Abstract We present an updated set of Carboniferous Sr, C and O isotope stratigraphies based on the existing literature, given the importance of chemostratigraphy for stratigraphic correlation in the Carboniferous. The Carboniferous 87 Sr/ 86 Sr record, constructed using brachiopods and conodonts, exhibits five first-order phases beginning with a rapid decline from a peak value of c. 0.70840 at the Devonian–Carboniferous boundary to a trough (0.70776–0.70771) in the Visean followed by a rise to a plateau ( c. 0.70827) in the upper Bashkirian. A decline to c. 0.70804 follows from the lowermost Gzhelian to the close of the Carboniferous. Contemporaneous carbonate δ 13 C records exhibit considerable variability between materials analysed and by region, although pronounced excursions (e.g. the mid-Tournaisian positive excursion and the end-Kasimovian negative excursion) are present in most records. Bulk carbonate δ 13 C records from South China and Europe, however, are generally consistent with those of brachiopod calcite from North America in terms of both absolute values and trends. Both brachiopod calcite and conodont phosphate δ 18 O document large regional variability, confirming that Carboniferous δ 18 O records are invalid for precise stratigraphic correlation. Nevertheless, significant positive δ 18 O shifts in certain intervals (e.g. mid-Tournaisian and the Mississippian–Pennsylvanian transition) can be used for global correlation.

The stratigraphic and regional distributions of paleosol morphology in latest Pennsylvanian through Early Permian strata in Colorado, Utah, Arizona, New Mexico, Texas, and Oklahoma are presented in this paper. This regional extent corresponds to a paleolatitudinal gradient spanning ~5°S to 10°N. Morphological trends from this region delineate significant and systematic temporal and spatial changes in Permian-Carboniferous paleoenvironment and paleoclimate. The inferred latest Pennsylvanian (Virgilian) through early Early Permian environmental pattern is complex, but it indicates persistently dry, semiarid to arid conditions in Colorado, Utah, and Arizona, at paleolatitudes north of ~2°N, whereas lower paleolatitude (~2°S to 2°N) tropical regions in New Mexico exhibit a stepwise shift from subhumid to semiarid and variably seasonal conditions throughout late Pennsylvanian and the first half of Early Permian (Virgilian through Wolfcampian) time, followed by a subsequent shift to more arid conditions during the latter part of the Early Permian (Leonardian). Notably, strata from the southernmost paleosites, in Texas and Oklahoma, exhibit the most significant and abrupt climate changes through this period; they show a rapid transition from nearly ever-wet latest Pennsylvanian climate (at ~5°S) to drier and seasonal climate across the Permian-Carboniferous system boundary, and finally to arid and seasonal climate by Leonardian time (at ~2–4°N). The inferred climate patterns show no robust long-term correlation with the high-latitude Gondwanan records of glaciation. Rather, the long-term record of Permian-Pennsylvanian climate indicators from the southwestern United States is most simply explained by an ~8° northward tectonic drift through (essentially) static climate zones over western tropical Pangea during the interval of study. However, the relatively rapid perturbations to climate recorded by these pedogenic archives appear to be too rapid for tectonic forces and might correspond to changes in climate drivers, such as atmospheric p CO 2 , atmospheric circulation, and glacial-interglacial cycles.

Abstract A numerical model is developed to calculate the rates at which minerals precipitate or dissolve in basin strata as groundwaters migrate along temperature and pressure gradients. The calculation is based on the assumption that minerals maintain local equilibrium with migrating groundwater and their solubilities depend only on temperature or temperature and pressure. The model integrates predicted groundwater flow patterns with geochemical reaction path modeling; this approach allows us to predict the rate at which minerals dissolve and precipitate in complex geochemical systems open to groundwater flow and mass transfer. The model is formulated and solved in geologic time and basin distance scales and can therefore be applied to study basin-wide diagenesis related to long-distance fluid migration. The calculation can adjust sediment porosity from the net volume of precipitation and dissolution, therefore accounting for feedback effects of chemical diagenesis on porosity, which in turn affects permeability and fluid flow. The model is used to study the rates and nature of diagenetic alteration in several hydrologic systems, including (1) diagenesis of quartz by flow through a wavy sandstone, a sloping aquifer and a faulted aquifer, (2) cementation of amorphous silica and its feedback effect on thermal convection, (3) cementation of anhydrite in the Lyons Sandstone, Denver basin, and (4) diagenesis by migrating brines in the deep aquifers of the Illinois basin. The sample calculations shed light on the rates and patterns of chemical diagenesis that Likely accompany fluid migration in sedimentary basins. When the predicted results can be compared to diagenetic patterns observed in basin strata, the model provides an interpretation for the origin of diagenetic alteration.

