Journal of Petrology Advance Access originally published online on February 4, 2005
Journal of Petrology 2005 46(5):921-944; doi:10.1093/petrology/egi005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Constraints from Thorium/Lanthanum on Sediment Recycling at Subduction Zones and the Evolution of the Continents
DEPARTMENT OF EARTH SCIENCES, BOSTON UNIVERSITY, BOSTON, MA 02215, USA
RECEIVED MARCH 17, 2003; ACCEPTED DECEMBER 1, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTh/La VARIATIONS IN MARINE...CREATING HIGH Th/La IN...CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Arc magmas and the continental crust share many chemical features, but a major question remains as to whether these features are created by subduction or are recycled from subducting sediment. This question is explored here using Th/La, which is low in oceanic basalts (<0·2), elevated in the continents (>0·25) and varies in arc basalts and marine sediments (0·09–0·34). Volcanic arcs form linear mixing arrays between mantle and sediment in plots of Th/La vs Sm/La. The mantle end-member for different arcs varies between highly depleted and enriched compositions. The sedimentary end-member is typically the same as local trench sediment. Thus, arc magmas inherit their Th/La from subducting sediment and high Th/La is not newly created during subduction (or by intraplate, adakite or Archaean magmatism). Instead, there is a large fractionation in Th/La within the continental crust, caused by the preferential partitioning of La over Th in mafic and accessory minerals. These observations suggest a mechanism of ‘fractionation & foundering’, whereby continents differentiate into a granitic upper crust and restite-cumulate lower crust, which periodically founders into the mantle. The bulk continental crust can reach its current elevated Th/La if arc crust differentiates and loses 25–60% of its mafic residues to foundering.
KEY WORDS: arc magmatism; continental crust; delamination; thorium; sediment subduction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTh/La VARIATIONS IN MARINE...CREATING HIGH Th/La IN...CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Continents currently grow at convergent margins through magmatic additions and island arc collisions. Not only is this an observable process, but it leaves a lasting mark in chemical similarities between arcs and the average continental crust (Taylor & McLennan, 1985
; Hofmann, 1988; Kelemen, 1995; Rudnick, 1995; Albarède, 1998). The preponderance of andesites at convergent margins originally led to the ‘Andesite Model’ for continental growth (Taylor, 1967), which creates continents of intermediate composition directly above subduction zones. Although some laboratory experiments have created silicic magmas by low-degree partial melting of peridotites (e.g. Baker et al., 1995) and melt–mantle reaction models can generate high Mg-number andesites in the mantle (Kelemen, 1990), most modern arcs require parental basalt (Kay, 1980; Arculus, 1981; Tatsumi et al., 1983; Pearcy et al., 1990; Holbrook et al., 1999). The net addition to arcs is predominantly basalt, with most andesites being created by fractional crystallization, assimilation and/or mixing. Thus, the andesitic continents cannot be generated by wholesale accretion of modern arcs, and this remains one of the major outstanding problems in geology (Rudnick, 1995; Davidson & Arculus, 2004).
Nonetheless, geochemical characteristics indicate a clear association between arcs and the continents. Certain trace-element ratios, such as Ce/Pb and Th/Nb, have similar values in arc magmas and the continental crust, although distinct from oceanic crust values. For example, with few exceptions, oceanic basalts [including mid-ocean ridge basalts (MORB) and ocean island basalts (OIB)] have Ce/Pb > 20 (Hofmann, 1988), whereas the average value for the bulk continental crust (BCC) is 3·3–4·1 (Rudnick & Fountain, 1995; Wedepohl, 1995; Gao et al., 1998; McLennan, 2000), which is well within the range of arc basalts (1–10; Noll et al., 1996). Low Ce/Pb in arc basalts derives from preferential partitioning of Pb with respect to Ce into slab aqueous fluids (Brenan et al., 1995; Kogiso et al., 1997), or from partial melting of subducting basalt and/or sediment (Tatsumi, 2000; Kelemen et al., 2003b). These processes through time can account for the low Ce/Pb of the continents and high Ce/Pb of the mantle (Hofmann, 1988; Miller et al., 1994; Chauvel et al., 1995).
Another feature shared by arc basalts and the continents is a depletion in Nb and Ta, usually termed the ‘Nb anomaly’. This depletion in Nb is usually identified on a mantle- or chondrite-normalized trace-element diagram by low abundance of Nb with respect to Th or La—elements which have similar partition coefficients to Nb during mantle melting. The Nb-anomaly is also reflected in the La/Nb ratio, which, in oceanic basalts, is commonly <2·5 and, in arc basalts, is typically >2 (Fig. 1). Recent estimates of Nb concentration in the bulk continental crust yield La/Nb of 2–3 (Gao et al., 1998; Plank & Langmuir, 1998; Barth et al., 2000). Albarède (1998) has emphasized the overlap between continents and oceanic mantle in La/Nb, but the separation in Th/Nb is very clear (Fig. 1). Indeed, Th/Nb is a better reflection of the Nb-depletion than La/Nb because the order of relative incompatibility during mantle melting is Th>Nb>La, and so mantle depletion leads to high La/Nb (with no Nb-anomaly), but low Th/Nb. Figure 1 shows the similarity in Th/Nb between the continents and arcs, and the difference from oceanic basalts.
|
; BCC Th/La = 0.27 ± 0.05). MORB data from Niu & Batiza (1997)The geochemical similarities between arcs and continents have been used to estimate the mass proportion of continents formed at subduction zones. Based on the average La/Nb of arc basalts and the bulk continents, Rudnick (1995)
calculated 85–90% of the continents being formed of arcs; recent revisions to Nb estimates for the bulk continents increase this value to as much as 95% (Plank & Langmuir, 1998; Barth et al., 2000). This mixing model assumes that high La/Nb is created at subduction zones, and then imparted to the continents. This simple calculation does not include, however, the effects of recycling sedimentary material at subduction zones. Figure 1 shows that marine sediments may also have high La/Nb and Th/Nb, like the continents from which sedimentary La, Nb and Th are primarily derived. Arc magmas may contain a significant portion of these continentally derived elements, via recycling of components from subducting sediment. Although studies comparing arc output and trench input find evidence for creation of the Nb-anomaly during sediment melting in the slab (Elliott et al., 1997; Turner et al., 1997; Class et al., 2000), some of the Nb-anomaly will derive from recycling of the sediment's initial Nb-anomaly (Johnson & Plank, 1999). In the end-member scenario, where high La/Nb is created in the Archaean and subduction since then has only recycled inherited La/Nb, there would be no sensitivity in mixing models to arc additions. Thus, it is important to separate the continental signatures that are created in the subduction zone from those that are inherited. How much of the similarity between arc magmas and the continents is a result of inheritance of the continental signature via recycling of subducted sediment, and how much is newly created in the subduction zone by slab dehydration/melting and then imparted to the continents? The answer to this question is fundamental to the origin of the continents, and, in particular, the proportion of the continents that were created from subduction-zone magmatism.
In order to evaluate this question, this paper compares the sediment input and arc output for several subduction zones, focusing on the element ratio Th/La. Although the Nb-anomaly is an excellent discriminator between MORB–OIB and arc lavas, it is caused by three independent processes: (1) prior mantle depletion in Nb (Ewart & Hawkesworth, 1987; Niu & Batiza, 1997; Lundstrom et al., 1998); (2) enrichment in Th and La relative to Nb in rutile-saturated partial melts from the slab (Ryerson & Watson, 1987; Kelemen et al., 1993; Elliott et al., 1997); (3) inheritance from subducting sediment. Focusing on Th and La alone simplifies this problem because both elements are generally enriched in arc basalts relative to MORB (unlike Nb), and so are less sensitive to assumptions about the prior state of mantle depletion. Another reason to focus on Th and La is that, unlike Pb, the enrichment is not complicated by major additions from the basaltic layer of the downgoing plate (Miller et al., 1994). Several lines of evidence suggest that excess Th and rare earth elements (REE) in most arc basalts derive largely from subducted sediment, with minor contributions from subducted MORB. These arguments include Th/Ce vs 143Nd/144Nd systematics (Hawkesworth et al., 1997), Ce-anomalies, U-series and Nd-isotope systematics (Elliott et al., 1997), the lack of Th- and La-enrichment during ocean crust alteration (Staudigel et al., 1996; Kelley et al., 2003), the lack of Th- and La-mobility during blueschist- and eclogite-facies metamorphism (Arculus et al., 1999; Becker et al., 1999, 2000; Spandler et al., 2003), and the MORB-like Th–Na intercept to the global arc array at zero Th-sediment flux (Plank & Langmuir, 1993). Exceptions include high-Mg-number andesites (or adakites) with high Th and light rare earth element (LREE) concentrations and high 143Nd/144Nd (Kelemen et al., 2003b). Thus, variations in Th/La in most arc basalts will depend largely on variations in the composition of subducting sediment, and less so on the composition of subducting altered basalt or background mantle, both of which are more difficult to constrain.
Although variations in Th/La are not as familiar to geochemists as the classic ratios that illustrate continental signatures (e.g. Ce/Pb, Nb/U and La/Nb), they are also useful in this regard. Figure 1 plots two measures of the Nb-anomaly (Th/Nb and La/Nb), as well as lines of constant Th/La, which clearly distinguish oceanic basalts (MORB and OIB) from the continental crust. The bulk and upper continental crust have Th/La 
0·3, whereas MORB are <0·1 (Fig. 1). OIB have Th/La 0·1, with some anomalously low ratios in Koolau basalts (Frey et al., 1994) and some high ratios in Society and Pitcairn basalts, which, based on other chemical arguments, may include recycled continental sources (White & Duncan, 1996; Eisele et al., 2002). Primitive mantle and chondrites have Th/La 0·12 (Hofmann, 1988; Sun & McDonough, 1989). Thus, even including the full range of mantle heterogeneity, the mantle has Th/La < 0·2. The continents are enriched with respect to average oceanic basalts by roughly a factor of three. Both arc basalts and marine sediments vary between mantle and continental values, and the next section focuses on how these variations are related. If arc basalts possess similar Th/La to the locally subducted sediment, then this argues for a signature inherited from subducting sediment, and the continental crust from which sedimentary Th and La are derived. If, on the other hand, arc and sediment variations are not related, or arcs have higher Th/La than the locally subducted sediment, then high Th/La would be created in the subduction zone by preferential enrichment of Th in slab fluids and/or melts. The difference between these two outcomes has fundamental implications for the origin of continental crust.
|
Th/La VARIATIONS IN MARINE SEDIMENTS AND ARC BASALTS |
|---|
|
TOPABSTRACTINTRODUCTIONTh/La VARIATIONS IN MARINE...CREATING HIGH Th/La IN...CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Th/La variations in marine sediments
Th/La in marine sediments varies with the proportions of terrigenous, hydrogenous and volcaniclastic sources. Plotted in Fig. 2 are Th/La vs Al2O3/P2O5 ratios for representative lithologies and average trench columns. The dominant control on sediment Th/La is mixing between terrigenous sediments, which have high Al2O3/P2O5 and upper-crustal Th/La of 0·3–0·4, and metalliferous sediments, which have low Th/La and Al2O3/P2O5 because of high REE concentrations in fish debris phosphate and hydrogenous Fe–Mn oxides [see discussion and references given by Plank & Langmuir 1998
]. The other dominant component is volcaniclastic material, which can have low Th/La, reflecting mantle values. These components combine to create subducting sedimentary columns which vary in Th/La by a factor of seven globally, from near average upper-crustal values at the Antilles, Java and northeast Honshu margins, to lower volcaniclastic values at Vanuatu, and even lower metalliferous–phosphatic values at the Marianas and Tonga margins. Thus, sediments with different Th/La subduct at different margins, and provide a signal of variation to look for in the corresponding arc volcanics.
|
Th/La variations in arc basalts
It is important to recognize, however, that arc basalts are mixtures of mantle and subducted sources. Whereas sediments may exert a strong control on mixed compositions, the final Th/La in the arc magma will depend on the proportion of sediment to mantle. A simple bulk-mixing calculation using normal-mid-ocean-ridge basalt (N-MORB) mantle (e.g. 0·014 ppm Th and 0·23 ppm La; Salters & Stracke, 2004
) and the southern Antilles subducting sediment (15 ppm Th, 48 ppm La; Plank & Langmuir, 1998) shows that Th/La of the mixture will approach within 95% of the sediment ratio after only 6% addition of this sediment to an N-MORB mantle source. Based on the Banda sediments, Vroon et al. (1995) performed a similar calculation, which was also constrained by Nd-isotopes. Other isotopic mixing arguments, as well as consideration of the total mass of sediment subducted relative to mantle mass, indicate that mixing proportions for individual arcs will be much less than 6%, and so many arcs should have Th/La between mantle and sediment values. Thus, these mixing calculations predict that in places where little sediment subducts, arcs should have approximately mantle Th/La, whereas, in places with a high flux of subducted sediment, the Th/La of the arc should approach the sediment ratio.
