USGS Science⌕ Search
Home/Archive/USGS

USGS · 70015418

Enrichment of trace elements in garnet amphibolites from a paleo-subduction zone: Catalina Schist, southern California

Sorena S. Sorensen·J. N. Grossman
View original source ↗Export metadata ↓

Abstract

The abundance, P - T stability, solubility, and element-partitioning behavior of minerals such as rutile, garnet, sphene, apatite, zircon, zoisite, and allanite are critical variables in models for mass transfer from the slab to the mantle wedge in deep regions of subduction zones. The influence of these minerals on the composition of subduction-related magmas has been inferred (and disputed) from inverse modelling of the geochemistry of island-arc basalt, or by experiment. Although direct samples of the dehydration + partial-melting region of a mature subduction zone have not been reported from subduction complexes, garnet amphibolites from melanges of circumpacific and Caribbean blueschist terranes reflect high T (>600°C) conditions in shallower regions. Such rocks record geochemical processes that affected deep-seated, high- T portions of paleo-subduction zones. In the Catalina Schist, a subduction-zone metamorphic terrane of southern California, metasomatized and migmatitic garnet amphibolites occur as blocks in a matrix of meta-ultramafic rocks. This mafic and ultramafic complex may represent either slab-derived material accreted to the mantle wedge of a nascent subduction zone or a portion of a shear zone closely related to the slab-mantle wedge contact, or both. The trace-element geochemistry of the complex and the distribution of trace elements among the minerals of garnet amphibolites were studied by INAA, XRF, electron microprobe, and SEM. In order of increasing alteration from a probable metabasalt protolith, three common types of garnet amphibolite blocks in the Catalina Schist are: (1) non-migmatitic, clinopyroxene-bearing blocks, which are compositionally similar to MORB that has lost an albite component; (2) garnet-amphibolite blocks, which have rinds that reflect local interaction between metabasite, metaperidotite, and fluid; and (3) migmatites that are extremely enriched in Th, HFSE, LREE, and other trace elements. These trace-element enrichments are mineralogically controlled by rutile, garnet, sphene, apatite, zircon, zoisite, and allanite. Alkali and alkaline earth elements are much less enriched in the solid assemblage, and thus appear to be decoupled from the other elements in the inferred metasomatic process(es). The compositions of migmatitic garnet amphibolite blocks seem to complement that of “average” island-arc tholeiite. Trace-element metasomatism reflects fluid-solid, rather than melt-solid, interaction. The metasomatic effects indicate that H 2 O-rich fluid, perhaps with a significant component of Na-Al silicate and alkalis, carried Th, U, Sr, REE, and HFSE. Fractionations of LREE in migmatites resemble those of migmatitic metasedimentary rocks underlying the mafic and ultramafic complex. “Exotic” LREE deposited in allanite in migmatites could have been derived from fluids in equilibrium with subducted sediment. If the paleo-subduction zone represented by the mafic and ultramafic complex of the Catalina Schist had continued its thermal and fluid evolution, a selvage of similarly enriched rocks might have been generated along the slab-mantle wedge contact between ~30 and 85 km depth. Rocks affected by “subduction-zone metasomatism,” although rarely recognized at the surface, could be volumetrically significant products of the initiation of subduction and may prove to be geochemical probes of convergent margins that approach the significance of xenoliths in the study of other magmatic environments.

Explore related subjects

Geochimica et Cosmochimica ActaArticle

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Sorena S. Sorensen, J. N. Grossman. 1989. Enrichment of trace elements in garnet amphibolites from a paleo-subduction zone: Catalina Schist, southern California. https://doi.org/10.1016/0016-7037(89)90096-3

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2026

Organic and isotopic indicators for sorting of sedimentary organic matter along a marginal submarine canyon

Submarine canyons are incised features of many continental margins that can have significant influence on the hydrodynamic distribution of sediments and organic matter eroded and deposited from the continents. Baltimore Canyon, on the mid-Atlantic margin of the United States, contains a complex set of sedimentary processes that simultaneously create unique benthic habitats and control the deposition of organic matter. Along the canyon axis, loci of net erosion, net deposition, and intense winnowing each host diverse faunal assemblages and varying mixtures of sedimentary organic matter derived both from production in the overlying water column and from mobilized sediments. Bioavailable components of this deposited organic matter sustain benthic communities, while recalcitrant components can contribute to long-term carbon burial in the deep sea. However, commonly employed bulk geochemical analyses provide little information about the relative bioavailability or depositional history of sedimentary organic matter. Here we employ a range of organic and isotopic analyses to explore in more detail how canyon-specific sediment dynamics determine the sorting of organic matter from shelf to open ocean. In combination with bulk geochemical characteristics, we subjected surface sediments from water depths of ∼200–1200 m in Baltimore Canyon to a sequential extraction procedure, isolating nonpolar and polar lipid classes, an acid-soluble fraction, and an acid-insoluble fraction. Each class was analyzed for carbon and nitrogen quantities and stable isotope ratios, and radiocarbon content where possible, along with compound-specific carbon and nitrogen isotope analysis of individual amino acids in the acid-hydrolysed fraction. We find different organic matter sources and depositional history recorded in the properties of younger, bioavailable organic matter components (polar lipids, amino acids) in comparison to the older, more recalcitrant components (nonpolar lipids, acid-insoluble fraction). These differences in source and bioavailability of organic matter vary along the canyon, correlating with grain size and erosion/deposition dynamics, and may help shape the benthic faunal assemblages. Additionally, our results suggest that determining the relative concentrations of acid-soluble and acid-insoluble organic matter may provide an easily accessible method to improve our understanding of the nutritional quality of sediments for benthic fauna than more commonly used bulk carbon or nitrogen concentrations.

