V13B-1461 1340h
Constraints on P-T Conditions and Evolution of Granulite and Ultramafic Xenoliths From Prindle Volcano, Alaska
Prindle volcano is an isolated, inactive basanitic cone located just across the Alaska-Yukon border (Lat. 63.72 °N; Long. 141.82 °W) at the northern edge of the northern Cordilleran volcanic province. The cone contains crustal and mantle xenoliths, including crustal granulites. The basement to Prindle volcano comprises Mesozoic metamorphic rocks (greenschist to amphibolite facies), but no granulites have been detected. The crustal xenoliths contain orthopyroxene (opx)-plagioclase (pl)-quartz (qtz) +/- mesoperthite (msp) and clinopyroxene (cpx). Geothermometry using opx-cpx yields T's ranging from 770 to 1015 C at 10 kbar. Integrated feldspar geothermometry yields 855 to 964 C at 10 kbar. Pl-cpx-qtz geobarometry yields P's ~10-11 kbar. Opx-cpx geothermometry on mantle ultramafic rocks yields 915-970 C at 15 kbar. Isochemical P-T phase diagrams for spinel lherzolite suggest a maximum P of 11.5 kbar at 950 C, but experimental work suggests higher pressures, ~15 kbar (+). Present Moho conditions are estimated to be 900 +/- 100 C and ~10 kbar. The calculated liquidus for the host basanite lava is ~1450 C at 15 kbar. Our results suggest that the crustal xenoliths were metamorphosed under relatively dry conditions with crustal thicknesses not significantly greater than present day; however, rocks at the present-day surface are high-pressure amphibolites. Geothermal gradients during Mesozoic metamorphism are near the higher estimates of the present geothermal gradient, ~25 +/- 3°C /km. P-T constraints from mantle xenoliths are consistent with present day geothermal gradients. Basanite liquidus temperatures, which are significantly higher than the lherzolite temperatures, are consistent with an asthenospheric source region.
V13B-1462 1340h
The Eruptive History of the Talpa-Mascota-San Sebastian Volcanic Field in Western Mexico.
The eruptive history of the Talpa-Mascota-San Sebastian (TMSS) volcanic field in the Jalisco Block (JB) of western Mexico is presented. The JB is bounded by the Tepic-Zacoalco and Colima grabens, as well as the Middle America Trench where the Rivera plate subducts beneath North America. The TMSS volcanic field spans $\sim$2030 km$^{2}$ and contains $\sim$123 small cones and flows of minette, absarokite, basic hornblende lamprophyre, basaltic andesite, and andesite. The petrology of these lavas is described in Lange and Carmichael (1990, 1991) and Carmichael et al. (1996). Of the $\sim$123 distinguishable eruptive units within this volcanic field, 26 samples have been dated by the $^{40}$Ar/$^{39}$Ar method, and are combined with 10 dates from a previous abstract and nine dates from the literature (for a total of 45). The oldest lavas (2.35 to 0.5 Ma) are found in the Talpa region, whereas the youngest lavas (predominantly $<$ 0.5 Ma) are found in the Mascota and San Sebastain regions to the north. There is thus a clear trend of volcanism becoming younger to the north, away from the trench. On the basis of these ages, field mapping, and the use of ortho airphotos and DEMs, it is estimated that a combined volume of $<$ 12 km$^{3}$ erupted in the last 1 Myr. The dominant lava type is basaltic andesite ($\sim$44 %), followed by minette ($\sim$20 %), basic, hornblende lamprophyre ($\sim$17 %), andesite ($\sim$13 %), and absarokite ($\sim$6 %). Thus more than half of the eruptive material (57 %) is andesite and basaltic andesite, which erupted in close spatial and temporal association with the highly potassic lavas. There is no time progression to the type of magma erupted. The volumes of the potassic lava types are dwarfed by the amount of intermediate, calc-alkaline magma ($\sim$360 km$^{3}$) that has erupted over the same time period ($<$ 1 Ma) within the Tepic-Zacoalco graben in western Mexico. These age results confirm that the potassic lavas of Mascota (not unlike those erupted 3-4 Myr ago in the Sierra Nevada batholith; Farmer et al., 2002) are part of line of similar volcanism that youngs to the N-NW. The Los Volcanes field was active from 3.4-1.5 Myr ago (Wallace and Carmichael, 1992), whereas two dates from the Ayutla volcanic field give ages of 4.4 and 4.5 Ma (Righter and Rosas-Elguera, 2001). It thus appears that this line of potassic volcanism within the interior of the Jalisco block is not a coeval "volcanic front", which raises the possibility that it may not be directly related to active subduction of the Rivera plate.
V13B-1463 1340h
Calc-alkalic vs. tholeiitic series revisited: new insight from isotopic micro-analyses of plagioclase phenocrysts
Major, trace and Sr isotopic compositions of plagioclase phenocrysts in calc-alkalic and tholeiitic rocks from Zao volcano, NE Japan were mapped with micro-analyses techniques. Plagioclase in tholeiitic rocks possesses a constant Sr isotopic ratio of ~0.7042, regardless with its Ca/Na ratio or the degree of magmatic differentiation. On the other hand, Sr isotopic ratios of plagioclase in calc-alkalic rocks are lower (0.7034-0.7041) than those for tholeiites, and more importantly, increase with decreasing Ca/Na ratios. Generally accepted mechanisms for production of those two types of arc magmas are fractional crystallization of a mantle-derived basalt magma for tholeiitic rocks and mixing between tholeiitic, mafic magmas and crust-derived, felsic magmas for calc-alkalic rocks. However, the above observation does not support this. Alternatively, tholeiitic magmas could be produced by anatexis of isotopically enriched, lower crustal material; calc-alkalic andesites are mixing products between a mantle-derived, isotopically depleted basalt magma and crust-derived, tholeiitic felsic magmas. Trace element characteristics of such mantle-derived, primary O^calc-alkalicO basalt are deduced based on plagioclase compositions and inferred distribution coefficients.
V13B-1464 1340h
Metallogeny of The Sierra de Guanajuato Range, Central M\'{e}xico
The Sierra de Guanajuato Range ({\bf SGR}), trending N315$\deg$ at Central M\'{e}xico, is an orographic feature extending over a distance of 80 km. {\bf SGR} comprises three well defined lithostratigraphic units: {\bf (1)} a cretaceous basement including an arc-derived terrane named Guanajuato Arc (GA) made of gabbro, diorite and basaltic pillowed lava, and volcano-sedimentary rocks belonging to Arperos fore-arc basin which are geochemically anomalous in Au (0.15 ppm), Ag (3 ppm), Cu (40 ppm), Pb (50 ppm) and Zn (15 ppm); {\bf (2)} Early Tertiary intrusive rocks, e.g., Comanja Granite which is affected by the presence of tourmalinized (sch\"{o}rl) aplito-pegmatite dykes mineralized with rare earths elements, and {\bf (3)} Eocene redbeds (1,500-2000 m) and Oligocene-Miocene volcanics cover. The metallogeny of the {\bf SGR} shows a multiple origin in time and space: volcano-sedimentary, granitic and volcanic, being possible to define three metallogenic epochs: cretaceous, paleocene and oligocene. {\bf Cretaceous epoch} includes: {\bf (a)} volcanogenic massive sulphide deposits ({\bf VMS}) of bimodal-siliclastic type belonging to Le\'{o}n-Guanajuato district; wallrock of {\bf VMS} is made of felsic-internediate volcanics and black argillite; at Los Gavilanes deposit paragenesis is next: chalcopyrite $>$ sphalerite $>$ galena, pyrite $>$ pyrrhotite $>$ marcasite; grade is as follows: Au: .02-.07 g/t; Ag: 157-18.5 g/t; Cu: 2.24-0.81%, Pb: 4.16-0.03%; Zn: 10.35-3.02 %; {\bf (b)} lens-shaped stratiform bodies of massive pyrite (i. ex., San Ignacio prospect; $\sim$ 4,000 ton) of exhalative-sedimentary origin with chalcopyrite and sphalerite microveins. {\bf Paleocene epoch} includes both quartz-cordierite-sanidine veins and replacement bodies of hydrothermal metamorphic filliation (W +Se-Bi, Pb, Zn, Cu), and pyrometasomatic bodies [Cu, Pb, Zn (Ag), W] which genetically are linked to Comanja Granite emplacement. The wallrock at El Maguey mine (35,000 ton; 0.6% WO3) is made of hornfel and the vein (1.8-3.2m width) has a banding structure made of : \{quartz & K-feldspar\}, \{(sch\"{o}rl) & specular hematite\} and epidote alternating bands; ore minerals are scheelite and tetradymite. {\bf Oligocene epoch} includes quartz-calcite-adulaire epithermal veins (Ag-Au) of geothermal-volcanic filliation. At Guanajuato mining District; ore minerals are: Au, electrum, acanthite, aguilarite, naumannite, polybasite, proustite, fischesserite (?); chalcopyrite, sphalerite and galena. Ore grade at Las Torres mine are: Ag 300 g/t, Au 2 g/t. At El Cubo mine because of the presence of rhyolitic domes gold grade reaches 100 g/t. Since Early Cretaceous Epoch, metallogenic concepts of heritage and permanence are valid in SG ore deposits.
