Journal of Petrology

This Article Abstract FREE Full Text (PDF) Supplementary data 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 (23) Request Permissions Google Scholar Articles by YANG, H.-J. Articles by CLAGUE, D. A. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Journal of Petrology | Volume 44 | Number 4 | Pages 603-627 | 2003
© Oxford University Press 2003

Constraints on the Source Components of Lavas Forming the Hawaiian North Arch and Honolulu Volcanics

H.-J. YANG^1,*, F. A. FREY² and D. A. CLAGUE³

¹ DEPARTMENT OF EARTH SCIENCES, NATIONAL CHENG-KUNG UNIVERSITY, TAINAN, TAIWAN 701
² DEPARTMENT OF EARTH, ATMOSPHERIC AND PLANETARY SCIENCES, MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE, MA 02139, USA
³ MONTEREY BAY AQUARIUM, RESEARCH INSTITUTE, MOSS LANDING, CA 95039-0628, USA

Telephone: 011-886-6-2757575, ext. 65429. Fax: 011-886-6-2740285. E-mail: hjyang{at}mail.ncku.edu.tw

RECEIVED NOVEMBER 2, 2001; ACCEPTED SEPTEMBER 30, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
Hawaiian volcanoes, dominantly shields of tholeiitic basalt, form as the Pacific Plate migrates over a hotspot in the mantle. As these shields migrate away from the hotspot, highly alkalic lavas, forming the rejuvenated stage of volcanism, may erupt after an interval of erosion lasting for 0·25–2·5 Myr. Alkalic lavas with geochemical characteristics similar to rejuvenated- stage lavas erupted on the sea floor north of Oahu along the Hawaiian Arch. The variable Tb/Yb, Sr/Ce, K/Ce, Rb/La, Ba/La, Ti/Eu and Zr/Sm ratios in lavas forming the North Arch and the rejuvenated-stage Honolulu Volcanics were controlled during partial melting by residual garnet, clinopyroxene, Fe–Ti oxides and phlogopite. However, the distinctively high Ba/Th and Sr/Nd ratios of lava forming the North Arch and Honolulu Volcanics reflect source characteristics. These characteristics are also associated with shield tholeiitic basalt; hence they arise from the Hawaiian hotspot, which is interpreted to be a mantle plume. Inversion of the batch melting equation using abundances of highly incompatible elements, such as Th and La, requires enriched sources with 10–55% clinopyroxene and 5–25% garnet for North Arch lavas. The ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios in lavas forming the North Arch and Honolulu Volcanics are consistent with mixing between the Hawaiian plume and a depleted component related to mid-ocean ridge basalts. Specifically, the enrichment of incompatible elements coupled with low ⁸⁷Sr/⁸⁶Sr and high ¹⁴³Nd/¹⁴⁴Nd relative to bulk Earth ratios is best explained by derivation from depleted lithosphere recently metasomatized by incipient melt (<2% melting) from the Hawaiian plume. In this metasomatized source, the incompatible element abundances, as well as Sr and Nd isotopic ratios, are controlled by incipient melts. In contrast, the large range of published ¹⁸⁷Os/¹⁸⁸Os data (0·134–0·176) reflects heterogeneity caused by various proportions of pyroxenite veins residing in a depleted peridotite matrix.

KEY WORDS: Hawaiian plume; Honolulu Volcanics; North Arch; plume–lithosphere interaction; rejuvenated stage; trace element geochemistry; alkalic lavas

	INTRODUCTION

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
After a period of quiescence and erosion, ranging from 0·25 to 2·5 Myr, volcanism at many Hawaiian volcanoes terminates with eruption of rejuvenated-stage lavas consisting of alkalic basalts, nephelinites and melilitites. Rejuvenated-stage or ‘post-erosional’ lavas occur on the shields of Haleakala (Hana Volcanics), Koolau (Honolulu Volcanics), Kauai (Koloa Volcanics) and Niihau (Kiekei Basalt). The lavas erupt from vents scattered on the tholeiitic shields.

Submarine lavas compositionally and isotopically similar to rejuvenated-stage lavas have erupted on the sea floor north of Oahu along the Hawaiian Arch (Fig. 1), which is a flexural arch that resulted from the loading of the Hawaiian Islands. This volcanic field covers an area of 25 000 km² at water depths of 3900–4380 m and is referred to as the North Arch Volcanic Field (Clague et al., 1990). The age of North Arch flows ranges from <0·5 to 1·5 Ma (Clague et al., 1990; Dixon et al., 1997; Clague & Dixon, 2000). This age range overlaps with the estimated age of the rejuvenated-stage Honolulu Volcanics on Koolau Volcano (Lanphere & Dalrymple, 1980; Clague & Frey, 1982). Although rejuvenated-stage lavas account for <1% of the mass of each volcano, alkalic lava apparently covers large areas of the Hawaiian Ridge.

View larger version (32K): [in this window] [in a new window]   Fig. 1. (a) Dredge locations (•, with lava flow numbers labelled) on the North Arch Volcanic Field, north of Oahu. Map modified from Clague et al. (1990). (b) Location map for samples from the Honolulu Volcanics located on eastern Oahu. Modified from Clague & Frey (1982).

 
The voluminous Hawaiian shields are generally interpreted as surface manifestations of the Hawaiian plume. However, the origin of the alkalic rejuvenated-stage and North Arch lavas is debated (Chen & Frey, 1983, 1985; Feigenson, 1984; Clague & Dalrymple, 1988). Compared with the dominantly tholeiitic Hawaiian shields, the rejuvenated-stage and North Arch lavas have higher ¹⁴³Nd/¹⁴⁴Nd and lower ⁸⁷Sr/⁸⁶Sr. This result is surprising in that, despite their relatively low Sm/Nd and high Rb/Sr, these alkalic lavas were derived from sources with higher time-averaged Sm/Nd and lower time-averaged Rb/Sr than the sources of the shield lavas. Two types of mixing mechanisms have been proposed to explain the paradox between Nd and Sr isotopic ratios and their corresponding parent/daughter (Sm/Nd and Rb/Sr) ratios. Chen & Frey (1983, 1985) showed that the inverse correlations between Rb/Sr and ⁸⁷Sr/⁸⁶Sr and Sm/Nd vs ¹⁴³Nd/¹⁴⁴Nd defined by the shield to rejuvenated-stage lavas of Haleakala volcano could be explained by mixing between incipient melts (<2%) from a depleted source related to mid-ocean ridge basalt (MORB) and melts from the mantle plume. The essential part of this model is that incipient melts of the depleted source inherit the isotopic ratios of the source, but they will have much higher Rb/Sr and lower Sm/Nd owing to low extents of melting. A similar mixing model was proposed for the generation of rejuvenated-stage lavas (Koloa Volcanics) on Kauai (Feigenson, 1984; Clague & Dalrymple, 1988).

An alternative mixing model was used to explain the compositional range and limited isotopic variations of the Honolulu Volcanics. Roden et al. (1984) proposed that the source of the Honolulu Volcanics formed by adding small amounts (2·5%) of melt derived from a MORB source by 0·3% melting to the plume source. Melting of this mixed source to varying extents then generated the Honolulu Volcanics. A similar model was proposed by Reiners & Nelson (1998) for the Koloa Volcanics on Kauai. In both types of mixing models garnet lherzolite is the major source for the magmas. Based on major and trace element compositions, volatile contents and limited isotopic data, a similar model has been proposed for the source of the North Arch lavas (Clague et al., 1990; Dixon et al., 1997; Frey et al., 2000). Specifically, the North Arch lavas were generated by variable but low extents of melting of a homogeneous garnet lherzolite.

In contrast, Lassiter et al. (2000) argued that Os isotopic ratios in rejuvenated-stage lavas of the Koloa and Honolulu Volcanics cannot be explained by plume–MORB source interaction, but that they are consistent with mixing between melts derived from lherzolite and pyroxenite forming the oceanic lithospheric mantle. This model differs significantly from previous models in two aspects: (1) the Hawaiian plume makes no contribution to the generation of rejuvenated-stage lavas; (2) an important role for pyroxenite is emphasized.

Discussions of the petrogenesis of Hawaiian rejuvenated-stage lavas, especially inversion approaches using trace element abundance data and melting models (i.e. Watson, 1993; Zou & Zindler, 1996; Sims & DePaolo, 1997), have relied heavily on the trace element compositions of the Honolulu Volcanics (Clague & Frey, 1982). However, data for some elements reported by Clague & Frey (1982) are not as precise and accurate as data obtained by current analytical techniques, and data for some elements, such as Nb, Y, Rb, Pb and U, are unavailable for most samples. In this paper, we present trace element data measured by inductively coupled mass spectrometry (ICP-MS) for 14 Honolulu Volcanics lavas and 21 North Arch samples, which were previously studied (Clague & Frey, 1982; Frey et al., 2000). Our goal is to compare the petrogenesis of lavas forming the North Arch and Honolulu Volcanics. In particular, we: (1) use variations in abundance of incompatible elements to constrain mineral proportions in the source; (2) use abundance ratios of incompatible elements with Sr and Os isotopic data to investigate the role of a plume component in the North Arch and rejuvenated- stage lavas; (3) evaluate the evidence for a pyroxenite source component.

	SAMPLING

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
A subset of previously analyzed lavas from the North Arch [Fig. 1a; 21 of 23 samples studied by Frey et al. (2000)] and Honolulu Volcanics [Fig. 1b; 14 of 32 samples studied by Clague & Frey (1982)] were selected for trace element analyses. The 14 samples from the Honolulu Volcanics comprise two alkali basalts, three basanites, five nephelinites and four nepheline melilitites. North Arch lavas are glassy or fine grained. Based on SiO₂ content, we classify all of the analyzed North Arch samples as alkali basalt and basanite.

	ANALYTICAL METHODS

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
The sample solutions were analyzed for Rb, Ba, Sr, Y, Nb, Ta, Zr, Hf, Pb, Th, U and 14 rare earth elements (REE) by ICP-MS at the Massachusetts Institute of Technology. For each sample, 100 mg of powder was digested using 3·5 ml HF (24N) and 0·5 ml HNO₃ (7N) at 250°F in a Teflon Savillex beaker for 48 h. The sample was then heated to dryness and fluxed with 6N HCl for 24 h. The HCl solution was then taken to dryness and converted to nitrate form using 3 ml of 7N HNO₃ and heated to dryness. The dried sample cake was then dissolved in 5 ml of 7N HNO₃, which was subsequently diluted to 250 ml in 2N HNO₃. US Geological Survey (USGS) standard samples BHVO-1, BCR-1, and AGV-1 were used to establish calibration curves. Additional details have been given by Huang & Frey (2003). The precision and accuracy of our analysis are addressed in the following section.

	RESULTS

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
Abundances of 25 trace elements are listed in Tables 1 and 2 together with the mean of 11 BHVO-2 (a USGS standard) replicates, which were analyzed during the course of this study (Huang & Frey, 2003). The complete major and trace element dataset may be downloaded from the Journal of Petrology website at http://www.petrology.oupjournal.org. Based on these replicates, the analytical precisions are 2% (1 SD) for most elements (Table 1). Duplicate analyses of eight North Arch samples agree within 3% for most elements. Relative to the previous instrumental neutron activation (INAA) data (Frey et al., 2000), the most significant improvement is in precision of the Th data; 2–3% for ICP-MS (Table 1) compared with 10–12% for INAA (Frey et al., 1990; Yang et al., 1996). Except for Th, the relative agreement of data obtained by INAA and ICP-MS is generally good (Appendix A). As expected, there are more significant discrepancies between the ICP-MS data and the older data for the Honolulu Volcanics reported by Clague & Frey (1982) (see Appendix A). To avoid systematic differences between data obtained by different analytical techniques, we use the new ICP-MS data to evaluate previous interpretations and to provide new constraints on the petrogenesis of the North Arch and Honolulu Volcanics lavas.

