Equilibrium thermodynamic modelling is commonly used to simulate garnet chemical zoning and reconstruct metamorphic pressure–temperature (P–T) histories, but the reaction kinetic effects may be overlooked. We integrate petrography, garnet chemical zoning, and thermodynamic calculations for metapelitic samples from the garnet and staurolite zones of Longmenshan (SW China) to examine the effects of delayed progress of the staurolite-forming reaction. In the staurolite-free garnet‑zone sample, suppressing staurolite yields a better fit to the chemical zoning and size of the largest garnet, whereas the staurolite-involved model predicts staurolite and fits poorly near the rim. In staurolite‑zone sample 20SC37, garnet records two growth segments: Grt 1 formed before staurolite and Grt 2 after staurolite, as revealed by two textural relationships: Grt 2 is absent on staurolite-contacting faces but preserved on matrix-facing faces, and staurolite inclusions only occur in Grt 2. The staurolite-involved model terminates Grt 1 growth at a smaller size than observed, whereas the staurolite-suppressed model better reproduces Grt 1 zoning and size. The initial low-Ca composition of Grt 2 cannot be reproduced from the Grt 1-fractionated bulk composition, indicating a change in effective reactive composition after staurolite formation. An adjacent staurolite-zone sample 20SC42 preserves the same Grt 1–staurolite–Grt 2 textural relation, and isopleths for the core composition of its largest garnet crystal already intersect in a staurolite-bearing equilibrium field. Conventional thermobarometry yields P–T conditions consistent with the phase equilibrium and garnet growth modelling. These results support overstepping of the staurolite-forming reaction and show that a calculated phase-in boundary should not be treated automatically as the point where the natural reaction proceeded. This highlights the kinetic effect of a later phase-in reaction on an already growing garnet population during metamorphism.
Partial melting of the lower continental crust during high-temperature, low-pressure metamorphism is commonly associated with the emplacement of felsic plutons in the middle to upper crust. However, the duration and conditions of anatexis, melt extraction, and emplacement processes in these settings remain uncertain. The El Oro Complex in southern Ecuador preserves a kilometre-scale, tilted crustal section comprising a structurally continuous Triassic sequence of metamorphosed sedimentary rocks and associated S-type granitoids. A large number per sample of combined in situ zircon U–Pb dates and trace-element compositions has been obtained by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) from both migmatites in the lower paleo-crust and granitoids in the middle to upper paleo-crust. This approach, combined with cathodoluminescence imaging, is used to distinguish different zircon populations within individual samples beyond U–Pb analytical uncertainty. All samples from both migmatites and granitoids mostly record zircon crystallisation ages between ca. 240 and 225 Ma. In the migmatites, trace-element heterogeneity in zircon is observed at the sample scale. Depending on the sample, this variability reflects either changes in the melting reactions or open-system behaviour involving melt influx from other sources at the scale of individual leucosomes. U–Pb dates and Zr-thermometry data from titanites using LA-ICP-MS were also collected from the Piedras metagabbro, commonly interpreted as the heat source for the sequence’s metamorphism. Titanite grains record a consistent age of 227.5 ± 4.5 (n = 94) Ma with a temperature of 720–740 °C, and are interpreted as dating the cooling of the gabbro emplacement related to the end of metamorphism. Together, these data confirm the co-genetic link between the migmatites and the granitoids of the El Oro complex and the interpretation that the Piedras unit is the main heat source for the metamorphism. This study showcases important trace-element variability in zircons at the sample scale, both in migmatites and granitoids, reflecting the interplay between partial melting, melt extraction, and subsequent crystallisation.
The Yarlung-Tsangpo ophiolites (YTO) preserve discontinuous oceanic crustal sections in which volcanic and intrusive rocks are locally associated with serpentinized mantle peridotites. Here we report field and petrographic observations, whole-rock geochemistry, mineral geochemistry, and Sr–Nd-Hf isotopic data for boninite-like high-Mg lavas and mafic rocks from the Deji ophiolite in the central YTO, southern Tibet. Most lavas in the Deji volcanic sequence have N-MORB-like compositions, whereas the boninite-like high-Mg lavas appear to form a subordinate component. These high-Mg lavas have relatively high MgO and SiO2, low TiO2, high Cr and Ni contents, and depleted incompatible element abundances. However, they contain abundant clinopyroxene phenocrysts and lack olivine and orthopyroxene phenocrysts, and are therefore referred to as boninite-like high Mg lavas rather than true boninites. Andesitic diabase sills intrude the volcanic sequence and have trace element characteristics with N-MORB affinity. Olivine gabbronorites preserve poikilitic textures and hydrous mineral assemblages, whereas mantle-hosted gabbroic veins record clinopyroxene-dominated crystallization and variable hydrothermal overprinting. Clinopyroxene equilibrium melts calculated from the volcanic and intrusive rocks show broadly N-MORB REE patterns, although those of the boninite-like high-Mg lavas have lower REE abundances. Whole-rock Nd-Hf isotopes indicate broadly similar depleted mantle sources, whereas variable Sr isotopes mainly reflect fluid-related overprinting. We interpret the Deji crustal rocks as products of an evolving magma-poor oceanic magmatic system dominated by N-MORB-like melts, with local generation of more depleted boninite-like high-Mg melts. In our preferred model, hydrothermal fluid infiltration along detachment-related faults hydrated the shallow mantle. Later asthenospheric upwelling and mafic melt input promoted localized deserpentinization, fluid release, and limited hydrous melting of refractory harzburgite. This process may have produced small volumes of boninite-like high-Mg melts without requiring the whole volcanic sequence to be boninitic. The Deji ophiolite thus provides a useful case study of coupled magmatic, tectonic, and hydrothermal processes during magma-poor oceanic crustal accretion.
Misti volcano in southern Peru is an active composite cone < 18 km away from the country’s second largest city, Arequipa (population > 1.1 million). Recent studies detail the volcano’s tephrostratigraphy, physical volcanology, and petrology during the Pacheco stage (≤ 21–2 ka) to better understand future signs of volcanic unrest and activity. Here, we apply Mg, Sr, and Ba diffusion chronometry to plagioclase in pumice from three tephra-fall deposits emplaced by Pacheco stage eruptions to estimate recharge-to-eruption timescales. Modeled chemical gradients identified in plagioclase zones near crystal rims in all three deposits are interpreted to have formed via magma recharge and mixing, which also provided the final impetus for eruption to occur. Tephra-fall deposits were selected to represent the range of inferred pre-eruptive processes during this stage, including melt homogenization between mixing magmas that is complete (magma hybridization, Autopista deposit) and incomplete (magma mingling, La Rosada and Ponche Iridescente deposits). Modeled Mg, Sr, and Ba timescales range from 2 h to 268 years (La Rosada; n = 80), 48 h to 201 years (Ponche Iridescente; n = 39), and 6 h to 327 years (Autopista; n = 54). Only 6