V13B-1465 1340h
Discovery of Late Cretaceous Granodiorites with Adakitic Geochemical Signatures from Southern Tibet: Petrogenesis and Tectonic Implications
The Transhimalayan magmatism and resultant Gangdese Batholith emplaced from the Cretaceous to Eocene in southern Tibet have been widely considered as products related to northward subduction of the Neo-Tethyan slab under Asia. Therefore, the Batholith has been conventionally thought as composed typically of calc-alkaline rocks. Here we report the discovery of granodioritic intrusions from the eastern Gangdese belt that instead show adakitic geochemical features, marked by apparently higher Sr/Y and La/Yb, coupled with lower Y and HREE, than the other Gangdese and common arc magmas. SHRIMP U-Pb zircon dating analyses of two such granodiorite samples yielded $^{206}$Pb/$^{238}$U dates of 79.9 \pm 0.7 and 83.8 \pm 1.5 (2\sigma) Ma, respectively, constraining their emplacement ages. The overall elemental characteristics of the intrusions suggest an origin by partial melting of garnet amphibolites, which we interpret on the basis of Sr and Nd isotope data [I$_{Sr} = 0.7044 \sim 0.7045; \epsilon$_{Nd}(T) = +1 \sim +3] to have been the juvenile lower crust that was produced by Cretaceous basaltic underplating above the mantle wedge and later thickened by tectonic contraction due to flattening of the Neo-Tethyan subduction, a process that eventually led to generation of the granodiorites. This interpretation is compatible with petrographic data, in particular the occurrence of magmatic epidote in the granodiorites, which imply the plutonism to have formed under moderately high pressures corresponding to middle to lower crustal depths in a fairly oxidizing environment. These furthermore allow us to infer that during the early Late Cretaceous southern Tibet was situated in an accretionary convergent margin where orogenic processes involving arc magma accretion, crustal thickening and rapid tectonic uplift may have been operating actively.
V13B-1466 1340h
Crystallization and differentiation of Cenozoic Granitic Melts, Taitao Peninsula, Chile
Geochemistry of the granitoid is significant for understanding the crustal growth on the earth. It is still controversial whether the characteristics originate from the distinctive composition of source, differences of pressure and temperature condition at the formation of melts, or differences of crystallized condition. One of the most serious problems in this issue is that the whole rock chemistry of such a coarse-grained rock is most unlikely to represent the nature of primary magma because they would have suffered from high fractionation during their fractionation. In this study, we present spot analysis of rare earth elements (REE) of relict igneous plagioclase, because the REE pattern of plagioclase reflects the melt composition, which the plagioclase was crystallized. Moreover, we estimated the primary magma of Phanerozoic granitoids to examine source, pressure and temperature path, which Taitao Peninsula granitoids went through. For this method, we conducted experiments to get the partition coefficient for REE between plagioclase and granitic melt. The partition coefficient is significant factor to estimate the melt composition. Nonetheless, previously reported partition coefficient for REE between plagioclase and melt were determined in implausible conditions for the Petrogenesis of granite (e.g., basaltic system, dry condition, or 1atm condition). We conducted experiments to synthesize plagioclase coexisting melt in a hydrous granitic system at 1.2 GPa and 900C in a piston-cylinder apparatus.
V13B-1467 1340h
Parental Magmas to the Alaskan-Type Platinum-Bearing Plutons: Evidence From Galmoenan Massif (Far East, Russia).
Alaskan-type platinum-bearing plutons and potassium-enriched mafic to ultramafic volcanic rocks are temporally and spatially associated within the Late Cretaceous-Paleocene Achaivayam-Valaginskii intraoceanic paleo-arc terrain, allochthonously present in the Koryak Highland and Kamchatka Peninsula (Far East Russia). The compositions of magmas parental to the Alaskan-type complexes were estimated using the Galmoenan plutonic complex as an example. This intrusive, composed of dunites, pyroxenites and minor gabbros, is the largest (~20 km3) in the system and best studied owing to associated platinum placer deposits. The compositions of principal mineral phases in the Galmoenan massif, olivine Fo$_{79-92}$, clinopyroxene (1-3.5 wt% Al2O3, 0.1-0.5 wt% TiO2, and incompatible trace element abundances) and Cr-spinel (5-15 wt% Al2O3 and 0.3-0.7 wt% TiO2), are typical of liquidus assemblage in primitive island arc magmas in the intraoceanic settings, and closely resembles the mineral compositions in the Achaivayam-Valaginskii ultramafic volcanic rocks. Temporal and spatial association of intrusive and extrusive units and remarkable similarity of their mineral compositions provide evidence that both suites were formed from the same parental magma. The composition of parental magmas for the Galmoenan plutonic rocks was defined using previously reported data on the Achaivayam-Valaginskii ultramafic volcanic rocks and their melt inclusions. The quantitative simulation of crystallisation in the Galmoenan magma chamber from the parental magma shows that the compositions of cumulate units are best modelled by fractional crystallisation with periodic magma replenishment. This model accounts reasonably well for large amounts of dunite present in the intrusive, extensive interval of co-crystallisation of olivine and clinopyroxene, and also explains the observed range of mineral compositions. Given estimated parameters of the parental magmas and their mantle sources, we can envisage that fluxing of highly refractory "boninite-type" mantle wedge by chlorine-rich aqueous fluids is primarily responsible for both extensive melting and geochemical characteristics of magmas, including their enrichment in platinum-group elements (PGE). Such ultramafic PGE-bearing magmas are proposed to be common in the formation of Alaskan-type intrusive complexes.
V13B-1468 1340h
Silicic Magmas Erupted From the Laguna de Bay Caldera, Macolod Corridor, Luzon, Philippines: Geochemistry and Origin
The Laguna de Bay Caldera is a depression 200 km$^{2}$ in diameter that occurs within the Macolod Corridor. The Macalod corridor is a NE-SW zone of rifting through the central part of Luzon that was the site of extensive Pliestocene to Holocene volcanism, including major pyroclastic eruptions from the Laguna de Bay Caldera. This caldera has erupted large volumes of pyroclastic material, with poorly constrained published ages from $<$ 27 ka to $>$ 50 ka. The range in pumice sample composition in these flow units is from 53 to 69 wt. % SiO$_{2}$. The abundant silicic compositions ($>$64 wt. % SiO$_{2}$) are the focus of this investigation. Published chemical data from two nearby and relatively young, subduction related, stratavolcanoes, Taal and Makiling, show that both also contain silicic deposits. A comparison of theses silicic deposits to the silicic samples from Laguna de Bay indicate that the Laguna de Bay pyroclastic deposits contain much higher K$_{2}$O/Na$_{2}$O ($<$ 1 for both Taal and Makiling and $>$6 for Laguna de Bay). Sr concentrations in the silicic samples from Laguna de Bay are high ($>$ 250 ppm), which precludes large amount of plagioclase fractionation. The small Eu anomaly is consistent with this interpretation. The REE element patterns for Laguna de Bay are LREE enriched with flat HREE. No depletion occurs in the middle REE. The lack of depletion in the middle REE is in contrast to a significant concave upward pattern for the Makiling samples, an indication of amphibole in the source (no REE data are available for Taal Volcano). Our preliminary conclusions are that the silicic samples from Laguna comprise distinct compositional groups, which may be interpreted as distinct magma batches. The very high K$_{2}$O/Na$_{2}$O values can be used to argue against the origin of these silicic magmas by fractional crystallization or partial melting of basaltic compositions. Melting or assimilation of a more evolved source must be involved. Evolved preexisting continental crust is absent in this area. Therefore we propose that the origin of the silicic magmas from the Laguna de Bay caldera is related to melting of preexisting subduction related, evolved crust.
