The timing of the India-Asia collision and the multi-stage tectonic control on the development of Himalayan sedimentary basins has remained controversial for over half a century. In particular, it remains unclear whether the stratigraphic record of the Indus Suture Zone (ISZ) in Ladakh, NW India, 1) preserves pre-collisional trench deposits, 2) whether the cessation of marine sedimentation reliably marks continental collision, and 3) how provenance and depositional ages inform on the geodynamic evolution of the orogeny. This study re-examines the tectonostratigraphic evolution of the Indus Basin in the region by leveraging new biostratigraphic data (radiolarians, dinoflagellates) and detrital zircon U-Pb ages. We further integrate wholerock and clay-size fraction geochemical and Nd isotopic data to constrain provenance changes and their relationship to major tectonic events pertaining to the collision between India and the Kohistan-Ladakh intra-oceanic arc (KLA). Results show that the stratigraphic record in the region exclusively preserves the Paleogene-Neogene syn- to post-collisional foreland sedimentation (< ca. 54 Ma). Deposition, sediment supply, and drainage patterns were largely controlled by 1) the Spongtang ophiolite obduction (64-42 Ma) onto the Indian margin, 2) the collisioninduced uplift of the KLA (53-40 Ma), and 3) the exhumation of the Tso Morari complex (45 Ma). Importantly, the diachronous cessation of plain marine conditions between the northern (50 Ma) and southern (40 Ma) margins of the Eocene foredeep warns against relying on the last occurrence of marine units as a uniquevocal marker for continental collision onset. Instead, the oldest foreland deposits only provide a minimum age constraint (ca. 54 Ma) for the India-KLA collision. These results highlight the relevance of applying a multi-proxy approach for reconstructing the evolution of collisional orogens. Furthermore, we demonstrate that palynological analyses in the Himalayas and adjacent basins constitute an effective dating method to accurately time tectonostratigraphic events.

Fossil arcs which expose the middle-lower crust have the potential to shed light on active arc processes because magmatic rocks from the entire history of the arc can be examined. We present new geochemical data from the middle crust of the Kohistan Arc and use existing geochemical and geochronological literature data from this arc to constrain the growth and chemical evolution of the arc crust by reconstructing the chemical evolution of intrusive and extrusive magmatism. The ultramafic-mafic lower crust of the Kohistan Arc is older than much of the intermediate-felsic middle-upper crust. The middle-upper crust of the Kohistan Arc has higher (Ce/Yb)N and lower 143Nd/144Nd ratios than the older lower crust. The later intermediate to felsic plutons of this arc were formed by fractional crystallization of basaltic magmas. The Kohistan Arc formed in two main stages: (a) subduction initiation and early intra-oceanic arc (e.g., Izu-Bonin-Mariana) magmatism (155-100 Ma) that formed the lower crust; and (b) mature intra-oceanic arc (85-60 Ma) magmatism (e.g., Cascade Arc) followed by collisional (60-40 Ma) magmatism that generated the upper-middle crust. Stage 1 volcanic and plutonic rocks have low (Ce/Yb)N ratios, whereas their Stage 2 equivalents have high (Ce/Yb)N ratios. The (Ce/Yb)N ratios and fractionation depths of the volcanic and plutonic rocks of the Kohistan Arc increased with decreasing age, reflecting the progressive maturation and crustal thickening of the arc. Lavas erupted at mature (similar to 100 Ma old) entirely intra-oceanic arcs are geochemically similar to those found at "continental" arcs (e.g., Cascade Arc).