Abstract Precambrian basins in Australia and Canada host massive unconformity-type uranium ore deposits, which constitute roughly 25% of the world’s known uranium resources and are the focus of international study as ancient analogs for nuclear waste repositories. Simulation of reactive mass transport and variable-density groundwater flow within these basins indicates that free convection of a uranium-bearing chloride brine at rates of about 1 ni/yr precipitated ore-grade quantities of uraninite (U0 2 ) near the unconformity between basin sandstones and graphiterich basement rocks within 100 to 1000 ky. Hydrothermal/diagenetic alteration surrounding the ore deposits resulted from mass transport through temperature gradients and across compositional boundaries. This alteration, consisting primarily of chlorite and muscovite precipitation, is strongly dependent on the pattern of groundwater flow and heat transport within the basins. The primary condition required for mineralization is the coincidence in space of graphite-rich basement rocks and upwelling limbs of free convection cells.

Abstract Models of burial diagenesis in the Gulf of Mexico sedimentary basin must explain a wide variety of phenomena, including: (1) uranium mineralization in volcanogenic Oligocene sandstones where the sandstones overlie growth fault zones in Eocene units, (2) lead-zinc mineralization in salt dome caprocks, with dissolved lead and zinc being known only from saline formation waters locally present in deeply buried Mesozoic reservoirs, (3) discharge of NaCl at the land surface, contributing to the dissolved chloride load of rivers, (4) natural seepage of oil and gas, which also leads to “vent” marine communities and IJ C-depleted CaCO,-cemented zones on the continental shelf and slope, (5) the presence of hydrocarbons above, and not uncommonly displaced laterally by ten’s or hundred’s of kilometers from mature source rocks, and (6) allochthonous, non-metalliferous saline water in Cenozoic clastic units. Fluid movement along faults is important in most, if not all of these processes. Convection within the self-fractured overpressured zone is inferred, based on the volumes of water necessary to: (1) remove Si0 2 and CaC0 3 from mudrocks and emplace authigenic quartz and calcite cements in sandstones, (2) transfer K 2 0 from feldspar dissolution in sandstones into mudrocks, where it is consumed by illitization, (3) remove sufficient volumes of hydrocarbons from “lean” Gulf Coast mudrocks as kerogen maturation proceeds and transport the hydrocarbons to permit the accumulation of significant volumes of oil and gas. Because the basement rocks beneath Gulf sediments are probably undergoing prograde metamorphism and devolatiUzation, material transfer into the sedimentary basin from the basement is inferred. The rate at which water (and C0 2 ) are added to the sedimentary basin from underlying rocks can potentally affect not only the volumes of water available for diagenesis but can maintain geopressures (and convection) within the sedimentary section long after pressure would normally decay back to hydrostatic values if compaction were the only operative process.

Abstract Most sedimentary basins contain close to 20% by volume pore water; much of which is of high ionic strength and is classified as brine. Although considerable variation occurs in the composition of basinal waters, some correlation exists between major ion concentrations and total dissolved solids. It is consequently possible to construct geochemical models for basinal waters that have general utility for investigating a variety of subsurface processes. In this paper, we examine the potential role of brine composition and movement on carbonate mineral mass transport in sedimentary basins up to pressure and temperature conditions of 300 bars and 100°C. Primary emphasis is placed on the impact of vertical and lateral migration of brines and the dispersive mixing of brines with waters containing widely varying concentrations of total dissolved solids. Model results indicate that mixing of subsurface waters may produce fluids which are highly supersaturated and at times slightly undersaturated with respect to calcite. This process may represent a major mechanism for production and destruction of carbonate cements in sediments. It may also offer an explanation as to how basinal scale mass transfer of carbonates can occur in waters that are close to equilibrium with respect to major sedimentary carbonate minerals.