V13B-1469 1340h
Geochemistry of pyroclastic deposits in Metro Manila, Philippines: insight on volcanic source
Most of the deposits in Metro Manila (MM) are primary pyroclastics and epiclastics included in the Diliman Tuff member of the Guadalupe Formation. Coring to a 40 m depth intersects a stratigraphic sequence consisting of, from top to bottom: a coarse grained pyroclastic flow deposit with basaltic to dacitic (51 to 66 wt.%\ SiO$_{2}$) clasts, a weathered ash - possibly a soil, a thick sequence of fines-rich surge and fall deposit, a fine grained pyroclastic flow consisting of andesitic (59 to 64 wt.%\ SiO$_{2}$) clasts and lastly another coarse grained pyroclastic flow deposit. Short cores (10 m) and outcrops in the area support this stratigraphy. The focus of this study is on the uppermost pyroclastic unit from MM. This pyroclastic flow deposit consists of mafic, felsic and banded pumice clasts. The pumices are glassy, finely vesiculated and with few plagioclase crystals. Geochemical variations in the pumice clasts of this MM deposit are different than those from the deposits erupted from the two nearest known calderas in the region, which are Laguna caldera, 40 km to the southeast, and Taal caldera, 60 km south. Both eruptions produced basaltic to dacitic pyroclastic flows. Chemical variation of the MM pyroclastics is distinctly different from Taal caldera deposits. The MM deposits have higher K$_{2}$O, Sr and Rb values compared to Taal. In terms of these chemical variables, the MM deposits are similar to the Laguna caldera deposits. Based on other chemical variables, however, MM pyroclastics also do not correlate with Laguna caldera. The pyroclastics in MM have lower TiO$_{2}$ and Zr and higher MgO concentrations than Laguna. Concentrations for these elements in MM pyroclastics are intermediate between Laguna and Taal caldera. The geochemical data from MM pyroclastic deposits are not consistent with MM magma sources being related by simple fractional crystallization from either Laguna or Taal caldera sources.
V13B-1470 1340h
Geochemical Evolution of Pre-caldera Magmas at Caviahue Caldera, Neuquen Province, Argentina
Caldera subsidence and glacial erosion at Caviahue, an upper Miocene to Pliocene volcanic center located in the Andean Southern Volcanic Zone (SVZ) at 37°50'S, has exposed a detailed cross-section of pre-caldera volcanic activity from the upper Miocene to the Pliocene. Caldera walls expose 500 to 800 m of ignimbrites, cinder cones, volcanic breccias, and lava flows, which range from 1 to nearly 100 m in thickness. Lavas erupted from the monogenetic pre-caldera volcanic field have compositions ranging from evolved basaltic andesites (4% MgO, 10% FeO) to trachytes. Strong Ni-depletion signatures and high Fe/Mg ratios indicate extensive geochemical modification of Caviahue lavas. Petrologic and geochemical analyses of major and trace element abundances in Caviahue lavas indicate cyclic fractionation and recharge in an upper-crustal magma chamber during pre-caldera volcanism. Compatible and incompatible element abundances (especially Ni, MgO, K, and Zr), plotted in stratigraphic succession, show at least six distinct fractionation trends occurred between emplacement of the oldest exposed lava flows and the eruption of the ignimbrite associated with caldera formation. Each fractionation trend is punctuated by the infusion of a volume of new, more primitive magma. Modeling of recharge events indicates that these introduced from less than half to several times the volume of the existing magma body of new, more primitive (but still evolved) magma to the chamber. Geochemical analyses of lavas deposited between intermittent periods of magma residence and volcanic eruptions show strong patterns of plagioclase, olivine, clinopyroxene, and oxide fractionation. Deposits recognized on the caldera floor thought to be associated with caldera collapse are correlated with extra-caldera trachytic ignimbrite deposits dated at 2.02 Ma, providing a late Pliocene age for caldera collapse. Post-caldera volcanism has been active until present, but has shifted to smaller polygenetic volcanic centers on the periphery of the Caviahue Caldera with the majority of volcanic activity at the historically active Volc n Copahue, located on the western rim of the caldera.
V13B-1471 1340h
Characterizing Sedimentary Carbon and Nitrogen Subduction Fluxes in the Central America Convergent Margin (ODP Legs 170, 205)
We present new data regarding the C and N contents and isotope ratios of the subducting calcareous sediments from the lower section (U3) of Site 1039 and the silty sediments from the accreted section sampled by Sites 1040, 1254, and 1255. Sediments in the calcareous section contain 62 to 866 ppm N with $\delta$$^{15}$N$_{Air}$ of 2.4 to 8.5 per mil and 0.09 to 0.73 wt.% TOC with $\delta$$^{13}$C$_{PDB}$ of -25.4 to -21.4 per mil. Integrating the upper diatomaceous ooze section (U1) and the silty clay section (U2) [Li et al., 2003], the subducting sediments along the 1100-km Central American convergent margin could carry inventories of 1.3\times10$^{10}$ g/yr N with average $\delta$$^{15}$N$_{Air}$ of 5.7 per mil, 1.4\times10$^{11}$ g/yr TOC with average $\delta$$^{13}$C$_{PDB}$ of -22.0 per mil and 1.5\times10$^{12}$ g/yr oxidized C with average $\delta$$^{13}$C$_{PDB}$ of 1.9 per mil. Sediments at Sites 1040, 1254 and 1255 contain 832 to 2398 ppm N with $\delta$$^{15}$N$_{Air}$ of 3.6 to 6.6 per mil, 0.85 to 3.06 wt.% TOC with $\delta$$^{13}$C$_{PDB}$ of -26.6 to -21.1 per mil, and $<$4 wt.% CaCO$_{3}$ with $\delta$$^{13}$C$_{PDB}$ of -26.1 to 4.1 per mil and $\delta$$^{18}$O$_{VSMOW}$ of 30.0 to 36.4 per mil. No obvious changes in N, TOC and carbonate contents and isotope concentrations are observed in sediments in and near the fault zones and the d\'{e}collement, indicating that fluid flow in the Central American convergent margin had a minimal effect on the TOC and N concentrations and isotopic compositions of the sediments. However, the impact of fluid processes on the sediment geochemistry could be indicated by extreme $^{13}$C-depletion in minor carbonate in the sediments; this depletion could have been produced biogenetically by incorporation of methane C. Regarding the mass-balance across the Central American subduction zone, if the C and N loss in sediments due to fluid processes and early devolatilization is not significant, the large amount of subducting C and N (even without considering the potentially comparable C and N inputs in altered oceanic crust and tectonically eroded material), relative to the smaller output fluxes in arc volcanic gases, indicates significant transfer to the deep earth, likely affecting the C and N isotopic characteristics of the mantle.
V13B-1472 1340h
What is the Significance of Adakitic Granitoids and Zircon Inheritance in Juvenile Arc Rocks of the Neoproterozoic Makkah Batholith, Saudi Arabia?
