.4 – 15.9 38.7Arn 52 4.7 40.4 5.4 0.3 – – 6.2 – 18.5 40.1Arn 53 2.4 35.1 4.6 – 0.4 0.9 6.4 – 14.6 46.5Arn 54 3.2 47.4 5.4 – 0.6 1.3 6.9 – 21.6 46.1Arn 55 2.9 34.7 5.8 – 0.8 1.1 5.8 – 11.8 45.3Arn 56 3.4 35.6 6.7 – – – 7.0 – 9.6 44.3

SW: spiculitic wackestone. BSW: bioturbated spiculitic wackestone. BC: phosphatic basal conglomerate.

56 H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

sulphate and carbonate diagenesis occur in both the distinguishedmicrofacies.

Biogenic skeletal opal-A becomes unstable when the organisms dieand readily dissolves when exposed to sea water under saturated withrespect to silica (Siever, 1962;Mizutani, 1970;Hurd, 1973;Kastner et al.,1977,Williams et al., 1985, Treguér et al., 1995;Maldonado et al., 2005).The biogenic opal dissolves either in the water column or in thesediment. In the Arnager Limestone the siliceous sponge spicules haveleft moulds indicating that dissolution took place in the sediment. Thelower abundance of moulds after spicules in the bioturbated spiculiticwackestone compared to the spiculitic wackestone indicates thatbioturbation enhanced dissolution prior to silicification.

The dissolution of sponge spicules resulted in supersaturation of thepore water with respect to opal-A, opal-CT and quartz and in theory allthreephaseswereable toprecipitate (Eqs. 1 and2) (Williamset al., 1985).However, opal-CT preferentially precipitated following initial dissolutionof sponge spicules. The concentration of dissolved silicic acid remainedstable as long as opal-A was still present. As opal-A disappeared, theconcentration of silicic acid in the porewater started to drop, and the lessordered opal-CT started to dissolve (Fig. 8).

The temperature for the transformation from opal-A to opal-CT canbe estimated on the basis of the d(101) value of opal-CT (Murata andNakata, 1974; Murata et al., 1977; Kano, 1983). The observed d(101)values, 4.09–4.11, indicate that the temperature during precipitation of

Fig. 6. Backscatter images of polished thin sections. Black areas are voids, whereas light grey is carbonate and dark grey is silicate. a) Spiculitic wackestone (Arn 38) showing highabundance of moulds after spicules (S). Some of the moulds are partly filled by opal-CT (O) as are several foraminifers (F) and calcispheres. Angular detrital quartz grains (D) arecommon. Note the vaguely sub-parallel orientation of moulds after spicules and fossil fragments. b) Spiculitic wackestone (Arn 23) with vaguely sub-parallel orientation of mouldsafter spicules but with fewer moulds than in a). Some of the foraminifers and moulds after spicules are filled with opal-CT, reducing the large pore spaces. Note the barite crystal (B)in the right part of the image. c) Bioturbated spiculitic wackestone (Arn 40) having a low content of large pores some of which have been almost totally filled by opal-CT. d) Arn 38showing a μm-sized carbonate grain-supported matrix (1500×). Micronmetre-sized pores are common but most are filled with opal-CT. Note the intact coccolith (Co). e) The matrixof Arn 23 shows high porosity compared to d) and carbonate fragments appear primarily to be embedded in opal-CT (1500×). Note the tabular clinoptilolite crystals (Cl). f) Markedlylower abundance of pores is registered in Arn 40 where many pores are filled with opal-CT. The carbonate fragments are primarily embedded in opal-CT (1500×).

Table 2Point counting.

Sample Type Spicules(vol.%)

Fossil fragments(vol.%)

Detrital grains(vol.%)

Glauconite(vol.%)

Phosphorite(vol.%)

