Journal of Petrology

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (29) Request Permissions Google Scholar Articles by MÜNKER, C. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Journal of Petrology Volume 41 Number 6 Pages 759-788 2000
© Oxford University Press 2000

The Isotope and Trace Element Budget of the Cambrian Devil River Arc System, New Zealand: Identification of Four Source Components

CARSTEN MÜNKER^1,2,*

¹INSTITUT FÜR GEOLOGIE UND DYNAMIK DER LITHOSPHÄRE AND GEOCHEMISCHES INSTITUT, GOLDSCHMIDTSTR. 3, 37077 GÖTTINGEN, GERMANY
²MAX-PLANCK-INSTITUT FÜR CHEMIE, ABTEILUNG GEOCHEMIE, POSTFACH 3060, 55020 MAINZ, GERMANY

Received December 18, 1998; Revised typescript accepted November 11, 1999

	ABSTRACT

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
The Takaka Terrane in the South Island of New Zealand contains a well-preserved Cambrian arc system (Devil River Volcanics) that displays a complete assemblage of interbedded low- to high-K arc rocks, back-arc rocks and boninites. Most volcanic rocks are mafic. A coherent dataset was obtained including major elements, trace elements and Sr–Nd–Pb isotope compositions from clinopyroxene and amphibole separates. With time, ²⁰⁷Pb/²⁰⁴Pb in the arc rocks become more unradiogenic and ¹⁴³Nd/¹⁴⁴Nd more radiogenic, and Th/Yb and La/Yb increase. La/Yb values range from one in the boninites and back-arc rocks to 30 in the high-K arc rocks. Corresponding Nd values range from -5 to +6. High field strength element systematics rules out an OIB source component in the mantle wedge. The isotope variations in the volcanic rocks indicate addition of subducted sediments derived from Archaean sources to the mantle source of the magmas. Simple two-component mixing between depleted mantle wedge and melt derived from subducted sediment, however, cannot accommodate all observed trace element and isotope features. Pb–Nd isotope and La/Yb mixing models provide evidence for mixing of at least four components in the mantle sources of the Devil River Volcanics. The components include the mantle wedge, melts from subducted sediment, mid-ocean ridge basalt (MORB)-derived slab fluid, and MORB-derived slab melt. The concomitant increase in Nd, Th/Yb and La/Yb in the arc rocks through time reflects an increasing contribution from MORB-derived slab melts. Addition of up to 0·5% melt from subducted sediment, 3% MORB-derived slab fluid and 6% MORB-derived slab melt (all in wt %) to the mantle wedge is required to explain all magmatic compositions. Such mixing proportions are consistent with those obtained for recent arc systems.

KEY WORDS: arc; Cambrian; isotopes; trace elements; mixing

	INTRODUCTION

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
Chemical variations in subduction-related magmas reflect the interplay of multiple sources in plate collision systems. These sources include the mantle wedge beneath the arc as well as fluids and melts derived from the subducting plate. Most recent arc magmas derive from the mantle wedge above the subducting plate (e.g. Ringwood, 1974; Gill, 1981; Wyllie, 1982; Arculus & Powell, 1986). The composition of the ‘primary’ mantle wedge is generally thought to be that of the mid-ocean ridge basalt (MORB) source or even more depleted mantle, although in some cases ocean island basalt (OIB) source components in the mantle wedge have been proposed (e.g. Morris & Hart, 1983; Woodhead, 1989; Crawford et al., 1995; Danyushevski et al., 1995; Peate et al., 1997). Fluids and melts derived from both the MORB and sediment portions of the downgoing slab have been known to modify the composition of the subarc mantle significantly (Tera et al., 1986; Ellam & Hawkesworth, 1988; von Huene & Scholl, 1993; Miller et al., 1994; Kamenetsky et al., 1997). Element transfer from the slab to the mantle wedge in most recent arc systems is the consequence of two major processes: (1) melting of subducted sediment; (2) dehydration of the MORB portion of the slab (Nichols et al., 1994; Elliot et al., 1997; Hawkesworth et al., 1997). Melting of the MORB portion requires unusually steep thermal gradients, and is thought to be limited at present-day geothermal gradients to young and relatively hot oceanic crust (Drummond & Defant, 1990; Peacock et al., 1994; Yogodzinski et al., 1995; Stern & Kilian, 1996). The composition of subduction-related melts can be further modified during their ascent to the surface. Reaction of the ascending melt with wall rock peridotite in the subarc mantle can change major and trace element compositions substantially and can even account for selective depletion of the high field strength elements (HFSE) (Navon & Stolper, 1987; Kelemen et al., 1993). Assimilation of continental crust and crystal fractionation can also modify the composition of subduction-related magmas at active continental margins through assimilation–fractional crystallization (AFC) or melting, assimilation, storage and homogenization (MASH) processes (DePaolo, 1981; Hildreth & Moorbath, 1988).

This study presents a coherent isotope, major and trace element dataset for a well-preserved Cambrian arc–back-arc system in the Takaka Terrane, New Zealand, to assess the relevance of the different arc-related processes for magma petrogenesis. The Devil River Volcanics of the Takaka Terrane are ideally suited to this task because they display several outstanding features:

1. An association of different magma types, such as back-arc tholeiites, low- to high-K arc rocks and boninites are exposed in a single stratigraphic succession.
   
2. Mafic compositions are dominant, which suggests a minor role of crustal contamination and thus allows insights into mantle source processes.
   
3. Mafic phenocrysts are abundant and well preserved because the metamorphic overprint of most of the Devil River Volcanics is limited to prehnite–pumpellyite facies. Analyses of these phenocrysts yield primary magmatic isotopic compositions, especially for the Rb–Sr and U–Pb systems.
   
4. Incompatible trace element abundances show significant variations and a large range of isotope compositions.
   
5. Incompatible trace element abundances display a systematic change of isotopic signatures as the arc evolved, thus providing a rare opportunity to study temporal variations in subarc mantle processes.
   

In modern arc settings, estimates of the extent of fluid and sediment transfer from the subducting plate to the mantle wedge have relied particularly on large ion lithophile element (LILE) systematics of the volcanic rocks (e.g. Ellam & Hawkesworth, 1988; Plank & Langmuir, 1988, 1993; Hawkesworth et al., 1997). Because low-grade metamorphism may have mobilized most LILE elements (K, Rb, Sr, Ba, Cs, U) in the Devil River Volcanics, these elements cannot be used reliably for petrogenetic modelling. Abundances of rare earth elements (REE), HFSE, Th and most major elements, however, vary systematically between the rock types and behave consistently within individual volcanic suites (Table 1). Any postmagmatic changes in these element abundances are therefore likely to be minor, thus allowing their use for petrogenetic evaluation together with Sr–Nd–Pb isotope data from pristine phenocrysts.

View this table: [in this window] [in a new window]   Table 1: Analytical results for rocks of the Devil River Volcanics

 

Particular attention is paid here to the large compositional spread observed in the arc rocks, which include interbedded low- to high-K types. Previous studies have suggested that such variations are caused by amounts and compositions of the sediment subducted beneath arcs (e.g. Plank & Langmuir, 1988, 1993), subarc slab melt enrichment (e.g. Stolz et al., 1996) or differences in the degree of melting (e.g. Pearce et al., 1995). Other studies suggest the presence of enriched ‘OIB source’ type mantle domains in the sources of high-K suites (e.g. Edwards et al., 1994). Compositional systematics of the Devil River Volcanics, as evaluated below, suggests mixing of mantle- and slab-derived components in their sources as the relevant process involved. On this basis, isotope and trace element systematics of the Devil River Volcanics are therefore explained here using multi-component mixing models. Each component is identified using specific trace element or isotope features of the volcanic rocks.

	GEOLOGICAL SETTING AND PETROLOGY

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
The Takaka Terrane in Northwest Nelson on the South Island of New Zealand (Fig. 1) is part of a voluminous, mainly intra-oceanic Early Palaeozoic arc–back-arc assemblage that is now dispersed as fragments in Australia, Antarctica and New Zealand (Crawford & Keays, 1978; Weaver et al., 1984; Cooper & Tulloch, 1992; Crawford & Berry, 1992; Münker & Cooper, 1995). These subduction-related rocks all originated from the collision of the palaeo-Pacific plate with the Gondwana continent, which lasted from the late Early to the Late Cambrian. Most of the Cambrian volcanic rocks in the Takaka Terrane have been grouped together as Devil River Volcanics and are interbedded with partly continent-derived turbidites (Haupiri Group, Münker & Cooper, 1999). Both Devil River Volcanics and Haupiri Group were deposited in submarine facies. The Devil River Volcanics were formed within 20 My as constrained by both palaeontological and isotopic age constraints (Münker & Cooper, 1999). U–Pb SHRIMP zircon and amphibole ⁴⁰Ar/³⁹Ar ages for the Devil River Volcanics vary from 490 to 515 Ma (C. Münker, T. Ireland, J. Wijbrans & S. D. Weaver, unpublished data, 1997).

View larger version (45K): [in this window] [in a new window]   Fig. 1. Location of the Takaka Terrane, New Zealand, within the Early Palaeozoic belts of SE Gondwana (after Cooper & Grindley, 1982; Cooper & Tulloch, 1992).

 

The Devil River Volcanics constitute a coherent arc–back-arc succession and are grouped into three major units: (1) calc-alkaline arc rocks of the Benson Volcanics; (2) back-arc tholeiites of the Mataki Volcanics; (3) boninitic intrusives of the Cobb Igneous Complex (Münker & Cooper, 1999). The Benson Volcanics are by far the most voluminous unit and are subdivided into nine suites, of which eight are predominantly of basaltic to basaltic–andesitic composition. One suite, the Heath volcanic suite, is andesitic, and is treated here separately, together with other rare felsic Benson Volcanics.

A subdution-related setting of the Benson Volcanics is indicated by the presence of the Balloon mélange in the Takaka Terrane. This soft sediment mélange of known Late Cambrian age (Münker & Cooper, 1999) has incorporated parts of all volcanic sequences, which indicates that a compressional regime was present until the Late Cambrian–Early Ordovician. In the Late Palaeozoic, the Cambrian sequence of the Takaka Terrane was faulted into 12 fault-bounded slices, following final amalgamation to the Australian Antarctic Gondwana margin. Stratigraphic overlap between the Cambrian fault slices [documented in detail by Münker & Cooper (1999)], however, demonstrates that all of the above volcanic units are part of one sequence and, consequently, have been temporally and spatially related.

