Scholarly article on topic 'Polybaric petrogenesis of Neogene alkaline magmas in an extensional tectonic environment: Viliga Volcanic Field, northeast Russia'

Polybaric petrogenesis of Neogene alkaline magmas in an extensional tectonic environment: Viliga Volcanic Field, northeast Russia Academic research paper on "Earth and related environmental sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Lithos
Keywords
{"Okhotsk–Chukotka Volcanic Belt" / "Northeast Russia" / "Continental margins" / "Intraplate alkaline volcanism" / "Extension-related melting"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Cornelius Tschegg, Theodoros Ntaflos, Vyacheslav V. Akinin

Abstract Neogene alkaline intraplate volcanic rocks from the Viliga Volcanic Field (NE-Russia) were studied both to precisely characterize their geochemical composition and to unravel their petrogenetic history. The sampled volcanic rocks crop out within the voluminous calc-alkaline sequences of the Cretaceous Okhostk–Chukokta Volcanic Belt, an Andean-type arc formed during subduction of the Paleo-Pacific Plate beneath modern far-east Asia. The mantle xenolith- and xenocryst-bearing basanites and nephelinites have intraplate ocean island basalt-type geochemical features. Sr and Nd isotopes combined with major and trace element systematics and rare-earth element modeling suggest polybaric melt generation of these alkaline magmatic rocks from a homogeneous garnet peridotite facies source with minor involvement of spinel peridotite facies partial melts. The basanite samples indicate segregation PT conditions around 1500°C at 33–38kbar whilst the nephelinites reflect smaller melt fractions segregated at over 40kbar. During ascent, olivine (up to 7%) was the main fractionating phase in the basanites; whereas in the nephelinites, both olivine and minor clinopyroxene fractionation occurred. Crustal contamination during ascent was insignificant. We argue that the melt generation of these alkaline magmas from the Viliga Volcanic Field was triggered by an extending lithosphere resulting in upwelling asthenosphere and decompression melting, analogous to geodynamic models of the coeval alkaline volcanic rocks along the adjacent North Pacific continental margins, rather than by subduction- or plume-related processes.

Academic research paper on topic "Polybaric petrogenesis of Neogene alkaline magmas in an extensional tectonic environment: Viliga Volcanic Field, northeast Russia"

Polybaric pedogenesis of Neogene alkaline magmas in an extensional tectonic environment: Viliga Volcanic Field, northeast Russia

Cornelius Tschegg a,*< Theodoros Ntaflos a, Vyacheslav V. Akinin b

a Department of Lithospheric Research, University of Vienna, Althanstr. 14, A-1090 Vienna, Austria b North East Interdisciplinary Scientific Research Institute, Russian Academy of Science, Magadan, Russia

ARTICLE INFO ABSTRACT

Neogene alkaline intraplate volcanic rocks from the Viliga Volcanic Field (NE-Russia) were studied both to precisely characterize their geochemical composition and to unravel their petrogenetic history. The sampled volcanic rocks crop out within the voluminous calc-alkaline sequences of the Cretaceous Okhostk-Chukokta Volcanic Belt, an Andean-type arc formed during subduction of the Paleo-Pacific Plate beneath modern far-east Asia. The mantle xenolith- and xenocryst-bearing basanites and nephelinites have intraplate ocean island basalt-type geochemical features. Sr and Nd isotopes combined with major and trace element systematics and rare-earth element modeling suggest polybaric melt generation of these alkaline magmatic rocks from a homogeneous garnet peridotite facies source with minor involvement of spinel peridotite facies partial melts. The basanite samples indicate segregation PT conditions around 1500 °C at 33-38 kbar whilst the nephelinites reflect smaller melt fractions segregated at over 40 kbar. During ascent, olivine (up to 7%) was the main fractionating phase in the basanites; whereas in the nephelinites, both olivine and minor clinopyroxene fractionation occurred. Crustal contamination during ascent was insignificant. We argue that the melt generation of these alkaline magmas from the Viliga Volcanic Field was triggered by an extending lithosphere resulting in upwelling asthenosphere and decompression melting, analogous to geodynamic models of the coeval alkaline volcanic rocks along the adjacent North Pacific continental margins, rather than by subduction- or plume-related processes.

© 2010 Elsevier B.V. All rights reserved.

Article history: Received 9 September 2010 Accepted 23 November 2010 Available online 30 November 2010

Keywords:

Okhotsk-Chukotka Volcanic Belt Northeast Russia Continental margins Intraplate alkaline volcanism Extension-related melting

1. Introduction

Primitive alkaline rocks with compositional features essentially unaffected by magmatic differentiation, crustal contamination or alteration can help unravel melt generating and petrogenetic processes in the Earth's mantle. The analysis of small volume high-alkaline magmas in intraplate continental and continental margin settings not only contributes to our understanding of magma generation, but also provides crucial information for developing a clearer geodynamic picture of an area. Investigations of the sporadically outcropping primitive lavas in northeastern Asia, including the region around Magadan, the Chukotka Peninsula and Kamchatka, will improve our understanding of the geological evolution of the northeastern Eurasian continental margin, which is composed of a complex assemblage of amalgamated and accreted terranes, overlain and intruded by volcanic arcs.

Within the spatially extensive Cretaceous continental arc-type volcanic successions of the Okhotsk-Chukotka Volcanic Belt (OCVB) at the southern continental margin of northeastern Russia, isolated

* Corresponding author. E-mail address: cornelius.tschegg@umvie.acat (C. Tschegg).

and little studied volcanic necks of Neogene alkaline lavas crop out. One of these, the Viliga Volcanic Field (VVF), lies in the Magadan area close to the Viliga River, and comprises mantle xenolith- and megacryst-bearing basanites as well as nephelinites. The initial discovery of the VVF magmas was made in 1943 by D. S. Kharkevitch, with a short description given by A. L. Krist in 1949. The first major element data on samples from the VVF were obtained by Ichetovkin et al. (1970). More details, including some trace element and Sr, Nd, and Pb isotopic data, were published by Akinin and Apt (1997) and Apt et al. (1998). These preliminary data were all obtained from samples from only one outcrop, in the Aliki River area (VL5, in our study). The most recent and more detailed studies focused on the mantle xenoliths and megacrysts (Akinin et al., 2005; Ntaflos et al., 2008; Tschegg et al., 2004), following an international expedition in 2003.

In this study, new data on the mineralogical and bulk major and trace element compositions and Sr and Nd isotope ratios from four individual localities of the VVF are presented. From these data, we propose a petrogenetic model of the magmatic suite, including mantle source characteristics, pressure/temperature estimates of melt generation, degrees of partial melting, and fractional crystallization processes. Further, the data are compared with Late Cenozoic alkaline lavas emplaced along the northern Pacific continental margin between South-East Asia and Alaska.

0024-4937/$ - see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.lithos.2010.11.009

2. Geological setting and sampling locations

The Okhotsk-Chukotka Volcanic Belt, an extensive Andean-type Early-Late Cretaceous continental volcanic arc that covers over 3000 km of Asia's eastern continental margin (Hourigan and Akinin, 2004) is one of the most voluminous magmatic provinces in northeastern Asia. At the beginning of the Cenozoic, all collision and accretion events had finished and the subduction front of the Paleo-Pacific oceanic plate moved eastwards in the direction of the modern Kamchatka Peninsula. Subsequently, the consolidated continental margin of northeastern Asia experienced intensive destruction along major fault systems and seismically active plate boundaries.

Sparsely distributed Neogene alkaline mafic rocks intrude through, and erupt over, the complete Cretaceous calc-alkaline volcanic successions of the OCVB. On the northern shore of the Sea of Okhotsk, near Magadan, these rocks are represented by basanites and nephelinites of the Viliga Volcanic Field and farther north, on the Chukotka peninsula, by basanites, tephrites and foidites of the Enmelen River Volcanic Field (ENM, Akinin and Apt, 1994; Fig. 1). Some of these primitive magmatic rocks, whose petrogenesis and regional implications in this continental arc-type environment are poorly understood, contain mantle xenoliths and upper mantle xenocrysts, which were studied by Apt et al. (1998), Akinin et al. (2005) and Ntaflos et al. (2008). Neither the Neogene volcanic rocks in the area, which post-date the Cretaceous Paleo-Pacific plate subduction, nor the underlying lithospheric mantle contains geo-

Fig. 1. Simplified geological map of NE-Russia (after Nokleberg et al., 2000; Fujita et al., 2009). Thick dotted line sketches the modern Eurasian-North-American Plate boundary (Chersky Seismic Belt). Thin lines show major regional faults, arrows indicate direction of tectonic movement; thin slashed line shows the location of the seismic profile C-C' from Goryachev et al. (2007). VVF and ENM indicate the location of the Viliga and Enmelen volcanic fields. Dashed field outlines the Balygychan-Sugoi graben; the extension of the Cretaceous Okhotsk-Chukotka Volcanic Belt (OCVB) is outlined in dark grey.

chemical features related to the previous subduction process (Ntaflos et al., 2008). Although the subduction front has moved eastwards since the Cretaceous, towards its present position at the Kamchatka Peninsula, the VVF is still seismically very active, with six recent earthquakes of magnitudes >3.5 and hypocenter depths of 30-35 km (Gunbina et al., 2007). This reflects its position adjacent to the southeastern branch of the dominant Chersky Seismic Belt, the modern boundary between the Eurasian and North-American continental plates and the Okhotsk Block (Fujita et al., 2009).

