Interpreting Volcaniclastic Sediments
Presiding: R Hickey-Vargas, Florida International University; I Savov, National Museum of Natural History, Smithsonian Institution
V52A-01 INVITED 10:30h
Fallout Tephra as Tool for Reconstructing the Geochemical Evolution of Arcs - a Case Study From Izu-Bonin-Mariana Arc/Trench System
While we are approaching an understanding of the currently active volcanic arcs, it is clear that major secular tectonic and chemical changes occur over the lifetime of arc systems. Such changes may substantially modify the arc outflux, and hence the impact of subduction zone volcanism on the global geochemical cycle. Volcanic arcs are mobile and dynamic systems that are subject to perpetual and rapid change. Therefore, a reliable reconstruction of the secular chemical evolution of arcs requires a temporally highly resolved rock record. Commonly, such a record cannot be obtained from lava series. However, it can be obtained from the distal fallout tephra that is generated from the explosive arc volcanism and that becomes embedded into the sediments of the adjacent basins. The unique geological potential of such fallout tephra is now evident from studies of the Izu Bonin Mariana arc (IBM), an intra oceanic arc in the NW Pacific that has been active since ~50 million years. Extensive drilling in the course of the DSDP and ODP Programs recovered both lava and tephra series. While the lavas record is disrupted by large gaps (>10 million years) of uncertain time and length, the fallout tephra provides a coherent, temporally highly resolved rock record (~1 Ma) of 40 million years. Fallout tephra are composed of variable proportions of volcanic glass shards, rock fragments (lithics) and phenocrysts. While alteration increases with time, fresh material is still present in tephra older than 40 million years. Careful separation of ash-sized particles prior to analysis and the usage of micro-analytical methods demonstrated that the basaltic through rhyolitic fallout tephra are the compositional counterparts of the IBM lavas series and thus provide a vital complement to studies of the temporal evolution of the IBM. The tephra profiles show that several influences are represented in the chemistry of the IBM. For one, steady-state processing is an important component of the IBM evolution, as shown by the constancy in major element oxides (except K2O) throughout most of the IBM's history. Notably, the major element chemistry appears to be largely unrelated to the expected crustal thickening with time. To the other, the temporal trends of trace elements and radiogenic isotopes demonstrate that the arc chemistry is responsive to changes of the arc input by slab and mantle wedge, and likely to transient tectonic changes as well (e.g. backarc basin formation). Furthermore, by means of the volcanic glass, we obtained data of elements and isotopes sensitive to the water fluxes through subduction zone (e.g. Cl, B, Li, B isotopes). Our results suggest that the time-dependent variations of these elements may allow for studying the role of 'subduction serpentinization', the subduction-related hydration of the upper mantle, in the waxing and waning of arcs. In summary, the IBM case study provides strong evidence that close linkages exist between the chemistry of arcs and the evolution of the Earth's surface reservoirs, and thus reinforces the role of arc volcanism in shaping the habitable environment of Earth.
V52A-02 10:45h
Tracing Mantle and Slab Contributions to the Proto Izu-Bonin-Mariana Arc Magmatism: Mineralogical and Geochemical Studies of a Volcaniclastic Sequence From the Palau-Kyushu Ridge, West Philippine Basin
The submerged Palau-Kyushu Ridge (PKP) is the remnant of a proto-Izu-Bonin-Mariana (IBM) island arc, which was active since about 48 to 35 Ma ago, contemporaneously to subduction of the Pacific plate beneath the spreading West Philippine basin. Volcanic activity along the PKR stopped in response to plate reorganization, and renewed subduction which initiated with an E-W rifting, separating the PKR from the Iwo Jima ridge in the north and the West Mariana Ridge in the south. A 400 m- thick sequence of turbidites containing volcaniclastic sediments, underlain by basin basalt with a magnetic age of 49 Ma, was drilled during ODP Leg 195 at Site 1201, located about 200 km off the western slope of the PKR. These volcaniclastics represent a unique opportunity to sample the past activity of the PKR. Therefore, the mineralogical and geochemical features of these volcaniclastics are used to investigate the petrogenesis of the proto-IBM arc magmas, and the temporal evolution of the currently active intra-oceanic IBM island arc and related back-arc basins. From the recovered turbidite sequence, many volcaniclasts (mm- to cm- sized) and fresh pyroxene and plagioclase grains (xenocrysts) were collected and investigated from a petrological point of view. Most of the clasts are angular (breccias), and no rounded grains occur in the xenocrysts population, suggesting that these materials were incorporated in the turbidites fairly quickly after removal from the slopes of the PKR. The analyzed clasts range in composition from basalt (one sample) to basaltic andesite (the majority); a few samples showing a high total alkalis content like mugearite, are classified as altered basalts enriched in alkalis by hydrothermal alteration. Trace element abundance and distribution are indicative of a calk-alkaline affinity, also evidenced by high LILE/HFSE and LREE/HREE ratios. Initial Sr and Nd isotope ratios indicate derivation of the parental magmas from an arc-like mantle source, suggesting the involvement of a slightly enriched subduction-derived component, in agreement with the intra-oceanic character of the PKR.
