Geochemistry at DSDP Leg 82 Holes

DSDP Leg 82 drilled nine sites to the southwest of the Azores Islands on the west flank of the Mid-Atlantic Ridge (MAR) in an attempt to determine the temporal and spatial evolution of the Azores "hot-spot" activity. The chemistry of the basalts recovered during Leg 82 is extremely varied: in Holes 558 and 561, both enriched (E-type: CeN/YbN = 1.5 to 2.7; Zr/Nb = 4.5 to 9.6) and depleted (or normal-N-type: CeN/YbN = 0.6 to 0.8; Zr/Nb > 20) mid-ocean ridge basalts (MORB) occur as intercalated lava flows. To the north of the Hayes Fracture Zone, there is little apparent systematic relationship between basalt chemistry and geographic position. However, to the south of the Hayes Fracture Zone, the chemical character of the basalts (N-type MORB) is more uniform. The coexistence of both E-type and N-type MORB in one hole may be explicable in terms of either complex melting/ fractionation processes during basalt genesis or chemically heterogeneous mantle sources. Significant variation in the ratios of strongly incompatible trace elements (e.g., La/Ta; Th/Ta) in the basalts of Holes 558 and 561 are not easily explicable by processes such as dynamic partial melting or open system crystal fractionation. Rather, the trace element data require that the basalts are ultimately derived from at least two chemically distinct mantle sources. The results from Leg 82 are equivocal in terms of the evolution of the Azores "hot spot," but would appear not to be compatible with a simple model of E-type MORB magmatism associated with upwelling mantle "blobs." Models that invoke a locally chemically heterogeneous mantle are best able to account for the small-scale variation in basalt chemistry.

Major ion concentrations and chlorine isotope composition of water-soluble chloride and structurally bound chloride in DSDP/ODP serpentinized ultramafic rocks

Eight DSDP/ODP cores were analyzed for major ion concentrations and d37Cl values of water-soluble chloride (d37Clwsc) and structurally bound chloride (d37Clsbc) in serpentinized ultramafic rocks. This diverse set of cores spans a wide range in age, temperature of serpentinization, tectonic setting, and geographic location of drilled serpentinized oceanic crust. Three of the cores were sampled at closely spaced intervals to investigate downhole variation in Cl concentration and chlorine isotope composition. The average total Cl content of all 86 samples is 0.26±0.16 wt.% (0.19±0.10 wt.% as water-soluble Cl (Xwsc) and 0.09±0.09 wt.% as structurally bound Cl (Xsbc)). Structurally bound Cl concentration nearly doubles with depth in all cores; there is no consistent trend in water-soluble Cl content among the cores. Chlorine isotope fractionation between the structurally bound Cl**- site and the water-soluble Cl**- site varies from -1.08? to +1.16?, averaging to +0.21?. Samples with negative fractionations may be related to reequilibration of the water-soluble chloride with seawater post-serpentinite formation. Six of the cores have positive bulk d37Cl values (+0.05? to +0.36?); the other two cores (173-1068A (Leg-Hole) and 84-570) have negative bulk d37Cl values (-1.26? and -0.54?). The cores with negative d37Cl values also have variable Cl**-/SO4**2- ratios, in contrast to all other cores. The isotopically positive cores (153-920D and 147-895E) show no isotopic variation with depth; the isotopically negative core (173-1068A) decreases by ~1? with depth for both the water-soluble and structurally bound Cl fractions. Non-zero bulk d37Cl values indicate Cl in serpentinites was incorporated during original hydration and is not an artifact of seawater infiltration during drilling. Cores with positive d37Cl values are most likely explained by open system fractionation during hydrothermal alteration, with preferential incorporation of 37Cl from seawater into the serpentinite and loss of residual light Cl back to the ocean. Fluid / rock ratios were probably low as evidenced by the presence of water-soluble salts. The two isotopically negative cores are characterized by a thick overlying sedimentary package that was in place prior to serpentinization. We believe the low d37Cl values of these cores are a result of hydration of ultramafic rock by infiltrating aqueous pore fluids from the overlying sediments. The resulting serpentinites inherit the characteristic negative d37Cl values of the pore waters. Chlorine stable isotopes can be used to identify the source of the serpentinizing fluid and ultimately discern chemical and tectonic processes involved in serpentinization.