This relationship is in fact observed. For high sediment flux margins (>0·32 Mg/yr/cm arc length), there is a good correlation between Th/La in arc basalts and Th/La in the local sediment column (filled symbols in Fig. 3; Table 1; see also Electronic Appendix A available at http://www.petrology.oupjournals.org). This sediment flux threshold would correspond to 3–6% mass of sediment relative to mass of mantle melted [mantle melting mass flux is calculated from arc growth rates of 30–60 km3/Ma/km arc length from Reymer & Schubert (1984) and Dimalanta (2002); 2·8 g/cm3 arc density, and 15% melting of the arc mantle]. At sediment fluxes below this threshold, the mantle should dominate the arc Th/La. Indeed, arcs corresponding to the lower sediment flux margins (open symbols in Fig. 3) generally fall below the sediment value, trending toward mantle values (Th/La < 0·1). Taking all of the arcs together, however, there is a clear influence on the arc from the composition of sediment subducted. Where the subducting sediment has Th/La > 0·2, the arcs have Th/La >0·15, and where the sediment has a value <0·2, the arcs are <0·15. This break in the sediment ratio corresponds to a predominance of terrigenous material (Th/La > 0·2; Fig. 2) vs slowly deposited metalliferous and phosphate-rich sediments (Th/La < 0·2). Thus, the sedimentological processes that mix different sedimentary components on the seafloor ultimately influence the composition of basalts erupted at nearby volcanic arcs. This provides further evidence, along with 10Be- (Tera et al., 1986), Nd- and Pb-isotopes (White et al., 1985), and trace-element fluxes (Plank & Langmuir, 1993), that local sediments are recycled through the subduction zone. Vroon et al. (1995) was perhaps the first to note that sediment trace-element ratios may be preserved in adjacent arc volcanics (in this case, Banda). I have extended this work to show that an element ratio can be used much like an isotope ratio to trace sediment recycling on a global basis.
|
|
Not only is there a relationship between Th/La input and output, but the slope of the correlation is near unity for the high sediment flux margins (Fig. 3). This result suggests that there is little change in Th/La during sediment recycling in subduction zones. Arcs with high sediment flux inherit the sediment ratio. This is not simply a matter of arc cannibalism, as local arc ash is a minor component of bulk sediment columns (Plank & Langmuir, 1998
). Although there is a good deal of scatter in Fig. 3, few of the data points indicate higher ratios in the arc than in the sediment. Thus, there is little evidence that the high Th/La of the continental crust is created by processes occurring in the subduction zone.
Unmixing the mantle and sediment contributions to arc Th/La
The relationship between arc and sediment Th/La (Fig. 3), although significant, is dependent on a small number of arcs with a large subducted sediment flux. For the other arcs, the mantle contribution obscures the sediment's Th/La. If there were a way to correct for the effect of mantle dilution, then the sedimentary mixing component could be examined for each arc to further test the correlation in Fig. 3. One way to accomplish this is to exploit the mixing relationships observed for individual samples within each arc, in particular, those for which high-quality trace-element data have been obtained (See Electronic Appendix A). Toward this end, new ICP-MS (inductively coupled plasma mass spectrometry) trace-element data are reported here for approximately 70 mafic samples from five arc systems (Electronic Appendix B, available at http://www.petrology.oupjournals.org). Several arc suites form trends consistent with variable amount of mixing between slab sediment and the mantle. For example, Fig. 4 shows Th/La vs Sm/La for basaltic samples from the Lesser Antilles arc. Both Th/La and Sm/La should vary, depending upon the proportion of sediment and mantle in basalt genesis, and, indeed, the Lesser Antilles data trend between a component with high Sm/La that could lie within the MORB array and a component with high Th/La that lies near the actual sediment subducting at the Lesser Antilles margin. Sm/La is a useful parameter because it is a measure of the extent of LREE depletion or enrichment and so is well understood from the REE pattern. Moreover, mixing lines are straight lines on plots of Sm/La vs Th/La, and so end-members are simple to calculate. Straight lines, however, could also result if both ratios were dominated by La variations, in which case Th/La and Sm/La would be highly positively correlated. This is clearly not the case, as MORB, OIB and arc data form inverse trends; in fact, for the Antilles, Th/La increases as both Th and La increase. These relationships are not consistent with La control, but are consistent with sediment–mantle mixing.
|
All of the arc data can be interpreted in this way, and the trends can be used to obtain independent estimates for the mantle and sedimentary mixing components for each arc. Returning to the case of the Lesser Antilles (Fig. 4), a line regressed through the data intersects the MORB–OIB mantle array at Sm/La of
0·9 [very near N-MORB at Sm/La = 1·0 (Sun & McDonough, 1989; Hofmann, 1988)]. This end-member approximates the mantle prior to slab additions. Each arc data point can then be projected from the mantle to calculate the high Th/La sediment component potentially involved. In practice, it is assumed that the high Th/La component has the same Sm/La as the bulk subducted sediment, but even projecting to the y-axis would not change the results significantly for the Lesser Antilles (Fig. 4). Experiments conducted at 2 GPa (Johnson & Plank, 1999) show that Sm/La is not fractionated during sediment dehydration at 700°C, but is lowered by a factor of 1·6 in sediment melts at 800°C. If Sm/La in the sedimentary component is reduced by this factor, then the resulting effect on Th/La of the sediment end-member is generally small (for Fig. 4, Th/La of 0·36 vs 0·34).
Figure 5 shows diverse mixing trends for the arcs considered here. For example, Marianas and eastern Aleutian basalts appear to mix toward different mantle compositions and different sedimentary components. The Marianas mantle is more depleted (higher Sm/La) than N-MORB (1·0), whereas the Aleutian mantle is comparable with a more enriched MORB (i.e. with a flat REE pattern), consistent with other inferences based on high-field-strength elements (Woodhead et al., 1993; Elliott et al., 1997; Pearce et al., 1999; Electronic Appendix B). On the other hand, the eastern Aleutians mixes toward sediment with higher Th/La than that for the Marianas arc, consistent with higher Th/La in the bulk sediment column subducting there (Fig. 5a). Despite complexities introduced by multiple components, crustal contamination and variable data quality (see notes in Electronic Appendix A), this procedure provides independent estimates of the mantle and sediment components for each convergent margin. This has been an elusive goal in studying subduction zone systems, where a minimum of four sources are often involved (mantle, slab sediment, slab basalt and upper plate lithosphere). Upper crust contamination is minimized here by examining only basalts, although contamination is clearly an issue for several arcs built on high Th/La continental material (see Honshu, Fig. 5h). Slab basalt aqueous contributions are also minimized by focusing on ‘fluid-immobile elements’ Th, La and Sm. Thus, this procedure attempts to isolate sediment and mantle components, and so provides a useful control in the multi-component mixing problem.
|
The composition of the sub-arc mantle
The sub-arc mantle composition, although not the focus of this paper, is an interesting by-product of the unmixing calculation. Studies of individual arcs have variously argued for N-MORB-type mantle (e.g. Stolper & Newman, 1994
), ultra-depleted mantle (e.g. Ewart & Hawkesworth, 1987; Woodhead et al., 1993) and enriched mantle (e.g. Morris & Hart, 1983; Reagan & Gill, 1986). Often, these conclusions are based on isotopic data, which can be ambiguous with respect to enriched mantle vs sediment, or high-field-strength elements, which reflect a combination of mantle depletion and slab enrichment (Elliott et al., 1997). The technique given here is useful in this regard, in that it explicitly accounts for slab enrichment (using Th/La). The primary assumption in this technique is that melting processes have a small effect on the Th/La and Sm/La of the mantle source. Strictly speaking, the mantle end-member identified here is a mantle melt and will include melting and melt-reaction processes. Partial melting will cause negligible variations in Th/La, because both elements are highly incompatible, but may contribute to variations in Sm/La. For example, 5% batch melting leads to 15% lower Sm/La than in the source (assuming DLa = 0·001 and DSm = 0·01), but this is a small effect compared with the range in MORB and arc basalts (>400%; Figs 4 and 6). Thus, Sm/La extrapolated from the arc data may be largely a function of the mantle-source composition.
Figure 6 illustrates differences in Sm/La of such inferred sub-arc mantle sources, from compositions approaching E-MORB (Java and Aleutians) to compositions more depleted than average MORB (Tonga and South Sandwich). Most melting processes will decrease Sm/La relative to the mantle source (as above), so Sm/La mantle end-members presented here are minima. For this reason, the following discussion focuses on high Sm/La (i.e. depleted mantle) regions. The inferred mantle compositions in Fig. 6 can be compared with those from Pearce & Parkinson (1993), who plotted Nb vs Y to identify depleted vs enriched mantle sources. Our results are consistent for depleted arcs (Tonga, Sandwich, Marianas), but not for arcs with higher Nb. For example, the Lesser Antilles arc plots in the enriched mantle field on Pearce & Parkinson's Nb/Y diagram, whereas the results here suggest that the Lesser Antilles mantle source may be depleted mantle typical of N-MORB. Nb may derive from enriched mantle and/or subducted sediment (Turner et al., 1996; Elliott et al., 1997), and so may be a less reliable indicator of mantle-source enrichment than projected Sm/La compositions, from which sediment contributions are explicitly removed.
|
Nonetheless, both techniques support the existence of highly (LREE) depleted mantle beneath some arcs. Several other studies have suggested that such depleted mantle may originate from melting in back-arc regions, which may deplete mantle that later melts beneath the arc (e.g. Ewart & Hawkesworth, 1987
; Woodhead et al., 1993). Among the five arc sources that are more depleted than average MORB (Tonga, South Sandwich, E. Sunda, Marianas and Banda, Fig. 6), three are in regions with nearby, recent back-arc spreading. For these three arcs, there is also a suggestion that the extent of mantle depletion in the arc correlates with spreading rate in the back-arc (i.e. Sm/La decreases from Tonga to South Sandwich to the Marianas arc as spreading rate decreases from Lau to Scotia to the Mariana back-arc). The very depleted mantle beneath Banda and E. Sunda is unexpected, as there is currently no back-arc spreading. It is possible that dynamic effects (stalling?) associated with continental margin subduction lead to progressive melt depletion beneath E. Sunda and Banda. Of the other seven arcs, which have normal MORB or more enriched mantle, there is no active back-arc spreading within 400 km of the arc. Thus, the mantle sources beneath arcs vary in composition, much like the MORB mantle, but more highly depleted sources may be generated by concurrent back-arc spreading.
Finally, this work, combined with others (Woodhead et al., 1993; Elliott et al., 1997; Pearce et al., 1999), argues strongly against OIB-type mantle in places such as the Marianas, where Sr-, Nd- and Pb-radiogenic isotopes overlap with some OIB (e.g. Morris & Hart, 1983; Stern & Ito, 1983). The preponderance of evidence now suggests that the Marianas mantle is more depleted than average MORB mantle, and that non-MORB isotopic ratios (e.g. 87Sr/86Sr, 143Nd/144Nd) and trace-element ratios (Th/La-, Th/Nb-, Ce-anomalies) derive from subducted sources.
Th/La in sedimentary components that recycle to the arc
The main purpose of the unmixing calculation is to provide estimates of Th/La in the sedimentary component that recycles to the arc. Figure 7 compares Th/La in the sedimentary column subducting at the trench with Th/La in the sedimentary component calculated from the arc mixing trends. The relationship is very close to one-to-one. Within uncertainties, most arcs mix toward a component that has the same Th/La as the sediment subducted.
|
Although the correlation is strong (R2 = 0·86), a few arcs are significantly displaced from the correlation line (Fig. 7). Most of the Honshu arc taps a component that has lower Th/La than the estimated bulk sediment. The ratio within the sedimentary column varies greatly, however, with values above 0·35 in the upper diatom and pelagic clay units, and values below 0·1 in the lower clay and chert unit (Cousens et al., 1994
; Plank & Langmuir, 1998). Thus, the average Th/La is very sensitive to the proportions of the different pelagic units. Although these pelagic units do not vary greatly in thickness on the seafloor, dynamic effects during subduction, such as frontal accretion or underplating (von Huene & Scholl, 1991), may affect the final subducted composition. Independent evidence for loss of the upper sedimentary unit in the Honshu fore-arc comes from a lack of 10Be in the arc (Tera et al., 1986), 10Be profiles in the fore-arc that are consistent with frontal accretion (Morris et al., 2002) and Pb-isotopic compositions that are shifted toward the lower pelagic units (Cousens et al., 1994; Gust et al., 1995). Loss of the upper 165 m of the Honshu sediment section will lower Th/La from 0·28 to 0·24, in better accord with the main arc trend (Fig. 5). Alternatively, Th/La will increase in the Honshu arc average if some of the anomalously high Th/La compositions shown in Fig. 5h are included, but these compositions are generally found in more silicic volcanics (>53 wt % SiO2), which are interpreted as crustally contaminated.
Arcs that plot significantly above the correlation line in Fig. 7 include Tonga, Guatemala, E. Sunda and the S. Antilles. The high S. Antilles values arise from the unusually high values in the southern end of the arc, in Kick-em-Jenny volcano and the M-series of Grenada (Fig. 5g). Basalts from these islands have higher 87Sr/86Sr (>0·7045) than those from the rest of the Lesser Antilles arc, and this feature has been interpreted as a product of crustal assimilation (Davidson, 1987; Thirlwall et al., 1996). If such high 87Sr/86Sr samples are excluded from the S. Antilles average (which excludes all samples from Kick-em-Jenny and St Lucia), then the S. Antilles sedimentary component shifts from 0·40 ± 0·02 to 0·36 ± 0·02, closer to the estimate for the subducting sedimentary column (0·31 ± 0·02). However, most of the very thick sediment wedge (>1500 m) subducting beneath the S. Antilles is unsampled, and so the sedimentary column is poorly known (Plank & Langmuir, 1998). Thus, better control on the sediments and volcanics in the southern portion of the Lesser Antilles arc are needed before determining whether this fractionation is real. Like Honshu, the E. Sunda sediment average depends critically on the proportion of the main lithologic units present (Plank & Langmuir, 1998), namely, Cretaceous clays (with Th/La = 0·25) and Cenozoic carbonate turbidites (with Th/La = 0·34). Nearby ODP Site 765 contains more carbonate turbidites than DSDP 261, and so has a bulk Th/La = 0·29 instead of 0·26 (Plank & Ludden, 1992), which is closer to the arc value and may be more typical of the Australian margin crust subducted beneath E. Sunda. Thus, given uncertainties in the layer thickness of the different units (particularly the carbonate turbidites), the E. Sunda arc may reflect the bulk sediment subducted. Indeed, increases in Th/La along the arc, from Java to E. Sunda to Banda, and from the northern to the southern Antilles, are qualitatively consistent with increasing proximity to continental crust (and high Th/La terrigenous material) in each of those cases. Unfortunately, proximity to continental crust leads to greater uncertainties in bulk sediment estimates, as a result of thick (and undrilled) sedimentary sections, as well as the dynamic complexities of incipient arc–continent collision.
