H33E-01 INVITED
The Start of Plate Tectonics in the Eoarchean: A Tribute to Gilbert N Hanson, Pioneer in Archean Geochemistry
The use of isotopic and trace element geochemistry and igneous petrology to understand the petrogenesis of Archean rocks was pioneered by Gilbert Hanson and Joseph Arth at SUNY Stony Brook in the 1970's. Extension of these approaches allows the onset of plate tectonics on Earth shortly after the end of the Hadean to be specified. Nb/Th and Th/U ratios of mafic-ultramafic rocks from the depleted upper mantle begin to change from 7 to 18.2 and 4.7 to 2.9 (respectively) at 3.6 Ga. This signals the appearance of subduction-altered slabs in general mantle circulation from subduction initiated at 3.8 Ga. Juvenile crustal rocks begin to show derivation from progressively depleted mantle with typical igneous εNd:εHf = 1:2 after 3.6 Ga. Cratons with stable mantle keels that have subduction imprints begin to appear at 3.5 Ga. These changes all suggest that extraction of continental crust by plate tectonic processes was progressively depleting the mantle from 3.6 Ga onwards. Neoarchean subduction appears largely analogous to present subduction except in being able to produce large cratons with thick mantle keels. The earliest Eoarchean juvenile rocks and Hadean zircons have compositions that reflect the integrated effects of separation of an early enriched reservoir and fractionation of perovskite from the Mars-size impact-derived magma ocean, rather than separation of voluminous continental crust or oceanic plate tectonics. Hadean zircons most likely were derived from a continent-absent, mafic to ultramafic protocrust that was multiply remelted between 4.4 and 4.0 Ga under wet conditions to produce evolved felsic rocks. If the protocrust was produced by global mantle overturn at ca 4.4 Ga, then the transition to plate tectonics resulted from radioactive decay-driven mantle heating. Otherwise, such protocrust would have been the typical product of mantle convection and the transition to plate tectonics resulted from cooling to the extent that large lithospheric plates stabilized.
H33E-02
Metasomatism of mantle-derived mafic and ultramafic rocks: Seawater? Subducted Slab? Sediment? Mantle wedge? or continental crust?
A variety of mantle-derived mafic and ultramafic rocks are exhumed as solid bodies as a result of tectonic processes occurring along all three plate boundary types. Mid Ocean Ridges (MOR), expose abyssal peridotite and serpentinite; subduction zones (SZ) contain serpentinite and blueschist-type or LT/HP eclogite; continental collisional zones (CC) are associated with "orogenic" serpentinite and spinel and/or garnet-bearing peridotite and eclogite and even transform faults (TF), such as the Motagua Fault Zone in Guatemala, display serpentinite, peridotite and eclogite. These rocks rarely reach the earth's surface without undergoing metasomatism. Metasomatism provides information on mantle and crustal processes while simultaneously obscuring information on the nature of the original mantle source. MOR metasomatism is the least complicated involving hydration and alteration by sea water. Subsequent SZ metasomatism occurs when fluids derived from sediments and hydrated basalts are added to the seawater signature. CC provides further complexity because orogenic peridotites and eclogites can be metasomatized in the mantle wedge by components derived from earlier MOR and SZ metasomatism and then can be metasomatized again when exposed to crustal fluids after they are transferred into the continental crust. However, the most complex metasomatism appears to have affected the serpentinites, eclogites and related rocks exposed on opposite sides of the Motagua Fault in Guatemala. Identical Cretaceous ages from eclogites on both sides require that they were exhumed recently (so that they are not offset) and high 87Sr/86Sr ratios and LIL concentrations from vein assemblages and retrograded eclogite requires that crustal metasomatism occurred during this exhumation. Cretaceous Sm-Nd ages and the enriched trace element chemistry and isotopic signature of eclogite facies minerals require prior metasomatism from sediment- and slab- derived fluids during Mesozoic subduction. Slightly enriched Sr and Nd isotope ratios from pyroxenes in serpentinized peridotite indicate metasomatism in a mantle wedge, presumably also during Mesozoic subduction. Finally, refractory trace element patterns (HFSE, HREE) and depleted Sr and Nd ratios indicate a MORB origin for eclogites and peridotites and serpentinites making it possible that some of the metasomatic signatures noted above occurred at a MOR. The Motagua Fault appears to have exploited a previous subduction zone resulting in an array of rocks with widely variable geochemical signatures.