Arn 4 SW 10.2 6.1 2.4 0.5 0.0Arn 20 SW 15.3 3.4 2.4 0.9 0.0Arn 21 SW 19.7 0.2 3.8 2.0 0.0Arn 23 SW 16.2 2.1 4.1 1.5 0.0Arn 27 SW 21.8 0.2 3.2 0.7 0.0Arn 37 SW 11.0 3.4 1.2 0.9 0.0Arn 38 SW 13.9 3.8 3.3 0.8 0.0Arn 39 SW 11.2 0.7 2.4 0.4 0.0Arn 41 SW 13.7 1.7 2.4 1.5 0.0Arn 42 SW 13.4 2.8 2.2 1.1 0.0Arn 43 SW 14.5 1.7 4.8 0.4 0.0Average SW 14.6 2.4 2.9 1.0 0.0Arn 5 BSW 7.5 2.2 1.0 0.3 0.0Arn 22 BSW 9.3 3.3 3.3 0.4 0.0Arn 24 BSW 7.4 0.4 3.0 0.6 0.0Arn 25 BSW 7.2 3.0 1.3 0.9 0.0Arn 26 BSW 8.1 1.7 2.4 0.4 0.0Arn 40 BSW 11.8 1.5 2.3 0.4 0.0Average BSW 8.5 2.0 2.2 0.5 0.0Arn 50B

BC 3.4 0.9 22.8 8.6 10.9

Arn 50A

BC 10.0 1.5 10.3 3.4 0.9

57H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

opal-CTwas in the range of 10–22°Cwith amean of 17°C. The presenceof small amounts of aluminium in the opal-CT structure indicatespresence of Al3+ in the solution which would have lowered thesolubility of SiO2 and catalyzed precipitation of opal-CT (Iler, 1973).

SiO2 þ 2H2O↔H4SiO4 ð1Þ

opal�A→opal�CT→quartz: ð2Þ

The precipitation of clinoptilolite, feldspar and smectite took placedeeper in the sediment than the initial opal-CT (Fig. 8), probably inthe upper 1–20 m, as reported from recent pelagic sediments(Stonecipher, 1976, 1978; Kastner and Stonecipher, 1978). Thetemperature estimates of 10–22°C is probably in the range of thatof the bottom sea waters, indicating precipitation of opal-CT near theseafloor. The coexistence of opal-CT and clinoptilolite indicatesprecipitation at temperatures below 50°C (Wilkin and Barnes,1998), and the clinoptilolite probably formed at only slightly highertemperatures than the 10–22°C indicated for the initial opal-CT.

The co-occurrence of clinoptilolite and opal-CT in moulds afterspicules shows that the formation of the two components was closelyrelated. It is likely that precipitation of opal-CT lowered the Si/Al ratioto a level that enabled precipitation of clinoptilolite and feldspars assuggested by Berger and von Rad (1972), Williams et al. (1985) andWilkin and Barnes (1998) (Fig. 8). The changes in pore waterchemistry also facilitated precipitation of minor amounts of smectite,

alkali feldspars and anorthite, depending on the relative concentra-tions of the ions in the pore water (Eq. 3). Clinoptilolite formation and

Fig. 7. SEM images of diagenetic minerals in the Arnager Limestone. a) Image showing several stages of opal-CT (O) formation. Note the 60° angle between the blades in thelepispheres. b) Pseudotetragonal clinoptilolite crystals (Cl) with monoclinic habit on top of crystals. c) Clinoptilolite on calcite crystal (C). Note that opal-CT does not occur on thecrystal surfaces of these minerals. d) Angular orthoclase (Or) in insoluble residue. Note the possible mark from a fossil on the crystal surface (arrow). e) Void-filling orthoclasetogether with opal-CT and clinoptilolite. f) Smectite (Sm) and opal-CT on a detrital quartz grain. g) Opal-CT lepisphere covered by granular quartz (Q). h) Transformation of opal-CTlepispheres, sitting on detrital quartz grain (DQ), to quartz. i) Dolomite crystal (D) with perfect rhombohedral shape with embedded carbonate fragments.

58 H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

stability are only favoured when silica activity exceeds the equilib-rium quartz solubility, so silica activity was still high in the pore waterduring this stage (Bowers and Burns, 1990). Feldspars with lower Si/Al ratios preferentially precipitated in the matrix. The higher Si/Alratio of the alkali feldspar in the moulds probably reflects excess silicafrom dissolution of the spicules.

