Constraining silica diagenesis in methane-seep deposits

Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo

Constraining silica diagenesis in methane-seep deposits D. Smrzka a, S.M. Kraemer b, J. Zwicker a, D. Birgel a, D. Fischer c,d, S. Kasten c,d, J.L. Goedert e, J. Peckmann a,⁎ a

Department für Geodynamik und Sedimentologie, Erdwissenschaftliches Zentrum, Universität Wien, 1090 Wien, Austria Department für Umweltgeowissenschaften, Erdwissenschaftliches Zentrum, Universität Wien, 1090 Wien, Austria Alfred-Wegener-Institut Helmholtz-Zentrum für Polar- und Meeresforschung, 27570 Bremerhaven, Germany d MARUM — Zentrum für Marine Umweltwissenschaften, Universität Bremen, 28334 Bremen, Germany e Burke Museum of Natural History and Culture, University of Washington, Seattle, WA 98195, USA b c

a r t i c l e

i n f o

Article history: Received 29 April 2014 Received in revised form 3 December 2014 Accepted 4 December 2014 Available online 11 December 2014 Keywords: Silicification Authigenic quartz Methane seeps Anaerobic oxidation of methane Geochemical modeling

a b s t r a c t Silicified fossils and authigenic silica are common in ancient seep limestones. Silicification of calcareous fossils facilitates the preservation of even fine details and is therefore of great interest to paleontologists, permitting a reliable taxonomic identification of the chemosynthesis-based taxa that lived at hydrocarbon seeps. Four methane-seep limestones of Paleozoic, Mesozoic, and Cenozoic age with abundant silica phases are compared in this study; one, an Eocene seep deposit on the north shore of the Columbia River at Knappton, western Washington State, USA, is described for the first time. Its lithology and fabrics, negative δ13Ccarbonate values as low as − 27.6‰, and 13C-depleted biomarkers of archaea involved in the anaerobic oxidation of methane (AOM) reveal that the carbonate rock formed at a methane seep. The background sediments of the studied Phanerozoic seep limestones contain abundant siliceous microfossils, radiolarian tests in case of the Carboniferous Dwyka Group deposits from Namibia and the Triassic Graylock Butte deposits from Oregon (USA), as well as diatom frustules in case of the Eocene Knappton limestone and an Oligocene seep deposit from the Lincoln Creek Formation (Washington State, USA). These microfossils are regarded as the source of dissolved silica, causing silicification and silica precipitation. Silica cements formed after AOM-derived cements ceased to precipitate but before equant calcite formed. Numerical experiments using the computer code PHREEQC confirmed that (1) AOM increases the pH of pore waters and that (2) this pH increase subsequently mobilizes biogenic silica, (3) followed by the re-precipitation of the dissolved silica in the periphery of the AOM hotspot. The experiments revealed that degassing of carbon dioxide has the potential to significantly increase the local pH of pore waters, exerting an even stronger control on the local pH and silica dissolution than the rate of AOM alone. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Marine cold seeps are sites on the seafloor where reduced, hydrocarbon-rich fluids of ambient temperature emanate into bottom waters. Methane and oil seeps occur along active and passive continental margins where the advection of fluids is facilitated by faults and conduits (Judd and Hovland, 2007). The microbially-mediated sulfatedependent anaerobic oxidation of methane (AOM) is the key biogeochemical process at seeps (Hinrichs et al., 1999; Boetius et al., 2000; Aloisi et al., 2002; Peckmann and Thiel, 2004). It is a consequence of the migration of methane and proceeds according to the overall reaction: 2−

−

−

CH4 þ SO4 →HCO3 þ HS þ H2 O:

ð1Þ

The production of bicarbonate raises the pore water alkalinity, commonly resulting in the precipitation of carbonate minerals (e.g., Ritger ⁎ Corresponding author. Tel.: +43 1 4277 53470. E-mail address: [email protected] (J. Peckmann).

http://dx.doi.org/10.1016/j.palaeo.2014.12.007 0031-0182/© 2014 Elsevier B.V. All rights reserved.

et al., 1987; Wallmann et al., 1997). The most common minerals observed at seeps are early diagenetic micrite (i.e., microcrystalline calcite) and fibrous aragonite cement, although microcrystalline aragonite or dolomite may be locally abundant too (Aloisi et al., 2000; Peckmann and Thiel, 2004; Campbell, 2006; Nöthen and Kasten, 2011). Micrite mostly makes up the matrix of seep limestones; aragonite cement commonly occurs as pore-filling. Other authigenic minerals that form at seeps include iron sulfide and sulfate precipitates (e.g., Aloisi et al., 2003; Peckmann et al., 2003; Torres et al., 2003; Kasten et al., 2012; Fischer et al., 2013). Silica minerals such as chalcedony and quartz, however, have to date only been recognized in ancient seep limestones (e.g., Goedert et al., 2000). The term chalcedony is used to describe a disordered, microcrystalline, fibrous variety of quartz (sensu Flörke et al., 1991). Silicification (i.e., replacement of other minerals or organic material by silica minerals) and silica precipitation are phenomena that have been reported for various seep deposits of different ages, and appear to have been common processes. For example, Late Carboniferous seep limestones in the glaciomarine Dwyka Group in southern Namibia (Himmler

14

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

et al., 2008) contain chalcedony and silica cements. The background sediments of the Dwyka Group seep deposits contain abundant radiolarian tests, which are also found within the seep limestones themselves (Himmler et al., 2008). Similarly, Jurassic–Cretaceous seep deposits enclose sponge spicules that are replaced by calcite (Hryniewicz et al., 2012). Their host sediments, on the other hand, have been found to contain siliceous sponge spicules (Nakrem and Kiessling, 2012). Another example are silicified brachiopod shells and silica phases enclosed in Late Triassic seep deposits from Eastern Oregon, USA (Peckmann et al., 2011). Abundant silica phases have also been recognized in Cretaceous methane-seep deposits from New Zealand (Kiel et al., 2013). Like for the other examples, silicification affected the early diagenetic banded and botryoidal cements that resulted from AOM. The silicified carbonate cements are commonly postdated by silica cement filling cavities (Himmler et al., 2008; Peckmann et al., 2011; Kiel et al., 2013). Most seep limestones that have been shown to feature silicified fossils and silica cements are of Cenozoic age, reflecting the overall abundance of recognized Cenozoic seep deposits (cf. Campbell, 2006). Prominent examples include Late Eocene to Early Oligocene seep limestones from Japan, which contain silica that replaces calcite in the interior of voids (Amano et al., 2013). The Oligocene part of the Lincoln Creek Formation in western Washington State, USA, also yielded several, partly silicified seep limestones with silicified mollusks (Kiel, 2010b) and silicified aragonite botryoids and aragonite needles preserved within later formed quartz crystals (Peckmann et al., 2002). Similarly, Goedert et al. (2000) described silicified aragonite botryoids surrounding worm tubes, authigenic quartz crystals, but also silicified tube walls of vestimentiferan-like worms. Finally, Kuechler et al. (2012) reported partially and completely silicified fossils and carbonate cements from the Miocene Astoria Formation, also located in western Washington State, USA. These authors highlighted the common patterns in the precipitation of silica phases, occurring after AOM-derived mineral phases ceased to form, but before late diagenetic calcite cements precipitated. Kuechler et al. (2012) emphasized the significance of silicification for the quality of fossil preservation and suggested a scenario for the silicification of seep limestones. Silicified fossils are of special interest for paleontologists, as fine details are preserved that allow for reliable identification. Likewise, Goedert et al. (2000) based their scenario for worm tube preservation on the engulfment of tubes by carbonate cements and subsequent replacement of tube walls and cements by silica minerals. Hikida et al. (2003) described tube worms filled by carbonate cement and chalcedony; tube walls recrystallized by chalcedony were reported to be especially well preserved. Similarly, Kaim et al. (2008) and Kiel (2010b) reported excellent preservation of gastropod shell ornamentations and, in particular, of protoconchs due to early silicification, enabling reliable taxonomical identification. The silicification of carbonate shells of seep biota is a critical process for the quality of fossil preservation, representing a significant constraint in the endeavor to unravel evolutionary patterns among chemosynthesis-based lineages. This study compares four ancient seep deposits in order to improve our understanding of the driving forces of the silicification of calcareous fossils and carbonate minerals. The lithology and geochemical patterns of one of the deposits, the Knappton deposit from western Washington, are described for the first time, the other three examples have been previously detailed. We combine petrology, stable isotope and lipid geochemistry, as well as paleontology to achieve a comprehensive description and interpretation of silicification processes in seep deposits. In order to arrive at a mechanistic understanding of how the formation of authigenic silica is linked to AOM, we performed numerical experiments using the computer code PHREEQC (Parkhurst and Appelo, 1999). We tested the hypothesis of Kuechler et al. (2012) that (1) anaerobic oxidation of methane leads to a local pH increase in pore waters and that (2) this increase leads to higher solubilities and dissolution rates of biogenic silica (i.e., radiolarian tests, diatom frustules, sponge spicules), followed by (3) the precipitation of silica due to a decrease