V52A-03 11:00h
Geochemistry of Oligocene volcaniclastic sediments from DSDP Site 296, Palau Kyushu Ridge
DSDP Site 296 is located on the northern Palau-Kyushu ridge (PKR), which is a remnant of the early Izu-Bonin-Mariana (IBM) arc. At about 30 Ma, early IBM arc activity ceased, and the arc was rifted lengthwise into remnant (PKR) and trenchward sections (the active IBM arc basement), separated by the Shikoku and Parece Vela back arc basins. Site 296 is the northernmost location on the PKR where early IBM arc volcanics have been sampled. Lapilli tuffs, volcanic sandstones and siltstones comprise a 634 meter thick, early to late Oligocene section that extends to 1087 meters below seafloor. A biostratigraphic age of early Oligocene was deduced for the oldest unit, but an Ar/Ar age of 47.5 Ma (middle Eocene) was reported for lapilli tuff near the bottom of the section, suggesting that that older, reworked material was incorporated into the sediment pile. Large clasts from a sandstone unit were isolated and analyzed for major and trace elements. All are andesite (58-60% SiO2), with distinct low-K and high-K features. High-K andesites are strongly incompatible element enriched, with Ba and Th abundances of 100 X primitive mantle values and normalized LREE/HREE = 7.5-8.0. Glass fragments and pyroxene grains were separated from 17 lapilli tuff and sandstone units. Glass compositions range from basalt (47% SiO2) to rhyolite (73% SiO2). Most glasses form a low to intermediate-K tholeiitic series, with 0.8 to 1.2% TiO2 in basaltic glasses, but a few units have clearly high-K characteristics. All units have both orthopyroxene (En80 to En50), and augite (Wo44:En52:Fs4 to Wo40:En38:Fs22). Based on the absence of En-rich orthopyroxene and the composition of volcanic glass, it appears that magnesium-rich boninites are not present in the section. The predominantly low-K character of the volcaniclasts is comparable to the depleted nature of present-day Izu-Bonin arc volcanics. The occurrence of high-K rocks and glass in the section indicates that more enriched magma sources were also present beneath the Izu-Bonin arc during the Oligocene.