Unlike the Antilles and Sunda cases, Guatemala and Tonga are difficult to reconcile with small differences in layer thicknesses or uncertainties in the sediment subducted. In both cases, sedimentary compositions are well known, and all of the arc values exceed the bulk sediment composition (Fig. 5b and c). Moreover, all of the Tonga sedimentary units have Th/La < 0·07, except for the upper brown clay unit (Th/La = 0·25). The thickness of the brown clay unit would have to be increased from 10 to 80 m in order to raise the bulk sediment Th/La to that of the Tonga arc (0·115). This seems unlikely, as it would require doubling the thickness of the sedimentary column and, whereas under-plating is always likely, spontaneous thickening is less so. Like Tonga, all of the Th/La ratios in the Guatemala arc data exceed the value in the bulk sediment. The Guatemala bulk sediment has very low Th/La because of the low ratio in the lower pelagic carbonate unit. This is the only dominantly carbonate section in all of the margins considered here, and so the shift of the Guatemala arc could derive from fundamentally different partitioning between fluids and carbonate vs siliciclastic lithologies. Alternatively, Carr et al. (1990) proposed that Sr- and Nd-isotopic compositions for some Guatemala volcanics have been affected by crustal contamination, and this could raise Th/La as well. Several basalts in Guatemala and El Salvador, however, which have high 143Nd/144Nd (>0·5129; Fuego, Conchagua) and do not appear crustally contaminated, still have higher Th/La than the bulk sediment subducted. Thus, this fractionation may be real and could arise from the different carbonate lithologies subducted at the Middle America trench. Finally, Tonga and Guatemala sediments are noteworthy in their high relative abundance of biogenic apatite, which may fractionate Th from La in the direction observed (see below).
To summarize this analysis, arc basalts clearly inherit the Th/La of the local sedimentary column subducted. Within the uncertainties of the calculation, the arc data trend toward a sedimentary mixing component that has the same Th/La as the bulk sediment at the trench. Apart from a few exceptions (Tonga and Guatemala), there is very little evidence that Th and La are fractionated in the subduction zone during sediment recycling to the arc. The predominance of sedimentary columns with bulk Th/La < 0·35 (Fig. 7) and the significant mantle component in most arcs (Fig. 3) means that arc addition has little effect on Th/La in the continental crust. The Th/La in average arc material is 0·25 ± 0·07 (Table 1), compared with 0·27 ± 0·05 for the bulk continental crust (Fig. 1). Although it may appear that the bulk continental crust inherits its Th/La from arcs, the reverse is the case. The prior analysis demonstrates that arcs inherit their Th/La from subducted sediment, and so they do not add ‘new’ high Th/La material to the continents. The net effect of subduction on Th/La in the continental crust is null: crust is removed as subducted sediment with Th/La of 0·24 ± 0·06 (Table 1), and replaced as arcs with Th/La of 0·25 ± 0·07. Subduction does not appear to create the high Th/La of the continents.
|
CREATING HIGH Th/La IN THE CONTINENTAL CRUST |
|---|
|
TOPABSTRACTINTRODUCTIONTh/La VARIATIONS IN MARINE...CREATING HIGH Th/La IN...CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
If the high Th/La of the continents is not created in subduction zones, then where is it created? Possibilities include other kinds of crustal additions (intraplate magmatism and Archaean magmatism) and other kinds of crustal losses (lower-crustal foundering/delamination), in combination with intracrustal processing. The first issue to consider, however, is the uncertainty in the estimate of the composition of the bulk continental crust.
Uncertainties in the composition of the bulk continental crust
The composition of the bulk continental crust is difficult to estimate with certainty because so little of it is exposed. Whereas the upper continental crust composition is fairly well constrained from average shale and loess compositions (Th/La = 0·33 ± 0·05, average UCC estimates in Fig. 1), the bulk composition of the lower crust is not well known. Current estimates are based on granulite xenoliths and exposed granulite terranes, for which geobarometry indicates equilibration at lower-crustal pressures (Rudnick & Presper; 1990; Gao et al., 1998). The database of Rudnick & Presper (1990), with over 2000 analyses of such granulites, provides a good sampling of the total range in granulite compositions, but there are large uncertainties in knowing how to weight the different compositions to obtain a lower-crustal average. The approach is typically to calculate medians or means from available analyses, and this has yielded estimates that are generally consistent with average P-wave velocities and heat-flow constraints (Rudnick & Fountain, 1995; Gao et al., 1998; Rudnick et al., 1998).
Given the range of lower-crustal compositions, however, is it possible to create a bulk continental crust with Th/La 0·1, similar to the primitive mantle? Figure 8a shows the trend in Th vs Th/La for granulite xenoliths (La vs Th/La is scattered). A minimum value for Th/La in the bulk crust can be calculated by taking the lowest Th/La regions (Chudleigh and McBride; Rudnick et al., 1986; Rudnick & Taylor, 1987; see Fig. 8a) and assuming they make up the entire lower and middle crust (70% of the crust). Even by using these extreme compositions, the bulk continental crust will still have Th/La > 0·2. There is poor leverage on this calculation, because the lowest Th/La granulites have the lowest Th concentrations, and so the least effect on the bulk crust. These calculations illustrate that the upper crust, which is the best-constrained crustal reservoir, is the dominant control on the composition of the continents because it has such high Th and La concentrations (10 s of ppm). Given the range of possible compositions for the lower crust (Fig. 8a), the bulk continental crust still has significantly higher Th/La than the mantle (0·27 ± 0·05; average BCC estimates in Fig. 1).
|
Another way to lower Th/La in the bulk continental crust is to add eclogite-facies material below the seismic Moho. Current estimates of the bulk continental crust are based to some degree on seismic velocity models, which may not include seismically fast pyroxenites within the petrologic crust. For example, xenolith suites in Miocene lavas from the central Sierras contain abundant garnet clinopyroxenites, which, based on geobarometry, are derived from below the seismic Moho (Ducea & Saleeby, 1996
). Sparse chemical data from these samples indicate that they are highly depleted in Th and La (0·1 ppm Th and 3 ppm La; Ducea, 2002) and so have little leverage on bulk continent compositions. Other sub-Moho garnet clinopyroxenites have Th/La 
0·1 (Griffin et al., 1988; Wendlandt et al., 1993), and so cannot drive the continents to mantle values <0·1. Even using the most extreme mafic granulite compositions (McBride xenoliths; Rudnick & Taylor, 1987), adding a 20 km eclogite-facies root to the entire average continental crust (as for the central Sierras) will still not create a bulk continental crust with Th/La < 0·2. Moreover, garnet pyroxenites are not abundant in most lower-crustal or mantle xenolith suites (Schulze, 1989; Rudnick & Presper, 1990) and so may not constitute a major reservoir beneath the continents. Thus, given that the major budget of Th and La resides in the upper crust, and given that the upper crust has a well-constrained Th/La 0·3, it is difficult to create a bulk continent with Th/La < 0·2, even by adding a ‘lowermost’ crust with extreme compositions.
Intraplate magmatism
Modern oceanic mantle melts (e.g. MORB and OIB) possess low Th/La, generally <0·1 (Fig. 4). Other kinds of intraplate magmas, however, contribute directly to the continents, and thus it is important to asses their Th/La. Indeed, O'Nions & McKenzie (1988) argued that the continents are formed from low-degree mantle melts, typical of those found in small, intraplate seamounts in the Pacific. These seamounts, however, have Th/La < 0·15 (Batiza & Vanko, 1984) and so this model fails to explain the high Th/La (>0·2) of the bulk continental crust. In fact, intraplate basaltic magmas with Th/La > 0·2 are rare. Most flood basalts overlap with OIB compositions, at Th/La < 0·15. Exceptions include some compositions between 0·15 and 0·25 that occur in the Siberian, Columbia River and Keweenawan provinces, where crustal or subduction contamination has been invoked (Carlson, 1984; Klewin & Berg, 1991; Wooden et al., 1993). Th/La ratios > 0·2 are uncommon in potassic rift volcanics, lamproites and nephelinites (Fraser et al., 1985; Anthony et al., 1989; Wilson, 1989). Although some of the Goudini carbonatites have high Th/La (>0·3), this is not a general feature of carbonatites (Nelson et al., 1988). Most carbonatites have Th/La < 0·10, and the few high ratios could represent recycled continental material (Nelson et al., 1988). Although many potassic volcanics do not have high Th/La (East African; Wilson, 1989), those from the Roman province can have Th/La > 0·5 (Rogers et al., 1985). These leucitites, however, may also reflect subducted sediment in their source (Rogers et al., 1987). Thus, in general, mantle-derived magmas do not have high Th/La, except when sediment subduction is independently implicated. The high Th/La of the continents does not appear to be produced by any mantle-melting process.
Archaean magmatism
Another possibility for explaining the high Th/La of the continental crust is to create this feature in the Archaean, with modern tectonic processes having little net effect. This mechanism is analogous to models that create the silicic crust in the Archaean by processes different from those today, such as tonalite genesis during slab melting in the hotter Archaean mantle (e.g. Martin, 1986; Rudnick, 1995). Th/La of the bulk Archaean crust is estimated at 0·17–0·19 [by Rudnick & Fountain (1995) and Taylor & McLennan (1985), respectively]. A more recent revision based on heat-flow constraints (McLennan & Taylor, 1996) yields Th/La of 0·19 (for a 1:1 mixture of felsic and mafic components). These low values for the Archaean continental crust are consistent with low Th/La in average Archaean mafic and felsic end-members (<0·22; Taylor & McLennan, 1985), in average Archaean felsic, andesitic and basaltic volcanics (<0·2; Condie, 1993), and in average lower crust from platform and shield areas (<0·15; Rudnick & Fountain, 1995). These observations are also consistent with the low Th/La of modern adakites (primitive andesites from the western Aleutians have Th/La < 0·15; Kelemen et al., 2003b), which are thought to be derived from melts of the basaltic portion of subducting slabs that interacted with the overlying mantle wedge during ascent. Finally, Condie (1993) showed that Archaean tonalites, granites and felsic volcanics have lower Th/La than their Phanerozoic counterparts. Thus, there is little evidence that the Archaean continents had as high Th/La as the modern bulk crust (0·17–0·19 vs 0·27 ± 0·05). The fractionation between the Archaean continental crust (Th/La 0·18) and the primitive mantle (Th/La 0·12) may be enough to explain the depletion of the MORB reservoir (Th/La 0·05), but it is not enough to explain Th/La in the modern continents.
Further evidence that the Archaean continents had low Th/La comes from the secular increase in this ratio recorded in independent estimates of upper-crust age provinces (Fig. 9). Both sedimentary rock averages (Taylor & McLennan, 1985) and areally averaged juvenile crustal provinces (Condie, 1993) show an increase in Th/La with time (Fig. 9). This observation provides further evidence that high Th/La was not an intrinsic feature of the Archaean continents, but rather that processes operating in the Proterozoic and Phanerozoic have resulted in an evolution toward higher continental Th/La.
|
Crustal fractionation and foundering
There is little evidence from the preceding discussion that magmatic additions from the mantle (at subduction zones, in intraplate settings, or during the Archaean) create the high Th/La of the modern continents. This leaves open the possibility of crustal subtractions—in the form of subducted sediment or of foundered/delaminated lower crust. The similarity in Th/La of subducting sediments (0·24 ± 0·06; Table 1), arc basalts (0·25 ± 0·07; Table 1) and the bulk continental crust (0·27 ± 0·05; Fig. 1) means that subduction recycling has no net effect on the continents. On the other hand, there is a large fractionation of Th/La within the continents (Fig. 8a), with the average lower crust (0·15) being more than a factor of two lower than the average upper crust [0·36; using self-consistent estimates of Rudnick & Fountain (1995)
]. Periodic loss of the lower crust during delamination or foundering episodes (see below) could remove low Th/La material, thus enriching the continents with time. This will not work if under-plated mafic material simply founders back into the mantle. Mafic material that enters the crust must differentiate into a low-Th/La lower crust and a high-Th/La upper crust for this process to have a net effect on the bulk continent. Several fractionation processes may occur to create a low-Th/La lower crust.
There is abundant evidence that the lower crust contains materials with very low Th/La (<0·1; Fig. 8a) and that some of this depletion is created in the crust. Lower-crustal xenoliths and regional averages are plotted in Fig. 8b to demonstrate that many of these compositions are unlike MORB, OIB or arc basalt. In particular, most mafic granulites and cumulates, as well as regional averages from Chudleigh, McBride and Navajo xenolith fields, are more depleted in Th/La for their Sm/La than common basaltic compositions, despite the fact that their protoliths are basaltic magmas. Some mafic granulites have Th/La within the MORB–OIB array (Fig. 8b) and, although these could reflect primary ratios, many of these granulites have been interpreted as arc basalts (Wendlandt et al., 1993; Mattie et al., 1997), and therefore have probably been shifted to lower Th/La compositions in the crust. These common low-Th/La compositions in the mafic granulite xenolith database could be a result of Th loss during granulite–facies metamorphism, as has been argued by Rudnick & Presper (1990) and Condie & Myers (1999), or could also be an intrinsic feature of plagioclase- and clinopyroxene-rich cumulates (Kelemen et al., 2003a). In fact, as many as three separate processes could create low Th/La in lower-crustal mafic granulites: (1) Th loss relative to La in melts during crustal anatexis; (2) Th loss relative to La in fluids during granulite-facies metamorphism; (3) cumulate processes, individually or in combination.
Fractionation of Th/La within the continental crust
Igneous processes are clearly major drivers in the differentiation of the continental crust, as evidenced by the abundance of granites and granodiorites in the upper crust. The prominent negative Eu-anomaly in the upper crust also points to its formation by intra-crustal melting or crystal fractionation, with plagioclase in the residue (Taylor & McLennan, 1985). The high Th/La of the upper crust also has its source in granitic material. Granite and granodiorite plutons and batholiths commonly have Th/La > 0·5, whether they be I-type, S-type or A-type (Winter, 2001). Some batholiths may be internally zoned, with high-Th/La material at the highest elevations (Sawka & Chappell, 1988), resulting from magmatic differentiation. The complement to the granitic upper crust must exist as restites and/or cumulates in the middle and lower crust. Restites have been identified in lower-crustal granulite xenolith suites (Hanchar et al., 1994) and most of them have low Th/La (Fig. 8a).