of pH values after AOM locally ceased. This effort provides the key to our understanding of how silica phases are mobilized, transported, and subsequently precipitated after diffusion into microenvironments typified by lower pH values. 2. Geologic setting 2.1. Knappton, “Siltstone of Shoalwater Bay”, western Washington, USA The tectonic unit in which most fossil seep deposits are found in western Washington is the Coast Range Terrane, consisting of Early Eocene basalts overlain by a thick cover of Middle Eocene to Early Miocene sediments (Brandon and Calderwood, 1990; Stewart and Brandon, 2004). The seep-bearing sediments of this terrane were deposited before and after the onset of the subduction of the Juan de Fuca plate beneath the North American plate around 35 million years ago, implying that the oldest seep carbonates were deposited on a passive continental margin, whereas all younger seep deposits formed in an active subduction-related setting (Kiel, 2010a). The subducting slab of the Juan de Fuca plate caused the uplift of the Olympic Mountains and induced compressive forces during the Eocene, which squeezed subsurface methane-rich waters toward the sediment–water interface (Kiel, 2010a). The informally named “Siltstone of Shoalwater Bay” (Wells, 1989) is part of a thick sequence of Eocene siltstone and sandstone units that have been collectively described as thin-bedded, laminated concretionary siltstones and fine micaceous sandstones. It is of Late Eocene age in the study area, containing dark gray, thin-bedded, laminated, indurated tuffaceous siltstones with thin tuff beds, minor thin-bedded feldspathic sandstone, and calcareous concretions (Wells, 1989), and extends from Knappton on the north shore of the Columbia River, northwesterly to Willapa Bay where minor amounts of seep limestone occur locally. The Knappton limestone deposit studied herein is derived from the “Siltstone of Shoalwater Bay” and will be referred to as the Knappton SSB limestone (Fig. 1). It crops out on both sides of the highway approximately 300 m north of the monument marking the former town-site of Knappton (Goedert and Benham, 2003), adjacent to numerous Holocene and Pleistocene landslide deposits comprised of poorly sorted colluvium and bedrock slump blocks (Wells, 1989). The Knappton SSB limestone contains few megafossils, mostly bathymodiolins (Kiel and Amano, 2013), and has well-developed cement crusts (Goedert and Benham, 2003). The Knappton SSB limestone should not be confused with seep limestones previously documented from the Lincoln Creek Formation just to the northeast of Knappton (Goedert and Squires, 1993; Kiel, 2010b) or those from the Astoria Formation farther east (Amano and Kiel, 2007; Kiel, 2010b; Kuechler et al., 2012). 2.2. Canyon River, Lincoln Creek Formation, western Washington, USA The Lincoln Creek Formation experienced a similar tectonic history to the aforementioned “Siltstone of Shoalwater Bay”. The Lincoln Creek Formation is Late Eocene to Early Miocene in age and is comprised of marine, dark to olive gray tuffaceous siltstone and feldspathic sandstones, as well as laminated, thin-bedded burrowed siltstone and interbeds of graded, fine to medium grained micaceous feldspathic sandstone (Wells, 1989). It has been cut through by several river valleys on the south side of the Olympic Mountains (Goedert et al., 2000). Numerous seep deposits are exposed along the banks of the Canyon and Satsop Rivers (Peckmann et al., 2002), where the formation reaches a thickness of about 3000 m (Rau, 1966). As the rivers have cut through the Lincoln Creek Formation almost perpendicularly to strike, a fairly continuous stratigraphic sequence is exposed (Prothero and Armentrout, 1985). Deposition was shown to have occurred at water depths of 600 to 800 m in cool to cold waters using foraminiferans (Rau, 1966). The seep deposits are mostly autochthonous, rather small and enclosed within thick-bedded siltstones and mudstones (Goedert et al., 2000; Peckmann et al., 2002; Nesbitt et al., 2013).

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

2.3. Graylock Butte limestones, Rail Cabin Member, eastern Oregon, USA The Graylock Butte limestones from eastern Oregon have been investigated by Peckmann et al. (2011). They are part of the Late Triassic Blue Mountain succession; this siliciclastic sequence is approximately 4000 m thick and was deposited in a forearc setting within a convergent margin (Dickinson and Thayer, 1978; Dickinson, 1979). The limestones are found within the 600 m thick mudstone unit known as Rail Cabin Member in the upper part of the Vester Formation (Blome et al., 1986). The mudstones are rich in radiolarians, suggesting a Late Carnian to Middle Norian age (Pessagno et al., 1979). However, these ages have been questioned since large parts of the Rail Cabin Member are poorly exposed (Pessagno and Whalen, 1982). With respect to the studied seep deposits, a Norian age was confirmed by the occurrence of a conodont of the genus Epigondolella (Peckmann et al., 2011).

15

The detailed procedure is described in Kiel et al. (2013) and references therein. The PHREEQC code (Parkhurst and Appelo, 1999) was used to simulate chemical reactions in a sediment column. We first conducted simple numerical batch-reaction experiments, followed by more complex, numerical reactive-transport experiments. These experiments incorporate aqueous equilibrium speciation, heterogeneous reactions, irreversible

2.4. Ganigobis limestones, Dwyka Group, Namibia The Dwyka Group sediments from Namibia are the lowest stratigraphic unit of the Karoo Supergroup, comprised mainly of siliciclastic deposits (Bangert et al., 2000). Sedimentation of the glaciomarine Dwyka Group began in the Late Carboniferous following the Damara orogenesis in southern Namibia and accumulated in a NW to SE striking intracontinental rift basin (Stollhofen, 1999). Dwyka Group sediments are well exposed in the Aranos Basin area, where they are characterized by a fining–upward sequence of glaciomarine deposits, possibly representing deglaciation sequences (Visser, 1997). Himmler et al. (2008) described seep limestones from the Ganigobis Shale Member, an 80 m thick mudstone unit deposited under full marine conditions in a water depth of about 600 m (Martin and Wilczewski, 1970). 3. Material and methods Rock samples of the Knappton SSB limestone were collected as lag material on river and shore-bank exposures and from strata above the beach on the west side of the highway (Goedert and Benham, 2003). Eleven rock samples were used for this study, all of which were collected at the sampling site close to Knappton (46° 16.490′ N, 123° 48.929′ W). Thin sections (150 x 100 mm) of each rock sample were prepared for standard petrographic and fluorescence microscopy. To discriminate carbonate minerals, some thin sections were partly stained with a mixture of potassium ferricyanide and alizarin red solution, dissolved in 0.1% hydrochloric acid. Some thin sections were partially stained with Feigl's solution to distinguish aragonite and calcite. Petrographic observations were carried out with a Nikon SMZ 1500 stereomicroscope coupled to a Prog.Res Speed XT core 5 camera, as well as a Nikon Optiphot-2 optical microscope. For image analysis and camera control, the software Prog.Res Capture pro 2.8 was used. Photographs under crossed-polarized light were made with a Leica DM 4500 P polarization microscope coupled to a Leica DFC 420 camera. For image analysis as well as camera control, the Leica Application Suite v. 3.2 software was used. Sample powder for X-ray diffraction was taken from polished slabs using a handheld microdrill. The powder was analyzed with a Panalytical PW 3040/60 X'Pert PRO (CuKα radiation, 40 kV, 40 mA, step size 0.0167, 5 s per step) diffractometer. The X-ray diffraction patterns were interpreted using the software “X'Pert High score plus” (Panalytical). Samples for oxygen and carbon stable isotopes from the Knappton SSB limestone were taken from polished slabs using a handheld microdrill at low rotational speed. Samples were analyzed at the light stable isotope laboratory of the Institute of Earth Sciences, KarlFranzens University, Graz. Effective accuracy was ±0.05 for δ13C values and ±0.11 for δ18O values. All values are reported in per mil relative to the Vienna PeeDee Belemnite (V-PDB) standard. Reproducibility was checked by replicate analysis of each sample. One limestone sample from Knappton was prepared and analyzed for its lipid biomarker inventory. Prior to analysis the limestone was decalcified and extracted .

Fig. 1. Sketch maps showing the locations of the studied methane-seep deposit close to the ghost town of Knappton derived from the “Siltstone of Shoalwater Bay” (SSB) and the Canyon River seep deposit with worm tubes (Goedert et al., 2000) used for comparison (A–C). (A) Study area in the northwestern USA. (B) Locations of the Knappton SSB and Canyon River seep deposits in western Washington State; modified after Goedert and Peckmann (2005). (C) Location of the Knappton seep deposits exposed on the northern shore of the Columbia River; modified after Kiel (2010b).

16

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

and kinetically controlled reactions, and calculate solution composition and saturation indices of minerals. Issues such as sedimentation, bioturbation, and local porosity and permeability changes of the sediment were not considered. Sensitivity analyses have been conducted to highlight the relevance of the respective processes in the system. Batch reaction experiments simulate a 1 dm3 cube of water of a given composition to which the reaction products of reaction (1) were added to solution. Sodium to calcium ratios corresponding to their concentration in seawater were added to account for charge balance of the solution. Solid phase assemblages of minerals were introduced corresponding to their mean contents in marine sediments. The processes under investigation included heterogeneous reactions between the solution and mineral phases (i.e., dissolution and precipitation reactions) and between the solution and a gas phase (i.e., the methane/carbon dioxide gas mixture characteristic for methane seeps). Reactive transport experiments simulate a volume equal to the batch calculation under similar conditions, but with fluid-flow through the reaction zone consistent with observed flow rates using data from Torres et al. (2002). Flow rates were recalculated in terms of the time required to completely exchange one pore volume of the reaction volume, the duration of one exchange was 66.3 days. The influent solution had the same composition as the initial pore water of the batch experiments discussed above. The total simulation time for the reactive-transport runs was set to twelve years. As opposed to the batch reaction calculations where precipitation of minerals is solely controlled by equilibrium thermodynamics, reaction kinetics for silica dissolution were applied during reactive transport calculations. Rates of AOM were modified over time to enable the construction of graphs, showing differences of parameters of interest as a function of AOM rates over the total simulation time. Simulated AOM rates span a wide range of magnitudes (Appendix, Table A.4) to better illustrate their effects on other parameters. The solid phase assemblage was adopted from the batch reaction calculations. The generation of alkalinity by AOM during the model runs led to ionic strengths (up to 2.9 M) that were significantly elevated compared to seawater. Therefore it was necessary to use Pitzer ionic strength corrections using the Thereda database (www.thereda.de). A detailed description of numerical batch and reactive-transport experiments including the modeling concept, boundary conditions, and input parameters is provided in the Appendix. 4. Results and discussion 4.1. Seep deposits with abundant silica 4.1.1. Knappton SSB limestone Samples of the Knappton SSB limestone are characterized by abundant large cavities, most of which are filled by cement. The most abundant cavity-filling phase is banded and botryoidal aragonite cement that forms the framework of Knappton SSB limestone (Figs. 2, 3). Therefore, the rock can be classified as cement framestone (sensu Flügel, 2004). Banded and botryoidal cement is characterized by aggregates of acicular crystals that form isopachous rims or botryoids. Micrite formed at different stages in the paragenetic sequence; it pre- and postdates banded and botryoidal cement. Micrite predating banded and botryoidal cement encloses abundant pyrite aggregates and detrital quartz grains. Occasionally, it exhibits a clotted or peloidal fabric (Fig. 3B). Micrite is commonly cross cut by a multitude of veins of variable size, which are mostly filled by carbonate cement. The banded and botryoidal aragonite cement is in places partly or completely replaced by silica. Different varieties of silica formed after banded and botryoidal cement, not only replacing this carbonate cement but also filling the remaining cavity space (Fig. 3C, D). Silica phases reveal two general types of textures, apparent in crossed-polarized light. The first variety is a very fine grained, microcrystalline texture. It is typified by grain sizes lower than 10 μm and is referred to as microquartz (sensu Knauth, 1994). The second variety is characterized by fibrous crystals arranged as botryoids. It forms crystal fans with