47Naþ þ 46K

þ þ 41Ca2þ þ 3Mg

2þ þ 135AlðOHÞ−4 þ 648SiðOHÞ4→20ðNa;K;CaÞ6Al6Si30O72 � 18H2O

ðclinoptiloliteÞþ 6NaAlSi3O8

ðAlbiteÞþ 6KAlSi3O8

ðOrthoclaseÞþ

3ðNa;CaÞ1=3ðAl;MgÞ2Si4O10ðOHÞ2 � nH2OðSmectiteÞ:

ð3Þ

The source for the Al3+ to form clinoptilolite, feldspar and smectiteoriginated either from volcanicmaterial (cf. Weaver, 1968; Berger andvon Rad, 1972) or fromdetrital smectite (cf. Berger and von Rad, 1972;Kastner and Stonecipher, 1978; Karpoff et al., 2007), whereas the silicawas sourced from biogenic siliceous skeletons. In pelagic sedimentswithout volcanic material, the presence of clinoptilolite has beenlinked to high biogenic opal-A content and alteration of detritalsilicates (Hay, 1977; Karpoff et al., 2007). No volcanic glass shards areobserved in the Arnager Limestone but detrital minerals and mouldsafter siliceous sponge spicules are common, suggesting that alterationof detrital smectite coupled with high biogenic productivity producedthe clinoptilolite. However, syn-depositional volcanic activity oc-curred in southern Sweden (Norin, 1933, 1934; Printzlau and Larsen,1972), northern Germany (Dorn and Bräutigam, 1959; Seibert and

Vortisch, 1979) and Poland (Solokowski, 1976). This indicates thatvolcanic fallout probably was incorporated in the sediment and thuscould have contributedwith Al3+ to formauthigenic smectite, feldsparand clinoptilolite (Mogensen, 1994). The ions Na+, K+ and Ca2+ werepartly sourced from the detrital and volcanic minerals but sea wateralso contributed with the Na+ and Ca2+.

The silica activity in the pore water of the Arnager Limestone musthave remained high for a long time to prevent dissolution of clinoptiloliteand opal-CT. The scattered occurrences of granular quartz on opal-CTlepispheres indicate lower silica activity during later diagenesis (Fig. 8).The transformation of opal-CT to quartz is interpreted to be controlled bythe same dissolution–precipitation process which initially formed theopal-CT (Eq. 2). Quartz formed on the expense of opal-CT as the silicaconcentration of the pore water was lowered below saturation withrespect to opal-CT. The localized occurrence of authigenic quartz and thecommon occurrence of clinoptilolite indicate that maximum tempera-tures never were very high and that the formation never was deeplyburied. However, the formation was buried to about 200–250 m depth,and the transformation to quartz was probably aided by higher pressureand temperature during maximum burial. The formation did not remainat maximum burial depth and temperature for long because of LateCretaceous–Paleogene inversion of the Arnager Block (Vejbæk, 1985;Graversen, 2004). The shallowburial explains thepresenceof opal-CTandclinoptilolite and the rarity of flint in the Arnager Limestone compared toother Upper Cretaceous chalks in NW Europe.

The presence of barite reflects the high content of biogenic opal-Aonce present in the sediment. A high Ba content is commonly linked to

Table 3Quantification of pores.

Sample Microfacies φlarge φsmall σ

21 SW 19.8 17.1 4.023 SW 11.5 30.4 0.826 BSW 11.3 27.0 2.938 SW 19.5 20.3 2.639 SW 14.4 27.5 2.840 BSW 14.6 17.5 2.5

Fig. 8. Precipitation of authigenic minerals, silica activity and Si/Al ration in pore wateras a function of depth.

59H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

the presence of biogenic opal and is considered to be a siliceousproductivity proxy (Kasten et al., 2001; Eagle et al., 2003; Karpoff et al.,2007). Degradation and dissolution of the siliceous sponge skeletonsreleased Ba2+ and Sr2+ into the pore water, whereas the sea watercontributed with the SO4

2− needed to form barite. The concentrations ofBa2+andSr2+ in theporewater are interpreted tohave increased locallyduring dissolution of the siliceous sponge skeletons, causing supersat-uration with respect to barite. Recrystallization probably occurredduringburial, resulting in larger barite crystals in the centreof pores. TheSO4

2−was also consumed by sulphate reducing bacteria forming FeS2 orH2S. Iron oxide stainings occur inmanymoulds after spicules, indicatingthat minor pyrite was once present but probably dissolved during lateroxidation. This indicates that Femost likelywas limited and thatmost ofthe sulphur was oxidized to form H2S.