H33E-03
Isotopic and Geochemical Tracers in Terrigenous Sediments: Applications to Tectonics, Crustal Evolution and Paleoclimate
The sources and sedimentary processes leading to deposition of terrigenous sedimentary rocks leave diagnostic compositional clues that make them valuable recorders of geological history. In the late 1980's through the 1990's several Stony Brook graduate students working in Gil Hanson's lab applied a variety of established petrographic and geochemical methods and developed new isotopic and geochemical approaches to study ancient tectonic histories preserved in several Paleoproterozoic and Paleozoic basins. These studies included testing Pb isotopes in detrital quartz and feldspar grains, using the Sm-Nd isotope system to identify isotopic resetting near the time of deposition, and the application of multiple tracers to unravel sources and processes in sedimentary systems. Since the early work of Goldschmidt, it has been appreciated that sedimentary rocks provide an "averaging" mechanism for large regions of exposed crust. Accordingly, sedimentary compositions, especially for elements not greatly mobilized by sedimentary processes such as weathering and diagenesis, can be used to estimate changes over geologic time in the composition of upper continental crust, and with this approach it appears that there was a profound change in the composition of exposed upper continental crust at about the Archean-post Archean boundary. Recently we have found a consistent case for this apparent change in a compilation of modern sediment data, using the Sm-Nd system as a measure of the continental age. The compositional changes that result from different source compositions and geological age as well as different degrees of chemical and mechanical weathering make the chemical and radiogenic isotope composition of sediments and their components valuable for paleoclimate studies as well. Thus the same approaches can be used to constrain such factors as wind and ocean circulation patterns and the dynamics of ice sheets in the past.
H33E-04
A Global Perspective on Nd-Hf Isotope Systematics in the Ocean From Marine Sediments and Ferromanganese Precipitates Around Antarctica
Neodymium-hafnium systematics of proximal terrigenous sediments around Antarctica are of major interest as they contain the power to address topics as diverse as crustal evolution, sedimentary processes, and the reconstruction of ocean circulation patterns. Here we present data on glacio-marine core-top sediments from around the perimeter of Antarctica. The results range from ε Hf = -30.1 to ε Hf = +0.8, reflecting the large range of basement ages and lithologies around the Antarctic continent. When plotted against Nd isotope results on the same sediments the data are collinear with the global silicate Earth array ("terrestrial array"). Hafnium and Nd isotopic compositions of Circum-Antarctic ferromanganese precipitates lie on the global "seawater array". The offset between these two arrays provides evidence for the sources and mechanisms by which these elements are added to ocean water. Lutetium-Hf and Sm-Nd evolution and mixing calculations provide compelling evidence that while hydrothermal contributions of Hf to the ocean may play a role, continental contributions are absolutely required to explain the variations in Hf isotope ratios in seawater. Furthermore, it can be shown that incongruent weathering of continental sources is alone sufficient, without any hydrothermal contributions, to create the observed offset of the Nd-Hf isotope array in seawater.
H33E-05
Tracing Provenance of Antarctic Glacio-marine Sediment Through Radiogenic Isotopes, Ar/Ar Hornblende Ages, and Bulk Geochemistry
We are applying a suite of sediment provenance tracing techniques to Antarctic marine diamict samples. The composition of sediments derived from the glacial erosion of Antarctica provides an important framework for evaluating iceberg contributions to the ocean over time. We targeted glacial diamict samples where available, or sandy mud otherwise. This strategy ensures sampling and homogenization over large areas, avoids possible biasing in limited and heterogeneous outcrops, and avoids the competing influences of sediment redistribution by ocean currents. When completed, this ongoing study will provide an integrated characterization of the lithologic clast composition, bulk sediment geochemistry, iron oxide mineralogy and geochemistry, Ar/Ar ages of hornblende grains, and radiogenic isotope signatures of a circum-Antarctic sample set. This approach will allow the identification of the geographic regions from which Antarctic sediment was supplied to the Southern Ocean, and thus constrain ice sheet dynamics and behavior of surface currents. Ice proximal samples from West Antarctica, East Antarctica and the Antarctic Peninsula were collected from fjords, bays, and inner shelf basins. Subsamples were taken from piston cores, kasten cores, and surface grabs collected over the past 40 years by the U.S. Antarctic Program. Here we present preliminary results from the Wilkes Land margin, Marie Byrd Land, and the western Antarctic Peninsula, which constitute an Archean craton, Paleozoic to Cenozoic sequences, and a mixed though relatively younger bedrock geology, respectively. The Nd isotope composition of the < 63- micron fraction generally conforms to the results from a survey of circum-Antarctic distal samples (Roy et al., Chemical Geology, submitted). However, the George V Coast samples (143 E longitude) yielded eNd values of approximately -24, the lowest values yet found in our survey, implying very old Nd crustal residence ages. The hornblende grains from these East Antarctic samples are irradiated and analyses are underway. The Roy et al. survey indicates that sediments from Wilkes Land are distinct from the rest of East Antarctic samples, with large peaks in the age spectrum at 1500-1800 Ma and 1100-1300 Ma, smaller peaks at ~400-600 Ma and 0-200 Ma. Ar/Ar ages of individual hornblendes separated from diamict samples from the West Antarctic margin and Antarctic Peninsula geologic terranes are younger than 250 Ma, and they cluster in populations that are distinct in different areas. Other analyses in progress include lithological clast identification, bulk geochemistry of the <63-micron fraction, and geochemical fingerprinting of detrital iron oxide grains. Accordingly it appears that with a more comprehensive survey it may be possible to discern in detail the regional contributions from different flow lines within Antarctica's ice sheets.
