Conceptual Models for Burrow-Related, Selective

Geobiology (2004), 2, 21–30

ConceptualBlackwellORIGINAL Publishing, models ARTICLE for Ltd. burrow-related dolomitization models for burrow-related, selective dolomitization with textural and isotopic evidence from the Tyndall Stone, Canada MURRAY K. GINGRAS, S. GEORGE PEMBERTON, KARLIS MUELENBACHS AND HANS MACHEL Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta, T6G 2E3, Canada

ABSTRACT

The formation of dolomite is generally explained using models that reflect larger-scale processes that describe the relationship between the supply and transport of Mg, and geochemical conditions that are amenable to the formation of dolomite. However, heterogeneities in the substrate, such as those made by bioturbating infauna, may play a more important role in dolomitization than has been previously considered. Burrow-facilitated dolomitization is evident in the Ordovician Tyndall Stone (Red River Group, Selkirk Formation) of central Canada. The diagenetic fabrics present are attributed to dolomitizing fluids that both flowed through and evolved within burrow networks. Petrographic analysis suggests that two phases of dolomite formation took place. The first formed a fine-grained, fabric-destructive type that probably accompanied early burial; the second is a fine- to medium-grained, locally sucrosic dolomite that is interpreted to have precipitated during later burial. Isotopic analysis supports the proposed paragenetic history: (1) an apparent linking of the stable isotopes 13C and 18O strongly suggests that the micrite matrix formed during very early diagenesis and was derived from seawater; (2) the initial phase of dolomitization is potentially microbially mediated, as evidenced by the enrichment of 13C; and (3) isotopic values for the second generation of dolomite reflect the mixing of ground water and resorbed early dolomite. This paper conceptualizes the physical and chemical conditions required for the formation of dolomite in association with burrow fabrics. The proposed model reveals a composite of biological and inorganic reactions that demonstrates the interdependence of sediment fabric, organic content and microbial interactions in the development of burrow-mottled dolomitic limestone. It is suggested that where burrow-associated dolomite occurs, it is most likely to develop in two stages: first, the byproducts of the degradation of organic material in burrows locally increase the permeability and porosity around burrow fabrics in shallow diagenetic depositional environments; and, second, the passing of burrowed media into deeper dysaerobic sediment is accompanied by the establishment of fermenting micro-organisms whose byproducts mediate dolomitization.

Received 17 November 2003; accepted 29 January 2004

Corresponding author: Dr Murray Gingras. E-mail: [email protected]

example of this is the Ordovician Tyndall Stone of the Red INTRODUCTION River Group in Manitoba, Canada: Tyndall Stone is discussed As discrete structures in calcareous strata, burrows effect in detail below. changes in many of the physical parameters manifested in Many models exist for the large-scale processes required to carbonate strata. These include permeability (Gingras et al., form dolomite in carbonate strata. These include seepage/ 1999), porosity, fabric and texture. Such changes are reflux, freshwater/seawater mixing, various burial settings, attributed to dissolution and precipitation of sedimentary sabkha dolomite and the Coorong model (e.g. Machel & minerals within and adjacent to trace fossils. Thus, Mountjoy, 1986; Morrow, 1990). On a much smaller scale, diagenetically altered zones commonly surround burrows in there are physicochemical heterogeneities that appear to calcareous – and siliceous – media. Dolomitization, for influence the distribution of ‘patchy’ dolomite. Variability on instance, can occur in diagenetic halos several millimetres in the centimetre scale is strongly influenced by the activities thickness around and within the burrows. An excellent of bioturbating infauna.

22 M. K. GINGRAS et al.

Many studies have suggested that primary fabrics and Kendall, 1977; Morrow, 1978; Jones et al., 1979; Chow & textural properties strongly influence common dolomitization Longstaffe, 1995). The objective of this paper is to discuss this patterns observed in the rock record (Beales, 1953; Murray & premise using the example of a bioturbated carbonate wackes- Lucia, 1967; Kendall, 1977; Morrow, 1978). As diagenesis tone, the Ordovician Tyndall Stone of Manitoba Canada. proceeds, the original texture of the rock deteriorates and becomes subordinate to the diagenetic texture; however, at GEOLOGICAL AND STRATIGRAPHIC SETTING the onset of diagenesis the primary sedimentary character of the deposit must exert a control on the patterns of remineral- Tyndall Stone is a dolomite-mottled Ordovician limestone ization and dissolution. The above studies demonstrate the that is quarried in Manitoba, Canada. The formal designation potential of determining dolomite distribution by understand- of this unit is the Selkirk Member of the Red River Formation ing rock selectivity and thereby suggest that early dolomitiza- (Cowan, 1971; Kendall, 1977). Shallow drilling indicates that tion can be related to the development of the initial porosity/ the mottled limestone occurs in the lower half of the Selkirk permeability network. Contrasts between the burrow fill and Member (Kendall, 1977). Similarly mottled limestone comprises matrix potentially develop notable anisotropic permeability the subsurface equivalent Yeoman Formation, which is present that can be exploited by dolomitizing fluids (Beales, 1953; in south-eastern Saskatchewan (Fig. 1; Kendall, 1975). Both

Fig. 1 Schematic summary of the stratigraphic relationships associated with the Red River Formation.