View this table: [in this window] [in a new window]   Table 1: Abundances of trace elements (ppm) analyzed with ICP-MS for lavas from the Hawaiian North Arch

 

View this table: [in this window] [in a new window]   Table 2: Abundances of trace elements (ppm) analyzed with ICP-MS for lavas from the Honolulu Volcanics, Hawaii

 
Incompatible element abundances in North Arch lavas
The data show that the abundances of highly and moderately incompatible elements, such as Th, U, light and middle rare earth elements (LREE and MREE), high field strength elements (HFSE), K, Rb, Ba and Sr, are positively correlated (Fig. 2a). Two samples from dredge 24 (24-3 and 24-6) have relatively lower TiO₂, Zr, Hf, K₂O and Rb concentrations compared with other samples with similar Th contents (Fig. 2a). The abundance of the highly incompatible element Th varies by almost a factor of three, whereas that of heavy rare earth elements (HREE) varies over a small range (e.g. a factor of 1·34 for Yb in Fig. 2a). Therefore, HREE were not highly incompatible during the petrogenesis of North Arch lavas, if these samples were derived from a common source as indicated by the similarities in their Sr, Nd and Pb isotopic ratios (Frey et al., 2000). On a primitive mantle normalized diagram, all the analyzed North Arch samples are depleted in Zr, Hf and K and enriched in Sr and Ba relative to their neighboring elements (Fig. 3a). Some samples also have small negative Ti anomalies and sample 34-2 shows a large positive Pb anomaly that is confirmed by duplicate analyses.

View larger version (49K): [in this window] [in a new window]   Fig. 2. (a) Abundances of La, Sm, Yb, TiO2, Nb, Zr, K2O, Rb, Ba and Sr vs Th content for samples from the Hawaiian North Arch Volcanic Field (all in ppm, except for TiO2 and K2O, which are in wt %). Two labelled samples are relatively depleted in TiO2, Zr, K2O and Rb. (b) Abundances of La, Sm, Yb, TiO2, Nb, Zr, K2O, Rb, Ba and Sr vs Th content for samples from the Honolulu Volcanics (all in ppm, except for TiO2 and K2O, which are in wt %). The labelled samples in the Th vs TiO2 and Nb and Zr plots are the ‘low-TiO2 group’ identified by Clague & Frey (1982), except for sample GMQ-9.

 

View larger version (47K): [in this window] [in a new window]   Fig. 3. Incompatible element abundances normalized to primitive mantle values (Sun & McDonough, 1989) for samples from the Hawaiian North Arch (a) and Honolulu Volcanics (b). In (a), representative samples are shown to cover the ranges for each of the three groups. In (b), all ICP-MS data are plotted.

 
Incompatible element abundances in Honolulu Volcanics lavas
The abundance of highly incompatible elements, such as Th, increases from alkali basalt to basanite to nephelinite to nepheline melilitite. For simplicity in the following discussion, alkali basalt and basanite refer to the four samples with lowest abundance of incompatible elements (<5 ppm Th in Fig. 2b). As in North Arch lavas, the abundance of Th is positively correlated with that of LREE, MREE, Ba and Sr. However, the Honolulu Volcanics extend to higher concentrations of incompatible elements (Fig. 2a and b). Although somewhat scattered, the abundance of Yb also increases with increasing Th content (Fig. 2b); this trend contrasts with the absence of a Yb–Th correlation in the North Arch lavas (Fig. 2a). Relative to trends for the North Arch lavas, plots of Th vs HFSE for the Honolulu Volcanics are more scattered with poorly defined positive trends. Four of our 14 analyzed samples have relatively lower TiO₂ contents at a given Th content. They belong to the ‘low-TiO₂ group' identified by Clague & Frey (1982) (Fig. 2b). The ‘low-TiO₂ group’ samples also have relatively low abundances of Zr and Nb at a given Th content, although the separations are not as obvious as in the TiO₂–Th plot (Fig. 2b). Sample GMQ-9 from vent 37 is distinct from other samples in its relatively high abundance of Sm, HFSE, Rb, K and Sr and low abundance of HREE (Fig. 2b), confirming the conclusion of Clague & Frey (1982, p. 486) that lava erupted from this vent has distinctive geochemical characteristics. Sample 66PY-1 is characterized by high HREE contents and sample 68KEE-1 has high Ba and low Sr relative to the general trends (Fig. 2b). When normalized to primitive mantle values, four samples of alkali basalt and basanite have relative depletions in Zr and Hf and enrichments in Ba and Sr, similar to the patterns of North Arch lavas (Fig. 3a). In addition to Zr and Hf depletions, nine nephelinites and nepheline melilitites also have negative Ti anomalies but their relative Sr abundances are variable (Fig. 3b).

	DISCUSSION

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
Compositional variations in lavas: source or process control
In a suite of primary lavas with limited variability in isotopic ratios and significant compositional variation, such as North Arch (Frey et al., 2000; Kani et al., 2000) and Honolulu Volcanics (Lassiter et al., 2000), the effects of variable extents of melting can be inferred by plotting abundance ratios involving incompatible elements against the concentration of a highly incompatible element. A constant abundance ratio over a wide range of incompatible element concentration indicates the source ratio (Hofmann et al., 1984), whereas a systematic increase or decrease in abundance ratio reflects the effect of the melting process, specifically the residual minerals that control the bulk-solid/melt partition coefficients (D). For the North Arch lavas, olivine is the only phenocryst, except for three samples, 15-1, 18-8 and 36, which also contain augite phenocrysts (Clague et al., 1990; Clague & Dixon, 2000). However, these three samples have CaO/Al₂O₃ ratios within the range of other North Arch lavas, indicating that their compositions were not affected by crystallization of augite; therefore, olivine was the only significantly fractionated silicate phase. Dixon et al. (1997) and Frey et al. (2000) inferred the compositions of primary melts for North Arch lavas by incrementally adding calculated equilibrium olivine [K_D = (Fe/Mg)_ol/(Fe/Mg)_melt = 0·3, using measured Fe₂O₃/FeO ratios] to the glass and whole-rock composition, respectively, until the lava composition is in equilibrium with Fo₉₁. Because the Honolulu Volcanics are also inferred to be near-primary magmas (Clague & Frey, 1982), we used the same approach for these lavas.

North Arch lavas
Based on the variations in abundances of major and trace elements and near isotopic homogeneity over the compositional range, Frey et al. (2000) proposed that North Arch lavas were derived from garnet lherzolite; phlogopite and amphibole were not residual minerals except for four samples with distinctively lower K/Ce ratios. We evaluate these inferences using the new ICP-MS data.

Garnet control on compositional variations. North Arch lavas, as a whole, define trends in Th vs Tb/Yb, Sr/Ce and Sr/Nd plots (Fig. 4). The positive Tb/Yb–Th correlation (Fig. 4a) requires that bulk solid/melt D_Tb < D_Yb. Because the experimentally determined D_Tb/D_Yb ratios for clinopyroxene/melt, amphibole/melt and phlogopite/melt partitioning are near unity (e.g. Hart & Dunn, 1993; Hauri et al., 1994; LaTourrette et al., 1995; Blundy et al., 1998; Lundstrom et al., 1998; Dalpé & Baker, 2000), these phases cannot effectively fractionate Tb from Yb. In contrast, it is well established the D_Tb/D_Yb ratios for garnet/melt partitioning are <1 (Philpotts et al., 1972; Shimizu & Kushiro, 1975; Johnson, 1998; van Westrenen et al., 2000). For example, the recent garnet/melt partitioning experiments of van Westrenen et al. (2000) show D_Tb/D_Yb ratios as low as 0·2. Although was not measured in the experiments of Johnson (1998), it is inferred to be 1·8 based on their and values, resulting in a of 0·24, consistent with the results of van Westrenen et al. (2000). Therefore, Tb is more incompatible than Yb in garnet and the positive Tb/Yb–Th correlation reflects the control of residual garnet.

View larger version (49K): [in this window] [in a new window]   Fig. 4. (a) Abundance ratios of incompatible elements vs the concentrations of Th and Ce adjusted to be in equilibrium with Fo91 for North Arch lavas. North Arch samples are classified into three groups (see discussion in text). The continuous lines indicate mixing between most enriched Group B and least enriched Group A samples. Each tick on the mixing lines indicates a 20% interval. PM indicates primitive mantle value. The error bars indicate 1 uncertainty. For panels with no error bars, the uncertainty is smaller than symbol size. (b) Abundance ratios of Ti/Eu, Zr/Sm, Zr/Hf and Zr/Tb vs the Th content adjusted to be in equilibrium with Fo91 for North Arch lavas. The symbols are as in (a).

 
In contrast to the positive Tb/Yb–Th correlation, Sr/Ce and Sr/Nd ratios are inversely correlated with Th contents (Fig. 4a) indicating that D_Sr > D_Ce and D_Nd. These trends do not reflect the effects of residual garnet which has D_Sr < D_Ce and D_Nd (e.g. Hauri et al., 1994; Johnson, 1998; van Westrenen et al., 2000) or residual clinopyroxene which has D_Sr D_Ce < D_Nd (e.g. Blundy et al., 1998, fig. 3). The inverse Sr/Ce and Sr/Nd vs Th trends are primarily defined by the relatively low Sr/Ce and Sr/Nd ratios of samples with Th contents >3 ppm (Fig. 4a). The high compatibility of Sr relative to Ce and Nd is a characteristic of phlogopite and amphibole (LaTourrette et al., 1995; Dalpé & Baker, 2000). The role of phlogopite and amphibole is evaluated in the next section using abundance ratios involving K, Rb and Ba.

The effects of residual K-bearing minerals on K, Rb and Ba. Residual amphibole and phlogopite are major hosts for K, Rb and Ba, which can result in low K/Ce, Rb/La and Ba/La ratios in equilibrium melts (e.g. Class & Goldstein, 1997). In contrast to the trends in the Th–Tb/Yb and Th–Sr/Ce plots (Fig. 4), Frey et al. (2000) noted the absence of a systematic variation in K/Ce vs Ce plot for the North Arch lavas and suggested that a K-bearing phase is not a residual phase, except for four samples with distinctively lower K/Ce ratios. After the effect of olivine fractionation is removed, 11 low-Ce samples (referred to as Group A) range to lower K/Ce than the six high-Ce samples (referred to as Group B) (Fig. 4a). Moreover, Group A and B lavas define positive and negative slopes, respectively, in Th vs Rb/La, Nb/La and Ba/La plots (Fig. 4a). The other four samples, 22-2, 24-3, 24-6 and 34-2 (referred to as Group C), have distinctively lower K/Ce, Rb/La and Nb/La ratios at a given Ce or Th content (Fig. 4a).

In Group B lavas, the increases in K/Ce, Rb/La and Ba/La ratios with increasing extent of melting, as indicated by decreasing Th content (Fig. 4a), are consistent with control by residual K-bearing minerals such as phlogopite or amphibole (Francis & Ludden, 1995; Class & Goldstein, 1997). This inference is consistent with their relatively lower Sr/Nd and Sr/Ce ratios (Fig. 4a). In contrast, these ratios decrease in Group A lavas with increasing extent of melting (Fig. 4a), thereby indicating that K, Rb and Ba were more incompatible than LREE during partial melting. Therefore, K-bearing minerals were not residual phases during generation of these lavas and the relative K depletion and high Ba/Th ratios (>110) of Group A lavas are source characteristics (Fig. 3a). The opposite slopes defined by Group A and B lavas in Th vs Rb/La, Ba/La and Nb/La plots suggest that these two groups are not related by magma mixing (Fig. 4a). However, the two groups could be explained by variable extents of partial melting of a common source with K-bearing residual minerals. That is, Group B lavas were derived from relatively low extents of melting with residual K-bearing minerals and Group A lavas were generated by relatively higher extents of melting that exhausted K-bearing minerals. This inference is consistent with the limited Sr, Nd, and Pb isotopic variations in these lavas (Fig. 5 and Frey et al., 2000; Kani et al. 2000).

View larger version (21K): [in this window] [in a new window]   Fig. 5. 206Pb/204Pb vs 87Sr/86Sr plot for North Arch lavas. , • and circle with a cross are dredged samples analyzed by Frey et al. (2000), which are classified into Group A, B and C, respectively, in this study. and are samples collected by Dive 6K502 and 6K503 of Shinkai 6500 submersible in 1999 (Kani et al., 2000). The error bars are indicated as ±2 (Frey et al., 2000; Kani et al., 2000). Dive 6K502 samples overlap with Group A, B and C samples within uncertainties. However, Dive 6K503 samples extend to lower 87Sr/86Sr and 206Pb/204Pb values.