Neoproterozoic plutonic rocks of the Makkah batholith along the northern margin of the Jiddah terrane in western Saudi Arabia record $>$50 Ma of arc magmatism. Results of the first detailed survey using U-Pb zircon geochronology by SHRIMP-RG and trace-element geochemistry by ICP-MS provide important new constraints on subduction-related magmatism in the Arabian-Nubian Shield (ANS), but have also illuminated new problems. Samples from several dozen composite plutons ranging in composition from gabbro to granodiorite or granite were analyzed. SHRIMP-RG data for the plutons reveal 3 magmatic pulses at ca. 800, 775, and 750 Ma. Phase I (ca. 800 Ma) produced plutons that yielded U-Pb zircon ages of 803+/-17 Ma (previously reported zircon age of 816+/-3 Ma) and 804+/-5 Ma. The 803-Ma granite is nonconformably overlain by a potentially glacial diamictite at the base of the arc volcanic sequence. Phase II (ca. 775 Ma) plutons yielded ages of 781+/-8 Ma (previously reported zircon age of 760+/-10 Ma), 782+/-7 Ma, 776+/-6 Ma, and 770+/-5 Ma (previously reported zircon age of 769+/-5 Ma). Phase III (ca. 750 Ma) plutons yielded ages of 755+/-5 Ma, 747+/-5 Ma, 750+/-5 Ma, and 747+/-9 Ma. SHRIMP-RG was used to overcome a long-recognized problem of inheritance in the region. Inherited zircons of Mesoproterozoic age occur in one arc suite and in post-tectonic, A-type alkalic granites, and zircons as old as Archean occur in the arc volcanic rocks. The source of the inheritance is enigmatic. Remnants of Archean-Paleoproterozoic continental crust are exposed in the Afif terrane far to the SE. Could arc plutons of the Makkah batholith have interacted with previously unrecognized Pre-Neoproterozoic continental crust beneath the Jiddah terrane? Alternatively, did shallow arc magmas interact with buried glacial deposits and inherit far-traveled detrital zircons? The plutons exhibit geochemical signatures expected for juvenile oceanic arc rocks. They are I-type granitoids of the low- to medium-K series, show strong subduction signatures, and plot as volcanic arc granites on discrimination diagrams. Nearly half the samples show characteristics of adakites (high Sr/Y $<$130, low Y, steep REE patterns). Adakitic rocks previously recognized in the ANS were attributed to the subduction and melting of young, hot oceanic crust. However, the production of adakites from slab-melting should wane with time. That adakitic rocks occur in all 3 magmatic pulses ($>$50 Ma) suggests they were produced by a more sustained mechanism, such as melting of thickened eclogitic or garnet-granulitic lower crust. Anatexis in the lower crust is an appealing explanation for the adakitic chemistries, and for the source of the inherited zircons, but is not consistent with the arc characteristics and juvenile isotopic signatures of the Jiddah terrane. Could subduction of young oceanic crust be sustained long enough to produce adakitic magmas for $>$50 Ma (e.g. by oblique subduction of ridge segments)? If so, could the subduction of ridge segments at ca. 25 Ma intervals have produced the 3 magmatic pulses? Also, could large volumes of juvenile arc magmas melt enough older crust to inherit abundant zircons yet retain their original arc characteristics and isotopic signatures? We are using this opportunity to invite constructive dialogue to help resolve these questions.
V13B-1473 1340h
Experimental Phase Equilibria and Trace Element Partitioning in Differentiated Hydrous Calc-Alkaline Melts at High Pressures
Garnet and amphibole fractionation from high-pressure, mantle-derived hydrous basaltic liquids exert an important control on the genesis of lower arc crust. Here we present experimental phase equilibria and partitioning of trace elements between garnet, amphibole and differentiated hydrous calc-alkaline melts at elevated pressures to evaluate the potential role of these minerals in generating specific trace element characteristics that have been attributed to their presence either during partial melting or during fractional crystallization. We carried out piston cylinder experiments at conditions of P=0.8-1.2 GPa and T=800-1000$\deg$C, using a synthetic H$_{2}$O under-saturated andesite composition that was derived by fractional crystallization experiments from a primary hydrous picrobasalt. Once the phase equilibria were established, the andesite starting material was doped with 20 trace elements. Run products, analysed by electron microprobe and ion microprobe, include amphibole, plagioclase, Fe-Ti-oxides, garnet, clinopyroxene and quenched liquid. Amphibole could successfully be analyzed (grain-size $>$10$\mu$m), whereas the grain-size of plagioclase and Fe-Ti-oxides was always too small for SIMS analysis ($\leq$5 $\mu$m). Large (up to 100 $\mu$m in diameter) and homogeneous garnets were produced, but clinopyroxene grain size was always smaller than 10 $\mu$m, just at the limit of successful analysis. The solid/liquid partition coefficients for garnet exhibit very steep variations, with LILE being highly incompatible (D$_{Ba}$=0.003) and heavy REE having very high D-values (D$_{Lu}$=163.8). The quantitative model proposed by van Westrenen et al., (2000, GCA 64) to calculate partition coefficient for garnets predicts much high values then we observe, indicating that models based on partition coefficients determined in anhydrous MORBs cannot adequately been used to predict garnet partition coefficients in hydrous, siliceous melts of calc-alkaline affinity. Amphiboles show basically constant HREE partition coefficients and higher but variable D's for HFSE (Nb=0.98, Ta=0.61) than published data for amphiboles in equilibrium with andesitic liquid compositions. Similarly to Tiepolo et al. (2000, EPSL 176), we obtain D$_{Nb}$ $>$ D$_{Ta}$. Our results indicate that amphibole fractionation from andesitic magma at elevated pressure producing amphibole-rich cumulates at the base of the arc crust is an additional scenario to explain sub-chondritic Nb/Ta ratio in differentiated island-arc magmas. The peraluminous to metaluminous restitic liquids are silica-rich andesites to dacites. High-pressure melts in equilibrium with garnet and cpx exhibit very low Y and HREE, very high Sr contents and lack Eu anomalies. High Sr/Y ratios of intermediate to acidic island-arc magmas are often regarded as indicative for so called adakites. The present data show that 10% fractionation of garnet and cpx at the base of a growing island arc results in a change from normal Sr/Y ratio of 20 for a basaltic andesite to values exceeding 200. Therefore, high Sr/Y ratios cannot be used as an unequivocal characteristic identifying "slab-melts"; identical or similar geochemical patterns can be produced through high-pressure fractionation of mantle-derived "normal" calc-alkaline basaltic primary magmas at the base of a growing island-arc.
V13B-1474 1340h
High SiO$_{2}$ (Rhyolitic) Melt Inclusions in Augite Phenocrysts from an Aleutian High Alumina Basalt: Evidence for a Magmatic Component or a Petrologic Process?
A high alumina basalt lava from the Aleutian Island of Kanaga has phenocrysts of sector zoned augite that contain numerous melt inclusions of highly evolved (greater than $>$70% SiO$_{2}$) rhyolitic glass. The inclusions range from 10 to 100 microns in their maximum dimension. All inclusions are anhedral and contain no evidence for post entrapment crystallization suggesting that they represent the composition of the liquid at the time of entrapment. A complete lack of textural or mineralogic evidence for a mixing origin coupled with a measured groundmass of 56% SiO$_{2}$ removes the possibility that the melt inclusions reflect either a rhyolitic body of magma (i.e. magmatic component) or the residual liquid during progressive basalt crystallization. These same augites have been previously explained (Brophy et al., 1999) as a consequence of rapid crystal growth due to decompression- induced volatile exsolution. This suggests that the melt inclusions might represent a thick (up to 100 microns) compositional boundary layer that developed adjacent to the crystal-liquid interface during rapid crystal growth. Given this possibility, analysis of the relative growth rates of the (111) and (010) sectors and crystallization-liquid diffusion modeling of the compositional boundary layer independently yield augite growth rate estimates of 10$^{-6}$ to 10$^{-7}$ cm/sec. Combining these estimates with a single augite CSD profile for the host lava yields an estimated "residence" time of 35 to 350 days. This implies that the entire process of magma ascent, volatile ex-solution, augite crystallization and eruption could have occurred quickly or even very quickly. This study also indicates that melt inclusions should be used with caution for deducing the presence of discrete magmatic components as they could very well be the reflection of a petrologic process.
V13B-1475 1340h
$^{40}$Ar/$^{39}$Ar Dating Reveals Multiple Generations of Mafic Dikes in Southeastern Alaska
Seven $^{40}$Ar/$^{39}$Ar whole-rock incremental heating dates of mafic dikes from the Wrangell transect in southeastern Alaska yield six concordant plateau dates of 14.7 $\pm$ 1.0 Ma, 24.7 $\pm$ 1.1 Ma, 30.1 $\pm$ 0.2 Ma, 41.8 $\pm$ 1.1 Ma, 107.5 $\pm$ 0.5 Ma, and 113.0 $\pm$ 0.7 Ma. One sample is discordant with most of the steps yielding dates $>$220 Ma. We also obtained a well-defined plateau date of 115.6 $\pm$ 0.3 Ma from a hornblende separate from the sample that yielded the 113.0 $\pm$ 0.7 Ma whole-rock date. The concordance of these two dates suggests that the whole-rock dates probably reflect a close approximation of the emplacement and crystallization ages of the dikes. The mafic dikes sampled for this study have distinct chilled margins and are 0.5-5 meter thick alkali-rich basalts and basaltic andesite. The dikes crosscut all ductile fabrics and define two main sets in orientation. One set strikes NNW and dips steeply to the east and the other set strikes NE with steep SE dips. Based on crosscutting relationships, the NNW dikes appear to be older and are mostly found west of Clarence Straight. The mid-Cretaceous dikes dated in this study intrude Silurian age rocks of the Alexander terrane and are confined to the western part of the Wrangell transect. The younger (14.7 - 41.8 Ma) dikes are found throughout the study area including within variably metamorphosed Jurassic-Cretaceous meta-sedimentary rocks, mid-Cretaceous plutons, and Tertiary volcanic rocks. These new dates show that mafic dikes intruded the Alexander terrane during mid-Cretaceous shortening in SE Alaska. This suggests that basaltic magmatism occurred in SE Alaska during the final accretion of the Alexander/Wrangellia terrane and during subsequent Cenozoic deformation along a transpressive to transtensional plate boundary.
