Variation of the Large Igneous Province (LIP) Record Through Time
Large igneous province (LIP) magmatism, involving the generation and transfer of enormous volumes (0.1-100 million km3) of mostly basaltic mantle melts, on relatively short time scales (1-10 Myr), represents a first-order geodynamic process. Magmatism at this scale is thought to affect many aspects of the Earth system and, over time, is likely to have been a significant contributor to crustal growth. LIP magmatism, whether driven by mantle plumes ascending from deep boundary layers or by plate rifting and upper mantle processes, is also a key diagnostic of deep Earth processes and of the overall rhythm of the supercontinent cycle. LIPs or their remnants occur in ocean basins and on the continents. A growing database of all such events through time lists well over 200 events. In the young record, this database captures oceanic LIPs (e.g., oceanic plateaus) as well continental LIPs (e.g., continental flood basalts), whereas in the older record we are generally dealing with LIP remnants preserved within the continents (e.g., remnant flood basalts, major dykes swarms and sill complexes, large layered intrusions, major basaltic to komatiitic volcanic packages preserved in greenstone belts). To assess whether there are major secular trends in this record, we have filtered the database for the more significant LIP events and binned these by 100 Myr intervals. To correct for uneven sampling over time and a more sparsely preserved geological record in the distant past, we have normalized the LIP count to the cumulative amount of continental crust extant at each point in time, as estimated from the approximate distribution of juvenile continental crust. Clearly such a correction is critical as young LIPs are fully preserved and can be sampled from essentially 100% of the Earth's surface area, whereas the preserved and accessible surface area of ancient rocks (e.g., 2.6-2.7 Ga) must represent a much smaller sample area (<10%). To make a smooth transition to the young record, we add in oceanic surface area and LIPs for the last 200 Myr. The emerging wiggle curve of LIPs through time shows several first-order features: 1) pronounced peaks and lows in LIP activity that strongly correlate with the putative supercontinent cycle; 2) a Phanerozoic background average of ~10-11 LIP events per 100 Myr interval, i.e. a major intraplate magmatic event every 10 million years; 3) an exponentially increasing average LIP frequence in more ancient record; 4) some pronounced peaks that are more significant than perhaps anticipated (2.0 to 2.3 Ga); and 5) a 2.7 Ga peak that is "off scale"; that is, an apparent singular event we refer to as the "2.7 Ga cataclysm", supporting ideas of a singular event such as a mantle overturn. Prior to 3 Ga, statistics are too poor, and extant crust too limited, such that few firm conclusions are possible at this stage. Interesting questions are: To what extent do we "double count" events in the older record due to continental breakup and dispersal? To what extent do we "overcorrect" by normalizing to preserved sample area? Initial answers to these questions will be discussed. Overall, the strongest signal seems to be a drastic decline (from peak to next low) following each major (super)continental aggregation.
Formation of the Permian Panzhihua Magmatic Fe-Ti Deposit and Associated A-type Granite of the Emeishan Large Igneous Province (SW China): Implications for the Daly Gap and the Hypabyssal-Volcanic Transition.
The Late Permian (260 Ma) Emeishan large igneous province of SW China contains numerous magmatic Fe- Ti oxide deposits. The Fe-Ti oxide deposits are within the lower portions of evolved layered gabbroic intrusions which are spatially and temporally associated with A-type granitic plutons. The geologic association has led to speculation that the layered gabbroic intrusions and the A-type granitic rocks are somehow linked to the formation of the Fe-Ti oxide deposits. The 260 Ma Panzhihua layered gabbroic intrusion hosts one of the largest magmatic Fe-Ti oxide deposits in China and is coeval with a peralkaline A-type granitic pluton. The granite has intruded the overlying Emeishan flood basalts and fed surface flows of columnar jointed peralkaline rhyolites. The presence of trachydacites between the layered gabbro and granite is evidence for compositional evolution from mafic to intermediate to felsic rocks. The trachydacites have intermediate to acidic composition with SiO2 = 60 to 64 wt%, MgO = 0.5 to 0.7 wt% and CaO = 1.5 to 2.5 wt% as compared to the granite SiO2 = 65 to 72 wt%, MgO = 0.1 to 0.5 wt%, CaO = < 1.0 wt%. Primitive mantle normalized incompatible element plots show corresponding reciprocal patterns between the mafic and felsic rocks. The chondrite normalized REE patterns show Eu anomalies changing from positive (Eu/Eu* = 1.1 to 2.6) in the gabbroic intrusion, to negative in the trachydacites (Eu/Eu* = 0.75 to 0.83), granites and rhyolites (Eu/Eu* = 0.55-0.87). Published εNd(T) values from clinopyroxenes (εNd(T) = +1.1 to +3.2) of the gabbroic intrusion match the whole rock values of the trachydacite (εNd(T) = +2.4), granite and rhyolite (εNd(T) = +2.2 to +2.9) suggesting that all rock types originated from the same mantle source. MELTS and trace element modeling studies have confirmed that all rock types can be generated by fractional crystallization of a common parental magma similar to high-Ti Emeishan flood basalt. The jump in SiO2 from the gabbro to the trachydacite is attributed to the en masse crystallization of the Fe- Ti oxides. The geological and geochemical data indicate that fractionation crystallization of a common parental magma produced the layered gabbroic intrusion and Fe-Ti oxide deposit, the trachydacites, granites and rhyolites of the Panzhihua region and thus form a coherent plutonic-volcanic igneous complex. The results suggest that the Daly Gap is closely related to the formation of Fe-Ti oxide deposits and that A-type granites and their volcanic equivalents are derived from enriched mantle basalts.