Abstract Burial diagenesis of platform carbonates is a very complex process occurring over tens of millions of years, encompassing several different tectonic settings, exhibiting many diagenetic products and involving waters of diverse origins and compositions. To facilitate understanding of these processes, burial diagenesis of platform carbonates is divided into three hydrotectonic realms: (1) passive margin burial diagenesis, (2) collision margin burial diagenesis, and (3) post-orogenic burial diagenesis. The passive margin burial diagenetic realm, exemplified by the northern Gulf of Mexico, is characterized by extensional tectonics, growth faulting, a relatively uniform subsidence rate, slow upward flow of compaction-driven fluids, and increases in temperature, pressure and salinity of pore waters with burial depth. The passive margin burial diagenetic realm can be divided into three subrealms: (1) pre-oil window diagenesis (<100°C), (2) oil-window diagenesis (100-140°C) and (3) gas-window diagenesis (140-200°C). Quantitative analyses of Upper Jurassic calcite ooid grainstones that have experienced all 3 subrealms indicate that, of the original 45% porosity, an average of 12% can be filled by carbonate cement, whereas 33% is lost by compaction and pressure solution. Unless a well-developed convection system exists, low HC0 3 contents of formation waters and slow fluid flow rates in passive margin settings suggest that most carbonate cements are derived locally from pressure solution of host carbonates. In passive margin settings, Mg 2 + contents of burial calcite cements should initially increase and then decrease with burial. This trend develops because initial increases in Mg 2+ distribution coefficient (DMg 2+ ) with temperature outweigh decreases in the mMg 2 V mCa 2+ ratio of pore waters, resulting in increases in Mg 2+ contents of calcite cements. But later, continued lowering of mMg 2+ /mCa 2+ ratios reverses this trend. Data suggest that Sr 2+ contents of calcite cements, initially very low, increase with temperature during deep burial. Wide variations in mFe^/mCa 2 * and mMn 2+ /mCa 2+ ratios of formation waters indicate that concentrations of Fe 2+ and Mn 2+ in carbonate cements are of limited value beyond local scales. The 8 ,8 0 compositions of calcite cements first decrease and then increase during passive margin burial. This trend develops because temperature-dependent calcite-water fractionation produces an initial decrease in 8 18 0 values of cements; but later, progressive increases in 8 ,8 0 values of formation waters overcome temperature fractionation effects, resulting in increases in 8 ,s O composition of carbonate cements. The 8 ,3 C values of carbonate cements can follow either one of two paths. In the absence of hydrocarbons, 8 I3 C compositions of pore waters are buffered by host carbonates and change little with burial. Where hydrocarbons are present, their decarboxylation at temperatures above 150°C promotes precipitation of carbonate cements with low 8 ,3 C values. Therefore, two burial 8 18 0-d l3 C trends are generated: (1) a “C trend” when organic-derived C0 2 is not available and (2) a “D trend” when hydrocarbon destruction generates significant amounts of CO,. The collision margin burial diagenetic realm, exemplified by the Ouachita collision belt, is characterized by compressional tectonics, thrust faulting, variable uplift/subsidence rates and episodic focused expulsion of tectonic fluids toward the craton. Three subrealms are recognized based on the intensity and types of diagenetic alterations: thrust belt diagenetic zone, foreland diagenetic zone, and craton margin diagenetic zone. The more important carbonate diagenetic events in collision margin settings are: (1) extensive pressure solution and fracture-fi11 carbonate cementation, (2) Mississippi Valley-type (MVT) mineralization, and (3) K-feldspar and magnetite precipitation. As much as 50% (by volume) of carbonates can dissolve by tectonic pressure solution and be available to precipitate in tectonic fractures. H 2 S produced by thermochemical sulfate reduction in deep carbonate reservoirs during passive margin settings is expelled during collision margin diagenesis to react with base metal-rich fluids in shallow depths leading to the formation of MVT ore deposits. Sulfide mineralization generates significant amounts of acids which can dissolve syndepositional dolomites and reprecipitate them as burial dolomites. K-feldspars form when K + -rich pore waters (generated by K-feldspar dissolution during passive margin diagenesis) are flushed to shallow parts of the basin. The post-orogenic burial diagenetic realm, exemplified by the Madison aquifer in the U.S. midcontinent, is characterized by a lack of tectonic activity, dominance of topographically-driven fluid flow and high fluid flow rales. Rainwater charged with soil C0 2 enters aquifers in highland recharge areas dissolving carbonates and evaporites along its flow paths. Temperature, pressure, alkalinity, pH, Ca 2 + , Mg 2+ , Cl + and Na + concentrations of groundwater increase and Eh decreases toward the discharge area. Important diagenetic processes in this realm are evaporite and carbonate dissolution, dedolomitization, calcite precipitation and bacterial sulfate reduction.