Isotopic geochemistry of granitoids in the Jinshajiang suture zone, SW China

The Jinshajiang suture zone, located in the eastern part of the Tethyan tectonic domain, is noticeable for a large-scale distribution of Late Jurassic to Triassic granitoids. These granitoids were genetically related to the evolution of the Paleo-Tethys Ocean. The Beiwu, Linong and Lunong granitoids occur in the middle zone of the Jinshajiang Suture Zone, and possess similar geochemical features, indicating they share a common magma source. SIMS zircon U-Pb dating reveals the Beiwu, Linong and Lunong granitic intrusions were emplaced at 233.9±1.4 Ma (2 sigma), 233.1 ±1.4 Ma (2 sigma) and 231.0±1.6 Ma (2 sigma), respectively. All of these granitoids are enriched in abundances of Si (SiO2 =65.2-73.5 wt.%), and large-ion-lithophile-elements (LILEs), but depleted in high-field-strength-elements contents (HFSEs, e.g., Nb, Ta, Ti). In addition, they have low P2O5 contents (0.06-0.11 wt.%), A/CNK values ([molecular Al2O3/(CaO+Na2O+K2O)], mostly<1.1) and 10000Ga/Al ratios (1.7-2.2), consistent with the characteristics of I-type granites. In terms of isotopic compositions, these granitoids have high initial 87Sr/86Sr ratios (0.7078-0.7148), Pb isotopic compositions [(206Pb/204Pb)t=18.213-18.598, (207Pb/204Pb)t=15.637-15.730 and (208Pb/204Pb)t=38.323-38.791], zircon d18O values (7. per mil-9.3 per mil) and negative eNd(t) values (-5.1 to -6.7), suggesting they were predominantly derived from the continental crust. Their Nb/Ta ratios (average value=8.6) are consistent with those of the lower continental crust (LCC). However, variable ?Hf(t) values (-8.6 to +2.8) and the occurrences of mafic microgranular enclaves (MMEs) suggest that mantle-derived melts and lower crustal magmas were involved in the generation of these granitoids. Moreover, the high Pb isotopic ratios and elevated zircon d18O values of these rocks indicate a significant contribution of the upper crustal composition. We propose a model in which the Beiwu, Linong and Lunong granitoids were generated under a late collisional or post-collisional setting. It is possible that this collision was completed before Late Triassic. Decompression induced mantle-derived magmas underplated and provided the heat for the anatexis of the crust. Hybrid melts including mantle-derived and the lower crustal magmas were then generated. The hybrid melts thereafter ascended to a shallow depth and resulted in some degree of sedimentary rocks assimilation. Such three-component mixing magmas source and subsequent fractional crystallization could be responsible for the formation of the Beiwu, Linong and Lunong granitoids.

Major and trace element abundances of igneous rocks from DSDP Site 37-334

REE abundances in gabbros and peridotites from Site 334 of DSDP Leg 37 show that these rocks are cumulates produced by fractional crystallization of a primitive oceanic tholeiite magma. They may be part of a layered oceanic complex. The REE distributions in the residual liquids left after such a fractionation are similar to those of incompatible element-depleted oceanic tholeiites. The REE data indicate that the basalts which overlie the gabbro-peridotite complex, are not genetically related to plutonic rocks.

Geochemistry of basalts fropm DSDP Leg 65 basalts

Samples of basalt collected on Leg 65 near 22°N on the East Pacific Rise all display the depleted light rare-earth pattern of "normal" oceanic crust. Consequently the La/Ta ratio is close to 18, as opposed to the value of 9 associated with the flat or enriched patterns found along parts of the Mid-Atlantic Ridge and the Emperor Seamount chain. The Leg 65 samples are chemically similar to those from the CYAMEX area at 21 °N and to the Leg 54 samples from 9°N, suggesting homogeneity of the upper mantle under the northern part of the East Pacific Rise over a minimum distance of about 1500 km. The geochemistry of the rocks and their field relationships with respect to depth and distance from the axis of the Rise show no pattern of distribution linked to the degree of fractional crystallization and thus cast doubt on any possible model involving large, long-lived magma chambers at the axis of the Rise.

