C. Brenhin Keller

Statistical geochemistry reveals disruption in secular lithospheric evolution about 2.5 Gyr ago

The Earth has cooled over the past 4.5 billion years (Gyr) as a result of surface heat loss and declining radiogenic heat production. Igneous geochemistry has been used to understand how changing heat flux influenced Archaean geodynamics, but records of systematic geochemical evolution are complicated by heterogeneity of the rock record and uncertainties regarding selection and preservation bias. Here we apply statistical sampling techniques to a geochemical database of about 70,000 samples from the continental igneous rock record to produce a comprehensive record of secular geochemical evolution throughout Earth history. Consistent with secular mantle cooling, compatible and incompatible elements in basalts record gradually decreasing mantle melt fraction through time. Superimposed on this gradual evolution is a pervasive geochemical discontinuity occurring about 2.5 Gyr ago, involving substantial decreases in mantle melt fraction in basalts, and in indicators of deep crustal melting and fractionation, such as Na/K, Eu/Eu* (europium anomaly) and La/Yb ratios in felsic rocks. Along with an increase in preserved crustal thickness across the Archaean/Proterozoic boundary, these data are consistent with a model in which high-degree Archaean mantle melting produced a thick, mafic lower crust and consequent deep crustal delamination and melting—leading to abundant tonalite–trondhjemite–granodiorite magmatism and a thin preserved Archaean crust. The coincidence of the observed changes in geochemistry and crustal thickness with stepwise atmospheric oxidation at the end of the Archaean eon provides a significant temporal link between deep Earth geochemical processes and the rise of atmospheric oxygen on the Earth.

Volcanic–plutonic parity and the differentiation of the continental crust

The continental crust is central to the biological and geological history of Earth. However, crustal heterogeneity has prevented a thorough geochemical comparison of its primary igneous building blocks—volcanic and plutonic rocks—and the processes by which they differentiate to felsic compositions. Our analysis of a comprehensive global data set of volcanic and plutonic whole-rock geochemistry shows that differentiation trends from primitive basaltic to felsic compositions for volcanic versus plutonic samples are generally indistinguishable in subduction-zone settings, but are divergent in continental rifts. Offsets in major- and trace-element differentiation patterns in rift settings suggest higher water content in plutonic magmas and reduced eruptibility of hydrous silicate magmas relative to dry rift volcanics. In both tectonic settings, our results indicate that fractional crystallization, rather than crustal melting, is predominantly responsible for the production of intermediate and felsic magmas, emphasizing the role of mafic cumulates as a residue of crustal differentiation.

A record of deep-ocean dissolved O 2 from the oxidation state of iron in submarine basalts

The oxygenation of the deep ocean in the geological past has been associated with a rise in the partial pressure of atmospheric molecular oxygen (O2) to near-present levels and the emergence of modern marine biogeochemical cycles. It has also been linked to the origination and diversification of early animals. It is generally thought that the deep ocean was largely anoxic from about 2,500 to 800 million years ago, with estimates of the occurrence of deep-ocean oxygenation and the linked increase in the partial pressure of atmospheric oxygen to levels sufficient for this oxygenation ranging from about 800 to 400 million years ago. Deep-ocean dissolved oxygen concentrations over this interval are typically estimated using geochemical signatures preserved in ancient continental shelf or slope sediments, which only indirectly reflect the geochemical state of the deep ocean. Here we present a record that more directly reflects deep-ocean oxygen concentrations, based on the ratio of Fe3+ to total Fe in hydrothermally altered basalts formed in ocean basins. Our data allow for quantitative estimates of deep-ocean dissolved oxygen concentrations from 3.5 billion years ago to 14 million years ago and suggest that deep-ocean oxygenation occurred in the Phanerozoic (541 million years ago to the present) and potentially not until the late Palaeozoic (less than 420 million years ago).