V52A-04 11:15h
Correlation and Analysis of Volcanic Ash in Marine Sediments From the Peru Margin
While land studies have identified the major volcanic centers of historic eruptions and active to recent volcanism within the Central Volcanic Zone (CVZ) of the Central Andes, the tephrachronologic records are disturbed by the high erosion rates of this arid region. However, volcanic material frequently occurs in marine sediment as discrete ash-fall layers and, or disseminated ash accumulations. Cores from three Peru Margin sites sites(1227, 1228, and 1229) drilled during Ocean Drilling Program (ODP) Leg 201 have been studied to determine the occurrence of volcanic ash layers and ash accumulations within marine sediments along the Peru shelf. The thickness of each ash layer and accumulations has been measured and the volumes calculated in order to decipher the episodicity of explosive volcanic activity in the North-Central Andes recorded in the off shore sediments. The geographic distribution of the sites (over 3 degrees of latitude and from 50 to 300 km offshore) and correlation of ash units between sites form the basis for minimal estimates of explosive volcanic activity in the region (only eruptions large enough to deposit ash in excess of 100 km from source are represented). Pouclet et al., (1990) estimated the minimum explosive activity along the Andean Arc from ash-bearing sediments and ash layers within cores from sites along the Peru margin collected during ODP Leg 112. As a result of better recovery (as much as ten times more core recovery in many intervals) and decreased disturbance in cores recovered during Leg 201, our documentation of ash content in cores from Leg 201 has led to a more complete record of the explosive volcanic activity along the Andean Arc. For example, Pouclet, et al., (1990) reports four ash layers from Sites 684, 680, and 681, whereas forty ash layers have been documented from cores recovered from the same locations (Sites 1227, 1228, and 1229 respectively). Our stratigraphic record agrees with Pouclet, et al., (1990), suggesting that explosive activity began in the early Eocene (35 Ma) and continued with explosive pulses during the Miocene. The greatest explosive activity occurred within the past 5 million years, with peak activity in the late Pliocene to early Pleistocene. Based on petrographic and geochemical analysis, most of the volcanic ash within cores from Leg 201 was derived from the Andean volcanic arc. These plinian eruptions produced acidic glasses and ash layers with abundant feldspar and quartz, and minor amounts of hornblende and biotite. Pouclet, et al., (1990) reports a transition from andesitic volcanism in the Middle to Late Miocene to a more shoshonitic composition through the late Pliocene to early Pleistocene during peak volcanic activity. Ash layers from at least one drilling location (Site 1228) bracketed by biostratigraphic and oxygen isotope dates may correlate with the 254 kyr eruption of Taupo, New Zealand. Previous studies have recognized the presence of tephra from this eruption in cores as far away as 1100 km from the source and suggest that ash from this volcano may occur in sediments off South America (Froggatt, et al., 1986). REFERENCES Froggatt, P.C., Nelson, C.S., Carter, L., Griggs, G., Black, K.P., 1986. An exceptionally large late Quaternary eruption from New Zealand. Nature (London), 319;6054 578-582. Pouclet, A., Cambray, H., Cadet, J.P., Bourgois, J., De Wever, P., 1990. Volcanic ash from Leg 112 off Peru. Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 112: 465-480.
V52A-05 11:30h
Regional Jurassic Submarine Arc-Apron Complex in the Northern Sierra Nevada, California
The Tuttle Lake Formation (TLF), a distinctive unit forming part of the wall rocks to the Mesozoic Sierra Nevada batholith in northern California, is interpreted to have developed within a major island arc fringing the western margin of North America during the Jurassic. It extends for 75 km along strike, from mountainous terrain NW of Truckee to the Mt. Tallac pendant SW of Lake Tahoe. Superb glaciated exposures at various locations along strike provide a window into the proximal parts of a submarine Jurassic arc-apron complex. The TLF is >4 km thick and consists mainly of massively bedded, matrix-supported, polymict volcanic breccias containing poorly vesicular, subangular to angular basaltic to andesitic clasts up to 2 m in length. Characteristics of the polymict breccias indicate deposition from submarine debris flows derived from slumping of near-vent accumulations of lithic debris or sector collapse of parts of the volcanic edifice. Interbeds of finer grained andesitic and silicic turbidites and ash-fall tuffs occur sparsely within the debris-flow sequence, as do volumetrically minor pillow-hyaloclastite breccias, recording local extrusion of lavas on the seafloor. Coarse-grained TLF debris-flow deposits abruptly overlie the Early to Middle Jurassic Sailor Canyon Formation, which consists dominantly of andesitic volcanic sandstones and mudstones deposited from