H33E-06
Use of Single Grain 40Ar/39Ar Mica Ages for Provenance Studies of Loess, Long Island, New York
In this provenance study we used single-step laser fusion 40Ar/39Ar ages of single grains of muscovite and biotite in loess and sand to evaluate its potential for distinguishing provenance sources. Long Island provides a good setting to evaluate this method for provenance because, in the potential local sources of loess north of Long Island, the cooling ages for mica in the basement rocks change systematically from 200 Ma in eastern Massachusetts and Rhode Island, to 300 and 400 Ma in western Connecticut and western Massachusetts, to 800 Ma in New York, and to older than 1,000 Ma further to the west in North America. Muscovite is common in our loess samples and there is reasonable published coverage of 40Ar/39Ar and K/Ar ages of bedrock in the region. The deposition of loess on Long Island probably started as soon as the ice sheet retreated. The most likely sources of the sediment were stream deposits on the exposed lake beds of the drained Proglacial Lake Connecticut and Lake Hitchcock north of Long Island. Loess and sand samples were collected along the north shore and on the south fork of Long Island. Additionally, a sample was collected from the Lake Hitchcock bottom deposit. 40Ar/39Ar ages of muscovite in loess and eolian sand from the north shore of Long Island have a mode between 300 Ma and 400 Ma corresponding to the source rock in western Connecticut and western Massachusetts. In contrast, muscovite from the South Fork of Long Island has a mode between 200 Ma and 300 Ma suggesting derivation from a more easterly source in eastern Massachusetts and Rhode Island. Muscovite from Lake Hitchcock bottom deposits has a bimodal distribution of 40Ar/39Ar ages with a mode younger than 300 Ma and a mode between 300 Ma and 400 Ma. Biotite 40Ar/39Ar ages from the same sample, however, are overwhelmingly younger than 320 Ma.
H33E-07
Linking the End of Glaciation in Gondwana to Aridity in the Tropics: Coupled Sr Chemostratigraphy and Cyclostratigraphy From the Permian Basin, Texas
Crowell (1978) suggested the Earth was poised on the brink of glaciation throughout the Paleozoic and that closure of the low-latitude seaway between North America and Europe was responsible for diversion of moisture- laden currents to the south to feed the long-lived Carboniferous-Permian glaciers. This same diversion of currents also produced dramatic aridity in the tropics. This is seen in changes in paleosol types as well as in widespread loess and erg deposits. It also coincides with a dramatic decline in Sr isotopes across the Carboniferous-Permian boundary. Sr chemostratigraphy from cores taken across this boundary on the Central Basin Platform (CBP) of the Permian Basin in Texas show the same dramatic shift as seen in data from the type sections in the Urals across this boundary. This confirms that the fusulinid based stage boundaries for the Permian Basin are correct. On the CBP, high-frequency high-amplitude cycles end at a major stepback of the shelf margin at Abo/Clearfork time (Sakmarian). We suggest this transgression represents the end of major Gondwanan glaciation. If we assume the decline in 87Sr/86Sr is reflecting the aridity of the topics and calculate the reduction in continental Sr flux to the oceans, we can relate this to decreased silicate weathering and the consequent increase in atmospheric CO2. Assuming crustal values of [Sr] of 350 ppm, congruent weathering, that the total number of moles of Sr in the ocean is the same as today, and that all of the change is due to a change in flux of Sr from the continents, the change from a late Carboniferous 87Sr/86Sr high of 0.7082 to a value of 0.7078 (the end of cyclothems), equates to an increase in the ocean-atmosphere system of approximately 4 * 106 gigatons of carbon. Preindustrial carbon concentration in the atmosphere is estimated at 578 gigatons. Obviously there are feedbacks that will remove some of the carbon from the ocean-atmosphere system, perhaps to the biosphere, but this simple calculation shows that no special circumstances are needed to account for the increase in pCO2 shown by published proxy data. It also demonstrates why a model that only considers tectonic changes might fail to reproduce this part of the Sr curve. However, the 87Sr/86Sr continues to decrease to a late Permian low of 0.7069 without reversal while the proxy data show a decrease in pCO2 in the middle Permian to near Carboniferous values. Nevertheless, the change is coincident with a flattening in the most recently published Permian Sr curve.