Chemical and isotopic compositions of rocks and minerals from the Franz Josef Land Archipelago

Rb, Sr, Sm, Nd, U, and Pb contents and Sr, Nd, and Pb isotopic composition were determined in tholeiite and subalkaline basalts (in both whole-rock samples and individual minerals) from the Franz Josef Land Archipelago. Isotopic data obtained for the Arctic basin are similar to those for islands from the Pacific, Atlantic, and Indian oceans. The assimilation of crustal (sedimentary) rocks by primary depleted material makes isochron determination of basalt age difficult or impossible. The subalkaline basalts (basaltic andesites) were presumably formed by the metasomatic introduction of incompatible elements in tholeiitie basalts and, only partially, through crustal contamination and fractional crystallization.

Measurements of trace metal abundances and isotopes in carbonate sediments, Exuma, Bahamas

In order to validate the use of 238U/235U as a paleoredox proxy in carbonates, we examined the incorporation and early diagenetic evolution of U isotopes in shallow Bahamian carbonate sediments. Our sample set consists of a variety of primary precipitates that represent a range of carbonate producing organisms and components that were important in the past (scleractinian corals, calcareous green and red algae, ooids, and mollusks). In addition, four short push cores were taken in different depositional environments to assess the impact of early diagenesis and pore water chemistry on the U isotopic composition of bulk carbonates. We find that U concentrations are much higher in bulk carbonate sediments (avg. 4.1 ppm) than in primary precipitates (avg. 1.5 ppm). In almost all cases, the lowest bulk sediment U concentrations were as high as or higher than the highest concentrations found in primary precipitates. This is consistent with authigenic accumulation of reduced U(IV) during early diagenesis. The extent of this process appears sensitive to pore water H2S, and thus indirectly to organic matter content. d238/235U values were very close to seawater values in all of the primary precipitates, suggesting that these carbonate components could be used to reconstruct changes in seawater U geochemistry. However, d238/235U of bulk sediments from the push cores was 0.2-0.4 per mil heavier than seawater (and primary precipitates). These results indicate that authigenic accumulation of U under open-system sulfidic pore water conditions commonly found in carbonate sediments strongly affects the bulk U concentrations and 238U/235U ratios. We also report the occurrence of dolomite in a tidal pond core which contains low 234U/238U and 238U/235U ratios and discuss the possibility that the dolomitization process may result in sediments depleted in 238U. From this initial exploration, it is clear that 238U/235U variations in ancient carbonate sediments could be driven by changes in global average seawater, by spatial and temporal variations in the local deposition environment, or subsequent diagenesis. To cope with such effects, proxies for syndepositional pore water redox conditions (e.g., organic matter content, iron speciation, and trace metal distributions) and careful consideration of possible post-deposition alteration will be required to avoid spurious interpretation of 238U/235U data from ancient carbonate sediments.

Geochemistry at DSDP Hole 57-439

Acidic to intermediate volcanic rocks were obtained as boulders, pebbles, and clasts with intercalated matrix sediments near the Japan Trench. A 47.5-meter conglomerate bed unconformably overlies acoustic basement consisting of Upper Cretaceous siltstone and is overlain in turn by massive coarse-sandstone and siltstone beds with many fossil mollusks. The volcanic cobbles and boulders in the conglomerate show pronounced porphyritic texture. Their phenocrysts are plagioclase, hornblende, and biotite; the groundmass consists of plagioclase, K-feldspar, quartz, iron oxide, and altered interstitial glass. The Plagioclase content of these volcanic rocks is very high, whereas iron oxide minerals are rare. The chemical composition of these volcanic rocks was analyzed to determine the rock series. Matrix sediments were also analyzed chemically, and their chemical composition was found to be similar to that of volcanic rocks, except for a lower CaO content. SiO2 content of the volcanic rocks ranges from 60.23 to 73.90, corresponding to that of andesite to rhyolite. All the samples show extremely high Al2O3 content, which reflects the high amounts of modal plagioclase. These volcanic rocks belong to both the calc-alkalic and tholeiitic rock series, and the differentiation trend is controlled by fractional crystallization, mainly of plagioclase, K-feldspar, and hornblende. The assemblage of calc-alkalic and tholeiitic rock series is frequently observed in island arcs and active continental margins. These volcanic rocks are derived from the Oyashio ancient landmass, which is a slightly matured island arc.