An overview of terrestrial trachyte-phonolite magmatism
Active trachytic-phonolitic magmatism is limited on Earth today with the exception of a few presently on-going eruptions (e.g., Mount Erebus). However, the planet features a number of extensional tectonic settings (e.g., East Africa, Rio Grande, British Columbia) and volcanic islands (e.g., Canary Islands, Azores, Hawaii) that have potentially active trachyte-phonolite systems. Furthermore, trachytes and chemically-equivalent 'silicic' phonolites (>58 wt percent SiO2) are notable in that: i) they have widespread spatial and temporal distributions on Earth, ii) they may be good candidates for Martian volcanoes, and iii) they span a compositional space that permits one of the widest ranges of rheological properties of any magma type. The variation of their rheological behaviour leads to a diverse suite of eruptive products, from viscous domes and thick lava flows to rheomorphic lava flows produced from low viscosity fire-fountaining. Although also chemically and mineralogically diverse, the trachyte-phonolite suite of rocks is generally characterized by a predominance of alkali feldspars (sanidine and/or anorthoclase), Na- and Fe-rich amphiboles and pyroxenes, and Fe-oxides. Subordinate phases can include quartz or nepheline, and less common phases such as fayalitic olivine. The predominance of Fe-end members of ferromagnesian solid solutions is consistent with their petrological position in petrogeny's residua. Surveys of chemical variations from large databases (e.g., NAVDAT: 2000 plus samples from North America) show that the composition space of trachytes is characterized by the following average oxide abundances and ranges: SiO2: 65.2 (range is 57.8-68.99), Al2O3: 16.2 (9.0-24.8), FeOT: 3.7 (0-14.5), MgO: 1.0 (0-16.0), CaO: 2.5 (0-16.7), Na2O: 4.5 (0-9.7), K2O: 4.8 (0.5-13.8). As reflected in their mineralogy, the trachyte-phonolite group also spans the silica saturated-undersaturated and metaluminous-peralkaline petrochemical divides. Convention suggests that trachyte/phonolite can form by three different paths: closed-system fractionation of a basaltic parent (e.g. several ocean island systems); open system fractionation accompanied by assimilation (e.g. Hoodoo Mtn, BC); and closed system partial melting of basaltic parent (e.g. East Africa). Each of these paths has different petrochemical, volcanological and thermodynamic consequences that can possibly be used to descriminate between paths, which is critical for understanding the petrotectonic settings for trachyte-phonolite magma generation, storage and eruption.