Abstract Upper Cambrian and Lower Ordovician sedimentary rocks in southeastern Missouri host the world-class Mississippi Valley-type (MVT) lead-zinc deposits of the region. Sedimentary facies of the lower part of the Upper Cambrian section are dominated by distinct clastic and carbonate facies belts associated with a high relief Cambrian topography. The upper part of the Upper Cambrian (post Davis Formation) and the Lower Ordovician section were deposited under epeiric sea conditions on a low relief topography. These latter rocks are characterized by cyclic sequences of shallow-water platform carbonates, with good lateral continuity of facies. The Davis Formation is composed of interbedded carbonates and shales and forms an effective aquiclude separating the upper and lower parts of the section into two distinct aquifers. Petrographic and cathodoluminescent studies of epigenetic dolomite cements in these Cambro-Ordovician sedimentary rocks document that: (1) dolomite cements of the Bonneterre Dolomite (lower aquifer) in the Viburnum Trend and the Old Lead Belt, which are related closely to mineralization, have a relatively complex, four zone cathodoluminescence (CL) pattern; (2) dolomite cements, in the Bonneterre and Davis Formations, which are not related spatially to mineralization commonly display less complex CL patterns; and (3) dolomite cements in the post- Davis part of the Cambrian and in the lower Ordovician section (upper aquifer) display a CL stratigraphy unrelated to that observed lower in the section. Carbon isotope compositions of host dolomite show two types of statistical variation. From the bottom of the Bonneterre Dolomite to the top of the Davis Formation, 8 I3 C values become higher (from -2.5 toward +3.0%»). Above the Davis to the Lower Ordovician, the trend reverses and 8 n C values become lower (toward — 3.0%o). A similar trend exists for 5 I8 0 values in the Bonneterre and Davis, as values become higher up section. However, above the Davis Formation, 8 1S 0 values for host dolomites display no statistical trends. The pattern of upwardly decreasing 8 I3 C values in post-Davis rocks may be the result of a secular trend in ocean carbon. The trend of upwardly increasing S' 3 C and 8 I8 0 values in the Bonneterre Dolomite and Davis Formation is likely the result of interaction with hydrothermal fluids emanating from the underlying Lamotte Sandstone, reflecting increased buffering by host dolomite as l2 C- and l6 0-enriched light fluids moved higher in the section. Distribution of sedimentary facies had a profound effect on the hydrological framework of southern Missouri during mineralization. Linear facies belts that developed on high-relief topography during deposition of the Bonneterre and Davis strata resulted in focused fluid flow and a greater degree of alteration of host dolomite. Broad, laterally continuous distribution of sedimentary facies in post-Davis rocks resulted in less focused fluid flow and less alteration of the host dolomite. The distinct C and O isotopic trend observed in the Bonneterre-Davis Formations versus that observed in post-Davis Formation rocks, coupled with differences in CL microstratigraphies of dolomite cements, indicate that the these two parts of the section acted as distinct aquifers, with relatively little fluid communication during mineralization.