Tellurium is a highly volatile, chalcophile and moderately siderophile trace element that is strongly enriched in volcanic gases relative to crustal rocks. Like mercury, tellurium concentrations in sediments can therefore represent a proxy for past volcanic activity, allowing the timing of LIP volcanism relative to environmental and biotic change during mass extinction events to be determined. Previous studies reported high Te contents in sedimentary rocks at the Permian-Triassic, Cretaceous-Paleogene and Paleocene-Eocene boundaries, which may be linked to eruption of the Siberian, Deccan, and North Atlantic flood basalts, respectively.Due to the low abundance of Te in most geological materials, and the relatively high ionization energy of Te, this element is rarely analyzed and its geochemical behavior is poorly understood. We have developed methods for analysis of nanogram amounts of Te (and other trace elements) using desolvating nebulizer ICP-MS. Addition of a single-step cationic exchange preconcentration allows analysis of samples containing ppt levels of Te. Using these methods, we carried out analyses of different geological materials, in order to advance our understanding of the behavior of Te in volcanic and sedimentary systems and assess its potential as a proxy for volcanic activity.Glacial diamictite composites, previously used to estimate the average composition of the Upper Continental Crust (UCC), yield an average Te concentration of 36.7 ± 0.5 ng/g. Assuming this is representative of average UCC, this enrichment in Te relative to estimates of the primitive mantle (silicate Earth) of about 12 ng/g, despite tellurium’s moderately compatible behavior during mantle melting, may indicate that Te has been concentrated in the UCC due to volcanic and hydrothermal processes.Deccan flood basalts that have not fractionated sulfide, have low Te concentrations (average 0.94 ppb, n=12) relative to MORB (3 – 5 ppb), suggesting that Te was largely degassed during emplacement of the subaerial Deccan lavas at 66.5 – 65.5 Ma. By contrast, the red boles (fossil soil horizons) interbedded with Deccan lavas, have high Te concentrations of up to 2200 ppb, indicating that significant amounts of Te were released during volcanism, some of which was deposited close to the site of volcanism. This observation agrees with data of several thousand sedimentary rocks from profiles across the K-Pg boundary in Italy, Egypt, Morocco, Turkey and Spain, thus supporting the use of Te as a geochemical proxy for LIP volcanism.

Granitoid magmas are abundant in subduction zones and form large portions of the upper continental crust. However, the formation of granitoid magmas remains debated, with models proposing (1) evolution from mantle-derived mafic melts by assimilation-fractional crystallization (AFC), (2) partial melting of lower to middle continental crust induced by mafic underplating, and (3) partial melting of metasomatized pyroxenite in the mantle. Since the Oligocene, slab rollback and trench retreat have caused the Aegean subduction zone to migrate approximately 350 km southwestward, resulting in extensive mafic to felsic magmatism with numerous granitoid intrusions in the region. We present new whole-rock major and trace element data together with Sr-Nd-Pb isotope compositions for the 15 to 8 Ma granitoids from Tinos, Mykonos, Naxos, Paros, Lavrion, and Serifos, as well as metasedimentary rocks of the Cycladic Blueschist Unit (CBU) from Tinos, Syros, Andros, and Sifnos. The CBU metamorphic rocks comprise low-grade metamorphic schists, marbles, and high-pressure mélanges and were subducted at the Aegean subduction zone. The metasediments received a high-pressure metamorphic imprint between 55 and 30 Ma. They exhibit element compositions similar to modern Eastern Mediterranean sediments, but many have higher initial 207Pb/204Pb and 208Pb/204Pb than the sediments from the Hellenic Trench. These differences indicate that the composition of subducted sediments changed over time at the Aegean subduction zone. Most granitoids display geochemical signatures characteristic of arc magmas and represent an isotopic end-member of Aegean magmatism in Sr-Nd-Pb isotope space. The isotopic compositions of many granitoids overlap with those of sediments and CBU metasediments, whereas others display distinctly more radiogenic (Pb) signatures. The Cyclades Continental Basement has much higher Sr isotope ratios than the granitoids. Consequently, the isotope composition of the granitoids does not support partial melting of lower continental crustal rocks. Partial melting of metasomatized pyroxenite is unlikely, as most granitoids lie on fractional crystallization trends. The high Th/Nd and low Ce/Pb of the granitoids indicate a fractionation of these elements by accessory minerals during partial melting of the upper crustal rocks. We propose that most granitoid magmas in the Aegean form by fractional crystallization of mafic magmas derived from mantle sources modified by subducted upper continental crustal components. The granitoids require a more radiogenic (Sr and Pb) subducted component than observed in the CBU metasediments or modern sediments, possibly related to the subduction of microcontinental fragments.