Most Devil River Volcanics are overprinted by prehnite–pumpellyite facies metamorphism. They contain abundant relict phenocrysts of clinopyroxene, amphibole, plagioclase and magnetite. The Benson Volcanics can be classified into low- to high-K groups based on their distinctive ranges in Th/Yb (Fig. 2). A systematic enrichment of the rocks in Th/Yb through time is also evident (Fig. 2). The boninitic samples come from gabbros, cumulates and melt layers within shallow layered magma chambers, which were emplaced into Haupiri Group sediments and back-arc sequences of the Mataki Volcanics (Münker & Cooper, 1999). Mg-rich clinopyroxenes and refractory compositions of chromite [Cr/(Cr + Al) > 0·6, Hunter, 1977) indicate the boninitic compositions of the parental melts.

View larger version (36K): [in this window] [in a new window]   Fig. 2. Plot of Th/Yb vs Ta/Yb for all analysed samples of the Devil River Volcanics. Fields for N-MORB, enriched mantle and subduction-related rocks are from Pearce (1983); TH, CA and S stand for tholeiitic, calc-alkaline and shoshonitic arc rocks, respectively. The arrow indicates a systematic change in volcanism through time towards more incompatible trace element enriched compositions.

 

The oldest (early Mid to middle Mid Cambrian) section of the Devil River Volcanics comprises the Mataki Volcanics (back-arc), interbedded with rare low-K arc rocks of the Benson Volcanics. The lower part of the Mataki Volcanics and interbedded Haupiri Group sediments are intruded by boninitic intrusives of the Cobb Ignous Complex (515 ± 7 Ma, U–Pb SHRIMP). This observed co-occurrence of boninites and back-arc rocks is known from present-day nascent back-arc basins, where such magmatism is caused by the ascent of MORB source mantle diapirs into refractory arc source mantle (e.g. Crawford et al., 1989). The youngest part of the Devil River Volcanics (late Middle to early Late Cambrian) consists of medium- to high-K rocks of the Benson Volcanics (490–500 Ma, U–Pb SHRIMP, Ar–Ar). High-K rocks are predominant in the uppermost part of this section, but also occur interbedded with medium-K rocks. Because all these volcanic units occur together in a single sequence and were both erupted over a longer time interval (10 My), a genetic link between them seems plausible.

	RESULTS

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
Major and trace element analyses and Sr–Nd–Pb isotope data for volcanic rocks and sediments of the Takaka Terrane are listed in Tables 1 and 2. Major and trace elements were analysed by X-ray fluorescence (XRF) (Phillips PW 1480) and inductively coupled plasma mass spectrometry (ICPMS) (VG Plasma Quad II+) on a total of 120 samples (both at the Geochemisches Institut, Universität Göttingen). Particular attention was paid to an accurate and precise determination of HFSE. Details of the analytical procedure have been given by Münker (1998). H₂O contents were analysed by Karl-Fischer titration, and CO₂ analyses were performed using an Eltra CS 1000RF IR absorption spectrometer. Major element contents of all samples, as reported below, were recalculated on an H₂O- and CO₂-free basis.

View this table: [in this window] [in a new window]   Table 2: Results of Sr–Nd–Pb isotope composition and Rb–Sr, Sm–Nd and U–Pb isotope dilution measurements

 

⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd data were obtained for 20 clinopyroxene and amphibole separates and six whole-rock samples covering all the above volcanic units. The whole-rock samples comprise three high-K rhyolites and three samples from the boninitic Cobb Igneous Complex, of which one (NZ-376/1) is a late-stage plagiogranite. Additional Sr–Nd data for 15 Haupiri Group sediments have been given by Wombacher & Münker (1998). Twelve representative mineral separates and seven Haupiri Group sediments were also analysed for their Pb isotope composition. All isotope values were calculated back to an age of 500 Ma using the measured U–Pb, Rb–Sr and Sm–Nd abundances in the mineral separates and rocks, respectively. The uncertainties in the ages (500 ± 15 Ma) have a negligible effect on the recalculated initial isotopic ratios and are within the analytical uncertainty.

SiO₂ contents are consistent with modal compositions of the examined samples (Table 1, and Münker & Cooper, 1999) and show the predominance of basalts and basaltic andesites. Andesitic to rhyolitic compositions are rare and are mainly confined to the Heath volcanic suite of the Benson Volcanics. The arc rocks (Benson Volcanics) are generally rich in Al₂O₃ (14–21 wt %) and poor in TiO₂ (0·4–1·5 wt %). By comparison, back-arc tholeiites of the Mataki Volcanics contain less Al₂O₃ (13–16 wt %) and much more TiO₂ (1·4–3·5 wt %). MgO, Ni and Cr contents of both units cover wide ranges (0·5–11 wt %, 40–400 ppm and 5–100 ppm, respectively), reflecting olivine and clinopyroxene fractionation.

The Benson Volcanics display an extremely wide range in their incompatible trace element abundances [e.g. Th/Yb 0·3–11 (Fig. 2) and La/Yb 2–30]. Th/Yb, Ta/Yb and La/Yb are strongly correlated (Fig. 2 and Table 1) and increase through time. Zr and Y abundances are in the range 20–150 ppm and 8–23 ppm, respectively. Back-arc samples are confined to lower Th/Yb (0·3–0·8) and La/Yb (2–6) but show larger ranges in Zr (80–230 ppm) and Y (20–50 ppm) than the arc rocks. Nb abundances in both units also vary considerably and range from 0·7 to 30 ppm.

The boninitic samples of the Cobb Igneous Complex show extreme variations in their MgO contents from 0·6 wt % (plagiogranite) to >40 wt % (dunitic cumulate). Ni and Cr vary over similarly large ranges (7–470 and 6–3450 ppm). Two samples from melt layers (NZ-330, NZ-333, Table 1) that are intercalated with the gabbros and cumulates show major element compositions that are similar to those of present-day low-Ca boninites (Crawford et al., 1989). Even the most differentiated samples of the Cobb Igneous Complex are extremely depleted in incompatible trace elements. Compositions are in the range 0·3–4 ppm for Nb, 4–60 ppm for Zr and 0·5–12 ppm for Y. Th/Yb and La/Yb also show large variations (0·4–0·9 and 1–12, respectively).

⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd compositions at 500 Ma of the Devil River Volcanics (Fig. 3) cover an extremely large range from -5 (boninites) to +6 (back-arc) in Nd and from 0·703 to 0·712 in ⁸⁷Sr/⁸⁶Sr with no systematic difference between the groups. Arc rocks of the Benson Volcanics cover a large range in Nd₅₀₀ from -1·5 to +4, and display a systematic increase in Nd₅₀₀ with increasing enrichment in their incompatible trace element contents (Fig. 3, and Fig. 10b below). Nd₅₀₀ values for the low-K group range from -1·5 to +2 and are much higher for the medium-K (+2 to +3·5) and high-K groups (+3 to +4). Sr isotope compositions do not show this trend and overlap between the different groups, except for sample NZ-113b, which has an anomalously high (⁸⁷Sr/⁸⁶Sr)₅₀₀ of 0·71158 (Fig. 3). The three high-K dacite–rhyolite samples have Nd₅₀₀ from -2 to +2 and (⁸⁷Sr/⁸⁶Sr)₅₀₀ from 0·70442 to 0·70587, lying between the compositions of the mafic high-K rocks and those of the Haupiri Group sediments of the Takaka Terrane (-11 to +2, Wombacher & Münker, 1998).

View larger version (33K): [in this window] [in a new window]   Fig. 3. Plot of 87Sr/86Sr vs Nd (both recalculated at 500 Ma) for all analysed mineral separates of the Devil River Volcanics. Nd isotope compositions become increasingly radiogenic from the boninites to the high-K and back-arc rocks.

 

View larger version (44K): [in this window] [in a new window]   Fig. 10. Calculated three-component Pb–Nd isotope and La/Yb mixing curves in comparison with the Devil River Volcanics. Mixing components are depleted mantle wedge (DMM), a 15% sediment melt and a fluid derived from the MORB portion of the subducting slab (Table 3). La/Yb8/10 indicates fractionation-corrected La/Yb values (Fig. 6). Pb isotope ratios are extremely sensitive to contamination by subducted sediment (a). MORB-derived fluids from the slab cause a selective enrichment of the mantle in MORB-derived Pb, thus lowering the Pb isotope ratio without changing the Nd isotope ratio significantly. The high La/Yb8/10 of the medium- to high-K arc rocks, however, cannot be explained by a three-component model (b).

 

View this table: [in this window] [in a new window]   Table 3: Mixing endmembers used for model calculations

 

View larger version (48K): [in this window] [in a new window]   Fig. 6. Plots of Nb (highly incompatible), La/Yb (highly/moderately incompatible ratio) and Y (moderately incompatible) vs MgO for the major groups of the Devil River Volcanics. Primitive melt compositions (grey boxes) are assumed by interpolation of fractionation trends (continuous lines) to MgO 8 wt % (10% for boninites; see text). Except for Nb in the high-K rocks, the arc rocks (Benson Volcanics) do not show clear incompatible trace element enrichment by fractionation, suggesting these element abundances to be rather controlled by source processes. Conversely, trace element abundances in the boninites (Cobb Igneous Complex), the back-arc tholeiites (Mataki Volcanics) and in the andesitic to rhyolitic high-K arc rocks (Heath volcanic suite and rhyolites of the Benson Volcanics) vary significantly with degree of fractionation.

 
Pb isotope compositions of the Devil River Volcanics differ significantly from those of the interbedded Haupiri Group sediments (Fig. 4). Initial clinopyroxene/amphibole compositions range from 18·4 to 19·3 in (²⁰⁶Pb/²⁰⁴Pb)₅₀₀ and from 15·60 to 15·70 in (²⁰⁷Pb/²⁰⁴Pb)₅₀₀. Whole-rock samples from the interbedded Haupiri Group sediments show significantly less radiogenic compositions from 17·4 to 18·1 in (²⁰⁶Pb/²⁰⁴Pb)₅₀₀ and from 15·58 to 15·62 in (²⁰⁷Pb/²⁰⁴Pb)₅₀₀.