Conceptual models for burrow-related dolomitization 23 the Selkirk and Yeoman Formations are overlain by evaporitic interpenetrating Thalassinoides-like burrows (Fig. 2A). Upon dolomicrites that are in turn overlain by evaporites (Kendall, close scrutiny it is apparent that the Thalassinoides appearance 1975, 1977). is normally due to a dolomitic halo, 2–10 mm thick, Selkirk and Yeoman Formation strata were deposited as an encompassing causative burrows 2–3 mm in diameter (locally epicontinental carbonate platform in the Williston Basin. Palaeophycus and Chondrites; Fig. 2A). The dolomite is typi- Their distinct sedimentological and ichnological signature are cally fine grained, although sucrosic crystals may be adjacent reminiscent of other Palaeozoic epicontinental carbonates, to the causative burrows. Many of the burrows have been most notably the Devonian Palliser Formation of the Western internally filled by calcite cement, which forms an interlocking Canadian Sedimentary Basin (Beales, 1953), and the Upper crystal mosaic (Fig. 2A,D). Staining from the original organic Ordovician Bighorn Facies from Williston Basin deposits in material is locally present at the causative burrow margins. the USA (Zenger, 1996a,b). Kendall (1977) noted that the degree of dolomitic mottling varies locally, but is typically confined to areas adjacent to burrows. Less commonly, dolomite has destructively replaced METHODS aragonitic allochems and infilled rare, small vugs (Fig. 2F). The lithological database consists of 15 samples from outcrop Petrographic analysis reveals that four different forms of near Tyndall, Manitoba, and eight core fragments from a calcite or dolomite are commonly observed in the Tyndall single well penetration in southern Saskatchewan. All samples Stone. These include: micrite (C1), or microcrystalline calcite were thin sectioned and stained with Alizarin red. (Fig. 2B); very fine-grained dolomite (D1), characterized by Samples for isotopic analysis were extracted from both out- subplanar subhedral crystals that form an idiotopic mosaic crop and subsurface datasets. Each sample was taken from a texture with calcite allochems dispersed throughout – D1 discrete (dolomitized) burrow-mottle and the calcite matrix domains in C1 lend the mottled appearance to the Tyndall adjacent to it. Where practical, 20–50-mg isotope samples Stone (Fig. 2C); fine- to medium-grained euhedral dolomite were extracted using small, high-speed rotary saws. Small areas (D2), which forms isopachous to undulatory rinds in and isolated within and near dolomitized trace fossils were sampled adjacent to the causative burrow structure, and within rare, using 1-mm-diameter coring bits: the drilled samples typically millimetre-scale vugs (Fig. 2D); and a calcite mosaic cement weighed between 1 and 3 mg. (C2), infilling allochems, open burrows and voids (Fig. 2E). The samples were subsequently prepared for analysis by reacting them with 100% H PO at 25 °C using the technique 3 4 Isotopic data of Epstein et al. (1964). The isotopic composition of calcite and dolomite were estimated by collecting and analysing The isotopic dataset is shown in Fig. 3. Micrite C1 plots a crudely evolved CO2 after reaction times of 1 h and 1 week, respec- linear trend, whereas calcite C2 generates a slightly enriched tively. Isotopic analyses were made at the University of Alberta cluster (Fig. 3). Isotopic ratios from C1 range between δ13C with a Finnigan-MAT 252 mass spectrometer operated in the of −2.2‰ and δ18O of −10.5‰ to δ13C of 0.7‰ and δ18O of dual inlet mode. −7.8‰. By contrast, isotope values for C2 are constrained within a δ13C range of −0.1 to 0.9‰, and δ18O values of −9.1 to −6.50‰. Two datapoints for calcite C2 fall well within the PETROGRAPHY AND ISOTOPIC ANALYSIS OF isotopic field occupied by dolomites D1 and D2: these are TYNDALL STONE denoted as C2x on Fig. 3 and discussed below. Dolomites D1 and D2 exhibit δ13C values of −1.4 to 0.8‰: δ18O is somewhat Lithological description more variable (between −10.4 and −7.3‰). The Tyndall Stone is a fossiliferous, cream-coloured limestone Potentially significant trends include: (1) the micrite (C1) with buff-coloured dolomitic mottling. The matrix of this unit trend, which is almost linear; (2) D1, D2 and C2 are typically typically consists of a crinoid–bryozoan–brachiopod wackes- enriched in 13C compared with C1; (3) with the exception of tone, locally grading to mudstone. Sharp-edged patches of one data point, dolomitic cements show less variability in δ13C crinoid–brachiopod packestone are present. Large fossils com- than in δ18O; and (4) a broader scattering of values for calcite prise abundant gastropods, cephalopods, corals, receptaculites mosaic C2. and stromatoporoids. Cemented mud, in situ precipitate and/ or recrystallized matrix is present as micrite. Most of the INTERPRETATION large fossils are calcite. Many of the fossils are micritized or have micritized rinds. Allochems are generally not abraded, Paragenesis although disarticulation is common. Larger allochems are typically intact and unabraded. The first stage of the paragenetic sequence is represented by The most distinctive feature of the Tyndall Stone is a C1 (Fig. 2B). Its association with micrite envelopes around ramifying burrowed pattern, which consists of branching and allochems suggests that this calcite developed during early