The VVF, which is the only known Late Neogene intraplate alkali basaltic volcanic field on the northern shore of the Sea of Okhotsk, lies on the east side of the Viliga River, at the junction of the Kananyga and the Aliki Rivers. The VVF is located at the intersection of two major fault systems; a series of NE-trending normal faults that stretch from Magadan along onshore Okhotsk, controlling the development of several neotectonic basins, and the sinistral Ulakhan strike-slip fault, which is perpendicular to the normal faults and thus parallel to the extension-direction (Fig. 1).

Isolated basaltic necks, remnants of cinder cones, dikes and lava flows crop out in the VVF. The largest outcrop is represented by a 7080 m thick lava flow covering an area of around 1.5 km2. Other dikes and cones cover 0.1 to 0.4 km2 each. The VVF cuts through and partly overlies Cretaceous andesites, rhyolites, and dacites of the subduc-tion-related calc-alkaline Okhotsk-Chukotka Volcanic Belt. More specifically, the VVF is located in the northeast trending Balygy-chan-Sugoi graben that is mostly filled with Cretaceous rhyolites and dacite ignimbrites. The graben developed during the early Cretaceous, within gently folded and weakly metamorphosed marine Triassic to Jurassic slates, shales, sandstones (Matveenko, 1957).

The age of VVF rocks is estimated as being Late Miocene; Ar-Ar and K-Ar ages range from 8 to 10 Ma (Akinin et al., 2005; 2008). On the Chukotka Peninsula, lavas of the same age and similar composition crop out in the Enmelen (ENM) volcanic fields (see Fig. 1; Apt et al., 1998).

The four outcrops of the VVF investigated include the previously unstudied locations VL1, VL2 and VL4 in the Kanayga River area (VL1 — N 61°12'46.80", E 155°9'10.8"; VL2—N 61°10'58.80", E 155°24'0.00"; VL4—N 61°10'51.60", E 155°22'48.00") and the locality in the Aliki River area (VL5—N 61°23'34.80", E 155°33'36.00"). The lavas from VL1 and VL5 contain mantle xenoliths (10-25 cm sized spinel-lherzolites) and megacrysts (augites of up to 5 cm diameter; Ntaflos et al., 2008; Akinin et al., 2005), whilst only small xenoliths and xenocrysts occur in VL2 and VL4.

3. Analytical techniques

Polished thin-sections were carbon-coated and analyzed with a Cameca SX 100 electron-probe microanalyzer, equipped with one energy-dispersive and four wavelength-dispersive spectrometers (Department of Lithospheric Research, University of Vienna). Working conditions, measurement procedures and analytical errors are described in more detail in Tschegg et al. (2010); plagioclase, feldspathoid and glass measurements were analyzed using a defocused 6 |jm beam technique. For bulk geochemical analysis, included xenocrysts were sequentially separated during crushing. Major and trace element abundances were analyzed on a PW 2400 X-ray fluorescence spectrometer (Department of Lithospheric Research, University of Vienna) and a Perkin Elmer ELAN 6100 DRC Inductively Coupled Plasma Mass-spectrometer (Division of Analytical Chemistry, University of Natural Resources and Applied Life Sciences, Vienna). Sample preparation and analytical methodology are described in Tschegg et al. (2008). An assessment of analytical precision and accuracy for XRD and ICP-MS measurements are given in the supplementary Table 5. Sr and Nd isotope analyses were performed on a Thermo-Finnigan Triton TI Tims (Department of Lithospheric Research, University of Vienna) following the procedures described in

Thoni et al. (2008). Analytical errors of ±0.000003 for 87Sr/86Sr and 143Nd/144Nd were assessed based on NBS987 and LaJolla international standard measurements.

4. Petrography and mineral chemistry

Representative analyses of each major mineral group forming the respective host lavas are given in the online supplement Table 1 (olivine), Table 2 (clinopyroxene), Table 3 (plagioclase, feldspathoids and glass) and Table 4 (ilmenite and Ti-magnetite). All rock samples show porphyritic textures with unaltered olivine phenocrysts up to 100 |am in size. Small clinopyroxene phenocrysts (<50 |jm) occur very rarely in lavas from VL4 and VL5.

Rocks from VL1 and VL2 contain well equilibrated, unzoned, rather homogeneously composed and mostly euhedral olivine phenocrysts (~Fo77) with a composition similar to the groundmass olivine (Fo76). These phenocrysts show minor evidence of solution at their outermost rim but without any compositional alteration of the rim. The olivine phenocrysts in VL4 and VL5 are slightly more ferrous and show minor zoning, from Fo76 in the core to Fo69 at the rim and from Fo73 to Fo70, respectively. Groundmass olivines have Fo67 in VL4 and Fo69 in VL5. The fairly small and very rare clinopyroxene phenocrysts in VL4 and VL5 are titaniferous augites with compositions of En39Wo50Fs11 and Mg# (100xmolar MgO/[MgO + FeOt]) ranging from 78 to 79. Titano-magnetite inclusions in augite phenocrysts are heterogeneously distributed in their host grain and have the same composition as the groundmass titano-magnetites.

The groundmass minerals are in general fine grained (10-30 |jm) and comprise olivine, clinopyroxene, nepheline, leucite, plagioclase, Ti-Fe oxides and glass. The commonly observed groundmass clinopyroxenes have a very small range of compositions (En39_38Wo50Fsi2_n) with Mg# 77-79 and TiO2 concentrations between 2.6 and 2.8 wt.%, nearly identical to the cpx-phenocrysts in VL4 and VL5. Feldspathoids, nepheline, and leucite are compositionally similar in all samples of the suite. Leucites are all compositionally very similar, with Ne0Ks20_19Q81_80 whereas neph-eline shows slight compositional differences between rocks from localities VL1 and VL2 (Ne24Ks4Q72) compared to those from VL4 and VL5 (Ne26_25Ks5Q70_69). Groundmass plagioclases are rare, occurring only in VL1 and VL2, and have bytownite compositions (Ab2iAn78Or-i). Titano-magnetites have TiO2 varying from 22 to 25 wt.% and FeO from 65 to 67 wt.% in VL1 and VL2, whereas in VL4 and VL5 their composition ranges from 23 to 27 wt.% TiO2 and 66_69 wt.% FeO, respectively. Al2O3 in the titano-magnetites ofVL1 and VL2 is systematically higher (~ 1.7 wt.%) compared to VL4 and VL5 (~0.2 wt.%). Ilmenites were only found in VL1 and VL2 and have 39_41 wt.% FeO and 51_51 wt.% TiO2. Interstitial glasses vary in composition between the analyzed samples with the most significant difference shown between VL1 _VL2 (basanites) and VL4_VL5 (nephelinites). Whereas glasses from VL1 and VL2 have an orthoclase-like chemistry (SiO2: 57_59wt.%, Al2O3: 23_26wt.% and K2O: 7 wt.%), the glasses in VL4 and VL5 have a more silica-undersaturated composition (SiO2: 45_48wt.%, Al2O3: 23_26wt.%, K2O: 5_6wt.% and CaO: 5 wt.%, with BaO concentrations between 2 and 3 wt.%). Accessory apatite was also observed in all studied samples.

Apart from the phenocrysts present, petrographic evidence indicates a crystallization sequence in the groundmass passing from olivine through clinopyroxene, Ti_Fe oxides, leucite, nepheline, plagioclase and finally to glass.

5. Major and trace element and Sr-Nd isotope chemistry

5.1. Classification

In the TAS (total alkalis vs. silica; Le Maitre et al., 1989) diagram (Fig. 2a,), the samples from VL1 and VL2 (SiO2: 43_45 wt.%) all plot in the tephrite/basanite field. Due to their high olivine contents (normative olivine: 14_19%), they can all be classified as basanites

(Le Maitre, 2002). The VL4 and VL5 samples (SiO2: 42_43 wt.%) are transitional between basanites and foidites. Since they are rich in nepheline (normative ne: >20%), we used the term nephelinites (after Le Maitre, 2002) for these rocks in the following text. In Fig. 2a, for comparison, rock compositions from the VVF and ENM, as well as from Neogene and Paleogene volcanic rocks of northeastern Russia and Alaska are plotted.

5.2. Major elements

Major and trace element concentrations of all analyzed samples are given in Table 1; major element abundances are illustrated in Fig. 2, trace elements in Fig. 3. Within the basanites from VL1 and VL2, MgO varies from 10.3 (VL2-T2) to 12.8 wt.% (VL1-40) whilst the studied nephelinites (VL4 and VL5) have lower MgO concentrations (8.8_9.7 wt.%). Total alkali contents vary only marginally within both the basanites (K2O = 2.5_2.9 wt.% and Na2O = 4.1 _4.3 wt.%) and the nephelinites (K2O = 2.9_3.1 wt.%, Na2O = 4.7_5 wt.%), with the latter having slightly higher amounts of alkalis. Both alkalis increase in concentration with decreasing concentrations of MgO (Fig. 2b). SiO2 is systematically higher in the basanite samples (43.5_44.6 wt.%) compared to the nephelinites (42.4_42.9 wt.%). Al2O3 shows a linear increase in concentration, from 13 to 14 wt.%, in the basanites with decreasing MgO but remains constant, at around 13.7 wt.%, in the nephelinites (Fig. 2c, d). CaO concentrations also increase, from 8.9 to 10 wt.%, in the basanite samples with decreasing MgO but remain constant, around 10 wt.%, in the nephelinites (Fig. 2e). FeO and TiO2 are enriched (12.2_12.7 wt.% and ~3 wt.% respectively) in the nephelinites compared to the basanites (10.9_11.2 wt.% and 2.5_ 2.8 wt.%). The basanites again show linear trends for these elements, when plotted against MgO (Fig. 2f). The CaO/Al2O3 ratios (Fig. 2g) remain fairly constant for the basanites (0.67_0.71) and but show a narrow range and higher values for the nephelinites (0.73). The calculated Saturation Index SI (Fitton and James, 1991; Fig. 2h) underlines the slight silica undersaturation of the basanites (SI: — 4.6 to - 7.7) and the significantly more undersaturated nature of the nephelinites (SI: —10.5 to —14.7).