 
Compared with Group A and B samples, the four Group C samples have lower abundances of alkali metals relative to LREE and Th (Fig. 4a). Three of the four have relatively lower Ti/Eu, Zr/Sm and Zr/Tb ratios at a given Th content (Fig. 4b). Moreover, two Group C samples, 24-3 and 24-6, deviate from the general Th–X correlations (Fig. 2a). Consequently, Group C samples are distinct from other samples in their relative depletions of alkali metals, Ti, Zr and Hf. These complexities are likely to reflect source characteristics, although the only Group C sample analyzed for isotopes has ⁸⁷Sr/⁸⁶Sr, ¹⁴³Nd/¹⁴⁴Nd and ²⁰⁶Pb/²⁰⁴Pb ratios that overlap with those of Group A samples (Fig. 5).

Source mineralogy constrained by the compatibility of Zr. Group A and B lavas also differ in the compatibility of Zr. Specifically, Group A lavas define a negative Zr/Sm–Th trend (Fig. 4b) and have relatively constant Zr/Tb ratios, implying D_Sm < D_Zr D_Tb. In contrast, Group B lavas define a positive Zr/Tb–Th correlation and have rather uniform Zr/Sm ratios, suggesting D_Sm D_Zr < D_Tb (Fig. 4b). Therefore, the relative compatibility of Zr and Sm is very different in Group A and Group B lavas; hence the Zr depletion in the primitive mantle normalized plot is probably caused by the melting process (Fig. 3a). However, D_Sm < D_Zr is not a characteristic of either residual clinopyroxene (e.g. Hart & Dunn, 1993; Hauri et al., 1994; Blundy et al., 1998; Lundstorm et al., 1998) or amphibole (LaTourrette et al., 1995; Dalpé & Baker, 2000). Fe–Ti oxides, such as ilmenite and rutile, have high D_Zr/D_Sm; i.e. is 0·3 (McCallum & Charette, 1978; Pearce & Norry, 1979; McKay et al., 1986), significantly higher than (<0·01 based on McKay et al., 1986; Nielsen et al., 1992) and is 300 (Foley et al., 2000, fig. 1). Also, in contrast to the typically chondritic Ti/Eu in MORB and ocean island basalt (OIB) (e.g. Garcia et al., 1996; Rhodes & Hart, 1996; Yang et al., 1996; Niu et al., 1999), North Arch lavas have Ti/Eu ratios slightly lower than the primitive mantle ratio of 7738 (Fig. 4b). In addition to Fe–Ti oxides, experimentally produced garnet with grossularite content >20% have D_Sm < D_Zr (Hauri et al., 1994; van Westrenen et al., 1999). Compared with Group B lavas, Group A lava could be explained by derivation from a source with more Fe–Ti oxides or Ca-rich garnet. If Group A and B lavas were derived from a common source, this inference requires that the amounts of residual garnet or Fe–Ti oxides increase as extent of melting increases. This is inconsistent with the experimentally determined melting stoichiometry showing that garnet preferentially enters melts (Walter et al., 1995; Walter, 1998). The behavior of Fe–Ti oxides during partial melting is not well constrained; however, spinel, a ubiquitous oxide phase in the shallow mantle, decreases in abundance as melting extent increases (Kinzler, 1997). If both garnet and Fe–Ti oxides preferentially enter the melt, the greater compatibility of Zr in Group A lavas which were derived by larger extents of melting than Group B lavas is inconsistent with derivation of Group A and B lavas from a common source. Despite the isotopic similarity of these groups, we infer that the sources for these lavas differed in their mineral proportions.

Honolulu Volcanics
Clague & Frey (1982) concluded that lavas forming the Honolulu Volcanics were derived from a garnet lherzolite mantle source. They also inferred that K-bearing minerals and Ti oxides were important residual minerals whose proportions varied in the source. Here we use our new trace element abundance data to evaluate and augment these conclusions.

Evidence for residual garnet. We find a positive correlation between the abundance of Th and Tb/Yb ratio (Fig. 6a) confirming the control of residual garnet, consistent with the conclusion of Clague & Frey (1982). Compared with North Arch lavas the Honolulu Volcanics define a lower slope in Tb/Yb vs Th plot (Fig. 6a), thereby reflecting a smaller proportion of garnet in the source of the Honolulu Volcanics. This inference can also be readily derived from the Th–Yb variation; the concentration of Yb in the Honolulu Volcanics is positively correlated with Th content (Fig. 2b), whereas Yb abundance in North Arch lavas is buffered to within 30% (Fig. 2a).

View larger version (66K): [in this window] [in a new window]   Fig. 6. (a) Abundance ratios of incompatible elements vs the concentration of Th adjusted to be in equilibrium with Fo91 for samples from the Honolulu Volcanics. Alkalic basalt and basanite have adjusted Th contents <5 ppm, whereas nephelinite and nepheline melilitite have adjusted Th contents >5 ppm. The shaded fields indicate Group A and B North Arch lavas. The circle with a cross indicates North Arch Group C lavas. The error bars indicate 1 uncertainty. For panels with no error bars, the uncertainty is smaller than symbol size. (b) Ti/Eu, Zr/Sm, Nb/La, Nb/U and Zr/Hf ratios vs Th content adjusted to be in equilibrium with Fo91 for samples from the Honolulu Volcanics. Symbols are as in (a). The labelled samples, except for GMQ-9, are the low-Ti samples identified by Clague & Frey (1982). The curve represents mixing between samples with highest and lowest Th contents. The ticks on the mixing line indicate 20% interval.

 
Evaluating the role of phlogopite. Both K and Rb are mobile during subaerial alteration of Hawaiian lavas, with Rb being more mobile than K, thereby leading to unusually high K/Rb (e.g. Feigenson et al., 1984). However, among the Honolulu Volcanics the strong correlation between K and Rb, with a nearly constant ratio of 317 ± 26 for 12 of 14 analyzed samples, suggests that K and Rb contents in these lavas were not significantly affected by alteration. Alkali basalt 65KPO-1 has an unusually high K/Rb ratio of 543 and a relatively low Rb/La and K/Ce ratio compared with samples with similar Th contents (Fig. 6a), possibly reflecting Rb and K loss during alteration. Basanite 65PAL-2 has a low K/Rb ratio of 229, which might also reflect alkali mobility.

The broad negative correlations in plots of Th vs Sr/Ce, Sr/Nd, K/Ce, Rb/La, Ba/La and Ba/Th ratios (Fig. 6a) are consistent with control by residual phlogopite and/or amphibole (LaTourrette et al., 1995; Dalpé & Baker, 2000). If the Honolulu Volcanics were derived from a common source as indicated by their homogeneous Sr and Nd isotopic ratios (Roden et al., 1984; Lassiter et al., 2000), the large variation in K/Ce, Rb/La and Ba/La ratios, by factors of 2–3, can be explained by residual phlogopite which has D_K,Rb,Ba/D_LREE,Th ratios >85 (LaTourrette et al., 1995). Although amphibole also has D_K,Rb,Ba > D_LREE,Th, its D_K,Rb,Ba/D_LREE,Th ratios (<10) are not large enough to compensate the effect of clinopyroxene which has D_K,Rb,Ba/D_LREE,Th <0·1 (e.g. Hart & Dunn, 1993; Lundstrom et al., 1998) and is more abundant than K-bearing minerals. The low abundances of incompatible elements and high K/Ce, Rb/La, Ba/La and Ba/Th ratios in alkali basalt and basanite (Th < 4 ppm) indicate that these lavas were derived from relatively larger extents of melting with no residual phlogopite. Therefore, like Group A North Arch lavas, the relative depletion in K and enrichment in Ba of these lavas (Fig. 3b) reflect source characteristics. Although some alkali basalts and basanites of the Honolulu Volcanics overlap with North Arch lavas in Th vs K/Ce, Rb/La and Rb/Ce plots, they are distinct from North Arch lavas in their higher Ba/La, Ba/Th and Sr/Th ratios at a given [Th]_Fo91 (Fig. 6a).

Variations of HFSE: implications on source characteristics. The Ti/Eu, Zr/Sm, Nb/La and Nb/U ratios in the Honolulu Volcanics are inversely correlated with Th content and vary much more than in the North Arch lavas; e.g. Zr/Sm, which is similar to the primitive mantle ratio in most oceanic basalts (Sun & McDonough, 1989), varies by a factor of 2·3 in the Honolulu Volcanics (Fig. 6b). At a given Th content the low-TiO₂ group has particularly low ratios of Ti/Eu, Zr/Sm and Nb/U (Fig. 6b). The high-Th melts, inferred to have formed by the lowest extents of melting, also have prominent negative Ti anomalies in a primitive mantle normalized diagram (Fig. 3b) consistent with control by residual Fe–Ti oxides. With high D_Ti/D_Eu, D_Zr/D_Sm, D_Nb/D_U and D_Nb/D_La (McCallum & Charette, 1978; Pearce & Norry, 1979; McKay et al., 1986; Nielsen et al., 1992), residual Fe–Ti oxides also result in decreasing Ti/Eu, Zr/Sm, Nb/La and Nb/U ratios with decreasing melting extent (Fig. 6b) and relative depletion in Zr and Hf (Fig. 3b).

The Zr/Hf ratio and Th content are positively correlated. The samples derived from largest extent of melting have Zr/Hf ratios approaching the chondritic value (Fig. 6b), supporting the interpretation that high Zr/Hf ratios in OIB reflect the melting process (David et al., 2000). The high Zr/Hf ratio at low extents of melting may be a consequence of equilibrating with Fe–Ti oxides. Clinopyroxene, however, has D_Zr < D_Hf (e.g. Hart & Dunn, 1993; Hauri et al., 1994; Blundy et al., 1998; Lundstrom et al., 1998; Salters & Longhi, 1999); therefore at low extents of melting magmas in equilibrium with clinopyroxene will have relatively high Zr/Hf.

Alkali basalt and basanite, however, also show prominent Zr and Hf depletions in a primitive mantle normalized diagram (Fig. 3b). Although they do not have negative Ti anomalies, three of the four have Ti/Eu <6500, lower than the primitive mantle value (7738). These lavas, like North Arch Group A lavas, may be derived from a source containing small amounts of Ti oxides or Ca-rich garnet. However, the alkali basalt and basanites of the Honolulu Volcanics have Nb/U ratio 50, overlapping with the field for North Arch lavas (Fig. 6b). Such values are typical for OIB and MORB (Hofmann et al., 1986). Hence both Nb/U and Zr/Hf in the source of the North Arch and Honolulu Volcanics are similar to the sources of other OIB, and the variable ratios in the lavas were caused by the melting process.

The unusual composition of sample GMQ-9. Sample GMQ-9 collected from vent 37 is compositionally similar to another sample from vent 37, 65MOIL-2, studied by Clague & Frey (1982). Compared with other Honolulu Volcanics lavas, these two samples have lower HREE and Sc contents and higher HFSE, Sr and Rb contents (Table 2, Fig. 4; Clague & Frey, 1982). Given the similarity in Sr and Nd isotopic ratios between samples from vent 37 and from other vents of Honolulu Volcanics (Roden et al., 1984), a different residual mineralogy is required for vent 37 lavas (Clague & Frey, 1982).

Summary
Variations in abundance ratios of incompatible elements in lavas forming the North Arch and Honolulu Volcanics (Figs 4 and 5) were primarily controlled by residual phases such as garnet (causing variable Tb/Yb), clinopyroxene (causing variable Zr/Hf), Ti oxides (causing variable Zr/Sm, Nb/La, Ti/Eu and Nb/U) and phlogopite (causing variable K/Ce, Rb/La, Sr/Ce, Sr/Nd and Ba/La). In contrast, some distinctive features of these lavas, such as very high Ba/Th (Figs 3 and 6a), the small positive Sr anomalies in primitive mantle normalized diagrams (Fig. 3) and the negative K anomalies in North Arch Group A (Fig. 3a) reflect source characteristics.