Fig. 2. Typical fabric of the Knappton SSB seep limestone with abundant banded and botryoidal aragonite cement (bbc), scanned thin section; m = micrite, ya = yellow aragonite, fms = fibrous microcrystalline silica, ec = equant calcite cement, r = embedding resin. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

oscillatory extinction under crossed-polarized light (Fig. 3B, D) and is referred to as fibrous microcrystalline silica (sensu Knauth, 1994). Under plane-polarized light, fibrous microcrystalline silica exhibits a pale yellowish to brownish color (Fig. 3C). It occurs as aggregates of fibrous crystals with a very high length to width ratio (Fig. 3B) and as botryoidal aggregates made of acicular crystals with a somewhat lower length to width ratio (Fig. 3B, D). Although fibrous microcrystalline silica comes in a wide range of slightly different textures (Fig. 3B), it can be collectively classified as chalcedony. Silica phases have been recognized in every rock sample, but represent only a small fraction of the rock volume, corresponding to approximately 5%. The latest phase in the paragenetic sequence is equant calcite spar, postdating silica in all samples (Fig. 3A). It is a minor phase, representing a late-stage cavity filling. Various cleavage domains within individual crystals are apparent under crossed-polarized light. 4.1.2. Lincoln Creek Formation deposit Micrite is the most abundant phase in the limestone with worm tubes from the Lincoln Creek Formation. It forms the matrix of the rock, commonly exhibiting a clotted fabric (Fig. 4A), and contains abundant pyrite crystals. Banded and botryoidal aragonite cement is also a common rock constituent. However, it is much less abundant than in the Knappton SSB limestone. The cement aggregates are made of fibrous crystals that grew toward the center of cavities (Fig. 4B) and show the typical rolling extinction under crossed-polarized light. Banded and botryoidal cement postdates micrite, but predates silica phases (Fig. 4B). Silica fills cavities, vugs, and cracks in the form of microquartz, microcrystalline fibrous silica, and megaquartz (Fig. 4). As noted for the Knappton SSB limestone, fibrous microcrystalline silica comes in a variety of slightly different radial textures, all revealing rolling extinction under crossed-polarized light. Megaquartz (sensu Knauth, 1994) reveals a macrocrystalline, drusy texture made up of crystals that reach several millimeters in diameter (Fig. 4A, C). Siliceous body fossils other than locally occurring hexactinellid sponges (Rigby and Goedert, 1996) are not commonly found in seep limestones of the Lincoln Creek Formation (Kuechler et al., 2012). However, molecular fossils of diatoms extracted from the limestones (Peckmann et al., 2002) reveal that diatoms have been present in the sediment even in cases where microfossils are lacking. The existence of abundant silica phases in the paragenetic sequence of the seep limestones is consequently best explained by dissolution of diatom frustules. The studied limestone from the Lincoln Creek Formation is characterized by abundant worm tubes (Goedert et al., 2000). Most tubes are

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

17

Fig. 3. Photomicrographs of the Knappton SSB limestone; cm = clotted micrite, bbc = banded and botryoidal aragonite cement, miq = microquartz, fms = fibrous microcrystalline silica, ec = equant calcite cement. (A) Banded and botryoidal aragonite cement, fibrous microcrystalline silica, and equant calcite filling a large cavity, plane-polarized light. (B) Two varieties of fibrous microcrystalline silica filling a small crack cutting through clotted micrite, crossed-polarized light. (C) Fibrous crystals of banded and botryoidal aragonite pointing toward the center of a cavity filled by fibrous microcrystalline silica, plane-polarized light. (D) Same detail as (C), crossed-polarized light.

silicified. Tube walls consist of microquartz, which is postdated by microcrystalline fibrous silica and megaquartz (Fig. 4C). Microcrystalline fibrous silica commonly nucleated at the inner tube walls, forming rims consisting of banded and botryoidal aggregates. Megaquartz crystals occur in the center of worm tubes, featuring various crystal sizes and diffuse grain boundaries (Fig. 4C). The intense silicification of the tubular fossils promoted the excellent preservation of tube wall delamination (Fig. 4C). This delamination fabric was not recognized by Goedert et al. (2000). Delamination has been suggested to represent a taphonomic feature of the primarily organic tube walls of vestimentiferan worms (Peckmann et al., 2005; Haas et al., 2009), agreeing with the tentative taxonomic assignment put forward by Goedert et al. (2000); but see Kiel and Dando (2009) for a note of caution. Irrespective of the taxonomic affiliation of the Lincoln Creek Formation worms, the preservation of such fine details confirms the great potential of silicification as an asset to fossil identification. 4.1.3. Graylock Butte deposits The limestones of the two Graylock Butte deposits are typified by large amounts of mostly homogenous micrite, which makes up the matrix of the rock characterized by the mass occurrence of the dimerelloid brachiopod Halorella. The limestones are pervaded by numerous smaller and larger cracks and veins. Some samples are rich in radiolarian tests, which prominently occur within the micritic matrix. Tests are commonly filled by small pyrite aggregates or micro- to cryptocrystalline silica (Fig. 5A). Smaller cavities within the rock matrix are filled by microquartz and equant calcite spar to different degrees. Larger cavities up to 5 mm in diameter are much less frequent compared to the limestones from the other localities in this study. Large cavities are mostly filled by equant calcite spar, however, microquartz, fibrous microcrystalline silica, and megaquartz, all predating equant calcite cement, occur as well (Fig. 5B, C). The sequence in which the silica varieties formed is identical to that of silica phases in the limestone with worm tubes from the Lincoln Creek Formation. The outer rims of

the cavities are mostly lined by microquartz, followed by fibrous silica, which grades into megaquartz (Fig. 5C). Cracks and veins are predominantly filled by fibrous microcrystalline silica apart from equant calcite spar; they seem to be devoid of large megaquartz crystals (Fig. 5B). Several brachiopod shells have been replaced by microquartz (Fig. 5D). 4.1.4. Dwyka Group deposits The matrix of the Ganigobis Shale Member limestones is dominated by clotted micrite containing pyrite, detrital grains, and abundant radiolarians (Fig. 6A). The originally siliceous radiolarian tests have been replaced by calcite (Fig. 6B) or pyrite (Fig. 6C). The limestones are further characterized by large cavities filled by banded and botryoidal cement, making this the second most abundant mineral phase after micrite. Additional minor carbonate phases are yellow calcite, spheroidal calcite, and equant calcite spar. Yellow calcite is often intercalated with banded and botryoidal cement. Silica is present mainly in larger cavities and postdates the carbonate phases except for equant calcite cement. Its growth nucleated on micrite, banded and botryoidal cement, yellow calcite, or spheroidal calcite (Fig. 6D, E). Silica occurs as microquartz, fibrous microcrystalline silica, and megaquartz. Microquartz and fibrous microcrystalline silica are both prominent in large cavities, where they occur in close proximity (Fig. 6E). The paragenetic sequence of the silica phases is the same as in the other deposits of this study, with megaquartz representing the latest silica phase within former cavities (Fig. 6F). In places, microcrystalline fibrous silica replaced banded and botryoidal carbonate cement. In contrast, the megaquartz of the Ganigobis Shale Member limestones does not seem to have replaced carbonate cements. 4.2. Isotope geochemistry of carbonate phases and molecular fossils — constraints on the nature of the Knappton SSB limestone Individual carbonate phases of the Knappton SSB limestone were analyzed for stable carbon and oxygen isotope compositions (Fig. 7). Micrite revealed δ13C values between −7.8 and 1.0‰ and δ18O values

18

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

δ13Ccarbonate values alone. The bathymodiolin mussels preserved in the limestone are the same species that is found in other seep deposits throughout western Washington (Kiel and Amano, 2013), agreeing with the envisaged seep setting. In order to further constrain the composition of seepage fluids, one sample of the Knappton SSB limestone was analyzed for its lipid biomarker inventory. The hydrocarbon fraction of the seep limestone is dominated by a suite of n-alkanes with carbon skeletons ranging from C15 to C33. Yet, the overall most abundant compound is the regular isoprenoid phytane, accompanied by the isoprenoids pentamethylicosane (PMI) and biphytane. The low δ13C values of PMI (− 113‰) and biphytane (− 107‰) reveal that these molecular fossils derive from methanotrophic archaea involved in AOM (cf. Peckmann and Thiel, 2004). The circumstance that phytane (− 94‰) is only slightly less 13 C-depleted indicates that it predominantly derived from precursor lipids of methanotrophic archaea too, rather than from phototrophic organisms (cf. Birgel et al., 2008). The presence of AOM biomarkers confirms that the Knappton site seepage fluids contained methane. The recognition of AOM as the key biogeochemical process at the Knappton seep is of immediate significance for the considerations on silicification processes outlined below. Similarly, it has been shown that the fluids that formed the Dwyka Group and Graylock Butte deposits contained methane (Birgel et al., 2008; Himmler et al., 2008; Peckmann et al., 2011; Table 1). Based on the lack of biomarker data and the lowest δ13 Ccarbonate value of −23.1‰ (Table 1), the composition of the seepage fluids that formed the Lincoln Creek Formation deposit with worm tubes remains elusive. But given that other close-by limestone bodies of the Lincoln Creek Formation have been shown to represent methane-seep deposits (Peckmann et al., 2002), it seems likely that the worm tube deposit formed at a methane seep as well. 4.3. Numerical experiments — formation of authigenic silica as a consequence of anaerobic oxidation of methane