24 M. K. GINGRAS et al.

Fig. 2 Cement types in the Tyndall Stone. (A) Stained (Alizarin red) thin section in which the burrow mottling is unstained (dolomite D1 = lighter grey) and the matrix is stained (micrite C1 = darker grey). Causative burrows are present near the middle of most dolomite patches (white arrows). A black arrow shows the locally gradational boundary between C1 and D1. (B) Micrite C1 in wackestone part of matrix. (C) Dolomite D1 is a fine-grained dolomite, comprised of subplanar/ subhedral crystals forming a idiotopic mosaic texture. This cement gives the Tyndall Stone its mottled appearance. (D) Fine- to medium-grained euhedral dolomite (D2), which develops a sucrosic texture and is present as isopachous to undulatory rind adjacent to, and inside, the causative burrow structure. Also present is a calcite mosaic cement (C2), which infills dissolved allochems, open burrows and voids. (E) A calcite crinoid ossicle present in a dolomitized zone. This indicates that dolomite patches do not represent burrow fill, but are best explained as diagenetic halos. (F) White arrows indicate crinoid fragments that are partially dissolved at the margin of the dolomite halo. (G) Partially dissolved allochem in dolomitized zone (white arrow) and small vugs rimmed with sucrosic dolomite D2 (thin bright zone indicated with black arrows).

Conceptual models for burrow-related dolomitization 25

Fig. 3 Analytical isotopic data. Fields for C1 and C2 are shown. Dolomites D1 and D2 are plotted in the same field but are discriminated from each other by symbols (see key for explanation). Isotope ratios for cements sampled next to each other are joined with a thin line. Two outlying values for C2 are not included in the C2 zone: these are labelled as C2x. diagenesis in a shallow marine environment, possibly penecon- networks. Dolomite D1, which cross-cuts C1 and may impinge temporaneous with deposition. Isotopic values (Fig. 3) are on allochem boundaries (Fig. 2F), at least locally replaced consistent with other Ordovician limestones (Railsback et al., original calcite/aragonite C1. Dolomite D1 is also present in 1990; Hudson, 1977; Tobin & Walker, 1997). The correla- the calcite matrix as isolated or clustered subhedral crystals tion (or trend) of δ13C and δ18O may indicate a response to that are interpreted to have formed concurrently with the slight changes in the temperature of the overlaying water column pervasively dolomitized (burrow) zone. at the time of deposition, may be linked to depositional The relative enrichment of δ13C in D1 is consistent with bathymetry or could represent a recrystallization trend that is dolomite D1 representing a biogenically mediated phase. δ13C in equilibrium with marine-derived porewater. enrichment can be attributed to dolomite formation in a Partial dissolution of various allochems marginal to dolo- geochemical environment influenced by the byproducts of mite patches (Fig. 2F,G) and the presence of locally open fermentation (Reitsema, 1980; Dimitrakopoulos & Muehlen- tubes within the original bioturbation structures suggest that bachs, 1987). This probably occurred as the sediment passed porosity and permeability of C1 were enhanced adjacent to further into the historical layers, below the levels of sulphate burrow fabrics. This is attributed to the presence of high CO2 reduction – or at least as sulphate reduction waned. Notably, concentrations associated with the oxidation of organic mate- the amount of 13C enrichment is small: this is explained if the rial present in the matrix and concentrated in the burrows. It biological precipitation of dolomite occurs only in the earliest is important to note, however, that the burrows themselves stages of dolomite formation, providing a substrate for ensuing may have initially contained burrow fills characterized by dolomitization. Moreover, it is difficult to explain the variance higher permeability than the surrounding matrix. between the dolomite and calcite cements as due to isotope The primary phase of dolomitization (fine grained, D1) is fractionation. Figure 3 shows connected datapoints that concentrated around the fluid conduits made available by the represent cements sampled adjacent to each other (2–5 mm higher permeabilities present in association with the burrow apart). If fractionation were the primary cause of isotope