Source characteristics inferred from batch melting inversion
Assessment of inversions of batch melting equation
Batch melting is an end-member model that may be appropriate for alkalic lavas formed by relatively small extents of melting. Clague & Frey (1982) used the inversion of the batch melting equation to constrain bulk-solid/melt partition coefficients of element i () by plotting the concentration of an incompatible element (i.e. Th) vs ratios of element abundances (i.e. Th/La) (see Appendix B for the derivation of equations). Based on the calculated , they further inferred that the source of Honolulu Volcanics is garnet peridotite. However, Sims & DePaolo (1997) pointed out that this approach has statistical limitations as a result of (1) the small variation in the abundance ratio x/y relative to the large range of concentration x and (2) both axes depending upon x. To overcome these limitations, Sims & DePaolo (1997) proposed to constrain D₀ⁱ by performing linear regression using two statistically independent variables; 1/x vs 1/y (Appendix B).

Another model for partial melting is fractional melting. When fractional melts are pooled to form an accumulated fractional melt, the resulting melt composition is similar to a batch melt (e.g. Langmuir et al., 1992). Figure 7 shows model 1/Th vs 1/La and 1/Sm trends for batch and accumulated fractional melts for the range of 1–22% melting. Over this range the trends for both melting models are very similar for 1/La vs 1/Th whereas for 1/Sm vs 1/Th, they differ significantly at higher extents of melting. Therefore, for inversion calculations we emphasize highly incompatible elements because they are insensitive to melting models. However, using this approach to constrain residual phases requires robust information on the partition coefficients (D) of these elements. At this time, the experimentally determined D values for highly incompatible elements in mantle phases vary by large factors, more than a factor of 10 [e.g. the difference in between Blundy et al. (1998) and Salters & Longhi (1999)]. The effect of this uncertainty on the estimated mineral proportions is addressed.

View larger version (16K): [in this window] [in a new window]   Fig. 7. (a) Modelled 1/Th vs 1/Sm, and (b) modelled 1/Th vs 1/La plots. The abundances of Th, La and Sm in melts are calculated from non-modal batch melting () and non-modal fractional melting (; accumulated melts) with melting extents varying from 2 to 22%. The model parameters are from Clague & Frey (1982, table 5). The initial concentrations for Th, La and Sm are 0·2, 2·67 and 0·71 ppm, respectively. The D° for Th, La and Sm are 0·0025, 0·008 and 0·047, respectively. The P values for Th, La and Sm are 0·038, 0·012 and 0·071, respectively. Each symbol indicates an increment of 2% melting.

 
Estimating the proportions of clinopyroxene and garnet in sources
An important objective is to evaluate the role of clinopyroxene and garnet in the sources of lavas forming the North Arch and Honolulu Volcanics. Garnet lherzolite has been proposed as the source for the Honolulu Volcanics (Clague & Frey, 1982) and North Arch (Frey et al., 2000), but recent Os isotopic data have been used to argue that pyroxenite forms an important part of the source of Honolulu Volcanics (Lassiter et al., 2000). During melting of anhydrous garnet lherzolite and pyroxenite the abundances of incompatible elements are dominantly controlled by residual clinopyroxene and garnet, which are more abundant in pyroxenite than in peridotite. Here, we use the inversion of batch melting equation [equation (6) in Appendix B] to determine the bulk partition coefficients of LREE (), which are then used to infer proportions of clinopyroxene and garnet in the source.

For this approach to be valid, there are several requirements. First, primary melt compositions must be estimated; second, the isotopic and major element characteristics of the lavas must be consistent with derivation from a common source; third, the mineral assemblage in the source must remain essentially unchanged during melting so that P^x and P^y in equation (5) of Appendix B are constant or much less than unity; fourth, the samples must be evenly distributed over a range of 1/x to ensure that the regression analysis is statistically meaningful. Based on their variations in trace element abundance ratios, Group A North Arch lavas and the seven lavas from the Honolulu Volcanics with Th content >5 ppm meet these requirements. Their systematic 1/Th–1/La variations are used for the inverse calculation.

In addition to the slope (S) and intercept (I) in a 1/x–1/y plot, four more variables are required to quantify bulk [equation (6) in Appendix B]. For mantle lithology, P^Th and P^LREE (see Appendix B for notations) are <<1; therefore, they can be ignored. Consequently, the major sources of uncertainties are and . Bulk is dominated by clinopyroxene and garnet. The experimentally determined varies from 0·0013 to 0·036, with most data within a range of 0·01–0·014 (LaTourrette & Burnett, 1992; Beattie, 1993a, 1993b; Hauri et al., 1994; Lundstrom et al., 1998; van Westrenen et al., 2000; Landwehr et al., 2001). For melts that equilibrated with both clinopyroxene and garnet, is lower than and ranges from 0·0013 to 0·0033 (LaTourrette & Burnett, 1992; Beattie, 1993a, 1993b; LaTourrette et al., 1993; Hauri et al., 1994). In contrast, Salters & Longhi (1999) used their experimental data to derive a lower of 0·006 and a higher of 0·008 for low-degree melting of garnet peridotite. They argued, on the basis of major element composition of the bulk system, that the partition coefficients determined from their experiments are suitable for peridotite and those from previous experiments are most applicable to garnet pyroxenite and eclogite. Consequently, we infer that the range of bulk varies from 0·009 for pyroxenite (70% clinopyroxene + 30% garnet) to 0·002 for peridotite (25% clinopyroxene + 10% garnet). This value for garnet peridotite is very similar to those calculated (0·0022) by Landwehr et al. (2001, table 4). With this range of , the Th contents in the primary North Arch Group A lavas can be derived by 4% melting from primitive mantle (PM) or 8–14% melting from enriched mantle with a Th concentration three times PM value. Higher Th content in the source, for example four times PM, requires 12–18% melting, which is unlikely for these highly alkalic lavas. With constraints of (0·002–0·009) and (1–3 times PM value), we then calculated bulk using the slopes and intercepts in the 1/Th–1/La plot.

In a 1/Th–1/La plot, Group A North Arch lavas form a trend with intercept/slope ratio (I/S) of 0·093 (Fig. 8). If the source of these lavas has a of 0·085–0·26 ppm (1–3 times the PM value) and a of 0·002–0·009 as we inferred above, the bulk in this source can vary from 0·011 to 0·034 based on equation (6) in Appendix B (stippled area in Fig. 9a). Using a of 0·08 (Blundy et al., 1998) and a of 0·0164 (Hauri et al., 1994), this range of bulk corresponds to 10–35% clinopyroxene in the source if the clinopyroxene/garnet ratio is two (Fig. 9b). A lower of 0·054 from Hart & Dunn (1993) increases the clinopyroxene proportion to 15–55% depending on the clinopyroxene/garnet ratios (Fig. 9b). Because the bulk is dominated by , variation in causes only <5% difference in clinopyroxene proportion estimates. The and from Salters & Longhi (1999) are <0·012, which are too low to account for the bulk (0·011–0·034). These estimated clinopyroxene and garnet modes extend to higher than those in typical garnet peridotite, but not as high as those in garnet pyroxenite. Therefore, the source of Group A North Arch lavas is inferred to be a mixture of peridotite and garnet pyroxenite.

View larger version (20K): [in this window] [in a new window]   Fig. 8. Plot of 1/Th vs 1/La for North Arch Group A lavas and the nephelinite and nepheline melilitites from the Honolulu Volcanics. The abundances of Th, La and Ce are corrected for olivine fractionation. Also indicated are regression lines through the two groups of lavas with the linear equations and R2 labelled.

 

View larger version (33K): [in this window] [in a new window]   Fig. 9. (a) Bulk vs bulk calculated for the source of North Arch lavas using inversion of batch melting [Appendix B, equation (6)]. The Th content in the source is considered to be 1–3 times the primitive mantle value (0·085) and the range of bulk is estimated to be within 0·002–0·009 (see text for derivation of these ranges). The stippled area indicates the range of bulk and corresponding bulk (0·011–0·034) estimated for the source of North Arch lavas. (b) Calculated bulk for the source of North Arch lavas vs clinopyroxene % (cpx%) estimated from the published for cpx and garnet (gt) using various cpx/gt ratios. (c) Bulk vs bulk calculated for the source of the Honolulu Volcanic field using inversion of batch melting [Appendix B, equation (6)]. The stippled area indicates the range of bulk and corresponding bulk estimated for the source of the Honolulu Volcanics. (d) Calculated bulk (0·004–0·012) for the source of the Honolulu Volcanic field vs required cpx%. The and cpx/gt ratios used to calculate the lower dash and dot lines are as indicated in (b).

 
Compared with Group A North Arch lavas, the seven Honolulu samples with Th contents >5 ppm define a much lower I/S ratio in a 1/Th–1/La plot (Fig. 8; 0·093 vs 0·0081). Because of the low I/S ratio, the bulk is insensitive to the term and is dominated by only the term [equation (6) in Appendix B]. Because a ranging from 0·002 to 0·009 leads to a bulk of 0·004–0·012 (Fig. 9c), the source of Honolulu lavas contains 5–20% of clinopyroxene based on a of 0·05–0·08 (Fig. 9d). Varying clinopyroxene/garnet ratio from one to three causes only <2% difference in the estimates for clinopyroxene proportion (Fig. 9d). The low and from Salters & Longhi (1999) require 30–90% clinopyroxene. Although the literature data for and result in a large variation in the proportion of clinopyroxene (Fig. 9c and d), most of the data for and indicate that the source of North Arch lavas contains 10–55% clinopyroxene and 5–25% garnet whereas that of Honolulu lavas has only 5–20% clinopyroxene and 5–8% garnet. This result is consistent with the inference from the Tb/Yb vs Th trend (Fig. 6a) that garnet was a more significant phase in the residue for North Arch lavas, and supports the conclusions of Clague & Frey (1982) and Frey et al. (2000) that another residual phase, such as olivine, is required for the source of Honolulu Volcanics and North Arch lavas.

Isotopic and compositional constraints on the sources of lavas forming North Arch and Honolulu Volcanics
Evaluating previous models
The low ⁸⁷Sr/⁸⁶Sr and high ¹⁴³Nd/¹⁴⁴Nd ratios in the Hawaiian North Arch and rejuvenated-stage lavas require derivation from a long-term depleted source. However, these lavas are relatively enriched in incompatible elements; i.e. they have high Rb/Sr and low Sm/Nd ratios compared with the tholeiitic basalts that form the Hawaiian shields. This apparent contradiction can be explained by three alternative models: (1) mixing between melts derived from enriched and depleted sources (Chen & Frey, 1985; Reiners, 2002); (2) extremely small extents of melting from depleted mantle sources (Frey & Roden, 1987; Sims et al., 1995); (3) larger extents of melting from a recently enriched source (Roden et al., 1984; Chen & Frey, 1985; Clague & Dalrymple, 1988; Class et al., 1998; Reiners & Nelson, 1998; Sims et al., 1999; Frey et al., 2000).

The observed linear correlations between abundances of incompatible elements (Fig. 2) could be explained by mixing between high-degree and low-degree melts from depleted mantle or mixing between melts derived from enriched and depleted sources (Reiners & Nelson, 1998; Lassiter et al., 2000). In North Arch lavas, however, the trends defined by mixing the extreme end-members of Groups A and B are nearly linear in K/Ce–Ce, Ba/La–Th and Nb/La–Th plots; these linear trends do not explain the different trends defined by Group A and B North Arch lavas (Fig. 4a). The absence of linear trends in Th vs Ti, Zr and Nb plots for the Honolulu Volcanics also argues against melt mixing (Fig. 2b). In addition, mixing between alkali basalt and nepheline melilitite from the Honolulu Volcanics forms a curve in the Th–Ti/Eu plot that is inconsistent with the linear array defined by lavas (Fig. 6b).