View larger version (52K): [in this window] [in a new window]   Fig. 4. 207Pb/204Pb vs 206Pb/204Pb plots for the Devil River Volcanics and Haupiri Group sediments of the Takaka Terrane, all recalculated at 500 Ma. MORB was recalculated using a µ value of eight, crustal reservoirs using both eight and 10. Pb compositions of the Devil River Volcanics cannot be explained by shallow-level mixing between a MORB source derived magma and Takaka Terrane sediment or continental crust of similar composition (e.g. LFB crust, a). The volcanic compositions can only be explained by mixing of MORB source mantle with subducted sediments derived from the Archaean cratons (b). Continuous lines are mixing curves of MORB source mantle with representative Archaean source sediment endmembers (Table 3). The stippled line indicates possible mixing of ‘high-Pb’ Archaean source-derived sediment with Takaka Terrane sediment before subduction (see text).

 

Two observations that need to be explained by any successful petrogenetic model are (1) the unusually radiogenic Pb isotope compositions in the Devil River Volcanics, and (2) their spread in Nd. The latter is unusual, because in recent intra-oceanic arc systems, Nd values are generally within the range +2 to +9 (White & Patchett, 1984; Arculus & Powell, 1986). A further exceptional feature is the observed increase in incompatible trace element ratios and abundances (e.g. Th/Yb, La/Yb, Fig. 2, and Fig. 10b below) with increasing Nd. Moreover, the medium- and high-K arc rocks of the Devil River Volcanics show a pronounced enrichment in HFSE (see Figs 7 and 8 below; Table 1) compared with average arc tholeiitic compositions (e.g. Ewart, 1982). Such a feature is often explained by an OIB component (high Nb/La and Nb/Th) in the source of arc magmas (e.g. Morris & Hart, 1983; Woodhead, 1989; Edwards et al., 1994; Danyushevski et al., 1995).

View larger version (39K): [in this window] [in a new window]   Fig. 7. Calculated Zr–Y and Nb–Yb compositions for melting of variably depleted mantle reservoirs (Appendix B) in comparison with fractionation-corrected contents in the Devil River Volcanics. Labels along melting curves indicate degree of partial melting (F; a value of 0·1 indicates 10% melting). Percentages indicate depletion of a primitive mantle composition [PRIMA, after Hofmann (1988)] before secondary melt extraction. Stippled lines are isograds of partial melting at different source depletions. Minimal melt compositions are marked with large dots. Melting of primitive mantle or its depleted residues fails to account for the Zr–Nb abundances in most Devil River Volcanics.

 

View larger version (23K): [in this window] [in a new window]   Fig. 8. Plots of Nb/Th vs Nb/Yb and of Zr/Nb vs La/Yb for all samples of the Devil River Volcanics. Symbols are as in Fig. 2. Arrows indicate the influence of various petrogenetic processes on compositional trends. An OIB component in the mantle source can be largely ruled out on the basis of Nb/Th (a) and Nb/Ta systematics (b). Zr/Nb ratios particularly decrease with increasing supply of slab melts to the mantle source because slab melts have low Zr/Nb, and transport Nb more efficiently to the wedge than slab fluids.

 

The trace element and isotope geochemistry of the Devil River Volcanics is first evaluated below by assessing the role of combined fractional crystallization and crustal assimilation processes (AFC). On this basis, the trace element and isotope characteristics of primary magmas are reconstructed by back-correction. In a second step, the influences of partial melting processes and mantle wedge depletion on incompatible element abundances and ratios are examined. Using HFSE systematics in the Devil River Volcanics, the possible presence of an OIB source mantle component in the magma sources is then discussed. In a final step, the nature of element transfer from the subducting slab to the mantle wedge (melt or fluid) is discriminated and the amount of slab component is quantified by multi-component modelling.

	DISCUSSION

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
Fractional crystallization–crustal assimilation processes
Fractionation of olivine, clinopyroxene and magnetite is evident from large variations in mg-number, Ni, Cr, Ti and Sc (Table 1, Fig. 5), and probably controls the compositions of major and compatible trace elements. Amphibole becomes important for the more differentiated rocks (e.g. Heath volcanic suite of the Benson Volcanics), in which it occurs as a phenocryst phase. The back-arc rocks show strong enrichment in TiO₂ with decreasing MgO, which suggests little or no magnetite fractionation and supports their classification as tholeiites. Conversely, the arc rocks show constant to slightly increasing TiO₂, suggesting a calc-alkaline trend, where magnetite fractionation controls the TiO₂ contents. High Al₂O₃ contents in the arc rocks (Benson Volcanics) indicate plagioclase accumulation as suggested for recent high-alumina rocks (Crawford et al., 1987).

View larger version (55K): [in this window] [in a new window]   Fig. 5. Plots of Nd and 87Sr/86Sr (500 Ma) vs mg-number for all examined mafic and andesitic samples of the Devil River Volcanics. Average compositions of primitive N-MORB melts are plotted for comparison. Arrows indicate schematic magmatic trends for pure (closed system) fractional crystallization and for assimilation of older continental crust during fractional crystallization (AFC). Most examined samples lack any significant crustal contamination.

 

Correlations between isotope ratios and compatible elements (Fig. 5) help to delimit crustal assimilation processes. Constant values of Nd₅₀₀ and (⁸⁷Sr/⁸⁶Sr)₅₀₀ at different mg-numbers in most suites (Fig. 5) preclude significant shallow-level assimilation of continental crust during magma crystallization (AFC). Shallow-level assimilation of continental material, such as Haupiri Group sediment, would cause a decrease in Nd and an increase of ⁸⁷Sr/⁸⁶Sr in the magma suites as indicated by schematic arrows in Fig. 5. Some samples of the low- and high-K calc-alkaline groups, in fact, do show an increase in ⁸⁷Sr/⁸⁶Sr at relatively low mg-number (high-K group and NZ-113b from the low-K group). However, crustal contamination effects observed for Sr isotopes are not detectable for Nd (Fig. 5). Pb isotope compositions in the Devil River Volcanics are different from those of the Haupiri Group sediments and do not lie on a mixing array between the sediments and recalculated Cambrian MORB source mantle (Fig. 4a). The Pb isotope compositions in the analysed samples therefore appear not to be affected by shallow-level assimilation of Haupiri Group sediments. Only sample NZ 113b has anomalously low ²⁰⁶Pb/²⁰⁴Pb (Fig. 4a) and overlaps in its Pb composition with compositions obtained for Haupiri Group sediments. This observation is consistent with the elevated ⁸⁷Sr/⁸⁶Sr in this sample, thus confirming assimilation of Haupiri Group sediment in this case. In summary, variations in Pb and Nd isotopes are therefore much more suitable tracers for source processes than Sr isotopes because both appear not to be significantly modified by shallow-level crustal contamination. Models for the mantle sources of the Devil River Volcanics below are therefore limited to the Pb–Nd isotope systems.

Dacitic–rhyolitic samples from the high-K calc-alkaline group have Nd₅₀₀ and (²⁰⁶Pb/²⁰⁴Pb)₅₀₀ that are significantly lower than those of the mafic high-K rocks (Table 2 and Fig. 4a). Pb isotope compositions of rhyolites (e.g. NZ-380) plot in the sediment field (Fig. 4a), suggesting a major Pb contribution from Takaka Terrane sediment. Dacite–rhyolite generation therefore involved assimilation of large amounts of Takaka Terrane sediment by mafic melts rather than closed system fractional crystallization.

As shown in Fig. 6, primary melt compositions for the different groups of the Devil River Volcanics were obtained from the measured abundances by back-correction. Following Plank & Langmuir (1988), the fractionation-corrected trace element abundances are obtained by extrapolation of observed fractionation trends to 8 wt % MgO, indicated as continuous lines. The boninites were extrapolated to 10 wt % MgO, which is the highest MgO content found in samples with mg-numbers between 70 and 75 (i.e. in equilibrium with olivine). These higher initial Mg contents in the boninites reflect larger (>30%) degree of melting (Fig. 7). Except for highly incompatible elements (e.g. Nb) in the high-K rocks, all mafic arc rocks of the Benson Volcanics lack any significant incompatible element enrichment by fractionation. In contrast, all abundances and ratios in the boninites, back-arc tholeiites and felsic high-K rocks increase steadily with increasing degree of fractionation. Corrected trace element abundances (e.g. Nb_8·0, Zr_8·0) are used for these suites below.

Partial melting, wedge depletion and the HFSE budget
Partial melting processes involved in generating subduction-related magmas are mainly controlled by (1) adiabatic mantle decompression and (2) the hydration of the mantle through slab-derived fluids, which lower the mantle solidus (Plank & Langmuir, 1988; Pearce & Parkinson, 1993). In a case study on the Mariana arc, Stolper & Newman (1994) showed a strong correlation between the degree of partial melting and the fluid flux from slab to mantle wedge. In their study, the large-scale influx of highly incompatible elements from slab-derived fluids or melts is reflected in positive correlations for all LILE (soluble in fluids) with H₂O contents in fresh volcanic glasses. On the other hand, Ti, Y and Zr (relatively insoluble in fluids) show negative correlations with H₂O. This trend indicates an increase in the degree of partial melting with increasing H₂O flux in the mantle wedge because Ti, Zr and Y are moderately incompatible. The budget of highly incompatible elements (LILE) in the magmas is therefore controlled by fluid influx, whereas the budget of the moderately incompatible elements [heavy REE (HREE), Zr, Y, Ti] is controlled by partial melting processes and mantle depletion. Empirically and experimentally determined slab–fluid partition coefficients (McCulloch & Gamble, 1991; Brenan et al., 1995b; Keppler, 1996; Stalder et al., 1998) confirm this element behaviour. On the basis of this systematics, Pearce & Peate (1995) proposed the classification of the above element groups as ‘conservative’ (HFSE and HREE) and ‘nonconservative’ (LILE) with respect to their mobilization by slab-derived fluids or melts in subduction systems. Conservative elements are thus preferred here to constrain variations in partial melting, whereas variations in nonconservative elements and isotope ratios (Sr and Pb isotopes, Th) are used to quantify slab-derived inputs.

Some approaches for estimating the degree of melting and wedge depletion in subduction-related rocks have been presented by Pearce & Parkinson (1993) and Woodhead et al. (1993). These studies demonstrate that compatible elements such as Sc or Cr (D_mantle/melt > 1) are buffered by the residual mantle and remain virtually constant over large variations in degree of melting. Moderately incompatible elements, such as Y, Yb or Ti (D_mantle/melt = 0·1–1), are more sensitive to the degree of melting, but are still buffered sufficiently by the mantle residuum to be relatively insensitive to the degree of mantle depletion. These elements are therefore well suited to monitor variations in degree of partial melting; furthermore, the abundances of these elements differ little between the different mantle sources of arc and back-arc magmas. More incompatible elements (D_mantle/melt < 0·1), such as Nb, are extremely sensitive to depletion events because they enter the melt very efficiently. Fractionation-corrected Nb, Zr, Ti and Y abundances (all conservative elements) have therefore been used for recent arc rocks to estimate degrees of melting and wedge depletion (Pearce & Parkinson, 1993; Woodhead et al., 1993). A strictly conservative behaviour of HFSE in the Devil River Volcanics, however, is questioned by Nb/Ta systematics, which suggests that Nb and Ta were slightly enriched in the mantle source by slab components (Münker, 1998). To clarify the isssue, variations in Nb (highly incompatible) vs Yb (moderately incompatible) and of Zr/Y vs Y are therefore modelled in a two-stage melting model (Appendix A; Fig. 7). Zr has been chosen, because its mobility in slab components and its compatibility during melting are transitional between Nb and Y–Yb.

