Source, composition and distribution of the fluid in the Kurile mantle wedge: Constraints from across-arc variations of B/Nb and B isotopes
Introduction
It is generally accepted that island arc magma is dominated by melts generated in the mantle wedge, which is metasomatized by fluid or melt derived from the subducting oceanic slab 1, 2, 3. Except for some convergent margins where high sheer stress operates and/or very young crust is subducting, the mass transfer from the slab to the mantle wedge is considered to be fluid dominated [4]. In this process, oceanic sediment and altered oceanic crust (AOC), constituting the uppermost part of the oceanic slab, have particular importance because not only are they two of the most significant fluid sources but also the transport of fluid-mobile elements accompanied with their progressive dehydration may affect the genesis of the chemical characteristics of arc magmas (e.g. 1, 5). Data of 10Be 6, 7 and Pb isotopes 8, 9 and mass balance calculation between sediment input and volcanic output of trace elements at global subduction zones [10] provide strong evidence for sediment subduction into subarc mantle. However, it is still controversial how much sediment is actually subducting into the subarc depth without being removed by accretion and/or underplating at shallower depths and to what extent sediment is important as a fluid source compared with underlying AOC 7, 10, 11, 12. Lack of this knowledge makes it difficult to deduce how and by what process the fluid is liberated, transferred and evolved, what chemical composition the fluid has, and how the fluid relates to the genesis of complex chemical characteristics of arc lavas.
Investigation of across-arc and along-arc variations of fluid-mobile elements and isotopes in arc lavas (e.g. 13, 14, 15) is one of the most effective means to address this problem because the results may reflect spatial distribution and composition of the fluid in the mantle wedge, resulting from integrated slab–mantle interaction. B is perhaps one of the elements best suited to elucidate fluid effects because it is extremely mobile during slab dehydration 16, 17 and on that account highly abundant in arc lavas compared with mid-ocean ridge basalt (MORB) and ocean island basalt (OIB) 11, 18. In addition, B isotope compositions of subducting sediment and AOC are generally distinguishable from each other 19, 20. Recent development of B/Nb–B isotope systematics [12], in which B content and isotope ratio are combined with fluid-immobile Nb, enabled investigation of across-arc variations with noticeably high sensitivity. This provides one of the most powerful tools for investigating the fluid-mediated processes at subduction zones.
In this paper, we present across-arc and along-arc variations in B/Nb and B isotope ratios observed in the Kurile arc. This arc is seismologically well characterized, and across-arc major and trace element compositions 13, 21, 22 and isotope ratios [23] have been determined. Based on B/Nb-B isotope systematics, together with previously reported data of trace elements (especially K) and Sr, Nd and Be isotopes, we seek to estimate spatial distribution and isotopic composition of the slab fluid in the Kurile mantle wedge and identify the fluid source. The ultimate purpose of this study is to provide a more generalized model for mass transfer in the arc system.
Section snippets
Geologic settings and geochemical feature of the Kurile arc
The Kurile arc, located on the northwestern Pacific margin, extends for 1150 km between the Kamchatka arc and the Hokkaido–Northeast Japan arc (Fig. 1). Beneath the arc a relatively old and cold Pacific Plate (90–118 Ma) is subducting at the rate of 8.6 cm/yr, and is seismologically visible down to 650 km [24]. The Quaternary volcanic front, represented by a row of major islands, from Paramushir in the north to Kunashir in the south, corresponds to a slab depth of 120–150 km. Behind the
Samples
The samples used in this study are lavas of basalt, basaltic andesite and andesite collected from a wide region of the Kurile arc (Table 1 and Fig. 1). They cover slab depth between 120 km and 220 km. The samples used are aliquots of interior pieces of fresh lava blocks prepared by Ryan et al. [22] and Tera et al. [28]. Rock chips were repeatedly cleaned by diluted HCl and deionized water and then powdered using an alumina ceramics or tungsten carbide mill.
B and Nb concentrations
B and Nb concentrations were determined by inductively coupled plasma–atomic emission spectrometry (ICP–AES). Generally, the difficulties in ICP analysis of B and Nb lie in B volatilization during chemical treatment, spectral overlap of the B and Fe lines and a relatively high detection limit for Nb (0.01 ppm in solution). In order to solve these problems, we developed a simple chemical procedure for the combined separation of B and Nb in a high state of purification. Typically, a sample of
B/Nb and K/Nb ratios
B and Nb concentrations obtained in this study are listed in Table 1. Because both are highly incompatible elements, smaller degrees of partial melting of mantle source or larger degrees of fractional crystallization generally result in elevated B and Nb in magma, as typically seen in L'vinaya Past' Caldera lavas. Considering the frequent occurrence of relatively high SiO2 lavas at the volcanic front compared with the rear arc, the majority of the B data show nearly constant concentration level
Estimated composition of the Kurile fluid
Besides the fundamental parameters of partial melting, crystal fractionation and crustal contamination, other superimposed factors may influence across-arc and along-arc variations in trace element and isotope compositions of the lavas. These can be summarized as: (1) systematic change in fluid flux across the arc; (2) incorporation of spatially or temporally heterogeneous fluid; and (3) source mantle heterogeneity.
In order to evaluate the contribution of these factors, the Kurile data are
Acknowledgements
We are deeply indebted to A. Tsvetkov for providing samples. We thank L. Brown and R. Carlson for technical assistance in mass spectrometry. We are grateful to A. Fujiyoshi for the opportunity to use ICP. We also thank J. Gill, M. Schmidt and N. Hemming for their helpful comments. This work was supported by the Carnegie Postdoctoral Fellowships and a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of the Japanese Government. [CL]
References (67)
- et al.
Partial melting of subducting oceanic crust
Earth Planet. Sci. Lett.
(1994) - et al.
Chemical characteristics of island arc basalts: implications for mantle sources
Chem. Geol.
(1980) - et al.
Sediment incorporation in island-arc magmas: inference from 10Be
Geochim. Cosmochim. Acta
(1986) - et al.
Pb and Sr isotopes in volcanic rocks form the Aleutian islands and Pribilof islands, Alaska
Geochim. Cosmochim. Acta
(1978) - et al.
Nd and Sr isotopic study of volcanic rocks from Japan
Earth Planet. Sci. Lett.
(1981) - et al.
Geochemistry of the Quaternary volcanic rocks of the northeast Japan arc
J. Volcan. Geotherm. Res.
(1986) - et al.
Boron depletion during progressive metamorphism: implications for subduction processes
Earth Planet. Sci. Lett.
(1992) - et al.
B–Be systematics in subduction-related metamorphic rocks: characterization of the subducted component
Geochim. Cosmochim. Acta
(1993) - et al.
The systematics of boron abundances in young volcanic rocks
Geochim. Cosmochim. Acta
(1993) - et al.
Boron isotope systematics of marine sediments
Earth Planet. Sci. Lett.
(1993)