The major difference between the models (2) and (3) is source composition, depleted or enriched in incompatible elements. Using the inversion approach based on the batch melting model, we infer an enriched source. Equation (3) in Appendix B indicates that the abundance ratio of two highly incompatible elements in the source, x/y, is approximated as the slope in the 1/x–1/y plot using lava compositions. This approach results in a La/Ce ratio of 0·47 for the sources of North Arch and Honolulu Volcanics. Clague & Frey (1982) reached a similar conclusion for the source of the Honolulu Volcanics using x–x/y plots. The inferred La/Ce ratio of 0·47 is significantly higher than the primitive mantle value (0·387) and is consistent with an enriched source. Watson (1993) and Zou & Zindler (1996) applied different inversion approaches using all REE and they concluded that an enriched source was required for the source of the Honolulu Volcanics. Moreover, extremely low extents of melting [model (2)] result in non-linear variations between abundances of two incompatible elements rather than the observed linear trends (Reiners & Nelson, 1998; Frey et al., 2000). Consequently, neither melting trajectories in x–y plots nor inferred source compositions favor model (2), derivation from a depleted source by small extents of melting. As a result, model (3), variable extent of melting from an enriched source, is the best model for generation of rejuvenated-stage and North Arch lavas.

Characteristics of the enriched source: constraints from the relationship between ⁸⁷Sr/⁸⁶Sr ratios and abundance ratios of incompatible elements
Two types of enriched mantle have been proposed for Hawaiian rejuvenated-stage lavas: (1) addition of low-degree melts from depleted mantle into an enriched source; (2) addition of low-degree melts from an enriched source into depleted mantle. Distinguishing between these two models is important. Is the major melting regime in the depleted mantle or the enriched mantle plume (Class & Goldstein, 1997)? If melting occurred in depleted mantle, what was the enriched component—mantle plume, E-MORB or non-plume metasomatic agent? If a mantle plume, most probably the Hawaiian plume, is the main source, why do these lavas have near-MORB isotopic signatures? To investigate these two types of enriched sources, we first identify the mixing end-members based on lava compositions, and then model the two types of metasomatized sources using mixing calculations. An appropriate source is determined by comparison between calculated and observed ⁸⁷Sr/⁸⁶Sr ratios and abundance ratios of incompatible elements.

In a ⁸⁷Sr/⁸⁶Sr vs ¹⁴³Nd/¹⁴⁴Nd plot, North Arch and rejuvenated-stage lavas plot between the fields for East Pacific Rise MORB and Hawaiian shield lavas, consistent with mixing between depleted mantle and a more enriched component (e.g. Lassiter et al., 2000, Fig. 1). We assume that the depleted component is similar to the source of MORB, but we note that Kani et al. (2000) inferred from Sr–Pb isotopic correlations that the depleted component in North Arch lavas differs from the source of Pacific MORB. The geochemical characteristics of MORB sources are relatively well constrained. In contrast, the composition of the enriched component is an issue of debate. The Hawaiian plume, a Cretaceous mantle plume, primitive mantle, enriched lithospheric mantle and carbonatitic melts have been proposed (Chen & Frey, 1985; Clague & Dalrymple, 1988; Salters & Zindler, 1995; Class & Goldstein, 1997; Reiners & Nelson, 1998; Frey et al., 2000; Lassiter et al., 2000). For example, Salters & Zindler (1995) speculated that carbonatite metasomatism may occur in the Hawaiian lithosphere, based on relative Zr and Ti depletion in clinopyroxenes from mantle xenoliths from Salt Lake Crater, a vent of the Honolulu Volcanics. Although some carbonatitic melts have Ba/Th >150, they are also characterized by depletion in Nb relative to La (i.e. Nelson et al., 1988; Hauri et al., 1993; Ionov et al., 1993; Toyoda et al., 1994). The high Nb/La ratios of North Arch lavas (Fig. 4a) are inconsistent with control by carbonatite.

Incompatible element abundance ratios of North Arch and rejuvenated-stage lavas, in particular, the unusually high Ba/Th (110–200 compared with 82·2 for primitive mantle; see Fig. 10) enable us to identify the enriched source component. The systematic decrease in Ba/Th with increasing Th content reflects the control of phlogopite at low extents of melting and suggests that the high Ba/Th ratios at high extents of melting (low Th content) are source characteristics (Fig. 6a). Most OIB have Ba/Th <100, but most Hawaiian and Icelandic lavas have Ba/Th >100 (Fig. 10a). Normal MORB typically have Ba/Th ratio lower than the chondritic value (i.e. Hofmann, 1988; Niu et al., 1999; Meurer et al., 2001). Also, the E-MORB average of Sun & McDonough (1989) has a Ba/Th of 95. More recent analyses of E-MORB and near ridge seamounts (Mühe et al., 1997; Niu & Batiza, 1997) also yield Ba/Th ratios (<100), consistent with this average. Although six of the 24 E-MORB samples from the East Pacific Rise analyzed by Niu et al. (1999) have Ba/Th ratios >110, they suggested that these samples contain a Hawaiian plume component. The rarity of high Ba/Th ratios (>100) in OIB and MORB suggests that high Ba/Th is a unique characteristic of the Hawaiian and Icelandic plumes. Consequently, the high Ba/Th ratios in North Arch and rejuvenated-stage lavas indicate that the enriched component in the sources of these lavas was the Hawaiian plume. A Nb/Y vs Zr/Y plot, which has been used to distinguish plume and MORB lavas (Fitton et al., 1997), also indicates that North Arch and rejuvenated-stage lavas contain a plume component (Fig. 10b).

View larger version (51K): [in this window] [in a new window]   Fig. 10. (a) Ba/Th vs Sr/Nd for Hawaiian and other OIB lavas. The data sources are indicated in the legend. The field encloses Hawaiian lavas, except for one Honolulu Volcanics sample, which has an unusually high Ba content. PM indicates the value for primitive mantle. The stars are averages of N-MORB and E-MORB from the compilation of Sun & McDonough (1989). All Hawaiian lavas have Ba/Th >110, except for some North Arch and Honolulu Volcanics lavas, which equilibrated with phlogopite and/or amphibole (see discussion in text). (b) A Nb/Y vs Zr/Y discrimination plot proposed by Fitton et al. (1997) to distinguish MORB from lavas derived from the Icelandic plume (the shaded area). The Hawaii data lead to two important observations: (1) Hawaiian shields straddle the lower boundary line, perhaps because the source of geochemically extreme shields, such as Koolau, contained a sedimentary component; (2) lavas from the North Arch and Honolulu Volcanics straddle the upper boundary overlapping with Icelandic lavas and are distant from the MORB field.

 
The relationships between ⁸⁷Sr/⁸⁶Sr and Ba/Th, Sr/Nd and La/Ce ratios can further constrain the role of the Hawaiian plume and depleted mantle in formation of the metasomatized source. Two possible mixing processes are evaluated: (1) addition of incipient melt (<2%) of depleted mantle to the Hawaiian plume; (2) addition of incipient melt (<2%) of the Hawaiian plume to depleted mantle. For modeling these processes we assume that: (1) the depleted component is lithospheric mantle with a ⁸⁷Sr/⁸⁶Sr ratio of 0·7025, whose incompatible element contents are those of a residue formed after extraction of a 3% MORB melt; (2) the enriched plume component has a ⁸⁷Sr/⁸⁶Sr ratio of 0·7036, a value typical of Hawaiian shield lavas (i.e. Stille et al., 1983, 1986; Tatsumoto et al., 1987; Chen et al., 1991; Garcia et al., 1993; Lassiter et al., 1996; Pietruszka & Garcia, 1999) with abundances of La, Ce, Sr and Nd in the enriched (plume) source of 2·0, 1·9, 1·7 and 1·6 times primitive mantle, respectively. The Ba (12·4 ppm) and Th (0·106 ppm) contents are inferred from the average Ba/La and Th/La ratios in Hawaiian shield lavas. Melting, 10–15%, of a peridotite source with these abundances will generate Hawaiian shield lavas (e.g. Feigenson et al., 1996). Partition coefficients from Johnson (1998), which are similar to those of many other workers (e.g. Hart & Dunn, 1993; Hauri et al., 1994; Lundstrom et al., 1998), were used to calculate the compositions of the incipient melts (Table 3).

View this table: [in this window] [in a new window]   Table 3: Source compositions (in ppm) and partition coefficients for the modeling compositions of incipient melts from the depleted mantle (DM) lithosphere and Hawaiian mantle plume

 
We first evaluate mixing models involving addition of incipient melts derived from depleted mantle lithosphere to the Hawaiian plume. This model was examined by Chen & Frey (1985) using a La/Ce vs ⁸⁷Sr/⁸⁶Sr plot and was favored for the generation of Koloa Volcanics on Kauai volcano (Clague & Dalrymple, 1988; Reiners & Nelson, 1998). Compared with the results of Chen & Frey (1985), our calculation shows lower La/Ce ratios in the mixed source [0·37–0·44 in Fig. 11 compared with 0·36–0·55 as given by Chen & Frey (1985)]. This discrepancy results from the difference in the D_La/D_Ce ratio between these two models. Chen & Frey (1985) used a low D_La/D_Ce of 0·5 to obtain higher La/Ce ratios in the incipient melts (<2%) from a MORB source. However, most recent experimental data result in higher D_La/D_Ce ranging from 0·58 to 0·70 (e.g. Hart & Dunn, 1993; Hauri et al., 1994; Lundstrom et al., 1998). Our modeling shows that a mixed source with ⁸⁷Sr/⁸⁶Sr = 0·7031 can explain the Th content inferred for the sources of North Arch and Honolulu Volcanics lavas, but it cannot explain the high La/Ce, Sr/Nd and Ba/Th ratios (Fig. 11). Very low extents of melting, <0·1%, of depleted mantle lithosphere are required to generate the high La/Ce, Sr/Nd and Ba/Th ratios in the source of lavas forming the North Arch and Honolulu Volcanics.

View larger version (35K): [in this window] [in a new window]   Fig. 11. Evaluation of mixing models involving addition of incipient melts of a depleted mantle source (DM) to the plume using 87Sr/86Sr vs La/Ce, Sr/Nd and Ba/Th and Th concentration. The sources of lavas forming the North Arch (NA) and Honolulu Volcanic (HV) are the labelled filled rectangles (ratios) and the stippled fields for Th. These sources were calculated from inversion of the batch melting equation. The vertical dotted line emanating from the DM source indicates the trajectory of melts formed by variable amounts of partial melting. The continuous lines are mixing lines between partial melts derived from DM and the plume source. These lines define a range of possible source compositions for NA and HV lavas that are created by mixing of incipient melts of a DM source with the plume. Labelled dotted lines connecting continuous lines indicate mixing proportions. In general, these model sources cannot explain the La/Ce, Sr/Nd and Ba/Th ratios inferred for NA and HV sources. The partition coefficients and the compositions of plume source and depleted mantle are listed in Table 3. The olivine:orthopyroxene:clinopyroxene:garnet proportions in source and melt are 0·55:0·25:0·15:0·05 and –0·1:0·3:0·4:0·4, respectively. The 87Sr/86Sr ratios are from Roden et al. (1984), Frey et al. (2000) and Lassiter et al. (2000). The Ba/Th ratio for the source of the Honolulu Volcanics is not plotted because the 1/Th vs 1/Ba plot results in a relatively low correlation coefficient.

 
In contrast, a mixed source containing 99·5–98% depleted mantle lithosphere and 0·5–2% melt derived from 2% partial melting of plume source has ⁸⁷Sr/⁸⁶Sr, Ba/Th, Sr/Nd and La/Ce ratios consistent with those inferred for the sources of the North Arch and Honolulu Volcanics lavas (Fig. 12). These low-degree melts of the plume source might be generated with residual K-bearing phases, thereby resulting in depletion of K₂O in the mixed source. Residual phlogopite is not expected for melting of high-temperature plumes (Class & Goldstein, 1997); however, it may persist during incipient melting of a plume at the water-saturated solidus (Sato et al., 1997; Wallace, 1998). Therefore, the K₂O depletion in Group A North Arch lavas, which do not have K/Ce and Rb/Ce variations consistent with residual K-bearing minerals (Fig. 4a), may reflect K depletion in the mixed source. This model also gives Th contents ranging from 0·08 to 0·20 in the metasomatized source, which overlaps with our estimate based on the batch melting equation. Consequently, we favor this model for generation of North Arch and Honolulu Volcanics lavas.