5.3. Trace elements

The compatible trace elements Ni (Fig. 3a) as well as Co and Cr (not shown) correlate positively with MgO. In the basanites (VL1 and VL2), Ni contents range from 201 to 320 ppm and in the nephelinites (VL4 and VL5) from 145 to 160 ppm. Co and Cr vary from 36 to 46 ppm and 252 to 414 ppm, respectively, in the basanites and from 38 to 40 ppm and 160 to 197 ppm in the nephelinites.

The LILE (Large Ion Lithophile Elements) Ba and Rb are present in lower amounts in the basanites (Ba: 493_583 ppm, Rb: 37_46 ppm) compared to the nephelinites (Ba: 628_655 ppm, Rb: 63_79ppm); neither shows a linear correlation with MgO. Concentrations of Sr and the HFSE (high field strength elements) Zr, Nb, Hf, Th, Ta and the REE (rare-earth elements) are quite homogeneous in the basanites of VL1 and VL2-T3, whereas VL2-T2 and VL2-T4 are more enriched in these elements. All the basanites (VL1 and VL2) show a negative correlation with MgO. The nephelinites (VL4 and VL5)systematically show the highest abundances in HFSE except for Th. Zr/Y ratios remain almost constant, varying only between 8.26 and 9.25 in the basanites but are higher in the nephelinites (10.0_10.5; Fig. 3b). La/Nb ratios range from 0.58 to 0.63 in the basanites (VL1 and VL2), whereas in the nephelinites (VL4 and VL5) the values are lower, varying from 0.52 to 0.57.

All samples show sub-parallel REE patterns (Fig. 3c) and with general LREE (light rare-earth element) as well as HREE (heavy rare-earth element) enrichment when normalized to chondrite composition. The basanites from VL1 are generally slightly less REE enriched (La: 175_201 xC1, Lu: 7.9_8.6xC1) compared to the nephelinites (La:

Fig. 2. Bulk major element compositions of the lavas from the VVF. Black arrows show calculated olivine fractionation trends starting at the most primitive sample VL1-40 (MgO: 12.8 wt.%). Black dots on the arrow indicate 3.2% olivine fractionation at MgO: 11.7 wt.% and 6.9% olivine fractionation at MgO: 10.3 wt.%. (a) TAS (total alkaline vs. silica) diagram after Le Maitre (1989); for comparison, data from the VVF, ENM (Aptet al., 1998) and both Neogene and Palaeogene volcanic rocks from northeastern Russia (Akinin etal., 2008) are plotted. Late Cenozoic alkaline basalts from the Bering Sea Volcanic Province (Wirth etal., 2002) are shown (dashed field) as well as two rocks from the Alaskan Prindle Volcano (PV, crosses; Andronikov and Mukasa, 2010); (b) MgO vs. total alkalis; (c) MgO vs. SiO2; (d) MgO vs. Al2O3; (e) MgO vs. CaO; (f) MgO vs. FeOtotal; (g) MgO vs. CaO/Al2O3; (h) MgO vs. Saturation Index (SI; Fitton and James, 1991).

Bulk major, trace and Sr and Nd isotope compositions of the alkaline magmas from the Viliga Volcanic Field.

Basanites Nephelinites

VL1-31 VL1-36 VL1-37 VL1-38 VL1-39 VL1-40 VL2-T2 VL2-T3 VL2-T4 VL4-T1 VL4-T2 VL5-T1 VL5-T2

XRF analyses [wt.%] SiO2 43.5 44.3 44.4 44.0 44.6 44.0 44.0 43.8 43.8 42.7 42.9 42.4 42.4

TiO2 2.78 2.73 2.61 2.72 2.49 2.68 2.67 2.84 2.65 2.99 3.00 3.02 3.01

M2O3 13.1 13.5 13.4 13.4 13.5 13.0 14.0 13.2 14.0 13.7 13.7 13.7 13.7

FeO1 11.1 11.0 11.0 11.1 10.9 11.0 11.1 11.2 11.0 12.2 12.4 12.7 12.7

MnO 0.19 0.19 0.19 0.19 0.19 0.19 0.21 0.20 0.22 0.25 0.24 0.25 0.25

MgO 12.7 12.1 12.6 12.0 11.7 12.8 10.3 12.4 10.6 9.58 9.72 8.75 9.21

CaO 8.91 9.05 9.05 9.20 9.44 8.92 9.96 9.18 9.79 10.0 9.94 9.97 10.0

Na2O 4.19 4.17 4.15 4.23 4.21 4.11 4.34 4.17 4.25 4.94 4.77 5.03 4.68

K2O 2.84 2.86 2.48 2.75 2.56 2.83 2.68 2.71 2.68 3.12 2.86 2.95 2.93

P2O5 1.06 1.01 0.98 1.03 0.99 0.96 1.46 1.12 1.45 1.26 1.37 1.42 1.34

TOTAL 100.4 100.9 100.8 100.5 100.6 100.7 100.7 100.8 100.5 100.8 100.8 100.1 100.2

LOI 0.72 0.65 0.37 0.61 0.01 - 0.01 1.16 0.86 1.02 1.23 1.27 1.13 0.97

SI - 7.7 - 6.4 - 4.6 - 6.9 - 5.3 - 6.5 - 7.2 - 6.3 - 6.5 - 14.7 - 10.9 - 12.8 - 10.5

XRF analyses [ppm] Nb 79.5 74.6 70.2 70.1 65.5 74.3 95.3 74.9 93.9 105 104 103 101

Zr 240 224 218 224 211 222 280 221 280 311 317 320 316

Sr 1144 1076 1051 1072 1061 1088 1353 1137 1342 1323 1317 1387 1465

Ga 17.8 18.1 17.7 18.1 17.2 17.1 18.1 17.4 19.8 20.1 20.4 20.9 20.8

Rb 45.0 41.0 37.4 40.0 37.3 42.4 38.2 46.0 37.3 75.9 63.0 78.7 77.3

Zn 96.5 101.2 97.8 101 97.3 100 112 112 112 116 115 113 116

Cu 27.2 44.5 67.7 50.4 46.6 47.9 43.6 47.5 43.4 39.8 40.1 43.2 40.5

Ni 323 280 298 268 243 320 201 299 210 163 160 145 138

Co 46.6 43.5 46.5 44.5 41.7 45.5 35.5 45.3 39.3 37.6 39.8 39.3 37.8

Cr 394 391 403 413 371 414 252 389 260 197 196 160 180

Sc 26.0 28.2 25.6 28.1 29.0 26.0 26.8 25.6 26.2 27.9 28.3 27.9 28.0

V 242 248 240 251 234 245 214 225 218 247 244 243 245

Ba 535 535 497 495 493 538 517 583 521 642 630 655 628

Mo 8.9 4.9 5.9 5 5.1 6.5 10 6.5 10.2 4.6 10.3 7.9 8.2

¡CP-MS analyses [ppm] Y 25.96 24.97 24.76 25.91 25.54 24.92 30.44 25.94 30.43 29.73 31.08 31.33 31.4

Cs 0.61 0.58 0.55 0.57 0.52 0.58 0.68 0.59 0.65 0.69 0.72 0.73 0.74

La 47.68 43.31 41.50 44.02 41.46 43.62 57.91 45.85 57.08 54.23 56.80 57.23 57.66

Ce 92.3 84.3 81.4 86.3 80.4 85.3 115.4 89.3 113.6 106.7 112.3 113.3 114.2

Pr 10.92 10.11 9.70 10.39 9.66 10.19 13.62 10.66 13.45 12.87 13.57 13.63 13.72

Nd 45.61 41.89 40.44 43.12 40.17 42.22 56.31 44.37 55.57 53.26 56.04 56.59 57.08

Sm 8.94 8.36 8.26 8.77 8.20 8.48 11.16 8.86 11.05 10.54 11.17 11.22 11.28

Eu 2.74 2.60 2.55 2.69 2.55 2.57 3.34 2.72 3.31 3.22 3.34 3.37 3.40

Gd 8.04 7.62 7.47 7.84 7.50 7.75 10.25 7.94 10.15 9.48 10.02 10.00 10.20

Tb 1.05 1.01 0.99 1.05 1.00 1.02 1.32 1.05 1.32 1.23 1.31 1.30 1.30

Dy 5.17 5.04 4.90 5.19 5.04 5.04 6.58 5.18 6.39 6.01 6.33 6.42 6.37

Ho 0.93 0.87 0.88 0.91 0.89 0.88 1.17 0.92 1.16 1.05 1.10 1.12 1.10

Er 2.26 2.16 2.15 2.26 2.22 2.16 2.73 2.26 2.68 2.56 2.68 2.71 2.73

Tm 0.28 0.26 0.27 0.27 0.27 0.26 0.32 0.27 0.31 0.31 0.32 0.32 0.31

Yb 1.56 1.58 1.53 1.56 1.63 1.55 1.83 1.56 1.80 1.75 1.90 1.85 1.85

Lu 0.21 0.21 0.20 0.21 0.22 0.20 0.25 0.21 0.25 0.23 0.24 0.24 0.25

Hf 4.79 4.54 4.44 4.67 4.28 4.48 5.74 4.73 5.66 5.64 5.95 6.03 6.02

Ta 2.96 2.77 2.24 2.74 2.25 2.74 3.63 2.85 2.95 3.60 4.38 3.98 3.85

Th 5.78 5.28 5.05 5.20 4.95 5.26 7.22 5.49 7.12 6.36 6.65 6.73 6.60

U 1.9 1.8 1.7 1.7 1.7 1.7 2.4 1.8 2.4 2.2 2.5 2.2 2.1

Pb 2.2 1.8 1.8 1.8 1.3 2.2 2.3 2.0 2.3 1.6 1.9 2.0 1.2

Measured isotope ratios 87Sr/86Sr 0.703065 0.703154 0.703088 0.703123 0.703119

143Nd/144Nd 0.513039 0.513028 0.513035 0.513035 0.513037

FeOt = Total amount of iron; LOI = Loss on ignition.