Accessory minerals appear to control the budget and fractionation of Th and the LREE during igneous fractionation (Sawka & Chappell, 1988). Although apatite, titanite, allanite, monazite, thorite and zircon may all be formed during igneous differentiation (Bingen et al., 1996; Bea & Montero, 1999), only apatite and zircon are common accessory minerals in granulite xenoliths from the lower crust (Leyreloup et al., 1982; Rudnick & Taylor, 1987; Roberts & Ruiz, 1989; Rudnick & Presper, 1990; Wendlandt et al., 1993). Based on sparse partitioning data (from phenocryst-matrix data; no experimental Th-partitioning data exist for either apatite or zircon), apatite appears to favour La > 10x over Th, whereas zircon favours Th > 10x over La, with respect to silicic melts (Mahood & Hildreth, 1983; Mahood & Stimac, 1990; Bea et al., 1994). These inferences are also consistent with measured Th and La in apatite and zircon from metapelite and granodiorite whole rocks (Sawka & Chappell, 1988; Bea & Montero, 1999). Thus, the effect on the bulk partition coefficients will depend on the relative abundance of the two minerals. For most lower-crustal compositions, apatite will be >10x more abundant than zircon (based on the P and Zr contents of these rocks; Rudnick & Presper, 1990) and so apatite will control the relative partitioning of Th and La, producing high-Th/La melts and low-Th/La residues. Adding other potentially dominant residual silicate phases (clinopyroxene, orthopyroxene, plagioclase, garnet and amphibole), most of which partition La over Th (Green, 1994), accentuates this effect. If 0·4% apatite and 0·01% zircon exist in the residue (conservative averages based on the range of P and Zr in lower-crustal granulite xenoliths), Th/La will increase in the melt by a factor of 1·8–4·1 upon 25% melting (or 75% crystallization), using D values reported in Table 2. Thus, the observed high Th/La of felsic igneous rocks (granites and granodiorites), the low Th/La of lower-crustal rocks and the theoretical igneous fractionation of these elements support igneous differentiation as one likely cause of the large Th/La fractionation between the upper and lower crust.
|
Although igneous processes clearly occur and can create the correct sense of Th/La fractionation, some lower-crustal granulites possess extreme Th-depletion, which is difficult to explain by igneous differentiation. Examples include mafic granulites from the Fraser complex, which show >3-fold depletion in Th with no change in the REE pattern (Condie & Myers, 1999
). Felsic gneisses from the Dharwar craton in India also show a 100-fold decrease in Th from amphibolite- to granulite-facies metamorphism, with little change in major-element content (Hansen et al., 1995; Fig. 8a). The latter study, in particular, argues for a low-aH2O fluid (CO2- and alkali-rich ‘salt melt’) as the agent of Th-depletion. This process is more difficult to model, because the mechanism of Th-depletion has yet to be quantitatively constrained. Nonetheless, the extreme Th-depletion observed in some granulites argues for a distinct fluid-mediated process that accompanies granulite-facies metamorphism in many lower-crustal regions (e.g. Dharwar, Lewisian and Lofoten granulites; Hansen et al., 1995).
Given that many lower-crustal granulite xenoliths are interpreted as mafic cumulates (Wilkinson & Taylor, 1980; Rudnick et al., 1986; Halliday et al., 1993; Lee et al., 1993; Kelemen et al., 2003a), crystal fractionation is another mechanism that may cause internal differentiation of the continents. In general, however, crystal fractionation of mafic to intermediate compositions creates mafic cumulates with low Th/La, but has little effect on Th/La in the liquid. For example, 70% fractional crystallization of plagioclase, clinopyroxene and olivine will cause Th/La in the liquid to increase by only 11%, whereas Th/La in the solid residue will be less than half that of the starting composition (using D values in Table 2). This calculation is in accord with observations from comagmatic volcanic suites, which show little change in Th/La in liquids produced by crystal fractionation (of olivine, clinopyroxene, plagioclase, orthopyroxene, oxide), up to the point of apatite saturation (Fig. 10). For example, the Th-enrichment in Anatahan andesites relative to basalts suggests >70% crystal fractionation, during which Th/La varies by <10% (Fig. 10). Not until apatite saturates in silicic compositions (63–64 wt % SiO2 for Anatahan) does Th/La begin to increase at a higher rate. Thus, the first 50–70% of crystal fractionation will drive compositions from basalt to andesite and create a large mass of low Th/La cumulates, but the low overall D values (Table 2) produce little effect on Th/La in the fractionated liquids. The high Th/La of the upper crust is generated predominantly in silicic liquids with high D values (Table 2), reflecting saturation in LREE-enriched phases such as apatite.
|
In summary, basaltic magmas will enter the crust and crystallize, leaving low-Th/La cumulates in the lower crust, and slightly higher Th/La in the melt. Further crystal fractionation, partial melting and/or granulite-facies metamorphism (involving REE-rich accessory phases) is then required to create a high-Th/La upper crust. The final step in enriching the bulk continents in Th/La is then to lose the low-Th/La material from the bottom.
Lower-crustal foundering
Many studies have outlined the geochemical evidence and physical mechanisms for lithospheric delamination or foundering. Bird (1979) and Houseman et al. (1981) made the case for foundering of mantle lithosphere, but lower-crustal foundering may also occur because of the negative buoyancy of ultramafic cumulates and garnet-bearing mafic lithologies (e.g. Herzberg et al., 1983; Arndt & Goldstein, 1989; Turcotte, 1989; Kay & Kay, 1991; Glazner, 1994; Jull & Kelemen, 2001). Such dense lower crust may develop convective instabilities that sink into the mantle, especially for high temperatures at the Moho (>700°C), high background strain rates, wet olivine mantle rheology and/or thickened crust (Jull & Kelemen, 2001). Cumulate pyroxenites and gabbronorites are particularly susceptible to such a process, and may founder shortly after they form (<10 Myr; Jull & Kelemen, 2001). Cold cratonic foundering episodes are unlikely to occur, due to the high viscosity of the underlying mantle at low temperature, whereas arc, rift or plume settings may all promote lower-crustal loss (Jull & Kelemen, 2001). Partial melting may then occur within the foundered crust (e.g. Ducea & Saleeby, 1998; Kelemen et al., 2003a), or subsequently within the crust as hot mantle asthenosphere replaces foundered lithosphere (e.g. Kay & Kay, 1993; Samson et al., 1995). Thus, lower-crustal foundering provides an ideal mechanism to both generate high-Th/La melts and remove low-Th/La cumulates and restites from the continental crust.
Given the kinds of processes outlined above, it is possible to model quantitatively the observed increase in continental Th/La since the Archaean. The purpose here is not to argue for a given crustal growth model, but to demonstrate that within reasonable parameters, the continents can evolve in Th/La to reach the present composition of the BCC. Models begin at 3 Ga, with a BCC Th/La of 0·17 [i.e. Archaean bulk crust from Rudnick & Fountain (1995)]. Arc magma is added at a constant rate, undergoes crystal fractionation which creates arc andesites and mafic cumulates (using ‘mafic’ D values in Table 2) and then partially melts to create granites and restite (using ‘felsic’ D values in Table 2). The distinction between cumulates and restites is somewhat artificial, but is introduced to accommodate the changing D values in a simple conceptual way. After igneous differentiation, a portion of the cumulates and restite is lost to the mantle (see Fig. 11 for schematic model). The leftover arc crust is mixed into the existing BCC, thus raising its Th/La. In accord with results from above, arc magma additions attain their Th and La primarily from recycled material from the upper continental crust. In practice, Th and La are extracted from an upper-crustal mass equal to the mass of new arc growth (after foundering), such that there is no net continental growth in this model. Upper-crustal Th and La is mixed with a mass of mantle Th and La [based on depleted mantle values of Salters & Stracke (2004)] required to supply the arc magma mass from 15% mantle melting. Lower mantle-melt fractions lead to a dominance of mantle Th and La (with low Th/La), whereas higher melt fractions may not be typical of average arcs.
|
Figure 12 shows model output, for a wide range of cumulate and restite fractions created and lost. At zero loss of mafic residues (cumulates or restite), the continents actually evolve to lower Th/La than the Archaean starting point, because of the net input of mantle material with low Th/La. The BCC only evolves to higher Th/La by processing the mantle and recycled crustal material through igneous differentiation, and loss of low-Th/La residues. As discussed above, the largest effect on Th/La occurs during felsic differentiation, as a result of the effect of an LREE-enriched accessory mineral phase such as apatite. The model is thus particularly sensitive to the proportion of restite lost, because the restites have low Th/La (<0·1) yet higher concentrations than the cumulates (because of the difference in D values for the two steps; Table 2). For this reason, loss of cumulates is less efficient than loss of restite, and so drives up total crustal losses without a comparable effect on BCC Th/La (Fig. 12). Models that successfully create today's BCC generally require 60–100% loss of restite and, depending on the additional loss of cumulates, 25–60% total loss of mafic residues (e.g. cumulates plus restite). The lower limit (25%) derives from a minimum of 10% cumulate loss [model (a) in Fig. 12]; it seems unreasonable to founder 60–75% restite but only 10% cumulates, although this could be consistent with the lower-crustal melting and foundering mechanism envisioned by Ducea (2001)
. The upper limit (60%) derives from a maximum of 100% restite loss [model (e) in Fig. 12].
|
The model results in Fig. 12 are also dependent on the initial arc magma flux rate, which is 7% of the continental mass, each 100 Myr. Lowering the rates will drive results to 100% restite loss. The 7% rate is >3x higher than the current frequency of juvenile ages over the past 100 Myr, which is closer to 2% (Condie, 1998
), or than Reymer & Schubert's (1984) magma-addition rates to the continents, which is also close to 2% of the modern continent mass over the past 100 Myr. The net arc addition rate (after 50% foundering), however, is closer to these estimates. Models with net crustal growth also require larger residue losses. Increasing the mass of the continental crust ultimately requires larger inputs from the mantle, which bring more low-Th/La material to the continents. Such mantle-derived material will have to be highly processed to increase Th/La to that of the BCC. For example, an OIB magma would require 20% partial melting (using felsic D values), and then 90% loss of restite to raise Th/La from 0·1 (mantle value) to 0·27 (average BCC value). Thus, net-growth models require more processing and foundering than no-growth models that strongly recycle upper continental crust.
This modelling exercise demonstrates that the bulk continental crust can evolve to high Th/La (0·27 ± 0·05) for a wide range of parameters, which ultimately require 25–60% loss of the arc crust through foundering of mafic cumulates and restite. This range is somewhat of a minimum, given that lower mantle-melting fractions, lower magmatic-flux rates and net crustal growth all require larger restite and total residue losses. Is this magnitude of foundering reasonable or observed? Ducea (2002) has made several arguments that the Sierra Nevada batholith has experienced massive foundering on a grand scale. The current 35 km of granitoid crust appears to have survived nearly 100% foundering of 35–50 km of lower-crustal pyroxenite residues, for total crustal losses of 50–60%. Jull & Kelemen (2001) demonstrated how arc gabbronorites at >25 km depth have a positive density contrast with respect to the mantle, and that effect may limit the thickness of arc and continental crust. Thus, 25–60% foundering of arc crust may be reasonable for continental arcs originally >33–62 km thick, such as the Andes, where delamination has been proposed (e.g. Kay & Kay, 1993), or for mountain belts in general (Nelson, 1992). A study of the Irish Caledonides (Draut et al., 2002) also makes geochemical and rheological arguments for large-scale crystal fractionation and loss of up to 50% of mass to foundering.
Other predictions of lower-crustal foundering
Thus far, minor elements (Th and La) have been used to identify a major process operating on the continents. Others have argued for the same mechanism to explain systematics for the most important major element—silicon (see reviews of Taylor & McLennan, 1985; Rudnick, 1995; Kelemen et al., 2003a; Davidson & Arculus, 2004). The fractionation & foundering model described here provides one way to create high silica in the bulk continental crust (e.g. 60 wt % SiO2; Rudnick & Fountain, 1995) from low-silica magmatic additions from the mantle (e.g. basalts with 50 wt % SiO2). For example, if arc magmas are added to the continents with 51% SiO2, then 50% foundering of restites and cumulates with 43% SiO2 will create a bulk continent with 59% SiO2. This calculation is compatible with the model results above (25–60% foundering calculated from Th/La systematics), but it demonstrates that the residue must have very low silica. Although such low silica is well below the average of the lower-crustal xenolith population (49–51 wt % SiO2), approximately 6% of the xenolith database consists of rocks with 
43 wt % SiO2, in a variety of assemblages (including olivine-rich cumulates and garnet-rich restites). It is possible that such low-silica samples are underrepresented in the current lower crust because they are gravitationally unstable and have foundered, although this is admittedly an unsatisfying conclusion. On the other hand, the Sierras again offer an example consistent with silica, Th/La and mass-flux constraints. The average of Sierran lower-crustal pyroxenite xenoliths, which are viable residues for the Sierran batholith, is 44·8 wt % SiO2 (Ducea, 2002); 55% foundering of such lower crust (Ducea estimated 50–60%) could create a continent with 59% SiO2 from a mafic mantle melt (51% SiO2). Thus, to the extent to which such Sierran-type episodes are significant in the history of continental evolution, the same model that satisfies Th/La may also satisfy silica.