Sulphur geochemistry of whole-rock samples from ODP Hole 176-735B

Sulfide petrography plus whole rock contents and isotope ratios of sulfur were measured in a 1.5 km section of oceanic gabbros in order to understand the geochemistry of sulfur cycling during low-temperature seawater alteration of the lower oceanic crust, and to test whether microbial effects may be present. Most samples have low SO4/Sum S values (<= 0.15), have retained igneous globules of pyrrhotite ± chalcopyrite ± pentlandite, and host secondary aggregates of pyrrhotite and pyrite laths in smectite ± iron-oxyhydroxide ± magnetite ± calcite pseudomorphs of olivine and clinopyroxene. Compared to fresh gabbro containing 100-1800 ppm sulfur our data indicate an overall addition of sulfide to the lower crust. Selection of samples altered only at temperatures <= 110 °C constrains microbial sulfate reduction as the only viable mechanism for the observed sulfide addition, which may have been enabled by the production of H2 from oxidation of associated olivine and pyroxene. The wide range in d34Ssulfide values (-1.5 to + 16.3 per mil) and variable additions of sulfide are explained by variable epsilon sulfate-sulfide under open system pathways, with a possible progression into closed system pathways. Some samples underwent oxidation related to seawater penetration along permeable fault horizons and have lost sulfur, have high SO4/Sum S (>= 0.46) and variable d34Ssulfide (0.7 to 16.9 per mil). Negative d34Ssulfate-d34Ssulfide values for the majority of samples indicate kinetic isotope fractionation during oxidation of sulfide minerals. Depth trends in sulfide-sulfur contents and sulfide mineral assemblages indicate a late-stage downward penetration of seawater into the lower 1 km of Hole 735B. Our results show that under appropriate temperature conditions, a subsurface biosphere can persist in the lower oceanic crust and alter its geochemistry.

Geochemical composition of gabbroic rocks from ODP Hole 179-1105A, southwest Indian Ridge

Oxide-free olivine gabbro and gabbro, and oxide olivine gabbro and gabbro make up the bulk of the gabbroic suite recovered from Ocean Drilling Program (ODP) Leg 179 Hole 1105A, which lies 1.2 km away from Hole 735B on the eastern transverse ridge of the Atlantis II Fracture Zone, Southwest Indian Ridge. The rocks recovered during Leg 179 show striking similarities to rocks recovered from the uppermost 500 m of Hole 735B during ODP Leg 118. The rocks of the Atlantis platform were likely unroofed as part of the footwall block of a large detachment fault on the inside corner of the intersection of the Southwest Indian Ridge and the Atlantis II Transform at ~11.5 Ma. We analyzed the lithologic, geochemical, and structural stratigraphy of the section. Downhole lithologic variation allowed division of the core into 141 lithologic intervals and 4 main units subdivided on the basis of predominance of oxide gabbroic vs. oxide-free gabbroic rocks. Detailed analyses of whole-rock chemistry, mineral chemistry, microstructure, and modes of 147 samples are presented and clearly show that the gabbroic rocks are of cumulate origin. These studies also indicate that geochemistry results correlate well with downhole magnetic susceptibility and Formation MicroScanner (FMS) resistivity measurements and images. FMS images show rocks with a well-layered structure and significant numbers of mappable layer contacts or compositional contrasts. Downhole cryptic mineral and whole-rock chemical variations depict both "normal" and inverse fine-scale variations on a scale of 10 m to <2 m with significant compositional variation over a short distance within the 143-m section sampled. A Mg# shift in whole-rock or Fo contents of olivine of as much as 20-30 units over a few meters of section is not atypical of the extreme variation in downhole plots. The products of the earliest stages of basaltic differentiation are not represented by any cumulates, as the maximum Fo content was Fo78. Similarly, the extent of fractionation represented by the gabbroic rocks and scarce granophyres in the section is much greater than that represented in the Atlantis II basalts. The abundance of oxide gabbros is similar to that in Hole 735B, Unit IV, which is tentatively correlated as a similar unit or facies with the oxide gabbroic units of Hole 1105A. Oxide phases are generally present in the most fractionated gabbroic rocks and lacking in more primitive gabbroic rocks, and there is a definite progression of oxide abundance as, for example, the Mg# of clinopyroxene falls below 73-75. Coprecipitation of oxide at such early Mg#s cannot be modeled by perfect fractional crystallization. In situ boundary layer fractionation may offer a more plausible explanation for the complex juxtaposition of oxide- and nonoxide-bearing more primitive gabbroic rocks. The geochemical signal may, in part, be disrupted by the presence of mylonitic shear zones, which strike east-west and dip both to the south and north, but predominantly to the south away from the northern rift valley where they formed. Downhole deformation textures indicate increasing average strain and crystal-plastic deformation in units that contain oxides. Oxide-rich zones may represent zones of rheologic weakness in the cumulate section along which mylonitic and foliated gabbroic shear zones nucleate in the solid state at high temperature, or the oxide may be a symptom of former melt-rich zones and hypersolidus flow, as predicted during study of Hole 735B.