Complementary crystal accumulation and rhyolite melt segregation in a late Miocene Andean pluton

High-silica granites are hypothesized to form via fractionation in the shallow crust, yet the predicted residues are rarely identified and can be difficult to distinguish within plutons whose rocks otherwise plot along liquid lines of descent. Bulk-rock compositional mass balance in the late Miocene Risco Bayo–Huemul plutonic complex (Chile) suggests that lithological differences within the Huemul pluton reflect residual crystal concentration in response to melt extraction. A compositional gap from 70 to 75 wt% SiO2 and strong depletion in Ba and Eu suggest that Huemul alkali feldspar (Afs) granites are frozen remnants of highly evolved rhyolitic melt extracted from a mush. Quartz monzonites enriched in Zr and Ba with Eu/ Eu* near unity are interpreted to represent the complementary residual silicic cumulates of this fractionation process. Compositional variations in Afs granite zircon (Eu/Eu*, Dy/ Yb) further support extraction of this melt from a zircon-saturated mush. U-Pb zircon dates indicate that Huemul rocks evolved ~800 k.y. after initial crystallization of more mafic Risco Bayo rocks, likely precluding their evolution via fractionation from mafic forerunners. This pluton records a means to produce rhyolite in the upper crust, which has propelled large silicic eruptions during the Quaternary within the Andean subduction zone.

Zircon age-temperature-compositional spectra in plutonic rocks

Geochronology can resolve dispersed zircon dates in plutonic rocks when magma cooling time scales exceed the temporal precision of individual U-Pb analyses; such age heterogeneity may indicate protracted crystallization between the temperatures of zircon saturation (Tsat) and rock solidification (Tsolid). Diffusive growth models predict asymmetric distributions of zircon dates and crystallization temperatures in a cooling magma, with volumetrically abundant old, hot crystallization at Tsat decreasing continuously to volumetrically minor young, cold crystallization at Tsolid. We present integrated geochronological and geochemical data from Bergell Intrusion tonalites (Central Alps, Europe) that document zircon compositional change over hundreds of thousands of years at the hand-sample scale, indicating melt compositional evolution during solidification. Ti-in-zircon thermometry, crystallization simulation using MELTS software, and U-Pb dates produce zircon mass-temperature-time distributions that are in excellent agreement with zircon growth models. These findings provide the first quantitative validation of longstanding expectations from zircon saturation theory by direct geochronological investigation, underscoring zircon’s capacity to quantify supersolidus cooling rates in magmas and resolve dynamic differentiation histories in the plutonic rock record. INTRODUCTION Zircon is an inimitable chronicler of the temporal, thermal, and compositional evolution of many crustal magmatic systems. U-Pb zircon geochronology by chemical abrasion–isotope dilution–thermal ionization mass spectrometry (CA-ID-TIMS; Mattinson, 2005) is the most robust method available for constraining the tempo of pluton assembly and the longevity of magma reservoirs in deep time, having been applied to discern the incremental nature of composite intrusive suites (Coleman et al., 2004). As a result of steadily improved analytical precision, CA-ID-TIMS can resolve heterogeneous U-Pb zircon dates, or zircon spectra, in many igneous rocks at the hand-sample scale (Miller et al., 2007; Schoene et al., 2012; Broderick et al., 2015). Such studies demonstrate that zircon age distributions can reflect protracted zircon crystallization time scales, supplanting convention that zircon retains a singular emplacement age of its host rock (e.g., see discussion in Samperton et al., 2015). Therefore, zircon age information combined with complementary petrologic data from zircon and other phases can be used to constrain the time scales of magma emplacement, calculate supersolidus cooling rates, and track magma compositional evolution. An important step in using zircon to further quantify magmatic processes is to compare zircon geochronology of natural systems with theoretical and experimental models for zircon saturation and growth (Watson, 1996; Harrison et al., 2007; Ferry and Watson, 2007; Boehnke et al., 2013; Bindeman and Melnik, 2016). Ab initio diffusive saturation calculations predict noninstantaneous, nonlinear volumetric zircon growth during monotonic magma cooling, with an asymmetric distribution of zircon mass crystallized as a function of time (t) or temperature (T) (Watson, 1996). This model predicts high initial crystallization rates at the system’s zircon saturation temperature (Tsat) and continuous, near-exponentially decreasing growth to the solidus temperature (Tsolid). While the treatment of Watson (1996) has been employed to describe zircon dissolution during thermal rejuvenation (Frazer et al., 2014), no attempt has been made to quantitatively compare theoretical expectations with measured zircon age spectra. If a zircon population is the product of crystallization in a cooling, closed system, then the following criteria may be observed: (1) resolvable dispersion in U-Pb zircon dates, reflecting protracted crystallization time scales; (2) trends in zircon composition through time, reflecting evolving melt composition during fractional crystallization; (3) a systematic decrease in Ti-in-zircon crystallization temperature with time, reflecting magma undercooling (Ferry and Watson, 2007); and (4) asymmetric distributions of zircon mass crystallized as functions of time and temperature (Watson, 1996). While such features have been previously documented individually (e.g., Ickert et al., 2011; Tierney et al., 2016), no prior study has demonstrated these criteria simultaneously. As such, the geochronological, experimental, and modeling perspectives of zircon crystallization have yet to be concisely unified, and our ability to link U-Pb geochronology, petrology, and numerical modeling in magmatic systems thus remains limited. Here we address this shortcoming through geochronological, geochemical, and thermometric characterization of zircon from mid-crustal granitoids. PREVIOUS GEOCHRONOLOGY AND HYPOTHESIS We use the Bergell Intrusion, Central Alps, Europe (Fig. 1), as an excellent locale to test the four criteria presented here. Thermal insulation of *Current address: Nuclear and Chemical Sciences Division, Lawrence Livermore National Laboratory, 7000 East Avenue, Livermore, California 94550, USA; E-mail: samperton1@llnl.gov. †Current address: Berkeley Geochronology Center, 2455 Ridge Road, Berkeley, California 94709, USA. GEOLOGY, November 2017; v. 45; no. 11; p. 983–986 | Data Repository item 2017333 | doi:10.1130/G38645.1 | Published online 23 August 2017