The timing of the calcite cementation is well established; it formedafter opal-CT and simultaneously with and after clinoptilolite. Thisindicates that calcite began to precipitatewhen the silica activity startedto diminish (Fig. 8). Up to 74 sponges per m2 has been counted onbedding planes indicating that organicmatterwas abundant at the timeof deposition (Noe-Nygaard and Surlyk, 1985).Most organicmatterwasoxidized and recycled at the seafloor but substantial amounts of organicmaterial probably reached the methanogenic zone. Thus the precipita-tion of calcite probably occurred in the reduced sediment where CO3

2−

concentration was higher because of methane oxidation (cf. Canfieldand Raiswell, 1991).

The dolomite occurs in thematrix and the timing of its formation isnot well established. Local dolomitization is commonly observed inbio-siliceous sediments and has been described to take place before,during, or after silicification (Jacka, 1974; Dapples, 1979; Bernoulli andGunzenhauser, 2001). It is likely that the dolomite in the limestoneformed during microbial decomposition of both the organic tissue in

sponges in the oxic sediment and in the methanogenic zone formingreduced environments. This occurred during but also after precipita-tion of the silicates. Themicrobial decomposition resulted in increasedpH, dissolution of carbonates, release of Ca2+, Mg2+, CO3

2− andlowering of SO4

2− in the pore water, enabling the precipitation ofdolomite (Vasconcelos et al., 1995; Vasconcelos and McKenzie, 1997;Wright, 2000; Wright and Wacey, 2005).

The diagenesis clearly affected the porosity and permeability.Analysis of back scatter images of thin sections at 50× and 1500×magnifications was used to quantify large pores (φlarge), small pores(φsmall), and the amount of cement and possibly to determine therelationships between porosity and permeability (Fig. 6, Table 3). Ingeneral high porosity and permeability occur in the spiculiticwackestone rich in moulds after spicules, whereas samples withhigh amounts of opal-CT cement in thematrix have low porosity. Highφlarge is primarily due to the presence of uncemented moulds afterspicules as indicated by the point counting data (Table 2). The φsmall

varies from 17 to 30% (Fig. 6d–f, Table 3) and mainly reflects variablecementation of the matrix by opal-CT and clinoptilolite.

7. Conclusions

The silica, sulphate and carbonate diagenesis of the ArnagerLimestone started in the upper metres of the sediment below theseafloor and continued during shallow burial to depths of about200–250 m. The diagenesis resulted in precipitation of opal-CT,clinoptilolite, feldspar, smectite, barite and dolomite. Quartz formedduring later dissolution–precipitation processes, but was probablyalso aided by higher pressure and temperature during maximumburial.

The presence of opal-CT, clinoptilolite and barite is interpreted toreflect high bio-siliceous productivity. Initial dissolution of spongespicules took place close to the sea floor. It caused a rise in silica activityand precipitation of opal-CT at temperatures around 17°C. It resulted inlowering of the Si/Al ratio in the pore water and precipitation of morealuminium-rich minerals such as clinoptilolite, feldspar and smectite.The precipitation of Sr-rich barite and dolomite was facilitated bydissolution of siliceous sponge spicules and microbial decomposition ofsponge tissue.

In bio-siliceous chalk like the Arnager Limestone, high porosityand permeability are likely to occur in areas dominated by a highnumber of dissolved spicules, whereas low porosity and permeabilityreflect more intense opal-CT and clinoptilolite cementation of thematrix.

Acknowledgements

This studywas financially supported by ConocoPhillips, Norway andthe Danish Natural Science Research Council (FNU). Anna Katarinopou-lou is acknowledged for valuable help with Rietveld analyses. NielsSpringer, Peter Frykman and Ida Lykke Fabricius are thanked for fruitfuldiscussion.We thank the journal reviewer Ian Jarvis and an anonymousreviewer for comments.

60 H.B. Madsen et al. / Sedimentary Geology 223 (2010) 51–60

References

Aase, N.E., Bjørkum, P.A., Nadeau, P.H., 2001. The effect of grain-coating microquartz onpreservation of reservoir porosity. American Association of Petroleum GeologistsBulletin 80, 1654–1673.

Abercrombie, H.J., Hutcheon, I.E., Bloch, J.D., Decaritat, P., 1994. Silica activity and thesmectite — illite reaction. Geology 22, 539–542.