In Fig. 7, only the boninites, some low-K arc rocks and some back-arc rocks fall within the field for the calculated mantle melts. In most arc rocks, Nb is enriched by up to a factor of 10 compared with primitive mantle melts at similar degrees of melting (and Yb contents). Similarly, most samples of the Devil River Volcanics plot at higher Zr/Y than the primitive mantle melting curve. Moreover, Zr/Y ratios are strongly correlated with Zr contents, but not with Y contents, which overlap in the arc magmas. This observation is important for two reasons:

1. the influence of residual garnet in the magma sources on Zr/Y can be largely excluded because of its high partition coefficient for Y (D_garnet/melt = 4–11, Jenner et al., 1994) relative to Zr (D_garnet/melt = 0·4–0·7, same reference). Varying amounts of residual garnet in the sources would change the Y contents in the magmas and Zr/Y would correlate with Y.
   
2. The Nb and Zr contents of most Devil River Volcanics can therefore not be explained by melting of primitive mantle or its depleted residues. Contribution of the ‘excess’ Nb and Zr to the mantle source is required, either from an ‘OIB’ source component or from the subducted slab.
   

The Zr–Nb systematics of the Devil River Volcanics therefore rules out any quantification of mantle wedge depletion for this arc system. However, higher mg-numbers and Cr abundances and low CaO/Al₂O₃ in the boninitic melts (Fig. 5, Table 3) suggest that the boninitic mantle source was significantly more depleted than those of the arc and back-arc rocks. This is consistent with recent arc systems where subarc mantle wedge compositions are known to be variably depleted, often more than typical MORB source mantle (e.g. Woodhead et al., 1993). Trace element modelling for the boninites below therefore needs to take into account a more depleted mantle source than for the arc and back-arc rocks.

Fractionation-corrected Y–Yb abundances suggest extraction of the back-arc rocks after 5–15% melting and of the arc rocks after 15–35% melting. Y–Yb abundances of the boninites suggest >30% partial melting. Assuming a pooled fractional melting process as above, such estimates of course only represent averages, but are strikingly similar to estimates obtained for recent arc and back-arc settings (Plank & Langmuir, 1988; Pearce & Parkinson, 1993; Woodhead et al., 1993; Stolper & Newman, 1994).

OIB mantle source vs slab component
In particular, high Nb abundances have been used as major evidence for an OIB component in the mantle source of subduction-related rocks [e.g. for high-K rocks in Indonesia (Edwards et al., 1994)]. Such an approach, however, has been challenged by, amongst others, Stolz et al. (1996) and Hoogewerff et al. (1997), who have attributed the high Nb abundances in the Indonesian high-K rocks to addition of melts derived from subducted sediment to the mantle wedge. Stolz et al. (1996) have reported extremely high Nb/Ta in Indonesian high-K rocks (>20). Because such high Nb/Ta are not known for OIB (Jochum et al., 1996), it has been suggested that Nb and Ta in the arc rocks originate from melts derived from the subducting slab. Apart from Nb/Ta ratios, Nb anomalies are another important discriminant between subduction components (negative anomaly) and ‘OIB’ components (no or positive anomaly). In contrast to Nb/Th and Nb/Ta, Sr–Nd–Pb isotope ratios of these two potential components partially overlap (White & Patchett, 1984; Hofmann, 1997) and are therefore a less suitable discriminant.

Figure 8a shows that Nb/Th in the arc rocks of the Devil River Volcanics ranges between one and six and is not correlated with Nb/Yb. Such systematics rules out any significant role of an OIB component in the source because OIB generally has much higher Nb/Th (>10, Jochum et al., 1991). A correlation of Nb/Th with Nb/Yb would be expected in the case of mixing between MORB source and OIB source mantle. Nb–Ta enrichment by a subduction component in the Devil River Volcanics is also suggested by Nb/Ta ratios ranging up to 25 ± 2 (Münker, 1998). Nb/Ta ratios in OIB are limited to chondritic (17·5 ± 1) or lower values (Jochum et al., 1996). Correlations of Nb/Ta with Th/Yb and La/Yb (both ‘nonconservative’) in the Devil River Volcanics (Münker, 1998) furthermore indicate that significant amounts of Nb and Ta were added by the slab component. Because rutile in the subducted slab has D_Nb–Ta > 100 with fluids and melts (Green & Pearson, 1987; Brenan et al., 1994), even small amounts of this mineral can control the Nb–Ta budget in the slab component. Subchondritic ratios in the boninites and low-K arc rocks (<17·6) can reflect fluid enrichment because the partitioning of Nb between rutile and fluid is higher than for Ta (Brenan et al., 1994). Superchrondritic ratios in most high-K rocks (>17·6) suggest melt enrichment as a result of lower partitioning of Nb compared with Ta between rutile and melt (Green & Pearson, 1987).

In Fig. 8b, samples with high La/Yb and high Nb enrichment relative to Zr (low Zr/Nb) always display Nb/Ta higher than chondrite. Zr/Nb ratios of the Devil River Volcanics lie below the values for average N-MORB [Zr/Nb =30, after Hofmann (1988)]. Addition of slab components to the mantle wedge can explain this feature: Nb is much more incompatible than Zr and, hence, much more enriched than Zr in small degree melts such as slab melts (see also Appendix B). Addition of slab melts to the depleted mantle wedge therefore causes extreme enrichment in Nb relative to Zr and a decrease in Zr/Nb (Fig. 8c). Alternatively, the Nb–Ta budget of samples having low Zr/Nb and low La/Yb at subchondritic Nb/Ta (Fig. 8b) must be controlled by slab fluids. Recent experimental work (Brenan et al., 1995b; Stalder et al., 1998) reports a higher solubility of Nb in subduction fluids relative to Zr.

Evidence for source contamination by subducted sediment
Constant Nd at different mg-numbers within single suites of the Devil River Volcanics together with predominantly mafic compositions (Fig. 5) rule out AFC (e.g. Hildreth & Moorbath, 1988). Figure 4a also shows that AFC contamination of magmas derived from ancient MORB source mantle with Haupiri Group sediment alone [or a Lachlan Fold Belt basement (Culloch & Woodhead, 1993)] cannot account for the Pb compositions observed in the Devil River Volcanics. Such contamination would result in significantly less radiogenic compositions, as indicated by the schematic mixing trends shown in Fig. 4a. Mixing between MORB and different OIB reservoirs (Hofmann, 1997), calculated to 500 Ma (Fig. 4b), also cannot explain the high ²⁰⁷Pb/²⁰⁴Pb observed in some samples of the Devil River Volcanics.

Hence, source contamination by subducted sediment is a preferred hypothesis. This subducted sediment must have extremely radiogenic Pb and unradiogenic Nd, not known from Palaeozoic foldbelts of southeastern Australia (McCulloch & Woodhead, 1993; Turner et al., 1993). Subduction of sediment beneath the Devil River arc that was at least in part derived from Archaean sources is strongly supported by the Cambrian palaeogeography (Dalziel, 1997; Münker, 2000): large areas of Archaean crust were exposed along the rims of the palaeo-Pacific ocean. These areas include Archaean provinces in Antarctica (DePaolo et al., 1982) and North America (Wyoming and Slave Provinces, Bowring et al., 1989; Mueller et al., 1993). Extremely radiogenic (²⁰⁷Pb/²⁰⁴Pb)₅₀₀ (15·5–17), highly variable (²⁰⁶Pb/²⁰⁴Pb)₅₀₀ and unradiogenic Nd₅₀₀ (-15 to -40) (Fig. 4b) are typical features of these Archaean domains.

²⁰⁷Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb mixing lines between depleted MORB source mantle and subducted sediment are shown as continuous lines in Fig. 4b. Bulk mixing of subducted sediment derived from Archaean sources with mantle wedge material was assumed. Compositions of the endmembers used are listed in Table 3. Three representative compositions of Archaean sources, named ‘low-’, ‘medium-’ and ‘high-Pb sediment’, referring to their enrichment in radiogenic Pb, were used. Admixture of only 0·1% of these sediment endmembers to the mantle wedge is sufficient to account for the compositions of the Devil River Volcanics, as calculated for recent arcs (e.g. White & Patchett, 1984; White & Dupré, 1986). This amount, however, is much lower than up to 2% added sediment obtained by Nd mixing calculations (not shown); therefore refinement of the above mixing model is required.

Mixing of sediment derived from Archaean sources on the subducting plate with Takaka Terrane sediment in the subduction trench, a geologically most realistic case, would significantly lower the ²⁰⁷Pb/²⁰⁴Pb composition of the contaminant. It is furthermore likely that the sediment on the subducting plate is not purely ‘Archaean’ in its provenance but also contains components derived from younger sources with unradiogenic Pb. Hence, a mixture of pure Takaka Terrane sediment and sediment derived from Archaean sources of ‘high-Pb’ composition is calculated (dotted line in Fig. 4b). A mixture of 70% average Takaka Terrane sediment (Table 3) and 30% ‘high-Pb’ Archaean source-derived sediment can best explain the volcanic compositions. Only such a mixture has the appropriate ²⁰⁶Pb/²⁰⁴Pb value (18·9, Fig. 4b) to account for the volcanic compositions by mixing with MORB source mantle. Pb–Nd isotope compositions of this sediment mixture are listed in Table 3.

Entrainment of subducted sediment into the mantle wedge occurs as melt rather than by bulk mixing (Nichols et al., 1994; Elliot et al., 1997; Hawkesworth et al., 1997). Melts of sediments can have significantly more fractionated trace element ratios than their protoliths, thus resulting in different mixing proportions with the mantle wedge. The composition of melts derived from average pelagic sediment were therefore calculated (Appendix B; Table 3). Plots of La/Yb, Pb/Nd, and Nd vs degree of melting for the calculated melts are shown in Fig. 9. Pb/Nd of the melts are close to those of the bulk sediment (Table 3), but La/Yb and Nd contents might increase by a factor of 10 at low degrees of melting. In what follows, a 15% sediment melt was used as representative of an (upper) average composition.