Abstract An integrated stratigraphic, petrographic and petrophysical study of closely spaced cores of the Lower Ordovician Upper Knox Group from the greater central Tennessee region documents stratigraphic trends in the distribution of early and late diagenetic replacement dolomites, secondary porosity and MVT mineralization that correlate with the distribution of depositional facies and cycle stacking patterns and the long-term accommodation history they record. Thin to thick (5 to 30 m) intervals of nonporous to porous, late diagenetic replacement dolomites occur with limestones in transgressive facies of large-scale depositional sequences that formed during long-term increases in accommodation space. In contrast, all facies within regressive cycles of large-scale depositional sequences are near completely replaced by nonporous, early diagenetic dolomites; minimal limestone or late diagenetic dolomite occurs within regressive cycles. Early diagenetic dolomites formed syndepositionally during tidal-flat progradation and exposure governed by short-term sea-level falls and were extensively developed in regressive cycles that formed during overall longer-term decreases in accommodation space. Shortened periods of progradation and exposure during short-term sea-level falls superimposed on longer-term increases in accommodation space resulted in partial dolomitization of transgressive cycles and retention of variable amounts of host limestone. Early diagenetic dolomites have minimal present-day porosity (avg. of 2.4%; range of 0.5 to 8.5% porosity) and permeability (avg. of 0.02 md; range of 0.00 to 0.14 md) but are interpreted to have been porous and permeable during deposition and early burial. Continued early diagenetic dolomitization coupled with varying degrees of diagenetic modification by burial dolomitizing fluids transformed porous and permeable early diagenetic dolomites into tightly interlocking, nonporous dolomite mosaics prior to maximum burial. Transgressive cycles contain minor (< 10%) to abundant (>50%) volumes of present-day, nonporous to porous (avg. of 8.2%; range of 2.2 to 15.6% porosity) and impermeable to permeable (avg. of 70.2 md; range of 0.02 to 1030 md) late diagenetic dolomites. Late diagenetic dolomite replacement of early diagenetic dolomites and limestones is interpreted to have occurred between latest Devonian to Carboniferous time at intermediate to deep burial depths. Limestones were extensively neomorphosed and cemented by Middle Ordovician meteoric waters and early burial fluids and have negligible present-day porosity (avg. of 1.1%) and permeability (avg. of 0.00 md). Moderate to extensive intercrystalline, moldic, vug and channel porosity developed within late diagenetic dolomites during advanced stages of dolomitization, whereas limestones and early diagenetic dolomites were unaffected by burial-related dissolution. Intervals of laterally continuous, porous late diagenetic replacement dolomites were established as hydrologically well-connected conduits in transgressive facies by intermediate to deep burial of Knox carbonates; moderate to thick intervals of early diagenetic replacement dolomites in regressive facies and stratigraphically isolated limestone intervals served as aquitards between conduits. Multiple generations of late diagenetic carbonate cements and MVT ore minerals and hydrocarbons are preferentially developed within late diagenetic replacement dolomites, which coupled with isotopic ages of 260 (±40) to 39 Ma for some late diagenetic minerals, indicate that dolomite conduits focused late diagenetic and mineralizing brines through the Knox regional aquifer for 100’s of million years. The systematic distribution of diagenetic facies with respect to depositional facies and cycle stacking patterns suggests a more complex internal architecture for the Knox regional aquifer than that implied by schematic models and hydrogeologic simulations of basinwide fluid migration. The three-dimensional distribution of this vertically stratified network of multiple, yet laterally continuous, fluid flow conduits and potential reservoirs shows a close relationship to the sequence stratigraphic framework and was ultimately governed by the long-term accommodation history of these cyclic carbonates.