The other question is whether foundered lower crust ever reappears in mantle-derived materials, such as ocean island or ocean flood basalts. Arndt & Goldstein (1989) proposed this to explain some aspects of mantle heterogeneity. They noted how old, depleted lower crust could evolve to isotopic compositions (e.g. moderate 87Sr/86Sr, low 143Nd/144Nd and low 206Pb/204Pb; Rudnick & Goldstein, 1990), typical of EM-1-type mantle (Zindler & Hart, 1986). Although most recent studies favour subducted sedimentary components in the source of EM-1 and EM-2 mantle (e.g. Weaver, 1991; White & Duncan, 1996; Eisele et al., 2002), it is difficult to discount lower crust as also contributing to mantle enrichment. Notably, lower crust has radiogenic Os (Saal et al., 1998), low Ce/Pb (Rudnick & Fountain, 1995) and lies predominantly within the terrestrial Hf–Nd array (Vervoort et al., 2000). These characteristics are consistent with EM-type mantle. Th/La in lower crust is also compatible with EM-type mantle. Using the results of the fractionation & foundering model above, the lower-crustal flux to the mantle would have Th/La of 0·08–0·12—significantly lower than bulk continental crust, but higher than depleted mantle (0·05), and similar to many OIB. Thus, at least on chemical grounds, some mantle enrichment could derive from foundered lower crust. Consideration of mass fluxes, however, suggests that lower crust is a minor component in the mantle. Assuming 25–60% total crustal losses (Fig. 12) at the rates given in Fig. 11, the total amount of foundered residues in the mantle would constitute a mass of roughly 50–100% of the mass of the modern continents. Thus, foundered lower crust would constitute only 0·3–0·65% of the total mantle mass, compared with 3·5% for subducted slabs over the same 3 Gyr period [using slab fluxes of von Huene & Scholl (1991)]. From the mantle's perspective, subduction is probably one order of magnitude more important in enriching the mantle than lower-crustal foundering.
Lack of Th/La fractionation in subduction zones
The final consideration is why Th and La are fractionated from one another during intracontinental differentiation, but not during slab metamorphism or melting. Arc volcanic data show clearly that Th and La are not substantially fractionated from one another during the process that transports material from sediments on the subduction plate to the source region of arc magmas. The simplest explanation for this would be that sedimentary material is transported in bulk, and so the arc simply inherits the bulk sediment compositions. There are several problems with bulk mixing, however. First, it is difficult to imagine how the subducting sediment would not evolve fluids or melts during prograde slab P–T paths. Secondly, bulk mixing of sediment also fails to produce the correct isotopic and trace-element ratios in arc magmas, particularly those ratios involving Nb or Ta (Turner et al., 1996; Elliott et al., 1997; Class et al., 2000). Instead, it is much more likely that sedimentary material is lost from the slab as fluids or melts. Recent studies of high P–T, blueschist- and eclogite-facies metapelites find a general lack of mobility of Th and REE up to 650°C, and a lack of Th/La fractionation in prograde paths (Arculus et al., 1999; Spandler et al., 2003). This supports the lack of fractionation of these elements during subduction metamorphism, but many lines of evidence point toward slab sediment melting to explain the large amount of Th recycled to arcs (Johnson & Plank, 1999, and references therein). So, if sediments melt on the downgoing plate, why are Th and La not fractionated as they are during melting in the continental crust?
As argued above, accessory phases can dominant Th/La fractionation during crustal differentiation. As mantle-derived mafic melts evolve into silicic continental crust, apatite may, in large part, control Th/La fractionation. There are reasons to expect that apatite will also be found in equilibrium with silicic melts generated in the subducting plate. In addition to models that predict apatite saturation (Watson, 1980), the high-pressure sediment-melting experiments of Johnson & Plank (1999) show that apatite is stable on both sides of the solidus, from 600 to 900°C at 2 GPa. As expected, these melts show a factor of 1·5 increase in Th/La. However, this starting composition is unusual in that it contains >3 wt % fish teeth apatite (1·6 wt % P2O5), which has forced apatite saturation. A recent study (Hermann, 2002) has argued instead for the importance of allanite in controlling Th and LREE budgets in subducting materials. Allanite is equally problematic, as it would also strongly increase Th/La in coexisting melts (Hermann, 2002). It is not clear, however, that allanite is a relevant accessory mineral during sediment melting. The allanite-bearing eclogites studied by Hermann (2002) had alkaline mafic protoliths unlike marine sediments, and the experiments conducted were all P-free, and so unable to saturate phosphate minerals such as apatite or monazite.
In the absence of relevant experiments, other crustal analogues can be used to predict the accessory phase assemblage in equilibrium with slab-sediment melts. Subducting sediments are predominantly pelitic in composition and, in this way, they diverge dramatically from most igneous rocks in the continental crust. Pelites (shales) are notably poor in calcium; to first order, GLOSS consists of 7% biogenic calcite and 93% of a terrigenous component with 2·1 wt % CaO. Thirteen of the 28 trench sections considered by Plank & Langmuir (1998) are virtually calcite free and have <2·5 CaO. Melts of calcium-poor pelites are peraluminous, and generally contain the accessory minerals monazite + zircon, in contrast to calcium-rich melts, which contain apatite + allanite + zircon ± sphene (Lee & Silver, 1964; Rapp & Watson, 1986; Broska et al., 2000). If monazite, and not apatite or allanite, is the major REE phase in equilibrium with metapelite melts in the subducting slab, then it may have a dramatically different effect on Th and La partitioning. Monazite contains weight percent quantities of both Th and La, and evidence from natural monazites indicates that (Th/La)mon/(Th/La)melt is near unity [based on monazite–granite pairs of Wark & Miller (1993) and Bea & Montero (1999); and the granite-hosted monazite average of Rosenblum & Fleischer (1995), which has Th/La of 0·5, which is very near granites themselves]. If subducting sediments have the REE abundances of GLOSS, and generate peraluminous melts (Nichols et al., 1996; Johnson & Plank, 1999), then monazite solubility models (Rapp & Watson, 1986; Montel, 1993) predict monazite saturation at temperatures up to 775°C. Monazite has large but similar D values for Th and La, and so melts in equilibrium with monazite would have Th/La similar to their starting composition.
In summary, the pelitic composition of subducting sediments may stabilize different accessory minerals (e.g. monazite) from those (e.g. apatite and allanite) stable in higher-Ca igneous protoliths in the continental crust. These differences would explain the large fractionation of Th and La during continental differentiation, and the lack thereof during subduction. Exceptions may help to test these ideas. For example, all 13 trench sections considered here have low-Ca bulk compositions (<3·5 wt % CaO, or <10% calcium carbonate), except for Guatemala and E. Sunda (60 and 16% calcium carbonate, respectively). The dominance of monazite in low-Ca sediments and apatite–allanite in high-Ca sediments could explain the anomalously higher Th/La in the Guatemala and E. Sunda arcs (e.g. significantly above the line in Fig. 7). The high fish teeth apatite content of the Tonga sediments (with a factor of five higher P2O5 than the other trench sections) could also explain the anomalously higher Th/La in the Tonga arc, which is a factor of 1·6 higher than the bulk sediments—almost exactly the same factor predicted in the apatite-rich experiments of Johnson & Plank (1999) based on a Tonga sediment starting material. Thus, the few arcs that show a shift in Th/La are those where unusually apatite-rich or carbonate-rich sediments subduct. Nonetheless, the dominantly pelitic (low-Ca) composition of marine sediments leads in most cases to no significant fraction of Th and La in the subduction zone. Clearly, further experimental work on accessory mineral phase stability and partitioning is needed to clarify the effect of melting on Th/La in continental vs subducting slab conditions.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONTh/La VARIATIONS IN MARINE...CREATING HIGH Th/La IN...CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
This paper develops the elemental ratio of Th/La as a useful tracer of sediment recycling at subduction zones. Th/La can be traced from the subducting sedimentary package on the downgoing plate to arc magmas, and the combined dehydration and melting processes that deliver slab material to the arc source do not typically fractionate Th from La. Th/La is thus one of the best tracers of slab sediments in arc magmas, and should help to both mass balance modern fluxes as well as identify recycled sediment in ancient arc systems.
Although useful in studying arc systems, the lack of fractionation of Th from La during subduction presents problems for creating the modern continental crust by arc accretion alone. Most other proposed modes of continent formation (intraplate magmatism, small-degree melts, Archaean adakites) also fail to create a continent with higher Th/La than the mantle. Instead, several differentiation processes naturally occur within the continental crust (partial melting, crystal fractionation and granulite-facies metamorphism) to create a high-Th/La upper crust and low-Th/La lower crust. Foundering of 25–60% of lower-crustal cumulates and restite into the mantle can create a bulk continental crust that evolves to high Th/La (>0·2) in 3 Ga. Thus, Th/La provides evidence for continental fractionation & foundering, as well as a means to quantify this process. The weight of evidence from geochemical, geological and geodynamic arguments favours a major process of lower continental foundering throughout most of Earth history. This process is not that different from subduction itself (sinking of high-density, mafic crust into the mantle), but it has one order of magnitude less impact by mass on the mantle than subduction.
Although providing evidence for lower-crustal foundering, this work does not discount arc and intraplate magmatism as the main builders of continental crust. In fact, much of the intracrustal processing could happen within the lifetime of the magmatic arc, as magmas differentiate and residues founder. This work does demonstrate, however, that calculating the relative contribution of arc and intraplate magmas to the continents is not as simple as bulk mixing of average magma types (e.g. Rudnick, 1995; Barth et al., 2000) but, instead, requires consideration of the effects of both sediment recycling and fractionation & foundering on trace-element ratios. For example, even though average arc and intraplate magmas have distinct Th/La, a mixing calculation based on Th/La would be meaningless, as the ratio in arc magmas and continental crust is dominated by recycled sediment and intracrustal fractionation. Previous calculations based on the continent's Nb-anomaly will also need to be reconsidered, taking into account fractionations in the subduction zone as well as in the continental crust.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONTh/La VARIATIONS IN MARINE...CREATING HIGH Th/La IN...CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Supplementary data for this paper are available on Journal of Petrology online.
|
ACKNOWLEDGEMENTS |
|---|
This study grew from initial questions posed by Catherine Chauvel, during the author's visit to the University of Rennes. Special thanks to Bill White, Bob Kay, Mike Carr, Joe Devine and Simon Turner for providing samples for ICP–MS analysis; Michael Cheatham, Dan Gravatt, Nate Williams and Kato Dee for technical assistance; Roberta Rudnick and Bill McDonough for their xenolith databases; and Jeremy Boyce for some key papers. Thanks to Tim Elliott and Geoff Abers for help with envisioning processes and estimating uncertainties. I am grateful to Richard Arculus, S. Ross Taylor, Andy Kurtz, Ethan Baxter and Katie Kelley for very helpful reviews; and to Peter Kelemen for a particularly thoughtful and thorough review. This work was supported in part by NSF grants EAR-9506730 and OCE-9521717.
* E-mail: tplank{at}bu.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTh/La VARIATIONS IN MARINE...CREATING HIGH Th/La IN...CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Albarède, F. (1998). The growth of continental crust. Tectonophysics 296, 1–14.[CrossRef][Web of Science]
Anthony, E. Y., Segalstad, T. V. & Neumann, E. R. (1989). An unusual mantle source region for nephelinites from the Oslo Rift, Norway. Geochimica et Cosmochimica Acta 53, 1067–1076.[CrossRef][Web of Science]
Arculus, R. J. (1981). Island arc magmatism in relation to the evolution of the crust and mantle. Tectonophysics 75, 113–133.[CrossRef][Web of Science]
Arculus, R. J., Lapierre, H. & Jaillard, E. (1999). Geochemical window into subduction and accretion processes: Raspas metamorphic complex, Ecuador. Geology 27, 547–550.
Arndt, N. T. & Goldstein, S. L. (1989). An open boundary between lower continental crust and mantle: its role in crust formation and crustal recycling. Tectonophysics 161, 201–212.[CrossRef][Web of Science]
Bailey, J. C., Larsen, O. & Frolova, T. I. (1987). Strontium isotope variations in lower Tertiary–Quaternary volcanic rocks from the Kurile island arc. Contributions to Mineralogy and Petrology 95, 155–165.[CrossRef][Web of Science]
Baker, M. B., Hirschmann, M. M., Ghiorso, M. S. & Stolper, E. M. (1995). Compositions of near-solidus peridotite melts from experiments and thermodynamic calculation. Nature 375, 308–311.[CrossRef]
Barreiro, B. A. (1983). Lead isotopic compositions of South Sandwich Island volcanic rocks and their bearing on magmagenesis in intra-oceanic island arcs. Geochimica et Cosmochimica Acta 47, 817–822.[CrossRef][Web of Science]
Barth, M. G., McDonough, W. F. & Rudnick, R. L. (2000). Tracking the budget of Nb and Ta in the continental crust. Chemical Geology 165, 197–213.[CrossRef][Web of Science]
Batiza, R. & Vanko, D. (1984). Petrology of young Pacific seamounts. Journal of Geophysical Research 89, 11235–11260.[CrossRef]
Bea, F. & Montero, P. (1999). Behavior of accessory phases and redistribution of Zr, REE, Y, Th and U during metamorphism and partial melting of metapelites in the lower crust: an example from the Kinzingite Formation of Ivrea–Verbano, NW Italy. Geochimica et Cosmochimica Acta 63, 1133–1153.[CrossRef][Web of Science]
Bea, F., Pereira, M. D. & Stroh, A. (1994). Mineral/leucosome trace element partitioning in a peraluminous migmatite (a laser ablation ICP-MS study). Chemical Geology 117, 291–312.[CrossRef][Web of Science]
Becker, H., Jochum, K. P. & Carlson, R. W. (1999). Constraints from high-pressure veins in eclogites on the composition of hydrous fluids in subduction zones. Chemical Geology 160, 291–308.[CrossRef][Web of Science]
Becker, H. Jochum, K. P. & Carlson, R. W. (2000). Trace element fractionation during dehydration of eclogites from high-pressure terranes and the implications for element fluxes in subduction zones. Chemical Geology 163, 65–99.[CrossRef][Web of Science]
Ben Othman, D., White, W. M. & Patchett, J. (1989). The geochemistry of marine sediments, island arc magma genesis and crust–mantle recycling. Earth and Planetary Science Letters 94, 1–21.[CrossRef][Web of Science]
Bingen, B., Demaiffe, D. & Hertogen, J. (1996). Redistributions of REE, Th and U over accessory minerals in the course of amphibolite to granulite facies metamorphism: the role of apatite and monazite in orthogneisses from southwestern Norway. Geochimica et Cosmochimica Acta 60, 1341–1354.[CrossRef][Web of Science]
Bird, P. (1979). Continental delamination and the Colorado Plateau. Journal of Geophysical Research 84, 7561–7571.