View larger version (24K): [in this window] [in a new window]   Fig. 9. Variations of La/Yb, Pb/Nd and Nd abundances in sediment-derived and MORB-derived (gt amphibolite and eclogite) slab melts at different degrees of partial melting (Appendix B). MORB and sediment compositions are assumed as in Table 3.

 

Mixing of up to 0·4% sediment-derived melt with the mantle wedge is required to explain the Nd isotope compositions in the Devil River Volcanics (Fig. 10a), a value that is much lower than for bulk sediment mixing. The calculated Pb isotope compositions, however, are still too radiogenic to fit Nd compositions (Fig. 10a), as Pb/Nd is not changed much during sediment melting (Fig. 9). Neither variation in isotope composition of the sediment nor variation in the way in which the sediment is transferred to the mantle (melt or bulk mixing) can therefore explain the discrepancy between Pb and Nd results. Independent addition of a third component to the wedge, i.e. slab-derived fluid, is thus required to explain the observed isotope systematics in the Devil River Volcanics.

Three components—fluids derived from subducted MORB
Fluids derived from the MORB portion of the subducted slab have been invoked in many recent arc systems as additional mixing endmembers in addition to subducted sediment (Ellam & Hawkesworth, 1988; McCulloch & Gamble, 1991; Miller et al., 1994; Hawkesworth et al., 1997). Fluids generally do selectively enrich Pb in the mantle wedge compared with Nd (Table 3 and references therein). Addition of MORB-derived fluids to the mantle wedge can thus explain why the Pb isotope ratios in the Devil River Volcanics are too high at fitting Nd, if mixing of pure sediment with the wedge is modelled. Selective admixture of MORB (= mantle)-derived Pb by the fluid lowers the Pb isotope ratio of the mantle wedge although Nd is virtually unaffected.

Nd–Pb mixing lines between mantle wedge and sediment melt for different amounts of fluid addition are shown in Fig. 10a. The isotopic compositions of the Devil River Volcanics can be explained by admixture of up to 3% MORB-derived fluid and up to 0·4% melt of subducted sediment to the mantle. Both results are within the range observed for recent arc systems (e.g. White & Patchett, 1984; White & Dupré, 1986; Ellam & Hawkesworth, 1988). Calculated fluid fluxes for individual samples increase with calculated sediment contributions. These results therefore suggest a scenario where slab dehydration and sediment melting occur simultaneously.

Three-component mixing of wedge, MORB fluid and sediment melt can therefore explain the isotopic compositions of the Devil River Volcanics. In contrast, the enriched incompatible trace element compositions in most arc rocks of the Devil River Volcanics cannot be explained (Fig. 10b). The Nd would be too low at the La/Yb_8/10 (and Th/Yb) observed in the medium and high-K rocks. This is because the sediment melt and not the slab fluid controls the light REE (LREE) and Th budget of the slab component (Table 3).

A fourth component: melts derived from subducted MORB
High Nd at high La/Yb and Th/Yb in the high-K rocks strongly suggest addition of slab melt that is at least in part derived from subducted MORB (high Nd). Incorporation of such MORB-derived slab melts into the sources of recent subduction-related magmas has been reported by, amongst others, Kepezhinskas et al. (1995), Yogodzinski et al. (1995) and Stern & Kilian (1996).

A calculation of slab-derived MORB melts is difficult because the source compositions involved could range from purely eclogitic to amphibolitic, depending on how far the phase transformation in the subducting slab has progressed (Drummond & Defant, 1990). Therefore, melts from both eclogitic and garnet amphibolitic sources were calculated (Table 3, Fig. 9). Variations in La/Yb, Pb/Nd and Nd vs degree of melting are shown for the calculated melts in Fig. 9. The Pb/Nd ratios of both eclogite and garnet amphibolite melts remain more or less constant over a large range of melting and are similar to MORB. The La/Yb and Nd abundances in the melts are significantly higher than in MORB. Because of the strong effect of garnet on REE fractionation, La/Yb ratios and Nd abundances in the eclogitic melts are higher than those in the garnet amphibolite melts. As average MORB melt composition, a 5% melt from an eclogitic MORB residue (together with the 15% sediment melt) was chosen for modelling (Table 3). With respect to sediments, MORB eclogite or MORB garnet amphibolite has a much higher solidus (Nichols et al., 1994), so that such low degrees of MORB melting are realistic.

Figure 11a shows Nd vs La/Yb_8/10 for melts derived from different mixtures of MORB source mantle wedge, MORB-derived slab melts and sediment-derived slab melts. A comparison with compositions of the Devil River Volcanics shows that addition of an 80:20 MORB:sediment melt mixture to the mantle wedge can explain the Nd–La/Yb_8/10 ranges of the medium- to high-K groups. Only small changes in sediment contribution to the slab melt cause large shifts in Nd of the mantle source (Fig. 11b). The medium- and high-K rocks plot tightly along an array parallel to the slab melt mixing lines, suggesting that the slab melts mixed into the sources of all samples were homogeneous. Homogeneous slab melt compositions strongly suggest that the MORB and sediment components were already mixed shortly after slab melting and were not added to the wedge as separate components. The La/Yb_8/10 ratios of the volcanic rocks suggest addition of up to 6% slab melt (Fig. 11b). This estimate might vary between 3 and 20%, depending on the slab petrology and degree of slab melting chosen (Fig. 9).

View larger version (45K): [in this window] [in a new window]   Fig. 11. Calculated Nd isotope and La/Yb8/10 mixing curves for melts derived from different mantle wedge and slab melt mixtures in comparison with the Devil River Volcanics. Mantle wedge compositions used are a depleted MORB source mantle wedge (a) and a boninitic mantle source (b). In a first step, variable proportions of a 15% sediment-derived slab melt and a 5% melt derived from altered MORB eclogite (Appendix B; Table 3) were added to the wedge. In a second step, melts were calculated from all these sources, assuming 15% batch melting (Appendix B). Continuous lines are mixing curves between mantle source and variable mixtures of altered MORB and sediment melts. Dotted lines connect identical degrees of slab melts added to the mantle.

 
Using trace element abundances, it has been shown above that the mantle wedge component in the sources of the Devil River Volcanics must be (1) a MORB source mantle (arc rocks) or (2) an even more depleted source (boninites). The above mixing model is therefore expanded in Fig. 11b, where a very depleted (boninitic) mantle source instead of a MORB source is used. The isotope composition of this boninitic mantle source (Table 3) is assumed to be that of the boninites with the most unradiogenic Nd. Mixing of such a mantle source with slab melts having sediment/MORB ratios between 5:95 and 20:80 can explain most compositions of the Devil River Volcanics (Fig. 11b). Calculated amounts of slab melt addition increase from <1% (low-K) to 3% (high-K), slightly lower than average results obtained using a MORB source mantle (Fig. 11a). Variations in mantle wedge composition therefore do not affect the mass balance calculations for the Devil River Volcanics significantly. Hence, the Devil River Volcanics were the mixing product of four components: mantle wedge, sediment melt, MORB-derived slab fluid and MORB-derived slab melt.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
The Devil River Volcanics are derived from heterogeneous mantle sources that originated from mixing of a depleted mantle wedge with components derived from subducted sediment and MORB. Transfer of sediment and MORB components from the subducting slab occurred via separate melt and fluid phases. A quantification of slab contributions to the magma compositions suggests addition of up to 0·4% sediment melt, 3% MORB-derived slab fluid and 6% (3–20%) MORB-derived slab melt to the mantle wedge. The amount of melt derived from the MORB portion of the slab increases through time. Melting of subducted MORB is suggested to require a relatively young and hot subducted oceanic crust (Peacock et al., 1994). Current palaeogeographic reconstructions (e.g. Dalziel et al., 1997) suggest that the opening of the Pacific Ocean pre-dates eruption of the Devil River Volcanics by 100 Ma. Models for equivalents of the Devil River Volcanics in the Lachlan Fold Belt of Australia (e.g. Glen et al., 1997) suggest that, similar to the present western Pacific, several arc chains and spreading ridges might have been present. Subducted oceanic crust might therefore have been substantially younger than 100 Ma, thus providing an ideal scenario for subduction of ‘hot’ oceanic crust.

Systematics of the HFSE in the Devil River Volcanics suggests that this element group did not behave ‘conservatively’ and was in part contributed by slab components. Explanation of HFSE enrichment by an ‘OIB source’ rather than ‘MORB source’ mantle wedge composition can be ruled out by Nb/Th, Nb/Ta and Nb/Yb systematics in the magmas. Combined Nb/Th, Nb/Ta and Nb/Yb ratios are therefore a powerful discriminant between OIB source components and slab components in the sources of present-day arc rocks, in particular for high-K suites. The observed contribution of HFSE by the slab components prevents a precise mass balance of the trace element budget in the Devil River Volcanics. Such mass balances frequently rely on HFSE contents in the volcanic rocks as indicator for mantle wedge depletion and composition (e.g. McCulloch & Gamble, 1991). If the HFSE budget is at least in part controlled by the slab component, it is virtually impossible to constrain the depletion of the mantle wedge, in particular with respect to the highly incompatible trace elements. Mass balance calculations will, under these circumstances, slightly underestimate the amount of added slab component.

	APPENDIX A: ANALYTICAL METHODS

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
Handpicked mineral separates (20–100 mg) were leached in cold 2 N HCl for 15 min and washed in purified H₂O before dissolution. A stronger leaching procedure (e.g. using dilute HF) was specifically avoided. Stronger leaching might cause differential leaching of parent and daughter elements and could result in the wrong calculation of initial isotopic compositions. A duplicate analysis for sample NZ-311, run on leached and unleached splits of one clinopyroxene separate (Table 2), yielded a similar ⁸⁷Rb/⁸⁶Sr ratio. More importantly, recalculated (500 Ma) ⁸⁷Sr/⁸⁶Sr values for both splits lie only slightly outside analytical error. The age-corrected ¹⁴³Nd/¹⁴⁴Nd results (Table 2) also agree well. This shows that the HCl leaching procedure applied here apparently does not cause significant Rb/Sr fractionation in the residue.