View larger version (39K): [in this window] [in a new window]   Fig. 12. Evaluation of mixing models involving addition of incipient melts of the plume source to DM using 87Sr/86Sr vs La/Ce, Sr/Nd and Ba/Th and Th concentrations. The approach is similar to that for Fig. 11 except that the continuous mixing lines represent mixing between DM and incipient melts of the plume. In general, the sources for NA and HV lavas can be created by adding 0·5–2% of plume-derived melts (2% melting) to DM.

 
The small amount of plume-derived melt, however, does not explain the relatively high proportion of clinopyroxene (10–55%) and garnet (5–25%) inferred to be in the source of North Arch lavas. As discussed by Lassiter et al. (2000), the mantle lithosphere created at the ridge axis must locally contain pyroxenite, if extraction of MORB melts is incomplete.

Constraints from Os and Sr isotopic ratios
In contrast to the small variation of ⁸⁷Sr/⁸⁶Sr ratio (0·7030–0·7034), the ¹⁸⁷Os/¹⁸⁸Os ratio in Hawaiian rejuvenated-stage lavas varies over a large range of 0·134–0·175 (Lassiter et al., 2000). Unlike Hawaiian shield lavas, rejuvenated-stage lavas do not define a positive ¹⁸⁷Os/¹⁸⁸Os–⁸⁷Sr/⁸⁶Sr correlation (Hauri et al., 1996; Lassiter & Hauri, 1998; Lassiter et al., 2000). Eight of 20 rejuvenated-stage samples analyzed by Lassiter et al. (2000) have higher ¹⁸⁷Os/¹⁸⁸Os ratios than shield lavas, thereby implying that the Hawaiian plume alone is not an appropriate source for rejuvenated-stage lavas. Lassiter et al. (2000) argued against the involvement of a Hawaiian plume component in the generation of rejuvenated-stage lavas because these lavas do not fall on calculated mixing curves between shield lavas and depleted mantle in a ¹⁸⁷Os/¹⁸⁸Os vs ⁸⁷Sr/⁸⁶Sr plot (see Lassiter et al., 2000, fig. 5). They proposed that rejuvenated-stage lavas were derived from a marble-cake lithospheric mantle with variable proportions of pyroxenite veins in a lherzolite matrix. With this interpretation the high ¹⁸⁷Os/¹⁸⁸Os ratios in rejuvenated-stage lavas reflect the contribution from pyroxenite veins, which can have high ¹⁸⁷Os/¹⁸⁸Os ratio ranging up to six (Reisberg et al., 1991; Roy-Barman et al., 1996). Lassiter et al. (2000) suggested that the pyroxenite component formed near the mid-ocean ridge from E-MORB melts generated at the limbs of the MORB melting region. A difficulty with this model is that many E-MORB do not have sufficiently high ⁸⁷Sr/⁸⁶Sr ratios (rejuvenated-stage lavas have ratios of 0·7030–0·7034) even after 100 Myr residence in the oceanic lithosphere. For example, E-MORB from the East Pacific Rise (Niu et al., 1999) increase their ⁸⁷Sr/⁸⁶Sr ratios from 0·7028 to 0·7029 after 100 Myr; therefore, a component with a higher ⁸⁷Sr/⁸⁶Sr ratio is required for the rejuvenated-stage lavas.

Because of the similarly high Ba/Th and Sr/Nd ratios in shield and rejuvenated-stage lavas, we have suggested that Hawaiian plume is a suitable component. We have also demonstrated that the ⁸⁷Sr/⁸⁶Sr, Ba/Th, Sr/Nd and La/Ce ratios and Th content in the source of rejuvenated-stage lavas can be modeled by lithospheric mantle metasomatized by low-degree melts from the Hawaiian plume. In our model, the incompatible element abundances of the metasomatized sources are dominated by metasomatic agents. In contrast, we suggest that Os isotopic ratios are controlled by the marble-cake lithosphere that is heterogeneous in ¹⁸⁷Os/¹⁸⁸Os ratio because of heterogeneous pyroxenite distribution (Reisberg et al., 1991; Roy-Barman et al., 1996; Lassiter et al., 2000). The calculated mixing curves in the ¹⁸⁷Os/¹⁸⁸Os–⁸⁷Sr/⁸⁶Sr plot (Fig. 13) show that mixing lithospheric mantle with small amounts (<5%) of plume-derived melt does not change the ¹⁸⁷Os/¹⁸⁸Os but does increase the ⁸⁷Sr/⁸⁶Sr of the lithospheric mantle (Fig. 13). When the proportion of plume-derived component exceeds 5%, the metasomatized mantle has the ⁸⁷Sr/⁸⁶Sr ratio of plume-derived melt and its ¹⁸⁷Os/¹⁸⁸Os ratio starts to approach the value of plume-derived melt (Fig. 13). The ¹⁸⁷Os/¹⁸⁸Os and ⁸⁷Sr/⁸⁶Sr ratios of rejuvenated-stage lavas are consistent with derivation from a source containing 98% depleted lithospheric mantle and 2% plume-derived melt (Fig. 13). This mixing ratio is within the range of mixing proportions required to create the incompatible element abundance ratios in the source of North Arch lavas and the Honolulu Volcanics (e.g. La/Ce, Ba/Th and Sr/Nd in Fig. 12). In these sources formed by mixing, the incompatible element contents and Sr and Nd isotopic ratios are dominated by the metasomatic melt, whereas the ¹⁸⁷Os/¹⁸⁸Os ratio reflects the heterogeneity of the depleted source caused by varying proportions of lherzolite and pyroxenite.

View larger version (31K): [in this window] [in a new window]   Fig. 13. 187Os/188Os vs 87Sr/86Sr. The dotted field encloses the data of Lassiter et al. (2000) for the Honolulu Volcanics. The strong hyperbolic curves represent mixing between a plume-derived melt with homogeneous 187Os/188Os and 87Sr/86Sr ratios (1% melting) and depleted mantle lithosphere with heterogeneous 187Os/188Os ratio caused by various proportions of pyroxenite. The dashed lines indicate mixing proportions. The melt has 1240 ppm Sr and 0·25 ppb Os. The depleted lithospheric mantle has 9 ppm Sr and 2 ppb Os.

 

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
In addition to their alkalic composition, lavas forming the Hawaiian North Arch and rejuvenated-stage Honolulu Volcanics are distinct from tholeiitic Hawaiian shield lavas in their high abundance of incompatible elements, relatively low ⁸⁷Sr/⁸⁶Sr and high ¹⁴³Nd/¹⁴⁴Nd ratios, and a large variation in ¹⁸⁷Os/¹⁸⁸Os ratio, which is not correlated with ⁸⁷Sr/⁸⁶Sr. The compositional and isotopic variations in these two alkalic suites of lavas provide important geochemical constraints on their sources.

1. The positive Tb/Yb–Th correlation is consistent with the control of residual garnet. Variations in Sr/Nd, K/Ce, Ba/La and Ba/Th indicate that lavas formed by the lowest extents of melting, i.e. highest Th content, formed in equilibrium with phlogopite. Also, variations in Ti/Eu, Zr/Sm, Nb/La and Nb/U indicate that Fe–Ti oxides were important residual phases, especially for the Honolulu Volcanics with >5 ppm Th.
   
2. Based on inversion of the batch melting equation, the North Arch lavas with Th contents <3 ppm were derived from a source with 10–55% clinopyroxene and 5–25% garnet. Compared with that of North Arch lavas, the source of lavas for the Honolulu Volcanics with Th contents >5 ppm contains lower proportions of clinopyroxene (5–20%) and garnet (5–8%).
   
3. The high Nb/Y ratio at a given Zr/Y and the high Ba/Th ratio (significantly greater than the chondritic value) in lavas from the North Arch and Honolulu Volcanics indicate that their sources contain a component derived from the Hawaiian plume.
   
4. Model calculations show that the ⁸⁷Sr/⁸⁶Sr ratio, abundance ratios of incompatible elements, such as La/Ce, Sr/Nd and Ba/Th, and Th contents in these two suites of lavas can be explained by derivation from a long-term depleted source recently enriched by incipient melts (<2%) derived from the Hawaiian plume.
   
5. The abundances of incompatible elements in such a metasomatized mantle are dominated by the incipient melts. In contrast, the variable isotopic ratios of Os, an element that is compatible during mantle melting, reflect heterogeneity, varying proportions of pyroxenite, in the long-term depleted source.
   

	SUPPLEMENTARY DATA

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
Supplementary data for this paper are available on Journal of Petrology online.

	APPENDIX A: COMPARISONS BETWEEN NEW ICP-MS DATA AND PREVIOUS ANALYSES BY X-RAY FLUORESCENCE (XRF) AND INAA

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
For North Arch samples, with few exceptions our ICP-MS data agree with the INAA and XRF analyses of Frey et al. (2000) within ±5% for abundances of Sr, Zr, La, Nd, Sm, Eu and Hf, and within ±10% for Ba and Lu (Appendix Fig. 1a). Data for other elements in these two datasets show systematic differences. Compared with the ICP-MS data, the analyses of Frey et al. (2000) are, in general, higher in Ce and Rb by 1–10%, lower in Y, Nb, Tb, Yb and Lu by 1–15%, and lower in Th by 15–25% (Appendix Fig. 1a). As expected, there are more significant discrepancies between the ICP-MS data and the older data for the Honolulu Volcanics reported by Clague & Frey (1982). Specifically, the abundances of Rb, Sr, Zr, Ba, La, Ce, Nd, Sm, Eu, Tb, Yb, Hf and Ta differ by ±10%, with some exceptions having larger differences of ±20% (Appendix Fig. 1b). In general, the data of Clague & Frey (1982) are 10–25% lower in Y and Nb and 5–15% lower in Th (Appendix Fig. 1b).

View larger version (39K): [in this window] [in a new window]   Appendix Fig. 1.

 

	APPENDIX B: DERIVATION FOR INVERSION OF BATCH MELTING EQUATION

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
The equation for non-modal batch melting is

(1)

where and are concentrations of trace element, i, in melt and initial source, respectively, is the bulk-solid/liquid partition coefficient, F is the extent of melting and P is a bulk-solid/melt partition coefficient determined by the proportion of phases contributing to the melt. Treuil & Joron (1975) and Minster & Allègre (1978) pointed out that the abundance ratio of two elements x and y can be obtained from (1) by dividing to eliminate F so that

(2)

If D and P for element x and y are constant, is a linear function of . Therefore, the linear data array in x/y vs x plots can be used to infer source characteristics () and partition coefficient () provided that , , P^x and P^y are constant. This requirement can be approximated if element x is a highly incompatible element, such as Th (Clague & Frey, 1982; Hofmann & Feigenson, 1983). Equation (2) can be rearranged to the form of a 1/x-1/y linear correlation (Sims & DePaolo, 1997):

(3)

The intercept (I) and slope (S) are given by

(4)

and

(5)

Equations (4) and (5) can be combined to eliminate and obtain

(6)

	ACKNOWLEDGEMENTS

 
We thank S. Huang, M. Schmitz and B. Grant for their assistance in ICP-MS analysis. We also gratefully appreciate reviews by K. Rubin, M. Feigenson, C. Class and D. Geist, as well as discussions with J. Lassiter and S. Parmann. This project was supported by NSF Grant EAR 0105557 to F.A.F. and NSC Grant 91-2116-M-006-005 to H.-J.Y.