Berger, W.H., von Rad, U., 1972. Cretaceous and Cenozoic sediments from the AtlanticOcean. Initial Report DSDP 14, 787–954.

Bernoulli, D., Gunzenhauser, B., 2001. A dolomitized diatomite in an Oligocene–Miocenedeep-sea fan succesion, Gonfolite Lombarda Group, Northern Italy. SedimentaryGeology 139, 71–91.

Bowers, T.S., Burns, R.G., 1990. Activity diagrams for clinoptilolite: Susceptibility of thiszeolite to further diagenetic reactions. American Mineralogist 75, 601–619.

Brückner, A., 2006. Taxonomy and paleoecology of lyssacinosan Hexactinellida from theUpper Cretaceous (Coniacian) of Bornholm, Denmark, in comparison with otherPostpaleozoic representatives. Abhandlungen der Senckenbergischen Natur-forschenden Gesellschaft 564, 103.

Brückner, A., Janussen, D., 2005. Rossella bromleyi n. sp.: the first entirely preservedfossil sponge of the genus Rosella (Hexactinellida) from the Upper Cretaceous ofBornholm, Denmark. Journal of Paleontology 79, 21–28.

Canfield, D.E., Raiswell, R., 1991. Carbonate precipitation and dissolution: its relevanceto fossil preservation. In: Allison, P.A., Derek, E.G.B. (Eds.), Releasing the DataLocked in the Fossil Record. Plenum Press, New York, pp. 412–445.

Chaika, C., Dvorkin, J., 2000. Porosity reduction during diagenesis of diatomaceousrocks. American Association of Petroleum Geologists Bulletin 84, 1173–1184.

Christensen, W.K., 1985. The Albian to Maastrichtian of southern Sweden andBornholm, Denmark: a review. Cretaceous Research 5, 313–327.

Christensen, W.K., Schulz, M.-G., 1997. Coniacian and Santonian belemnite faunas fromBornholm, Denmark. Fossils and Strata 44, 1–73.

Dapples, E.C., 1979. Silica as an agent. In: Larsen, G., Chilingar, G.V. (Eds.), Diagenesis insediments and sedimentary rocks. Developments in Sedimentology, vol. 25. Elsevier,Amsterdam, pp. 99–141.

Dorn, P., Bräutigam, F., 1959. Hinwise auf Oberkreide vulkanismus in NW-Deutschland.Abhandlungen der Braunschweigischen Wissenschaftlichen Gesellschaft 11, 1–4.

Eagle, M., Paytan, A., Arrigo, K.R., van Dijken, G., Murray, R.W., 2003. A comparisonbetween excess barium and barite as indicators of carbon export. Paleoceanography18, 1021.

Fabricius, I.L., Borre, M.K., 2007. Stylolites, porosity, depositional texture, and silicatesin chalk facies sediments. Ontong Java Plateau – Gorm and Tyra fields North Sea.Sedimentology 54, 183–205.

Gingele, F.X., Schulz, H.D., 1993. Authigenic zeolites in Late Pleistocene sediments of theSouth Atlantic (Angola Basin). Marine Geology 111, 121–131.

Graversen, O., 2004. Upper Triassic–Cretaceous stratigraphy and structural inversionoffshore SW Bornholm, Tornquist Zone, Denmark. Bulletin of the Geological Societyof Denmark 51, 111–136.

Gravesen, O., Rolle, F., Surlyk, F., 1982. Lithostratigraphy and sedimentary evolution ofthe Triassic, Jurassic and lower Cretaceous of Bornholm, Denmark. Bulletin of theGeological Society of Denmark Series B 7, 1–51.

Hay, R.L., 1977. Geology of zeolites in sedimentary rocks. In: Mumpton, F.A. (Ed.),Mineralogy and Geology of Natural Zeolites. Mineralogical Society of America,Short Course Notes, vol. 4, pp. 53–64.

Huggett, J.M., Gale, A.S., Wray, D.S., 2005. Diagenetic clinoptilolite and opal-CT from theMiddle Eocene Wittering Formation, Isle of Wight, U.K. Journal of SedimentaryResearch 75, 585–595.

Hurd, D.C., 1973. Interaction of biogenic opal, sediment and seawater in the centralequatorial Pacific. Geochimica et Cosmochimica Acta 37, 2257–2282.