The fore-arc crust at subduction zones is generally thought to be considerably older than the overlying arc, and to have formed during subduction initiation. We present new Ar-40/Ar-39 ages as well as geochemical and petrological data for volcanic rocks from a 500 m-high section along the wall of a rifted basin within the northern Tonga Ridge fore-arc. Compared to lavas from the adjacent Tonga Arc, the samples have relatively high SiO2 and low TiO2 at a given MgO, and the least evolved samples have boninitic bulk-rock compositions. Ar-40/Ar-39 ages of three samples are about 1.3 Ma, which is younger than the initiation of subduction in northern Tonga (possibly 20 Ma). All lavas are depleted in incompatible trace elements but enriched in fluid-mobile elements as a result of slab dehydration beneath the fore-arc. Although lavas from the adjacent arc front at this latitude have been proposed to contain Pb from the subducting Louisville Seamount Chain (LSC), the unradiogenic Pb isotope compositions (Pb-206/Pb-204 < 18.9) of the older fore-arc lavas indicate that they do not contain a contribution from LSC materials. A dike intruding the lavas is relatively enriched in fluid-immobile Nb and Ta and has higher Nb/Zr ratios than N-MORB. The Nd and Hf isotope compositions resemble those of rejuvenated Samoa lavas, and we propose that Samoa mantle flowed into fore-arc mantle after eruption of the 61ROV lavas at about 1.3 Ma. Our results show that not all boninites (including high-Si boninites) in fore-arc locations are related to subduction initiation.

The interplay between Large Igneous Provinces (LIPs) and significant mass extinctions has been widely acknowledged for a long time (e.g., Courtillot and Renne, 2003). What makes the K/Pg boundary (KPB) extinction significant is the simultaneous occurrence of two major catastrophic events – the Deccan volcanism and the Chicxulub impact – within a very brief timeframe (Schoene et al., 2019). To better understand the link between volcanic eruptions and the resulting environmental stress, it's crucial to delve into the impact of the Deccan volcanic activity on the ecosystem. To uncover this phenomenon, our focus lies in detailing this relationship using detailed counts of species along with isotope and geochemical analyses conducted on two complete sections within the Mudurnu-Göynük and Haymana basins in central Anatolia (Turkey).Over the late Maastrichtian period, our examination of δ13C measurements in the Haymana Basin exhibits cyclical patterns that underscore the influence of precession cycles on the δ13C record. Intriguingly, each cycle concludes with a sudden cooling event (a positive shift in δ18O values). A quantitative assessment of planktic foraminifera, on the other hand, shows a continual decline in species diversity throughout the late Maastrichtian (Karabeyoglu et al., 2019). This decline seems to accelerate just before reaching the K/Pg boundary. In the Göynük and Okçular sections, this decline aligns with distinct intervals of low magnetic susceptibility, hinting at a possible event of ocean acidification during the late Maastrichtian.The K/Pg boundary itself is identifiable by a reddish oxidized layer measuring 2-3 mm in thickness. This layer provides evidence of a sequence of events: the abrupt disappearance of large, specialized ecological specialists (such as globotruncanids, racemiguembelinids, planoglobulinids), a surge in mercury (Hg), and increased levels of trace elements (e.g., Iridium (Ir), Tellurium (Te), Nickel (Ni), Chromium (Cr), and Cobalt (Co)). Notably, the correlation between Hg/Te suggests that Te might serve as a proxy for volcanic activity. In terms of the faunal record, we observed peaks in Thoracosphaera and Guembelitria cretacea, signifying an ecosystem collapse following the KPB.In summary, our comprehensive examination of paleontological, isotopic, and geochemical data indicates that the detrimental impacts of Deccan volcanism had already begun prior to the Chicxulub impact. This predisposed the fauna to an eventual extinction event at the K/Pg boundary.ReferencesCourtillot, V.E., Renne, P.R. 2003. On the ages of flood basalt events. Comptes Rendus Geoscience, 335, 113–140.Schoene, B., Eddy, M.P., Samperton, K.M., Keller, C.B., Keller, G., Adatte, T., Khadri, S.F.R. 2019. U-Pb constraints on pulsed eruption of the Deccan Traps across the end-Cretaceous mass extinction. Science, 363, 862-866.Karabeyoglu, A.U., Özkan-Altıner, S., Altıner, D. 2019. Quantitative analysis of planktonic foraminifera across the Cretaceous-Paleogene transition and observations on the extinction horizon, Haymana Basin, Turkey. Cretaceous Research, 104, 104169.

