Zell E. Peterman

Variations in Sr, Rb, K, Na, and Initial Sr87/Sr86 in Mesozoic Granitic Rocks and Intruded Wall Rocks in Central California

Initial Sr87/Sr86 of granitic rocks which are exposed north of the Garlock fault in California, and which represent the entire 130-m.y. time span of emplacement during the Mesozoic, ranges mainly from 0.7031 to 0.7082, with one value of 0.7094. A systematic areal variation, independent of age, exists for initial Sr87/Sr86 in these granitic rocks and is the same as the areal variation in initial Sr87/Sr86 of superjacent upper Cenozoic basalts and andesites. Two values of initial Sr87/Sr86, 0.7040 and 0.7060, mark natural separations of granitic rock data on K-Rb, K-Sr, and Rb/Sr-Rb variation diagrams, and also, when contoured, seem to represent geographic markers of paleo-geographic, geochemical, and physiographic significance. Upper Precambrian sedimentary and metamorphic rocks in California crop out only in the region where initial Sr87/Sr86 of granitic rocks is greater than 0.7060. A line of initial Sr87/Sr86 = 0.7060 is approximately coincident with the boundary between Paleozoic eugeosynclinal and miogeosynclinal rocks. Granitic rocks intruded into Paleozoic miogeosynclinal rocks have initial Sr87/Sr86 greater than 0.7060, whereas those intruded into eugeosynclinal Paleozoic rocks have initial Sr87/Sr86 less than 0.7060. The line of initial Sr87/Sr86 = 0.7040 is the eastern limit of principal exposures of ultramafic rocks, the western limit of Cretaceous granitic rocks, and is coincident with an abrupt change in “topographic expression” on the Bouguer gravity map of California. Correlation of the isotopic variations with these major crustal features suggests that there has been a sharp lateral contrast in crust-mantle chemistry across the region of study that has been fixed in position from the Precambrian to the present time. The chemical and isotopic variations observed are best explained if the parent magmas of the majority of granitic rocks investigated were derived in a region that was laterally variable in composition and in a zone of melting that intersected both upper mantle and lower crust. However, some igneous rocks, such as Jurassic volcanic rocks in wall rocks and roof pendants and some granitic rocks with high strontium concentrations and low Rb-Sr ratios, suggest that deeper sources are also involved in the total spectrum of igneous rocks in the region.

Isotopic composition of strontium in sea water throughout Phanerozoic time

Isotopic analyses of strontium in primary fossil carbonate reveal significant variations in Sr87Sr86 of sea water during the Phanerozoic. The strontium isotopic composition may have been uniform from the Ordovician through the Mississippian, with an average Sr87Sr86 of 0.7078. A subsequent decrease in this value into the Mesozoic is interrupted by two provisionally documented positive pulses in Sr87Sr86—one in the Early Pennsylvanian and one in the Early Triassic. The lowest observed value (0.7068) occurred in Late Jurassic time, and this was followed by a gradual increase to 0.7075 in the Late Cretaceous and a more rapid increase through the Tertiary to 0.7090 for modern sea water. These variations are thought to be the result of a complex interplay of periods of intense volcanism and epeirogenic movements of the continents on a worldwide scale.

Ostwald ripening of clays and metamorphic minerals.

Analyses of particle size distributions indicate that clay minerals and other diagenetic and metamorphic minerals commonly undergo recrystallization by Ostwald ripening. The shapes of their particle size distributions can yield the rate law for this process. One consequence of Ostwald ripening is that a record of the recrystallization process is preserved in the various particle sizes. Therefore, one can determine the detailed geologic history of clays and other recrystallized minerals by separating, from a single sample, the various particle sizes for independent chemical, structural, and isotopic analyses.

