Search

Marine carbonates record geochemical information on the state of Earth's surface environment in the geological past, but recrystallization during burial can make the records difficult to interpret. Here, we report shore to basin gradients in δ44Ca values in matrix limestone and burrow dolomite across a sequence of Late Ordovician (Katian Stage) burrow-mottled carbonates in the Williston Basin, an intracratonic epeiric marine basin in North America. Both gradients are interpreted to have diagenetic origins. The gradient in limestone δ44Ca values developed in a bathymetrically controlled sediment- to fluid-buffered diagenetic system operating between the center and the edges of the basin. Two dolomitizing events formed the gradient in dolomite δ44Ca values. The sediment-buffered end of the limestone gradient preserves the δ44Ca value of the original carbonate mud (∼–1.28 ± 0.10 ‰). The fluid-buffered end of the gradient conserves the δ44Ca value of contemporaneous seawater (–0.47 ± 0.10 ‰), because: (1) pore water exchange drivers operated with greater variety and intensity in shallow water settings, (2) shallow nearshore deposits exchanged more Ca with seawater during recrystallization to diagenetic calcite than deeper offshore deposits, and (3) diagenetic calcite formed with negligible fractionation under equilibrium conditions. The gradient conserves the local value of Δ sed (–0.81 ± 0.20 ‰) despite its diagenetic origin, and the magnitude of Δ sed is consistent with primary calcite precipitation in a 'calcite sea'. Field assessment of the equilibrium Ca isotope fractionation factor between calcite and dolomite of –0.41 ± 0.17 ‰ at ambient temperature is applied to correct basin-centered dolomite for fractionation, yielding –1.62 ± 0.27 ‰, which is very different from the –0.25 ± 0.07 ‰ value of Ca in evaporatively concentrated seawater from which the dolomite originally formed. Applying the same fractionation factor to correct literature data on early diagenetic dolomite in the Great Basin, North America, yields –0.69 ± 0.22 ‰ for the δ44Ca value of seawater, which is ∼0.2 ‰ lower than 'normal marine' waters in stratigraphically equivalent deposits in the Willison Basin, and 0.44 ‰ lower than evaporatively concentrated seawater during a brief period of basin restriction and anhydrite deposition. The nearly 0.5 ‰ variation in the δ44Ca value of Katian seawater falls within the range of scatter in the brachiopod reconstruction of seawater Ca isotopes. Here, we propose that the true ocean record of secular change lies towards the bottom of the range of scatter, and that the higher values in the record reflect local Ca cycling and circulation restriction in epeiric seas. Epeiric seas cultivate the fluid-buffered diagenetic conditions required to promote Ca isotope exchange between reactive carbonate sediments and seawater through carbonate mineral dissolution/recrystallization, crystal coarsening by way of Ostwald ripening, and replacive dolomitization. It is not the apportioning of kinetically fractionated aragonite and calcite in the carbonate deposition flux that drove secular changes in global value of Δ sed between 'calcite' and 'aragonite' seas, but rather the replacement of kinetic fractionations with equilibrium fractionations during fluid-buffered diagenesis with seawater. [ABSTRACT FROM AUTHOR]

Abstract The Late Devonian extinction ranks as one of the 'big five' Phanerozoic extinctions affecting up to 80% of marine species and is characterized by two major extinction events that are separated by <3 My. The leading hypotheses explaining both extinction events are global cooling and/or widespread marine anoxia. We test the marine anoxia hypothesis by analyzing uranium isotopes (δ 238 U) across a ∼7 My interval of well-dated Upper Devonian marine limestones from the Devil's Gate Formation in Nevada, USA. The measured δ 238 U curve shows no co-variation with local sediment-derived redox trends, water-depth dependent facies changes, redox-sensitive metals, TOC, or diagnostic elemental ratios. From these relationships, we infer that the δ 238 U curve represents global seawater redox conditions. Two negative δ 238 U excursions (indicating more reducing seawater) are observed with durations of ∼3.8 My (late Frasnian) and ∼1.1 My (early Famennian), respectively. Steady-state modeling of the observed −0.2 to −0.3‰ shifts in δ 238 U points to a ∼5–15% increase in the total area of anoxic seafloor during these excursions. The late Frasnian negative excursion is broadly coincident with the first extinction event (late rhenana Zone or lower Kellwasser event), whereas the early Famennian negative excursion (lower-middle triangularis zones) occurs after the most intense Frasnian–Famennian boundary (F–F) extinction event (upper Kellwasser event). Compilations of local sediment redox conditions from Upper Devonian marine deposits with conodont zone-level age control indicates that the extinction events were coincident with widespread anoxic deposits accumulating in subtropical epeiric sea and some open ocean settings, supporting previous interpretations that widespread marine anoxia had an important influence on the Late Devonian extinctions. The temporal relationships between global-ocean redox trends represented by the δ 238 U curve and the newly compiled subtropical marine redox sediment trends indicates that local epeiric seawaters carried a similar U-isotopic composition as the open ocean for the majority of the studied interval except for a brief interval (<500 ky) spanning the F–F boundary. Highlights • Late Devonian δ 238 U trends from marine limestones record two widespread anoxic events. • Late Frasnian global anoxia coincides with lower Kellwasser extinction event. • Epeiric sea anoxia but not global ocean anoxia associated with Upper Kellwasser extinction event. • Late Devonian epeiric seas in relatively good redox communication with open oceans. [ABSTRACT FROM AUTHOR]