26 M. K. GINGRAS et al. enrichment in the dolomite, the δ13C ratios for dolomite DISCUSSION cement should be consistently linked with δ13C indicated for the sample’s associated calcite. Finally, with D2 no clear sea- Biogenic sedimentary structures (ichnofossils) alter sediment water or equilibrium recrystallization trend is evident. Similar, substrates in various ways. In general, these include physical disassociated δ13C and δ18O trends have been attributed to manipulation and chemical modification. Open burrows, early dolomitization from anoxic, interstitial porewaters or in particular, act as open conduits to the sediment-water fluids that are actively circulating in the substrate (Wright, 1997). interface, but maintain a variety of chemical subenvironments. Sucrosic dolomite D2 (Fig. 2D) was deposited on the inter- For these reasons, biogenic sedimentary structures can provide nal surface of ichnofossils and open moulds, possibly during a substrate and the chemical conditions required to host later diagenesis. Some dissolution probably preceded the pre- various micro-organisms. cipitation of D2, as no vestigial fabric or allochems are Because ichnofossils introduce physical and chemical observed in areas where this cement is present. This is further heterogeneity into sediment substrates, they are commonly supported by the presence of abrupt to sharp contacts gener- the locus of early diagenesis. The range of sediment modifica- ally shared by C1 or D1, with D2 (Fig. 2B). The dolomite tion due to bioturbation and the significance of those proc- crystals are euhedral. esses to dolomitization of the Tyndall Stone are discussed The isotopic data for D2 are typically indistinguishable from below. those for D1 (Fig. 3). To resolve these data with the textural observations (i.e. no allochems within; sharp, irregular con- Biomechanical modification of sediment tact; cross-cutting C1 and D1), it is suggested that D2 was derived from dissolved D1: infilling of vuggy porosity by D2 Burrowing organisms significantly alter the physical character (Fig. 2G) supports this interpretation. Sucrosic dolomite D2 of a substrate. The changes involved include modification of probably resulted from a mixture of subsurface fluids and grain size, redistribution of grain size (vertically and laterally), dissolved D1. However, the isotopic similarity of D1 and D2, compaction and sorting (Bromley, 1996). Infauna mediate shown in Fig. 3, may indicate that the formation of D2 was changes in the physical parameters by burrowing through not strongly influenced by isotopically variable subsurface and ingesting the substrate. Burrowing is accomplished by fluids (i.e. more wholly derived from D1). one or more means of removing or pushing aside sediment. The final stage of the paragenetic sequence is represented These include inversion, compression, excavation or backfill by C2 (Fig. 2E), a void-filling calcite mosaic cement that (summarized in Bromley, 1996). The physical manifestations completely occludes burrows and pores where it is present – of these processes are variable because infauna can either some open voids and burrows are still present. A lack of increase or reduce the grain size, and sort or mix the sediment. geopetal or meniscate rinds implies that C2 precipitated in a Grain size reduction is thought to represent the most sig- phreatic environment. This is a primary cement, as evidenced nificant substrate modification (Chow & Longstaffe, 1995). by its sharp boundaries and the euhedral state of D2, which Reduction of substrate grain size permits a larger surface is normally enveloped by C2. area to interact with dolomitizing fluids and provides a The range of isotope ratios observed for mosaic calcite C2 larger number of nucleation sites per given volume; however, suggests that there was more than one source for the C2 stock. this process may reduce the overall porosity. An increase in Carbon and oxygen for the C2x samples, for example (Fig. 3), grain size may contribute to dolomitization in a substrate if may have been inherited from dissolved cements (D1 and D2) the burrow system then provides a conduit through which during partial dedolomitization of the Tyndall Stone; how- dolomitizing fluids may interact with the matrix. ever, no direct evidence for dedolomitization of D2 was observed. Clearly basinal fluids or meteoric water also contrib- Biochemical modification of sediment uted to the formation of C2. Calcite C2 therefore represents a combination of subsurface brines and possibly dissolved An important result of burrowing in substrates is the precursor cements (C1–D2). Notably, 18O is not strongly incorporation of localized, concentrated organic material in depleted in any of the samples, suggesting that meteroic waters the form of mucous or faecal material. Decomposing organic did not contribute significantly to the overall diagenesis. debris creates a geochemical microenvironment that may We summarize the paragenetic history as having occurred extend several centimetres into the substrate. Furthermore, in five distinct stages: (1) an initial micrite (C1); (2) a calcite Sander & Kalff (1993) showed that the presence of organic dissolution phase focused around burrow fabrics; (3) a fine- material is most important to the production and activity of grained dolomite (D1) that was developed in the sulphate- micro-organisms. In well-aerated sediments, burrow linings reducing zone and represents the primary, or at least most composed of agglutinated, faecal or primary organic material distinctive diagenetic stage; (4) formation of a medium-grained contain the most concentrated local sources of organic to sucrosic primary dolomite (D2); and (5) the precipitation material. These burrows therefore can provide excellent of a calcite mosaic. substrates for bacterial colonization.