Brenan, J. M., Shaw, H. F. & Ryerson, F. J. (1995). Experimental evidence for the origin of lead enrichment in convergent margin magmas. Nature 378, 54–56.[CrossRef]
Broska, I., Petrik, I. & Williams, C. T. (2000). Co-existing monazite and allanite in peraluminous granitoids of the Tribec Mountains, Western Carpathians. American Mineralogist 85, 22–32.
Carlson, R. W. (1984). Isotopic constraints on Columbia River flood basalt genesis and the nature of the subcontinental mantle. Geochimica et Cosmochimica Acta 48, 2357–2372.[CrossRef][Web of Science]
Carr, M. J. & Rose, W. I., Jr (1987). CENTAM: a data base of Central American volcanic rocks. Journal of Volcanology and Geothermal Research 33, 239–240.[CrossRef][Web of Science]
Carr, M. J., Feigenson, M. D. & Bennett, E. A. (1990). Incompatible element and isotopic evidence for tectonic control of source mixing and melt extraction along the Central-American arc. Contributions to Mineralogy and Petrology 105, 369–380.[CrossRef][Web of Science]
Chauvel, C., Goldstein, S. L. & Hofmann, A. W. (1995). Hydration and dehydration of oceanic crust controls Pb evolution in the mantle. Chemical Geology 126, 65–75.[CrossRef][Web of Science]
Cheatham, M. M., Sangrey, W. F. & White, W. M. (1993). Sources of error in external calibration ICP–MS analysis of geological samples and an improved nonlinear drift correction procedure. Spectrochimica Acta Part B—Atomic Spectroscopy 48, E487–E506.[CrossRef]
Class, C., Miller, D. L., Goldstein, S. L. & Langmuir, C. H. (2000). Distinguishing Melt and Fluid Components in Umnak Volcanics, Aleutian Arc. Geochemistry, Geophysics, Geosystems 1, 34 (paper number 1999GC000010).
Condie, K. C. (1993). Chemical composition and evolution of the upper continental crust: contrasting results from surface samples and shales. Chemical Geology 104, 1–37.[CrossRef][Web of Science]
Condie, K. C. (1998). Episodic continental growth and supercontinents: a mantle avalanche connection? Earth and Planetary Science Letters 163, 97–108.[CrossRef][Web of Science]
Condie, K. C. & Myers, J. S. (1999). Mesoproterozoic Fraser Complex: geochemical evidence for multiple subduction-related sources of lower crustal rocks in the Albany–Fraser Orogen, Western Australia. Australian Journal of Earth Sciences 46, 875–882.[CrossRef][Web of Science]
Cousens, B. L., Allan, J. F. & Gorton, M. P. (1994). Subduction-modified pelagic sediments as the enriched component in back-arc basalts from the Japan Sea: Ocean Drilling Program Sites 797 and 794. Contributions to Mineralogy and Petrology 117, 421–434.[CrossRef][Web of Science]
Davidson, J. P. (1987). Crustal contamination versus subduction zone enrichment: examples from the Lesser Antilles and implications for mantle source compositions of island arc volcanic rocks. Geochimica et Cosmochimica Acta 51, 2185–2198.[CrossRef][Web of Science]
Davidson, J. P. & Arculus, R. J. (2004). The significance of Phanerozoic arc magmatism in generating continental crust. In: Brown, M. & Rushmer, T. (eds) Evolution and Differentiation of the Continental Crust. Cambridge, MA: Cambridge University Press, in press.
Dimalanta, C. (2002). New rates of western Pacific island arc magmatism from seismic and gravity data. Earth and Planetary Science Letters 202, 105–115.[CrossRef][Web of Science]
Draut, A. E., Clift, P. D., Hannigan, R. E., Layne, G. & Shimizu, N. (2002). A model for continental crust genesis by arc accretion: rare earth element evidence from the Irish Caledonides. Earth and Planetary Science Letters 203, 861–877.[CrossRef][Web of Science]
Ducea, M. (2001). The California Arc: thick granitic batholiths, eclogitic residues, lithospheric-scale thrusting, and magmatic flare-ups. GSA Today 11, 4–10.
Ducea, M. N. (2002). Constraints on the bulk composition and root foundering rates of continental arcs: a California arc perspective. Journal of Geophysical Research 107, 2304, doi: 10.1029/2001JB000643.[CrossRef]
Ducea, M. N. & Saleeby, J. B. (1996). Buoyancy sources for a large, unrooted mountain range, the Sierra Nevada, California: evidence from xenolith thermobarometry. Journal of Geophysical Research B101, 8229–8244.
Ducea, M. N. & Saleeby, J. (1998). Crustal recycling beneath continental arcs: silica-rich glass inclusions in ultramafic xenoliths from the Sierra Nevada, California. Earth and Planetary Science Letters 156, 101–116.[CrossRef][Web of Science]
Dupuy, C., Dostal, J., Marcelot, G., Bougault, H., Joron, J. L. & Treuil, M. (1982). Geochemistry of basalts from central and southern New Hebrides arc: implications for their source rock composition. Earth and Planetary Science Letters 60, 207–225.[CrossRef][Web of Science]
Dupuy, C., Dostal, J. & Trineau, H. (1985). Geochemistry of volcanic rocks from Mt. Pelee, Martinique. Journal of Volcanology and Geothermal Research 26, 147–165.[CrossRef][Web of Science]
Edwards, C. M. H., Morris, J. D. & Thirlwall, M. F. (1993). Separating mantle from slab signatures in arc lavas using B/Be and radiogenic isotope systematics. Nature 362, 530–533.[CrossRef]
Eisele, J., Galer, S. J. G., Blichert-Toft, J., Devey, C. W. & Hofmann, A. W. (2002). The role of sediment recycling in EM-1 inferred from Os, Pb, Hf, Nd, Sr isotope and trace element systematics of the Pitcairn Hotspot. Earth and Planetary Science Letters 196, 197–212.[CrossRef][Web of Science]
Elliott, T., Plank, T., Zindler, A., White, W. & Bourdon, B. (1997). Element transport from subducted slab to volcanic front at the Mariana arc. Journal of Geophysical Research 102, 14991–15019.
Ewart, A. & Hawkesworth, C. J. (1987). The Pleistocene-Recent Tonga–Kermadec arc lavas: interpretation of new isotopic and rare earth data in terms of a depleted mantle source. Journal of Petrology 28, 495–530.
Ewart, A., Collerson, K. D., Regelous, M., Wendt, J. I. & Niu, Y. (1998). Geochemical evolution within the Tonga–Kermadec–Lau arc–back-arc systems: the role of varying mantle wedge composition in space and time. Journal of Petrology 39, 331–368.[CrossRef][Web of Science]
Fraser, K. J., Hawkesworth, C. J., Erlank, A. J., Mitchell, R. H. & Scott-Smith, B. H. (1985/6). Sr, Nd and Pb isotope and minor element geochemistry of lamproites and kimberlites. Earth and Planetary Science Letters 76, 57–70.
Frey, F. A., Garcia, M. O. & Roden, M. F. (1994). Geochemical characteristics of Koolau Volcano; implications of intershield geochemical differences among Hawaiian volcanoes. Geochimica et Cosmochimica Acta 58, 1441–1462.[CrossRef][Web of Science]
Gao, S., Luo, T.-C., Zhang, B.-R., Zhang, H.-F., Han, Y.-W., Zhao, Z.-D. & Hu, Y.-K. (1998). Chemical composition of the continental crust as revealed by studies in east China. Geochimica et Cosmochimica Acta 62, 1959–1975.[CrossRef][Web of Science]
George, R., Turner, S., Hawkesworth, C., Morris, J., Nye, C., Ryan, J. & Zheng, S.-H. (2003). Melting processes and fluid and sediment transport rates along the Alaska–Aleutian arc from an integrated U–Th–Ra–Be isotope study. Journal of Geophysical Research 108, 2252, doi: 10.1029/2002JB001916.[CrossRef]
GERM (Geochemical Earth Reference Model). http://www.earthref.org/GERM/main.htm.
Glazner, A. F. (1994). Foundering of mafic plutons and density stratification of continental crust. Geology 22, 435–438.
Gorton, M. (1977). The geochemistry and origin of Quaternary volcanism in the New Hebrides. Geochimica et Cosmochimica Acta 41, 1257–1270.[CrossRef][Web of Science]
Govindaraju, K. (1994). 1994 compilation of working values and sample description for 383 Geostandards. Geostandards Newsletter 18, 158 pp.
Gravatt, D. R. (1997). Magma genesis and evolution processes on the islands of Grenada and St. Lucia, Lesser Antilles. M.S. thesis, Cornell University, Ithaca, NY, 202 pp.
Green, T. H. (1994). Experimental studies of trace-element partitioning applicable to igneous petrogenesis: Sedona 16 years later. Chemical Geology 117, 1–36.[CrossRef][Web of Science]
Griffin, W. L., O'Reilly, S. Y. & Stabel, A. (1988). Mantle metasomatism beneath western Victoria, Australia; II, Isotopic geochemistry of Cr-diopside lherzolites and Al-augite pyroxenites. Geochimica et Cosmochimica Acta 52, 449–459.[CrossRef][Web of Science]
Gust, D. A., Arculus, R. J. & Kersting, A. B. (1995). Aspects of magma sources and processes in the Honshu Arc. Canadian Mineralogist 35, 347–365.
Halliday, A. N., Dickin, A. P., Hunter, R. N., Davies, G. R., Dempster, T. J., Hamilton, P. J. & Upton, B. G. J. (1993). Formation and composition of the lower continental crust: evidence from Scottish xenolith suites. Journal of Geophysical Research 98, 581–607.
Hanchar, J. M., Miller, C. F., Wooden, J. L., Bennett, V. C. & Staude, J.-M. G. (1994). Evidence from xenoliths for a dynamic lower crust, Eastern Mojave desert, California. Journal of Petrology 35, 1377–1415.
Hansen, E. C., Newton, R. C., Jamardhar, A. S. & Lindenberg, S. (1995). Differentiation of Late-Archaean crust in the eastern Dharwar craton, Krishnagiri–salem area, South India. Journal of Geology 103, 629–651.[Web of Science]
Hawkesworth, C. J., Turner, S. P., McDermott, F., Peate, D. W. & van Calsteren, P. (1997). U–Th isotopes in arc magmas: implications for element transfer from the subducted crust. Science 276, 551–555.
Hermann, J. (2002). Allanite: thorium and light rare earth element carrier in subducted crust. Chemical Geology 192, 289–306.[CrossRef][Web of Science]
Herzberg, C. T., Fyfe, W. S. & Carr, M. J. (1983). Density constraints on the formation of the continental Moho and crust. Contributions to Mineralogy and Petrology 84, 1–5.[CrossRef][Web of Science]
Hilton, D. R. & Craig, H. (1989). A helium isotope transect along the Indonesian archipelago. Nature 342, 906–908.[CrossRef]
Hochstaedter, A. G., Kepezhinskas, P., Defant, M., Drummond, M. & Koloskov, A. (1996). Insights into the volcanic arc mantle wedge from magnesian lavas from the Kamchatka Arc. Journal of Geophysical Research 101, 697–712.[CrossRef]
Hofmann, A. W. (1988). Chemical differentiation of the Earth: the relationship between mantle, continental crust and oceanic crust. Earth and Planetary Science Letters 90, 297–314.[CrossRef][Web of Science]
Holbrook, S. W., Lizarralde, D., McGeary, S., Bangs, N. & Diebold, J. (1999). Structure and composition of the Aleutian island arc and implications for continental crustal growth. Geology 27, 31–34.
Hoogewerff, J. A., van Bergen, M. J., Vroon, P. Z., Hertogen, J., Wordel, R., Sneyers, A., Nasution, A., Varekamp, J. C., Moens, H. L. E. & Mouchel, D. (1997). U-series, Sr–Nd–Pb isotope and trace-element systematics across an active island-arc continent collision zone: implications for element transfer at the slab–wedge interface. Geochimica et Cosmochimica Acta 61, 1057–1072.[CrossRef][Web of Science]
Houseman, G. A., McKenzie, D. P. & Molnar, P. (1981). Convective instability of a thickened boundary layer and its relevance for the thermal evolution of continental convergent belts. Journal of Geophysical Research 86, 6135–6155.
Jahn, B.-M., Gallet, S. & Han, J. (2001). Geochemistry of the Xining, Xifeng and Jixian sections, Loess Plateau of China: eolian dust provenance and paleosol evolution during the last 140 ka. Chemical Geology 178, 71–94.[CrossRef][Web of Science]
Johnson, M. C. & Plank, T. (1999). Dehydration and melting experiments constrain the fate of subducted sediments. Geochemistry, Geophysics &. Geosystems 1, paper number 1999GC000014.
Jull, M. & Kelemen, P. B. (2001). On the conditions for lower crustal convective instability. Journal of Geophysical Research 106, 6423–6446.[CrossRef]
Kay, R. W. (1980). Volcanic arc magmas: implication of a melting–mixing model for element recycling in the crust–upper mantle system. Journal of Geology 88, 497–522.[Web of Science]
Kay, R. W. & Kay, S. M. (1988). Crustal recycling and the Aleutian arc. Geochimica et Cosmochimica Acta 52, 1351–1359.[CrossRef][Web of Science]
Kay, R. W. & Kay, S. M. (1991). Creation and destruction of lower continental crust. Geologische Rundschau 80, 259–278.
Kay, R. W. & Kay, S. M. (1993). Delamination and delamination magmatism. Tectonophysics 219, 177–189.[CrossRef][Web of Science]
Kay, S. M. & Kay, R. W. (1994). Aleutian magmas in space and time. In: Plafker, G. & Berg, H. C. (eds) The Geology of Alaska, The Geology of North America, Vol. G-1. Boulder, CO: Geological Society of America, pp. 687–722.
Kelemen, P. B. (1990). Reaction between ultramafic wall rock and fractionating basaltic magma: Part I. Phase relations, the origin of calc-alkaline magma series and the formation of discordant dunite. Journal of Petrology 31, 51–98.