Abstract Within the Presqu’ile barrier of northeastern British Columbia, a Middle Devonian carbonate complex extends over a lateral distance of 400 km in the subsurface of the Western Canada Sedimentary Basin (WCSB). Dolomite cements have decreasing 87 Sr/ 86 Sr ratios (0.7106 to0. 7081), decreasing fluid inclusion homogenization temperatures (178° to 92°C) and increasing δ 18 O values (— 16‰ to — 7‰PDB), northeastward. These trends suggest that hot, radiogenic basinal fluids moved updip (northeast) along the Presqu’ile Barrier and mixed with cooler ambient formation waters. This barrier behaved as a deeply buried regional conduit system in focussing and channeling basinal fluids. The geochemistry of the Rimbey-Meadowbrook reef trend, another extensive subsurface carbonate system in the WCSB, lacks the distinctive regional trends exhibited by the Presqu’ile barrier. With increasing depth southward along the Rimbey-Meadowbrook reef trend, dolomite and calcite cements have a slight decrease in δ 18 0 values (4 to — 7‰ PDB) and slightly more radiogenic Sr/Sr ratios, although no systematic relation between Sr and O isotopic compositions is observed. Minimum fluid inclusion homogenization temperatures (uncorrected for depth and pressure) in most dolomite and calcite cements from the Rimbey-Meadowbrook reef trend are 30° to 40°C higher than is estimated to have been reached during maximum burial in Early Tertiary time. The minimum homogenization temperatures (T h ) of inclusions in intermediate-diagenetic dolomite cements do not vary with depth suggesting that they may have formed from hydrothermal fluids during shallow burial. Late-diagenetic pre-thermal sulphate reduction (TSR) calcite cements (deep burial) have minimum T h that occur along a 40°C/km geothermal gradient or higher and may also indicate the involvement of hydrothermal fluids. However when corrected for depth and pressure they plot close to a 30°C/km gradient. Latediagenetic calcites associated with TSR exhibit a more variable pattern and follow geothermal gradients between 20° and 25°C/km. Replacement dolomites and dolomite cements from Frasnian Leduc and Famennian Wabamun strata near the southeast Peace River Arch formed from saline hydrothermal fluids (between 100° and 200°C) that moved upwards along an extensive fault and fracture conduit system (in part formed by extensive subsurface solution), possibly as early as Early Carboniferous time. Fluid inclusion homogenization temperatures are best explained in two cases by means of the net lateral flow of hydrothermal fluids along dolomitized conduits, or in one case by upward flow along a fault-fracture conduit system. Data from intermediate-diagenetic dolomite cements from the Peace River and the Rimbey-Meadowbrook reef trend suggest that fluid movements probably occurred during Late Paleozoic shallow burial, possibly related to the Antler orogeny to the west. This requires the movement of hot fluids at shallow depths both across parts of the basin and vertically. Temperatures of later calcite cements suggest that hydrothermal fluid movement took place prior to and after hydrocarbon maturation, probably close to maximum burial during the Laramide orogeny. The movements of hydrothermal fluids in the WCSB, in part focussed through subsurface conduit systems, appear to be mainly related to orogenic compression and sedimentary loading.

Abstract In the Alberta Basin, carbonates of the 120-km-long Upper Devonian Bashaw reef complex and the nearby 320-km-long Upper Devonian Rimbey-Meadowbrook reef trend, together with the underlying Cooking Lake platform, act as regional subsurface conduits for oil, gas and water. Evidence of regional fluid migration in these carbonates includes: (1) huge deposits of hydrocarbons that originate from a single source rock that have migrated up-dip for hundreds of km, (2) basin-scale studies of water movement indicating that these carbonates funnel fluids from nearby formations through the deeper parts of the basin towards its edge and (3) extensive pervasive dolomitization of the carbonate “eservoir rocks by a mechanism attributed to long-distance fluid migration. Analysis of potentiometric surfaces, pressure-depth plots and formation-fluid chemistries reveal that regional fluid flow occurs through two main aquifers. One is the Cooking Lake-Leduc aquifer that underlies and comprises the Rimbey-Meadowbrook trend, where fluids migrate laterally northward. The other is the Bashaw reef complex, where fluids move vertically upward across the Ireton aquitard and into the overlying Nisku aquifer. Regionally, fluids in the Nisku aquifer move up-dip to the northeast, except over the Bashaw area, where they are met from below by the ascending fluids from the Leduc aquifer. The critical control on petroleum trapping in both the Leduc and Nisku formations appears to be the thickness of the intervening Ireton aquitard. Geological mapping has identified 28 wells in the Bashaw area where the Ireton aquitard is very thin or absent, creating breaches that allow for cross-formational movement of water and hydrocarbons. Trapping in both the Leduc and Nisku formations can be correlated with the thickness of the Ireton aquitard. In the southern part of the Rimbey-Meadowbrook reef trend, trapping conditions are slightly different with many Leduc reefs filled to capacity. However, some hydrocarbons must have migrated across the Ireton aquitard, because there has been scattered production from the Nisku Formation. The results of this study demonstrate the role regional fluid flow and cross-formational flow play in hydrocarbon migration and entrapment in the subsurface. Considering the findings from the Bashaw area, it may be possible to apply the concept of vertical migration through thin shales to explore for other traps above Leduc reefs.