Valuable understanding regarding the behaviour of chalcophile elements in magmatic systems can be gained by examining magmatic sulphide droplets, which are the solidified remnants of previously immiscible sulphide liquids [1-2]. We analysed the trace element composition of sulphide droplets by LA-ICP-MS, creating a comprehensive data set for volcanic and plutonic rocks from intra-oceanic arcs.We observed a significant enrichment of elements like Zn, Cd, Sn, Te, and Bi in sulphide droplets from lava samples compared to those hosted by gabbro xenoliths, which cannot be attributed to the fractionation of olivine, spinel, or magnetite [3-5]. These compositional differences are likely the result of changing sulphide droplet composition during cooling and solidification of the silicate melt. This process involves continuous sulphide segregation or re-equilibration with the silicate melt. A key aspect of our suggested model is sulphide droplets' resorption or partial remelting, particularly of the Cu-Fe-rich intermediate solid solution proportion. This process is driven by pressure decrease during magma ascent, leading to an increase in sulphur solubility in the silicate melt [6], which potentially liberates elements like Cu, Au, Zn, Bi, Te, and Ag from the sulphide droplet to the silicate melt. Our findings further suggest that subsequent magma stagnation and fractional crystallisation lead to a second stage of sulphide saturation, likely dominated by the elements previously liberated from the intermediate solid solution.The complex crystallisation history indicates that sulphide droplet formation during silicate melt evolution in subduction-related settings is a non-equilibrium process. We further propose that volatile saturation preceding the second stage of sulphide segregation from a silicate melt enriched in chalcophile elements liberated from intermediate solid solution could result in particularly metal-rich fluids (e.g., Cu, Au, Bi, Te) with a high ore-forming potential in magmatic-hydrothermal environments. [1] Wood, B. J. and Kiseeva, E. S. (2015), Earth and Planetary Science Letters, 424, 280-294. [2] Patten, C. et al. (2013), Chemical Geology, 358, 170–188. [3] Distler, V. V. et al., (1983), Initial Reports of the Deep Sea Drilling Project, 69, 607-617. [4] Keith, M. et al. (2017), Chemical Geology, 451, 67–77. [5] Schäfer, W. et al., in prep. [6] Peach et al. (1990), Geochimica et Cosmochimica Acta, 12, 3379-3389.

The mafic lavas erupted at active subduction zones have higher P/Nd ratios than the bulk continental crust. This difference has been explained by the delamination of apatite-rich lower crust, transferring P from the crust to the mantle at subduction zones. Here we compile geochemical data from arc basalts (5.5-6.5 wt% MgO) that have formed in the last 2 Ga and show that Proterozoic arcs erupted basalts with lower average P6/Nd6 ratios than those of Phanerozoic fossil and active arcs. The P6/Nd6 ratios of active arc basalts are negatively correlated with Th6/La6 and positively correlated with Sr6/Th6, Ba6/Th6 and 143Nd/144Nd, indicating that high P6/Nd6 ratios in arc lavas are not inherited from subducted sediments, which generally have low P/Nd ratios. Phosphatised basaltic oceanic crust has high P contents and P/Nd ratios due to interaction with seawater, and metalliferous sediments deposited at active spreading ridges have very high P contents (2000-16,000 ppm) and P/Nd ratios (100-1456) due to coprecipitation of P with Fe-oxyhydroxides formed when reducing hydrothermal vent fluids mix with oxygenated deep ocean water. We propose that the lower P/Nd ratios of Proterozoic mafic arc lavas are the result of lower concentrations of P in seawater at that time. An increase in the P content of seawater at the end of the Proterozoic, coincident with the Neoproterozoic Oxygenation Event, is reflected in lower concentrations of P in marine sediments before 800 Ma, and in 1.3 Ga metalliferous sediments. The high P/Nd ratios of many young arc lavas are therefore not representative of those formed during most of Earth's history. Since most of the continental crust was formed before the end of the Proterozoic, less delamination of apatite-bearing lower crust may be needed to the explain the low P/Nd ratio of the bulk continental crust.