distal turbidity currents in a long-lived, deep marine basin. This marked lithologic change records rapid influx of coarse-grained volcanogenic detritus into the Sailor Canyon basin, related to a major shift in position of volcanic centers. Available data show that the TLF accumulated in a narrow time frame in the Middle Jurassic, just prior to regional tilting and batholith emplacement at ~165 Ma. Coeval basaltic to andesitic hypabyssal intrusions typically compose >15% of the exposed area of the TLF. They have identical major- and trace-element compositions and REE patterns to clasts within the host debris-flow sequence, and all analyses plot as a tight group in calc-alkaline, volcanic-arc fields on standard discrimination diagrams. The intrusions range from 3 km in length down to smaller pods and intrusive pillows a few meters across, many of which appear to be tubular feeder conduits in 3D. Marginal peperites indicate that intrusion occurred while the host sediments were still wet and unconsolidated. Abundant, isolated pockets of globular and blocky peperite are inferred to have been supplied by conduits extending from larger intrusions. Magma/sediment interaction was generally non-explosive, but steam explosions locally played a role in generating dispersed peperite. The overall characteristics of the TLF support a model in which coarse-grained volcaniclastic deposits accumulated rapidly in a proximal submarine setting around a major vent complex within the Middle Jurassic arc. Magma/wet-sediment interaction occurred over large areas within proximal parts of the arc apron, where uprising batches of magma were unable to penetrate the thick volcaniclastic sequence to undergo extrusion. Instead, an extensive network of hypabyssal intrusions and peperite formed at shallow levels beneath the sea floor. Complex hypabyssal networks of this type are probably common in submarine arc sequences, but detailed mapping is required to document their full extent and significance.
V52A-06 11:45h
Hydrothermal Input into Volcaniclastic Sediments of the SuSu Knolls Hydrothermal Field, Eastern Manus Basin, Bismarck Sea, Papua New Guinea
Short sediment cores were examined from the active SuSu Knolls hydrothermal field in the eastern Manus back-arc basin in order to understand the origin of the hydrothermal component in sediments surrounding volcanogenic massive sulfide deposits. Their mineralogical and geochemical composition displays various inputs of intra-basin volcaniclastic, hydrothermal, terrigenous and biogenic components. A 40 cm-thick sediment recovered from the base of a core proximal to the Suzette chimney site consists of blocky nonvesicular to elongate vesicular volcanic glass fragments at different stages of alteration intermixed with pyrite, chalcopyrite, barite, gypsum, atacamite, illite, Fe oxyhydroxide, quartz, cristobalite, plagioclase and alunite. The composition indicates that the sediment was derived from erosion of volcanic edifices and old oxidized chimneys. Geochemical indicators for the mass wasting event are the extremely high concentrations of Cu (up to 2.3%) and Au (up to 3.5 ppm), elevated concentrations of As, Ba, Zn and Fe, as well as a positive Eu anomaly. The strong Cu-Au positive correlation suggests that chalcopyrite and gold-rich chimneys of the Suzette site are the source of hydrothermal detritus. 14C dating of foraminifera points to an approximate age of the beginning of the strongest mass wasting event at about 2050 ybp. This event was interrupted by deposition of a widespread apron of volcaniclastic sediment overlying the SuSu Knolls volcanic rocks. The volcaniclastic sediment consists of dacite fragments with plagioclase and pyroxene microlites, angular grains of Ca-rich plagioclase and clino- and orthopyroxenes, glass shards, cristobalite, aggregates of Si-dominated amorphous material and illite, alunite, pyrite, magnetite and barite. Based on the compositional similarity between the components of the volcaniclastic sediment and plagioclase-pyroxene porphyric dacite lavas building the SuSu Knolls together with the products of their hydrothermal alteration and mineralization, it is suggested that the volcaniclastic sediment originated from violent hydrothermal eruptions at the SuSu Knolls field. The incorporation of a hydrothermal component representing altered volcanics creates elevated concentrations of Au, Cu and Ba. On the other hand, the fast sedimentation of volcaniclastic material could obscure the geochemical signal, and in particular REE patterns, resulting from fallout of particles from a hydrothermal plume in the overlying water column. Deposition of plume derived hydrothermal material proximal to the venting source is inferred for the surface sediments of the cores adjacent to the Suzette site. Indicators are barite aggregates and Cu-containing particles of probable bacterial origin, which account for elevated concentrations of Ba and Cu. The Mn enrichment in the cores far from SuSu Knolls is considered to represent more distant deposition of the particulate fallout.