V13B-1476 1340h
Tertiary volcanic activity at Sonora Pass, CA: arc and non-arc magmatism in the central Sierra Nevada
The volume and composition of Tertiary volcanic rocks in the Sierra Nevada of California changes dramatically near Sonora Pass (latitude $38\deg$ N). North of Sonora Pass is a large volume of volcanic rocks petrographically and chemically linked to subduction in the Cascade arc. South of Sonora Pass these calc-alkaline rocks are lacking and the only preserved volcanic rocks are small-volume mafic to intermediate potassic lavas that may have been generated by Pliocene lithospheric delamination (e.g. Farmer et al 2002). We have undertaken geologic mapping and geochemical and isotopic analysis of rocks near Sonora Pass at the boundary between these two magmatic provinces. At Sonora Pass, the 16-10 m.y.-old Relief Peak Formation and its hypabyssal equivalents are dominated by hornblende-phyric andesite lava flows and mudflow breccias (a stratovolcano assemblage) with marked high field-strength element (HFSE) depletions relative to large-ion lithophile elements (LILE), high Sri ($\sim$ 0.7056), and low $\epsilon$Nd (-1.4 $>$ $\epsilon$Nd $>$ -2.5). The overlying Stanislaus Group ($\sim$10-8 my old) has elevated HFSE and LILE relative to the Relief Peak Formation, anhydrous mineralogy, and similar isotope ratios (Sri $\sim$ 0.7056, -1.9 $>$ $\epsilon$Nd $>$ -3.4). The overlying Disaster Peak Formation is petrographically similar to the Relief Peak Formation. Lavas of the Relief Peak Formation may have been derived from the hydrated, LILE-rich and HFSE-poor mantle wedge above the subducting Juan de Fuca plate as part of the ancestral Cascade arc. The eruption of the dry, HFSE- and LILE-richer Stanislaus Group from vents near those of the Relief Peak Formation and to the east during an apparent pause in Relief Peak-type activity may represent a temporary shift to a dry, more fertile, isotopically enriched source in the mantle. The shift coincides with the arrival of the subducted Mendocino Fracture Zone (MFZ) beneath Sonora Pass, and the change in plate stress as the MFZ traversed the region may have influenced this shift. The Disaster Peak Formation probably represents the final episode of Cascade arc magmatism near Sonora Pass, ending about 5 Ma. No Pliocene or younger rocks derived from the low $\epsilon$Nd, high-K source tapped in the 3.5 Ma lithospheric delamination event to the south were found.
V13B-1477 1340h
Melting Of Amphibole-Bearing Wehrlites: An Experimental Study On The Origin Of Ultracalcic Nepheline-Normative Melts In Arc Settings.
Ultramafic olivine+clinopyroxene$\pm$amphibole cumulates have been widely documented in island arc settings (Alaskan-type ultramafic-mafic complexes, xenoliths in basaltic arc lavas) and probably constitute the lower part of the arc crust. Due to the low melting temperature of amphibole ($\sim$1100 $\deg$C), such cumulates may melt during intrusion of primary mantle magmas. Accordingly, we have investigated melting conditions and melt compositions for a model amphibole-wehrlite, to assess the possible role of such cumulates in island arc magma genesis. First melts are controlled by pargasitic amphibole breakdown, are strongly nepheline-normative and Al$_{2}$O$_{3}$-rich. With increasing melt fraction, melts become ultracalcic while remaining strongly nepheline-normative. The ultracalcic melts coexist with olivine and clinopyroxene and are generated at temperatures of $>=$ 1200 $\deg$C (1.0 GPa), at least 100 $\deg$C lower than temperatures previously required to generate ultracalcic liquids from other lithologies. The melt fraction required to obtain ultracalcic liquids depends on the amount of amphibole in the source, and may amount to only a few percent in cumulates containing interstitial amphibole only. The experimental melts have strong compositional similarities to natural nepheline-normative ultracalcic melt inclusions and whole rocks from arc settings (Schiano et al., G3, 2000). The high CaO contents (up to 18 wt%), low SiO$_{2}$ contents (down to 43 wt%) and high CaO/Al$_{2}$O$_{3}$ ratios (up to 1.2) of these natural ultracalcic melts span the range of experimental partial melts from amphibole-wehrlites. Phase relations suggest that they originate by melting according to amphibole + clinopyroxene = melt + olivine in the arc crust. Pargasitic amphibole is the key phase in this process, as it lowers melting temperatures and imposes the nepheline-normative signature. It is proposed that nepheline-normative ultracalcic melt inclusions and lavas in island arc settings are tracers of interaction between high-MgO basalts and the arc crust as the temperature difference between hot (1200-1300 $\deg$C) incoming picro-basaltic magma and the necessary melting temperature is sufficient for generation of nepheline-normative ultracalcic melts from amphibole-bearing cumulates.
V13B-1478 1340h
The Ancestral Cascades Arc, northern California/western Nevada: Spatial and Temporal Variations in Volcanism and Geochemistry
Voluminous Miocene through Pliocene volcanic rocks and debris flow deposits blanket the northern Sierra Nevada of California and adjacent western Nevada. These rocks represent the extinct southern extension of the late Tertiary Cascades continental volcanic arc, termed the Ancestral Cascades arc. In western Nevada, Ancestral Cascades arc volcanism began as early as ~30 Ma, much earlier than previously recognized. In the Lake Tahoe area of northern California, ages range from as old as 16 Ma to approximately 3 Ma. In contrast to the pre-Miocene ignimbrite flareup in Nevada, arc magmatism was concentrated in the belt that was to become the Walker Lane; magmatic heating and weakening of the lithosphere may have been a factor in focusing strike-slip faulting. Arc magmatism generated numerous stratovolcanoes throughout the region, notably near the Comstock lode and Virginia City as well as along the Twin Peaks - Squaw Valley - Donner Pass region. Erupted lavas are primarily andesitic, consisting of moderately to coarsely plagioclase porphyritic rocks with accessory hornblende, pyroxene and rare biotite. Dome collapse events were common, producing unilithologic debris flows. A huge volume of volcanic material was remobilized and transported westward through paleovalleys in the Sierra Nevada (Mehrten Fmn.). More mafic (basalt to basaltic andesite) volcanism began at least as early as 15 Ma in broadly the same region as the intermediate activity. Ancestral Cascades arc lavas are subalkaline and follow a calc-alkaline fractionation trend, identical to that of the modern south Cascades. However, the Ancestral Cascades differ from their modern counterpart in that a) high-alumina olivine tholeiites (HAOT) lavas are unknown, b) all lavas are light REE-enriched and REE patterns are concave upwards, c) all lavas have 87Sr/86Sr greater than 0.7040 and 1434Nd/144Nd less than 0.5128 and overlap with alkaline lavas of the Western Great Basin well to the south. La/Sm, CaO/Al2O3, 87Sr/86Sr, and 18O/16O are all correlated with SiO2, suggesting that primary magmas were modified by interaction with Sierra Nevada crust. This hypothesis is being tested by conducting a transect across the Ancestral Cascades at approximately latitude 39N from Nevada City, CA. east to the Stillwater Range, NV. In addition, we are investigating geochronological and compositional variations along a wide corridor from southwestern Lake Tahoe to Lassen National Park to determine where the influence of Sierran crust wanes. Post-arc (lees than 3Ma) volcanism in the Lake Tahoe-Reno area is constrained to a narrow east-west corridor from north of Lake Tahoe (Truckee area) east to the Fallon, NV. area. Although these lavas tend to be mildly alkaline in composition and are not as depleted in HFSE, they are otherwise virtually indistinguishable from Ancestral Cascades lavas and have chemically similar mantle sources. The orientation of this corridor, roughly parallel to the proposed southern edge of the subducting Juan de Fuca slab, combined with a recent magma injection event into the deep crust beneath the north shore of Lake Tahoe, is interpreted to reflect asthenospheric upwelling along the edge of the slab.