Delaware, New Jersey↗
USGS2025

Bayesian calibration of the 40K decay scheme with implications for 40K-based geochronology

The K/Ar and 40 Ar/ 39 Ar geochronometers are based on the naturally occurring radionuclide 40 K. Their precision and accuracy are limited by uncertainties on the 40 K decay constants and, in the case of the 40 Ar/ 39 Ar geochronometer, the isotopic composition of neutron fluence monitors. To address these limitations, we introduce a Bayesian calibration of the 40 K decay scheme. We formulate robust priors for all model parameters including partial 40 K decay constants, 238 U and 235 U decay constants, and age offset parameters to account for phenomena that can perturb apparent U-Pb and 40 Ar/ 39 Ar ages. We then harness a set of complementary 40 Ar/ 39 Ar, 238 U/ 206 Pb, and 235 U/ 207 Pb data from well- characterized geological samples with ages from 1.919 ka to 2000 Ma to derive Bayesian estimates of the 40 K decay constants. Posterior values for the partial 40 K decay constants are λ β - "> λ β - = (4.9252 ± "> ± 0.0054) × "> × 10 −10 yr −1 , λ β + "> λ β + = (5.6658 ± "> ± 0.1543) × "> × 10 −15 yr −1 , λ EC ∗ "> λ EC0 = (5.7404 ± "> ± 0.0053) × "> × 10 −11 yr −1 , and λ EC 0 "> λ EC0 = (4.9060 ± "> ± 0.2942) × "> × 10 −13 yr −1 (uncertainties reported at the 68 % (1 σ "> σ ) credible interval). These combine to a total 40 K decay constant λ tot "> λ tot = (5.5042 ± "> ± 0.0054) × "> × 10 −10 yr −1 . Model estimates of the 238 U and 235 U decay constants are statistically indistinguishable from those reported by Jaffey et al. (1971) . Posterior values of the 40 K decay constants and the 40 Ar*/ 40 K isotopic composition of Fish Canyon sanidine (FCs) define a K/Ar FCs age of 28.183 ± "> ± 0.017 Ma (1 σ "> σ ). Significantly, Bayesian calibrated 40 Ar/ 39 Ar ages align with astronomically tuned ages throughout the Cenozoic and with 238 U/ 206 Pb and 235 U/ 207 Pb ages in the Mesozoic, Paleozoic, and Proterozoic, as well as having comparable precision to the 238 U/ 206 Pb method. Thus, Bayesian calibration of the 40 K decay scheme and the K/Ar age of FCs reconciles the 40 Ar/ 39 Ar, U-Pb, and astronomical chronometers.

Geochimica et Cosmochimica Acta↗
USGS2024

Characterizing sulfur redox state and geochemical implications in deep-time using mineral chemistry network analysis

Sulfur (S) is a central element in global biogeochemical cycling and Earth’s redox evolution. Minerals that contain S are an important record of local environmental conditions at the time of their formation based on chemical speciation and redox. However, the oxidation state of S for hundreds of different S-containing minerals and thousands of S-containing mineral localities is unknown, largely sulfides and sulfosalts, and the redox state alone does not fully capture mineral chemistry diversity, thus limiting understanding of S redox evolution. Here, we use mineral chemistry network analysis and the weighted Mineral Element Electronegativity Coefficient of Variation (wMEE CV ) metric to investigate the element interactions and localities of S-containing minerals from the Mineral Evolution Database (MED) to infer the redox state of S in minerals where the redox state is unknown (S U ). Louvain community detection of the S mineral chemistry redox network reveals that there are three main network communities that are separated by redox state. The S 6+ community includes minerals that contain the S 6+ redox state and a small number of S 4+ and S 2+ minerals, the S 2− community includes S 2− -containing minerals, and the S U community includes minerals in which the redox state of S is unknown. The wMEE CV values of the S U community closely overlap with the wMEE CV values of the S 2− community, and do not overlap with the wMEE CV values of the S 6+ community, indicating the S U community minerals contain predominately reduced S. Assuming that S U community minerals contain reduced S, as supported by their network chemical associations and wMEE CV values, then reduced S-containing minerals make up approximately 81 % of S-containing mineral localities in the S mineral chemistry network, even though the majority of all mineral localities (S-containing and non-S-containing) are oxygen (O)-containing minerals. Additionally, reduced S-containing minerals make up the majority (∼75 %) of all non-O containing mineral localities in the MED, representing the importance of reduced S as an electron source and substrate in the evolution of microbial metabolic networks. The range wMEE CV values of S 6+ community minerals expands through time due primarily to formation of chemically diverse sulfate minerals, coinciding with crustal oxidation from the late Proterozoic to Phanerozoic and the expansion of the marine sulfate reservoir. The intersection of shared constituent elements among reduced and oxidized S in the mineral chemistry network represents redox convergence of weathered S in the geosphere that was crucial in the formation of natural resource deposits and the evolution of biogeochemical cycles.

Geochimica et Cosmochimica Acta↗