Carbon cycle-climate dynamics were nonlinear through Earth’s history, driven by changes in internal and external forcing processes acting on various geological timescales. This study focuses on determining the relationship between volcanism, orbital parameters, and organic carbon burial during the Aalenian (Middle Jurassic) - a pivotal time at the dawn of the Mesozoic Marine Revolution, marked by a disruption of the carbon cycle and major climate shifts. Here, new high-resolution magnetic susceptibility and trace elements data are combined with previously published organic carbon isotopes and total organic carbon data from two sites in France and Chile. Our dataset shows for the first time a temporal coincidence between the major carbon cycle perturbation during the middle–late Aalenian and the onset of enhanced volcanic activity, suggesting a causality link. We propose that volcanic activity triggered a transient warming episode within the long-term Middle Jurassic coldhouse and played a key role in shifting organic carbon burial from the ocean to terrestrial settings. This period therefore contrasts with other Mesozoic carbon cycle perturbations, which generally record enhanced marine organic matter burial in oxygen-depleted environments during volcanism-triggered warming events.

Abstract Late Maastrichtian to Early Eocene sediments from Wadi Nukhul, Egypt were deposited between about 67 and 55.5 Ma during the eruption of the Deccan Traps and the North Atlantic Igneous Province (NAIP). We use Te and Hg as proxies for volcanism to constrain the timing of flood basalt volcanism relative to environmental perturbations and extinction events during this period. We find enrichment in Te in the latest Maastrichtian and earliest Danian and in the Late Paleocene to Early Eocene, which result from enhanced volcanic input of Te. An increase in volcanic Te during the Late Maastrichtian coincides with the Late Maastrichtian Warming Event. A second larger peak in Te begins about 120–80 kyr before the Cretaceous/Paleogene boundary (KPB) and continues into the Danian, with highest values about 70–30 kyr prior to the KPB, potentially related to eruptions of the voluminous Deccan Wai subgroup. The Chicxulub impact, therefore, did not trigger these eruptions, and Deccan volcanism likely led to climate instability, which may have amplified the environmental effects of the impact. A 6 Myr period of low Te during the Paleocene is followed by an increase in Te starting at 57.5 Ma and peaking at the Paleocene‐Eocene boundary (PEB; 56 Ma) during the opening of the North Atlantic and the highest eruptive rates of the NAIP. In contrast, Hg variations over the same time period are less systematic. Our results show that Te in sediments may be a robust proxy, complementary to Hg, for large volcanic events.

The Cretaceous-Paleogene (K-Pg; ~66 Ma) extinction and the Paleocene Eocene Thermal Maximum (PETM; ~56 Ma) were two major environmental and biotic crises in Earth’s history. While both coincide with the emplacement of Large Igneous Provinces (LIPs), uncertainties persist about the role of the Deccan Traps and the North Atlantic Igneous Province (NAIP) respectively in these events. In both cases, the reconstruction of cause and effect of environmental perturbations is hampered by the difficulty in determining the timing of volcanism relative to environmental change and extinction. The main phase of the Deccan volcanism initiated at ~66.5 Ma in C24n and lasted for about 1 Myr, overlapping with the Chicxulub impact at ~66.03 Ma and the Cretaceous-Paleogene boundary (KPB) extinction. In case of the NAIP, linking the eruptive history to the light carbon excursion (CIE) and rapid warming at the P-E boundary (~56.01 Ma) remains difficult, thus raising questions about the trigger for the CIE. Some studies found volcanic degassing to be sufficient to account for the CIE, in which case the main period of volcanic activity was short, and initiated at the Paleocene-Eocene boundary. Alternatively, the light carbon originates from magmatic tapping of carbon-rich sediments or destabilization of methane clathrates. In order to tie Large Igneous Province (LIP) volcanism to consequent environmental perturbations, we present tellurium (Te), mercury (Hg) and other trace element proxies from a complete sedimentary profile at Wadi Nukhul, Egypt spanning about 12 Myr from the late Maastrichtian to the early Eocene. A peak in Te in the latest Maastrichtian corresponds to the Late Maastrichtian Warming Event, and may be coincident with early Deccan volcanism on the Malwa Plateau. Te concentrations rise again up to 467 ppb immediately before the KPB, possibly reflecting eruptions of the massive Wai Formation at the Deccan Traps. Te concentrations are close to crustal average values (~10 ppb) throughout most of the Paleocene, and another spike of 465 ppb can be recognized at Paleocene – Eocene (PE) boundary. In contrast, the Hg record for this period is less clearly influenced by volcanism. Hg/TOC ratios peak in the Late Maastrichtian and earliest Eocene, but similarly high values occur throughout the section. Using a vast set of trace elements, we rule out changes in lithology, accessory mineral content, or changing redox and productivity conditions as controlling factors on Te concentrations. We use nannofossil zone ages to calculate sedimentation rates and elemental fluxes. We show that an increase in Te during the late Cretaceous coincides with the late Maastrichtian warming event and that the eruption of the Wai group initiated before the Chicxulub impact. Deccan volcanism likely contributed to climate instability and may have amplified the effects of bolide impact on the biotic crisis. Furthermore, an abrupt increase in Te concentrations coinciding with the opening of the North Atlantic at the P-E boundary, suggest highest volcanic degassing concurrent with the CIE. We, therefore, conclude that atmospheric injection of volcanic CO2 may have been the major driver of the negative CIE.