Conceptual models for burrow-related dolomitization 27

Fig. 4 A schematic diagram illustrating the geochemical characteristics of the oxic and anoxic zones in a subaqueous (marine) substrate.

2+ 2+ 2− ↔ Incorporation of organic matter into the substrate substan- Ca + Mg + 2(CO 3 ) CaMg(CO3)2 or 2+ 2− ↔ tially alters the local geochemistry. A signiﬁcant chemical CaCO3(solid) + Mg + 2(CO 3 ) byproduct in oxygenated substrates is CO2, which is produced CaMg(CO3)2(solid)(Lippmann, 1973) during the oxidization of organic material. Depending on the geochemical conditions, increased CO2 activities can enhance but not necessarily the replacement of calcite by dolomite. carbonate dissolution in the early diagenetic environment. Slaughter & Hill (1989) suggest that organogenic dolomitiza-

Thus, burrow-derived CO2 could generate an anisotropic, tion results from the transfer of neutral ion pairs and to and heterogeneous network that can be exploited by early diage- from a construction site on the crystal surface. For this reason 2– netic fluids. Sulphate ions (SO4 ) are also present in oxidizing they conclude that dolomitization requires the concentration 2– – subenvironments. These ions not only inhibit the dissolution of CO3 to be high and exceeding that of HCO3. For these and precipitation of dolomite (Reddy, 1977; Baker & Kastner, conditions to be met, the pH and carbonate alkalinity must be 1981), they may also bond with Mg2+ and thus lower the Mg/ high, such as in the burrow microenvironment. Ca ratio (Fig. 4). In reducing substrates, sulphate ions are consumed by It is known that dolomite formation preferentially occurs sulphate-reducing bacteria and the inhibitory effects of these under reducing conditions. In burrow microenvironments ions are removed from the system. The combination of that are characterized by reducing conditions, the geochemical elevating pH and reducing sulphate levels may be key to the processes and byproducts are much different from those found precipitation of dolomite during early diagenesis. Brown & in oxygenated substrates – this is in part due to the presence Farrow (1978), studying dolomite concretions surrounding 2– of microbes, the products of which can promote dolomitiza- modern crustacean burrows, proposed that CO3 ions were tion (Warthmann et al., 2000; Van Lith et al., 2003a,b). produced during sulphate reduction: Notably, only burrow forms that sequester organic material 2− ↔ 2− are likely to develop a comparatively reducing microenviron- SO4 + 2C + H2O CO 3 + CO2 + H2S and 2− ↔ 2− ment. Such burrows have organic linings or faecal fills, like SO4 + 2H2 + H2O CO3 + 3H2O + H2S. the ichnofossils Chondrites and Palaeophycus. Although CO2 is a common product in this geochemical setting, NH3 is The authors postulated that because of the ubiquity of Mg in also abundant (Slaughter & Hill, 1989). Ammonia com- marine environments, the geochemical microenvironment ↔ + – bines with H2O (NH3 + H2O NH4 + OH ) and effec- was the most important parameter for dolomite cementation tively raises the pH of interstitial fluids in and around the of modern burrows. They went on to note ‘early diagenetic burrow networks. High pH increases the activity of the bicar- events may be common in temperate waters and may be – bonate ion (HCO3), which favours the direct precipitation of associated with organic activity – burrowing and bacteria’. dolomite: We further suggest that dolomitization may occur below the