V13B-1479 1340h
Middle Miocene to Quaternary Primary Basalt Magmas of North Hokkaido, Japan: Segregation Depths and Degrees of Partial Melting of Mantle Peridotite
On the basis of Sr and Nd isotopic studies on middle Miocene to Quaternary basaltic rocks from north Hokkaido (NH basaltic rocks) and the back-arc side of the NE Japan arc (BA basaltic rocks), the generation of both NH and BA basaltic rocks is considered to be closely related to mantle evolution which involves upwelling of asthenosphere into the continental lithosphere beneath north Hokkaido and the back-arc side of the NE Japan arc, at the same time as spreading of the Kurile back-arc basin and Japan Sea back-arc basin. Most of the NH basaltic magmas younger than 12 Ma and the BA basaltic magmas younger than 15 Ma were derived from the upwelled asthenospheric mantle which is slightly different from MORB source in terms of major and trace element characteristics such as TiO2/K2O and Zr/Y as well as Sr and Nd isotopes. Using the olivine maximum fractionation method, primary magma compositions were calculated for the NH primitive basalts (younger than 12 Ma) from eight locations in north Hokkaido, Japan. The modelled primary basalt magma compositions suggest segregation depths of about 23-56 km, with no across-arc variation in terms of segregation depth. Based on Fo versus NiO relations for measured and calculated equilibrium olivine compositions, the relation between Cr/(Cr+Al) ratios in chromian spinel inclusions and Fo content in host olivine phenocrysts, and the N-MORB normalized pattern of HFS elements, and the chondrite normalized patterns of REE of calculated primary magmas, most of the 10-12 Ma, 7-9 Ma, and $<$ 3 Ma primary magmas were probably derived from a common mantle peridotite source, which has depleted Sr, Nd isotope ratio, but by different degrees of partial melting, with the degree of partial melting decreasing with age. Generation of these primary magmas has been associated with twice asthenospheric mantle upwellings accompanied by spreading of the Japan Sea and Kurile back-arc basins. The older basalt magmas have been produced from high temperature mantle material beneath the northwest Hokkaido during upwelling at 10-12 Ma, by greater degrees of partial melting. The younger basalt magmas were generated from similar material under the northeast Hokkaido during upwelling at 7-9 Ma, at lower temperatures and by smaller degrees of partial melting. The older Ohmu magma may be derived from the subcontinental lithospheric mantle, which was thinned by the upwelling of the asthenospheric mantle, which is characterized by undepleted Sr and Nd isotopic compositions, and lower Zr/Y ratios than the upwelled asthenospheric mantle material.
V13B-1480 1340h
The Hf-Nd isotopic diversity of subducting oceanic sediments
We report Lu-Hf and Sm-Nd isotopic data for sediments from several DSDP and ODP sites from the Atlantic (South Sandwich, ODP 801), Indian (Nicobar Fan, DSDP 211) and Pacific Oceans (Aleutian site 183; W. Alaskan abyssal plain, DSDP 178; Astoria Fan, DSDP 174; Central America, DSDP 495, ODP 844; S. American, DSDP 321; Tonga, 595/6; Marianas, DSDP 801; Phillippines/Ryuku, DSDP 291, 294/295). These sediments span the compositional variation that exists in the ocean basins and that are actively being transported to subduction zones on the present-day Earth. The compositions of these sediments lie between three end-member groups. One group of sediments (Central America; Aleutians) has radiogenic Hf and Nd isotopic compositions ($\epsilon$$_{Hf}$ =$^{+}$10; $\epsilon$$_{Nd}$ =$^{+}$5) that are largely controlled by ash inputs from proximal volcanic arcs. A second group of sediments (Astoria and Nicobar fans) has very unradiogenic Hf and Nd isotopic compositions ($\epsilon$$_{Hf}$ =$^{-}$20; $\epsilon$$_{Nd}$ =$^{-}$15) that reflect their derivation from mature continental crust. Both these end-members, and intermediate compositions between them, lie within the crust-mantle Hf-Nd array and have Lu/Hf and Sm/Nd ratios typical of crustal rocks. This variation in isotopic composition mirrors to a large extent the range that exists in continental sediments. The third end-member group of sediments, composed of slowly accumulating metalliferous clays, is highly anomalous in terms of both its trace element and isotopic composition. These pelagic clays are the dominant components of the Tonga and Ryuku sites and are also important in the Philippine site and, locally, the Marianas site. The clays are highly enriched in REEs (many with Nd concentrations 100 to 400 ppm), but have average concentrations of Hf. This results in extremely high, superchondritic Lu/Hf ratios ($^{176}$Lu/$^{177}$Hf up to 0.18), but Sm/Nd ratios that are typically subchondritic. In addition, nearly all of these sediments have anomalously radiogenic Hf compared to Nd ($\epsilon$$_{Hf}$ =$^{+}$10 to $^{+}$15; $\epsilon$$_{Nd}$ =$^{-}$5), which causes them to plot distinctly above the crust-mantle array. In summary, there is a wide range of isotopic compositions being delivered to subduction zones that is a function of the diversity of sediment types in the world's oceans. Although the REE rich pelagic clays have unique Hf-Nd isotopic compositions, their restricted global distribution will limit their widespread application as an isotopic tracer.
V13B-1481 1340h
The Activity Of The Colima Volcano From 1999 To The 2003
The Colima Volcano has shown intense activity since the 10th of February 1999. This explosive activity of 1999 and 2000 generated an elliptical crater of 260 x 265 m, which began to be filled in by a Dome from October 2001, at February 2002 the volume of the Dome was of approximately 2x106 m3 spreading over the edges of the crater and starting to flow during the following 11 months, in this period small lobes formed on the flanks of the volcano. Constants landslides originated in these lobes filled ravines of San Antonio, El Cordovan, El Muerto, El Cafesito and Atenquique (subsequent to the earthquake of January of the 2003) with non consolidated materials, increasing the hazard of lahares during the rainy season. Beginning February 2003 the explosive activity increased, most significantly from April to August, when the plumes reached heights over 2000 meters above the crater, occasionally small pyroclastic flows were observed. The explosive events continue to date. We mapped the most significant morphological changes produced at the summit by the activity described, using three photogrammetric flights conducted by INEGI (2003) and CARTODATA (2002 and 2003). These were data complemented by a very large number of photographs taken on helicopter flights undertaken during these months. Both the photographs and the digital mapping have provided detailed information to quantify the geomorphologic evolution of the superior section of the volcano, in the course of the last five years.
V13B-1482 1340h
Magmatic Differentiation in the Accreted Talkeetna Arc, South-Central Alaska
Geochemical studies of island-arc magmatism have principally focused on the volcanic carapace of arcs because of the limited exposures of intrusive rocks in active arc settings. As a result, our understanding of how mantle-derived magmas evolve within island arcs is largely based on geochemical trends of arc volcanic suites. The accreted Talkeetna arc, however, exposes a nearly complete crustal section and provides an opportunity to compare both the plutonic and volcanic trends from a Jurassic intraoceanic arc. Our ongoing research on the Talkeetna arc includes U/Pb zircon dating and geochemical analyses that provide the basis for modeling magmatic differentiation within the arc. U/Pb zircon ages from the Chugach Mountains record active Talkeetna arc magmatism from 201-180 Ma, which is consonant with biochronology that constrains arc-related volcanism to Late Triassic to early Bajocian (206-169 Ma). Plutonic ages from the Alaska Peninsula (183-164 Ma) and the Talkeetna Mountains (177-156 Ma) suggest a northward shift in arc magmatism at $\sim$180 Ma, with continued plutonism into the Late Jurassic. Radiogenic isotope data from the Chugach Mountains record limited isotopic variation of $^{87}$Sr/$^{86}$Sr$_{initial}$ = 0.7032-0.7037 and $^{143}$Nd/$^{144}$Nd$_{initial}$ = 0.512707-0.512743. The restricted range in the isotopic ratios is similar to primitive modern arcs such as the Izu-Bonin-Mariana arc and demonstrates the juvenile nature of the main section of Talkeetna plutonic rocks. Isotopic ratios from the Talkeetna Mountains define a wider range of $^{87}$Sr/$^{86}$Sr$_{initial}$ = 0.7035-0.7053 and $^{143}$Nd/$^{144}$Nd$_{initial}$ = 0.512627-0.512735. The geographic distribution of the data suggests that the more evolved ratios may reflect assimilation of an older crustal component. The primitive isotopic ratios within the Talkeetna arc are consistent with an oceanic origin, while the slightly enriched isotopic ratios in the Talkeetna Mountains may reflect either a collisional event or a pre-existing crustal boundary that was intruded by arc magmas. The rare exposures in the Talkeetna arc provide direct constraints on models of magmatic differentiation. Greene et al. (2002) used least-squares calculations to reproduce Talkeetna arc lava compositions with 48-57 wt% SiO$_{2}$ by fractional crystallization of mineral compositions found in arc gabbronorites. We will use similar least-squares models to determine whether more evolved lava and plutonic compositions (57-77 wt% SiO$_{2}$) can also be explained by simple crystal fractionation of observed cumulate phases, or whether these compositions require multistage differentiation.