Kelemen, P. B. (1995). Genesis of high Mg# andesites and the continental crust. Contributions to Mineralogy and Petrology 120, 1–19.[Web of Science]
Kelemen, P. B., Shimizu, N. & Dunn, T. (1993). Relative depletion of Nb in some arc magmas and the continental crust: partitioning of K, Nb, La and Ce during melt/rock reaction in the upper mantle. Earth and Planetary Science Letters 120, 111–133.[CrossRef][Web of Science]
Kelemen, P. B., Hanghoj, K. & Greene, A. R. (2003a). One view of the geochemistry of subduction-related magmatic arcs, with an emphasis on primitive andesite and lower crust. In: Rudnick, R. L. (ed.) The Crust, Vol. 3. In: Holland, H. D. & Turekian, K. K. (eds) Treatise on Geochemistry. Oxford: Elsevier–Pergamon, pp. 593–659.
Kelemen, P. B., Yogodzinski, G. M. & Scholl, D. W. (2003b). Along-strike variation in the aleutian island arc: genesis of high-Mg# andesite and implications for continental crust. In: Eiler, J. (ed.) Inside the Subduction Factory. Geophysical Monograph, American Geophysical Union 138, 223–276.
Kelley, K. A., Plank, T., Ludden, J. N. & Staudigel, H. (2003). The composition of altered oceanic crust at ODP sites 801 and 1149. Geochemistry, Geophysics, Geosystems 4, doi: 10.1029/2002GC000435.
Klewin, K. W. & Berg, J. H. (1991). Petrology of the Keweenawan Mamainse Point lavas, Ontario: petrogenesis and continental rift evolution. Journal of Geophysical Research 96, 457–474.
Kogiso, T., Tatsumi, Y. & Nakano, S. (1997). Trace element transport during dehydration processes in the subducted oceanic crust 1: experiments and implications for the origin of ocean island basalts. Earth and Planetary Science Letters 148, 193–205.[CrossRef][Web of Science]
Lee, C.-Y., Chung, S.-L., Chen, C.-H. & Hsieg, Y.-L. (1993). Mafic granulite xenoliths from Penghu islands: evidence for basic lower crust in SE China continental margin. Journal of the Geological Society of China 36, 351–379.
Lee, D. E. & Silver, L. T. (1964). Accessory minerals in some granitic rocks in California and Nevada as a function of calcium content. American Mineralogist 49, 1660–1670.[Web of Science]
Leyreloup, A., Bodinier, J. L., Dupuy, C. & Dostal, J. (1982). Petrology and geochemistry of granulite xenoliths from central Hoggar (Algeria): implications for the lower crust. Contributions to Mineralogy and Petrology 79, 68–75.[CrossRef][Web of Science]
Lundstrom, C. C., Shaw, H. F., Ryerson, F. J., Williams, Q. & Gill, J. (1998). Crystal chemical control of clinopyroxene–melt partitioning in the Di–Ab–An system: implications for elemental fractionations in the depleted mantle. Geochimica et Cosmochimica Acta 62, 2849–2862.[CrossRef][Web of Science]
Mahood, G. & Hildreth, W. (1983). Large partition coefficients for trace elements in high-silica rhyolites. Geochimica et Cosmochimica Acta 47, 11–30.[CrossRef][Web of Science]
Mahood, G. & Stimac, J. A. (1990). Trace element partitioning in pantellerites and trachytes. Geochimica et Cosmochimica Acta 54, 2257–2276.[CrossRef][Web of Science]
Martin, H. (1986). Effect of steeper Archaean geothermal gradient on geochemistry of subduction-zone magmas. Geology 14, 753–756.
Mattie, P. D., Condie, K. C., Selverstone, J. & Kyle, P. R. (1997). Origin of the continental crust in the Colorado Plateau: geochemical evidence from mafic xenoliths from the Navajo Volcanic Field, southwestern USA. Geochimica et Cosmochimica Acta 61, 2007–2021.[CrossRef][Web of Science]
McLennan, S. M. (2000). Relationships between the trace element composition of sedimentary rocks and upper continental crust. Geochemistry, Geophysics, Geosystems 2, 2000GC000109.
McLennan, S. M. & Taylor, S. R. (1996). Heat flow and the chemical composition of continental crust. Journal of Geology 104, 369–377.[Web of Science]
Miller, D. M., Goldstein, S. L. & Langmuir, C. H. (1994). Cerium–lead and lead-isotope ratios in arc magmas and the enrichment of lead in the continents. Nature 368, 514–520.[CrossRef]
Montel, J.-M. (1993). A model for monazite/melt equilibrium and application to the generation of granitic magmas. Chemical Geology 110, 127–146.[CrossRef][Web of Science]
Morris, J. D. & Hart, S. R. (1983). Isotopic and incompatible element constraints on the genesis of island arc volcanics from Cold Bay and Amak island, Aleutians, and implications for mantle structure. Geochimica et Cosmochimica Acta 47, 2015–2030.[CrossRef][Web of Science]
Morris, J., Valentine, R. & Harrison, T. (2002). 10Be imaging of sediment accretion and subduction along the northeast Japan and Costa Rica convergent margins. Geology 30, 59–62.
Nelson, K. D. (1992). Are crustal thickness variations in old mountain belts like the Appalachians a consequence of lithospheric delamination? Geology 20, 498–502.
Nelson, D. R., Chivas, A. R., Chappell, B. W. & McCulloch, M. T. (1988). Geochemical and isotopic systematics in carbonatites and implications for the evolution of ocean-island sources. Geochimica et Cosmochimica Acta 52, 1–17.[CrossRef][Web of Science]
Nichols, G. T., Wyllie, P. J. and Stern, C. R. (1996). Experimental melting of pelagic sediment, constraints relevant to subduction. In: Bebout, G. E., Shock, D. W., Kirby, S. H. & Platt, J. P. (eds) Subduction: Top to Bottom. Washington, DC: American Geophysical Union, pp. 293–298.
Niu, Y. & Batiza, R. (1997). Trace element evidence from seamounts for recycled oceanic crust in the eastern Pacific mantle. Earth and Planetary Science Letters 148, 471–483.[CrossRef][Web of Science]
Noll, P. D., Newsom, H. E., Leeman, W. P. & Ryan, J. G. (1996). The role of hydrothermal fluids in the production of subduction zone magmas: evidence from siderophile and chalcophile trace elements and boron. Geochimica et Cosmochimica Acta 60, 587–611.[CrossRef][Web of Science]
O'Nions, R. K. & McKenzie, D. P. (1988). Melting and continent generation. Earth and Planetary Science Letters 90, 449–456.[CrossRef][Web of Science]
Pearce, J. A. & Parkinson, I. J. (1993). Trace element models for mantle melting: application to volcanic arc petrogenesis. In: Prichard, H. M., Alabaster, T., Harris, N. B. W. & Neary, C. R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications 76, 373–403.
Pearce, J. A., Baker, P. E., Harvey, P. K. & Luff, I. W. (1995). Geochemical evidence for subduction fluxes, mantle melting and fractional crystallization beneath the South Sandwich island arc. Journal of Petrology 36, 1073–1109.
Pearce, J. A., Kempton, P. D., Nowell, G. M. & Noble, S. R. (1999). Hf–Nd element and isotope perspective on the nature and provenance of mantle and subduction components in western Pacific arc–basin systems. Journal of Petrology 11, 1579–1611.
Pearcy, L. G., DeBari, S. & Sleep, N. H. (1990). Mass balance calculations for two sections of island arc crust and implications for the formation of continents. Earth and Planetary Science Letters 96, 427–442.[CrossRef][Web of Science]
Peate, D. W., Pearce, J. A., Hawkesworth, C. J., Colley, H., Edwards, C. M. H. & Hirose, K. (1997). Geochemical variations in Vanuatu arc lavas: the role of subducted material and a variable mantle wedge composition. Journal of Petrology 38, 1331–1358.[CrossRef][Web of Science]
PetDB (Petrological Database of the Ocean Floor), http://petdb.ldeo.columbia.edu/petdb/.
Plank, T. & Langmuir, C. H. (1993). Tracing trace elements from sediment input to volcanic output at subduction zones. Nature 362, 739–743.[CrossRef]
Plank, T. & Langmuir, C. H. (1998). The chemical composition of subducting sediment: implications for the crust and mantle. Chemical Geology 145, 325–394.[CrossRef][Web of Science]
Plank, T. & Ludden, J. N. (1992). Geochemistry of sediment in the Argo Abyssal Plain at Site 765: a continental margin reference section for sediment recycling in subduction zones. In: Gradstein, F. M., Ludden, J. N., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 123. College Station, TX: Ocean Drilling Program, pp. 167–189.
Rapp, R. P. & Watson, E. B. (1986). Monazite solubility and dissolution kinetics: implications for the thorium and light rare earth element chemistry of felsic magmas. Contributions to Mineralogy and Petrology 94, 304–316.[CrossRef][Web of Science]
Reagan, M. K. & Gill, J. B. (1986). Coexisting calcalkaline and high-niobium basalts from Turrialba Volcano, Costa Rica: implications for residual titanates in arc magma sources. Journal of Geophysical Research 94, 4619–4633.[CrossRef]
Reymer, A. & Schubert, G. (1984). Phanerozoic addition rates to the continental crust and crustal growth. Tectonics 3, 63–77.[Web of Science]
Roberts, S. J. & Ruiz, J. (1989). Geochemistry of exposed granulite facies terrains and lower crustal xenoliths in Mexico. Journal of Geophysical Research 94, 7961–7974.
Rogers, N. W., Hawkesworth, C. J., Parker, C. J. & Marsh, J. S. (1985). The geochemistry of potassic lavas from Vulsini, central Italy and implications for mantle enrichment processes beneath the Roman region. Contributions to Mineralogy and Petrology 90, 244–257.[CrossRef][Web of Science]
Rogers, N. W., Hawkesworth, C. J., Mattey, D. P. & Harmon, R. S. (1987). Sediment subduction and the source of potassium in orogenic leucitites. Geology 15, 451–453.
Rosenblum, S. & Fleischer, M. (1995). The distribution of rare-earth elements in minerals of the monazite family. US Geological Survey Bulletin 2140, 62 pp.
Rudnick, R. L. (1995). Making continental crust. Nature 378, 571–578.[CrossRef]
Rudnick, R. L. & Fountain, D. M. (1995). Nature and composition of the continental crust: a lower crustal perspective. Reviews of Geophysics 33, 267–309.[CrossRef][Web of Science]
Rudnick, R. L. & Goldstein, S. L. (1990). The Pb isotopic compositions of lower crustal xenoliths and the evolution of lower crustal Pb. Earth and Planetary Science Letters 98, 192–207.[CrossRef][Web of Science]
Rudnick, R. L. & Presper, T. (1990). Geochemistry of intermediate- to high-pressure granulites. In: Vielszeuf, D. & Vidal, P. (eds) Granulites and Crustal Evolution. Norwell, MA: Kluwer, pp. 523–550.
Rudnick, R. L. & Taylor, S. R. (1987). The composition and petrogenesis of the lower crust: a xenolith study. Journal of Geophysical Research 92, 13 981–14 005.[CrossRef]
Rudnick, R. L., McDonough, W. F., McCulloch, M. T. & Taylor, S. R. (1986). Lower crustal xenoliths from Queensland, Australia: evidence for deep crustal assimilation and fractionation of continental basalts. Geochimica et Cosmochimica Acta 50, 1099–1115.[CrossRef][Web of Science]
Rudnick, R. L., McDonough, W. F. & O'Connell, R. J. (1998). Thermal structure, thickness and composition of continental lithosphere. Chemical Geology 145, 395–411.[CrossRef][Web of Science]
Ryan, J. G. (1995). Cross-arc geochemical variations in the Kurile Arc as a function of slab depth. Science 270, 625–627.
Ryerson, F. J. & Watson, E. B. (1987). Rutile saturation in magmas: implications for Ti–Nb–Ta depletion in island arc magmas. Earth and Planetary Science Letters 86, 225–239.[CrossRef][Web of Science]
Saal, A. E., Rudnick, R. L., Ravizza, G. E. & Hart, S. R. (1998). Re–Os isotope evidence for the composition, formation and age of the lower continental crust. Nature 393, 58–61.[CrossRef]
Salters, V. J. M & Stracke, A. (2004). Composition of the depleted mantle. Geochemistry, Geophysics, Geosystems, 5, doi: 10.1029/2003GC000597.
Samson, S. D., Coler, D. G. & Speer, J. A. (1995). Geochemical and Nd–Sr–Pb isotopic composition of Alleghenian granites of the Southern Appalachians: origin, tectonic setting, and source characterization. Earth and Planetary Science Letters 134, 359–376.[CrossRef][Web of Science]
Sawka, W. N. & Chappel, B. W. (1988). Fractionation of U, Th and REE in a vertically zoned granodiorite: implications for heat production distributions in the Sierra Nevada batholith, California, USA. Geochimica et Cosmochimica Acta 52, 1131–1143.[CrossRef][Web of Science]
Schulze, D. J. (1989). Constraints on the abundance of eclogite in the upper mantle. Journal of Geophysical Research 94, 4205–4212.
Spandler, C., Hermann, J., Arculus, R. & Mavrogenes, J. (2003). Redistribution of trace elements during prograde metamorphism from lawsonite blueschist to eclogite facies: implications for deep subduction-zone processes. Contributions to Mineralogy and Petrology 146, 205–222.[CrossRef][Web of Science]
Staudigel, H., Plank, T., White, W. M. & Schmincke, H. (1996). Geochemical fluxes during seafloor alteration of the upper oceanic crust: DSDP Sites 417 and 418. In: Bebout, G. E. & Kirby, S. H. (eds) SUBCON: Subduction from Top to Bottom. Geophysical Monograph, American Geophysical Union 96, 19–38.