229-243x01, Lu: 9,1-9,7x01), with the exception of basanite samples VL2-T2 and VL2-T4 (La: 241-244x01, Lu: 9,8x01) which have similar REE abundances as VL4 and VL5 lavas, Beyond this, VL1 basanites have a minor spread of LREE concentrations, shown by varying (La/Yb)N ratios (18,2-22) whilst in the nephelinites the concentrations do not vary significantly ((La/Yb)N: 22,2-22,4), All samples show very similar HREE patterns, with (Tb/Yb)N ranging from 2,8 to 3,3, Eu anomalies are generally absent,

The primitive mantle (PM) normalized multi-element diagram (Fig, 3d) again illustrates the higher abundances of the highly to

moderately incompatible elements in the nephelinites and the slightly lower concentrations in the basanites, again with the exception of VL2-T2 and VL2-T4. In general, all analyzed samples have PM-normalized patterns with slight negative anomalies in Th and Zr as well as positive anomalies of Nb, Sr and P with respect to their neighboring elements. The nephelinites differ from all the basanites (including VL2-T2 and VL2-T4) in being enriched in trace elements, except for Rb, Ba, K and Ti. In all samples, a general trend of increasing normalized element concentration with increasing incompatibility is observed until K; for the most

Fig. 3. Bulk trace and rare-earth elements of the VVF alkaline lavas. (a) MgO vs. Ni; (b) MgO vs. Zr/Nb; (c) REE diagram normalized against chondrite composition (C1; Sun and McDonough, 1989); (d) multi-element diagram normalized against primitive mantle composition (PM; McDonough and Sun, 1995); Late Cenozoic alkaline rocks from the Bering Sea Volcanic Province (dashed field) and basanites from the Prindle Volcano (PV, crosses) are given for comparison (Wirth et al., 2002; Anrdonikov and Mukasa, 2010).

incompatible elements (Nb-Rb), the gradient remains fairly constant (Fig. 3d).

5.4. Sr-Nd isotope chemistry

Measured 78Sr/86Sr and 143Nd/144Nd isotope ratios are reported in Table 1; initial ratios are plotted in Fig. 4, calculated to 9 Ma for all

Fig. 4. Sr and Nd isotope compositions of the rocks from the VVF. 87Sr/86Srj plotted against 143Nd/144Ndi, based on an age of 9 Ma. DMM, PREMA, HIMU and BSE as well as OIB and MORB fields are after Zindler and Hart (1986), Hofmann (1997) and Stracke et al. (2005). Isotopic signatures of lavas from former studies in the VVF and the Enmelen Volcanic Fields (Apt et al., 1998) are also plotted for comparison as well as Sr and Nd isotopes of spinel-lherzolite xenoliths (Sp-Lh) from the VVF (Ntaflos et al., 2008) and Late Cenozoic alkaline volcanic basalts (PV) from the Bering Sea Volcanic Province Anrdonikov and Mukasa (2010) and Pliocene to Holocene basaltic volcanics (PI) from the Pribilof Islands, Alaska (Chang et al., 2009).

lavas. Both basanites and nephelinites plot within a very small field, with 78Sr/86Sr varying from 0.703065 to 0.7030154 and 143Nd/144Nd from 0.513028 to 0.513039. Epsilon Nd (sNd) ranges from 7.7 to 7.9. The more isotopically depleted samples plot close to the PREMA (prevalent mantle) field, between HIMU (high and DMM (depleted MORB mantle), at the intersection of MORB (mid ocean ridge basalt)-and OIB (ocean island basalt)-type melts (Hofmann, 1997; Stracke et al., 2005; and Zindler and Hart, 1986). Apt et al. (1998) reported similar Sr and Nd isotopic ratios from the Viliga and Enmelen volcanic fields. The clinopyroxene isotopic signatures from spinel-lherzolite xenoliths in the VVF are not related to the investigated melts as they are more depleted in radiogenic isotopes (143Nd/144Nd: 0.5132340.513622, Ntaflos et al., 2008).

6. Pedogenesis of the rocks from the Viliga Volcanic Field

The basanites and nephelinites from the VVF indicate relatively primitive compositions, having both high compatible element contents and mantle xenoliths that appear almost unaffected by the host lavas. However, the major and trace element compositions indicate that the rocks do not represent primary magmas and that there were minor but significant differences in the melt generating processes between basanites and nephelinites. Petro-graphic, mineral and bulk geochemical evidence suggests that the basanites are a co-genetic suite of rocks, while the nephelinites represent melt from a slightly deeper but very similar source. In the following sections, both the petrogenetic processes, including the degree and depth of partial melting, fractional crystallization and crustal contamination and the mantle source that produced the lavas are considered.

6.1. Crustal contamination

Prior to the assessment of source characteristics and subsequent magmatic differentiation processes, the amount of crustal contamination affecting the samples must be evaluated. The primitive basanitic and nephelinitic magmas of the VVF have Zr/Nb ratios between 2.9 and 3.2, typical for alkaline rocks with OIB character (Weaver, 1991). Due to higher Zr/Nb ratios in crustal rocks, alkaline lavas contaminated by continental crust would have significantly higher values (Taylor and McLennan, 1985). La/Nb and K/P ratios in magmas affected by substantial degrees of crustal components normally reach 1.5 and 7 respectively (Hart, 1988), in contrast to the rocks of this study, which have much lower values (La/Nb: ~0.6 and K/P: ~4.6). Fig. 5 (Nb vs. Nb/U) clearly shows the similarity of the basanites and nephelinites from the VVF to OIB-type, mantle derived lavas. The range in Nb/U ratios (39-49) of the VVF lavas does not show any evidence for crustal contamination. Moreover, a positive linear correlation between SiO2 and 87Sr/86Sr, commonly regarded as key for crustal contamination (Chang et al., 2009), is completely absent in the studied rocks from the VVF.

The geochemical evidence clearly indicates that no or at least only insignificant assimilation of continental crustal rocks occurred during the rapid ascent of the lavas. This includes the ratios of the trace elements discussed previously, the mantle source characteristics (high Nb, low Zr, Th and K,), the depleted 87Sr/86Sr and 143Nd/144Nd isotopic signature, and also the lack of crustal xenoliths or xenocrysts in the rocks.

6.2. Fractional crystallization

The petrographic investigation of the studied samples suggests that olivine fractionation was the main magma differentiation process in the basanites, while in the nephelinites the fractionation of olivine together with minor clinopyroxene phenocrysts was important. Theoretical olivine fractionation trends were calculated for the major elements (after Pearce, 1978; plotted in Fig. 2), starting at sample VL1-40, which has the most primitive composition (MgO: 12.8 wt.%). For the most evolved sample from group VL1 (MgO: 11.7 wt.%), modeling suggests that 3.2 % olivine has been fractionated; the most differentiated sample fromVL2 (MgO: 10.34 wt.%) requires 7% olivine fractionation. With ongoing differentiation, the increasing degree of olivine extraction led to the observed enrichment of incompatible elements from VL1 to VL2 (Fig. 3c, d). Linearly decreasing amounts of Ni and near-constant Zr/Y at decreasing MgO, both features of the studied basanites, also suggest olivine fractionation. Mineral analyses show that during differentiation from VL1 to VL2, the core composition of olivine phenocrysts evolves to

1 1 1 1 ii111 1 1 1 1 1 1 ll| 1 iii 1 1 11| :

_ MORB oib :i

- c-chondrites -

l____1

continental

vl2 □

- vl4 q -

........1 i i i i i i i il i iii .....

,1 1 10 100 Nb [ppm]

Fig. 5. Nb/U vs. Nb for the basanites and nephelinites from the Viliga Volcanic Field. Figure modified after Hofmann et al. (1986).

lower MgO core concentrations, which is consistent with the observed bulk geochemical trends.

Compared to the basanites, the nephelinites have low SiO2 and Al2O3 as well as high CaO, FeO and TiO2 contents. These compositions cannot be explained by olivine fractionation from the basanitic magma suite. This is confirmed by the CaO/Al2O3 ratios in the basanites, which are both slightly lower than in the nephelinites and show a flat trend when plotted against MgO (Fig. 2g). Although petrographically some rare clinopyroxene phenocryst grains can be found in the nephelinites, no geochemical evidence (decreasing CaO/ Al2O3 or Sc, or increasing Zr/Nb at decreasing MgO) for clinopyroxene fractionation has been observed.