Two additional mineral separates (NZ-198, NZ-113b) were handpicked twice and analysed for Rb–Sr in duplicate. Duplicate analyses for sample NZ-198 gave ⁸⁷Sr/⁸⁶Sr at 500 Ma of 0·70501 and 0·70497, which are within analytical error. The duplicate separates of sample NZ-113b differed in their purity. One handpicked separate included 50% milky clinopyroxene, and the other consisted entirely of glassy clinopyroxene. The glassy separate gave (⁸⁷Sr/⁸⁶Sr)₅₀₀ of 0·71158 and is preferred over the milky one, which gave (⁸⁷Sr/⁸⁶Sr)₅₀₀ of 0·71223. This difference is 22 times the external reproducibility, but is none the less negligible in the petrogenetic context discussed below, where variations in (⁸⁷Sr/⁸⁶Sr)₅₀₀ are much larger. Nevertheless, only glassy grains were handpicked from all other mineral separates.

Pb and Sr–Nd isotope analyses were generally performed on different splits of a once homogenized mineral separate or whole-rock powder. All samples were dissolved in HF–HNO₃. From the Pb splits, aliquots of 5 wt % were taken after sample dissolution and separately spiked with ²⁰⁸Pb and ²³⁵U spikes. In contrast, Sr–Nd splits were spiked with mixed ¹⁵⁰Nd–¹⁴⁹Sm and ⁸⁴Sr–⁸⁵Rb spikes without further sample aliquoting (total spiking).

Blanks for Sm and Nd were less than 25 pg and 75 pg, respectively, and are negligible. Rb and Sr blanks were less than 200 pg and 60 pg, respectively. Pb blanks, run in parallel with each batch of samples, were <150 pg (n = 7), excluding two outliers at 218 and 333 pg, and are negligible for the isotope composition (IC) splits. Results of isotope dilution measurements of Pb (5% aliquots of the IC solutions) needed to be corrected for Pb blanks. Blank correction is always <15% of the total amount in the sample and <5% in most cases.

Isotope measurements were performed on a Finnigan MAT 261 mass spectrometer at the Max-Planck-Institut für Chemie in Mainz (White & Patchett, 1984). Sr–Nd results have all been normalized to ⁸⁶Sr/⁸⁸Sr of 0·1194 and ¹⁴⁶Nd/¹⁴⁴Nd of 0·7219. Nd values have been calculated at 500 Ma relative to a present-day CHUR of 0·512638 for ¹⁴³Nd/¹⁴⁴Nd and of 0·1966 for ¹⁴⁷Sm/¹⁴⁴Nd (Jacobsen & Wasserburg, 1980). ⁸⁷Sr/⁸⁶Sr obtained for the NBS 987 standard was 0·710247 ± 29 (2, n = 15). Measurements of the La Jolla standard yielded 0·511850 ± 23 (2, n = 16) for ¹⁴³Nd/¹⁴⁴Nd. Four whole-rock powders of Devil River Volcanics were dissolved and analysed in duplicate; each yielded reproducibilites for ¹⁴³Nd/¹⁴⁴Nd better than 0·000013 (n = 3) and for ⁸⁷Sr/⁸⁶Sr better than 0·000031 (n = 4, Table 2). Measured Pb isotope ratios were corrected for mass fractionation by 1·3 per a.m.u. relative to the values of Todt et al. (1996), based on 18 analyses of the NBS 982 standard (external reproducibility 0·36 per a.m.u., 2). Pb isotope compositions for each mineral separate were analysed in duplicate, with reproducibilities better than 0·36 per a.m.u. and most values below 0·2 per a.m.u. (2 errors).

	APPENDIX B: TRACE ELEMENT AND ISOTOPE MODELLING CALCULATIONS

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
Mantle melting models
Previous melting–depletion models of Pearce & Parkinson (1993) and Woodhead et al. (1993) are modified here, with the following improvements:

1. residues of first-order depletion events are calculated from a primitive mantle composition (Hofmann, 1988) to avoid uncertainties in estimating a depleted MORB source mantle;
   
2. recently determined partition coefficients were used, of which many were obtained by microbeam techniques and are of higher accuracy.
   

Partition coefficients used throughout the modelling calculations are those of Hart & Dunn (1993) for clinopyroxene and Kennedy et al. (1993) for olivine and orthopyroxene. For Nb, a maximum D_mineral/melt of 0·01 was used for olivine and orthopyroxene. Spinel partition coefficients are from Horn et al. (1994) for Nb and Zr, whereas the partition coefficient for Yb is from Nagasawa et al. (1980). The Y partition coefficient for spinel was chosen to be equal to that of Yb. A spinel lherzolitic mantle composition of 60% olivine, 10% orthopyroxene, 25% clinopyroxene and 5% spinel was assumed as starting material. A non-modal melt extraction model was assumed using a composition of 10% olivine, 20% orthopyroxene and 70% clinopyroxene. These compositions were selected according to petrological observations that suggest clinopyroxene is present during extraction of up to 30% melt from peridotite (e.g. Hirose & Kushiro, 1993). In a first step, residues after 5, 10, 15, 20 and 25% melt extraction were calculated. Non-modal batch melting was assumed during the first depletion event using the equation

where C_S is the element concentration in the residue, C₀ is the element concentration in the primitive mantle, D is the bulk partition coefficient for the residue, P is the bulk partition coefficient for the phases entering the melt and F is the degree of melting. In a second step, melts were then extracted from the different residues assuming non-modal accumulated fractional melting, where several melt increments have collected together. The equation

was used, where C_L is the element concentration in the melt.

Melts derived from subducted sediment
Sediment-derived melts were calculated from an average pelagic sediment composition taken from White et al. (1985). Available petrological data (Irifune et al., 1994; Nichols et al., 1994; Plank & Johnson, 1997) suggest an assemblage of clinopyroxene + garnet + amphibole + muscovite + biotite + coesite at the solidus. Partitioning of incompatible elements into coesite is negligible. For clinopyroxene and amphibole, partitioning of REE and Pb can be assumed to be approximately identical (i.e. Hart & Dunn, 1993; Brenan et al., 1995a; LaTourette et al., 1995). A generalized modal sediment composition of 40% clinopyroxene, 40% garnet and 20% biotite [following Stern & Kilian (1996)] is therefore used. Partition coefficients used are those from Hart & Dunn (1993) for clinopyroxene, Johnson (1994) for garnet [Pb from Hauri et al. (1994)] and LaTourette et al. (1995) for phlogopite. Melting is assumed to be accumulated fractional melting in modal mode using the equation

Fluids derived from subducted altered MORB
A realistic calculation of element abundances in slab fluids is not straightforward because fluid–mineral partitioning data are scarce. Furthermore, the possible influences of complexing agents or fluid matrices (e.g. Cl, F, Si) and the physical fluid properties on partitioning are poorly understood at present. Available fluid–mineral or fluid–slab partitioning data (McCulloch & Gamble, 1991; Brenan et al., 1995a; Keppler, 1996; Stalder et al., 1998), however, give surprisingly consistent results for Sr, Pb and REE. Slab–fluid partition coefficients used here (Table 3) are average values derived from the above references. A fluid in equilibrium with average N-MORB (Hofmann, 1988) was calculated (Table 3) using the simplified equation

This equation is used, because the amount of equilibrating MORB remains virtually constant during dehydration (F = 0), thus simplifying equation (3) for partial melting.

Melts derived from subducted altered MORB
Modal compositions of subducted MORB were assumed to be either (1) 50% clinopyroxene and 50% garnet (eclogite), or (2) a 10% garnet amphibolite. Melting of the garnet amphibolite was assumed to be plagioclase free [10% garnet, 90% amphibole; after Drummond & Defant (1990)]. Partition coefficients used were the same as those used for the sediment melts. Amphibole partition coefficients used were taken from LaTourette et al. (1995). Average N-MORB (Hofmann, 1988) was used as starting composition and accumulated fractional melting was assumed [equation (3)].

Mixing endmembers
The depleted MORB source mantle wedge composition used (DMM) is an N-MORB source average from Wood (1979), Sun & McDonough (1989) and depleted lherzolite compositions of Jagoutz et al. (1979). Nd isotope compositions of the MORB source were recalculated at 500 Ma using the depleted mantle values of Goldstein et al. (1984). The Pb isotope composition for the MORB source is taken from the radiogenic end of the 500 Ma MORB field in Fig. 4b, and the composition for the sediment component is calculated by mixture of 30% average Takaka Terrane sediment and 70% of ‘high-Pb’ Archaean source-derived sediment. An average sediment composition is estimated from values given by DePaolo et al. (1982), White et al. (1985), McCulloch (1987), Ellam & Hawkesworth (1988) and Turner et al. (1993). Calculation of an Nd for the subducted sediment is based on Pb mixing proportions in Fig. 4 (70% Takaka Terrane and 30% Archaean source derived sediment). As slab melts, a 15% sediment melt and a 5% MORB melt are used. A ‘boninite source’ mantle wedge composition is estimated using the boninite source with the lowest Nd ( -5). Trace element concentrations of this mantle source were calculated as the source for NZ-330, assuming 15% non-modal batch melting and D and P values of Fig. 7.

	ACKNOWLEDGEMENTS

 
This study was sponsored by the DFG project II C6—Schn 392/2-1 ‘Zur Geodynamik des kambrischen Nordostrandes Gondwanas am Beispiel des neuseländischen Takaka Terrans’. Constructive comments and support by G. Wörner, S. Galer, K. Mezger, A. Schneider and D. Kamenetsky are appreciated. A. W. Hofmann is thanked for the hospitality at MPI Mainz. K. Simon, B. Dietrich and E. Schiffczyk supported laboratory work in Göttingen. T. Funk, N. v. Wolfersdorff, C. Siebert, F. Wombacher, I. Hertel and U. Hinrichs are thanked for laboratory assistance work. The Department of Conservation, New Zealand, provided invaluable support during sampling. Journal reviews by I. Nicholls, G. Yogodzinski and three anonymous reviewers are appreciated. G. Bergantz and J. Davidson are thanked for editorial handling.

	FOOTNOTES

 
^*Present address: Zentrallaboratorium für Geochronologie, Institut für Mineralogie, Corrensstr. 24, 48149 Münster, Germany. e-mail: muenker{at}nwz.uni-muenster.de

	REFERENCES

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING AND PETROLOGY RESULTS DISCUSSION CONCLUSIONS APPENDIX A: ANALYTICAL METHODS APPENDIX B: TRACE ELEMENT... REFERENCES

 
Arculus, R. J. & Powell, R. (1986). Source component mixing in the regions of arc magma generation. Journal of Geophysical Research 91, 5913–5926.