Fig. 4. Photomicrographs of the Canyon River seep deposit from the Lincoln Creek Formation (cf. Goedert et al., 2000), crossed-polarized light; m = micritic matrix, bbc = banded and botryoidal aragonite cement, fms = fibrous microcrystalline silica, miq = microquartz, mq = megaquartz. (A) Fibrous microcrystalline silica and megaquartz filling a crack within the micritic matrix. (B) Cavity filled by banded and botryoidal aragonite cement, fibrous microcrystalline silica, and microquartz. (C) Fibrous microcrystalline silica and megaquartz within a worm tube. White arrow denotes preserved delamination of the tube wall.

between −10.2 and −8.9‰. The δ13C values of yellow aragonite range from −26.6 to −23.2‰, δ18O values range from −6.9 to −6.6‰. Clear, fibrous aragonite forming banded and botryoidal crystal aggregates yielded δ13C values ranging from − 27.6 to − 12.3‰ and δ18O values ranging from −12.4 to −9.3‰. The δ13C values of equant calcite fall between 4.4 and 7.1‰, δ18O values fall between −11.7 to −11.1‰. The lowest δ13Ccarbonate value of −27.6‰ is not negative enough to exclude carbon sources other than methane in addition to marine carbonate (cf. Peckmann and Thiel, 2004). The large amount of 13C-depleted authigenic carbonate forming the Knappton SSB limestone is best explained by hydrocarbon seepage, but the composition of fluids – methane vs. oil-dominated – cannot be assessed with certainty based on

The purpose of the numerical experiments is to identify key processes and parameters that have been at work in the ancient seep system and affected silica stability. In order to relate the experiments accomplished here to the conditions that prevailed in the past, petrographic and geochemical results are compared to the outcome of the experiments, allowing us to ascertain the relevance of particular geochemical processes. The main goal is to attain a mechanistic understanding of the processes that led to silica mobilization and subsequent re-precipitation. In order to assess the sensitivity of environmental conditions in a fixed volume of sedimentary pore water over time to individual geochemical processes, a series of batch calculations was performed first. By comparing the outcome of numerical experiments that either included or excluded individual geochemical processes, we arrived at a better understanding of the sensitivity of the system toward these processes. 4.3.1. Batch reaction experiments The first experiment was conducted to investigate the effect of AOMgenerated alkalinity increase on the pH during calcite precipitation over a simulation period of one year. The reaction products of AOM, including the stoichiometric sodium to calcium ratio of seawater for charge balance, were added to solution in 20 consecutive steps. Simultaneously, calcite was allowed to precipitate. In the course of the experiment calcium concentration decreased in solution (Fig. 8A), whereas the amount of authigenic calcite increased in the solid phase assemblage (Fig. 8B). Both effects are caused by the precipitation of calcite from solution during the entire simulation time according to the equation: −

HCO3 þ Ca

2þ

þ

→CaCO3 þ H :

ð2Þ

The pH of the solution does not change significantly over a period of one year due to high alkalinity and carbonate precipitation efficiently

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

19

Fig. 5. Photomicrographs of the Graylock Butte seep deposits (cf. Peckmann et al., 2011); m = micritic matrix, bbc = banded and botryoidal cement, miq = microquartz, fms = fibrous microcrystalline silica, mq = megaquartz; ec = equant calcite cement. (A) Radiolarians (one specimen denoted by white arrow) within the micritic matrix, plane-polarized light. (B) Fibrous microcrystalline silica and microquartz filling a large vein surrounded by equant calcite, stained thin section, crossed-polarized light. (C) A large cavity filled by banded and botryoidal cement, microquartz, fibrous microcrystalline silica, and megaquartz crystals, crossed-polarized light. (D) Silicified brachiopod shell fragments (white arrow) within a micrite matrix, crossed-polarized light.

buffering the pH in the system (Fig. 8A, B). As calcium is continuously removed from solution, calcite precipitation declines during each reaction step and pH changes are small. This experiment shows the effect of heterogeneous reactions on solution composition and pH, considering reactions (1) and (2). In seep environments, calcium would be partially replenished by fluid flow and diffusion toward the zone of AOM, a process that is not considered in the closed batch system models, but is accounted for in the Reactive-transport experiments (see below). A potentially important heterogeneous equilibrium is the removal of carbon dioxide by degassing favored by the seepage itself. This process may become especially relevant at active seeps typified by advective transport, where large amounts of methane oversaturate the solution, resulting in the formation of gas bubbles. Torres et al. (2002) reported methane fluxes of up to 106 mmol/m2/d along thrust faults and up to 100 mmol/m2/d at bacterial mats. The amount of methane escaping into the water column is controlled by the amount of methane delivered and the efficiency of the methane-oxidizing microbial community to remove methane from pore waters (Reeburgh, 2007; Knittel and Boetius, 2009; Mogollón et al., 2009). The efficiency of microorganisms to oxidize methane can be as high as 70 to 80% (Wallmann et al., 2006) and may reach more than 90% in depositional environments controlled by diffusive transport (Niewöhner et al., 1998). However, if methane concentrations and fluxes are high enough and if methane transport occurs in the form of free gas, greater portions of methane may escape into the water column. If carbon dioxide cannot escape the sediment and is trapped, this may lead to microenvironments in which pH values remain constantly low. If, however, gaseous methane is transported rapidly through sediments along faults or conduits or even produces in situ brecciation, carbon dioxide may in fact be efficiently removed. Römer et al. (2012) reported up to 0.7 mol% carbon dioxide in gas phases from high-flux seep

areas in the Black Sea. The effect of carbon dioxide degassing into methane gas bubbles is shown in the reaction: þ

−

H þ HCO3 ↔H2 CO3 ↔CO2ðgÞ þ H2 O:

ð3Þ

According to this equation, the degassing of carbon dioxide from solution consumes protons and therefore results in a shift toward higher pH values. In our experiments, interactions with gas phases at seeps were simulated by introducing heterogeneous reactions (i.e., degassing) with a gas phase of a fixed composition in the batch calculation. In a next step, this degassing effect on local pH and calcite precipitation was tested. We allowed constant carbon dioxide degassing corresponding to concentrations determined by Römer et al. (2012) over the simulated period of one year. In the course of the experiment the effect of carbon dioxide degassing resulted in a pH increase from 7.3 to 8.3. Apparently, not only calcite precipitation, but also carbon dioxide degassing exerts a strong influence on solution compositions and pH values. 4.3.2. Reactive-transport experiments Limitations of the batch calculations are the accumulation of AOM reaction products (e.g., alkalinity) and the depletion of solutes by precipitation (e.g., calcium) within the reaction volume. Under natural conditions, advective and diffusive processes lead to the replenishment of reactants and the removal of products generated by AOM. Taking into account flow rates consistent with field observations (e.g., Torres et al., 2002) allows for a more realistic simulation of the geochemical environment in the reaction volume. Two numerical reaction transport experiments were performed, one with and one without carbon dioxide degassing (applied AOM rates are listed in the Appendix). The presence of carbon dioxide in methane gas at seeps (Römer et al., 2012) indicates that degassing of carbon dioxide occurs under high-flow conditions.

20

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

Fig. 6. Photomicrographs of the Dwyka Group seep deposits (cf. Himmler et al., 2008); cm = clotted micrite, bbc = banded and botryoidal cement, sc = spheroidal calcite, miq = microquartz, fms = fibrous microcrystalline silica, mq = megaquartz; ec = equant calcite cement. (A) Banded and botryoidal carbonate cement filling a small crack through clotted micrite; white arrow denotes one radiolarian fossil, plane-polarized light. (B) Radiolarian tests (arrows) replaced by calcite. (C) Radiolarian test replaced by pyrite. (D) Larger vein filled by fibrous microcrystalline silica and equant calcite surrounded by banded and botryoidal cement, crossed-polarized light. (E) Microquartz and fibrous microcrystalline silica enclosed by banded and botryoidal cement and spheroidal calcite, crossed-polarized light. (F) Large cavity filled by fibrous microcrystalline silica and megaquartz, crossed-polarized light.

However, in systems where low methane concentrations result in its complete oxidation, alkalinity increase is small and if hydrostatic pressure is high enough to inhibit the degassing of pure carbon dioxide, the pH of the system will not increase significantly as indicated by the numerical batch reaction experiments. These differences were further explored in numerical reactive-transport experiments (Fig. 9). AOM rates were increased stepwise over a period of eight years (Fig. 9A, Appendix), resulting in AOM rates 80 times higher than initial rates reported by Wallmann et al. (2006). Similarly high rates have been reported from incubation experiments with AOM enrichment cultures by Holler et al. (2011). The change of pH values varied from 0.6 to 1.2 units at moderate and high AOM rates in our experiments, respectively (Fig. 9A); the greatest pH increase occurred in the experiments with carbon dioxide degassing. This led to a pH maximum of 9.1 at the time of maximum AOM rates. Over the last four years of the experiment, AOM rates were lowered to zero, resulting in lower pH values close to values of the influent solution. The effect of pH on silica dissolution rates is of particular interest in this study, since AOM, carbonate precipitation, and particularly carbon dioxide degassing exert control over pore water pH as revealed by the

previous experiments. Silica concentrations in pore waters of marine sediments are controlled by the rate of silica dissolution (Van Cappellen and Qiu, 1997b), which in turn depends on the pH of the solution (Hurd, 1972; Crerar and Dove, 1990; Dove, 1994; Dove and Rimstidt, 1994; Van Cappellen and Qiu, 1997a; Schlüter and Rickert, 1998). Dissolution rates also depend on the reactive surface area of the mineral, which decreases with ongoing burial (Van Cappellen, 1996; Dixit and Van Cappellen, 2002). This aging process – the reduction of reactive surface sites with time – is not considered in our experiments. However, in order not to overestimate dissolution rates, a low surface area was assumed for biogenic silica, corresponding to a state where silica particles have already been modified in the water column and by early diagenetic processes prior to the onset of AOM-related geochemical changes (cf. Van Cappellen et al., 2002). Our experiments revealed that dissolution rates and concentrations of dissolved silica increase with increasing pH (Fig. 9B). At high rates of AOM and high dissolution rates, high ionic strength and high dissolved silica concentration lead to supersaturation of amorphous silica phases between years seven and nine (Fig. 9C). However, without carbon dioxide degassing, the solution is constantly under-saturated with respect to amorphous