© 2004 Blackwell Publishing Ltd, Geobiology, 2, 21–30 28 M. K. GINGRAS et al. sulphate-reducing zone if sulphate has been mostly removed by Gunatilaka et al. (1987) from Quaternary siliciclastic tidal- from the sediment. bar deposits in northern Kuwait. In both examples, it was sur- The role of micro-organisms in mediating dolomitization mised that organic material mediated the dolomitization and cannot be underestimated (Burns et al., 2000; Van Lith et al., that the burrows had served as a fluid conduit. 2003a). Using bacteria-inoculated brines, Van Lith et al. (2003b) successfully precipitated Mg-calcite under reducing Metal ion enrichment in burrows conditions. They surmise that ‘active metabolism of sulphate reducing bacteria is essential to induce low-temperature It is proposed that cation enrichment on burrow linings, along dolomite formation in hypersaline solutions under reducing with Mg from inerstitial (marine) waters, provides a plausible conditions’ (Van Lith et al., 2003b, p. 77). Van Lith et al. source of Mg for dolomitization (Fig. 5). This is supported by (2003b) further suggest that the requisite geochemical condi- earlier research suggesting that metals present in organic tions include sulphate removal, initial pH increase, carbonate material facilitate dolomitization (Mirsal & Zankl, 1985). alkalinity increase, concentration of bivalent cations and Enrichment of trace metals in burrow walls is well documented. chemical diffusion gradients. These conditions are similar to Over (1990) found trace metals were preferentially concen- those outlined above for burrow microenvironments, partly trated in burrow linings (primarily Ophiomorpha nodosa). Iron, because those environments are microbially influenced. Mn, Cu, Ni and Zn were concentrated in burrow walls in three Van Lith et al. (2003b) and Rogers et al. (2003) both con- ways: (i) adsorption, as oxide or oxyhydroxide coatings on to sider the initial phase of dolomitization to occur as a microbial the high-energy surfaces provided by fine-grained particles precipitate; this process is also viable in the burrow environ- (silt and clay); (ii) as sulphide or phosphate phases under ment. Biological precipitates potentially act as a seed crystal reducing conditions; or (iii) associated with organic material under amenable bulk (pore)water conditions. (as adsorbed coatings or forming organometallic complexes). With regard to methanogenesis, it is uncertain if micro- The relative importance of each type of enrichment mechanism organisms engaged in fermentation of organic material was not addressed. Significantly, metal enrichment occurred precipitate dolomite crystals: to answer this question, much before burial, but its imprint persisted into the rock record. more research is required. However, the overall geochemical environment is at least geochemically amenable to dolomite The ichnology of the Tyndall Stone precipitation. There are examples of dolomitized burrows from modern The diagenetic fabric of the Tyndall Stone strongly reflects the sedimentary environments. Garrison & Luternauer (1971), original sedimentological and ichnological character of the working at the Fraser River Delta, British Columbia, noted deposit. Pronounced dolomitized halos surrounding causative the occurrence of cemented sand-filled burrows postulated burrows, locally abundant open tubular cavities and the selective to have formed under reducing conditions. An example of dolomitization in association with those artefacts strongly imply dolomitization in and around arthropod burrows was provided that dolomitizing fluids were channelled primarily through

Fig. 5 Metallic concentration in burrow walls results from steep geochemical gradients that are present in the burrow microenvironment.