Bowring, S. A., King, J. E., Housh, T. B., Isachsen, C. E. & Podosek, F. A. (1989). Neodymium and lead isotope evidence for enriched early Archean crust in North America. Nature 340, 222–225.

Brenan, J. M., Shaw, H. F., Phinney, D. L. & Ryerson, F. J. (1994). Rutile–aqueous fluid partitioning of Nb, Ta, Hf, Zr, U and Th: implications for high field strength element depletions in island-arc basalts. Earth and Planetary Science Letters 128, 327–339.

Brenan, J. M., Shaw, H. F., Ryerson, F. J. & Phinney, D. L. (1995a). Experimental determination of trace element partitioning between pargasite and a synthetic hydrous andesitic melt. Earth and Planetary Science Letters 135, 1–11.

Brenan, J. M., Shaw, H. F., Ryerson, F. J. & Phinney, D. L. (1995b). Mineral–aqueous fluid partitioning of trace elements at 900°C and 2·0 GPa: constraints on the trace element geochemistry of mantle and deep crustal fluids. Geochimica et Cosmochimica Acta 59, 3331–3350.[Web of Science]

Cooper, R. A. & Grindley, G. W. (1982). Late Proterozoic to Devonian sequences of southeastern Australia, Antarctica and New Zealand and their correlation. Geological Society of Australia Special Publication 9, 90 pp.

Cooper, R. A. & Tulloch, A. J. (1992). Early Paleozoic terranes in New Zealand and their relationship to the Lachlan Fold Belt. Tectonophysics 214, 129–144.

Crawford, A. J. & Berry, R. F. (1992). Tectonic implications of Late Proterozoic–Early Paleozoic igneous rock associations in western Tasmania. Tectonophysics 214, 37–56.[Web of Science]

Crawford, A. J. & Keays, R. R. (1978). Cambrian greenstone belts in Victoria: marginal sea-crust slices in the Lachlan Fold Belt of southeastern Australia. Earth and Planetary Science Letters 41, 197–208.

Crawford, A. J., Falloon, T. J. & Eggins, S. (1987). The origin of high-alumina basalts. Contributions to Mineralogy and Petrology 97, 417–430.

Crawford, A. J., Falloon, T. R. & Green, D. H. (1989). Classification, petrogenesis and tectonic setting of boninites. In: Crawford, A. J. (ed.) Boninites and Related Rocks. Boston, MA: Unwin Hyman, pp. 2–49.

Crawford, A. J., Briqueu, L., Laporte, C. & Hasenaka, T. (1995). Coexistence of Indian and Pacific oceanic upper mantle reservoirs beneath the central New Hebrides island arc. In: Taylor, B. & Natland, J. P. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union 88, 199–217.

Dalziel, I. W. D. (1997). Neoproterozoic–Paleozoic geography and tectonics: review, hypothesis, environmental speculation. Geological Society of America Bulletin 109, 16–42.[Abstract/Free Full Text]

Danyushevski, L. V., Sobolev, A. V. & Falloon, T. J. (1995). North Tongan high-Ca boninite petrogenesis: the role of Samoan plume and subduction zone–transform fault transition. Journal of Geodynamics 20(3), 219–241.

DePaolo, D. J. (1981). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallisation. Earth and Planetary Science Letters 53, 189–202.[Web of Science]

DePaolo, D. J., Manton, W. I., Grew, E. S. & Halpern, M. (1982). Sm–Nd, Rb–Sr and U–Th–Pb systematics of granulite facies rocks from Fyfe Hills, Enderby Land, Antarctica. Nature 298, 614–618.

Drummond, M. S. & Defant, M. J. (1990). A model for trondhjemite–tonalite–dacite genesis and crustal growth via slab melting: Archean to modern comparisons. Journal of Geophysical Research 95, 21503–21521.[Web of Science]

Edwards, C. M. H., Menzies, M. A., Thirlwall, M. F., Morris, J. D., Leeman, W. P. & Harmon, R. S. (1994). The transition to potassic alkaline volcanism in island arcs: the Ringgit–Beser complex, east Java, Indonesia. Journal of Petrology 35, 1557–1595.[Abstract/Free Full Text]

Ellam, R. M. & Hawkesworth, C. J. (1988). Elemental and isotopic variations in subduction related basalts: evidence for a three component model. Contributions to Mineralogy and Petrology 98, 72–80.[Web of Science]

Elliot, T., Plank, T., Zindler, A., White, W. M. & Bourdon, B. (1997). Element transport from subducted slab to volcanic front at the Mariana arc. Journal of Geophysical Research 102(B7), 14991–15019.

Ewart, A. (1982). The mineralogy and petrology of Tertiary–recent orogenic volcanic rocks with special reference to the andesitic–basaltic compositional range. In: Thorpe, R. S. (ed.) Andesites: Orogenic Andesites and Related Rocks. New York: John Wiley, pp. 25–95.

Gill, J. B. (1981). Orogenic Andesites and Plate Tectonics. Berlin: Springer-Verlag.

Glen, R. A., Crawford, A. J. & Scheibner, E. (1997). In the wake of Laurentia: conversion of a Late Proterozoic–Cambrian passive margin into a Cambrian active margin along the southeastern edge of Gondwanaland. In: Bradshaw, J. D. & Weaver, S. D. (eds) Terrane Dynamics 97. Christchurch, New Zealand: Canterbury University, pp. 75–76.

Goldstein, S. L., O’Nions, R. K. & Hamilton, P. J. (1984). A Sm–Nd study of atmospheric dusts and particulates from major river systems. Earth and Planetary Science Letters 70, 221–236.[Web of Science]

Green, T. H. & Pearson, N. J. (1987). An experimental study of Nb and Ta partitioning between Ti-rich minerals and silicate minerals at high pressure and temperature. Geochimica et Cosmochimica Acta 51, 55–62.

Hart, S. R. & Dunn, T. (1993). Experimental cpx/melt partitioning of 24 trace elements. Contributions to Mineralogy and Petrology 113, 1–8.[Web of Science]

Hauri, E. H., Wagner, T. P. & Grove, T. L. (1994). Experimental and natural partitioning of Th, U, Pb and other trace elements between garnet, clinopyroxene and basaltic melts. Chemical Geology 117, 149–166.[Web of Science]

Hawkesworth, C. J., Turner, S., Peate, D., McDermott, F. & van Calsteren, P. (1997). Elemental U and Th variations in island arc rocks: implications for U-series isotopes. Chemical Geology 139, 207–221.

Hildreth, W. & Moorbath, S. (1988). Crustal contributions to arc magmatism in the Andes of central Chile. Contributions to Mineralogy and Petrology 98, 455–489.[Web of Science]

Hirose, K. & Kushiro, I. (1993). Partial melting of dry peridotites at high pressures: determination of compositions of melts segregated from peridotite using aggregates of diamond. Earth and Planetary Science Letters 114, 477–489.[Web of Science]

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.[Web of Science]

Hofmann, A. W. (1997). Mantle geochemistry: the message from oceanic volcanism. Nature 385, 219–229.

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.

Horn, I., Foley, S. F., Jackson, S. E. & Jenner, G. A. (1994). Experimentally determined partitioning of high field strength- and selected transition elements between spinel and basaltic melt. Chemical Geology 117, 193–218.

Hunter, H. W. (1977). Geology of the Cobb Intrusives, Takaka Valley, North-west Nelson, New Zealand. New Zealand Journal of Geology and Geophysics 20, 469–501.

Irifune, T., Ringwood, A. E. & Hibberson, W. O. (1994). Subduction of continental crust and terrigenous and pelagic sediments: an experimental study. Earth and Planetary Science Letters 126, 351–368.

Jacobsen, S. B. & Wasserburg, G. J. (1980). Sm–Nd isotopic evolution of chondrites. Earth and Planetary Science Letters 50, 139–155.[Web of Science]

Jagoutz, E., Palme, H., Baddenhausen, H., Blum, K., Cendales, M., Dreibus, G., Spettel, B., Lorenz, V. & Wänke, H. (1979). The abundances of major, minor and trace elements in the Earth’s mantle as derived from primitive ultramafic nodules. Proceedings of the 10th Lunar and Planetary Science Conference. pp. 2031–2050.

Jenner, G. A., Foley, S. F., Jackson, S. E., Green, T. H., Fryer, B. J. & Longerich, H. P. (1994). Determination of partition coefficients for trace elements in high pressure–temperature experimental run products by laser ablation microprobe–inductively coupled plasma-mass spectrometry (LAM-ICP-MS). Geochimica et Cosmochimica Acta 58, 5099–5103.

Jochum, K. P., Arndt, N. T. & Hofmann, A. W. (1991). Nb–Th–La in komatiites and basalts: constraints on komatiite petrogenesis and mantle evolution. Earth and Planetary Science Letters 107, 272–289.[Web of Science]

Jochum, K. P., Hofmann, A. W. & Stolz, A. J. (1996). Nb and Ta in Ocean Island Basalts. American Geophysical Union Fall Meeting Abstracts. EOS Supplement 77(46), 785.

Johnson, K. T. M. (1994). Experimental cpx/melt and garnet/melt partitioning of REE and other trace elements at high pressures: petrologic implications. Mineralogical Magazine 58A, 454–455.

Kamenetsky, V. S., Crawford, A. J., Eggins, S. & Mühe, R. (1997). Phenocryst and melt inclusion chemistry of near-axis seamounts, Valu Fa Ridge, Lau Basin: insight into mantle wedge melting and the addition of subduction components. Earth and Planetary Science Letters 151, 205–223.

Kelemen, P. B., Shimizu, N. & Dunn, T. (1993). Relative depletion of niobium 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, 111–134.[Web of Science]

Kennedy, A. K., Lofgren, G. E. & Wasserburg, G. J. (1993). An experimental study of trace element partitioning between olivine, orthopyroxene and melt in chondrules; equilibrium values and kinetic effects. Earth and Planetary Science Letters 115, 177–195.[Web of Science]

Kepezhinskas, P. K., Defant, M. J. & Drummond, M. S. (1995). Na metasomatism in the island-arc mantle by slab melt–peridotite interaction: evidence from mantle xenoliths in the North Kamchatka arc. Journal of Petrology 36, 1505–1527.

Keppler, H. (1996). Constraints from partitioning experiments on the composition of subduction-zone fluids. Nature 380, 237–240.