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

21

the pH of interstitial waters. Degassing of carbon dioxide, on the other hand, is preceded by protonation of carbonate and bicarbonate ions and a concomitant pH increase. Although methane makes up the better part of gases escaping from modern cold seeps (over 95 mol%, Römer et al., 2012), gases emanating from active seeps contain small amounts of carbon dioxide (e.g., Pape et al., 2010). Similarly, Kim et al. (2012) reported carbon dioxide concentrations in headspace gases of up to 4.5 mol% from the Ulleung Basin, South Korea. The corresponding low carbon dioxide partial pressure in the gas phase provides a driving force for degassing. Local calcite precipitation in seep sediments adds protons to solution, reacting with bicarbonate and producing carbonic acid according to reaction (3). If during more vigorous seepage methane gas is being transported upward toward the sediment–water interface via advection, carbon dioxide gas can be purged efficiently from the system by methane degassing. Such a scenario is very plausible as studies have repeatedly shown that methane migrating at seeps is partly present as free gas (Flemings et al., 2003; Haeckel et al., 2004; Milkov et al., 2004; Torres et al., 2004, 2011; Liu and Flemings, 2006; Fischer et al., 2013; Römer et al., 2014). This process may be additionally enhanced by episodic release of free gas due to gas hydrate destabilization, which occurs frequently at seeps (e.g., Bohrmann et al., 1998; Berndt et al., 2014). The removal of carbon dioxide via gas seepage can consequently account for a significant local pH increase. It may also accelerate AOM-driven carbonate formation, as it increases the saturation state with respect to carbonate minerals (Mackenzie and Morse, 1990). In this context it is interesting to note that Pape et al. (2010) reported slight carbon dioxide enrichments in shallow gases at seeps in the eastern Black Sea; the gas composition was affected by AOM and carbonate precipitation. Although AOM is common in most marine sediments (Knittel and Boetius, 2009), interestingly, carbonate minerals do not precipitate everywhere and significant precipitation – in particular of aragonite – is only encountered during times of active methane seepage (e.g., Nöthen and Kasten, 2011).

Fig. 7. Cross plot showing stable carbon and oxygen isotope compositions of authigenic carbonate phases of the Knappton SSB limestone.

silica during the entire time of the simulation. Variations in pH also affect the amount of solid amorphous silica present in the solid-phase assemblage. When allowing for carbon dioxide degassing, the silica content is reduced to a greater extent than without degassing. 4.4. Implications of numerical experiments on silicification in methane-seep deposits 4.4.1. pH and carbon dioxide degassing The experiments confirmed that increasing alkalinity produced by AOM triggers carbonate precipitation (cf. Ritger et al., 1987). They also revealed that pH changes do occur without the removal of carbon dioxide, but given the required high increase in the rate of AOM to obtain this effect, AOM alone is likely to play a minor role in increasing Table 1 Lowest δ13C values of authigenic carbonate phases of the Knappton SSB (this study), Lincoln Creek Formation (Goedert et al., 2000), Graylock Butte (Peckmann et al., 2011), and Dwyka Group (Himmler et al., 2008) seep limestones. Site

Lowest δ13C values

Knappton Lincoln Creek Fm. Graylock Butte Dwyka Group

−27.6‰ −23.1‰ −35.6‰ −51.4‰

4.4.2. AOM and silica dissolution The influence of the advent, spatial shift, and cessation of AOM within seep sediments on silica transformation has been discussed by Kuechler et al. (2012). During the twelve-year period of our simulation the quantity of silica that was dissolved amounts to 0.077 and 0.049 mol with and without carbon dioxide degassing, respectively. This corresponds to dissolution of 3 and 2% of the total initial solid silica in the solid-phase assemblage, respectively. It would take approximately 412 years with and 640 years without carbon dioxide degassing to dissolve the complete solid amorphous silica pool with the applied AOM rates, ignoring the effect of decreasing dissolution rates with decreasing surface area (cf. Van Cappellen et al., 2002). In order to test the sensitivity of the system toward AOM rates, ten simulations were conducted in which AOM rates were kept constant. Each run was calculated with and without carbon dioxide degassing (Table 2). If carbon dioxide degassing in the course of seepage is eliminated, nearly twice as much silica is dissolved if high AOM rates are applied (311 mmol/m2/d; Table 2) compared to the absence of AOM. Likewise, the time required for total depletion of the solid silica pool within the sediment column is twice as long without AOM. However, if carbon dioxide degassing is allowed, the AOM rate exerts a relatively lesser control on the rate of silica dissolution (Table 2). Based on these experiments it appears that the system reacts more sensitive to the presence or absence of carbon dioxide than to the rate of alkalinity increase depending on the rate of AOM. 4.4.3. Silica diffusion and precipitation The numerical experiments revealed that silica dissolution occurs at higher rates at high pH, and high AOM rates produce locally increased concentrations of dissolved silica. Local carbon dioxide partial pressures exert an even stronger effect on pH and, subsequently, on local dissolved silica concentrations in the pore waters. The outcome of these effects is twofold, resulting in (1) a local increase of the saturation

22

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

Fig. 8. Results from batch reaction calculations. (A) pH and calcium in solution without carbon dioxide degassing and calcite precipitation. (B) Analog to (A) showing pH and calcite within the solid phase assemblage. (C) pH and calcium in solution with carbon dioxide degassing and calcite precipitation. (D) Analog to (C) showing pH and calcite within the solid phase assemblage.

state of silicate minerals and (2) a dissolved silica-concentration gradient away from the site of most intense AOM activity, referred to as the AOM hotspot. The input solution of the experiments used data acquired for pore-water samples collected during the RV METEOR cruise M76/3b (Zabel et al., 2008). During this cruise pore water concentrations of silica in the periphery of an active methane seep were determined (see Appendix for details). Dissolved silica concentrations in the surrounding sediments range from 0.3 to 0.6 mmol/L, whereas calculated dissolved silica concentrations allowing for degassing of carbon dioxide lie between 0.9 and 1.2 mmol/L at no and high AOM rates, respectively (Fig. 9B). This suggests that dissolved silica concentrations could be up to three times higher at the AOM hotspots than in peripheral sediments. We conclude that degassing of carbon dioxide and AOM produce local microenvironments where pH and dissolved silica concentrations are higher than in the surrounding sediments, resulting in diffusion of dissolved silica away from AOM hotspots and toward preexisting carbonate crusts should these be present. Subsequently, pore waters rich in dissolved silica enter the porous crust and precipitate as chalcedony and quartz (cf. Kuechler et al., 2012). 4.4.4. Silica diagenesis and phase transitions at methane seeps Most marine sediments contain amorphous silica or opal-A (Williams et al., 1985). During early diagenesis, this primary phase is transformed into a less soluble cristobalite–tridymite phase referred to as opal-CT (Jones and Sengit, 1971). Opal-CT subsequently transforms into chalcedony (disordered, microcrystalline, fibrous variety of quartz), leading to the formation of the thermodynamically most stable silica phase, quartz (Murata and Norman, 1976). These phase transformations occur via dissolution and re-precipitation reactions, which are

dependent on solubility, surface area, ionic strength of the solution, and the presence of detrital minerals in the sediment (Williams et al., 1985). The sequence of transformations is partly reflected in the paragenetic sequence of silica observed in the studied seep limestones. In sedimentary environments the transition from opal-A to opal-CT occurs when increasing burial causes temperatures to rise (Williams and Crerar, 1985; Knauth, 1994). Opal-A is believed to completely convert to opal-CT at temperatures as low as 18 °C (Pisciotto, 1981). The transition from opal-CT to quartz, although less well constrained, may proceed at temperatures between 30° and 165 °C (Murata and Larsen, 1975; Murata et al., 1977). The stable oxygen isotope compositions of the studied seep limestones suggest that all samples have been heated to temperatures well above 18 °C (cf. Goedert et al., 2000; Himmler et al., 2008; Peckmann et al., 2011; this study), explaining why amorphous opal-A and opal-CT have converted to chalcedony or microand megaquartz. Provided that all of the latter varieties of silica occur within a cavity, chalcedony always precedes megaquartz. In several cavities the walls are lined with chalcedony, which grades into quartz (Figs. 4B, C, 5C, 6F). Such textural changes have been interpreted to result from varying dissolved silica concentrations in the parent solution, rather than deviations in local temperature and pressure conditions (Heaney, 1993). Chalcedony is more soluble than quartz (Walther and Helgeson, 1977), and higher dissolved silica concentrations are required to precipitate it from solution than are required for quartz. The repeated occurrence of chalcedony within the seep deposits suggests that the parent solutions were highly supersaturated with respect to silica polymorphs, allowing chalcedony formation over quartz. As a consequence of chalcedony precipitation, the silica concentration is subsequently lowered to a point where chalcedony precipitation is no longer favored.

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

23

Fig. 9. Results from reactive transport calculations comparing parameters with and without the effect of carbon dioxide degassing over a time period of twelve years. Solid brown line in (A) shows the stepwise increase in AOM rates during reactive transport experiments (see Appendix for details). Note that in the last year of the experiment the AOM rate is set to zero. (A) pH change compared to varying AOM rates. (B) Dissolved silica in solution. (C) Saturation index of amorphous silica. (D) Amorphous silica within the solid phase assemblage. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Quartz may still precipitate from such solutions, however, its growth is extremely slow and produces large, drusy megaquartz crystals. Heaney (1993) described similar silica phase transitions in agates and also suggested that this change is related to a decreasing silica polymerization (i.e., decreasing dissolved silica concentrations) of the parent fluid due to chalcedony precipitation. The transition from chalcedony to megaquartz in ancient seep deposits is, thus, probably also related to decreasing silica concentrations in response to progressive carbonate crust formation, which decreases permeability and gradually impedes the flow of silica-rich diagenetic fluids through cracks and cavities.