Iler, R.K., 1973. Effect of adsorbed alumina on the solubility of amorphous silica inwater. Journal of Colloid and Interface Science 43, 399–408.

Jacka, A., 1974. Replacement of fossils by length-slow chalcedony and associateddolomitization. Journal of Sedimentary Petrology 41, 1045–1058.

Jensen, J.B., Hamann, N.E, 1988. Geological mapping of Mesozoic deposits along theeastern margin of the Rønne Graben, offshore Bornholm, Denmark. Bulletin of theGeological Society of Denmark 37, 237–260.

Kano, K., 1983. Ordering of opal-CT in diagenesis. Geochemical Journal 17, 87–93.Karpoff, A.M., Destrigneville, C., Stille, P., 2007. Clinoptilolite as a new proxy of

enhanced biogenic silica productivity in lower Miocene carbonate sediments of theBahamas platform: isotopic and thermodynamic evidence. Chemical Geology 245,285–304.

Kasten, S., Haese, R.R., Zabel, M., Ruhlmann, C., Schulz, H.D., 2001. Barium peaks atglacial terminations in sediments of the equatorial Atlantic Ocean — relicts ofdeglacial productivity pulses? Chemical Geology 175, 635–651.

Kastner, M., 1981. Authigenic silicates in the deep-sea sediments: formation anddiagenesis. In: Emiliani, C. (Ed.), The Sea. JohnWiley & Sons, NewYork, pp. 915–980.

Kastner, M., Keene, J.B., Gieskes, J.M., 1977. Diagenesis of siliceous oozes, I. controls onthe rate of opal-A to opal-CT transformation — an experimental study. Geochimicaet Cosmochimica Acta 41, 1041–1059.

Kastner, M., Stonecipher, S.A., 1978. Zeolites in pelagic sediments of the Atlantic, Pacific,and Indian Oceans. In: Sand, L.B., Mumpton, F.A. (Eds.), Natural Zeolites. PergamonPress, New York, pp. 199–220.

Kennedy, W.J., Christensen, W.K., 1991. Coniacian and Santonian ammonites fromBornholm, Denmark. Bulletin of the Geological Society of Denmark 38, 203–244.

Liboriussen, J., Ashton, P., Tygesen, T., 1987. The tectonic evolution of the Fennoscan-dian Border Zone in Denmark. Tectonophysics 137, 21–29.

Maldonado, M., Carmona, C., Velasquez, Z., Puig, A., Cruzado, A., Lopez, A., Young, C.M.,2005. Siliceous sponges as a silicon sink: an overlooked aspect of benthopelagiccoupling in the marine silicon cycle. Limnology and Oceanography 50, 799–809.

Maliva, R.G., Dickson, J.A.D., 1992. Microfacies and diagenetic controls of porosity inCretaceous/Tertiary chalks, Eldfisk, Norwegian North Sea. American Association ofPetroleum Geologists Bulletin 76, 1825–1838.

Mizutani, S., 1970. Silica minerals in the early stage of diagenesis. Sedimentology 15,419–436.

Mogensen, A.T., 1994. Diagenese af kiselholdigt skrivekridt og dannelsen af silika,heulandit, feldspater og phyllosilikater i Arnager Kalk Formationen. UnpublishedThesis, Geological Institute, University of Copenhagen, Copenhagen, Denmark.

Murata, K.J., Nakata, J.K., 1974. Cristobalitic stage in the diagenesis of diatomaceousshale. Science 184, 567–568.

Murata, K.J., Friedman, I., Gleason, J.D., 1977. Oxygen isotope relations betweendiagenetic silica minerals in Monterey Shale, Temblor Range, California. AmericanJournal of Science 277, 259–272.

Noe-Nygaard, N., Surlyk, F., 1985. Mound bedding in a sponge-rich Coniacian chalk,Bornholm, Denmark. Bulletin of the Geological Society of Denmark 34, 237–249.

Norin, R., 1933. Mineralogische und petrographische Studien an den Basalten Schonens.Geologiska Föreningens i Stockholm Förhandlinger 55, 101–149.

Norin, R., 1934. Zur Geologie der südschwedischen Basalte. Meddelanden från LundsGeologisk-Mineralogiska Institution. 57, 1–174.