Stern, R. J. & Ito, E. (1983). Trace-element and isotopic constraints on the source of magmas in the active Volcano and Mariana island arcs, western Pacific. Journal of Volcanology and Geothermal Research 18, 461–482.[CrossRef][Web of Science]
Stolper, E. & Newman, S. (1994). The role of water in the petrogenesis of Mariana Trough magmas. Earth and Planetary Science Letters 121, 293–325.[CrossRef][Web of Science]
Stolz, A. J., Varne, R., Davies, G. R., Wheller, G. E. & Foden, J. D. (1990). Magma source composition in an arc–continent collision zone: the Flores–Lembata sector, Sunda arc, Indonesia. Contributions to Mineralogy and Petrology 105, 585–601.[CrossRef][Web of Science]
Sun, S. & McDonough, W. F. (1989). Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A. D. & Norrey, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications 42, 313–345.
Tatsumi, Y. (2000). Slab melting: its role in continental crust formation and mantle evolution. Geophysical Research Letters 27, 3941–3944.[CrossRef][Web of Science]
Tatsumi, Y., Sakuyama, M., Fukuyama, H. & Kushiro, I. (1983). Generation of arc basalt magmas and thermal structure of the mantle wedge in subduction zones. Journal of Geophysical Research 88, 5815–5825.
Taylor, S. R. (1967). The origin and growth of continents. Tectonophysics 4, 17.[CrossRef][Web of Science]
Taylor, S. R. & McLennan, S. M. (1985). The Continental Crust: its Composition and Evolution. Oxford: Blackwell Scientific 312 pp.
Tera, F., Brown, L., Morris, J., Sacks, I. S., Klein, J. & Middleton, R. (1986). Sediment incorporation in island-arc magmas: inferences from 10Be. Geochimica et Cosmochimica Acta 50, 535–550.[CrossRef][Web of Science]
Thirlwall, M. F. & Graham, A. M. (1984). Evolution of high-Ca, high-Sr C-series basalts from Grenada, Lesser Antilles: the effects of intra-crustal contamination. Journal of the Geological Society, London 141, 427–445.
Thirlwall, M. F., Graham, A. M., Arculus, R. J., Harmon, R. S. & Macpherson, C. G. (1996). Resolution of the effects of crustal assimilation, sediment subduction, and fluid transport in island arc magmas: Pb–Sr–Nd–O isotopes geochemistry of Grenada, Lesser Antilles. Geochimica et Cosmochimica Acta 60, 4785–4810.[CrossRef][Web of Science]
Till, R. (1974). Statistical Methods for Earth Scientists: An Introduction. New York: John Wiley, pp. 83–103.
Turcotte, D. L. (1989). Geophysical processes influencing the lower continental crust. In: Mereu, R. F., Mueller, S. & Fountain, D. M. (eds) Properties and Processes of Earth's Lower Crust. Geophysical Monograph, American Geophysical Union, 321–329.
Turner, S. T., Hawkesworth, C., van Calsteren, P., Heath, E., Macdonald, R. & Black, S. (1996) U-series isotopes and destructive margin magma genesis in the Lesser Antilles. Earth and Planetary Science Letters 142, 191–207.[CrossRef][Web of Science]
Turner, S., Hawkesworth, C. J., Rogers, N., Bartlett, J., Worthington, T., Hergt, J., Pearce, J. & Smith, I. (1997). U-238–Th-230 disequilibria, magma petrogenesis, and flux rates beneath the depleted Tonga–Kermadec island arc. Geochimica et Cosmochimica Acta 61, 4855–4884.[CrossRef][Web of Science]
Varne, R. & Foden, J. D. (1986). Geochemical and isotopic systematics of eastern sunda arc volcanics: implications for mantle sources and mantle mixing processes. In: Wezel, F. C. (ed.) Origin of Arcs. Amsterdam: Elsevier, pp. 159–189.
Vervoort, J. D., Patchett, P. J., Albarède, F., Blichert-Toft, J., Rudnick, R. & Downes, H. (2000). Hf–Nd isotopic evolution of the lower crust. Earth and Planetary Science Letters 181, 115–129.[CrossRef][Web of Science]
von Huene, R. & Scholl, D. W. (1991). Observations at convergent margins concerning sediment subduction, subduction erosion, and the growth of continental crust. Reviews of Geophysics 29, 279–316.[Web of Science]
Vroon, P. Z. (1992). Subduction of continental material in the Banda Arc, Eastern Indonesia: Sr–Nd–Pb isotope and trace-element evidence from volcanics and sediment. Ph.D. dissertation, University of Utrecht.
Vroon, P. Z., van Bergen, M. J., Klaver, G. J. & White, W. M. (1995). Strontium, neodymium, and lead isotopic and trace-element signatures of the East Indonesian sediments: provenance and implications for Banda Arc magma genesis. Geochimica et Cosmochimica Acta 59, 2573–2598.[CrossRef][Web of Science]
Wade, J. A., Plank, T., Stern, R. J., Tollstrup, D. L., Gill, J. B., O'Leary, J. C., et al. (2005). The May 2003 eruption of Anatahan volcano, Mariana Islands: geochemical evolution of a silicic island arc volcano. Journal of Volcanology and Geothermal Research, in press.
Wark, D. A. & Miller, C. F. (1993). Accessory mineral behavior during differentiation of a granite suite: monazite, xenotime and zircon in the Sweetwater Wash pluton, southeastern California, USA. Chemical Geology 110, 49–67.[CrossRef][Web of Science]
Watson, E. B. (1980). Apatite and phosphorus in mantle source regions: an experimental study of apatite/melt equilibria at pressures to 25 kbar. Earth and Planetary Science Letters 51, 322–335.[CrossRef][Web of Science]
Weaver, B. L. (1991). The origin of ocean island basalt end-member compositions: trace element and isotopic constraints. Earth and Planetary Science Letters 104, 381–397.[CrossRef][Web of Science]
Wedepohl, K. H. (1995). The composition of the continental crust. Geochimica et Cosmochimica Acta 59, 1217–1232.[CrossRef][Web of Science]
Wendlandt, E., DePaolo, D. J. & Baldridge, W. S. (1993). Nd and Sr isotope chronostratigraphy of Colorado Plateau lithosphere: implications for magmatic and tectonic underplating of the continental crust. Earth and Planetary Science Letters 116, 23–43.[CrossRef][Web of Science]
White, W. M. & Duncan, R. A. (1996). Geochemistry and geochronology of the Society islands: new evidence for deep mantle recycling. In: Basu, A. & Hart, S. (eds) Earth Processes: Reading the Isotopic Code. Geophysical Monograph, American Geophysical Union 95, 183–206.
White, W. M. & Dupre, B. (1986). Sediment subduction and magma genesis in the Lesser Antilles: isotopic and trace element constraints. Journal of Geophysical Research 91, 5927–5941.
White, W. M. & Patchett, J. (1984). Hf–Nd–Sr isotopes and incompatible element abundances in island arcs: implications for magma origins and crust–mantle evolution. Earth and Planetary Science Letters 67, 167–185.[CrossRef][Web of Science]
White, W. M., Dupre, B. & Vidal, P. (1985). Isotope and trace element geochemistry of sediments from the Barbados Ridge–Demerara Plain region, Atlantic Ocean. Geochimica et Cosmochimica Acta 49, 1875–1886.[CrossRef][Web of Science]
Whitford, D. J. (1975). Geochemistry and petrology of volcanic rocks from the Sunda arc. Indonesia. Ph.D. thesis. Australian National University Canberra, 449 pp.
Wilkinson, J. F. G. & Taylor, S. R. (1980). Trace element fractionation trends of tholeiitic magma at moderate pressure: evidence from an Al-spinel ultramafic inclusion suite. Contributions to Mineralogy and Petrology 75, 225–233.[CrossRef]
Wilson, M. (1989). Igneous Petrogenesis. London: Chapman and Hall, 466 pp.
Winter, J. D. (2001). An Introduction to Igneous and Metamorphic Petrology. Upper Saddle River, NJ: Prentice Hall, p. 347.
Wooden, J. L., Czamanske, G. K., Fedorenko, V. A., Arndt, N. T., Chauvel, C., Bouse, R., King, B.-S., Knight, R. J. & Siems, D. F. (1993). Isotopic and trace-element constraints on mantle and crustal contributions to Siberian continental flood basalts, Noril'sk area, Siberia. Geochimica et Cosmochimica Acta 57, 3577–3704.
Woodhead, J., Eggins, S. & Gamble, J. (1993). High field strength and transition element systematics in island arc and back-arc basin basalts: evidence for multi-phase melt extraction and a depleted mantle wedge. Earth and Planetary Science Letters 114, 491–504.[CrossRef][Web of Science]
Zhou, L. & Kyte, F. T. (1992). Sedimentation history of the South Pacific pelagic clay province over the last 85 million years inferred from the geochemistry of DSDP Hole 596. Paleoceanography 7, 441–465.
Zindler, A. & Hart, S. (1986). Chemical geodynamics. Annual Review of Earth and Planetary Sciences 14, 493–571.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Georgiev, P. Marchev, C. A. Heinrich, A. Von Quadt, I. Peytcheva, and P. Manetti Origin of Nepheline-normative High-K Ankaramites and the Evolution of Eastern Srednogorie Arc in SE Europe J. Petrology, October 1, 2009; 50(10): 1899 - 1933. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. R. Johnson, P. J. Wallace, H. Delgado Granados, V. C. Manea, A. J. R. Kent, I. N. Bindeman, and C. S. Donegan Subduction-related Volatile Recycling and Magma Generation beneath Central Mexico: Insights from Melt Inclusions, Oxygen Isotopes and Geodynamic Models J. Petrology, September 1, 2009; 50(9): 1729 - 1764. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Boari, S. Tommasini, M. A. Laurenzi, and S. Conticelli Transition from Ultrapotassic Kamafugitic to Sub-alkaline Magmas: Sr, Nd, and Pb Isotope, Trace Element and 40Ar-39Ar Age Data from the Middle Latin Valley Volcanic Field, Roman Magmatic Province, Central Italy J. Petrology, July 1, 2009; 50(7): 1327 - 1357. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. Nutman, V. C. Bennett, C. R. L. Friend, F. Jenner, Y. Wan, and D. Liu Eoarchaean crustal growth in West Greenland (Itsaq Gneiss Complex) and in northeastern China (Anshan area): review and synthesis Geological Society, London, Special Publications, January 1, 2009; 318(1): 127 - 154. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Tiepolo and R. Tribuzio Petrology and U-Pb Zircon Geochronology of Amphibole-rich Cumulates with Sanukitic Affinity from Husky Ridge (Northern Victoria Land, Antarctica): Insights into the Role of Amphibole in the Petrogenesis of Subduction-related Magmas J. Petrology, May 1, 2008; 49(5): 937 - 970. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-T. A. Lee, D. M. Morton, M. G. Little, R. Kistler, U. N. Horodyskyj, W. P. Leeman, and A. Agranier Regulating continent growth and composition by chemical weathering PNAS, April 1, 2008; 105(13): 4981 - 4986. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Klimm, J. D. Blundy, and T. H. Green Trace Element Partitioning and Accessory Phase Saturation during H2O-Saturated Melting of Basalt with Implications for Subduction Zone Chemical Fluxes J. Petrology, March 1, 2008; 49(3): 523 - 553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Avanzinelli, T. Elliott, S. Tommasini, and S. Conticelli Constraints on the Genesis of Potassium-rich Italian Volcanic Rocks from U/Th Disequilibrium J. Petrology, February 1, 2008; 49(2): 195 - 223. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Bonadiman, M. Coltorti, S. Duggen, L. Paludetti, F. Siena, M. F. Thirlwall, and B. G. J. Upton Palaeozoic subduction-related and kimberlite or carbonatite metasomatism in the Scottish lithospheric mantle Geological Society, London, Special Publications, January 1, 2008; 293(1): 303 - 333. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. N. Ducea and M. D. Barton Igniting flare-up events in Cordilleran arcs Geology, November 1, 2007; 35(11): 1047 - 1050. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Marchesi, C. J. Garrido, D. Bosch, J. A. Proenza, F. Gervilla, P. Monie, and A. Rodriguez-Vega Geochemistry of Cretaceous Magmatism in Eastern Cuba: Recycling of North American Continental Sediments and Implications for Subduction Polarity in the Greater Antilles Paleo-arc J. Petrology, September 1, 2007; 48(9): 1813 - 1840. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. Richards and R. Kerrich Special Paper: Adakite-Like Rocks: Their Diverse Origins and Questionable Role in Metallogenesis Economic Geology, June 1, 2007; 102(4): 537 - 576. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Gao, Z. Hou, B. S. Kamber, R. Wei, X. Meng, and R. Zhao Lamproitic Rocks from a Continental Collision Zone: Evidence for Recycling of Subducted Tethyan Oceanic Sediments in the Mantle Beneath Southern Tibet J. Petrology, April 1, 2007; 48(4): 729 - 752. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Anderson The eclogite engine: Chemical geodynamics as a Galileo thermometer Geological Society of America Special Papers, January 1, 2007; 430(0): 47 - 64. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Toothill, C. A. Williams, R. MacDonald, S. P. Turner, N. W. Rogers, C. J. Hawkesworth, D. A. Jerram, C. J. Ottley, and A. G. Tindle A Complex Petrogenesis for an Arc Magmatic Suite, St Kitts, Lesser Antilles J. Petrology, January 1, 2007; 48(1): 3 - 42. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Guo, W.-M. Fan, X.-Y. Li, and C.-W. Li Geochemistry of Mesozoic mafic volcanic rocks from the Yanshan belt in the northern margin of the North China Block: relations with post-collisional lithospheric extension Geological Society, London, Special Publications, January 1, 2007; 280(1): 101 - 129. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-I. KIMURA and T. YOSHIDA Contributions of Slab Fluid, Mantle Wedge and Crust to the Origin of Quaternary Lavas in the NE Japan Arc J. Petrology, November 1, 2006; 47(11): 2185 - 2232. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Tollstrup and J. B. Gill Hafnium systematics of the Mariana arc: Evidence for sediment melt and residual phases Geology, September 1, 2005; 33(9): 737 - 740. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||