Table 2 Results from numerical experiments with varying AOM rates and carbon dioxide degassing conditions over a time period of twelve years, showing total moles of amorphous silica dissolved, the amount of total amorphous silica dissolved in percent, and the time required for complete depletion of the solid amorphous silica phase assemblage; am = amorphous. AOM rates [mmol/m2/d]

0

31.1

62.2

311

CO2 degassing

Moles SiO2 (am) dissolved

0.03 0.062 1.29 2.35 928 510

0.041 0.066 1.56 2.53 770 474

0.046 0.07 1.74 2.64 687 454

0.057 0.08 2.18 3.10 548 389

No Yes No Yes No Yes

Percent of total SiO2 (am) Years to SiO2 (am) depletion

5. A conceptual model for silicification of seep limestones We put forward that the observed paragenetic sequence of the studied seep limestones is the result of the following consecutive steps: 1. AOM triggers precipitation of early diagenetic carbonate minerals. Calcite and aragonite precipitation dominates over silica precipitation at prevailing pH values (~7.5) and high alkalinity. 2. AOM and associated precipitation of carbonate minerals alone are insufficient to induce a significant pH increase of interstitial pore waters. Advective methane flux favors high rates of AOM and produces gas bubbles, which allows for the degassing of carbon dioxide resulting from carbonate precipitation. 3. Carbon dioxide removal leads to (1) a pH increase from approximately 7.5 to around 8.5 to 9, (2) an increase of dissolved silica concentration caused by the dissolution of biogenic silica (radiolarian tests, diatom frustules, sponge spicules) at high pH, and (3) enhanced carbonate precipitation. 4. Silica either diffuses from the AOM hotspot and site of bubble formation or is transported with advecting fluids into the surrounding sediments including voids within preexisting carbonate crusts where it precipitates along a gradient of decreasing pH. Silica minerals either replace methane-derived carbonate cements or precipitate as cements. Concentrations of dissolved silica decrease with time due to chalcedony precipitation and a lowered fluid flow caused by progressive cementation. As a result, quartz precipitates as last

24

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

Fig. 10. Cartoon illustrating the conceptual model of silicification, no scale implied. (A) Methane rich fluids rising from depth, locally producing gas bubbles. Thin black arrows show possible fluid migration paths around carbonate crusts formed as a result of AOM. SWI = sediment–water interface. (B) Close-up of the red rectangle in (A). Conditions during moderate methane seepage, AOM and carbonate precipitation. The latter buffers pH at low levels compared to seawater. The carbonate crust is rimmed by newly precipitated carbonate enclosing a cavity (C), which is partly filled by AOM-derived carbonate cement (pink). The center of the cavity (white area) remains unfilled. (C) Same scenario as in (B) during periods of high methane flux. Methane supersaturation leads to bubble formation, enabling degassing of carbon dioxide. Carbonate precipitation is enhanced and pH increases, leading to dissolution of silica phases and diffusion of dissolved silica. (D) Dissolved silica diffuses into the carbonate crust, precipitating as chalcedony and subsequently quartz. See text for further details. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

silica phase before equant calcite cement fills remaining porosity during burial diagenesis. The processes outlined above are illustrated in a schematic cartoon in Fig. 10, which depicts the evolution from carbonate precipitation to silica dissolution and subsequent precipitation of silica within cavities. Petrographic observations revealed that silicification occurred after the formation of AOM-derived carbonate had ceased and before late-stage equant calcite cement precipitated (Goedert et al., 2000; Himmler et al., 2008; Peckmann et al., 2011; this study). This implies that dissolved silica diffused into partially consolidated carbonate crusts in response to a concentration gradient of dissolved silica, or was transported by fluid advection. Silica mobilization is ultimately triggered by spatial and temporal changes in methane concentrations, whereby high concentrations favor the formation of gas bubbles, promoting carbon dioxide removal, and increase AOM rates. It needs to be stressed that gas bubble formation and high AOM rates are both dependent on high methane concentrations and are, thus, ultimately linked. The scenario is therefore only valid for seeps where not all methane is oxidized by the microbial consortium, allowing for supersaturation and gas bubble formation (cf. Martens and Klump, 1980). However, sediments containing significant quantities of free gas are indeed widespread in continental margin sediments (e.g., Fleischer et al., 2001; Judd, 2003; Fischer et al., 2013; Römer et al., 2014), and are often typified by carbonate authigenesis. It is, thus, very likely that free gas

formation and resultant methane and carbon dioxide ebullition significantly affect the precipitation of minerals at seeps, also triggering silica dissolution with subsequent diffusion and re-precipitation. Our experiments provide a conceptual model for the mechanisms of silicification and silica precipitation at seeps. It seems plausible that biogeochemical processes apart from AOM have the potential to induce silicification in non-seep environments, provided that these processes impact the pH of the system in a way that allows for the dissolution of pre-existing silica phases at high pH, the transport of dissolved silica, and the re-precipitation of silica in microenvironments typified by lower pH values. Acknowledgments We thank Susanne Gier (Vienna, Austria) for XRD analysis, Sylvain Richoz (Graz, Austria) for stable isotope analysis of carbonate samples, Beatrix Bethke (Vienna, Austria) for lipid extraction and sample preparation, Birgit Wild and Andreas Richter (both Vienna, Austria) for help measuring compound-specific stable carbon isotopes, as well as Thomas Pape (MARUM and Department of Geosciences, University of Bremen, Germany) for data and discussion on pore water geochemistry. We are grateful to André Gassner, Jens Gröger, and Jörn Tönnius (MARUM, University of Bremen, Germany) for support with pore water sampling and analysis onboard RV METEOR and Matthias Zabel (MARUM, University of Bremen) for providing sulfate pore water data, David

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

Parkhurst (Denver, USA) for advice concerning PHREEQC, and an anonymous referee for comments that helped improve the manuscript. We thank Green Diamond Resource Company (Shelton, Washington) for access to their land in the Canyon River area. Appendix. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.palaeo.2014.12.007. References Aloisi, G., Pierre, C., Rouchy, J.M., Foucher, J.P., Woodside, J., MEDINAUT Scientific Party, 2000. Methane-related authigenic carbonates of eastern Mediterranean Sea mud volcanoes and their possible relation to gas hydrate destabilisation. Earth Planet. Sci. Lett. 184, 321–338. Aloisi, G., Bouloubassi, I., Heijs, S.K., Pancost, R.D., Pierre, C., Sinninghe Damsté, J.S., Gottschal, J.C., Forney, L.J., Rouchy, J.M., 2002. CH4-consuming microorganisms and the formation of carbonate crusts at cold seeps. Earth Planet. Sci. Lett. 203, 195–203. Aloisi, G., Wallmann, K., Bollwerk, S.K., Derkachev, A., Bohrmann, G., Suess, E., 2003. The effect of dissolved barium on biogeochemical processes at cold seeps. Geochim. Cosmochim. Acta 68, 1735–1748. Amano, K., Kiel, S., 2007. Fossil vesicomyid bivalves from the North Pacific region. Veliger 49, 270–293. Amano, K., Jenkins, R.G., Sako, Y., Ohara, M., Kiel, S., 2013. A Paleogene deep-sea methaneseep community from Honshu, Japan. Palaeogeogr. Palaeoclimatol. Palaeoecol. 387, 126–133. Bangert, B., Stollhofen, H., Geiger, M., Lorenz, V., 2000. Fossil record and high-resolution tephrostratigraphy of Carboniferous glaciomarine mudstones, Dwyka Group, southern Namibia. Communications of the Geological Survey of Namibia 12, pp. 235–245. Berndt, C., Feseker, T., Treude, T., Krastel, S., Liebetrau, V., Niemann, H., Bertics, V.J., Dumke, I., Dünnbier, K., Ferré, B., Graves, C., Gross, F., Hissmann, K., Hühnerbach, V., Krause, S., Lieser, K., Schauer, J., Steinle, L., 2014. Temporal constraints on hydrate-controlled methane seepage off Svalbard. Science 343, 284–287. Birgel, D., Himmler, T., Freiwald, A., Peckmann, J., 2008. A new constraint on the antiquity of anaerobic oxidation of methane: Late Pennsylvanian seep limestones from southern Namibia. Geology 36, 543–546. Blome, C.D., Jones, D.L., Murchey, B.L., Liniecki, M., 1986. Geologic implications of radiolarian-bearing Paleozoic and Mesozoic rocks from the Blue Mountains Province, eastern Oregon. In: Vallier, T.L., Brooks, H.C. (Eds.), Geology of the Blue Mountains region of Oregon, Idaho, and Washington: geologic implications of Paleozoic and Mesozoic paleontology and biostratigraphy. Blue Mountains Province, Oregon and Idaho. U.S. Geological Survey Professional Papers 1435, pp. 79–101. Boetius, A., Ravenschlag, K., Schubert, C.J., Rickert, D., Widdel, F., Gieseke, A., Amann, R., Jørgensen, B.B., Witte, U., Pfannkuche, O., 2000. A marine microbial consortium apparently mediating anaerobic oxidation of methane. Nature 407, 623–626. Bohrmann, G., Greinert, J., Suess, E., Torres, M.E., 1998. Authigenic carbonates from the Cascadia subduction zone and their relation to gas hydrate stability. Geology 26, 647–650. Brandon, M.T., Calderwood, A.R., 1990. High-pressure metamorphism and uplift of the Olympic subduction complex. Geology 18, 1252–1255. Campbell, K.A., 2006. Hydrocarbon seep and hydrothermal vent paleoenvironments and paleontology: past developments and future research directions. Palaeogeogr. Palaeoclimatol. Palaeoecol. 232, 362–407. Crerar, D.A., Dove, P.M., 1990. Kinetics of quartz dissolution in electrolyte solutions using a hydrothermal mixed flow reactor. Chem. Geol. 84, 301–304. Dickinson, W.R., 1979. Mesozoic forearc basin in central Oregon. Geology 7, 166–170. Dickinson, W.R., Thayer, T.P., 1978. Paleogeographic and paleotectonic implications of Mesozoic stratigraphy and structure in the John Day Inlier of central Oregon. In: Howell, D.G., McDougal, K.A. (Eds.), Mesozoic Paleogeography of the Western United States. Society of Economic Paleontologists and Mineralogists, Pacific Section, Paleogeography Symposium 2, pp. 147–161. Dixit, S., Van Cappellen, P., 2002. Surface chemistry and reactivity of biogenic silica. Geochim. Cosmochim. Acta 66, 2559–2568. Dove, P.M., 1994. The dissolution kinetics of quartz in sodium chloride solutions at 25° to 300 °C. Am. J. Sci. 294, 665–712. Dove, P.M., Rimstidt, J.D., 1994. Silica–water interactions. In: Heaney, P.J., Prewitt, C., Gibbs, G.V. (Eds.), Reviews in Mineralogy vol. 29: Silica — Physical Behavior, Geochemistry and Materials Applications. Mineralogical Society of America, Washington, D.C., pp. 259–308. Fischer, D., Mogollón, J.M., Strasser, M., Pape, T., Bohrmann, G., Fekete, N., Spiess, V., Kasten, S., 2013. Subduction zone earthquake as potential trigger of submarine hydrocarbon seepage. Nat. Geosci. 6, 647–651. Fleischer, P., Orsi, T.H., Richardson, M.D., Anderson, A.L., 2001. Distribution of free gas in marine sediments: a global overview. Geo-Mar. Lett. 21, 103–122. Flemings, P.B., Liu, X., Winter, W.J., 2003. Critical pressure and multiphase flow in Blake Ridge gas hydrates. Geology 31, 1057–1060. Flörke, O.W., Graetsch, H., Martin, B., Röller, K., Wirth, R., 1991. Nomenclature of microand non-crystalline silica minerals, based on structure and microstructure. Neues Jb. Mineral. Abh. 163, 19–42. Flügel, E., 2004. Microfacies of Carbonate Rocks. Springer-Verlag, Berlin, Heidelberg (976 pp.).