Plagioclase fractionation can be excluded for all investigated samples, because neither Eu anomalies nor decreasing Sr contents with decreasing SiO2 or Al2O3 have been found. Nor was Fe-Ti oxide fractionation important, since with lowering MgO concentrations, FeO and TiO2 concentrations remain constant. Minor amounts of low-pressure leucite fractionation, though not observable petrographical-ly, could be a reason for the lower K and Rb concentrations in the basanites compared to the nephelinites, as these elements partition into leucite (Platz et al., 2004 and references therein).

6.3. Mantle source of the Viliga Volcanic Field lavas

Silica-undersaturated alkaline magmas such as the basanites and nephelinites in this study are normally related to high pressure, small volume melting of a garnet peridotite source (Kushiro, 1996). Both garnet lherzolite and pyroxenite melting can produce alkaline magmas with low silica contents (Keshav et al., 2004; Walter, 1998) and the high LREE/HREE ratios (eg. La/Yb) of the studied lavas suggest that they were produced by low fraction melting of a garnet-bearing source. MREE/HREE ratios are even better proxies for garnet melting, such that, for example, (Dy/Yb)N ratios < 1.06 indicate melts derived from a spinel peridotite source, whereas ratios > 1.06 characterize melts from a garnet peridotite source (Blundy et al., 1998). Both the basanites and nephelinites from the VVF have (Dy/Yb)N ratios ranging between 2.1 and 2.4. The higher (Dy/Yb)N ratios are restricted to the nephelinites, indicating smaller melt fractions for those samples. This is in accordance with the lower SiO2 and higher CaO and TiO2 concentrations in the nephelinites, as well as their trace element abundances, which also indicate very small degrees of partial melting at even greater depths, than the basanites (Hirose and Kushiro, 1993). In terms of their trace element signatures both the nephelinites and the basanites resemble average OIB compositions, most likely HIMU (high |j)-OIB (Hofmann, 1997; Weaver, 1991). The Sr and Nd isotopes of the VVF lavas span a relatively small range, which is very similar to data presented by Apt et al. (1998) for the VVF and Enmelen volcanic fields (see Fig. 4). The isotopically relatively depleted lavas plot where present-day MORB and OIB overlap, close to the HIMU- and PREMA (prevalent mantle)-type sources. Trace element ratios such as Nb/U (see Fig. 5) but also Ba/Ce or Ba/Nb average around 5.8 (SD: 0.6) and 6.6 (SD: 0.75) respectively in the studied rocks, which are within the range of typical OIB values (Ba/Ce: 3.2-5.8; Ba/Nb: 5.6-10.2); MORB values, in contrast, range from 0.5 to 1.7 for Ba/Ce and from 1.9 to 5.1 for Ba/Nb (Halliday et al., 1995). Taking both trace element patterns and radiogenic isotopes into account, the lavas reflect melting of a single source similar to that of depleted continental OIB-type basalts, rather than the source of MORB (Hofmann, 1997; Stracke et al., 2005). The fertile spinel lherzolitic mantle xenoliths from the area give an even more depleted isotope signature that was interpreted to reflect a 2-3 Ga old mantle, rather than a source that is depleted due to MORB extraction (Ntaflos et al., 2008). Sr-Nd isotopic ratios are similar to those of rocks from the Bering Sea Basaltic Province (see Fig. 4— samples from the Prindle Volcano and the Pribilof Islands, both Alaskan, are plotted for comparison). Pb isotope systematics suggest that the isotopic composition of the young Alaskan basalts reflects a

MORB source that fairly recently (< 170 Ma) was metasomatized by a young HIMU component, and was not influenced by pelagic or continental sediments (Chang et al., 2009). In the Bering Sea, such a model is feasible, since subduction and the related dehydration of the slab has occurred continuously since the Paleocene.

6.4. Meltfacies and degree of partial melting

For qualitative and semi-quantitative source and partial melting modeling, REE systematics were used. Diagrams plotting LREE and HREE against MREE and HREE (e.g. La/Yb-Dy/Yb, see Fig. 6a) allow a distinction to be made between melting in the garnet or spinel peridotite stability fields (Baker et al., 1997). If both sources contribute to the melting process, the melts produced should lie on tie-lines connecting the two sources.

All samples indicate a strong fractionation of HREE (Dy/Yb >3), typical of magmas derived from garnet peridotite. The varying ratios of (La/Yb)N from 18.2 to 22.7 in the basanites are interpreted to be a result of slightly different degrees of partial melting of the same source, since all other trace element concentrations remain almost identical. Semiquantitative fractional melt modeling of a primitive mantle composition was carried out to determine whether different degrees of partial melting could have been responsible for the observed element trends (see Fig. 6). The observed linearity for La/ Yb vs. Dy/Yb (Fig. 6a) cannot be explained by fractional crystallization or crustal contamination processes. The positive correlation of the

analyzed samples indicates that all the basanites and nephelinites originated from a source having a very similar REE signature and, that the nephelinites with higher La/Yb and Dy/Yb represent smaller degrees of partial melting compared to most basanites. The highest CaO/Al2O3 values in the nephelinites also correlate with the lowest calculated melt fractions.

The melting models (after Shaw, 1970) presented in Fig. 6a and b were performed to assess the nature of partial melting that formed the VVF rocks as well as to theoretically assign the melts to distinct sources. The primitive mantle composition of McDonough and Sun (1995) was taken as a starting material that melted from 0.5 % to 20% under garnet as well as under the spinel peridotite field conditions; exact modal compositions of the source and melt proportions are given in the figure caption of Fig. 6. The La/Yb vs. Dy/Yb and La/Yb vs. Yb diagrams (Fig. 6a, b) indicate that the generation of the VVF rocks cannot be explained by variable degrees of partial melting of an exclusively garnet-bearing or a spinel-bearing lherzolite source. The suite lies on a near linear mixing line between theoretical garnet and spinel-bearing source melts, so that the simplest melt model involves an initial major partial melting in the garnet stability field followed by mixing of these melts with small melt fractions of spinel peridotite. This model, adapted from Baker et al. (1997), suggests that a 1 % melt of garnet peridotite mixed with a melt produced by 0.5 % melting of spinel peridotite. The theoretical mixing proportions lie between 9095% melt of garnet peridotite and 5-10% of spinel peridotite. The samples from VL1 indicate slightly higher degrees of partial melting

Fig. 6. (a, b) Qualitative and semi-quantitative partial melting models based on La/Yb vs. Dy/Yb and La/Yb vs. Yb. Fractional melts of garnet peridotite (0.598 olivine, 0.211 orthopyroxene, 0.076 clinopyroxene, 0.115 garnet) and spinel peridotite (0.578 olivine, 0.270 orthopyroxene, 0.119 clinopyroxene, 0.033 spinel) that melt in the proportions 0.05 olivine,0.20 orthopyroxene,0.3 clinopyroxene, 0.45 garnet and 0.1 olivine, 0.27 orthopyroxene, 0.5 clinopyroxene, 0.13 spinel (Thirlwall et al., 1994) with partition coefficients from McKenzie and O'Nions (1991). Tie lines indicate mixing between large fractions from the garnet-facies mantle with small melt fractions from the spinel-facies mantle. (c) (La/ Yb)N vs. P, calculated with three different geobarometers. Lines show linear regressions for melt segregation pressures of each series; full symbols outline the calculations after Scarrowand Cox (1995) including also the nephelinic lavas from VL4 and VL5. (d) T vs. P illustrates the melt segregation temperatures and pressures for the analyzed basanites based on two different geothermobarometers. Full symbols show PT estimates after Albarede (1992) and Putirka (2005) respectively; open symbols PT estimates after Lee et al., 2009. Dry lherzolite solidus from Katz et al. (2003).

and moreover, somewhat larger melt fractions from the spinel stability field. Mixing of these two upper mantle melts is also documented by the negative correlation of La/Yb with Yb (Fig. 6b) in the basanites, with the same range of mixing proportions. Due to higher Yb concentrations (1.8-1.9 ppm) in the nephelinites and the two most differentiated basanite samples, these rocks do not plot exactly on the mixing line defined by most of the basanites with Yb ranging between 1.5 and 1.6 ppm. Although these melting models are only theoretical approaches and strongly dependent on the starting mantle composition as well as the partition coefficients, we conclude that the analyzed lavas are products of small volume partial melting of a garnet peridotite source and that the melts mixed with melts from spinel-lherzolite. In addition, the nephelinites, which do not plot exactly on the melting trends, indicate lower degrees of melting, probably at higher pressures, compared to the basanites (see next Section 6.5).

Based on petrographic and geochemical observations, evidence for hydrous K-rich minerals (such as amphibole or phlogopite) that could have been present in the mantle source and contributed to the primary magma melting, has not been detected. Negative K anomalies in normalized incompatible multi-element diagrams are usually indicative of either K-rich phases that fractionated from the primary magmas or K-rich phases that remained in the source during melting. As neither phenocrysts of such hydrous phases nor negative K anomalies (with respect to the neighboring elements) are observed in the primitive mantle normalized diagram (Fig. 3d), the presence of amphibole or phlogopite during melting is unlikely. Rb/K2O values are relatively low and constant in the basanites (14-17) and systematically higher in the nephelinites (22-27); this is attributed to general source characteristics rather than to an amphibole-bearing residue. Following Herzberg and Asimow (2008) and Dasgupta et al. (2007), a minor CO2 influence (<2 wt.%) is implied by the calculated primary magmas of the VVF basanite suite, based on SiO2-CaO criteria. CO2 involvement can produce smaller melt fractions at lower temperatures than anhydrous peridotite, with a systematic increase of CaO at decreasing SiO2.