V13B-1483 1340h
Geochemistry of Concepcion and Maderas Volcanoes, Nicaragua, Central America
The occurrence of geochemical zoning along the Central American volcanic arc is due primarily to changes in mantle composition and both the composition and amount of incorporated slab components. The maximum slab signal, represented by the ratio of Ba/La, occurs in the Nicaraguan volcano Telica and decreases to both the northwest and southeast. Whereas continuous geochemical coverage between El Salvador and Nicaragua suggests that the transition is gradual, a gap in volcanic geochemical analyses between Nicaraguan and Costa Rican volcanoes hinders any conclusion as to whether the decrease is gradual here as well, or more abrupt. To this end, we analyze newly collected volcanic lavas from Isla de Ometepe in Lago de Nicaragua and northern Costa Rica using XRF and HR- ICP- MS techniques to quantify major oxides and trace elements. The island consists of two volcanoes: Concepcion and Maderas. Concepcion is presently active and records two periods of lava flows, one recent and one prehistoric. Here, our work on lavas from the eastern side of the island complements the recent work by Benjamin van Wyk de Vries on western tephra. Maderas, while not historically active, has long been a private coffee plantation granting public access only recently. These analyses will mark the first samples of this volcano within the geological community.
V13B-1484 1340h
Ar/Ar ages and Chemistry of Miocene "Adakites" East of the Chile Triple Junction: An Association between Slab-melts of the Nazca Plate and Patagonian Plateau Basalts
The origin of many 'adakites' ascribed to slab melting has been questioned with their formation being attributed to melting of thickened continental crust or forearc subduction erosion. Some of the best remaining candidates for a slab-melt origin are the Cerro Pampa 'adakites' ($\sim$48$\deg$S) in Patagonia, erupted east of the inactive volcanic arc as the Chile Ridge collided with the Chile Trench at ca. 12 Ma. Support for a slab melt origin for these hornblende-bearing dacites comes from high Sr (1330-2300 ppm) at 63-68% SiO$_{2}$, MORB-like $^{87}$Sr/$^{86}$Sr (0.7028-0.7031) and $^{143}$Nd/$^{144}$Nd $>$0.5129, and steep (heavy depleted) REE patterns (La/Yb = 30-37) in an area of thin crust. Two other 'adakite' localities to the southwest, also east of where the Chile ridge collided at ca. 12 Ma, have been recognized (65-66% SiO$_{2}$, 1370-1440 ppm Sr, $^{87}$Sr/$^{86}$Sr = 0.7032-7033, $^{143}$Nd/$^{144}$Nd $\sim$0.51289, La/Yb = 28-30). These 'adakites' at Puesto Neuvo ($\sim$49$\deg$S) and Chalt\'{e}n north of Lago Viedma ($\sim$49.5$\deg$S) lie east of young Austral Volcanic Zone 'adakites'. $^{40}$Ar/$^{39}$Ar plateau ages obtained via laser incremental heating on hornblendes show that the Chalt\'{e}n, Puesto Nuevo and Cerro Pampa 'adakites' decrease in age (14.50$\pm$0.29, 13.12$\pm$0.55, 11.39$\pm$0.61 Ma) consistent with melting of the trailing edge of the Nazca plate as ridge collision progressed to the north. Their chemistry contrasts with non-adakitic Miocene (?) backarc andesite/dacites further south, east of where the ridge collided at ca. 13 to 14 Ma. A major difference in the later region is the absence of late Miocene plateau basalts related to a slab window between the Nazca and Antarctic plates. This association suggests that Miocene slab melting in Patagonia required a thermal input from the asthenospheric mantle as well as a young, hot subducting plate. A similar case can be made for Austral Volcanic Zone "adakites" and Pliocene/Recent plateau basalts.
V13B-1485 1340h
El Jorullo Revisited: Petrology, Geochemistry & Volcanology
El Jorullo (western Trans Mexican Volcanic Belt) was produced by a monogenetic eruption between 1759 and 1774. It and Paricutin are the only two historic cones of the nearly 1000 volcanoes in the Michoacan-Guanajuato volcanic field. In a seminal study, Luhr & Carmichael (CMP 90, 1985) demonstrated that El Jorullo lavas erupted from a few cones along a 4km long NE-SW trending fissure and in a compositional sequence from mafic calc-alkaline basalts to basaltic andesites (SiO$_{2}$, Al$_{2}$O$_{3}$, MgO = 53, 16.8, 8.6 and 55.4, 19, 4, wt % respectively). They divided the lavas into 3 flow units and concluded on primarily mineralogic and major element grounds that the compositional evolution was due to crystal fractionation at lower crustal/upper mantle depths. They ruled out a significant role for crustal assimilation. Our recent field observations include a new interpretation of the flow field (8 flow units) and a greater proportion of late stage lavas. New petrologic, petrographic and radiogenic isotope (Sr, Nd, Pb) data have been collected. Major/trace element trends are similar to Luhr & Carmichael (1985), although the new sampling and flow unit subdivision reveals a smoother compositional evolution in time, akin to Paricutin. The earliest flow units (1-3) erupted relatively uniform mafic compositions comprising about half of the flow field volume; middle sequence lavas (units 4, 5) are slightly more evolved, account for 12% of the flow field volume and commonly contain partially melted/disaggregated crustal xenoliths (granodiorite and tonalite); latest, most evolved lavas (units 6-8) show greater compositional shifts both internally and relative to earlier units, and account for 35% of the erupted lava volume. The lavas are sparsely phyric and xenocrysts (plag, rare sphene and quartz) occur mainly in later state lavas. Variations in Sr and Pb isotope and trace element ratios throughout the sequence require up to 15% crustal assimilation and $\sim$25% crystal fractionation (AFC) relative to the mafic parent. The large volume fraction of the more contaminated late stage magma and the extent of the compositional shifts suggest that assimilation is a primary control on erupted magma compositions, much like at Paricutin (e.g., McBirney et al. CMP 95, 1987).
V13B-1486 1340h
Determining the Source for the Magmas of Monte Amiata (Central Italy) using Strontium, Neodymium, and Lead Isotope Analysis
The volcanoes of the Tuscan Magmatic Province (TMP) and the Roman Magmatic Province (RMP), situated on the western flank of Italy, record an extraordinarily complex subduction history. Although clearly subduction related, the volcanoes of the more northern TMP exhibit an abnormal degree of crustal influence. New isotope and trace element data may shed light on magma genesis of lavas at Monte Amiata, situated on the border between the TMP and RMP. The Monte Amiata lavas range in major element composition from trachydacites to olivine latites and form tight isotopic groupings. $^{87}$Sr/$^{86}$Sr ratios range from 0.712447 to 0.713068 and $^{143}$Nd/$^{144}$Nd ratios from 0.512094 to 0.512203. $^{207}$Pb/$^{204}$Pb ratios range from 15.6694 to 16.2330 and $^{206}$Pb/$^{204}$Pb ratios from 18.7125 to 19.0882. The isotopic data are remarkably homogeneous, revealing neither geographic nor temporal trends. The unusual isotope and trace element compositions of the Monte Amiata volcanics may have been produced by one of several different processes: 1) crustal contamination of the magma during ascension and emplacement, 2) magma mixing between the more silicic TMP magmas and a high-K series of the RMP, and 3) crustal contamination of the upper asthenosphere by marine sediments being carried down with the subducting slab under central Italy. The data appear to be closest in isotopic affinity to marine sediments from adjacent off-shore areas. The tightly constrained compositional range for both isotopic ratios and trace elements, including LREE would suggest a single magma source for these volcanics. We tentatively conclude that although some upper crustal contamination and magma mixing may be involved in the evolution of the Monte Amiata lavas, the high degree of isotopic similarity between samples of different ages and different bulk compositions suggests magma derivation from a metasomatized suprasubduction zone where the upper mantle has been altered by infiltration of fluids and partial melts from subducting marine sediments.