25

Goedert, J.L., Benham, S.R., 2003. Biogeochemical processes at ancient methane seeps: the Bear River site in southwestern Washington. Geological Society of America, Field Guide 4 (8 pp.). Goedert, J.L., Peckmann, J., 2005. Corals from deep-water methane-seep deposits in Paleogene strata of Western Oregon and Washington, USA. In: Freiwald, A., Roberts, M. (Eds.), Cold-water Corals and Ecosystems. Springer, Berlin, Heidelberg, pp. 27–40. Goedert, J.L., Squires, R.L., 1993. First Oligocene record of Calyptogena (Bivalvia: Vesicomyidae). Veliger 36, 72–77. Goedert, J.L., Peckmann, J., Reitner, J., 2000. Worm tubes in an allochthonous cold-seep carbonate from lower Oligocene rocks of western Washington. J. Paleontol. 74, 992–999. Haas, A., Little, C.T.S., Sahling, H., Bohrmann, G., Himmler, T., Peckmann, J., 2009. Mineralization of vestimentiferan tubes at methane seeps on the Congo deep-sea fan. Deep-Sea Res. I 56, 283–293. Haeckel, M., Suess, E., Wallmann, K., Rickert, D., 2004. Rising methane gas bubbles form massive hydrate layers at the seafloor. Geochim. Cosmochim. Acta 68, 4335–4345. Heaney, P.J., 1993. A proposed mechanism for the growth of chalcedony. Contrib. Mineral. Petrol. 115, 66–74. Hikida, Y., Suzuki, S., Togo, Y., Ijiri, A., 2003. An exceptionally well-preserved fossil seep community from the Cretaceous Yezo Group in the Nakagawa area, Hokkaido, northern Japan. Paleontol. Res. 7, 329–342. Himmler, T., Freiwald, A., Stollhofen, H., Peckmann, J., 2008. Late Carboniferous hydrocarbon-seep carbonates from the glaciomarine Dwyka Group, southern Namibia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 257, 185–197. Hinrichs, K.-U., Hayes, J.M., Sylva, S.P., Brewert, P.G., DeLong, E.F., 1999. Methaneconsuming archaebacteria in marine sediments. Nature 398, 802–805. Holler, T., Wegener, G., Niemann, H., Deusner, C., Ferdelman, T.G., Boetius, A., Brunner, B., Widdel, F., 2011. Carbon and sulfur back flux during anaerobic microbial oxidation of methane and coupled sulfate reduction. Proc. Natl. Acad. Sci. U. S. A. 108, 1484–1490. Hryniewicz, K., Hammer, Ø., Nakrem, H.A., Little, C.T.S., 2012. Microfacies of the Volgian– Ryazanian (Jurassic–Cretaceous) hydrocarbon seep carbonates from Sassenfjorden, central Spitsbergen, Svalbard. Nor. J. Geol. 92, 113–131. Hurd, D.C., 1972. Factors affecting solution rate of biogenic opal in seawater. Earth Planet. Sci. Lett. 15, 411–417. Jones, J.B., Sengit, E.R., 1971. The nature of opal, I. Nomenclature and constituent phases. J. Geol. Soc. Aust. 18, 57–68. Judd, A.G., 2003. The global importance and context of methane escape from the seabed. Geo-Mar. Lett. 23, 147–154. Judd, A.G., Hovland, M., 2007. Seabed Fluid Flow — The Impact of Geology, Biology and the Marine Environment. Cambridge University Press (492 pp.). Kaim, A., Jenkins, R.G., Warén, A., 2008. Provannid and provannid-like gastropods from Late Cretaceous cold seeps of Hokkaido (Japan) and the fossil record of the Provannidae (Gastropoda: Abyssochrysoidea). Zool. J. Linnean Soc. 154, 421–436. Kasten, S., Nöthen, K., Hensen, C., Spieß, V., Blumenberg, M., Schneider, R.R., 2012. Gas hydrate decomposition recorded by authigenic barite precipitates at pockmark sites of the Northern Congo Fan. Geo-Mar. Lett. 32, 515–524. Kiel, S., 2010a. An Eldorado for paleontologists: the Cenozoic seeps of western Washington State, USA. In: Kiel, S. (Ed.), Vent and Seep Biota: Aspects from Microbes to, Ecosystems, pp. 433–448. Kiel, S., 2010b. On the potential generality of depth-related ecologic structure in cold-seep communities: evidence from Cenozoic and Mesozoic examples. Palaeogeogr. Palaeoclimatol. Palaeoecol. 295, 245–257. Kiel, S., Amano, K., 2013. The earliest bathymodiolin mussels: an evaluation of Eocene and Oligocene taxa from deep-sea methane seep deposits in western Washington State, USA. J. Paleontol. 87, 589–602. Kiel, S., Dando, P.R., 2009. Chaetopterid tubes from vent and seep sites: implications for fossil record and evolutionary history of vent and seep annelids. Acta Palaeontol. Pol. 54, 443–448. Kiel, S., Birgel, D., Campbell, K.A., Crampton, J.S., Schiøler, P., Peckmann, J., 2013. Cretaceous methane-seep deposits from New Zealand and their fauna. Palaeogeogr. Palaeoclimatol. Palaeoecol. 390, 17–34. Kim, H.J., Torres, M.E., Choi, J., Bakh, J.J., Park, H.M., Hong, L.W., 2012. Inferences on gas transport based on molecular and isotopic signatures of gases at acoustic chimneys and background sites in the Ulleung Basin. Org. Geochem. 43, 26–38. Knauth, P.L., 1994. Petrogenesis of chert. In: Heaney, P.J., Prewitt, C.T., Gibbs, G.V. (Eds.), Reviews in Mineralogy vol. 29: Silica — Physical Behavior, Geochemistry and Materials Applications. Mineralogical Society of America, Washington, D.C., pp. 233–258. Knittel, K., Boetius, A., 2009. Anaerobic oxidation of methane: progress with an unknown process. Annu. Rev. Microbiol. 63, 311–334. Kuechler, R.R., Birgel, D., Kiel, S., Freiwald, A., Goedert, J.L., Thiel, V., Peckmann, J., 2012. Miocene methane-derived carbonates from southwestern Washington, USA and a model for silicification at seeps. Lethaia 45, 259–273. Liu, X., Flemings, P.B., 2006. Passing gas through the hydrate stability zone at southern Hydrate Ridge, offshore Oregon. Earth Planet. Sci. Lett. 241, 211–226. MacKenzie, F.T., Morse, J.W., 1990. Geochemistry of sedimentary carbonates. Dev. Sedimentol. 48, 724. Martens, C.S., Klump, J.V., 1980. Biogeochemical cycling in an organic rich coastal marine basin. 1. Methane sediment–water exchange processes. Geochim. Cosmochim. Acta 44, 471–490. Martin, H., Wilczewski, N., 1970. Paleoecology, conditions of deposition and the paleogeography of the marine Dwyka Beds of South West Africa. In: Haughton, S.H. (Ed.), International Gondwana Symposium, 2nd IUGS Gondwana Symposium. Pretoria, pp. 621–626.