Magmatism and Crustal Growth along the Central Proto-Andes
Current models for the stabilisation of cratons include orthogonal accretion of allochthonous crustal blocks along convergent plate margins, coupled with the emplacement of granitoid batholiths that are, at least in part, composed of mantle-derived magmas. The extent to which this juvenile addition takes place in long-lived orogens is difficult to estimate by elemental abundances alone because sedimentary systems are efficient mixers of crustal detritus (sialic end-member), while melting to produce emblematic granitic batholiths tends to vertically average large domains of middle to lower crust and uppermost mantle (basaltic end-member). On the other hand, attempts at assessing juvenile additions by conventional Sr-Nd-Pb isotope systematics are only successful if sufficient isotopic contrasts exist between invading magmas and the country rock. Moreover, the contribution of mantle-derived melts to the enlargement of continental crust along orogenic margins during episodes of extension is even less clear and has only recently received due attention. The granitoid intrusives of the Peruvian proto-Andes however, intrude the ancient (1.1-2.5 Ga) continental crust of Western Amazonia, display a range of isotopic signatures and are fortuitously situated along a segment of the cratonic margin that remained non-collisional, thus providing an ideal setting for a study of the importance of arc versus intra-plate magmatism in the formation of continental crust over an extended time period. We build upon the results of an extensive geochemical characterisation and an in situ U-Pb geochronological study of igneous zircon from granitoid batholiths that form the backbone of the Peruvian Eastern Cordillera, and combine them with in situ Lu-Hf isotopic tracing of dated zircon grains to identify the sources of consecutive magma pulses, and track crustal evolution of the proto-Andean margin of Western Amazonia during >1.1 Ga. The PEC Hf isotope systematics are characterised by a range in initial 176Hf/177Hf compositions for a given intrusive event suggesting mixing of material derived from the Paleoproterozoic crustal substrate and variable additions of juvenile sources from Neoproterozoic to Cenozoic time. Intrusives associated with phases of regional compressive tectonism correspond to mean initial epsilon Hf values of -6.73, -2.43, -1.57 for the Ordovician, Carboniferous-Permian and Late Triassic respectively, suggesting minimum crustal contributions between 40 and 74 percent by mass. The average initial Hf systematics from granitoids associated with periods of regional extension such as the middle Neoproterozoic, Permian-Triassic and Cenozoic Andean back arc plutonism are consistently shifted towards positive values (mean initial epsilon Hf = -0.7 to + 8.0) indicating systematically larger inputs of juvenile magma (25-38 mass percent of ancient crust). In the absence of evidence for significant lateral accretion of exotic crust, the time-integrated Hf record from the west-central proto-Andes of Amazonia suggests crustal reworking was the dominant process during episodes of arc magmatism. Crustal growth, if any, must have taken place vertically via underplating of isotopically juvenile, mantle-derived melts during intervals of lithospheric extension.
Geochemical Transition From Miocene to Quaternary Arc Volcanism in the Sierra Nevada, Northern California.
During the Miocene to Pliocene, the Ancestral Cascade Arc extended south past the Lassen volcanic centre (LVC) through the Reno to Lake Tahoe area where composite volcanoes erupted dominantly porphyritic andesites. Due to the northward migration of the subducting Juan de Fuca plate (Farallon plate during the Miocene) the southern boundary of the modern Cascade Arc now lies underneath the MVC. When comparing the geochemical, petrographical and isotopic characteristics of the Miocene to Pliocene volcanic rocks that are exposed in the Lake Tahoe region to those of the modern southern Cascade Arc, it is evident that the geochemical characteristics of the arc have changed. These differences can be seen when looking particularly at the isotopic data, as Lake Tahoe volcanics have a higher 87Sr/86Sr and lower 143Nd/144Nd compared to the southern Cascade Arc. Lake Tahoe lavas are also relatively enriched in the light rare earth elements and large ion lithophile elements. These geochemical signatures are indicative of an older, subduction modified lithospheric mantle source or additional slab components that have been added to the mantle wedge during the Miocene-Pliocene compared to present day magmas. Although much work has been done on the volcanic rocks around Lake Tahoe area and the LVC, very little work has been done on volcanic rocks in the region between the two. This study area therefore signifies a substantial gap in the understanding of the tectonic and volcanic evolution between the Ancestral and Modern Cascade Arcs. Petrographically, the basalts and basaltic andesites in the study area are dense, have a light to medium gray, aphanitic groundmass, with subhedral olivine (1 to 2mm) and plagioclase microlites. The andesites were porphyritic, with a phenocryst assemblage of fresh, euhedral hornblendes; blocky, euhedral, fresh to slightly weathered, 1 to 5mm plagioclase and 1 to 2mm, tabular clinopyroxenes. Biotite is only seen in the dacitic lavas and they occur as subhedral phenocrysts (0.5 to 2mm) and are usually quite fresh. Dacites also have a minor amount of plagioclase microlites in their aphanitic, grayish white groundmass. A comparison of 87Sr/86Sr isotope values between LVC and Lake Tahoe shows that the Ancestral Cascade volcanic rocks around Lake Tahoe have higher values (greater than 0.7050) than that of the modern Cascade volcanic rocks around LVC (less than 0.7045). In the Lake Tahoe area, both 87Sr/86Sr and 143Nd/144Nd isotope values vary across the entire range of SiO2. Early data from the volcanic rocks between Lake Tahoe and LVC show that both 206Pb/204Pb and 87Sr/86Sr are more homogeneous than Lake Tahoe, at 18.8 to 19.0 and 0.7038 to 0.7042 respectively. Meanwhile, preliminary 143Nd/144Nd values are slightly more variable with ranges from 0.5127 to 0.5129. Isotopically, these rocks have characteristics more similar to those of Quaternary LVC rocks then that of Miocene to Pliocene Lake Tahoe rocks.
Revisiting the link between adakites and TTG
Adakites are a group of arc-related rocks which are studied primarily for their similarities with the Archean TTG (tonalite-trondhjemite-granodiorite) series that represents the bulk of the Archean continental crust. Year after year, the name of adakite has been used for an ever-increasing range of rocks progressively more distinct from the adakites as originally defined; predictably, a correspondingly increasing range of petrogenetic models for adakites, and by inference, for TTG was also proposed. In the same way, about any Archaean plutonic or gneissic rock that is not clearly potassic has also been referred to as "TTG", regardless of the actual definition. Re-examining a database of over 4000 published analyses of TTG and adakites helps unraveling this complexity, by identifying the different rock types that have been lumped together under these two names. Besides the "true" adakites as originally defined (sodic dacites with fractionated REE and high Sr/Y: Defant and Drummond, 1990), the name has also been applied to high Sr/Y basalts ("low silica adakites"), to Archean sodic rhyolites, and to a random collection of rocks with medium to elevated Sr/Y ("continental", or "C-type adakites"). Likewise, the name of "TTG" has been applied to genuine sodic plutonic rocks, but also to comparable, but potassic, granitoids and to a range of minor components of Archean grey gneiss complexes including amphibolites, restites, metasediments and aplitic dykes. The only meaningful comparison that can be made is between the true adakites and the Archean sodic plutonic rocks - collectively representing only about 50% of the analyses published as "adakite" or "TTG". The sodic, relatively felsic and high La/Yb & Sr/Y nature of both groups requires a high pressure (> 10-15 kbar) origin or evolution, and certainly suggests melting of a mafic (not ultramafic nor felsic) source. Adakites that do match the original definition are indeed restricted to active subduction environments and are generally associated with the subduction of young, hot lithosphere - even though other high Sr/Y rocks can form in different contexts, they are not sodic dacites and shouldn't be called adakites. Even within a definition restricted to sodic, felsic high Sr/Y and La/Yb magmas, different rock types can be identified. The TTGs s.s. cover a larger range of compositions than adakites, with at least three distinct subseries (high Sr, Na, Al, low Y and Yb subseries; medium Sr, Na, Al subseries; and low Sr). The adakites are a closer match to the medium Sr subseries. The high Sr subseries, even though its characteristics suggest a higher pressure origin than the medium Sr group, appears to be a uniquely Archean rock type with no or rare modern counterpart. The low Sr subseries is closer to modern plagiogranites than to adakites, and it probably reflects non-subduction environments. Defant and Drummond, 1990. Nature 367:662-665
The Fiskenaesset Anorthosite Complex and Associated Amphibolites, SW Greenland: An Association of a Neoarchean Oceanic Island Arc
The Neoarchean Fiskenaesset Complex, SW Greenland, is characterized by about 550 meter thick layered anorthosites, leucogabbros, gabbros, and ultramafic rocks. The Complex is spatially associated with basaltic amphibolites displaying a tholeiitic composition and relict pillow structures. There are rare amphibolites with a calc-alkaline high-magnesian andesite composition. The Complex was emplaced as multiple sills into a basaltic oceanic crust. The Complex and the associated amphibolites were undergone several phases of deformation and amphibolite to granulite facies metamorphism, and were intruded by Neoarchean granitoids. Despite the deformation and metamorphism, primary cumulate textures and igneous layering are locally well preserved throughout the Complex. Lead isotope analyses (forty five samples) of the anorthosites, leucogabbros, gabbros and ultramafics have yielded an isochron age of 2948-/+38 Ma (MSWD=47) Ma for the Complex. The Fiskenaesset anorthosites display diverse REE systematics characterized by moderately depleted to strongly enriched LREE patterns and by flat to depleted HREE patterns. The anorthosites, leucogabbros, gabbros, ultramafics, basaltic amphibolites and high-magnesian andesites are all share the depletion of Nb, relative to Th and La, on primitive mantle-normalized diagrams. Similarly, on Nb/Yb versus Th/Yb petrogenetic discrimination diagram, the Fiskenaesset rocks plot within the magmatic arc field of Phanerozoic rocks, consistent with a subduction zone origin. The major and trace element compositions of the basaltic amphibolites suggest that they are petrogenetically related to the anorthosites, gabbros, leucogabbros and ultramafic rocks by fractional crystallization. On the basis of field relationships and geochemical characteristics, we interpret the Fiskenaesset Complex as a fragment of a Neoarchean island arc.