6.5. Temperature and pressure estimates of melting

Pressure/depth and temperature estimates ofmantle melting were carried out following the procedures described in Albarede (1992), Scarrow and Cox (1995), Putirka (2005—model A) and Lee et al. (2009). However, most of these barometers have not been calibrated for silica-undersaturated rocks. Due to this, only the Scarrow and Cox (1995) barometer was used to assess the nephelinite melt segregation pressures. For the basanites, which fulfil all requirements for the applied geothermobarometers, the yielded pressure estimates range from 33 to 37 kbar (Albarede, 1992), 34-38 kbar (Scarrow and Cox, 1995) and 44-50 kbar (± 2 kbar; Lee et al., 2009). All three show good positive linear correlations with each other. A particularly good overlap is obvious for the calculated pressure of Albarede (1992) and Scarrow and Cox (1995) while systematically ~10kbar higher pressures are given for the pressure estimates using Lee et al. (2009), (see Fig. 6c). Lee et al. (2009) noted that the pressure estimates from Albarede (1992) are underestimated by up to 10 kbar, when pressures >30 kbar are obtained. However, the studied basanites are close to nephelinite composition (undersaturated in SiO2), so that the calculated high pressures from Lee et al. (2009) are likely to be overestimated. Based on the good agreement of the two individual barometers of Albarede (1992) and Scarrow and Cox (1995), we infer a pressure range of 33-38 kbar for the basanites from the VVF. For the nephelinite samples of this study, Scarrow and Cox (1995) barometry gives 42-44 kbar; even when treated with caution, this clearly indicates a deeper source compared to the basanitic samples of the suite. Furthermore, the nephelinites have the highest FeO, CaO and trace element contents within the studied rock suite, at

substantially lower SiO2 concentrations compared to the basanites, which also reflect deeper melt segregation levels (Hirose and Kushiro, 1993).

The calculated temperatures for the basanites VL1 and VL2 give a narrow field of 1495-1509 °C according to Putirka (2005) and 15371563 °C according to Lee et al. (2009). Fig. 6d compares the PT conditions of the analyzed basanites, based on two different PT calculations. The pressure conditions based on Albarede (1992), which were converted to depth using the method of Scarrow and Cox (1995), yielded 104-116 km for the basanites (and 150-164 km for the nephelinites; see Fig. 6d). Such segregation temperatures and pressures in a continental intraplate setting indicate partial melting in the garnet peridotite field either caused by a deep-seated thermal anomaly (e.g. plume) or, as in this case, by extension of a thick continental lithosphere that either buffers heat beneath continents due to lithospheric insulation or mantle return flow due to the preceding subduction (Lee et al., 2009 and ref. therein). Even if a slight mixing with spinel peridotite facies melts can be modeled for the studied rocks, the lavas retain the major and primary melt conditions of the garnet peridotite facies, so that the calculated PT estimates may be underestimates.

6.6. Polybaric melt evolution

Plotting (La/Yb)N against the estimated P conditions (Fig. 6c) gives a positive linear trend, illustrating the correlation between the pressure of melting (based on major element mass-balance solutions) and the degree of fractional melting (based on REE systematics). The basanite petrogenesis, therefore, suggests that the samples segregated at deeper levels reflect lower degrees of partial melting, compared to the samples from shallower depths that indicate slightly larger melt fractions. This mirrors a polybaric melt evolution - with increasing degrees of partial melting at decreasing depth - of a common garnet peridotite source and apparently minor mixing with spinel peridotite partial melts during ascent, most likely en-route or during very shortlived "pooling" periods. The nephelinites ((La/Yb)N between 21 and 22) also indicate low fraction melting, at higher depths though, but still following the trend of higher melt fractions triggered through progressive pressure decrease. Since the isotopic Sr and Nd signal of the melts shows no significant differences, the melt fractions of the spinel peridotite field that contributed to the garnet-facies melts were either too marginal to influence the isotopic signature of the lavas or the melt contributing garnet- and spinel-facies mantle sources were isotopically fairly homogeneous.

7. Magmatism and regional tectonics

The Late Neogene mantle xenolith-bearing basanites and nephe-linites from the VVF are similar to the melanephelinites, tephrites and basanites analyzed by Apt et al. (1998) from the Viliga and the Enmelen (Chukotka Peninsula) volcanic fields in their major and trace element and Sr-Nd isotopic composition (Figs. 2a and 4). Additionally, the combined field of Late Neogene alkaline basalts from northeast Russia is similar in major element composition to the VVF lavas (Fig. 2a). The Neogene volcanic rocks, at least from the Viliga and Enmelen volcanic fields, formed within the same period and in a very similar tectonic setting. They also have an almost identical major element composition and Sr and Nd isotope signature, so that they are quite likely to be derived from the same or a very similar garnet peridotite source and erupted at contemporary associated volcanic events.

The trace element and isotope geochemistry of the young alkaline lavas from the VVF do not indicate any significant influence of Paleo-Pacific slab components, which previous studies and regional geological constraints suggest should be preserved below the surrounding Cretaceous arc (Hourigan and Akinin, 2004). In contrast,

late Cretaceous (~81-83 Ma) extrusive rocks (mainly basalts, basaltic andesites and rhyolites) from the Ola volcanic field (close to Viliga, approx, 130 km N from Magadan) directly post-dating the Paleo-Pacific subduction activity, clearly contain subduction-related components (Hourigan and Akinin, 2004; Leitner etal,, 2010), The slightly younger Mygdykit basalts (74-78 Ma), also cropping out within the O0VB, were formerly also attributed to late stage subduction-related volcanism, However, their distinctly younger age compared to the underlying calc-alkaline series and the fact that these capping basalts already have intraplate geochemical features indicate that they are more likely related to post-subduction extension than being directly subduction-related (Hourigan and Akinin, 2004 and ref, therein),

Based on our geochemical and radiogenic isotope observations as well as calculations of PT conditions of melting, we infer that the petrogenesis of the studied rock suite is related to post-subduction passive mantle upwelling in an extensional regime, rather than to melt generation through local heat anomalies, The cessation of the subduction (87 Ma, Hourigan and Akinin, 2004) and the following "piecemeal breakup" (Ntaflos et al,, 2008) of the slab, combined with extension due to subsidence and back-rotation of the oceanic crust, provided space between the slab pieces and thus facilitated adiabatic upwelling of the asthenosphere, combined with decompression melting and ascent of the studied lavas, The breakup of the northeastern Asian continental margin along seismically highly active belts and major faults, situated at plate boundaries and along the Sea of Okhotsk margin, occurred simultaneously with rare intraplate alkaline volcanism in the area, Seismic studies from the area indicate a thin crust (30-35 km) at the Okhotsk margin as well as beneath the OOVB and transparency on seismic profiles (lack of reflectivity) under the Omsukchan trough (Goryachev et al,, 2007; Nikolaevsky, 1967), The N-S trending rift-type structural lineament of the Omsukchan trough is closely related to the evolution of the OOVB (Sidorov et al,, 2009), as seen on seismic profiles from almost 200 km further north of the VVF (0-0' profile from Goryachev et al, (2007) is shown in Fig, 1), The lack of seismic reflection is interpreted as thinning of the lithosphere combined with the upwelling of asthenospheric material, Post-subduction extension related intraplate alkaline magmatism has been described in a number of plate margin settings (Fiji, Mexican Volcanic Belt, New Zealand, Western U,S,; Wirth et al,, 2002 and references therein), They are distinct from continental rift settings as they do not include felsic magmatism, 0omparing the lavas from this study to alkaline rocks (basanites and alkali-olivine basalts) from Late 0enozoic volcanic fields of the Bering Sea (Bering Sea Volcanic Province), a strong similarity is evident in both the major and trace elements as well as the Sr and Nd isotopes (see Figs, 2a, 3d and 4), These younger, isolated but widely dispersed intraplate volcanic centers in the Bering Sea region have been proposed to be the products of small fraction melts of a fertile mantle, Furthermore, they reflect a very similar composition and emplacement history as the Late 0enozoic volcanics of Southeast Asia (collectively termed "diffuse igneous province"), which are also related to tectonic extrusion and extension along regional faults (Hoang et al,, 1996; Wirth et al,, 2002), In the North 0ordilleran Volcanic Province, Alaska, 0enozoic intraplate volcanic rocks can be found that are compositionally analogous to the rocks from the VVF, Basanites from the Prindle Volcano (PV), for example, are slightly less alkaline in nature, but have almost identical trace element patterns and isotopic compositions (Figs, 2a, 3d, and 4), Based on their composition and their mantle xenolith-bearing nature, the lavas from Prindle Volcano are thought to be near-primary magmas (Anrdonikov and Mukasa, 2010), Late 0enozoic volcanism in this region is attributed to extension, triggered by the change of relative plate motion between the North-American and the Pacific Plates and related faulting, Extension ofthe lithosphere resulted in upwelling asthenosphere, and episodic decompression melting from the Late Tertiary to Quaternary, This melting was localized at plate boundary regions but also in interior Alaska

(Anrdonikov and Mukasa, 2010), Even though produced in an extensional tectonic environment due to decompression melting of a garnet-bearing asthenospheric source, a minor influence of subduction-related fluids of the North Pacific lithosphere beneath Alaska has not been entirely excluded during the genesis of the Prindle Volcano basanites (Anrdonikov and Mukasa, 2010),