Related Strontium Isotopic and Chemical Variations in Oceanic Basalts

Sr 87 /Sr 86 values in oceanic basalts range from 0.7012 to 0.7057 and correlate with basalt composition as measured by the ratio K 2 O/(Na 2 O + K 2 O). The distribution of data points on this plot can be approximated by the following ranges in Sr 87 /Sr 86 and K 2 O/(K 2 O + Na 2 O) respectively: (l) ocean ridge tholeiites—0.7020 to 0.7035 (one value 0.7012), 0.30. If the volcanism occurring throughout much of geologic time preferentially depleted rubidium and potassium relative to strontium in the mantle, preservation of the resultant heterogeneities is necessary to explain the isotopic and chemical differences among oceanic basalts. As a corollary to this long-term depletion of rubidium and potassium of the mantle, the primitive mantle or total crust-mantle system would have an Sr 87 /Sr 86 value higher than many oceanic basalts derived from zones that have undergone multistage histories. Therefore, we suggest that the potassic lavas with Sr 87 /Sr 86 higher than those of ocean ridge tholeiites and many island basalts represent the least depleted or most primitive mantle sampled by young oceanic volcanism.

The Great Lakes tectonic zone — A major crustal structure in central North America

The Great Lakes tectonic zone is a major Precambrian crustal feature more than 1,200 km long extending eastward from Minnesota into Ontario, Canada. It is a zone of distinctive tectonism, affecting both Archean and early Proterozoic rocks, along the northern margin of the early Proterozoic Penokean fold belt adjacent to the Archean Superior province. The zone coincides with the boundary between two Archean crustal segments recognized in the region: a greenstone-granite terrane (∼2,700 m.y. old) to the north (Superior province) and an older (in part 3,500 m.y. old) gneiss terrane to the south. Tectonism along the zone began in the late Archean, during the joining together of the two terranes into a single continental mass, and culminated in the early Proterozoic, when steep or northward-facing overturned folds were formed in the supracrustal rocks, and intense cataclasis and a penetrative cleavage developed in subjacent basement rocks of the greenstone-granite terrane. The Proterozoic deformation took place under low to intermediate pressures. Movement occurred along the Great Lakes tectonic zone through much of the Precambrian time recorded in the region. In the early Proterozoic, crustal foundering, which was parallel to the zone and was diachronous, initiated the structural basins in which the early Proterozoic sequences of the Lake Superior and Lake Huron regions were deposited. Later, during the Penokean orogeny (∼1,850 to 1,900 m.y. ago), compression deformed the sequences in both regions. Still later, intermittent (∼1,850 to 1,100 m.y. ago) crustal extension provided sites for emplacement of abundant mafic igneous rocks. There is no definite evidence that any of the extensional events progressed to the stage of development of oceanic crust; probably the zone has been wholly intracratonal since its inception in late Archean time. During the Phanerozoic, minor differential movements occurred locally in the Great Lakes tectonic zone, as recorded by the thinning of Cretaceous strata and their subsequent tilting and by historic earthquakes in Minnesota.

The 1.7- to 1.8-b.y.-old trondhjemites of southwestern Colorado and northern New Mexico: Geochemistry and depths of genesis

Four trondhjemitic bodies — three of intrusive and one of extrusive origin — 1.7 to 1.8 b.y. in age occur in the Precambrian rocks of northern New Mexico and southwestern Colorado. These are the metamorphosed plutonic or hypabyssal trondhjemite of Rio Brazos, New Mexico, the interlayered quartzofeldspathic and metabasaltic metavolcanic Twilight Gneiss of the West Needle Mountains, Colorado, the syntectonic Pitts Meadow Granodiorite of the Black Canyon of the Gunnison River, Colorado, and the late syntectonic to posttectonic Kroenke Granodiorite of the Central Sawatch Range, Colorado. From south to north, over a distance of 235 km, the four rock units show systematic increases in average Al2O3 from 13.7 to 16.1 percent, in K2O from 1.5 to 2.6 percent, in Rb from 28 to 76 ppm, and in Sr from 101 to 547 ppm. Initial Sr87/Sr86 ratios are low — 0.7015 to 0.7027 — and suggest a mafic or ultramafic source. All four trondhjemite bodies have similar light rare-earth element (REE) contents. The trondhjemites of Rio Brazos and the Twilight Gneiss have relatively flat patterns (Ce/Yb 10) with low heavy rare earth content and small or no Eu anomalies. Whole-rock δO18 values for siliceous rocks of three of the bodies range from 5.8 to 8.0 per mil, although the Pitts Meadow Granodiorite gives values of 8.5 to 9.4 per mil. The parent magmas for these bodies were probably generated from a parental basaltic source, either by partial melting or fractional crystallization. Fractional crystallization mechanisms would operate at crustal levels where crystallization of plagioclase and clinopyroxene or hornblende would produce the Rio Brazos and Twilight magmas, and crystallization of hornblende, plagioclase, and biotite would produce the Kroenke and Pitts Meadows magmas. The preferred partial melting mechanism would produce the Rio Brazos and Twilight magmas at shallow depth (< 50 km), leaving a residue of plagioclase and clinopyroxene or amphibole; the Pitts Meadow magma at 50 to 60 km, where hornblende, garnet, clinopyroxene, and plagioclase would be residual; and the Kroenke magma at greater than 60 km leaving a residue of garnet and clinopyroxene. The magmas probably formed in a ridge-and-basin complex that lay between the early Precambrian craton to the north and the contemporaneous quartzite-rhyolite-tholeiite terrane to the south. A northward-dipping subduction zone can be postulated from the variation in compositions and inferred depths of melting, but complete modern analogues of similar setting are not known. A better tectonic analogue might be the Archean regimes, in which vertical motion is dominant and trondhjemitic magmas may have formed by melting at the base of foundering thick volcanic piles.