V13B-1487 1340h
The Petrological Full Picture of Back-arc Volcanism along the en Echelon Seamounts: Mantle Heterogeneity and Across Arc Variations of Slab-derived Components beneath the Izu-Bonin Arc
Detailed petrological and geochemical analyses of sampled volcanic rocks from the en echelon seamounts has enabled the clarification of petrological full picture for back-arc volcanism associated with spreading system in an oceanic arc. The en echelon seamounts in the northern Izu-Bonin arc, a typical example of back-arc seamount chains, consist of three distinct volcanic rock suites (Machida & Ishii, 2003, G-Cubed; Machida et al., in prep). More Enriched Suites (MES) have enriched HFSE, higher Nb/Zr, and low Cr# in spinel. Less Enriched Suites (LES) have depleted HFSE, lower Nb/Zr, and high Cr#. More Depleted Suites (MDS) have the lowest Nb/Zr values, Nb and Zr contents, and slightly enriched LILE. Mineralogical and geochemical analyses show that petrological variations (basalt to rhyolite) within volcanic rock suites are explained by fractionation with open system magma mixing. However, magma mixing do not critically affect for distinction of three volcanic rock suites. HFSE enrichment and depletion cannot be explained by differences in degree of melting and/or subduction inputs by slab melting and dehydration. Instead, these results require distinct mantle sources having different fertility for the three volcanic rock suites: fertile mantle for MES lavas, less enriched mantle for LES lavas, and more depleted mantle for MDS lavas. Geochemical features indicate that slab derived components are Ba-poor slab-derived fluids and sediment melts. The chemical signatures of slab components are strongest in the MDS rocks, followed by LES, with the MES rocks having the weakest signal. Furthermore, evidence for the influence of a fluid component tends to increase toward the quaternary volcanic front. In contrast, sediment melt signature reach a maximum at a distance of about 120-130 km from the quaternary volcanic front, where K-rich magmas (Alkaline andesites) were erupted. We suggest that these across arc geochemical variations should illustrate dehydration and melting processes of deep-subducted slab beneath island arc.
V13B-1488 1340h
Olivine and Chromian Spinel in Primitive and Near-Primitive Garibaldi Belt lavas: Implications for Sub-arc Mantle Sources in the 'Hot' Northern Cascadia Subduction System
Low-volume mafic lavas of the Garibaldi volcanic belt exhibit a systematic along-strike change from high-alumina olivine tholeiite, LILE- and LREE-enriched, HFSE-depleted calc-alkaline basalts and magnesian andesites in northwestern Washington through transitional basalts to alkali-olivine basalts and trachybasalts in southwestern British Columbia. Northward variation in basalt type accompanies a decrease (approximately 7 m.y.) in age of young ($<$22 m.y.), relatively hot Juan de Fuca oceanic crust subducted beneath the volcanic front. Arc-parallel variations in both major- and trace element basalt chemistries, which have been related to impact of slab thermal structure on magma generation in sub-arc mantle, have suggested that mantle heterogeneity had only a limited influence on GVB basalt compositions, which were controlled largely by different degrees of mantle melting and inputs of slab-derived fluxes. Electron-microprobe analyses of olivine-spinel pairs in primitive and near-primitive GVB basalts (MgO $>$6.0, up to 11.28 wt %; $>$100 ppm Ni and Cr) are used to constrain further the nature and extent of source heterogeneity beneath the volcanic belt. The lavas are aphyric or sparsely porphyritic ($<$5% phenocrysts), with olivine as the only or dominant phenocryst and microphenocryst phase; plagioclase and extremely rare clinopyroxene are solely present as microphenocrysts. Olivines in all lavas show considerable compositional range, but most phenocryst cores exhibit compositions between Fo$_{82}$ and Fo$_{89}$, contain up to 0.35 wt % NiO and 0.1 wt % CoO, and crystallized at high temperatures ($>$ $1235\deg$ C). Chromian spinels included in olivines display a large range in Cr \# (22-69) that correlate negatively with Fo content of host olivine. Ti contents ($<$2 wt %) in spinels decrease with increasing Al (10-42 wt %), Fe$^{2+}$/Fe$^{Total}$ (0.53-0.74 based on stoichiometry) and Mg \# (42-65). Considering most primitive olivine-spinel pairs, (1) Cr \# ranges from 33 to 49 with no systematic along-strike compositional variation in spinels of forearc basalts from Glacier Peak northward to Meager Mountain; (2) spinels in Meager Mountain trachybasalt and behind-the volcanic-front trachyandesites/magnesian andesites in the central GVB have elevated Cr \# (57-62), suggesting more depleted source regions than associated basalts; and (3) spinels in alkalic basalts of northernmost GVB suites (Bridge River and Salal Glacier) have the lowest Cr \# (22-28) and highest Mg \# (54-64), consistent with slightly more enriched, possible deeper, source regions in the subarc mantle. Spinel compositions suggest that geochemical diversity of GVB basalts may reflect local, but not along-strike variability in fertility of mantle sources.
V13B-1489 1340h
Petrology of Early Miocene volcanic rocks from Okushiri Island, Japan: geochemical characteristics of lithospheric mantle beneath the back-arc side of the NE Japan arc.
The Sr and Nd isotopic studies on Tertiary to Quaternary basaltic rocks of the NE Japan arc have shown that isotopic characteristics of basaltic rocks found on the back-arc side of the NE Japan arc changed drastically from an undepleted isotopic signature (initial 87Sr/86Sr (SrI)=0.7040-0.7060 and initial 143Nd/144Nd (NdI)=0.51260-0.51284) to a depleted isotopic signature (SrI=0.7030-0.7040 and NdI=0.70284-0.51308) at around 15 Ma (Shuto et., 2004). This feature may have resulted from changes around 15 Ma in the isotopic compositions of the magma source beneath the back-arc side in the NE Japan arc due to the thinning of the undepleted subcontinental lithospheric mantle by upwelling of depleted asthenospheric mantle material during the opening of Japan Sea. Based on major and trace element data as well as SrI and NdI values for Early Miocene basaltic rocks from the back-arc side of the NE Japan arc, we examined geochemical characterization of the magma source (lithospheric mantle) for these basaltic rocks. Early Miocene (23-18 Ma) basalts and associated more felsic volcanic rocks form seven volcanic fields (Okushiri Is., Matsumae Pen., Fukaura, Oga Pen., Honjo, Atsumi and Sado Is.) delineating a 500 km-long array in the back-arc side of the NE Japan arc. In terms of major, trace element and Nd isotopic compositions, two groups of Early Miocene basalts can be distinguished. Group 1 is composed of tholeiitic basalts and alkali basalts, and is characterized by high TiO2 contents (1.5-2.5 %) and high (La/Yb)n ratios ($>$5.5), and high Zr/Y ratios ($>$6). These samples show the chondrite-normalized LREE-enriched patterns and have NdI values ranging from 0.51259 to 0.51282. Group 2 is composed of tholeiitic basalts, and is different from Group 1 by lower TiO2 contents ($<$1.5 %), lower (La/Yb)n ratios ($<$5) and lower Zr/Y ratios ($<$5.5). These samples show modelately LREE-enriched patterns and have NdI values ranging from 0.51250 to 0.51278. In contrast, Middle Miocene (after 15 Ma) basaltic rocks from the Niigata and Honjo areas in the back-arc side of the NE Japan arc have NdI values of 0.51286 to 0.51310, which are higher than those of Early Miocene basalts. These Middle Miocene samples are lower in TiO2 contents and (La/Yb)n and Zr/Y ratios than Early Miocene Group 1 samples, but partly overlap with Early Miocene Group 2 samples. The geochemical differences between Group 1 and 2 basalt magmas possibly originated from subcontinental lithospheric mantle may be due to differences in degrees of partial melting or heterogeneity of the mantle.