Climate models and paleo-reconstructions suggest that alterations in the Atlantic Meridional Overturning Circulation (AMOC) are not only indicators but also drivers of climate changes. Therefore, the AMOC is considered a critical tipping element within Earth’s climate system. Many lines of evidence indicate that the last glacial termination was characterised by large swings in AMOC strength, yet proxy evidence remains ambiguous about centennial-scale fluctuations during the Holocene. Inconsistencies persist regarding the timing, spatial pattern, and intensity of North Atlantic deep-water production. This study evaluates the variability of the AMOC during the Holocene based on several marine sediment cores covering the North Atlantic in high temporal resolution. For this, we exploit the 231Pa/230Th proxy, which indicates the bottom water advection strength. Additionally, past particle fluxes were reconstructed to determine a possible influence of particle composition and particle rain rate on the 231Pa/230Th signal. This study thus aims to extend existing paleo-circulation reconstructions of the AMOC from the last deglacial period with more recent analyses. Five new high-resolution 231Pa/230Th down-core records from different oceanographic settings and water depths in the North Atlantic consistently exhibit low variability throughout the entire Holocene. The 231Pa/230Th records generally display deviations of ± 10% from their respective Holocene mean. A generalised additive model (GAM) was fitted to the timeseries to detect mean North Atlantic trends within the different Holocene-normalised datasets. This model exhibits a virtually constant 231Pa/230Th level throughout the Holocene, interrupted by two time periods of slightly increased ratios, indicative of a weaker AMOC. The first time period is within the timeframe of the 8.2 ka event, characterised by a sudden cold spell across parts of the Northern Hemisphere. During this interval, four of the five timeseries show slightly elevated 231Pa/230Th ratios, although two records within this period hold a reduced sampling resolution. This limited temporal resolution and the shortness of the event make it challenging to decidedly conclude on the magnitude of the AMOC weakening during this time. The second period of higher 231Pa/230Th coincides with the 4.2 ka event and is only evident from the ODP 1063 data (Bermuda Rise). However, these higher 231Pa/230Th ratios can be explained by increased bottom scavenging of 231Pa presumably caused by benthic storms, induced by the transfer of eddy kinetic energy from the surface to the deep ocean. Consequently, atmospheric forcing during the 4.2 ka event seems to be a more plausible explanation than a paleoceanographic cause for the observed higher 231Pa/230Th. In conclusion, our study suggests that deep ocean circulation in the North Atlantic did not exhibit high variability on sub-millennial time scales, but has remained relatively stable throughout the Holocene.