Paleogene anatexis along the Gulf of Alaska margin

Early Tertiary plutons of biotite tonalite, granodiorite, and granite are found in a curvilinear 2,000-km-long belt along the margin of the Gulf of Alaska. These plutons intrude flyschoid rocks that were accreted to the continental margin during Late Cretaceous and/or early Tertiary time. Field, petrologic, age, and Rb-Sr data on plutonic and metamorphic rocks in eastern Chugach Mountains suggest that the granitic magmas were produced by partial melting of deeper parts of the accretionary prism after it was deformed against the continent. Such magmas may be a common product of heating at the termination of certain major accretionary episodes; they are not subduction-related magmas from subcrustal sources.

Nd isotopes and the origin of 1.9-1.7 Ga Penokean continental crust of the Lake Superior region

Nd isotopic data on 26 samples demonstrate the origin of the Early Proterozoic crust of the Penokean orogen. The three major components are (1) the Marquette Range Supergroup, a predominantly miogeoclinal continental-margin sequence deposited before 1.85 Ga on Archean basement; (2) ca. 1.88 Ga felsic metavolcanic rocks of the Wisconsin magmatic terrane to the south, of evolved island-arc affinity; and (3) 1.87-1.76 Ga granitoids that intrude the metavolcanic rocks. Initial ϵ Nd values for the metavolcanic rocks of the magmatic terrane range between 0.0 and +2.4 and T DM ages between 1.9 and 2.2 b.y. The volcanic rocks primarily represent new crustal material that had only a limited Archean input, probably through mixing of subducted sediments into the magma source area. ϵ Nd (T) values for sedimentary rocks in the lower part of the Marquette Range Supergroup indicate an Archean source, most likely the 2.7 Ga Superior province to the north. On the other hand, the sedimentary rocks in the upper part of the Marquette Range Supergroup (graywackes of the upper Michigamme Formation) have initial ϵ Nd values between -0.8 and +1.5, indicative of an Early Proterozoic source. These graywackes probably are foredeep deposits derived from the volcanic rocks of the Wisconsin magmatic terrane to the south. Thus, the time of deposition of the upper Michigamme Formation dates the final convergence of the Wisconsin magmatic terrane with the continental margin. The granitoids of the Wisconsin magmatic terrane have a wide range of ϵ Nd (T) values, from -4.5 to +4.0. They represent mixtures of variable amounts of new crustal material and recycled Archean detritus. The more negative ϵ Nd (T) values occur in samples close to the northern boundary of the Wisconsin magmatic terrane, which is marked by a mylonitic shear zone (the Niagara fault zone). We suggest that the miogeoclinal lower Marquette Range Supergroup rocks having an Archean Nd signature became involved in magma genesis close to the collisional boundary, whereas increasingly lesser amounts of this older material were available for mixing farther away from the suture. The 1.9-1.7 Ga Penokean events involved major growth of new crust from the mantle.