Petrology and geochemistry of the basement boninite series of ODP Site 125-786

The 720 m of igneous basement that was penetrated at Site 786 of Ocean Drilling Program Leg 125 consists of boninite-series volcanics. Bronzite andesites dominate the lithology and primitive magmas of high-Ca, intermediate-Ca, and low-Ca boninite are present in subordinate amounts. Sparsely phyric boninites typically contain olivine and orthopyroxene phenocrysts with Mg numbers [= Mg/(Mg + Fe) in moles] between 86% and 87%. Their high whole-rock Mg numbers, and the absence of zonation in the phenocrysts, imply equilibration at temperatures probably between 1200° and 1250°C, and 20° to 50°C below their liquidus. Equilibrium olivine and orthopyroxene have identical Mg numbers, and Mg/Fe partitioning between these minerals and the melt thus can be described with a single Kd. The invariably phenocryst-rich bronzite andesites contain Plagioclase that has spectacular zoning and mafic phases that can be as magnesian as those of the boninite parent. The most evolved melts are rhyolites with hypersthene, Plagioclase (An50), and magnetite. Eruption temperatures for the rhyolites are estimated at about 1000°C. Some magmas contain ferroactinolite in the groundmass, which is most likely a secondary, low-temperature phase. The locally large contrasts in degree of alteration are consistent with multiple episodes of magmatic activity. However, all igneous events produced boninite volcanics. Only the first, the edifice-building episode, gave rise to differentiated magmas. Differentiation of parental boninites took place by limited fractional crystallization, producing bronzite andesites. The erupted andesites, dacites and rhyolites are filter pressed extracts from these bronzite andesite magmas, which, as a result, have accumulated crystals. Subsequent younger igneous events produced high-Ca and intermediate-Ca boninites which intruded as dikes and sills throughout the basement sequence. The mineralogy of the dikes and sills reflects variable degrees of subliquidus cooling of the magma before emplacement.

(Table 1) Major and trace elements of basalts at DSDP Hole 72-516F

The geochemistry of basalts recovered during Leg 72 is described with emphasis on trace elements. Only Hole 516F penetrated basement; the basalts recovered are plagioclase-phyric and olivine-phyric and pervasively altered. Chemically, the basalts from Hole 516F are rather uniform in composition. However, four distinct geochemical units can be recognized, although the chemistry of two of the units appears to be controlled by chemical mobility associated with alteration. The two less-altered units cannot be related by fractional crystallization processes. Hole 516F basalts have a trace element chemistry characteristic of T-type mid-ocean ridge basalt; rare-earth element patterns (as indicated by Ce/Y ratios) are mildly fractionated flight rare-earth element enriched), and a number of incompatible element ratios are close to chondritic.

Compositions of Ni-rich magnesiowüstite and coexisting magnesioferrite spinel from sediments of DSDP Holes 86-577 and 91-596

We have found trace inclusions of Ni-rich magnesiowüstite within grains of magnesioferrite spinel recovered from Cretaceous/Tertiary boundary sediments from DSDP Site 596, South Pacific (23°51.20'S, 169°39.27'W) and DSDP Site 577, North Pacific (3°6.51'N, 157°43.40'E). Measured compositions of these inclusions range from (Mg_0.85Ni_0.74Fe_0.17)O to (Mg_0.74Ni_0.09Fe_0.17)O. Coexisting magnesioferrite and magnesiowüstite can only crystallize from ultramafic, refractory, Mg-rich liquids with Mg/Si > 2 (atom ratio). Such liquid compositions cannot form as a result of fractional crystallization and are unknown to occur as a result of terrestrial igneous processes or meteoroid ablation. We infer that these minerals crystallized from liquid droplets that equilibrated with silicate vapor at high temperatures (probably >2300°C), resulting in fractionation of volatile SiO2 from more refractory MgO. The most plausible source of this high-temperature vapor is in the fireball of the major impact event that terminated the Cretaceous.