© 2004 Blackwell Publishing Ltd, Geobiology, 2, 21–30 Conceptual models for burrow-related dolomitization 29 networks of worm tubes. Notably, the preserved ichnofossils, At least for the example of the Tyndall Stone, the initial Palaeophycus and Chondrites, commonly contain organic presence of organic material in the burrow networks has materials. profoundly influenced the likelihood of developing burrow The original diameter of the burrows ranged between 2 dolomite. The importance of organic concentration is and 4 mm. This is contrary to Kendall’s (1977) interpretation underscored by the following facts: (1) only Chondrites- and that suggested the mottled dolomite pattern was the result Palaeophycus-dominated fabrics – ichnofossils known to of larger-diameter arthropod burrows (Thalassinoides), rebur- sequester organic material – are observed at the centre of rowed by worms. However, the formation of dolomite within dolomite halos; (2) several ichnofossils that do not typically and around bioturbation structures is consistent with the sequester organic material, such as Planolites, are not dolomi- observations of Zenger (1996a,b), who described similar dolo- tized; and (3) a burrow with a faint organic core and elevated mitization patterns in the Upper Ordovician Bighorn Facies metal concentration (usually Fe) is commonly located near the from Williston Basin deposits in the USA. Zenger’s (1996a,b) centre of the dolomite patch. observations indicated that dolomite formed within and around Ultimately, dolomitization of burrow fabrics undoubtedly biogenic sedimentary structures that are similar to those pre- depends on several factors not detailed above. These include sented herein. In contrast to the Tyndall example, ichnofossils the depth of bioturbation, the oxygen content of the deposi- of the Bighorn Facies have a more homogeneous cement and tional waters, the depth to the redox discontinuity, sedi- may not have suffered a second phase of dolomitization. mentation rates and various infaunal behaviours not considered The role of organic material associated with the original here. However, quantification of those factors will require exten- trace fossil is probably crucial. In the example of the Tyndall sive research in modern sedimentary systems and will probably Stone, dolomite is primarily associated with Chondrites- and not be determined strictly from rock-record examples. Palaeophycus-dominated fabrics. Other ichnofossils that do not typically sequester organic material, such as Planolites, are ACKNOWLEDGEMENTS generally not dolomitized. Moreover, the faint organic lining of the ichnofossil is still present near the centre of some halos Astrid Arts and Jason Montpetit were inestimably helpful in (Fig. 2A,F). Is burrow-associated dolomitization a self- generating conceptual models of dolomitization. Brenda Hunda regulating process? It is reasonable to surmise that the whole conducted SEM analysis of burrow linings. Special thanks are rock has not been dolomitized because the ingredients due to Robert Grover, Brian Jones, Rozalia Pak, Tom Saunders, required to induce the diagenetic changes became depleted. John-Paul Zonneveld, Demian Robbins and Arjun Keswani There is, after all, only a limited quantity of organic material for their insights and comments. The reviewers and editors of and, perhaps, Mg sequestered in the trace fossil. Geobiology greatly improved the manuscript. Funding for this research was generously provided by the Natural Sciences Engineering and Research Council (NSERC) Operating CONCLUSIONS Grants to M.K.G., S.G.P., K.M. and H.M. The Tyndall Stone is an excellent example of how burrow structures facilitate dolomitizing processes in at least two ways: by enhancing the bulk fluid flow through the matrix; and by REFERENCES enriching the substrate in organic and metallic material that Baker PA, Kastner M (1981) Constraints on the formation of can be used to catalyse and construct dolomite. In many ways sedimentary dolomite. Science 213, 214–216. the dolomitization of the Selkirk Formation is comparable Beales FW (1953) Dolomitic mottling in Palliser (Devonian) with that of the Devonian Palliser Formation in the Western Limestone, Banff and Jasper National Parks, Alberta. Canadian Sedimentary Basin and Upper Ordovician rocks American Association of Petroleum Geologists Bulletin 37, 2281–2293. described by Zenger (1996a,b). In those cases dolomite Bromley RG (1996) Trace Fossils, Biology and Taphonomy, 2nd edn. formation has been closely linked to ichnofossils. Unwin-Hyman, London. It is important that the paragenetic sequence is tractable if Brown BJ, Farrow GE (1978) Recent dolomitic concretions of biological mediation of the sedimentary fabric and bulk crustacean burrow origin from Loch Sunant, west coast of composition are considered. Micrite C1 probably formed in a Scotland. Journal of Sedimentary Petrology 48, 825–834. Burns SJ, McKenzie JA, Vasconcelos C (2000) Dolomite formation shallow marine environment; this was followed by an episode and biogeochemical cycles in the Phanerozoic. Sedimentology 47, of metastable carbonate dissolution focused primarily around 49–61. ichnofossils; dolomite D1 probably developed in the sulphate- Chow N, Longstaffe F (1995) Dolomites of the Middle Devonian Elk reducing zone, mostly within the zone of the organic-rich Point Formation, Southern Manitoba: intrinsic controls on early bioturbation structures; dolomite D2 is sucrosic primary dolomitization. Bulletin of Canadian Petroleum Geology 43, 214–225. dolomite, which formed within dissolved burrow cores, Cowan J (1971) Ordovician and Silurian stratigraphy of the interlake possibly during burial diagenesis; and calcite C2 represents the area, Manitoba. In Geoscience Studies in Manitoba (ed. Turnock final void-filling cement. AC). Geological Association of Canada, Special Paper 9, 235–241.