	REFERENCES

TOP ABSTRACT INTRODUCTION SAMPLING ANALYTICAL METHODS RESULTS DISCUSSION CONCLUSIONS SUPPLEMENTARY DATA APPENDIX A: COMPARISONS BETWEEN... APPENDIX B: DERIVATION FOR... REFERENCES

 
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]

Beattie, P. (1993a). The generation of uranium series disequilibria by partial melting of spinel peridotite: constraints from the partitioning studies. Earth and Planetary Science Letters 117, 379–391.[CrossRef][Web of Science]

Beattie, P. (1993b). Uranium–thorium disequilibria and partitioning on melting of garnet peridotite. Nature 363, 63–65.[CrossRef]

Blundy, J. D., Robinson, J. A. C. & Wood, B. J. (1998). Heavy REE are compatible in clinopyroxene on the spinel lherzolite solidus. Earth and Planetary Science Letters 160, 493–504.[CrossRef][Web of Science]

Caroff, M., Maury, R. C., Guille, G. & Cotton, J. (1997). Partial melting below Tubuai (Austral Islands, French Polynesia). Contributions to Mineralogy and Petrology 127, 369–382.[CrossRef]

Chauvel, C. & Hémond, C. (2000). Melting of a complete section of recycled oceanic crust: trace element and Pb isotopic evidence from Iceland. Geochemistry, Geophysics, Geosystems 1, 1999GC000003.

Chen, C.-Y. & Frey, F. A. (1983). Origin of Hawaiian tholeiite and alkalic basalt. Nature 302, 785–789.[CrossRef]

Chen, C.-Y. & Frey, F. A. (1985). Trace element and isotopic geochemistry of lavas from Haleakala volcano, East Maui, Hawaii: implications for the origin of Hawaiian lavas. Journal of Geophysical Research 90, 8743–8768.

Chen, C.-Y., Frey, F. A. & Garcia, M. O. (1990). Evolution of alkalic lavas at Haleakala Volcano, east Maui, Hawaii. Contributions to Mineralogy and Petrology 105, 197–218.[CrossRef][Web of Science]

Chen, C.-Y., Frey, F. A., Garcia, M. O., Dalrymple, G. B. & Hart, S. R. (1991). The tholeiite to alkalic basalt transition at Haleakala Volcano, Maui, Hawaii. Contributions to Mineralogy and Petrology 106, 183–200.[CrossRef][Web of Science]

Clague, D. A. & Dalrymple, G. B. (1988). Age and petrology of alkalic postshield and rejuvenated-state lava from Kauai, Hawaii. Contributions to Mineralogy and Petrology 99, 202–218.[CrossRef][Web of Science]

Clague, D. A. & Dixon, J. E. (2000). Extrinsic controls on the evolution of Hawaiian ocean island volcanoes. Geochemistry, Geophysics, Geosystems 1, 1999GC000023.

Clague, D. A. & Frey, F. A. (1982). Petrology and trace element geochemistry of the Honolulu Volcanics, Oahu: implications for the oceanic mantle below Hawaii. Journal of Petrology 23, 447–504.[Abstract/Free Full Text]

Clague, D. A., Holcomb, R. T., Sinton, J. M., Detrick, R. S. & Torresan, M. E. (1990). Pliocene and Pleistocene alkalic flood basalts on the seafloor north of the Hawaiian islands. Earth and Planetary Science Letters 98, 175–191.[CrossRef][Web of Science]

Class, C. & Goldstein, S. L. (1997). Plume–lithosphere interactions in the ocean basins: constraints from the source mineralogy. Earth and Planetary Science Letters 150, 245–260.[CrossRef][Web of Science]

Class, C., Goldstein, S. L., Altherr, R. & Bachelery, P. (1998). The process of plume–lithosphere interactions in the ocean basins—the case of Grande Comore. Journal of Petrology 39, 881–903.[CrossRef][Web of Science]

Dalpé, C. & Baker, D. R. (2000). Experimental investigation of large ion lithophile element, high field strength element and rare earth element partitioning between calcic amphibole and basaltic melt: the effects of pressure and oxygen fugacity. Contributions to Mineralogy and Petrology 140, 233–250.[CrossRef][Web of Science]

David, K., Schiano, P. & Allègre, C. J. (2000). Assessment of the Zr/Hf fractionation in oceanic basalts and continental materials during petrogenetic processes. Earth and Planetary Science Letters 178, 285–301.[CrossRef][Web of Science]

Dixon, J. E., Clague, D. A., Wallace, P. & Poreda, R. (1997). Volatiles in alkalic basalts from the North Arch volcanic field, Hawaii: extensive degassing of deep submarine-erupted alkalic series lavas. Journal of Petrology 38, 911–939.[CrossRef][Web of Science]

Feigenson, M. D. (1984). Geochemistry of Kauai volcanics and a mixing model for the origin of Hawaiian alkalic basalts. Contributions to Mineralogy and Petrology 87, 109–119.[CrossRef][Web of Science]

Feigenson, M. D., Patino, L. C. & Carr, M. J. (1996). Constraints on partial melting imposed by rare earth element variations in Mauna Kea basalts. Journal of Geophysical Research 101, 11815–11829.[CrossRef]

Fitton, J. G., Saunders, A. D., Norry, M. J., Hardarson, B. S. & Taylor, R. N. (1997). Thermal and chemical structure of the Iceland plume. Earth and Planetary Science Letters 153, 197–208.[CrossRef][Web of Science]

Foley, S. F., Barth, M. G. & Jenner, G. A. (2000). Rutile/melt partition coefficients for trace elements and an assessment of the influence of rutile on the trace element characteristics of subduction zone magmas. Geochimica et Cosmochimica Acta 64, 933–938.[CrossRef][Web of Science]

Francis, D. & Ludden, J. (1995). The signature of amphibole in mafic alkaline lavas, a study in the Northern Canadian Cordillera. Journal of Petrology 36, 1171–1191.[Abstract/Free Full Text]

Frey, F. A. & Roden, M. F. (1987). The mantle source for the Hawaiian islands: constraints from the lavas and ultramafic inclusions. In: Menzies, M. A. & Hawkesworth, C. J. (eds) Mantle Metasomatism. London: Academic Press, pp. 423–463.

Frey, F. A., Wise, W. S., Garcia, M. O., West, H., Kwon, S.-T. & Kennedy, A. (1990). Evolution of Mauna Kea Volcano, Hawaii: petrologic and geochemical constraints on postshield volcanism. Journal of Geophysical Research 95, 1271–1300.

Frey, F. A., Garcia, G. A. & 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]

Frey, F. A., Clague, D. A., Mahoney, J. J. & Sinton, J. M. (2000). Volcanism at the edge of Hawaiian plume: petrogenesis of submarine alkalic lavas from the North Arch volcanic field. Journal of Petrology 41, 667–691.[Abstract/Free Full Text]

Garcia, M. O., Rhodes, J. M., Wolfe, E. W., Ulrich, G. E. & Ho, R. A. (1992). Petrology of lavas from episodes 2–47 of the Puu Oo eruption of Kilauea volcano: evaluation of magmatic processes. Bulletin of Volcanology 55, 1–16.[CrossRef][Web of Science]

Garcia, M. O., Jorgenson, B. A., Mahoney, J. J., Ito, E. & Irving, A. J. (1993). An evaluation of temporal geochemical evolution of Loihi summit lavas: results from Alvin submersible dives. Journal of Geophysical Research 98, 537–550.

Garcia, M. O., Foss, D. J. P., West, H. B. & Mahoney, J. J. (1995). Geochemical and isotopic evolution of Loihi volcano, Hawaii. Journal of Petrology 36, 1647–1674.[Abstract/Free Full Text]

Garcia, M. O., Rhodes, J. M., Trusdell, F. A. & Pietruszka, A. J. (1996). Petrology of lavas from the Puu Oo eruption of Kilauea Volcano: III. The Kupaianaha episode (1986–1992). Bulletin of Volcanology 58, 359–379.[CrossRef][Web of Science]

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

Hauri, E. H., Shimizu, N., Dieu, J. J. & Hart, S. R. (1993). Evidence for hotspot-related carbonatite metasomatism in the oceanic upper mantle. Nature 365, 221–227.[CrossRef]

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.[CrossRef][Web of Science]

Hauri, E. H., Lassiter, J. C. & DePaolo, D. J. (1996). Osmium isotope systematics of drilled lavas from Mauna Loa, Hawaii. Journal of Geophysical Research 101, 11793–11806.[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]

Hofmann, A. W. & Feigenson, M. D. (1983). Case studies on the origin of basalt. I. Theory and reassessment of Grenada basalts. Contributions to Mineralogy and Petrology 84, 382–389.[CrossRef][Web of Science]

Hofmann, A. W., Feigenson, M. D. & Raczek, I. (1984). Case studies on the origin of basalt. III. Petrogenesis of the Mauna Ulu eruption Kilauea, 1969–1971. Contributions to Mineralogy and Petrology 88, 24–35.[CrossRef][Web of Science]

Hofmann, A. W., Jochum, K. P., Seufert, M. & White, W. M. (1986). Nb and Pb in oceanic basalts: new constraints on mantle evolution. Earth and Planetary Science Letters 79, 33–45.[CrossRef][Web of Science]

Huang, S. & Frey, F. A. (2003). The trace element abundances of Mauna Kea basalt from phase 2 of the Hawaii Scientific Drilling Project: petrogenetic implications of correlations with major element content and isotopic ratios. Geochemistry, Geophysics, Geosystems (in press).

Ionov, D. A., Dupuy, C., O'Reilly, S. Y., Kopylova, M. G. & Genshaft, Y. S. (1993). Carbonated peridotite xenoliths from Spitsbergen: implications for trace element signature of mantle carbonate metasomatism. Earth and Planetary Science Letters 119, 283–297.[CrossRef][Web of Science]

Johnson, K. T. M. (1998). Experimental determination of partition coefficients for rare earth and high-field-strength elements between clinopyroxene, garnet, and basaltic melt at high pressures. Contributions to Mineralogy and Petrology 133, 60–68.[CrossRef][Web of Science]

Kani, T., Uto, K., Clague, D. A., & Naka, J. (2000). Isotopic study for alkalic lavas from the North Arch Volcanic Field, Hawaii. EOS Transactions, American Geophysical Union, Western Geophysics Meeting 81, 256.

Kinzler, R. J. (1997). Melting of mantle peridotite at pressures approaching the spinel to garnet transition: application to mid-ocean ridge basalt petrogenesis. Journal of Geophysical Research 102, 853–874.[CrossRef]

Landwehr, D., Blundy, J., Chamorro-Perez, E. M., Hill, E. & Wood, B. (2001). U-series disequilibria generated by partial melting of spinel lherzolite. Earth and Planetary Science Letters 188, 329–348.[CrossRef][Web of Science]

Langmuir, C. H., Klein, E. M. & Plank, T. (1992). Petrological systematics of mid-ocean ridge basalts: constraints on melt generation beneath ocean ridges. In: Phipps Morgan, J., Blackman, D. K. & Sinton, J. M. (eds) Mantle Flow and Melt Generation at Mid-Ocean Ridges. Geophysical Monograph, American Geophysical Union 71, 183–280.

Lanphere, M. A. & Dalrymple, G. B. (1980). Age and strontium isotopic composition of the Honolulu Volcanic Series, Oahu, Hawaii. American Journal of Sciences 280-A, 736–751.

Lassiter, J. C. & Hauri, E. H. (1998). Osmium-isotope variations in Hawaiian lavas: evidence for recycled oceanic lithosphere in the Hawaiian plume. Earth and Planetary Science Letters 164, 483–496.[CrossRef][Web of Science]

Lassiter, J. C., Depaolo, D. J. & Tatsumoto, M. (1996). Isotopic evolution of Mauna Kea volcano: results from the initial phase of the Hawaii Scientific Drilling Project. Journal of Geophysical Research 101, 11769–11780.[CrossRef][Web of Science]

Lassiter, J. C., Hauri, E. H., Reiners, P. W. & Garcia, M. O. (2000). Generation of Hawaiian post-erosional lavas by melting of a mixed lherzolite/pyroxenite source. Earth and Planetary Science Letters 178, 269–284.[CrossRef][Web of Science]

LaTourrette, T. Z. & Burnett, D. S. (1992). Experimental determination of U partitioning and Th partitioning between clinopyroxene and natural and synthetic basaltic liquid. Earth and Planetary Science Letters 110, 227–244.[CrossRef][Web of Science]

LaTourrette, T. Z., Kennedy, A. K. & Wasserburg, G. J. (1993). Thorium–uranium fractionation by garnet: evidence for a deep source and rapid rise of oceanic basalts. Science 261, 739–742.[Abstract/Free Full Text]

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

McCallum, I. S. & Charette, M. P. (1978). Zr and Nb partition coefficients: implications for the genesis of mare basalts, KREEP, and sea floor basalts. Geochimica et Cosmochimica Acta 42, 859–869.[CrossRef][Web of Science]

McKay, G. A., Wagstaff, J. & Yang, S.-R. (1986). Zirconium, halfnium, and rare earth element partition coefficients for ilmenite and other minerals in high-Ti lunar mare basalts: an experimental study. Proceeding 16th Lunar and Planetary Science Conference. Journal of Geophysical Research 91, Supplement, D229–D237.