Isotopic provenance of sandstones from the Eocene Tyee Formation, Oregon Coast Range

The Tyee Formation of Eocene age in the Oregon Coast Range has been studied by a variety of isotopic techniques in order to determine its provenance. Traditional basin analyses including paleocurrent measurements, lithofacies mapping, and study of sandstone compositions made previously suggest derivation from the Klamath Mountains, which lie to the south. In contrast, the isotopic compositions of whole-rock sandstone samples, white mica, and potassium feldspar separates preclude derivation solely from this local source area. Nd-Sm, Rb-Sr, K-Ar, 18 O/ 16 O, and D/H analyses of sandstones from the Tyee and related formations yield the following information about their source areas. (1) Whole-rock ϵ Nd values between −7.1 and −7.3 at the time of deposition indicate that an old crustal component (∼700 Ma) was incorporated in the source rocks. (2) Whole-rock Rb-Sr systematics implies an older age than those of sandstones clearly derived from the Klamath Mountains. These Rb-Sr values are similar to those of modern sands of the Columbia River that were derived from eastern source areas. (3) Either apparent Rb-Sr ages of potassium feldspars are too old or their initial 87 Sr/ 86 Sr ratios are too high to have been derived from plutonic rocks of the Klamath Mountains or Sierra Nevada. (4) White micas have a fairly consistent Late Jurassic Rb-Sr isochron but have K-Ar ages of 68 Ma, an overprint not recognized in the Klamath terranes. (5) White micas have δ 18 O values of ∼9.5, a value typical of S-type granges such as are found in the Idaho batholith, but too high for normal I-type granites such as in the northern Sierra Nevada and too low for metamorphic rocks in the Klamath Mountains. (6) White micas have δD values consistent with those observed for plutonic white mica in the Idaho batholith, but markedly lower than those of white mica from schist in the Klamath Mountains. (7) Potassium feldspars have δ 18 O values that vary widely and that mainly are not in oxygen-isotope equilibrium with coexisting white mica, suggesting that these minerals were not derived from the same source area. These results indicate that the provenance of these sandstones included S-type (two-mica) granites that formed in Late Jurassic time from sources that included an old crustal component. Minerals in the granites underwent subsequent thermotectonic age resetting in Late Cretaceous time. Rocks in the Klamath Mountains and northern Sierra Nevada do not possess these features and consequently are precluded from being major source areas for the Tyee Formation. The sandstones most likely were derived from the Idaho batholith. Abundant detritus from that source area is consistent with a model in which the Oregon Coast Range basin lay much farther east, closer to Idaho, during deposition and subsequently moved westward to its present position. Such major displacement is compatible with the tectonic-rotation history documented for the Oregon Coast Range that began during the time of deposition of the Tyee Formation.

Bimodal tholeiitic—dacitic magmatism and the Early Precambrian crust

Abstract Interlayered plagioclase-quartz gneisses and amphibolites from 2.7 to more than 3.6 b.y. old form much of the basement underlying Precambrian greenstone belts of the world; they are especially well-developed and preserved in the Transvaal and Rhodesian cratons. We postulate that these basement rocks are largely a metamorphosed, volcanic, bimodal suite of tholeiite and high-silica low-potash dacite—compositionally similar to the 1.8-b.y.-old Twilight Gneiss — and partly intrusive equivalents injected into the lower parts of such volcanic piles. We speculate that magmatism in the Early Precambrian involved higher heat flow and more hydrous conditions than in the Phanerozoic. Specifically, we suggest that the early degassing of the Earth produced a basaltic crust and pyrolitic upper mantle that contained much amphibole, serpentine, and other hydrous minerals. Dehydration of the lower parts of a downgoing slab of such hydrous crust and upper mantle would release sufficient water to prohibit formation of andesitic liquid in the upper part of the slab. Instead, a dacitic liquid and a residuum of amphibole and other silica-poor phases would form, according to Green and Ringwoods experimental results. Higher temperatures farther down the slab would cause total melting of basalt and generation of the tholeiitic member of the suite. This type of magma generation and volcanism persisted until the early hydrous lithosphere was consumed. An implication of this hypothesis is that about half the present volume of the oceans formed before about 2.6 b.y. ago.