The formation of the very young (Pliocene to Holocene) alkaline volcanism on the Pribilof Islands (Alaska) is reported as being both related to lithospheric extension inducing passive upwelling and decompression melting of a metasomatized upper mantle (0hang et al,, 2009), Even though the extensional peak in the Bering Sea region occurred around Middle- to Late-0retaceous times (Dumitru et al,, 1995), the regionally active tectonic movements (normal faults, Basin and Range type tectonics) were related to extension until recent times, As postulated for the magmatic activity in the VVF, mantle plume activity was unlikely to have been involved in the basalt formation on the Pribilof Islands (including also the rest of the Bering Sea Basaltic Province) but passive mantle upwelling controlled by the present ex- or transtensional geodynamic features (0hang et al,, 2009), Apart from their petrogenetic and geodynamic similarities, the basalts, trachybasalts and basanites from the Pribilof Islands are compositionally similar to the rocks from this study, mainly confirmed by their overlapping Sr and Nd istopic signatures (see Fig, 4),

The 0enozoic intraplate extensional tectonics and its related magmatism further south, along the whole East Asian continental margin (from northeast Russia to eastern 0hina), are also proposed to be linked to the subduction of the Pacific Plate under Eurasia (Pirajno et al,, 2009), However, the magmatism is not restricted to the East Asian continental margin but extends inland from the Eurasian Plate boundary, Here alkaline volcanic series are generally situated in subduction-linked, fault-dominated settings, suggesting intraplate mantle processes as a cause of their emplacement (Pirajno et al,, 2009), 0rustal extension and fault-related extrusion, lithospheric thinning and asthenospheric mantle upwelling undoubtedly are the key driving forces for plutonic and volcanic activity in the North Pacific adjacent to continental margin settings during the 0enozoic period, The magmatic activity in the Viliga Volcanic Field can be attributed to similar geodynamic features,

8. Concluding remarks

- According to their major and trace element geochemical composition, the lavas from the Viliga Volcanic Field represent a co-genetic suite of intraplate OIB-type basanites and nephelinites,

- Though intruding the calc-alkaline 0retaceous continental arc rocks of the Okhotsk-0hukotka Volcanic Belt, the Neogene alkaline lavas of the VVF do not contain subduction-related geochemical or isotopic features,

- Sr and Nd isotope ratios, as well as trace element systematics suggest melting of the studied lavas from a homogeneous garnet-bearing mantle source (at the depleted end of the OIB array), without a contribution from K-rich phases,

- Mantle melt modeling and PT estimates reveal that the basanites from the VVF were produced by partial melting under garnet peridotite facies conditions (33-38 kbar and ~1500 °0) and mixed with minor melt proportions from the spinel peridotite field, The nephelinites clearly reflect melting at deeper levels (>40 kbar),

- The VVF lavas were produced by passive mantle upwelling that triggered decompression melting, Evidence of polybaric melt formation can be observed in the studied rocks, Lavas with higher melt segregation pressures systematically reflect lower degrees of partial melting and co-genetic lavas with lower segregation pressures reflect larger melt fractions,

- En-route, the basanites experienced up to 7% olivine fractionation, Olivine together with some minor clinopyroxene fractionation is

observable only in the nephelinites. Contamination of the lavas during ascent by continental crust components cannot be verified.

- Similar to many of the other volcanic fields around the northern Pacific that are adjacent to continental margins, the genesis of the VVF was most probably triggered by post-subduction extension, crustal-scale plate reorganization, adiabatic asthenospheric mantle upwelling and decompression melting, resulting in the episodic and sporadic construction of intraplate volcanic provinces during Cenozoic. Large-scale trans- and extensional tectonics in the area around the Viliga Volcanic Field are still documented by a number of highly active faults and strong seismicity.

Supplementary materials related to this article can be found online at doi:10.1016/j.lithos.2010.11.009.

Acknowledgements

We thank Massimo Coltorti, Mikhail Parfenov and Vera Leonova for the productive collaboration in the field and the residents of the village Merenga for their invaluable hospitality after our disastrous truck breakdown in the Viliga River. Without the assistance of Evsukov Vyacheslav and Sidak Viketr, who helped save our samples from the drowned Ural truck, and the professional manor of our driver, Igor Cheremnych, who struggled for his and our lives, this study wouldn't have been possible. Sr and Nd isotope analyses from our colleagues Monika Horschinegg and Martin Thoni are also gratefully acknowledged. The work was supported by the FWF grant I2001-N22 (Th.N.) and the UNESCO grant IGCP-430 (Th.N.) as well as grant DVO 09-1-P16-11, RFBR 09-05-91005 (V.V.A.). We are grateful to Michael Roden and Szabolcs Harangi for their thoughtful reviews, to Andrew Kerr for his constructive comments, and to Hugh Rice for linguistic improvement of the original manuscript.

References

Akinin, V.V., Apt, J.E., 1994. Enmelen volcanoes, Chukchi Peninsula: petrology of alkaline lavas and deep-seated inclusions. Magadan NEISRI FEB RAS 97 p. (in Russian).

Akinin, V.V., Apt, Y.E., 1997. Late Cenozoic alkaline basaltic volcanism of the North East of Russia. In: Byalobzhesky, S.G. (Ed.), Magmatism and Ores of North East of Russia, pp. 153-172.

Akinin, V.V., Miller, E.L., Layer, P., 2005. Late Cretaceous modification of deep continental crust in the NE Paleo Pacific: additional evidence from Viliga lower crust xenoliths: EOS Transactions 86. Fall Meet. Suppl. 52, V51D-V1516D. Akinin, V.V., Evdokimov, A.N., Korago, E.A., Stupak, F.M., 2008. Recent volcanism of arctic margin of northern Eurasia. In: Kovalenko, V.I., Yarmolyuk, V.V., Boratikov, O. A. (Eds.), Changes of Natural Environment and Climate. : Recent Volcanism of Northern Eurasia: Evolution, Volcanic Hazard, and Connection with Deep Processes and Environmental and Climate Changes, Volume 2. IGEM RAN, Moscow, pp. 41 -80 (in Russian).

Albarede, F., 1992. How deep do common basaltic magmas form and differentiate? J.

Geophys. Res. 97, 10997-11009. Anrdonikov, V., Mukasa, S.B., 2010. 40Ar/39Ar eruption ages and geochemical characteristics of Late Tertiary to Quaternary intraplate and arc-related lavas in interior Alaska. Lithos 115, 1 -14. Apt, Y.E., Akinin, V.V., Wright, J.E., 1998. Sr, Nd, and Pb isotopes in Neogene melanephelinites and deep-seated xenoliths from northeast Russia. Geochem. Int. 36 (1), 24-33.

Baker, J.A., Menzies, M.A., Thirlwall, M.F., MacPherson, C.G., 1997. Petrogenesis of Quaternary intraplate volcanism, Sana'a, Yemen: implications for plume-lithosphere interaction and polybaric melt hybridization. J. Petrol. 38, 1359-1390. Blundy, J.D., Robinson, J.A.C., Wood, B.J., 1998. Heavy REE are compatible in clinopyroxene on the spinel lherzolite solidus. Earth Planet. Sci. Lett. 160,493-504. Chang, J.M., Feeley, T.C., Deraps, M.R., 2009. Petrogenesis of basaltic volcanic rocks from the Pribilof Islands, Alaska, by melting of metasomatically enriched depleted lithosphere, crystallization differentiation, and magma mixing. J. Petrol. 50 (12), 2249-2286.

Dasgupta, R., Hirschmann, M.M., Smith, N.D., 2007. Partial melting experiments on peridotite + CO2 at 3 GPa and genesis of alkalic ocean island basalts. J. Petrol. 48, 2093-2124.

Dumitru, T.A., Miller, E.L., O'Sullivan, P.B., Amato, J.M., Hannula, K.A., Calvert, A.T., Gans, P.B., 1995. Cretaceous to recent extension in the Bering Strait region, Alaska. Tectonics 13, 549-563.

Fitton, J.G., James, D., 1991. Basic magmatism associated with Late Cenozoic extension in the Western United States: compositional variations in space and time. J. Geophys. Res. 96 (B8), 13693-13711.

Fujita, K., Koz'min, B.M., Mackey, K.G., Riegel, S.A., McLean, M.S., Imaev, V.S., 2009. Seismotectonics of the Chersky Seismic Belt, eastern Sakha Republic (Yakutia) and Magadan District, Russia. Stephan Mueller Spec. Publ. Ser. 4,117-145.

Goryachev, N.A., Byalobzhesky, S.G., Kuznetzov, V.M., Palymsky, B.F., Feldman, L.L., 2007. Peculiarities of deep structure of the continental margin arcs on example from northern onshore of Okhotsk Sea. In: Salnikov, A.S. (Ed.), Structure and Composition of Earth Crust in Magadan Sector of Russia Using Geological and Geophysical Data. Nauka, Novosibirsk, pp. 118-133.

Gunbina, L.V., Leschuk, N.M., Kurtkin, S.V., 2007. North-East Russian and Chukotka. In: Malovichko, A.A. (Ed.), Earthquakes of the Russia in 2004. Geophysical Survey of Russian Academy of Science. Obninsk pp. C. 37-39 (in Russian).

Halliday, A.N., Lee, D.-C., Tommasini, S., Davies, G.R., Paslick, C.R., Fitton, J.G., James, D.E., 1995. Incompatible trace elements in OIB and MORB and source enrichment in the sub-oceanic mantle. Earth Planet. Sci. Lett. 133, 379-395.

Hart, S.R., 1988. Heterogeneous mantle domains: signatures, genesis and mixing chronologies. Earth Planet. Sci. Lett. 90, 273-296.

Herzberg, C., Asimow, P.D., 2008. Petrology of some oceanic island basalts: PRIMELT2. XLS software for primary magma calculation. Geochem. Geophys. Geosyst. 9 doi: 10.1029/2008GC002057.