Packer, S.R., Hart,M.B., 1994. Evidence for sea level change from the Cretaceous of Bornholm,Denmark. Geologiska Föreningens i Stockholm Förhandlinger 116, 167–173.

Printzlau, I., Larsen, O., 1972. K/Ar age determinations on alkaline olivine basalts fromScania, southern Sweden. Geologiska Föreningens i Stockholm Förhandlinger 94,259–269.

Ravn, J.P.J., 1918. Kridtaflejringerne paa Bornholms Sydvestkyst og deres Fauna.Danmarks geologiske Undersøgelse II. Række 31, 1–39.

Ravn, J.P.J., 1921. Kridtaflejringer på Bornholms sydvestkyst og deres fauna. II. Senonet.Kridtaflejringerne ved Stampe Aa. Danmarks geologiske Undersøgelse II. Række 32,1–52.

Røgen, B., Gommesen, L., Fabricius, I.L., 2001. Grain size distributions of chalk fromimage analysis of electron micrographs. Computers & Geosciences 27, 1071–1080.

Seibert, E., Vortisch, W., 1979. Zur Stratigraphie, Petrologie und Genese einer Bentonit-Lage aus dem Oberen Mittel-Turon (Oberkreide) des südöstlischen Münsterlandes.Geologisches Rundschau 68, 649–679.

Siever, R., 1962. Silica solubility, 0–200 °C and the diagenesis of siliceous sediments.Journal of Geology 70, 127–150.

Solokowski, S., 1976. The Cretaceous. In: Solokowski, S. (Ed.), Geology of Poland.Publishing House Wydawnictwa, Warsaw, p. 859.

Stonecipher, S.A., 1976. Origin, distribution and diagenesis of philipsite and clinopti-lolite in deep sea sediments. Chemical Geology 17, 307–318.

Stonecipher, S.A., 1978. Geochemistry and deep-sea philipsite, clinoptilolite and hostsediments. In: Sand, L.B., Mumpton, F.A. (Eds.), Natural Zeolites: Occurrence,Properties, Use. Pergamon, Oxford, pp. 221–234.

Treguér, P., Nelson, D.M., Van Bennekom, A.J., DeMaster, D.J., Leynaert, A., Queguiner, B.,1995. The silica balance in the world ocean: a reestimate. Science 268, 375–379.

Tröger, K.-A., Christensen, W.K., 1991. Upper Cretaceous (Cenomanian–Santonian)inoceramid bivalve faunas from the island of Bornholm, Denmark. Danmarksgeologiske Undersøgelse Serie A 28, 1–47.

Vasconcelos, C., McKenzie, J.A., 1997. Microbial mediation of modern dolomiteprecipitation and diagenesis under anoxic conditions (Lagoa Vermelha, Rio DeJaneiro, Brazil). Journal of Sedimentary Research 67, 378–390.

Vasconcelos, C., McKenzie, J.A., Bernasconi, S., Grujic, D., Tiens, A.J., 1995. Microbialmediation as a possible mechanism for natural dolomite formation at lowtemperatures. Nature 377, 220–222.

Vejbæk, O.V., 1985. Seismic stratigraphy and tectonics of sedimentary basins aroundBornholm, southern Baltic. Danmarks geologiske Undersøgelse Serie A 8, 1–30.

Weaver, C.E., 1968. Mineral facies in the Tertiary of the continental shelf and BlakePlateau. Southeastern Geology 9, 57–63.

Wilkin, R.T., Barnes, H.L., 1998. Solubility and stability of zeolites in aqueous solution: I.Analcime, Na- and K-clinoptilolite. American Mineralogist 83, 746–761.

Williams, L.A., Parks, G.A., Crerar, D.A., 1985. Silica Diagenesis, I. Solubility Controls.Journal of Sedimentary Petrology 55, 301–311.

Wright, D.T., 2000. Benthic microbial communities and dolomite formation in marineand lacustrine environments — a new dolomite model. In: Glenn, C., Prévôt-Lucas,R.L., Lucas, J. (Eds.), Marine Authigenesis: From Global to Microbial, vol. 66. SEPMSpecial Publication, Tulsa, pp. 7–20.

Wright, D.T., Wacey, D., 2005. Precipitation of dolomite using sulphate-reducingbacteria from the Coorong Region, South Australia: significance and implications.Sedimentology 52, 987–1008.