Mercury (Hg) and more recently tellurium (Te) are indicator of large-scale volcanism in marine sediments and provide valuable insights into relative timing between biological and environmental changes, mass extinctions and delayed recovery. Several studies evaluated the relationship between Hg anomalies in sediments and LIP activity across mass extinction horizons. The bulk (80%) of Deccan Trap eruptions occurred over a relatively short time interval in magnetic polarity C29r. U-Pb zircon geochronology reveals the onset of this main eruption phase 350 ky before the Cretaceous-Tertiary (KT) mass extinction. Maximum eruption rates occurred before and after the K-Pg extinction, with one such pulse initiating tens of thousands of years prior to both the bolide impact and extinction, suggesting a cause-and-effect relationship. We present a comprehensive high-resolution analysis of Deccan Traps Hg-Te loading, climate change and end-Cretaceous (KPB) mass extinction from a transect, which includes 30 sections deposited in both shallow and deep environments located in France, Spain, Italia, Denmark, Israel and Tunisia. In all sections, our findings indicate that Hg concentrations are more than 2 orders of magnitude greater during the final 100ky of the Maastrichtian up to the early Danian P1a zone (first 380 Ky of the Paleocene). Notably, Hg anomalies generally show no correlation with clay or total organic carbon contents, suggesting that the mercury enrichments resulted from higher input of atmospheric Hg species into the marine realm, rather than being driven by organic matter scavenging and/or increased run-off. Significant and coeval Hg enrichments are observed in multiples basins characterized by proximal and distal, as well as shallow and deep-water settings, supporting a direct fallout from volcanic aerosols. Hg enrichments are not observed in the Indian redboles, confirming that it is not a proximal proxy for volcanism. But significant Hg anomalies have been found in more distal intertrapeans sediments at Anjar (Gujarat), Daiwal and Podgavan (SW Nagpur, Maharashtra). Significant Hg anomalies are also found in the more distal Megalaya section. Hg isotope data from Bidart confirm a direct Hg fallout from volcanic aerosols. Furthermore, Te/Th ratios measured in the Goniuk (Turkey), Elles (Tunisia), Gubbio (Italy) and Wadi Nukhul (Egypt) sections show the same trend as Hg/TOC and are consistent with a volcanic origin, albeit a minor extraterrestrial contribution of Hg at the boundary cannot be excluded. Hg and Te maximum loadings coincide with time of maximum Deccan emission rates and volumes determined by zircon dating. Hg and Te concentrations within sediments in conjunction with Te/Th and Hg/TOC ratios are therefore robust and useful proxies to trace intensity of volcanism.

Important insights into the behaviour of chalcophile elements in magmatic systems can be obtained from the study of immiscible sulphide liquids (SL) and crystalline monosulphide solid solutions (MSS) that are preserved as magmatic sulphides. Here, we report on the trace element compositions of >100 magmatic sulphides and the host volcanic glass from mid-ocean ridges, back-arc basins and intra-oceanic arcs. Most chalcophile elements (e.g., Cu, Zn, Se, Pb, Tl) are enriched in MSS hosted by mafic silicate melts from intra-oceanic arcs compared to those from back-arc basins and mid-ocean ridges, suggesting a contribution of these elements by subduction input. Partition coefficients of SL (D-SL/SM) and MSS (D-MSS/SM) relative to mafic silicate melts (SM) from mid-ocean ridges and back-arc basins are similar to previously published values (e.g., Ni, Cu, Zn, Se, Mo, Ag, Cd, In, Sn, Bi), thereby indicating equilibrium conditions. The new partition coefficients obtained in this study suggest that mid-ocean ridge and back-arc basin magmas typically segregate minor amounts of SL and MSS relative to the silicate melt portion (0.001-1 %), which nevertheless significantly modifies the concentrations of strongly chalcophile elements (DSL-MSS/SM > 100) of mantle melts during cooling and crystallisation. In contrast, SL and MSS from more evolved subduction zone magmas (>51 wt% SiO2) typically do not reach equilibrium conditions due to higher melt viscosity and a complex sulphide segregation history. This results in unrealistically low DSL-MSS/SM values for moderately to strongly chalcophile elements (e.g., Co, Ni, Mo, Sn), which cannot be explained by silicate and oxide fractionation accompanying the sulphide segregation. Therefore, we conclude that sulphide saturation in subduction environments commonly reflects a continuous or multistage process that involves the remelting or resorption of early formed MSS during magma ascent, which liberates chalcophile elements (e.g., Cu, Au, Zn, Sn, Cd, Bi, Te, Ag) into the silicate melt. Accompanied fluid release from such magmas provides a potential mechanism supporting the formation of magmatic-hydrothermal ore deposits. Consequently, a simple single-stage sulphide segregation model at equilibrium conditions, as suggested for mafic mid-ocean ridge magmas, cannot explain the chalcophile element evolution of more evolved silicate melts and genetically related magmatic sulphides in subduction environments.