Abstract Diagenetic calcite, dolomite, pyrite, marcasite, and anhydrite of the Upper Devonian Nisku Formation in central Alberta, Canada, were analyzed for trace elements and sulfur isotopes by ion microprobe with sample spots <20 nm in size. Calcite samples exhibit significant trace element variations (of up to about 3 orders of magnitude), and disseminated sulfides display large sulfur isotope variations (8 34 S of up to about 60%o CDT) on a scale of < 100 nm. These variations cannot be resolved using conventional analysis of powdered samples. The data suggest that most of the analyzed calcites formed from marine pore fluids that evolved toward lower redox-potential. The sulfur isotope variations indicate intensive bacterial sulfate reduction and closed-system Rayleigh fractionation as the dominant diagenetic processes leading to iron sulfide formation. Together, carbonate and sulfide data indicate that the system was at least partially closed during shallow to intermediate burial diagenesis (up to about 1100 m) in the Late Devonian to Early Carboniferous. During the subsequent 300 my, the Nisku reefs in the study area were buried to more than 4000 m, with maximum burial in the Late Cretaceous to Early Tertiary. The existence and preservation of the observed geochemical variations further imply that water-rock interaction during this prolonged and deep burial was quite limited (i.e., the analyzed calcites and sulfides are remarkably resistant to compositional modification via recrystallization).

Abstract A succession of deep burial carbonate cements with two types of saddle dolomite and three types of blocky calcite was investigated in Permian to Tertiary sedimentary rocks in different nappes of the Eastern Alps. The first generation of saddle dolomite occurs only in rocks of Permian to Late Triassic/Early Jurassic age. All otlier carbonate cements occur within rocks of Permian to Early Tertiary age. Carbon and oxygen isotopic compositions of the carbonate cements and of the Triassic to Tertiary host rocks exhibit regional trends as well as trends to more negative δ 18 O with increasing burial. Fluid inclusion data show homogenization temperatures between 90° and 250°C for the carbonate cements. Temperatures decrease from the bottom to the top of the stratigraphic column, and regional trends are also exhibited. Calculated oxygen isotopic compositions of fluids precipitating the carbonate cements suggest strong positive 8’“0 values, which are characteristic of saline formation waters or metamorphic waters. The first generation of saddle dolomite is inferred to have formed in the same paleo-fluid system as Late Triassic/Early Jurassic Pb-Zn ores by fluid flow directed from the hinterland in the north (Vindelician high, Bohemian massif) to the East Alpine area. The fluids ascended and precipitated saddle dolomite and, under certain conditions, Pb-Zn ores. All other carbonate cements formed post-Oligocene time after the peak of metamorphism in the Central Eastern Alps and during uplift of this area. Results suggest that meteoric fluids descended and equilibrated with metamorphic rocks subsequently mixed with metamorphic waters in this uplifted area and flowed northwards and southwards, ascending through the rock pile. Fluid mixing with a second, near surface, meteoric groundwater system could explain a renewed decrease in carbonate cement δ 18 O values at the top of the sedimentary succession and in northern parts of the Alpine realm.