A New Workflow to Assess Emplacement Duration and Melt Residence Time of Compositionally Diverse Magmas Emplaced in a Sub-volcanic Reservoir

Construction durations of magma reservoirs are commonly inferred from U–Pb zircon geochronology using various statistical methods to interpret zircon U–Pb age spectra (e.g. weighted mean ages of concordant zircon populations). However, in compositionally different magmas, zircon saturation and crystallization are predicted to occur at different times relative to other mineral phases and the geological event of interest; for instance, magma emplacement. The timescales of these processes can be predicted by numerical modeling and measured using U–Pb zircon thermal ionization mass spectrometry (TIMS) geochronology, therefore creating an opportunity to quantify magma emplacement in space and time to constrain the size and longevity of magma reservoirs during pluton construction. The Jurassic tilted, bimodal (gabbroic and granitic) Guadalupe igneous complex (GIC) in the Sierra Nevada arc presents an exceptional opportunity to study the construction duration of a shallow (1–10 km) magma reservoir comprising multiple magma batches. We present a new workflow to constrain emplacement ages from zircon geochronology of compositionally different magma batches and evaluate melt-present timescales. High-precision U–Pb chemical ablation isotope dilution (CA-ID)-TIMS zircon ages are combined with MELTS modeling to calculate zircon saturation ages for each dated sample. Bayesian statistics are then used to compare calculated zircon saturation distributions with zircon age distributions from TIMS data to predict time, temperature, and melt fraction at zircon saturation and solidus. In addition, we use mineral thermometry and cooling rate calculations to relate zircon saturation ages to emplacement ages for felsic and mafic rocks, resulting in a best estimate for the total construction duration of 295 ± 110 kyr for the GIC. Rhyolites exposed at the top of the GIC are ∼2–3 Myr older and thus not part of the same magmatic system. The good agreement between Ti-in-zircon crystallization temperatures and calculated zircon saturation temperatures by MELTS implies that bulk-rock compositions of both mafic and felsic rocks are close to liquid compositions. Mafic and felsic magmas experienced extensive mingling at the emplacement level in a magma chamber (which, as defined here, has temperatures above the solidus of the respective rock composition) encompassing ∼60% of the exposed map area of the complex shortly after construction. Melt was present within the system for a total duration of ∼550 kyr as constrained by two-dimensional thermal finite-difference modeling using an incremental growth and sill emplacement model. The construction and melt-present timescales calculated in this study for the shallow GIC have implications for the potential of in situ differentiation, mixing and mingling timescales and eruption in shallow magmatic systems.