Hirose, K., Kushiro, I., 1993. Partial melting of dry peridotites at high pressures: determination of composition of melts segregated from peridotite using aggregate of diamonds. Earth Planet. Sci. Lett. 114,477-489.

Hoang, N., Flower, M.F.J., Carlson, R.W., 1996. Major, trace element, and isotopic compositions of Vietnamese basalts: interaction of enriched mobile asthenosphere with the continental lithosphere. Geochim. Cosmochim. Acta 60, 4329-4351.

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

Hofmann, A.W., Jochum, K.P., Seufert, M., White, W.M., 1986. Nb and Pb in oceanic basalts: new constraints on mantle evolution. Earth Planet. Sci. Lett. 79, 33-45.

Hourigan, J.K., Akinin, V.V., 2004. Tectonic and chronostratigraphic implications of new 40Ar/39Ar geochronology and geochemistry of the Arman and Maltan-Ola volcanic fields, Okhotsk-Chukotka volcanic belt, northeastern Russia. GSA Bull. 116 (5/6), 637-654.

Ichetovkin, N.V., Silinski, A.D., Fadeev, A.P., 1970. Quaternary alkaline basalts from Kananyga River and Viliga River (Okhotsk-Chukotsk volcanic belt). Geol. Geophysika 8,122-127 (in Russian).

Katz, R.F., Spiegelman, M., Langmuir, C.H., 2003. A new parameterization of hydrous mantle melting. Geochem. Geophys. Geosyst. 4 doi: 10.1029/2002GC000433.

Keshav, S., Gudfinnsson, G., Sen, G., Fei, Y., 2004. High-pressure melting experiments of garnet clinopyroxenite and the alkalic to tholeiitic transition in ocean-island basalts. Earth Planet. Sci. Lett. 223, 365-379.

Kushiro, I., 1996. Partial melting of a fertile mantle peridotite at high pressures: an experimental study using aggregates of diamond. In: Basu, A., Hart, S.R. (Eds.), Earth Processes: Reading the Isotopic Code. Geophysical Monograph: American Geophysical Union, 95, pp. 109-122.

Le Maitre, R.W., 2002. Igneous Rocks: A Classification and Glossary of Terms, Recommendations of the International Union of Geological Sciences Subcommission on the Systematics of Igneous Rocks, 2nd Edition. Cambridge University Press. 219 p.

Le Maitre, R.W., Bateman, P., Dudek, A., Keller, J., Lameyre, J., Le Bas, M., Sabine, P., Schmid, R., Sorensen, H., Streckeisen, A., Woolley, A., Zanettin, B., 1989. A classification of igneous rocks and glossary of terms. In: Le Maitre, R.W. (Ed.), Recommendations of the International Union of Geological Sciences Subcommission on the Systematics of Igneous Rocks. Blackwell. 193 p.

Lee, C.-T.A., Luffi, P., Plank, T., Dalton, H., Leeman, W.P., 2009. Constraints on the depth and temperatures of basaltic magma generation on Earth and other planets using new thermobarometers for mafic magmas. Earth Planet. Sci. Lett. 279, 20-33.

Leitner, J., Ntaflos, Th., Akinin, V.V., Tschegg, C., 2010. Geochemical and petrologic investigation of the Ola Plateau-basalts from the Okhotsk-Chukotka Volcanic Belt (NE Russia). Geophys. Res. Abstr. 12,12951.

Mackey, K.G., Fujita, K., Gunbina, L.V., Kovalev, V.N., Imaev, V.S., Kozmin, B.M., Imaeva, L.P., 1997. Seismicity of the Bering Strait region: evidence for a Bering block. Geology 25 (11), 979-982.

Matveenko, V.T., 1957. Petrology and basic metallogenesis of Omsukchan ore center (North-East Russia). Magadan VNII1 31 100 p. (in Russian).

McDonough, W.F., Sun, S., 1995. The composition of the Earth. Chem. Geol. 120, 223-253.

McKenzie, D., O'Nions, R.K., 1991. Partial melt distributions from inversion of rare earth element concentrations. J. Petrol. 32, 1021 -1091.

Nikolaevsky, A.A., 1967. Deep Structure of the Okhotsk-Chukotka Volcanic Belt. Regional Geophysical Investigations in Siberia. Nauka, Novosibirsk, pp. 137-148 (in Russian).

Nokleberg, W.J., Parfenov, L.M., Monger, J.W.H., Norton, I.O., Khanchuk, A.I., Stone, D.B., Scotese, C.R., Scholl, D.W., Fujita, K., 2000. Phanerozoic Tectonic Evolution of the Circum-North Pacific. Professional Paper 1626, U.S. Geological Survey. 122 p.

Ntaflos, Th., Tschegg, C., Coltorti, M., Akinin, V.V., Kosler, J., 2008. Asthenospheric signature in fertile spinel lherzolites from the Villiga Volcanic Field in NE Russian. In: Coltorti, M., Gregoire, M. (Eds.), Metasomatism in Oceanic and Continental Lithospheric Mantle Geological Society of London: Special Publications, 293, pp. 57-81.

Pearce, T.H., 1978. Olivine fractionation equation for basaltic and ultrabasic liquids. Nature 276, 771-774.

Pirajno, F., Ernst, R.E., Borisenko, A.S., Fedoseev, G., Naumov, E.A., 2009. Intraplate magmatism in Central Asia and China and associated metallogeny. Ore Geol. Rev. 35,114-136.

Platz, T., Foley, S.F., André, L., 2004. Low-pressure fractionation of the Nyiragongo volcanic rocks, Virunga Province, D.R. Congo. J. Volcanol. Geoth. Res. 136,269-295.

Putirka, K.D., 2005. Mantle potential temperatures at Hawaii, Iceland, and the midocean ridge system, as inferred from olivine phenocrysts: evidence for thermally driven mantle plumes. Geochem. Geophys. Geosyst. 6, 5 doi: 10.1029/2005GC000915.

Scarrow, J.H., Cox, G., 1995. Basalts generated by decompressive adiabatic melting of a mantle plume: a case study from the Isle of Skye, Nw Scotland. J. Petrol. 36 (1), 3-22.

Shaw, D.M., 1970. Trace element fractionation during anatexis. Geochim. Cosmochim. Acta 34, 237-243.

Sidorov, A.A., Volkov, A.V., Belyi, V.F., Alekseev, V.Y., Kolova, E.E., 2009. The Gold-Silver Okhostk-Chukotka Volcanic Belt. Geol. Ore Deposits 51 (6), 441-455.

Stracke, A., Hofmann, A.W., Hart, S.R., 2005. FOZO, HIMU, and the rest of the mantle zoo. Geochem. Geophys. Geosyst. 6, 5 doi: 10.1029/2004GC000824.

Sun, S., McDonough, W.F., 1989. Chemical and isotopical systematic of oceanic basalts: implications for mantle composition and processes. In: Saunders, A.D., Norry, M.J. (Eds.), Magmatism in the Ocean Basins. Geological Society Special Publication, Vol. 42, pp. 313-345.

Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: Its Composition and Evolution. Blackwell Scientific Publications. 312p.

Thirlwall, M.F., Upton, B.G.J., Jenkins, C., 1994. Interaction between continental lithosphere and the Iceland plume—Sr-Nd-Pb isotope chemistry of Tertiary basalts, NE Greenland. J. Petrol. 35, 839-897.

Thoni, M,, Miller, 0,, Blichert-Toft, J,, Whitehouse, MJ,, Konzett, J,, Zanetto, A,, 2008, Timing of high-pressure metamorphism and exhumation of the eclogite type-locality (Kupplerbrunn-Prickler Halt, Saualpe, south-eastern Austria): constraints from correlations of the Sm-Nd, Lu-Hf, U-Pb and Rb-Sr isotopic systems, J, Metamorph, Geol, 26, 561-581, Tschegg, 0,, Ntaflos, Th,, 0oltorti, M,, Akinin, V,V,, 2004, Mantle xenoliths from the Viliga volcanic field, N-E Siberia: evidence for a fertile subcontinental lithosphere, 32nd International Geological 0ongress Abstracts, Volume 2, p, 1153, Tschegg, 0,, Hein, I,, Ntaflos, Th,, 2008, State of the art multi-analytical geoscientific approach to identify 0ypriot Bichrome Wheelmade Ware reproduction in the Eastern Nile delta (Egypt), J, Archaeol, Sci, 35 (2008), 1134-1147, Tschegg, 0,, Ntaflos, Th,, Kiraly, F,, Harangi, Sz,, 2010, High temperature corrosion of olivine phenocrysts in Pliocene basalts from Banat, Romania, Austrian J, Earth Sci, 103/1,

Walter, MJ,, 1998, Melting of garnet peridotite and the origin of komatiite and depleted

lithosphere, J, Petrol, 39, 29-60, Weaver, P,P,E,, 1991, The origin of ocean island basalt end-member composition: trace-

element and isotopic constraints, Earth Planet, Sci, Lett, 104, 381-397, Wirth, K,R,, Grandy, J,, Kelley, K,, Sandofsky, S,, 2002, Evolution of crust and mantle beneath the Bering Sea region: evidence from xenoliths and late 0enozoic basalts, In: Miller, E,L,, Grantz, A,, Klemperer, S,L, (Eds,), Tectonic and Evolution of the Bering Shelf-0hukchi Sea-Arctic Margin and Adjacent Landmasses: Geological Society of America Special Papers, 260, pp, 167-193, Zindler, A,, Hart, S,, 1986, 0hemical Geodynamics, Annu, Rev, Earth Planet, Sci, 14, 493-571,