26

D. Smrzka et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 420 (2015) 13–26

Milkov, A.V., Dickens, G.R., Claypool, G.E., Lee, Y.J., Borowski, W.S., Torres, M.E., Xu, W., Tomaru, H., Tréhu, A.M., Schultheiss, P., 2004. Co-existence of gas hydrate, free gas and brine within the regional gas hydrate stability zone at Hydrate Ridge (Oregon margin): evidence from prolonged degassing of a pressurized core. Earth Planet. Sci. Lett. 222, 829–843. Mogollón, J.M., L'Heureux, I., Dale, A.W., Regnier, P., 2009. Methane gas-phase dynamics in marine sediments: a model study. Am. J. Sci. 309, 189–220. Murata, K.J., Larsen, R.R., 1975. Diagenesis of Miocene siliceous shales, Temblor Range, California. J. Res. U. S. Geol. Surv. 3, 553–566. Murata, K.J., Norman, M.B., 1976. An index of crystallinity for quartz. Am. J. Sci. 276, 1120–1130. Murata, K.J., Friedman, I., Gleason, J.D., 1977. Oxygen isotope relations between diagenetic silica minerals in Monterey Shale, Temblor Range, California. Am. J. Sci. 277, 259–272. Nakrem, H.A., Kiessling, W., 2012. Late Jurassic (Volgian) radiolarians from Central Spitsbergen — a preliminary study. Nor. J. Geol. 92, 149–155. Nesbitt, E.A., Martin, R.A., Campbell, K.A., 2013. New records of Oligocene diffuse hydrocarbon seeps, northern Cascadia margin. Palaeogeogr. Palaeoclimatol. Palaeoecol. 390, 116–129. Niewöhner, C., Hensen, C., Kasten, S., Zabel, M., Schulz, H.D., 1998. Deep sulfate reduction completely mediated by anaerobic methane oxidation in sediments of the upwelling area off Namibia. Geochim. Cosmochim. Acta 62, 455–464. Nöthen, K., Kasten, S., 2011. Reconstructing changes in seep activity by means of pore water and solid phase Sr/Ca and Mg/Ca ratios in pockmark sediments of the Northern Congo Fan. Mar. Geol. 287, 1–13. Pape, T., Bahr, A., Rethemeyer, J., Kessler, J.D., Sahling, H., Hinrichs, K.-U., Klapp, S.A., Reeburgh, W.S., Bohrmann, G., 2010. Molecular and isotopic partitioning of lowmolecular-weight hydrocarbons during migration and gas hydrate precipitation in deposits of a high-flux seepage site. Chem. Geol. 269, 350–363. Parkhurst, D.L., Appelo, C.A.J., 1999. A User'S Guide to PHREEQC (version 2) — A Computer Program for Speciation, Batch-reaction, One-dimensional Transport and Inverse Geochemical Calculations. U.S. Geological Survey, Denver Colorado (326 pp.). Peckmann, J., Thiel, V., 2004. Carbon cycling at ancient methane-seeps. Chem. Geol. 205, 443–467. Peckmann, J., Goedert, J.L., Thiel, V., Michaelis, W., Reitner, J., 2002. A comprehensive approach to the study of methane-seep deposits from the Lincoln Creek Formation, western Washington State, USA. Sedimentology 49, 855–873. Peckmann, J., Goedert, J.L., Heinrichs, T., Hoefs, J., Reitner, J., 2003. The Late Eocene “Whiskey Creek” methane-seep deposit (western Washington State). Part II: petrology, stable isotopes, and biogeochemistry. Facies 48, 241–254. Peckmann, J., Little, C.T.S., Gill, F., Reitner, J., 2005. Worm tube fossils from the Hollard Mound hydrocarbon-seep deposit, Middle Devonian, Morocco: Palaeozoic seeprelated vestimentiferans? Palaeogeogr. Palaeoclimatol. Palaeoecol. 227, 242–257. Peckmann, J., Kiel, S., Sandy, M.R., Taylor, D.G., Goedert, J.L., 2011. Mass occurrences of the brachiopod Halorella in Late Triassic methane-seep deposits, eastern Oregon. J. Geol. 119, 207–220. Pessagno Jr., E.A., Whalen, P.A., 1982. Lower and Middle Jurassic Radiolaria (multicytrid Nassellariina) from California, east-central Oregon and the Queen Charlotte Islands, B.C. Micropaleontology 28, 111–169. Pessagno Jr., E.A., Finch, W., Abbott, P.L., 1979. Upper Triassic Radiolaria from the San Hipólito Formation, Baja, California. Micropaleontology 25, 160–197. Pisciotto, K.A., 1981. Diagenetic trends in the siliceous facies of the Monterey Shale in the Santa Maria region, California. Sedimentology 28, 547–571. Prothero, D.R., Armentrout, J.M., 1985. Magnetostratigraphic correlation of the Lincoln Creek Formation, Washington: implications for the age of the Eocene/Oligocene boundary. Geology 13, 208–211. Rau, W.W., 1966. Stratigraphy and Foraminifera of the Satsop River Area, southern Olympic Peninsula, Washington. State of Washington Division of Mines and Geology Bulletin 53 (66 pp.). Reeburgh, W.S., 2007. Oceanic methane biogeochemistry. Chem. Rev. 107, 486–513. Rigby, J.K., Goedert, J.L., 1996. Fossil sponges from a localized cold-seep limestone in Oligocene rocks of the Olympic Peninsula, Washington. J. Paleontol. 70, 900–908. Ritger, S., Carson, B., Suess, E., 1987. Methane-derived authigenic carbonates formed by subduction-induced pore-water expulsion along the Oregon/Washington margin. Bull. Geol. Soc. Am. 98, 147–156.

Römer, M., Sahling, H., Pape, T., Bahr, A., Feseker, T., Wintersteller, P., Bohrmann, G., 2012. Geological control and magnitude of methane ebullition from a high-flux seep area in the Black Sea — the Kerch seep area. Mar. Geol. 319–322, 57–74. Römer, M., Torres, M., Kasten, S., Kuhn, G., Graham, A.G.C., Mau, S., Little, C.T.S., Linse, K., Pape, T., Geprägs, P., Fischer, D., Wintersteller, P., Marcon, Y., Rethemeyer, J., Bohrmann, G., Shipboard scientific party ANT-XXIX/4, 2014. First evidence of widespread active methane seepage in the Southern Ocean, off the sub-Antarctic island of South Georgia. Earth Planet. Sci. Lett. 403, 166–177. Schlüter, M., Rickert, D., 1998. Effect of pH on the measurement of biogenic silica. Mar. Chem. 63, 81–92. Stewart, R., Brandon, M., 2004. Detrital-zircon fission-track ages for the “Hoh Formation”: implications for late Cenozoic evolution of the Cascadia subduction wedge. Geol. Soc. Am. Bull. 116, 60–75. Stollhofen, H., 1999. Karoo synrift-deposition and its tectonic control at the evolving continental margin of Namibia. Z. Dtsch. Geol. Ges. 149, 519–632. Torres, M.E., McManus, J., Hammond, D.E., de Angelis, M.A., Heeschen, K.U., Colbert, S.L., Tryon, M.D., Brown, K.M., Suess, E., 2002. Fluid and chemical fluxes in and out of sediments hosting methane hydrate deposits on Hydrate Ridge, OR, I: hydrological provinces. Earth Planet. Sci. Lett. 201, 525–540. Torres, M.E., Bohrmann, G., Dubé, T.E., Poole, F.G., 2003. Formation of modern and Paleozoic stratiform barite at cold methane seeps on continental margins. Geology 31, 897–900. Torres, M.E., Wallmann, K., Tréhu, A.M., Bohrmann, G., Borowski, W.S., Tomaru, H., 2004. Gas hydrate growth, methane transport, and chloride enrichment at the southern summit of Hydrate Ridge, Cascadia margin off Oregon. Earth Planet. Sci. Lett. 226, 225–241. Torres, M.E., Kim, J.H., Choi, J., Ryu, B.J., Bahk, J.J., Riedel, M., Collett, T.S., Hong, W.L., Kastner, M., 2011. Occurrence of high salinity fluids associated with massive nearseafloor gas hydrate deposits. Proceedings of the 7th International Conference on Gas Hydrates (ICGH2011), Edinburgh, Scotland, United Kingdom (65 pp.). Van Cappellen, P., 1996. Reactive surface area control of the dissolution kinetics of biogenic silica in deep-sea sediments. Chem. Geol. 132, 125–130. Van Cappellen, P., Qiu, L., 1997a. Biogenic silica dissolution in sediments of the Southern Ocean. I. Solubility. Deep-Sea Res. II 44, 1109–1128. Van Cappellen, P., Qiu, L., 1997b. Biogenic silica dissolution in sediments of the Southern Ocean. II. Kinetics. Deep-Sea Res. II 44, 1129–1149. Van Cappellen, P., Dixit, S., van Beusekom, J., 2002. Biogenic silica dissolution in the oceans: reconciling experimental and field-based dissolution rates. Glob. Biogeochem. Cycles 16, 1075. http://dx.doi.org/10.1029/2001GB001431. Visser, J.N.J., 1997. Deglaciation sequences in the Permo-Carboniferous Karoo and Kalahari Basins of southern Africa: a tool in the analysis of cyclic glaciomarine basin fills. Sedimentology 44, 507–521. Wallmann, K., Linke, P., Suess, E., Bohrmann, G., Sahling, H., Schlüter, M., Dählmann, A., Lammers, S., Greinert, J., Von Mirbach, N., 1997. Quantifying fluid flow, solute mixing, and biogeochemical turnover at cold vents of the eastern Aleutian subduction zone. Geochim. Cosmochim. Acta 61, 5209–5219. Wallmann, K., Drews, M., Aloisi, G., Bohrmann, G., 2006. Methane discharge into the Black Sea and the global ocean via fluid flow through submarine mud volcanoes. Earth Planet. Sci. Lett. 248, 544–559. Walther, J.V., Helgeson, H.C., 1977. Calculation of the thermodynamic properties of aqueous silica and the solubility of quartz and its polymorphs at high pressures and temperatures. Am. J. Sci. 277, 1315–1351. Wells, R.E., 1989. Geologic map of the Cape Disappointment–Naselle River area, Pacific and Wahkiakum counties, Washington. United States Geological Survey Miscellaneous Investigations Series Map I-1832. Williams, L.A., Crerar, D.A., 1985. Silica diagenesis: II. General mechanisms. J. Sediment. Petrol. 55, 312–321. Williams, L.A., Parks, G.A., Crerar, D.A., 1985. Silica diagenesis: I. Solubility controls. J. Sediment. Petrol. 55, 301–311. Zabel, M., Boetius, A., Emeis, K.C., Ferdelman, T.G., Spieß, V., 2008. RV METEOR Cruise Report M76/L1, M76/L2, M76/L3a + b, PROSA Process Studies in the Eastern South Atlantic. DFG Senatskommission für Ozeanographie (223 pp.).