Geochemistry and minerals at DSDP LEg 58 Holes

Leg 58 successfully recovered basalt at Sites 442, 443, and 444, in the Shikoku Basin, and at Site 446 in the Daito Basin. Only at Site 442 did penetration reach unequivocal oceanic layer 2; at the other sites, only off-axis sills and flows were sampled. Petrographic observations indicate that back-arc basalts from the Shikoku Basin, with the exception of the kaersutite-bearing upper sill at Site 444, are mineralogically similar to basalts being erupted at normal mid-ocean ridges. However, the Shikoku Basin basalts are commonly very vesicular, indicating a high volatile content in the magmas. Site 446 in the Daito Basin penetrated a succession of 23 sills which include both kaersutite-bearing and kaersutite-free basalt varieties. A total of 187 samples from the four sites has been analyzed for major and trace elements using X-ray-fluorescence techniques. Chemically, the basalts from Sites 442 and 443 and the lower sill of Site 444 are subalkaline tholeiites and resemble N-type ocean-ridge basalts found along the East Pacific Rise and at 22° N on the Mid-Atlantic Ridge (MAR), although they are not quite as depleted in certain hygromagmatophile (HYG) elements. They do not show any chemical affinities with island-arc tholeiites. The basalts from Site 446 and from the upper sill at Site 444 show alkaline and tholeiitic tendencies, and are enriched in the more-HYG elements; they chemically resemble enriched or E-type basalts and their differentiates found along sections of the MAR (e.g., 45°N) and on ocean islands (e.g., Iceland and the Azores). Most of the intra-site variation may be attributed to crystal settling within individual massive flows and sills, to high-level fractional crystallization in sub-ridge magma chambers, or, where there is evidence of a long period of magmatic quiescence between units, to batch partial melting. However, the basalts from Sites 442 and 443 and from the lower sill at Site 444 cannot easily be related to those from Site 446 and the upper sill at Site 444, and it is possible that the different basalt types were derived from chemically distinct mantle sources. From comparison of the Leg 58 data with those already available for other intra-oceanic back-arc basins, it appears that the mantle sources giving rise to back-arc-basin basalts are chemically as diverse as those for mid-ocean ridges. In addition, the high vesicularity of the Shikoku Basin basalts supports previous observations that the mantle source of back-arc-basin basalts may be contaminated by a hydrous component from the adjacent subduction zone.

Chemical and isotope composition of rocks from the Omgon Range, Western Kamchatka coastal area

Hypabyssal rocks of the Omgon Range, Western Kamchatka that intrude Upper Albian-Lower Campanian deposits of the Eurasian continental margin belong to three coeval (62.5-63.0 Ma) associations: (1) ilmenite gabbro-dolerites, (2) titanomagnetite gabbro-dolerites and quartz microdiorites, and (3) porphyritic biotite granites and granite-aplites. Early Paleocene age of ilmenite gabbro-dolerites and biotite granites was confirmed by zircon and apatite fission-track dating. Ilmenite and titanomagnetite gabbro-dolerites were produced by multilevel fractional crystallization of basaltic melts with, respectively, moderate and high Fe-Ti contents and contamination of these melts with rhyolitic melts of different compositions. Moderate- and high-Fe-Ti basaltic melts were derived from mantle spinel peridotite variably depleted and metasomatized by slab-derived fluid prior to melting. The melts were generated at variable depths and different degrees of melting. Biotite granites and granite aplites were produced by combined fractional crystallization of a crustal rhyolitic melt and its contamination with terrigenous rocks of the Omgon Group. The rhyolitic melts were likely derived from metabasaltic rocks of suprasubduction nature. Early Paleocene hypabyssal rocks of the Omgon Range were demonstrated to have been formed in an extensional environment, which dominated in the margin of the Eurasian continent from Late Cretaceous throughout Early Paleocene. Extension in the Western Kamchatka segment preceded the origin of the Western Koryakian-Kamchatka (Kinkil') continental-margin volcanic belt in Eocene time. This research was conducted based on original geological, mineralogical, geochemical, and isotopic (Rb-Sr) data obtained by the authors.