Meurer, W. P., Sturm, M. A., Klein, E. M. & Karson, J. A. (2001). Basalt compositions from the Mid-Atlantic Ridge at SMARK area (23°30'N to 22°50'N)—implications for parental liquid variability at isotopically homogeneous spreading centers. Earth and Planetary Science Letters 186, 451–469.[CrossRef][Web of Science]

Minster, J. F. & Allègre, C. J. (1978). Systematic use of trace elements in igneous processes. Part III: Inverse problem of batch melting in volcanic suites. Contributions to Mineralogy and Petrology 68, 37–52.[CrossRef][Web of Science]

Mühe, R., Bohrmann, H., Garbe-Schönberg, D. & Kassens, H. (1997). E-MORB glasses from the Gakkel Ridge (Arctic Ocean) at 87°N: evidence for the Earth's most northerly volcanic activity. Earth and Planetary Science Letters 152, 1–9.[CrossRef][Web of Science]

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]

Nielsen, R. L., Gallahan, W. E. & Newberger, F. (1992). The partitioning of Sc, Y, and the REE between olivine, low-Ca pyroxene, ilmenite, magnetite and natural silicate magmas. Contributions to Mineralogy and Petrology 110, 488–499.[CrossRef][Web of Science]

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]

Niu, Y., Collerson, K. D., Batiza, R., Wendt, J. I. & Regelous, M. (1999). Origin of enriched-type mid-ocean ridge basalt at ridges far from mantle plume. Journal of Geophysical Research 104, 7067–7087.[CrossRef]

Pearce, J. A. & Norry, M. J. (1979). Petrogenetic implications of Ti, Zr, Y, and Nb variations in volcanic rocks. Contributions to Mineralogy and Petrology 69, 33–47.[CrossRef][Web of Science]

Philpotts, J. A., Schnetzler, C. C. & Thomas, H. M. (1972). Petrogenetic implications of some new geochemical data on eclogitic and ultrabasic inclusions. Contributions to Mineralogy and Petrology 36, 1131–1166.

Pietruszka, A. J. & Garcia, M. O. (1999). A rapid fluctuation in the mantle source and melting history of Kilauea volcano inferred from the geochemistry of its historical summit lavas. Journal of Petrology 40, 1321–1342.[CrossRef][Web of Science]

Reiners, P. W. (2002). Temporal–compositional trends in intraplate basalt eruptions: implications for mantle heterogeneity and melting processes. Geochemistry, Geophysics, Geosystems 3, U1–U30.

Reiners, P. W. & Nelson, B. K. (1998). Temporal–compositional–isotopic trends in rejuvenated-stage magmas of Kauai, Hawaii, and implications for mantle melting process. Geochimica et Cosmochimica Acta 62, 2347–2368.[CrossRef][Web of Science]

Reisberg, L. C., Allègre, C. J. & Luck, J.-M. (1991). The Re–Os systematics of the Ronda ultramafic complex of southern Spain. Earth and Planetary Science Letters 105, 196–213.[CrossRef][Web of Science]

Reynolds, R. W. & Geist, D. J. (1995). Petrology of lavas from Sierra Negra volcano, Isabela Island, Galápagos archipelago. Journal of Geophysical Research 100, 24537–24553.[CrossRef]

Rhodes, J. M. (1996). Geochemical stratigraphy of lava flows sampled by the Hawaii Scientific Drilling Project. Journal of Geophysical Research 101, 11729–11746.[CrossRef]

Rhodes, J. M. & Hart, S. R. (1996). Episodic trace element and isotopic variations in historical Mauna Loa lavas: implications for magma and plume dynamics. In: Rhodes, J. M. & Lockwood, J. P. (eds) Mauna Loa Revealed. Geophysical Monograph, American Geophysical Union 92, 263–288.

Roden, M. F., Frey, F. A. & Clague, D. A. (1984). Geochemistry of tholeiitic and alkalic lavas from the Koolau Range, Oahu, Hawaii: implications for Hawaiian volcanism. Earth and Planetary Science Letters 69, 141–158.[CrossRef][Web of Science]

Roy-Barman, M., Luck, J.-M. & Allègre, C. J. (1996). Os isotopes in orogenic lherzolite massifs and mantle heterogeneities. Chemical Geology 130, 55–64.[CrossRef][Web of Science]

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]

Salters, V. J. M. & Longhi, J. (1999). Trace element partitioning during the initial stages of melting beneath mid-ocean ridges. Earth and Planetary Science Letters 166, 15–30.[CrossRef][Web of Science]

Salters, V. J. M. & Zindler, A. (1995). Extreme ¹⁷⁶Hf/¹⁷⁷Hf in the sub-oceanic mantle. Earth and Planetary Science Letters 129, 13–30.[CrossRef][Web of Science]

Sato, K., Katsura, T. & Ito, E. (1997). Phase relations of natural phlogopite with and without enstatite up to 8 GPa: implication for mantle metasomatism. Earth and Planetary Science Letters 146, 511–526.[CrossRef][Web of Science]

Shimizu, N. & Kushiro, I. (1975). The partitioning of rare earth elements between garnet and liquid at high pressures: preliminary experiments. Geophysical Research Letters 2, 413–416.[Web of Science]

Sims, K. W. W. & DePaolo, D. J. (1997). Inferences about mantle magma sources from incompatible element concentration ratios in oceanic basalts. Geochimica et Cosmochimica Acta 61, 765–784.[CrossRef][Web of Science]

Sims, K. W. W., DePaolo, D. J., Murrell, M. T., Baldridge, W. S., Goldstein, S. J. & Clague, D. A. (1995). Mechanism of magma generation beneath Hawaii and mid-ocean ridges: uranium–thorium and samarium/neodymium isotopic evidence. Science 267, 508–512.[Abstract/Free Full Text]

Sims, K. W. W., DePaolo, D. J., Murrell, M. T., Baldridge, W. S., Goldstein, S. J., Clague, D. A. & Jull, M. (1999). Porosity of the melting zone and variations in the solid mantle upwelling rate beneath Hawaii: inferences from ²³⁸U–²³⁰Th–²²⁶Ra and ²³⁵U–²³¹Pa disequilibria. Geochimica et Cosmochimica Acta 63, 4119–4138.[CrossRef][Web of Science]

Späth, A., le Roex, A. P. & Duncan, R. A. (1996). The geochemistry of lavas from the Comores archipelago, western Indian Ocean: petrogenesis and mantle source region characteristics. Journal of Petrology 37, 961–991.[Abstract/Free Full Text]

Stille, P., Unruh, D. M. & Tatsumoto, M. (1983). Pb, Sr, Nd and Hf isotopic evidence of multiple sources for Oahu, Hawaii basalts. Nature 304, 25–29.[CrossRef]

Stille, P., Unruh, D. M. & Tatsumoto, M. (1986). Pb, Sr, Nd and Hf isotopic constraints on the origin of Hawaiian basalts and evidence for a unique source. Geochimica et Cosmochimica Acta 50, 2303–2319.[CrossRef][Web of Science]

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

Tatsumoto, M., Hegner, E. & Unruh, D. M. (1987). Origin of the West Maui volcanic rocks inferred from Pb, Sr and Nd isotopes and a multicomponent model for oceanic basalt. US Geological Survey Professional Paper 1350, 723–744.

Toyoda, K., Horiuchi, H. & Tokonami, M. (1994). Dupal anomaly of Brazilian carbonatites: geochemical correlations with hotspots in the South Atlantic and implications for the mantle source. Earth and Planetary Science Letters 126, 315–331.[CrossRef][Web of Science]

Treuil, M. & Joron, J.-L. (1975). Utilisation des elements hygromagmatophiles pour la simplification de la modelisation quantitative des processus magmatiques. Exemples de L'Afar et de la Dorsale Medioatlantique. Societa Italiana Mineralogia e Petrologia 31, 125–174.

van Westrenen, W., Blundy, J. D. & Wood, B. J. (1999). Crystal-chemical controls on trace element partitioning between garnet and anhydrous silicate melt. American Mineralogist 84, 838–847.[Abstract]

van Westrenen, W., Blundy, J. D. & Wood, B. J. (2000). Effect of Fe²⁺ on garnet–melt trace element partitioning: experiments in FCMAS and quantification of crystal-chemical controls in natural systems. Lithos 53, 189–201.[CrossRef][Web of Science]

Wallace, P. J. (1998). Water and partial melting in mantle plumes: inferences from the dissolved H₂O concentrations of Hawaiian basaltic magmas. Geophysical Research Letters 25, 3639–3642.[CrossRef][Web of Science]

Walter, M. J. (1998). Melting of garnet peridotite and the origin of komatiite and depleted lithosphere. Journal of Petrology 39, 29–60.[CrossRef][Web of Science]

Walter, M. J., Sisson, T. W. & Presnall, D. C. (1995). A mass proportion method for calculating melting reactions and application to melting of model upper mantle lherzolite. Earth and Planetary Science Letters 135, 77–90.[CrossRef][Web of Science]

Watson, S. (1993). Rare earth element inversions and percolation models for Hawaii. Journal of Petrology 34, 763–783.[Abstract/Free Full Text]

White, W. M. & Duncan, R. A. (1996). Geochemistry and geochronology of the Society Island: new evidence for deep mantle recycling. In: Basu, A. & Hart, S. R. (eds) Earth Processes: Reading the Isotopic Code. Geophysical Monograph, American Geophysical Union 95, 183–206.

Yang, H. J., Frey, F. A., Rhodes, J. M. & Garcia, M. O. (1996). Evolution of Mauna Kea volcano: inferences from lava compositions recovered in the Hawaii Scientific Drilling Project. Journal of Geophysical Research 101, 11747–11767.[CrossRef][Web of Science]

Zou, H. & Zindler, A. (1996). Constraints on the degree of dynamic partial melting and source composition using concentration ratios in magmas. Geochimica et Cosmochimica Acta 60, 711–717.[CrossRef][Web of Science]

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

This article has been cited by other articles:

				M. Portnyagin, D. Savelyev, K. Hoernle, F. Hauff, and D. Garbe-Schonberg Mid-Cretaceous Hawaiian tholeiites preserved in Kamchatka Geology, November 1, 2008; 36(11): 903 - 906. [Abstract] [Full Text] [PDF]

				R. Dasgupta, M. M. Hirschmann, and N. D. Smith Partial Melting Experiments of Peridotite + CO2 at 3 GPa and Genesis of Alkalic Ocean Island Basalts J. Petrology, November 1, 2007; 48(11): 2093 - 2124. [Abstract] [Full Text] [PDF]

				D. J. Frost The Stability of Hydrous Mantle Phases Reviews in Mineralogy and Geochemistry, January 1, 2006; 62(1): 243 - 271. [Full Text] [PDF]

				R. N. THOMPSON, C. J. OTTLEY, P. M. SMITH, D. G. PEARSON, A. P. DICKIN, M. A. MORRISON, P. T. LEAT, and S. A. GIBSON Source of the Quaternary Alkalic Basalts, Picrites and Basanites of the Potrillo Volcanic Field, New Mexico, USA: Lithosphere or Convecting Mantle? J. Petrology, August 1, 2005; 46(8): 1603 - 1643. [Abstract] [Full Text] [PDF]

				G. Sen, S. Keshav, and M. Bizimis Hawaiian mantle xenoliths and magmas: Composition and thermal character of the lithosphere American Mineralogist, May 1, 2005; 90(5-6): 871 - 887. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Supplementary data 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 (23) Request Permissions Google Scholar Articles by YANG, H.-J. Articles by CLAGUE, D. A. 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