Dimitrakopoulos R, Muehlenbachs K (1987) Biodegredation of Over DJ (1990) Trace metals in burrow walls and sediments, Georgia petroleum as a source of 13C-enriched carbon dioxide in the Bight, USA. Ichnos 1, 31–41. formation of carbonate cement. Chemical Geology 65, 283–291. Railsback LB, Ackerly SC, Anderson TF, Cisne JL (1990) Epstein S, Graf DL, Degans ET (1964) Oxygen isotope studies on Palaoentological and isotope evidence for warm, saline, deep the origin of dolomites, Chapter 13. In: Isotopic and cosmic waters in Ordovician oceans. Nature 343, 156–159. chemistry (Eds Craig H, Miller SL, Wasserburg JG). Amsterdam, Reddy MM (1977) Crystallization of calcium carbonate in the Netherlands, North-Holland Publishing Co. 169–180. presence of trace concentrations of phosphorous-containing Garrison RE, Luternauer JL (1971) Textures of calcitic cements anions. Journal of Crystal Growth 41, 287–295. formed during early diagenesis, Fraser Delta, British Columbia. Reitsema RH (1980) Dolomite and nahcolite formation in organic In Carbonate Cements (ed. Bricker OP). Johns Hopkins Press, rich sediments – isotopically heavy carbonates. Geochimica et Baltimore, MD, pp. 151–154. Cosmochimica Acta 44, 2045–2049. Gingras MK, Pemberton SG, Mendoza C, Henk F (1999) Modeling Rogers JR, Bennett PC, MacPherson GL (2003) Microbial fluid flow in trace fossils; assessing the anisotropic permeability of precipitation of dolomite in groundwater. In Goldschmidt Glossifungites surfaces. Petroleum Geoscience 5, 349–357. Conference Abstracts A400. The Geological Society (Department of Gunatilaka A, Al-Zamel A, Shearman AJ, Reda A (1987) A spherulitic Earth & Planetary Sciences, Washington University, St Louis, MO, fabric in selectively dolomitized siliciclastic crustacean burrows, USA). northern Kuwait. Journal of Sedimentary Petrology 57, 922–927. Sander CB, Kalff J (1993) Factors controlling bacterial production in Hudson JD (1977) Stable isotopes and limestone lithification. marine and freshwater sediments. Microbial Ecology 26, 79–99. Geological Society of London 133, 637–660. Slaughter M, Hill RJ (1989) The influence of organic matter in Jones B, Oldershaw AE, Narbonne GM (1979) Nature and origin of organogenic dolomitizion. Journal of Sedimentary Petrology 61, rubbly limestone: the Upper Silurian Read Bay Formation of Arctic 296–303. Canada. Sedimentary Geology 24, 227–252. Tobin KJ, Walker KR (1997) Ordovician oxygen isotopes and Kendall AC (1975) Anhydrite replacements of gypsum (satin-spar) paleotempuratures. Palaeogeography, Palaeoclimatology, veins in the Mississippian caprocks of southeastern Saskatchewan. Palaeoecology 129, 269–290. Canadian Journal of Earth Sciences 12, 1190–1195. Van Lith Y, Warthmann R, Vasconcelos C, McKenzie JA (2003a) Kendall AC (1977) Origin of dolomite mottling in Ordovician Sulphate-reducing bacteria induce low-temperature Ca-dolomite limestones from Saskatchewan and Manitoba. Bulletin of Canadian and high Mg-calcite formation. Geobiology 1, 71–79. Petroleum Geology 25, 480–504. Van Lith Y, Warthmann R, Vasconcelos C, McKenzie JA (2003b) Lippmann F (1973) Sedimentary Carbonate Minerals. Springer- Microbial fossilization in carbonate sediments; a result of the Verlag, New York. bacterial surface involvement in dolomite precipitation. Machel HG, Mountjoy EW (1986) Chemistry and environments Sedimentology 50, 237–245. of dolomitization; a reappraisal. Earth-Science Reviews 23, Warthmann R, Van Lith Y, Vasconcelos C, McKenzie JA (2000) 175–222. Bacterially induced dolomite precipitation in anoxic culture Mirsal IA, Zankl H (1985) Some phenomenological aspects of experiments. Geology 28, 1091–1094. carbonate geochemistry: the control effect of transition metals. Wright DT (1997) An organogenic origin for widespread dolomite Geologische Rundschau 74, 367–377. in the Cambrian Eilean Dubh Formation, Northwest Scotland. Morrow DW (1978) Dolomitization of Lower Paleozoic burrow- Journal of Sedimentary Research 67, 54–64. fillings. Journal of Sedimentary Petrology 48, 295–306. Zenger DH (1996a) Dolomitization patterns in widespread Morrow DW (1990) Synsedimentary dolospar cementation: a ‘Bighorn Facies’ (Upper Ordovician), Western Craton, USA. possible Devonian example in the Camsell Formation, Northwest Carbonates and Evaporites 11, 219–225. Territories, Canada. Sedimentology 37, 763–773. Zenger DH (1996b) Dolomitization of the ‘C’ zone, Red River Murray RC, Lucia FJ (1967) Cause and control of dolomite Formation (Upper Ordovician) in deep core, Williston Basin, distribution by rock selectivity. Geological Society of America Richland County, eastern Montana. Contributions to Geology 31, Bulletin 78, 21–36. 57–75.