LaTourette, T., Hervig, R. L. & Holloway, J. R. (1995). Trace element partitioning between amphibole, phlogopite and basanite melt. Earth and Planetary Science Letters 135, 13–30.[Web of Science]

McCulloch, M. T. (1987). Sm–Nd isotopic constraints on the evolution of Precambrian crust in the Australian continent. In: Kröner, A. (ed.) Proterozoic Lithospheric Evolution. Geodynamics Series, American Geophysical Union 17, 115–130.

McCulloch, M. T. & Gamble, J. A. (1991). Geochemical and geodynamical constraints on subduction zone magmatism. Earth and Planetary Science Letters 102, 358–374.[Web of Science]

McCulloch, M. T. & Woodhead, J. D. (1993). Lead isotopic evidence for deep crustal-scale fluid transport during granite petrogenesis. Geochimica et Cosmochimica Acta 57, 659–674.

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.

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.

Mueller, P. A., Shuster, R. D., Wooden, J. L., Erslev, E. A. & Bowes, D. R. (1993). Age and composition of Archean crystalline rocks from the southern Madison Range, Montana: implications for crustal evolution in the Wyoming craton. Geological Society of America Bulletin 105, 437–446.[Abstract/Free Full Text]

Münker, C. (1998). Nb/Ta fractionation in a Cambrian arc/back-arc system, New Zealand: source constraints and application of refined ICPMS techniques. Chemical Geology 144, 23–45.

Münker, C. (2000). Pb–Nd isotopes indicate the origin of island arc terranes in the Early Paleozoic Pacific. Journal of Geology (in press).

Münker, C. & Cooper, R. A. (1995). The island arc setting of a New Zealand Cambrian volcano-sedimentary sequence: implications for the evolution of the SW Pacific Gondwana fragments. Journal of Geology 103, 687–700.

Münker, C. & Cooper, R. A. (1999). The Cambrian arc complex of the Takaka Terrane, New Zealand: an integrated stratigraphical, paleontological and geochemical approach. New Zealand Journal of Geology and Geophysics 42(3), 415–446.

Nagasawa, H., Schreiber, H. D. & Morris, R. V. (1980). Experimental mineral/liquid partition coefficients of the rare earth elements (REE), Sc and Sr for perovskite, spinel and melilite. Earth and Planetary Science Letters 46, 431–437.[Web of Science]

Navon, O. & Stolper, E. (1987). Geochemical consequences of melt percolation: the upper mantle as a chromatographic column. Journal of Geology 95, 285–307.[Web of Science]

Nichols, G. T., Wyllie, P. J. & Stern, C. R. (1994). Subduction zone melting of pelagic sediments constrained by melting experiments. Nature 371, 785–788.

Peacock, S. M., Rushmer, T. & Thompson, A. B. (1994). Partial melting of subducting oceanic crust. Earth and Planetary Science Letters 121, 227–244.[Web of Science]

Pearce, J. A. (1983). The role of sub-continental lithosphere in magma genesis at destructive plate margins. In: Hawkesworth, C. J. & Norris, M. J. (eds) Continental Basalts and Mantle Xenoliths. Nantwich: Shiva, pp. 230–249.

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. & Neary, C. R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publication 76, 373–403.

Pearce, J. W. & Peate, D. W. (1995). Tectonic implications of the composition of volcanic arc magmas. Annual Review of Earth and Planetary Sciences 23, 251–285.[Web of Science]

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.[Abstract/Free Full Text]

Peate, D. W., Pearce, J. A., Hawkesworth, C. J., Colley, V., 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.

Plank, T. & Johnson, M. (1997). Experimental study of trace-element partitioning in fluids and melts generated during sediment subduction. In: 7th Annual V. M. Goldschmidt Conference. Houston, TX: Lunar and Planetary Institute, Vol. 921, p. 164.

Plank, T. & Langmuir, C. H. (1988). An evaluation of the global variations in the major element chemistry of arc basalts. Earth and Planetary Science Letters 90, 349–370.

Plank, T. & Langmuir, C. H. (1993). Tracing trace elements from sediment input to volcanic output at subduction zones. Nature 362, 739–743.

Ringwood, A. E. (1974). The petrological evolution of island arc systems. Journal of the Geological Society, London 130, 183–204.[Abstract/Free Full Text]

Stalder, R., Foley, S. F., Brey, G. P. & Horn, I. (1998). Mineral–aqueous fluid partitioning of trace elements at 900–1200°C and 3·0–5·7 GPa: new experimental data for garnet, clinopyroxene, and rutile, and implications for mantle metasomatism. Geochimica et Cosmochimica Acta 62(10), 1781–1801.[Web of Science]

Stern, C. R. & Kilian, R. (1996). Role of the subducted slab, mantle wedge and continental crust in the generation of adakites from the Andean Austral Volcanic Zone. Contributions to Mineralogy and Petrology 123, 263–281.[Web of Science]

Stolper, E. & Newman, S. (1994). The role of water in the petrogenesis of the Mariana trough magmas. Earth and Planetary Science Letters 121, 293–325.

Stolz, A. J., Jochum, K. P., Spettel, B. & Hofmann, A. W. (1996). Fluid and melt related enrichment in the subarc mantle: evidence from Nb/Ta variations in island arc basalts. Geology 24, 587–590.[Abstract/Free Full Text]

Sun, S. S. & McDonough, W. F. (1989). Chemical and isotopic systematics of oceanic basalts. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publication 42, 313–345.

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.

Todt, W., Cliff, R. A., Hanser, A. & Hofmann, A. W. (1996). Evaluation of a ²⁰²Pb–²⁰⁵Pb double spike for high precision lead isotope analysis. In: Basu, R. & Hart, S. R. (eds) Earth Processes: Reading the Isotopic Code. American Geophysical Union Monograph 95, 429–437.

Turner, S., Foden, J., Sandiford, M. & Bruce, D. (1993). Sm–Nd isotopic evidence for the provenance of sediments from the Adelaide Fold Belt and Southeastern Australia with implications for episodic crustal addition. Geochimica et Cosmochimica Acta 57, 1837–1856.

von Huene, R. & Scholl, D. W. (1993). The return of sialic material to the mantle indicated by terrigenous material subducted at convergent margins. Tectonophysics 219, 163–175.

Weaver, S. D., Bradshaw, J. D. & Laird, M. G. (1984). Geochemistry of Cambrian volcanics of the Bowers Supergroup and implications for the Early Paleozoic tectonic evolution of northern Victoria Land, Antarctica. Earth and Planetary Science Letters 68, 128–140.

White, W. M. & Dupré, 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, P. J. (1984). Hf–Nd–Sr isotopes and incompatible element abundances in island arcs: implications for magma origin and crust–mantle evolution. Earth and Planetary Science Letters 67, 167–185.

White, W. M., Dupré, 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.

Wombacher, F. & Münker, C. (1998). Isotope and trace element patterns of Cambrian sediments from New Zealand: a record of island arc and SE-Australian continental provenance. Mineralogical Magazine 62A, 1662–1663.

Wood, D. A. (1979). A variably veined suboceanic upper mantle—genetic significance for mid-ocean ridge basalts from geochemical evidence. Geology 7, 499–503.[Abstract/Free Full Text]

Woodhead, J. D. (1989). Geochemistry of the Mariana arc (western Pacific): source composition and processes. Chemical Geology 76, 1–24.

Woodhead, J. D., Eggins, S. & Gamble, J. (1993). High field strength and transition element systematics in island arc and back-arc basin basalts: evidence for a multiphase melt extraction and a depleted mantle wedge. Earth and Planetary Science Letters 114, 491–504.

Wyllie, P. J. (1982). Subduction products according to experimental prediction. Geological Society of America Bulletin 93, 468–476.[Abstract/Free Full Text]

Yogodzinski, G. M., Kay, R. W., Volynets, O. N., Koloskov, A. V. & Kay, S. M. (1995). Magnesian andesite in the western Aleutian Komandorsky region: implications for slab melting and processes in the mantle wedge. Geological Society of America Bulletin 107, 505–519.[Abstract/Free Full Text]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				S. Schuth, C. Munker, S. Konig, C. Qopoto, S. Basi, D. Garbe-Schonberg, and C. Ballhaus Petrogenesis of Lavas along the Solomon Island Arc, SW Pacific: Coupling of Compositional Variations and Subduction Zone Geometry J. Petrology, May 1, 2009; 50(5): 781 - 811. [Abstract] [Full Text] [PDF]

				B. Dhuime, D. Bosch, C. J. Garrido, J.-L. Bodinier, O. Bruguier, S. S. Hussain, and H. Dawood Geochemical Architecture of the Lower- to Middle-crustal Section of a Paleo-island Arc (Kohistan Complex, Jijal-Kamila Area, Northern Pakistan): Implications for the Evolution of an Oceanic Subduction Zone J. Petrology, March 19, 2009; (2009) egp010v1. [Abstract] [Full Text] [PDF]

				J.-H. ZHAO, R. HU, M.-F. ZHOU, and S. LIU Elemental and Sr Nd Pb isotopic geochemistry of Mesozoic mafic intrusions in southern Fujian Province, SE China: implications for lithospheric mantle evolution Geological Magazine, November 1, 2007; 144(6): 937 - 952. [Abstract] [Full Text] [PDF]

				M. Gutjahr, J. D. Bradshaw, S. Weaver, C. Munker, and T. Ireland Provenance of Cambrian conglomerates from New Zealand: implications for the tectonomagmatic evolution of the SE Gondwana margin Journal of the Geological Society, December 1, 2006; 163(6): 997 - 1010. [Abstract] [Full Text] [PDF]

				N. Jiang Petrology and geochemistry of the Shuiquangou syenitic complex, northern margin of the North China Craton Journal of the Geological Society, January 1, 2005; 162(1): 203 - 215. [Abstract] [Full Text] [PDF]

				Z. GUO, J. HERTOGEN, J. LIU, P. PASTEELS, A. BOVEN, L. PUNZALAN, H. HE, X. LUO, and W. ZHANG Potassic Magmatism in Western Sichuan and Yunnan Provinces, SE Tibet, China: Petrological and Geochemical Constraints on Petrogenesis J. Petrology, January 1, 2005; 46(1): 33 - 78. [Abstract] [Full Text] [PDF]

				null Zhu Yongfeng, Y. ZHU, S. SUN, L. GU, Y. OGASAWARA, N. JIANG, and H. HONMA Permian volcanism in the Mongolian orogenic zone, northeast China: geochemistry, magma sources and petrogenesis Geological Magazine, March 1, 2001; 138(2): 101 - 115. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (29) Request Permissions Google Scholar Articles by MÜNKER, C. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Online ISSN 1460-2415 - Print ISSN 0022-3530

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics