Plume–lithosphere interaction beneath Mt. Cameroon volcano, West Africa: Constraints from 238U–...

www.elsevier.com/locate/gca

Geochimica et Cosmochimica Acta 71 (2007) 1835–1854

Plume–lithosphere interaction beneath Mt. Cameroonvolcano, West Africa: Constraints from 238U–230Th–226Ra and

Sr–Nd–Pb isotope systematics

Tetsuya Yokoyama a,*, Festus T. Aka a,b, Minoru Kusakabe a, Eizo Nakamura a

a Institute for Study of the Earth’s Interior, Okayama University, Tottori-ken 682-0193, Japanb Institute of Mining and Geological Research, Center for Volcanological and Geophysical Research (IRGM/ARGV) Ekona,

P.O. Box 370 Buea, Cameroon

Received 24 August 2006; accepted in revised form 9 January 2007; available online 17 January 2007

Abstract

Precise measurements of 238U–230Th–226Ra disequilibria in lavas erupted within the last 100 yr on Mt. Cameroon are pre-sented, together with major and trace elements, and Sr–Nd–Pb isotope ratios, to unravel the source and processes of basalticmagmatism at intraplate tectonic settings. All samples possess 238U–230Th–226Ra disequilibria with 230Th (18–24%) and 226Ra(9–21%) excesses, and there exists a positive correlation in a (226Ra/230Th)–(230Th/238U) diagram. The extent of238U–230Th–226Ra disequilibria is markedly different in lavas of individual eruption ages, although the (230Th/232Th) ratiois constant irrespective of eruption age. When U-series results are combined with Pb isotope ratios, negative correlationsare observed in the (230Th/238U)–(206Pb/204Pb) and (226Ra/230Th)–(206Pb/204Pb) diagrams. Shallow magma chamber processeslike magma mixing, fractional crystallization and wall rock assimilation do not account for the correlations. Crustal contam-ination is not the cause of the observed isotopic variations because continental crust is considered to have extremely differentPb isotope compositions and U/Th ratios. Melting of a chemically heterogeneous mantle might explain the Mt. Cameroondata, but dynamic melting under conditions of high DU and DU/DTh, long magma ascent time, or disequilibrium mineral/meltpartitioning, is required. The most plausible scenario to produce the geochemical characteristics of Mt. Cameroon samples isthe interaction of melt derived from the asthenospheric mantle with overlying sub-continental lithospheric mantle which haselevated U/Pb (>0.75) and Pb isotope ratios (206Pb/204Pb > 20.47) due to late Mesozoic metasomatism.� 2007 Elsevier Ltd. All rights reserved.

1. INTRODUCTION

Intraplate basaltic magmatism generally has a widechemical diversity especially in its trace element and radio-genic isotope compositions when compared to rather chem-ically uniform mid-ocean ridge basalt (MORB). The

0016-7037/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.gca.2007.01.010

* Corresponding author. Present address: Department of Geol-ogy, University of Maryland, College Park, MD 20742, USA. Fax:+1 301 405 3597.

E-mail addresses: [email protected] (T. Yokoyama),[email protected] (F.T. Aka), [email protected] (M. Kusakabe), [email protected] (E. Naka-mura).

variation is considered to arise from chemically heteroge-neous plume sources deep in the mantle (e.g., Strackeet al., 2005 and references therein), but the interaction be-tween upwelling plume and the lithospheric mantle or thecrust is another possibility to produce chemical diversities(e.g., Class and Goldstein, 1997; Bourdon et al., 1998; Classet al., 1998; Claude-Ivanaj et al., 1998; Macdonald et al.,2001; Lundstrom et al., 2003; Rankenburg et al., 2005),making it much more complicated to fully understand mag-ma generation processes in intraplate tectonic settings.

Uranium series disequilibrium in volcanic rocks is apowerful geochemical tracer for understanding the originof magmas and the dynamics of melting including the time-scales of magma transport from the source to surface, so

mailto:[email protected]


mailto:mhk2314@hotmail.


Fig. 1. Geological maps of (a) the Cameroon Volcanic Line and (b)Mt. Cameroon volcano showing lavas (dark shading) erupted inthe last 100 yr.

1836 T. Yokoyama et al. / Geochimica et Cosmochimica Acta 71 (2007) 1835–1854

U-series systematics has vigorously been applied to variousbasaltic samples of intraplate magmatism (e.g., Sims et al.,1995, 1999; Turner et al., 1997; Thomas et al., 1999; Lund-strom et al., 2003; Bourdon et al., 2005). These studies have,however, focused on samples erupted in different stages of asingle volcano or samples from different volcanoes belong-ing to a single hotspot region, thus their chemical character-istics could be affected by large-scale compositionaldifferences in the plume source. In contrast, a detailed studyof compositional variations in primitive mantle-derivedmagmas erupted in a single volcanic stage has the potentialto unravel the dynamics of melt migration from the sourceto the surface more precisely than in conventional ways(Reiners, 2002). Such an approach has recently been carriedout on Kilauea lavas (1985–2001) by applying U-series dis-equilibria to investigate the mechanism of melt transport inthe Hawaiian mantle plume (Pietruszka et al., 2006).

In this study, we have investigated 238U–230Th–226Rasystematics of volcanic rocks of well characterized erup-tions within the last 100 yr (1909–2000) of Mt. Cameroon.The frequent eruption of this volcano, one of the mostactive in Africa, gives us an excellent opportunity to applyU-series systematics to better understand the source anddynamics of magma processes in this intraplate tectonic set-ting. In addition to U-series data, we also report major ele-ments, trace elements and Sr–Nd–Pb isotopes of thevolcano to give further constraints on the magmageneration.

2. CAMEROON VOLCANIC LINE MAGMATISM AND

MT. CAMEROON

The Cameroon volcanic line (CVL) is a chain of 12Cenozoic volcanoes running for approximately 1600 kmfrom the island of Annobon in the Gulf of Guinea tothe Biu Plateau (Nigeria) located in the continental interi-or of West Africa (Fig. 1a). The CVL can be divided intothree zones: the oceanic sector (Annobon, Sao Tome andPrincipe), continent/ocean boundary (c.o.b.: Bioko, Etındeand Mt. Cameroon) and the continental sector (Manen-gouba, Bambouto, Oku, Ngaoundere Plateau, MandaraMountains and Biu Plateau). Alkaline mafic rocks (bas-alts, basanites, trachytes and phonolites) are predominantin the volcanism along the CVL. Fitton and Dunlop(1985) found geochemical similarity in trace elementsand Sr isotopes for basalts from both the oceanic and con-tinental sectors, and suggested that these magmas are de-rived from sub-lithospheric depths without furtherinteraction with the overlying lithosphere. Halliday et al.(1988, 1990), however, reported anomalously high206Pb/204Pb ratios (up to 20.5) for the c.o.b. volcanoes,compared to relatively lower (19–20) 206Pb/204Pb ratiosfor the oceanic and continental sector volcanoes. Withthe combination of other geochemical tracers such as Sr,Nd and O isotopes, they demonstrated that such a high206Pb/204Pb signature was created by the remelting andU/Pb fractionation during melt migration in the St. Hele-na fossil plume head as it cooled after emplacement at125 Ma. The study concluded that Cameroon line magmasare currently derived from a zone in the upper portions of

the fossil plume in the lithospheric mantle. This model im-plies that the observed Pb isotope anomaly did not derivefrom the heterogeneity of asthenospheric mantle but de-rived from metasomatized lithosphere. Recent researchon helium isotopes for the CVL showed MORB-like3He/4He ratios for the oceanic and the continental sectorsand HIMU–OIB-like ratios for the c.o.b. (Aka et al.,2004), which can consistently be explained by the fossilplume remelting model.

Mt. Cameroon (Lat. 4�200N, Long. 9�170E) is a strato-volcano 4095 m above sea level with a volume of approxi-mately 1200 km3 (Suh et al., 2003). It is located on thecontinental side of the c.o.b. zone (Fig. 1a) and is the only

Plume–lithosphere interaction beneath Mt. Cameroon volcano 1837

currently active volcano of the CVL. It is not clear when thevolcanic activity of Mt. Cameroon actually started. Theearliest record of eruptive activity (around 500 BC) wasmade by mariners. There is a lack of well-documented his-toric record before the 20th century, but it is believed thatthe volcano erupted eight times in the 19th century (Geze,1953). Seven eruptions occurred in the 20th century(1909, 1922, 1954, 1959, 1982, 1999 and 2000). The 1954eruption was a strombolian explosive activity from thesummit without any lava emission. The rest of the eruptionswere accompanied by massive basaltic lava flows from thesummit and flanks. More details about each of the erup-tions, including petrology and geochemistry, can be foundin Suh et al. (2003) and references therein.

3. SAMPLE DESCRIPTION

We collected fresh lava samples from six eruptions (1909,1922, 1959, 1982, 1999 and 2000; Fig. 1b). When sampledfrom flow sides, only the fresh interiors of lavas were collect-ed, particularly for vegetation-colonized flows (1909, 1922,1959) so as to avoid surface alteration by weathering and veg-etation as reported in Chauvel et al. (2005). In this study, 26samples were analyzed for major and trace elements, Sr–Nd–Pb isotopes and U-series disequilibria. The 1909 eruptiontook place in a crater in the NE flank and emitted�5 · 106 m3 of lava (Geze, 1953). Six samples collected alongthe length of the flow (from crater to lava terminus) were ana-lyzed in this study. The 1909 samples contain olivine and clin-opyroxene as phenocrysts, but plagioclase andtitanomagnetite are rarely found. Total volume of pheno-crysts is usually <5%. The 1922 eruption took place at thesummit and in the WSW flank with the emission of�107 m3 of lava (Geze, 1953). We analyzed two samples col-lected from the WSW lava flow in this study. Olivine and clin-opyroxene are the major phenocrysts (�15 vol %) in thesesamples, with a small amount of plagioclase (1–2 vol %).The 1959 eruption was located at ENE with the eruption vol-ume of �5 · 106 m3 (Suh et al., 2003). Ten samples collectedalong the lava flow were analyzed in this study. The 1959 la-vas have 5–10 vol % of phenocrysts with a mineral assem-blage of olivine, clinopyroxene and plagioclase. The 1982eruption occurred SW of the summit and emitted �107 m3

of lava (Fitton et al., 1983). Two 1982 samples were used inthis study. The 1982 lavas are aphyric with very little amountof olivine and clinopyroxene phenocrysts (both <1%) so aredifferent from the other eruptives. The 1999 and 2000 erup-tions, which are considered as a series of a single magmaticactivity, occurred in sites locating from the summit to SWdirection. The volume of the 1999 and 2000 lavas is�6.6 · 107 m3, which exceeds the total lava volume from1909 to 1982 (Suh et al., 2003). Four 1999 samples and two2000 samples were analyzed. The 1999 and 2000 samples con-tain �10 vol % of phenocrysts which are mainly olivine andclinopyroxene with minor volume (1–2%) of plagioclase.

4. ANALYTICAL TECHNIQUES

All the analyses were carried out at the PheasantMemorial Laboratory, Okayama University at Misasa

(Nakamura et al., 2003). Trace element and isotope analy-ses were handled under class 100 clean conditions. Lavasamples were crushed and pulverized into fine powderswith grain size under #200 as described in Yokoyamaet al. (2003). Analytical techniques are basically the sameas described in Yokoyama et al. (1999a,b, 2001, 2003,2006), and are briefly summarized as follows. Major ele-ments were determined by XRF (Philips PW2400). Traceelements were measured by a quadrupole type ICP-MS(inductively coupled plasma mass spectrometry: Agilent7500cs) following the methods of Makishima and Nakam-ura (2006) for Li, Rb, Sr, Y, Cs, Ba, REE and Pb, and Luet al. (2007) for B, Zr, Hf, Nb and Ta. Typical analyticalreproducibilities (2r) are 5% for Li, Y and Ta, 4% forNb and Pb and <3% for the rest of elements. To avoidfluoride coprecipitation of Zr and Hf in the HFSE analysis,an Al-solution was added for some Ca-enriched samplesbefore acid digestion following the method of Tanakaet al. (2003). Sr and Nd isotopes were measured by a TIMS(thermal ionization mass spectrometry: Finnigan MAT262) with internal fractionation correction using86Sr/88Sr = 0.1194 and 146Nd/144Nd = 0.7219 as normali-zation factors. Instrumental discrimination of the Sr andNd isotope analyses was corrected by repeated analysesof standard materials during the same analytical campaign,which were finally normalized to the values described inMakishima and Masuda (1994) (87Sr/86Sr = 0.710240,NIST SRM987; 143Nd/144Nd = 0.511839, La Jolla). Pb iso-tope ratios were precisely measured by the ‘normal doublespike method’ described in Kuritani and Nakamura (2003)using a TIMS (Finnigan MAT 261). Typical analyticalreproducibility was 0.005%, 0.005% and 0.01% for Sr, Ndand Pb isotopes, respectively (2r).

Isotope analyses of U, Th and Ra were performed byusing a TIMS (Finnigan MAT262 with RPQplus). Ra wasmeasured by a total evaporation TIMS (TE-TIMS) tech-nique except for three samples (CA060, CA04 andCA076) which were measured by a conventional method.The TE-TIMS technique enables to obtain 2–3 timesmore precise data than the conventional one (Yokoyamaand Nakamura, 2004). Replicate analyses of a standardrock sample JB-2 (tholeiitic basalt of Izu-Oshima, Japan)obtained from the Geological Survey of Japan gave(230Th/232Th) = 1.250 ± 0.013, Th = 0.2561 ± 0.0022 (lg/g),(234U/238U) = 1.001 ± 0.002, U = 0.1536 ± 0.0013 (lg/g)and 226Ra = 81.1 ± 0.5 (fg/g), respectively (errors are2r: Yokoyama and Nakamura, 2004; Yokoyama et al.,2006). Analytical uncertainty for the (226Ra/230Th) ratiodetermined from repeated analyses of the standard rock(JB-2) by the conventional method is estimated to be2%. Total procedural blanks were <50 pg for Th andU, and �0.03 fg for Ra, and they are negligible in thisstudy.

Isotope ratios in parentheses represent activity ratiosthroughout this paper unless noted otherwise. Decay con-stants of U, Th and Ra nuclides used for calculations in thisstudy are: k238U = 1.55125 · 10�10, k235U = 9.8485 · 10�10

k234U = 2.8263 · 10�6, k232Th = 4.9475 · 10�11, k230Th =9.158 · 10�6 and k226Ra = 4.332 · 10�4 (Le Roux andGlendenin, 1963; Jaffey et al., 1971; Cheng et al., 2000).


5. RESULTS

5.1. Major and trace elements

Major and trace element data for 1909–2000 lavas ob-tained in this study are presented in Table 1. All the sampleshave a loss on ignition (LOI) lower than �0.3, showing nosignificant chemical change by weathering after eruption(Chauvel et al., 2005). Fig. 2a shows the total alkali-silica(TAS) diagram of Le Bas et al. (1986), plotted together withpreviously published data (Fitton et al., 1983; Suh et al.,2003). All the samples measured in this study are classifiedas basanite, with SiO2 concentrations that range from43.5% to 46.6%, although some of the 1999–2000 samplesanalyzed in a previous study are hawaiite which might havederived from a small pocket of residual fractionated magma(Suh et al., 2003). Relatively low Mg# (49.9–62.9; Table 1)of the Mt. Cameroon samples compared to primary mag-mas formed by partial melting of source peridotite (e.g.,Green et al., 1974) implies that these rocks have undergonesome degree of fractionation. Fig. 2b–g show selected majorelement oxides and Ni content plotted against the MgOcontent for the 1909–2000 samples including previouslypublished data. In these figures, our data generally showgood agreement with previous data. As MgO decreases,Fe2O3

�, CaO and Ni decrease while SiO2, Al2O3 and P2O5

increase. The CaO/Al2O3 decreases as FeO/MgO increases(Fig. 2h). Such characteristics suggest the removal of olivineand clinopyroxene by fractionation. In terms of samplesfrom a single eruption, the 1999–2000 samples show thelargest variation of major element concentrations.

A bold line in Fig. 2g shows 5% mantle melting calculat-ed following the method described in Hart and Davis (1978).This line does not change significantly even when the degreeof melting is decreased down to 1%. Thin curves representthe removal of olivine from the primary melt by fractionalcrystallization, calculated as described in Claude-Ivanajet al. (1998). The Mt. Cameroon data have a shallow lineartrend almost parallel to the mantle melting line, which can-not be explained solely by olivine fractionation. The trendmight correspond to a two-stage fractionation process; oliv-ine fractionation (�7%) from a primary melt with 13% MgOand 330 lg/g Ni followed by fractionation of0.85cpx + 0.15ol (curve A). In this case, the amount of oliv-ine and clinopyroxene crystallized in the second fraction-ation stage reaches up to 40%. Otherwise samples ofindividual eruption ages might have been created by follow-ing different fractionation paths (e.g., curves B and C).

Fig. 3 shows the trace element pattern for Mt. Camer-oon samples normalized to primitive mantle (Sun andMcDonough, 1989). A strong enrichment in incompatibleelements and a gradual decrease of this enrichment alongwith an increase of compatibility are ‘‘typical’’ signatureof OIB (Sun and McDonough, 1989). Unlike ‘‘typical’’OIB, however, the Mt. Cameroon lavas show convex pat-terns with a maximum value at Nb (Nbsample/NbPM � 150)and strong depletions in K and Pb, which are similar toHIMU basalts (Willbold and Stracke, 2006). As a result,the Mt. Cameroon lavas have high Ce/Pb and U/Pb ratios(Ce/Pb = 41, U/Pb = 0.57) and low Ba/Nb and Rb/Nb ra-

tios (Ba/Nb = 5.0 and Rb/Nb = 0.42), all of which arecharacteristics of HIMU basalts and discriminate themfrom EM-type basalts (Willbold and Stracke, 2006).

Fig. 4a and b show temporal variation in La/Yb andU/Pb ratios for the Mt. Cameroon samples measured in thisstudy. These ratios are constant (within analytical error) inlavas of any single eruption, but change, though not mono-tonically, throughout the 100-year eruption period. Europi-um anomaly is absent for all samples (Table 1), suggestingplagioclase is not a major fractionation phase.

5.2. Sr–Nd–Pb isotopes

Sr–Nd–Pb isotopic compositions of the 1909–2000 sam-ples are listed in Table 1. As previously reported (e.g., Hall-iday et al., 1988), Mt. Cameroon samples have slightly high87Sr/86Sr ratios (0.70318–0.70332) and low 143Nd/144Nd ra-tios (0.51277–0.51281) compared to depleted MORB, butthey plot on the left side of the ‘‘mantle array’’. Such a sig-nature is quite similar to HIMU basalts, but when com-pared to St. Helena data (87Sr/86Sr = 0.7028–0.7031,143Nd/144Nd = 0.5128–0.5130: Stracke et al., 2003), theMt. Cameroon samples have slightly radiogenic 87Sr/86Srratio at a given 143Nd/144Nd ratio.

The Mt. Cameroon samples have extremelyradiogenic Pb isotope signatures (206Pb/204Pb = 20.382–20.427, 207Pb/204Pb = 15.656–15.663, 208Pb/204Pb = 40.173–40.199). Fig. 5a is a 206Pb/204Pb–207Pb/204Pb plot of theMt. Cameroon data together with St. Helena data (Strackeet al., 2003). In this figure, the Mt. Cameroon data have alinear correlation, plotting parallel to, and below theNHRL (Northern Hemisphere Reference Line: Hart,1984), also referred as the negative D7/4Pb signature (e.g.,Thirlwall, 1997). Mt. Cameroon samples have lower207Pb/204Pb and higher 208Pb/204Pb ratios at a given206Pb/204Pb ratio when compared to St. Helena data(206Pb/204Pb = 20.40–20.90, 207Pb/204Pb = 15.71–15.81,208Pb/204Pb = 39.74–40.17: Stracke et al., 2003). Rather,their Pb isotope characteristics are close to the FOZO com-ponent redefined in Stracke et al. (2005).

The 87Sr/86Sr ratios of the Mt. Cameroon samples showa variation that slightly exceeds the range of analyticalreproducibilities. There is no trend with eruption age. Thevariation of 143Nd/144Nd ratios is within analytical repro-ducibility. In contrast with Sr–Nd isotopes, Pb isotope ra-tios of the Mt. Cameroon lavas have variations thatexceed analytical error, especially for the 206Pb/204Pb ratios,although they do not show any systematic change with erup-tion age (Fig. 4c). It should be noted that the variation of Pbisotopes within any single eruption are comparable to ana-lytical error. Of all the samples in this study, those of 1909have the most while those of 1982 have the least radiogenicPb isotope ratios. 206Pb/204Pb ratios correlate positivelywith U/Pb ratios (Fig. 5b), although they do not show anycorrelation with Pb and SiO2 concentrations (Fig. 5c).

5.3. 238U–230Th–226Ra disequilibria

U–Th–Ra isotope compositions for the Mt. Cameroonsamples are listed in Table 1. All the samples are in

Table 1Major elements, trace elements and Sr, Nd, Pb, U, Th and Ra isotope data

Sample 1909-2 1909-6 1909-7 1909-8 1909-9 1909-10 CA060 1922-2 CA04 C20 C21 C23 C25Age 1909 1909 1909 1909 1909 1909 1922 1922 1959 1959 1959 1959 1959

(wt %)SiO2 45.5 45.6 45.5 45.6 45.6 45.5 43.5 43.5 46.3 46.5 46.2 46.3 46.1TiO2 3.43 3.44 3.43 3.43 3.45 3.46 3.42 3.28 3.31 3.27 3.31 3.30 3.35Al2O3 15.8 15.9 15.8 15.9 15.8 15.9 12.5 12.3 16.0 16.0 15.9 16.0 16.1Fe2O3

� 12.5 12.4 12.4 12.4 12.5 12.5 14.0 13.8 11.9 11.7 12.0 11.9 12.1MnO 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.20 0.20 0.20 0.20 0.20MgO 5.58 5.47 5.44 5.43 5.53 5.44 10.04 10.03 5.26 5.39 5.44 5.34 5.23CaO 10.59 10.62 10.54 10.44 10.63 10.53 11.86 11.71 10.12 10.00 10.16 10.00 10.22Na2O 4.18 4.21 4.24 4.25 4.21 4.19 3.31 3.34 4.49 4.49 4.42 4.51 4.44K2O 1.68 1.69 1.70 1.72 1.68 1.71 1.48 1.45 1.82 1.84 1.82 1.84 1.82P2O5 0.82 0.83 0.83 0.84 0.82 0.83 0.78 0.81 0.86 0.86 0.85 0.86 0.85LOI �0.68 �0.66 �0.69 �0.47 �0.53 �0.48 �0.65 �0.62 �0.59 �0.32 �0.59 �0.54 �0.61Total 99.6 99.7 99.5 99.8 99.9 99.8 100.6 99.8 99.8 100.0 99.8 99.7 99.8Mg#a 51.0 50.6 50.6 50.5 50.8 50.3 62.5 62.9 50.8 51.7 51.4 51.1 50.1

(lg/g)Cr2O3 100 108 97 95 111 97 595 625 116 79 117 114 96NiO 58 54 54 53 58 57 216 216 58 46 66 67 61

(lg/g)Li 6.48 7.58 7.61 7.18 7.10 7.11 6.20 6.47 7.75 7.31 7.70 8.02 7.91B 2.71 2.86 2.75 2.92 2.75 2.86 2.60 1.69 2.54 2.85 2.80 3.01 2.90Rb 48.1 43.2 43.1 43.5 44.4 41.9 44.2 39.9 40.4 44.7 45.9 46.8 49.8Sr 1163 1215 1267 1218 1165 1194 1104 1125 1205 1189 1189 1276 1223Y 36.5 38.2 39.7 39.1 37.7 37.6 34.3 35.4 38.0 39.6 41.0 40.9 39.7Zr 384 394 388 390 387 389 343 355 406 413 403 417 412Nb 104 106 105 106 105 105 88 91 112 113 112 115 114Cs 0.476 0.497 0.497 0.488 0.478 0.480 0.438 0.460 0.504 0.522 0.535 0.542 0.522Ba 512 525 528 525 511 508 473 485 552 553 547 578 567La 77.1 79.7 79.2 78.6 76.7 77.5 77.5 80.2 82.2 83.1 84.6 87.5 85.0Ce 159 165 165 162 159 161 164 166 172 170 174 181 177Pr 18.8 19.0 19.1 19.2 18.8 18.6 18.7 19.4 19.7 19.8 20.2 20.9 20.3Nd 73.6 75.8 74.4 73.9 73.1 72.9 73.7 75.3 76.9 77.6 77.9 79.0 78.3Sm 12.4 12.7 12.9 12.7 12.4 12.7 12.2 12.6 12.6 13.1 13.1 13.3 13.1Eu 3.85 3.89 3.98 3.88 3.79 3.81 3.71 3.83 4.03 4.04 4.02 4.14 4.07Gd 10.4 10.7 10.4 10.5 10.3 10.6 10.1 10.3 10.4 10.7 10.8 11.3 11.0Tb 1.41 1.43 1.45 1.42 1.39 1.41 1.34 1.39 1.47 1.49 1.45 1.49 1.47Dy 7.24 7.29 7.34 7.20 7.22 7.14 6.74 6.93 7.40 7.51 7.47 7.81 7.40Ho 1.34 1.35 1.35 1.34 1.33 1.31 1.23 1.26 1.37 1.40 1.39 1.41 1.38Er 3.06 3.14 3.16 3.11 3.04 3.04 2.89 2.85 3.24 3.25 3.22 3.31 3.30Tm 0.423 0.428 0.430 0.431 0.417 0.422 0.390 0.404 0.440 0.455 0.445 0.468 0.463Yb 2.71 2.75 2.69 2.71 2.62 2.62 2.39 2.51 2.83 2.88 2.86 2.95 2.86Lu 0.362 0.366 0.359 0.357 0.356 0.360 0.337 0.336 0.385 0.374 0.369 0.387 0.395

(continued on next page)

Plu

me–lith

osp

here

interactio

nb

eneath

Mt.

Cam

eroo

nvo

lcano

1839

Table 1 (continued)

Sample 1909-2 1909-6 1909-7 1909-8 1909-9 1909-10 CA060 1922-2 CA04 C20 C21 C23 C25Age 1909 1909 1909 1909 1909 1909 1922 1922 1959 1959 1959 1959 1959

Hf 7.82 8.08 7.92 7.92 7.96 7.90 7.50 7.76 8.26 8.35 8.23 8.54 8.35Ta 5.68 5.74 5.67 5.74 5.70 5.64 5.79 5.98 6.10 6.15 6.05 6.26 6.21Pb 3.75 3.91 3.78 3.90 3.78 3.91 3.63 3.89 4.15 4.34 4.31 4.52 4.31Th (TIMS) 8.222 8.232 8.345 8.267 8.116 8.173 7.91 8.153 8.993 9.018 9.012 9.046 9.016U (TIMS) 2.229 2.258 2.266 2.274 2.227 2.249 2.14 2.242 2.417 2.467 2.446 2.474 2.439Eu* 1.00 0.99 1.01 1.00 0.99 0.98 0.99 0.99 1.04 1.01 1.00 1.00 1.0187Sr/86Sr 0.703254 0.703262 0.703219 0.703261 0.703179 0.703266 0.703320 0.703307 0.703267 0.703263 0.703208 0.703276 0.703281143Nd/144Nd 0.512786 0.512782 0.512798 0.512766 0.512799 0.512773 0.512771 0.512781 0.512781 0.512777 0.512789 0.512779 0.512771206Pb/204Pb 20.427 20.425 20.426 20.424 20.423 20.423 20.402 20.404 20.413 20.413 20.414 20.413 20.414207Pb/204Pb 15.663 15.662 15.663 15.662 15.661 15.662 15.661 15.656 15.661 15.661 15.661 15.659 15.661208Pb/204Pb 40.196 40.193 40.197 40.193 40.190 40.190 40.199 40.183 40.188 40.187 40.192 40.187 40.191(230Th/232Th) 0.992 1.002 0.991 0.997 0.993 0.997 0.987 0.988 0.984 0.997 0.997 0.999 0.991±2r 0.004 0.004 0.002 0.004 0.003 0.003 0.003 0.003 0.003 0.003 0.004 0.004 0.003(238U/232Th) 0.823 0.832 0.824 0.835 0.833 0.835 0.819 0.834 0.815 0.830 0.823 0.830 0.821(230Th/238U) 1.206 1.204 1.203 1.194 1.193 1.195 1.204 1.184 1.207 1.201 1.211 1.204 1.208(234U/238U) 1.002 1.003 1.001 1.003 1.001 1.002 1.000 1.002 1.001 1.002 1.000 1.002 0.999±2r 0.002 0.001 0.001 0.001 0.002 0.001 0.001 0.002 0.001 0.002 0.001 0.001 0.002226Ra (fg/g) 1017 1031 1041 1041 1018 1026 959 976 1130 1136 1144 1151 1132± (2r %) 0.3 0.2 0.3 0.3 0.4 0.3 0.4 0.3 0.5 0.5 0.4 0.4 0.3(226Ra/230Th) 1.121 1.124 1.131 1.136 1.135 1.132 1.104b 1.089 1.148b 1.136 1.144 1.145 1.138(226Ra/230Th)0 1.13 1.13 1.14 1.14 1.14 1.14 1.11 1.09 1.15 1.14 1.15 1.15 1.14

Sample C27 C28 C29 C33 C36 1982-1 1982-2 CA076 1999-1 1999-2 1999-3 2000-1 2000-2Age 1959 1959 1959 1959 1959 1982 1982 1999 1999 1999 1999 2000 2000

(wt %)SiO2 46.1 46.2 46.0 46.1 46.3 44.7 44.8 46.2 46.2 45.8 46.6 45.4 45.5TiO2 3.36 3.34 3.35 3.33 3.30 3.53 3.59 3.25 3.18 3.24 3.13 3.30 3.26Al2O3 16.0 15.9 15.9 16.0 16.0 15.0 15.4 15.5 15.8 15.4 16.2 15.1 15.0Fe2O3

� 12.2 12.1 12.1 12.0 11.9 13.6 13.6 12.3 12.0 12.4 11.5 12.8 12.7MnO 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20MgO 5.42 5.50 5.38 5.36 5.36 6.09 5.81 6.03 5.87 6.34 5.27 6.42 6.58CaO 10.24 10.29 10.25 10.19 10.08 11.80 11.46 10.57 10.14 10.59 9.78 10.89 10.99Na2O 4.37 4.39 4.40 4.43 4.49 3.59 3.76 4.20 4.30 4.05 4.55 3.91 3.86K2O 1.80 1.79 1.81 1.82 1.84 1.35 1.44 1.67 1.74 1.62 1.86 1.55 1.54P2O5 0.84 0.84 0.84 0.85 0.86 0.60 0.64 0.75 0.76 0.72 0.82 0.69 0.68LOI �0.61 �0.69 �0.61 �0.57 �0.61 �0.75 �0.79 �0.33 �0.58 �0.56 �0.44 �0.61 �0.67Total 99.8 99.9 99.7 99.8 99.7 99.8 99.9 100.3 99.6 99.8 99.6 99.7 99.6Mg#a 51.0 51.4 50.8 50.9 51.2 51.0 49.9 53.4 53.3 54.3 51.6 53.9 54.7

(lg/g)Cr2O3 (lg/g) 105 106 97 96 111 108 72 201 157 206 103 229 254NiO 66 68 63 64 62 78 70 82.0 82 100 61 99 104

1840T

.Y

ok

oyam

aet

al./

Geo

chim

icaet

Co

smo

chim

icaA

cta71

(2007)1835–1854

(lg/g)Li 7.87 7.68 7.64 7.57 7.58 6.06 6.18 7.03 7.26 6.80 7.88 6.68 6.64B 3.02 2.92 3.00 3.03 2.88 2.23 2.38 2.68 2.81 2.65 2.92 2.61 2.70Rb 46.4 46.3 46.5 47.0 48.4 34.1 35.8 41.0 45.1 39.0 45.7 41.7 39.2Sr 1221 1213 1219 1188 1213 1020 1050 1087 1156 1087 1181 1089 1099Y 38.7 38.1 38.6 38.8 39.9 33.1 34.1 35.1 37.1 34.9 38.3 36.4 34.0Zr 408 405 406 417 413 320 331 372 391 370 409 356 353Nb 112 111 112 115 116 81.4 85.7 103 108 102 115 97.3 95.7Cs 0.516 0.503 0.512 0.512 0.520 0.371 0.394 0.463 0.488 0.458 0.520 0.454 0.447Ba 571 545 555 556 574 421 445 492 526 477 549 487 482La 83.4 83.7 85.5 84.0 85.8 62.2 66.8 72.8 80.2 74.5 84.4 73.5 72.8Ce 175 168 171 171 177 131 139 153 163 150 170 150 148Pr 20.1 19.2 19.8 19.6 20.0 15.1 16.2 17.3 18.7 17.1 19.6 17.5 17.1Nd 78.5 74.7 76.0 75.8 78.1 61.0 63.8 67.0 71.9 67.8 75.2 67.1 66.9Sm 13.2 12.9 12.9 13.0 13.3 11.0 11.4 11.5 12.2 11.6 12.5 11.3 11.3Eu 3.90 3.86 3.95 4.01 4.04 3.34 3.48 3.53 3.72 3.46 3.87 3.56 3.51Gd 10.7 10.8 10.7 10.6 10.9 9.13 9.39 9.70 10.2 9.39 10.1 9.55 9.62Tb 1.45 1.42 1.47 1.48 1.53 1.29 1.32 1.30 1.37 1.31 1.42 1.31 1.29Dy 7.57 7.50 7.46 7.65 7.70 6.56 6.74 6.74 7.22 6.79 7.31 6.81 6.77Ho 1.39 1.35 1.36 1.37 1.39 1.22 1.25 1.25 1.31 1.24 1.35 1.24 1.24Er 3.23 3.07 3.28 3.23 3.28 2.81 2.96 2.95 3.15 2.85 3.20 2.92 2.86Tm 0.452 0.438 0.451 0.452 0.456 0.391 0.402 0.405 0.432 0.400 0.448 0.399 0.400Yb 2.91 2.85 2.88 2.84 2.91 2.43 2.53 2.59 2.72 2.57 2.88 2.57 2.53Lu 0.376 0.384 0.377 0.386 0.387 0.317 0.331 0.350 0.370 0.338 0.374 0.343 0.339Hf 8.33 8.33 8.35 8.46 8.39 7.09 7.24 7.72 7.88 7.62 8.18 7.40 7.41Ta 6.07 6.14 6.22 6.35 6.24 4.53 4.68 5.62 5.76 5.54 6.10 5.25 5.18Pb 4.22 4.21 4.20 4.38 4.31 3.27 3.44 3.73 3.89 3.78 3.99 3.74 3.69Th (TIMS) 8.942 8.891 8.964 8.970 9.134 6.506 6.975 8.31 8.574 7.931 9.225 7.941 7.788U (TIMS) 2.417 2.395 2.422 2.428 2.449 1.713 1.833 2.24 2.303 2.144 2.501 2.129 2.092Eu* 0.97 0.97 1.00 1.01 0.99 0.99 1.00 0.99 0.99 0.98 1.02 1.01 1.0087Sr/86Sr 0.703287 0.703194 0.703195 0.703275 0.703232 0.703213 0.703203 0.703282 0.703213 0.703203 0.703253 0.703243 0.703257143Nd/144Nd 0.512776 0.512786 0.512805 0.512766 0.512808 0.512809 0.512797 0.512779 0.512807 0.512805 0.512798 0.512797 0.512797206Pb/204Pb 20.415 20.413 20.412 20.414 20.414 20.382 20.390 20.406 20.406 20.404 20.404 20.407 20.406207Pb/204Pb 15.661 15.660 15.659 15.660 15.660 15.658 15.659 15.661 15.660 15.660 15.659 15.660 15.659208Pb/204Pb 40.191 40.187 40.185 40.188 40.188 40.173 40.179 40.190 40.188 40.186 40.185 40.192 40.189(230Th/232Th) 0.987 0.990 0.984 0.991 0.987 0.989 0.991 0.982 0.990 0.992 0.986 0.991 0.990±2r 0.003 0.002 0.004 0.004 0.005 0.003 0.003 0.004 0.002 0.003 0.004 0.002 0.004(238U/232Th) 0.820 0.817 0.820 0.821 0.814 0.799 0.797 0.817 0.815 0.820 0.823 0.814 0.815(230Th/238U) 1.203 1.212 1.200 1.206 1.213 1.238 1.243 1.202 1.214 1.209 1.198 1.218 1.214(234U/238U) 0.999 1.000 1.001 1.003 1.003 1.002 1.002 1.003 1.002 1.000 1.002 1.003 1.002±2r 0.002 0.002 0.002 0.002 0.002 0.001 0.002 0.001 0.001 0.002 0.002 0.001 0.001226Ra (fg/g) 1126 1123 1124 1137 1146 864 916 1056 1099 1022 1176 1017 998± (2r %) 0.2 0.4 0.4 0.5 0.2 0.3 0.6 0.4 0.3 0.7 0.6 0.2 0.2(226Ra/230Th) 1.148 1.147 1.145 1.150 1.143 1.207 1.191 1.164b 1.164 1.168 1.162 1.162 1.163(226Ra/230Th)0 1.15 1.15 1.15 1.15 1.15 1.21 1.19 1.16 1.16 1.17 1.16 1.16 1.16

a Supposing Fe2O3 ¼ 0:15� Fe2O�3.b Analytical error is estimated to be 2%.

Plu

me–lith

osp

here

interactio

nb

eneath

Mt.

Cam

eroo

nvo

lcano

1841

a b

c d

e f

g h

Fig. 2. Major element and Ni variation diagrams for Mt. Cameroon lavas. Large symbols are samples in this study while small, open symbolsare published values (Fitton et al., 1983; Suh et al., 2003) of which eruption ages correspond to large symbols in the same shape. (a)Classification on the TAS diagram. (b–f) Major element oxides plotted against MgO concentration. (g) Ni–MgO diagram. A bold lineindicates 5% mantle melting calculated after Hart and Davis (1978). Curves A–C represent olivine fractionation and basaltic(0.85cpx + 0.15ol) fractionation trends, which are calculated by using KMgO

ol=melt ¼ 4; KMgOcpx=melt ¼ 2:5 and KNi

cpx=melt ¼ 2: KNiol=melt is a

function of MgO concentration in melt (Hart and Davis, 1978). Tick marks indicate the amount of crystals fractionated. (h) CaO/Al2O3-FeO/MgO diagram. FeO/MgO is calculated assuming Fe2O3

� ¼ Fe2O3� � 0:15.


Fig. 3. Trace element patterns normalized to primitive mantle (Sunand McDonough, 1989) for the Mt. Cameroon lavas, ‘‘typicalOIB’’ (dashed line: Sun and McDonough, 1989) and St. Helena(gray zone: Willbold and Stracke, 2006). Note that Li data is notgiven for St. Helena.

a b

ed

Fig. 4. Temporal variation of representative trace element and isotopic rlines show average of all the Mt. Cameroon data obtained in this study, asame as Fig. 2.


238U–234U equilibrium within analytical error. This, inaddition to low LOI values, shows that the effect ofalteration by weathering is negligible. All the samples have238U–230Th–226Ra disequilibria, showing (230Th/238U)ratios greater than 1 with 18–24% enrichments of 230Thand (226Ra/230Th)0 ratios (eruption time-corrected(226Ra/230Th) ratio) greater than 1 with 9–21% excess226Ra. Our data are consistent with a limited number ofliterature values for the 1982 lavas ((230Th/238U) = 1.32Williams and Gill, 1992; (230Th/238U) = 1.245,(226Ra/230Th) = 1.19: Chabaux and Allegre, 1994).

Fig. 6a is a U–Th equiline diagram of the Mt. Cameroonlavas and previously published MORB and OIB data. Mt.Cameroon samples plot on the left of the equiline in themiddle of the OIB field, which can be attributed to theexistence of garnet in the source region The variation in(230Th/232Th) ratio for the Mt. Cameroon samples(0.982–1.002) is very small and comparable to the analyticalreproducibility (Fig. 4d). In contrast, the 1982 sampleshave distinctively high (230Th/238U) ratios (1.238–1.243)

c

f

atios in lava erupted on Mt. Cameroon within the last 100 yr. Boldnd thin lines represent analytical uncertainties (2r). Symbols are the

a

b

c

Fig. 5. (a) 206Pb/204Pb–207Pb/204Pb diagram for the Mt. Cameroonlavas plotted together with St. Helena (Stracke et al., 2003), FOZO(Stracke et al., 2005) and Nigeria basement rock data (Dickin et al.,1991; Dada et al., 1995). Bold line represents the NorthernHemisphere Reference Line (Hart, 1984). Dashed line is aregression line for the Mt. Cameroon lavas for which the slopecorresponds to 2.43 Ga of Pb–Pb age. ‘‘A’’ is data of Biu Plateauultramafic xenolith (Lee et al., 1996), and ‘‘B’’ and ‘‘C’’ representmegacrysts from Biu and Jos Plateaux, respectively (Rankenburget al., 2005). (b and c) Variation diagram of 206Pb/204Pb againstU/Pb and SiO2 for the Mt. Cameroon lavas. Symbols are the sameas Fig. 2.

a

b

Fig. 6. (a) U–Th equiline diagram for the Mt. Cameroon lavasplotted together with MORB and OIB data. Dark gray areaindicates Pan-African granites of north-central Nigeria (Dadaet al., 1995). Dashed line is a regression line of the Mt. Cameroondata. (b) (226Ra/230Th)0–(230Th/238U) diagram for the Mt. Camer-oon lavas. Symbols are the same as Fig. 2.


compared to the rest of the samples (1.184–1.218) (Fig. 4e).The variation of (226Ra/230Th)0 ratio is larger than that ofthe (230Th/238U) ratio. Similar to Pb isotope ratios,(226Ra/230Th)0 ratios show limited variations in samplesof a single eruption and they have distinctive valuesdepending on the eruption age (Fig. 4f). The degree of226Ra–230Th disequilibrium decreases from 1909 towards1922, increases towards 1982, and then decreases to 2000.The maximum (226Ra/230Th)0 is observed in the 1982 sam-ples (1.19–1.21), while the 1922 samples have the lowest val-ues (1.09–1.11).

When (226Ra/230Th)0 ratios are plotted against(230Th/238U) ratios, there exists a positive correlation hav-ing distinctive values in individual eruption ages (Fig. 6b).In a (226Ra/230Th)0–206Pb/204Pb diagram, a negative linearcorrelation is observed excluding the 1922 data (Fig. 7a).The (230Th/238U) and 206Pb/204Pb ratios are also negativelycorrelated (Fig. 7b). Similar to 206Pb/204Pb ratios, however,(226Ra/230Th)0 and (230Th/238U) ratios are not well correlat-ed with U, Th and SiO2 concentrations (not shown).

6. DISCUSSION

One of the key observations is the geochemical varia-tions in the Mt. Cameroon lavas, especially for the changeof (230Th/238U), (226Ra/230Th)0 and 206Pb/204Pb ratios. Weconsider that this is caused by the interaction of melts de-rived from the asthenospheric mantle with overlying sub-continental lithospheric mantle. Before discussing such ascenario, we examine other possibilities that may explainthe observed variations. We then discuss the origin and pro-cesses of basaltic magmatism beneath Mt. Cameroon basedon the plume–lithosphere interaction model.

a

b

Fig. 7. Variation diagrams of (a) (226Ra/230Th)0 and (b)(230Th/238U) plotted against 206Pb/204Pb for the Mt. Cameroonlavas. Dashed lines are regression lines for the Mt. Cameroon data.Symbols are the same as Fig. 2.


6.1. Shallow magma chamber processes

The simplest mechanism for the correlations shown inFigs. 6 and 7 would be binary mixing of two chemically dif-ferent components (excluding 1922). Actually, Fitton et al.(1983) demonstrated that magma mixing played an impor-tant role in causing compositional diversity in the 1982eruptives. However, lack of correlations between U–Th–Ra–Pb isotopes and SiO2, U, Th and Pb concentrationsin the Mt. Cameroon samples (e.g., Fig. 5c) rejects the pos-sibility of simple two-component mixing in a shallow mag-ma chamber.

It has been reported that the (226Ra/230Th)0 ratio canvary in lavas collected from a single volcano (e.g., Condo-mines et al., 1995; Sigmarsson, 1996; Asmerom et al.,2005; Sigmarsson et al., 2005; Pietruszka et al., 2006). Sucha variation is, in some cases, observed in lavas effused by asingle eruption (Vigier et al., 1999; Yokoyama et al., 2006).Because the half-life of 226Ra is relatively short (1600 yr),226Ra decay by ageing of magma accompanied by fraction-al crystallization in a magma chamber can produce a mag-ma which has a different (226Ra/230Th) ratio from that ofthe undifferentiated magma. Mixing of the ‘‘old’’ magmawith newly injected juvenile magma can therefore createthe variation in (226Ra/230Th)0 ratio. However, such a pro-cess alone cannot explain the variation in 206Pb/204Pb ratiosthat would not change during fractional crystallization.

Yokoyama et al. (2006) found a systematic variation in(226Ra/230Th)0 ratio of volcanic rocks from Miyakejima,

Izu arc, Japan, and demonstrated that fractional crystalli-zation accompanied by assimilation of magma chamberwall rock modified the original magma composition andcreated the observed (226Ra/230Th)0 variation. This processmight explain the Mt. Cameroon data when the assimilanthas Pb and U-series isotope compositions different from theinjected magma. If assimilation of wall rock occurred in themagma chamber beneath Mt. Cameroon at the final stagejust before eruption, then there should be a correlation be-tween (226Ra/230Th) ratio and indices of magma differenti-ation such as MgO, SiO2 and Th (Yokoyama et al.,2006). As previously mentioned, this is not observed inthe Mt. Cameroon samples. Actually, 238U–230Th–226Radisequilibria and 206Pb/204Pb ratios of the 1999–2000 sam-ples remain constant while these lavas show wide variationsin terms of major elements (Fig. 2) that were presumablycreated by fractionation in the final stage just beforeeruption.

There was an increase in the seismicity of Mt. Cameroona few days before the onset of the 1999 eruption, with theearthquakes occurring at depths of 30–55 km below theEarth’s surface (Suh et al., 2003). This depth is below thegeophysical Moho beneath Mt. Cameroon (20–22 km),and is located in the lithospheric mantle (Ambeh et al.,1989). From the geophysical information and the petrolog-ical and geochemical data, Suh et al. (2003) suggested thatprimitive magmas of Mt. Cameroon come from theasthenospheric mantle and the deep earthquakes associatedwith volcanic activity could reflect magmas invading thelithospheric mantle. This would support the lack of well-es-tablished shallow magma chambers beneath Mt.Cameroon.

On the basis of discussions above, we propose that thevariations of U–Th–Ra–Pb isotopes observed in Mt. Cam-eroon samples do not result from shallow magma chamberprocesses (binary mixing of chemically distinct magmas,fractional crystallization involving 230Th and 226Ra decay,and wall rock assimilation), but are linked to magma pro-cesses that took place before final fractional crystallizationhas occurred at a yet undefined depth beneath the volcano.

6.2. Effect of continental crust

Previous studies (e.g., Halliday et al., 1988, 1990) foundno evidence for contamination by ancient radiogenic rocks(basement Pan-African granites) in the CVL c.o.b. samples.Lack of well-established shallow magma chambers beneathMt. Cameroon is further evidence against crustal contami-nation of the samples. Nevertheless, because the Pb isotopicvariation observed in these samples (206Pb/204Pb = 20.382–20.427) is subtle compared to the variation so far reportedfor c.o.b. zone volcanoes (206Pb/204Pb = 19.983–20.522), itis worth reevaluating whether the Pb and U-series isotopicvariations could come from crustal contamination. If crust-al contamination is the cause of the Pb and U-series isotopevariations, it is straightforward to postulate that the extentof this process is larger for the 1909 and 1959 lavas than the1982 lavas, because the 1982 lavas have higher MgO,(226Ra/230Th) and (230Th/238U) ratios and lower SiO2, sug-gesting they are more primitive than the 1909 and 1959

Fig. 8. Gd/Yb–La/Yb diagram for the Mt. Cameroon lavasplotted together with calculated curves for partial melting of aprimitive mantle source (open circle in the lower left). Symbols arethe same as Fig. 2. The mineral assemblage of the source wasdetermined by the mixture of garnet lherzolite (ol:opx:cpx:grt =54:17:9:20) and spinel lherzolite (ol:opx:cpx:sp = 46:28:18:8) withstoichiometric reaction to form olivine and garnet from orthopy-roxene, clinopyroxene and spinel (Workman et al., 2004). Twocurves labeled Grt 4% and 8% represent the abundance of garnet inthe source, corresponding to 20:80 and 40:60 proportions of Grt:Splherzolite. Accumulated fractional melting is assumed and partitioncoefficients reported in Halliday et al. (1995) are used in thecalculation.


samples. Then, as is clear from Fig. 7a, the contaminant,namely ancient crustal material, should have more radio-genic 206Pb/204Pb ratio than the Mt. Cameroon samples.This is obviously inconsistent with the fact that Nigerianbasement rocks (granulites, gneisses migmatites and gran-ites) as analyzed by Dickin et al. (1991) and Dada et al.(1995) have less radiogenic 206Pb/204Pb ratios, opposite tothe direction of possible contaminant suggested by thetrend of Mt. Cameroon data (Fig. 5a). A similar discussionis possible for 207Pb/206Pb ratios. Although Nigerian base-ment rocks have very high 207Pb/206Pb ratios (>0.8), the1909 lavas have the lowest 207Pb/206Pb ratio (0.7668) whilemore primitive 1982 lavas have the highest 207Pb/206Pb(0.7681) ratio of the Mt. Cameroon samples.

Another argument against crustal contamination comesfrom the U–Th dataset. From the discussion above, thepossible contaminant should be on the extrapolation ofthe Mt. Cameroon trend, which is the right side of the1909 and 1959 samples in the U–Th equiline diagram(Fig. 6a). It is worth noting that the composition of the con-taminant does not necessarily lie on the equiline even if thecrustal material is sufficiently old as to be in 238U–230Thequilibrium. This is because assimilation of a crustal partialmelt which could be in 238U–230Th disequilibrium might oc-cur instead of bulk assimilation. Continental crust is gener-ally more enriched in Th than U (e.g., Rudnick and Gao,2003). Actually, Pan-African granites of north-centralNigeria have Th/U ratios ranging from 4.6 to 19.8 (Dadaet al., 1995), corresponding to (238U/232Th) of 0.16–0.68(Fig. 6a). Such compositions are too far off the possiblecrustal contamination trend, even if partial melting of crust-al material is taken into account. It is reasonable to con-clude therefore that the Pb and U-series isotopicvariations observed in Mt. Cameroon lavas is not due tocontamination by Pan-African crust, unless the magmahas selectively reacted with some yet undiscovered phasein the crust.

6.3. Melting processes in the mantle source

6.3.1. Mantle source heterogeneity

Melting conditions of the Mt. Cameroon samples can beinferred from variations in the rare earth elements: LREE/HREE ratios are sensitive to the (accumulated) extent ofmelting, while MREE/HREE ratios are sensitive to theamount of residual garnet in the source. Fig. 8 shows aGd/Yb–La/Yb diagram for the Mt. Cameroon samplesplotted together with calculated curves for partial meltingof lherzolites that have different mineral assemblages buthave a chemically homogeneous composition (primitivemantle: Sun and McDonough, 1989). The Mt. Cameroondata require 6–8% of residual garnet in the source, whichis consistent with the fact that the samples have238U–230Th disequilibrium with 230Th excess. As previouslymentioned, these rocks would have experienced the frac-tional crystallization of olivine and clinopyroxene beforeeruption (Section 5.1). However, REE are incompatible toolivine, and clinopyroxene crystallization would not effec-tively fractionate MREE and HREE (DGd/DYb = 1.02Halliday et al., 1995). Actually, fractionation of olivine

and clinopyroxene deduced from curve A of Fig. 2g chang-es Gd/Yb ratio by <3%. Therefore, the variation of Gd/Ybratios in the Mt. Cameroon samples is not due to fractionalcrystallization, nor can it be explained by different degreesof melting of a homogeneous source mantle. This, in addi-tion to the variations in Pb isotope ratios (Fig. 5), would re-quire chemical (and mineralogical) heterogeneity in thesource mantle if the variations observed in Fig. 4 are ofsource mantle origin.

The existence of chemically enriched pyroxenite in themantle has long been discussed for the generation of ocean-ic basalts (Stracke et al., 1999 and references therein). Dif-ferent degrees of melting of pyroxenite-bearing peridotitemight produce variations in trace element ratios and radio-genic isotopes. However, pyroxenite melt is expected tohave Lu/Hf and 143Nd/144Nd ratios markedly differentfrom peridotite melt, but these ratios do not show any dif-ference in the Mt. Cameroon samples. Thus, chemical het-erogeneity of the Mt. Cameroon source is, if it truly exists,derived from materials other than enriched pyroxenite oreclogite.

As previously noted, the Pb isotope signature of the Mt.Cameroon samples is close to the end component of FOZOas redefined by Stracke et al. (2005) (Fig. 5), who insistedthat FOZO is a ubiquitous component in MORB andOIB sources produced by continuous recycling of oceaniccrust. The global array of (230Th/232Th) with 87Sr/86Srand 143Nd/144Nd ratios for MORB and OIB samples showclear hyperbolic mixing trends between DMM (depletedMORB mantle) and EM2, suggesting coherent mantle dif-ferentiation over Earth’s history regarding U/Th, Rb/Srand Sm/Nd (Sims and Hart, 2006). However, similar to


samples from Canary Islands and Mt. Erebus (Antarctica),Mt. Cameroon lavas have markedly low (230Th/232Th) ra-tios at given 87Sr/86Sr ratios compared to MORB andOIB samples. Sims and Hart (2006) suggested that such asignature is due to mixing of an additional (HIMU-like)source component. Apart from Pb, there are no obviouschanges in the other radiogenic isotopes (Sr, Nd and Th)of the Mt. Cameroon samples. If the observed Pb isotopevariation is of asthenospheric mantle origin, the sourcemantle should have homogeneous Sr, Nd and Th but heter-ogeneous Pb isotope compositions as a result of mantleevolution via recycling of ‘ubiquitous’ oceanic crust withslight contribution of a specific material (subduction-modi-fied oceanic crust). Although difficult to envision, this sce-nario is possible because the global array of Pb with Th,Sr and Nd isotopes require not two but four mixing compo-nents (mantle tetrahedron), which suggests that the U–Pbisotopic system is strongly decoupled from the U/Th,Rb/Sr, and Sm/Nd isotopic systems (Sims and Hart, 2006).

6.3.2. Dynamic melting

Contrary to stable trace element ratios, U-series disequi-libria are especially useful in evaluating mantle melting con-ditions because the source mantle is, irrespective of

Fig. 9. Curves represent (226Ra/230Th) and (230Th/238U) ratioscalculated by the dynamic melting model (Williams and Gill, 1989)when source mantle contains 6% garnet. Vertical trends areactivities for constant malting rate (numbers in boxes) rangingfrom 10�6 to 10�2 kg m�3 yr�1. Horizontal trends are activities forconstant porosity ranging from 0.001 to 0.01. Also plotted is theMt. Cameroon data with eruption time correction for(226Ra/230Th). Note that the Mt. Cameroon data are out of thepossible values. Symbols are the same as Fig. 2.

Table 2Partition coefficients of U, Th and Ra used in dynamic melting and AFC

Olivine Cpxa Cpxb Opx

DU 6 · 10�5 1.5 · 10�2 5.3 · 10�3 5.2 · 10�

DTh 1 · 10�5 1.2 · 10�2 5.6 · 10�3 2.0 · 10�

DRa 5 · 10�8 1 · 10�5 1 · 10�5 9 · 10�8

All the data are taken from Blundy and Wood (2003).a Partition coefficients at 3 GPa which were used in the dynamic meltib Partition coefficients at 1.5 GPa which were used in the AFC calcula

chemical (isotopic) and mineralogical heterogeneity, con-sidered to be in 238U–230Th–226Ra equilibrium before theonset of melting. As shown in Fig. 8, �2–3% of accumulat-ed partial melts is expected to produce the La/Yb andGd/Yb ratios of the Mt. Cameroon lavas. However, it isobvious that a fractional melting model cannot explainthe elevated (230Th/238U) ratios with �20% excess 230Thin the samples unless we assume an unrealistically low(<0.1%) degree of accumulated partial melting which, inturn, conflicts with the La/Yb and Gd/Yb ratios observed.Such inconsistencies would be resolved by adopting thedynamic melting model (McKenzie, 1985; Williams andGill, 1989) that considers continuous extraction and accu-mulation of melt from upwelling mantle accompanied byin-growth of daughter nuclides in the residual solid. Fig. 9shows the calculated (226Ra/230Th) and (230Th/238U) ratiosfor dynamic melting of a dry mantle source containing6% residual garnet (ol:opx:cpx:grt:sp = 48:25:15:6:6) bychanging the rate of melting from 1 · 10�6 to 1 · 10�2

kg m�3 yr�1 and the matrix porosity from 1 · 10�3 to5 · 10�2, the parameters which cover the possible range inthe actual mantle melting conditions (Bourdon and Sims,2003). Partition coefficients (D) of U, Th and Ra were takenfrom Blundy and Wood (2003) (Table 2). For garnet andcpx, partition coefficients of MORB–pyrolite source at3 GPa were used. Such dynamic melting can create238U–230Th–226Ra disequilibria with large 230Th and 226Raexcesses (meshed zone in Fig. 9). However, the Mt. Cameroondata have low (226Ra/230Th) ratios at given (230Th/238U)ratios, that fall outside the possible 238U–230Th–226Ra com-bination in the dynamic melting model. Increase of garnetmodal abundance in the source mantle can shift the meshedzone downward in Fig. 9. In this case, however, more than14% of garnet in the source is needed to account for the ob-served Mt. Cameroon data, which contradicts the plausiblegarnet modal abundance in the source (�6%) as estimatedfrom Fig. 8. Chromatographic melting (Spiegelman andElliott, 1993) would draw the meshed zone in(226Ra/230Th)–(230Th/238U) space similar to that shown inFig. 9 (Bourdon and Sims, 2003), but the Mt. Cameroondata still would lie out of the zone.

Several possibilities could account for the observeddecoupling. First, choice of D values for U, Th and Ra dif-ferent from those used above would produce higher(230Th/238U) and/or lower (226Ra/230Th) ratios in the pri-mary melt with given physical parameters. For Ra, a twoorders of magnitude higher DRa

bulk=melt value (3 · 10�4) thanused above is needed to account for the Mt. Cameroondata. This is possible only when the source mantle has min-

calculations

Garnet Spinel Amphibole Phlogopite

3 1.9 · 10�2 1 · 10�5 2 · 10�2 2 · 10�4

3 4.7 · 10�3 1 · 10�6 2 · 10�2 2 · 10�4

7 · 10�9 0 3.2 · 10�2 1.5

ng calculation.tion.


erals with very high DRa such as amphibole and phlogopite,but these hydrous minerals are generally not stable at thethermal condition of asthenospheric mantle (Class andGoldstein, 1997). On the other hand, minerals having highDU/DTh (high DU) ratios can effectively fractionate U andTh and would produce high (230Th/238U) ratio while keep-ing the (226Ra/230Th) ratio low as shown in the Mt. Camer-oon samples. U and Th are preferentially partitioned intododecahedral X-site of garnet which is principally occupiedby Ca, Mg and Fe2+, and thereby chemical variation in gar-net can drastically change DU

grt=melt and DThgrt=melt. In many

cases, however, experimentally determined Dgrt/melt are notsufficiently high for DU/DTh and DU to produce the Mt.Cameroon data by dynamic melting (Beattie, 1993;LaTourrette et al., 1993; Hauri et al., 1994; Salters andLonghi, 1999; van Westrenen et al., 2000; Klemme et al.,2002; Salters et al., 2002). The only exception is grossular-rich garnet (grossular >60%, pyrope <40%) that hasextremely high DU

grt=melt (>0.3) (van Westrenen et al.,1999), which is not conceivable as garnet that commonlyexists in asthenospheric mantle. High-aluminous clinopy-roxene also preferentially incorpotares U than Th, com-pared to diopside-rich clinopyroxene (DU = 0.023,DTh = 0.019 at 1.9 GPa; Wood et al., 1999). However, theMt. Cameroon data cannot be reproduced even when allthe clinopyroxene in the source mantle is high-aluminousclinopyroxene.

A long magma transfer time from melt segregation toeruption (>2500 yr) would cause 226Ra decay in the meltwhich explains the low (226Ra/230Th) ratios observed in theMt. Cameroon samples. Although it is difficult to completelyreject this possibility, this would not be the main reason ofthe low (226Ra/230Th) ratios because (1) some Mt. Cameroonlavas contain cm-size xenoliths (Suh et al., 2003) which is anindication of rapid magma ascent rate as commonly suggest-ed for alkali basalts (McKenzie, 2000) and (2) well-estab-lished shallow magma chambers are possibly absentbeneath Mt. Cameroon (see Section 6.1). Otherwise it wouldrequire an alternative approach in the modeling of mantlemelting processes which involves disequilibrium mineral/melt partitioning for U-series nuclides as recently discussedby Bourdon and Van Orman (2006).

In summary, supposing the observed variations in traceelements, Pb isotopes and U-series disequilibria in the Mt.Cameroon samples are all of asthenospheric mantle origin,it requires some chemical heterogeneity in the source man-tle. Dynamic melting is required to produce the large238U–230Th disequilibrium with �20% 230Th excess,although a specific condition (extremely high DU/DTh andDU, long magma transfer time, or disequilibrium partition-ing) is needed to account for the observed high (230Th/238U)and low (226Ra/230Th) ratio.

6.4. Plume–lithospheric mantle interaction

6.4.1. Model calculations

Finally, we advocate and evaluate the possibility thatcontamination of primary melts (asthenospheric origin)by melting of sub-continental lithospheric mantle (SCLM)created the variation of trace elements, Pb isotopes and

U-series disequilibria observed in the Mt. Cameroon sam-ples. This scenario does not require chemical heterogeneityin the asthenospheric mantle. Such a plume–lithosphereinteraction model has been proposed to explain trace ele-ment and isotope signatures of some intraplate alkali bas-alts such as Grande Comore Island (Class and Goldstein,1997; Bourdon et al., 1998; Class et al., 1998; Claude-Ivanajet al., 1998), Canary Islands (Lundstrom et al., 2003), Ken-ya Rift (Macdonald et al., 2001) and the CVL (Marzoliet al., 2000; Rankenburg et al., 2005). This process has beenquantitatively modeled by Bourdon et al. (1998) in which amodified assimilation-fractional crystallization (AFC)model was applied. In the modified AFC model, they con-sidered partial (batch) melting of the lithosphere, in whichthe concentration of an element in the melt (Cm) is given by

Cm ¼ Cm0 � F �z þ r

r � 1

� �� Ca

0

zðDa þ f ð1� DaÞÞ� ð1� F �zÞ

ð1Þ

where Cm0 and Ca

0 are initial concentrations in melt andassimilant, respectively, F is the fraction of melt remaining,f is the degree of melting of the lithosphere, r is the ratio ofassimilation rate and crystallization rate, z = (r + Dc � 1)/(r � 1), Dc and Da are solid/melt partition coefficients forcrystallizing solid and assimilant, respectively. When f isclose to 1, it implies bulk assimilation has occurred in thelithospheric mantle.

We have applied Eq (1) to our Mt. Cameroon data. Inthe calculation, we assumed that all AFC parameters butF values are common for the Mt. Cameroon samples. Inother words, variations of trace elements and Pb and U-se-ries isotopes are produced only by changing F in the AFCprocess. We also considered that a magma with a composi-tion similar to sample #1982-1 in terms of Pb isotope ratiosand U-series disequilibria interacted with the SCLM via theAFC process because the 1982 samples have the highest(230Th/238U) and (226Ra/230Th)0 and the lowest 206Pb/204Pbratios. It should be noted that major and trace elementcompositions of the injected magma do not exactly matchthose of #1982-1 owing to the subsequent fractionation be-fore eruption (Fig. 5c). The AFC calculation is, therefore,focused mainly on Pb and U-series isotope ratios whichare not changed largely by fractional crystallization. TheU/Pb ratio is also used in the calculation because this ratiochanges only by 0.4% for even 50% of fractional crystalliza-tion with 0.85cpx + 0.15ol.

We tested three different mineralogical compositions forthe SCLM: anhydrous, amphibole-bearing and phlogopite-bearing spinel lherzolite. Partition coefficients summarizedin Blundy and Wood (2003) were used for U, Th and Ra(Table 2), and that of Halliday et al. (1995) were appliedfor Pb. The crystallizing phase was determined to be0.85cpx + 0.15ol as described in Section 5.1. A subtlechange of the crystallization phase does not affect the resultsignificantly. For simplification, all U-series nuclidesincluding 226Ra were treated as stable isotopes, assumingthat the duration of AFC was very short compared to thehalf-life of 226Ra (1600 yr). We also assumed that theSCLM was in 238U–230Th–226Ra equilibrium prior to theonset of the AFC process. This implies that Ca

0 for 230Th


and 226Ra are automatically determined once Ca0 for U is

given. Then, unknown parameters in the AFC calculationare r, f, Ca

0 for U and Pb, and Pb isotope ratios of theSCLM.

Fig. 10 represents the AFC trajectories plotted on the(226Ra/230Th)0–(230Th/238U) diagram by changing F and f

values from 1 to 0.5 and from 0.001 to 1, respectively.The Ca

0 for U and the r value are fixed at 1 lg/g and 0.2,respectively. Note that Pb concentration and isotopic com-position of SCLM do not affect the trajectories in this fig-ure. We found no combination of AFC parameters underanhydrous SCLM condition that reconcile AFC trajecto-ries, which suggests the existence of accessory mineral(s)that have very high DRa such as amphibole and phlogopitein the SCLM. The minimum amount of the accessory

a

b

c

Fig. 10. AFC trajectories in (226Ra/230Th)0–(230Th/238U) diagramfor different mineral compositions of SCLM: (a) anhydrous spinellherzolite (ol:opx:cpx:sp = 62:24:12:2), (b) amphibole-bearing spi-nel lherzolite (ol:opx:cpx:sp:amp = 62:24:7:2:5) and (c) phlogopite-bearing spinel lherzolite (ol:opx:cpx:sp:phl = 62:24:11:2:1). Urani-um concentration in SCLM and r value are fixed as 1 lg/g and 0.2,respectively. Values of F and f are changed from 1 to 0.5 and from0.001 to 1, respectively. Gray zone in (b) is the case for 3%amphibole-bearing SCLM (ol:opx:cpx:sp:amp = 62:24:9:2:3).

mineral required to account for the Mt. Cameroon data is5% and 0.1% for amphibole and phlogopite, respectively.The AFC trajectories fall outside the range of Mt. Camer-oon data when the abundance of the hydrous mineral is low(Fig. 10b). It is important to note that the slopes of theAFC trajectories in Fig. 10 are insensitive to Ca

0 for Uand r values. By applying least squares calculation in thisdiagram, we can obtain the f value.

The rest of the AFC parameters were determined byleast squares calculation in (226Ra/230Th)0–(230Th/238U)–206Pb/204Pb–U/Pb space at a given mineralogical composi-tion of the SCLM. Table 3 summarizes the results of theAFC calculation when the modal abundance of amphiboleand phlogopite in the SCLM is varied from 5% to 10% and0.5% to 3%, respectively. Fig. 11 shows the AFC trajecto-ries when the SCLM contains 2% of phlogopite, r = 0.2and Ca

0 for Pb is 0.2 lg/g. Similar to Fig. 10, at givenAFC parameters, the slopes of the AFC trajectories inFig. 11 are insensitive to r value. The slopes are also insen-sitive to Ca

0 for Pb so long as U/PbSCLM is constant, imply-ing that r and Ca

0 for U and Pb are not uniquelydetermined by least squares calculation. Instead, the shapeof the AFC trajectory is strongly controlled by f as well asPb isotope composition, U/Pb ratio and mineralogical com-position of the SCLM. The deviation of the 1922 samplesfrom the main AFC trend can be explained by applying dif-ferent AFC parameters (Fig. 11). However, it could be pos-sible that primary magma for the 1922 samples derivedfrom the asthenospheric mantle had a chemical composi-tion different from the rest of the samples. This is becausethe 1922 samples are the most mafic, with the highestLa/Yb and Gd/Yb ratios and the lowest extent of U–Th–Radisequilibria, and intermediate 206Pb/204Pb ratios of theMt. Cameroon samples studied.

The determined f values are not close to 1 but small (<0.1;Table 3). Such small degree melts are not enriched in 226Raowing to high DRa of the hydrous SCLM, of which the(226Ra/230Th) is <1 in some cases. This helps to decrease(226Ra/230Th) ratio in the magma drastically via the AFCprocess. The involvement of the small degree melt potentiallyenhances La/Yb ratio in the magma because the SCLM has alow DLa/DYb (�0.1), which is consistent with the observation(Fig. 8). However, the initial La/Yb ratio of the SCLM is dif-ficult to be estimated because it is also influenced by cpx crys-tallization in the AFC process. Again, the change of majorand trace element concentrations in the magma via SCLMinteraction is further difficult to be quantitatively evaluatedby AFC calculation owing to subsequent magma differentia-tion that possibly occurred before eruption (Fig. 5c).

6.4.2. Origin of SCLM beneath Mt. Cameroon

The constraints obtained in the previous section give usthe chance to further speculate on the origin of the SCLMbeneath Mt. Cameroon. One of the most intriguing featuresof the AFC calculation is that accessory mineral(s) havingvery high DRa such as amphibole and phlogopite arerequired in the SCLM. The existence of hydrous mineralsin the lithospheric mantle where asthenospheric meltinfiltrates and reacts has previously been proposed for thepetrogenesis of some intraplate volcanic rocks, based on

Table 3Optimized AFC parameters for amphibole- and phlogopite-bearing SCLM

Amp (%) 206Pb/204Pb 207Pb/204Pb 208Pb/204Pb U/Pb f l Tm (Ma)

Amphibole-bearing SCLM

5 20.47 15.66 40.22 0.75 0.0029 50 1416 20.50 15.67 40.23 0.80 0.0035 54 1307 20.52 15.67 40.24 0.85 0.0039 57 1238 20.54 15.67 40.25 0.88 0.0044 59 1189 20.56 15.67 40.26 0.91 0.0048 61 113

10 20.58 15.67 40.27 0.94 0.0052 63 110

Phl (%) 206Pb/204Pb 207Pb/204Pb 208Pb/204Pb U/Pb f l Tm (Ma)

Phlogopite-bearing SCLM

0.5 20.79 15.69 40.38 1.23 0.019 83 871.0 20.85 15.69 40.41 1.29 0.034 87 841.5 20.91 15.70 40.43 1.40 0.048 94 792.0 20.93 15.70 40.44 1.42 0.062 96 782.5 20.94 15.70 40.45 1.45 0.075 98 773.0 20.97 15.70 40.47 1.52 0.088 103 73

207Pb/204PbSCLM and 208Pb/204PbSCLM were determined by extrapolating the trends of the Mt. Cameroon samples in the207Pb/204Pb–206Pb/204Pb and 208Pb/204Pb–206Pb/204Pb diagrams, respectively.


the depletion in K, Ba and Rb (Class and Goldstein, 1997;Class et al., 1998; Macdonald et al., 2001), on melting mod-els using Gd/Yb–La/Yb ratios (Marzoli et al., 2000) and onthe variation of Nb/U ratios (Lundstrom et al., 2003).Compared to these approaches, our results, as well asClaude-Ivanaj et al. (1998) and Bourdon et al. (1998), areinstead based on U-series disequilibria that are not affectedby chemical composition of the asthenospheric mantle.McDonough (1990) estimated the chemical compositionof the continental lithospheric mantle by compiling pub-lished mantle xenolith data. He found that the median com-position of U and Pb would provide more reasonableestimate than the average composition, and a U/Pb =0.25 was obtained for the continental lithospheric mantle.Thus, the hydrous SCLM beneath Mt. Cameroon is sup-posed to have elevated U/Pb ratio (>0.75) compared tocommon continental lithospheric mantle, together withradiogenic Pb isotope ratios (206Pb/204Pb > 20.47) whichis obviously out of the range of FOZO (Table 3; Fig. 5a).This highlights the peculiarity regarding the Pb isotopecomposition in the SCLM, and it requires a certain eventto have increased the Pb isotope ratios. Radiogenic Pb(206Pb/204Pb = 21.00, 207Pb/204Pb = 15.57, 208Pb/204Pb =40.68: ‘‘A’’ in Fig. 5a) in ultramafic xenolith of Biu Plateau,northern CVL, is considered to derive from metasomatizedSCLM (Lee et al., 1996). Lee et al. (1996) also reported veryhigh U/Pb ratios (maximum value = 1.90) in some SCLM-derived xenoliths obtained from the CVL, although suchsamples have less radiogenic 206Pb/204Pb ratios than theMt. Cameroon lavas. Recently, Rankenburg et al. (2005)reported enriched Pb isotope signature in some megacrystsfrom Biu and Jos Plateaux, northern CVL (206Pb/204Pb =19.82–20.90, 207Pb/204Pb = 15.67–15.76, 208Pb/204Pb =39.56–40.95: ‘‘B’’ and ‘‘C’’ in Fig. 5a). They insisted thatthese megacrysts crystallized from magmas that had beencontaminated by metasomatically enriched SCLM viaassimilation. It is, therefore, possible that the SCLMbeneath Mt. Cameroon also had radiogenic Pb isotope

composition and high U/Pb ratio inherited from previousmetasomatic event(s). Our estimate of Pb isotope ratiosfor the SCLM lies on the regression line in Fig. 5a, and isdifferent from values measured in enriched xenoliths andmegacrysts from northern CVL. This presumably indicatesthat the SCLM underlying the CVL c.o.b. and its continentsector is metasomatized but is heterogeneous in Pb (andprobably other long-lived radiogenic) isotopes. Actually,megacrysts from Biu Plateau have more radiogenic Pb iso-topes compared to Jos Plateau (Rankenburg et al., 2005).

The timing and enrichment process of CVL SCLM isstill controversial. Halliday et al. (1990) proposed that thehigh 206Pb/204Pb ratios for the CVL c.o.b. arises from re-cent (125 Ma) variable fractionation of U/Pb by theemplacement of the St. Helena fossil plume, and that mag-mas of CVL volcanism are derived from remelting of thefossil plume. As shown in Fig. 5a, our Mt. Cameroon datahave a steep slope which corresponds to 2.43 Ga isochron.If radiogenic Pb isotope ratios in the Mt. Cameroon sam-ples were inherited from melting of the metasomatizedSCLM without interaction with melt from the underlyingasthenosphere, then the 2.43 Ga would represent the timeof SCLM metasomatism. This isochron age is far greaterthan the timing suggested by Halliday et al. (1990). We pro-pose instead that the trend of the Mt. Cameroon samples inFig. 5a is an AFC trajectory and that the isochron age isonly apparent and does not represent the timing of SCLMmetasomatism.

Supposing the initial Pb isotopic ratios of the SCLMlaid on the NHRL at the time of U/Pb fractionation, wecan calculate the age of SCLM metasomatism (Tm) by solv-ing the following equations:

R64 ¼ R640 þ R84 � ðek238U�T m � 1Þ ð2Þ

R74 ¼ R740 þ R54 � ðek235U�T m � 1Þ ð3Þ

R640 þ R84

0 � ðek238U�T m � 1Þ ¼ 0:1084 � fR740 þ R54

0�ðek235�T m � 1Þg þ 13:491 ð4Þ

a

b

c

Fig. 11. AFC trends in (a) (226Ra/230Th)0–206Pb/204Pb, (b)(230Th/238U)–206Pb/204Pb and (c) U/Pb–206Pb/204Pb diagrams forthe Mt. Cameroon lavas when the SCLM contains 2% phlogopite.Pb concentration of the SCLM and r are fixed as 0.2 lg/g and 0.2,respectively. Bold curves are the AFC trends optimized for thesamples excluding 1922 lavas (206Pb/204PbSCLM = 20.93, f = 0.062and U/PbSCLM = 1.42) and dashed curves are those for 1922 lavas(206Pb/204PbSCLM = 20.56, f = 0.037 and U/PbSCLM = 1.02).


where R64, R74, R84, and R54 are 206Pb/204Pb, 207Pb/204Pb,238U/204Pb and 235U/204Pb ratios of present SCLM, R withsuperscript 0 represents the ratio of SCLM before metaso-matism. The R84

0 and R540 are calculated by assuming that

the U/Pb ratio of unmetasomatized SCLM is 0.25 (McDon-ough, 1990). The rest of R values are taken from Table 3.The calculated Tm ranges from 73 to 141 Ma, and doesnot differ significantly when the U/Pb0 ratio is changedfrom 0.2 to 0.3. Consequently, we conclude that the U-se-ries and Pb isotope signatures in the Mt. Cameroon sam-ples were produced by the interaction of asthenosphericmelt with the SCLM metasomatized in late Mesozoic, pre-sumably related to ancient St. Helena plume activity. Itshould be noted, however, that our conclusion is based

on many assumptions which involve some uncertainties:the exact composition of the SCLM is not well constrained,the SCLM is assumed to have evolved following the NHRLuntil the time of metasomatism, and the NHRL is not anarrow line but actually the average of various MORBand OIB data (Hart, 1984). Contrary to the recent metaso-matic model, Rankenburg et al. (2005) proposed multiplemetasomatic events in the old (Proterozoic) lithosphere toproduced the enriched isotope signature of the SCLM.Unfortunately, such multiple metasomatic events are notpossible to inversely follow using our data estimated forthe SCLM. No matter the timing of SCLM metasomatism,it is important to note that our Mt. Cameroon data requirethe involvement of asthenospheric melt with the metasoma-tized SCLM having high Pb isotope and high U/Pb ratios.

As previously mentioned, but for Pb, there are no obvi-ous changes in the other radiogenic (Sr, Nd and Th) iso-topes of the Mt. Cameroon samples. This is presumablydue to the fact that isotopically homogeneous melt derivedfrom the asthenospheric mantle interacted with the SCLMwhich has Sr, Nd and Th isotope compositions similar tothe injected magma and/or has Sr, Nd and Th compositionsthat are too low to modify the original isotopic composi-tions. We thus consider that Sr, Nd and Th isotope compo-sitions of the Mt. Cameroon samples are inherited fromthose of the asthenospheric mantle which has been influ-enced by a slight contribution from a HIMU-like compo-nent (Section 6.3.1). If this is the case, the asthenosphericmantle should have a radiogenic Pb isotopic compositioncompared to ubiquitous mantle component. Such a signa-ture is strengthened by the interaction with SCLM whichhas even more radiogenic Pb isotopic composition owingto the SCLM metasomatism.

7. CONCLUSION

(1) Lavas of Mt. Cameroon erupted within the last100 yr possess 238U–230Th–226Ra disequilibria with 230Th(18–24%) and 226Ra (9–21%) excesses, and there exists a po-sitive correlation in a (226Ra/230Th)0–(230Th/238U) diagram.The extent of 238U–230Th–226Ra disequilibria is markedlydifferent in lavas of individual eruption ages, although the(230Th/232Th) ratio is constant irrespective of eruptionage. When U-series results are combined with Pb isotoperatios, negative correlations are observed in (230Th/238U)–(206Pb/204Pb) and (226Ra/230Th)0–(206Pb/204Pb) diagrams.

(2) Shallow magma chamber processes (binary mixing ofchemically different magmas, fractional crystallizationinvolving 230Th and 226Ra decay, and wall rock assimila-tion) cannot explain the data. Crustal contamination isnot the cause of the Pb and U-series isotope trends becausecontinental crust is considered to have extremely differentPb isotope composition and U/Th ratios.

(3) If the observed variations in trace elements, Pb iso-topes and U-series disequilibria are all of asthenosphericmantle origin, then this requires chemical heterogeneity inthe source mantle which has a slight contribution ofHIMU-like component to ubiquitous FOZO component.Dynamic melting would produce the large 238U–230Th dis-equilibrium, although a particular condition (extremely


high DU/DTh and DU,, long magma transfer time or dis-equilibrium partitioning) is required to account for the ob-served high (230Th/238U) and low (226Ra/230Th) ratios.

(4) The most favorable model to explain the Mt. Camer-oon data is interaction of asthenosphere-derived melts withthe SCLM. A model calculation requires the existence of anaccessory mineral phase having high DRa such as amphi-bole and/or phlogopite in the SCLM. Such SCLM wouldhave radiogenic Pb (206Pb/204Pb > 20.47) as well as elevatedU/Pb (>0.75) ratios, created by metasomatism presumablyin late Mesozoic and related to St. Helena plume activity.

ACKNOWLEDGMENTS

We thank A. Makishima, R. Tanaka, T. Kuritani and all themember of PML for their technical help and discussion. M. Men-zies, B. Bourdon and two anonymous reviewers are thanked forhandling and reviewing the manuscript. This research was support-ed by grants from MEXT of the Japanese Government to E.N. andJSPS grants for ‘‘Asia-Africa Science Platform Program (Geo-chemistry of the Lake Nyos gas disaster, Cameroon VolcanicLine—Rift Valley volcanoes and the underlying mantle)’’ and‘‘Center of Excellence in the 21st Century in Japan’’ to the Institutefor Study of the Earth’s Interior, Okayama Univ. Japan. The sam-pling was supported by a grant from MEXT to M.K. (#16403012).AFT gratefully acknowledges support from the administration andcolleagues of IRGM/ARGV.

REFERENCES

Aka F. T., Nagao K., Kusakabe M., Sumino H., Tanyileke G.,Ateba B., and Hell J. (2004) Symmetrical helium isotopedistribution on the Cameroon Volcanic Line, West Africa.Chem. Geol. 203, 205–223.

Ambeh W. B., Fairhead J. D., Francis D. J., Nnange J. M., andDjallo S. (1989) Seismicity of the Mount Cameroon Region,West-Africa. J. African Earth Sci. 9(1), 1–7.

Asmerom Y., DuFrane S. A., Mukasa S. B., Cheng H., andEdwards R. L. (2005) Time scale of magma differentiation inarcs from protactinium-radium isotopic data. Geology 33,

633–636.

Beattie P. (1993) Uranium-thorium disequilibria and partitioningon melting of garnet peridotite. Nature 363, 63–65.

Blundy J. and Wood B. (2003) Mineral-melt partitioning ofuranium, thorium and their daughters. In Uranium-series

Geochemistry, vol. 52 (eds. B. Bourdon, G.M. Henderson,C.C. Lundstrom, and S.P. Turner). Geochemical Society,Mineralogical Society of America, pp. 59–123.

Bourdon B., Joron J.-L., Claude-Ivanaj C., and Allegre C. J. (1998)U–Th–Pa–Ra systematics for the Grande Comore volcanics:melting processes in an upwelling plume. Earth Planet. Sci.

Lett. 164, 119–133.

Bourdon B. and Sims K.W.W. (2003) U-series constraints onintraplate basaltic magmatism. In Uranium-series Geochemistry,vol. 52, pp. 215–254. Geochemical Society, MineralogicalSociety of America.

Bourdon B., Turner S. P., and Ribe N. M. (2005) Partial meltingand upwelling rates beneath the Azores from a U-series isotopeperspective. Earth Planet. Sci. Lett. 239, 42–56.

Bourdon B., and Van Orman J. (2006) 226Ra deficits in OIB: a keyto the rate of melt extraction in the mantle. Geochim.

Cosmochim. Acta 70, A61.

Chabaux F., and Allegre C. J. (1994) 238U–230Th–226Ra disequi-libria in volcanics: a new insight into melting constraints. Earth

Planet. Sci. Lett. 126, 61–74.

Chauvel C., Dia A. N., Bulourde M., Chabaux F., Durand S.,Ildefonse P., Gerard M., Deruelle B., and Ngounouno I. (2005)Do decades of tropical rainfall affect the chemical compositionsof basaltic lava flows in Mount Camerooon? J. Anal. Atom.

Spectrom. 141, 195–223.

Cheng H., Edwards R. L., Hoff J., Gallup C. D., Richards D. A.,and Asmerom Y. (2000) The half-lives of uranium-234 andthorium-230. Chem. Geol. 169, 17–33.

Class C., and Goldstein S. L. (1997) Plume-lithosphere interactionsin the ocean basins: constraints from the source mineralogy.Earth Planet. Sci. Lett. 150(3–4), 245–260.

Class C., Goldstein S. L., Altherr R., and Bachelery P. (1998) Theprocess of plume–lithosphere interactions in the ocean basins—the case of Grande Comore. J. Petrol. 39(5), 881–903.

Claude-Ivanaj C., Bourdon B., and Allegre C. J. (1998) Ra–Th–Srisotope systematics in Grande Comore Island: a case study ofplume–lithosphere interaction. Earth Planet. Sci. Lett. 164,

99–117.

Condomines M., Tanguy J.-C., and Michaud V. (1995) Magmadynamics at Mt Etna: constraints from U–Th–Ra–Pb radioac-tive disequilibria and Sr isotopes in historical lavas. Earth

Planet. Sci. Lett. 132, 25–41.

Dada S. S., Briqueu L., Harms U., Lancelot J. R., and Matheis G.(1995) Charnockitic and Monzonitic Pan-African series fromNorth-Central Nigeria—trace-element and Nd, Sr, Pb isotopeconstraints on their petrogenesis. Chem. Geol. 124(3-4), 233–252.

Dickin A. P., Halliday A. N., and Bowden P. (1991) A Pb, Sr andNd isotope study of the basement and Mesozoic ring complexesof the Jos Plateau, Nigeria. Chem. Geol. 94(1), 23–32.

Fitton J. G., and Dunlop H. M. (1985) The Cameroon Line, West-Africa, and its bearing on the origin of Oceanic and Continentalalkali basalt. Earth Planet. Sci. Lett. 72(1), 23–38.

Fitton J. G., Kilburn R. J., Thirlwall M. F., and Hughes D. J.(1983) 1982 eruption of Mount Cameroon, West Africa. Nature

306, 327–332.

Geze B. (1953) Les Volcans du Cameroon occidental. Bull.

Volcanol. 13, 63–92.

Green D. H., Edgar A. D., Beasley P., Kiss E., and Ware N. G.(1974) Upper mantle source for some Hawaiites, Mugearitesand Benmoreites. Contrib. Mineral. Petrol. 48(1), 33–43.

Halliday A. N., Davidson J. P., Holden P., DeWolf C., Lee D.-C.,and Fitton J. G. (1990) Trace-element fractionation in plumesand the origin of HIMU mantle beneath the Cameroon line.Nature 347, 523–528.

Halliday A. N., Dickin A. P., Fallick A. E., and Fitton J. G. (1988)Mantle dynamics: a Nd, Sr, Pb and O isotopic study of theCameroon line volcanic chain. J. Petrol. 29, 181–211.

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

Hart S. R. (1984) A large-scale isotope anomaly in the SouthernHemisphere mantle. Nature 309, 753–757.

Hart S. R., and Davis K. E. (1978) Nickel partitioning betweenolivine and silicate melt. Earth Planet. Sci. Lett. 40(2), 203–219.

Hauri E. H., Wagner T. P., and Grove T. L. (1994) Experimentaland natural partitioning of Th, U, Pb and other trace elementsbetween garnet, clinopyroxene and basaltic melts. Chem. Geol.

117, 149–166.

Jaffey A. H., Flynn K. F., Glendenin L. E., Bentley W. C., andEssling A. M. (1971) Precision measurements of half-lives andspecific activities of 235U and 238U. Phys. Rev. C 4, 1889.

Klemme S., Blundy J. D., and Wood B. J. (2002) Experimentalconstraints on major and trace element partitioning duringpartial melting of eclogite. Geochim. Cosmochim. Acta 66(17),

3109–3123.


Kuritani T., and Nakamura E. (2003) Highly precise and accurateisotopic analysis of small amounts of Pb using 205Pb–204Pb and207Pb–204Pb, two double spikes. J. Anal. Atom. Spectrom. 18,

1464–1470.

LaTourrette T. Z., Kennedy A. K., and Wasserburg G. J. (1993)Thorium–uranium fractionation by garnet: evidence for a deepsource and rapid rise of oceanic basalts. Science 261, 739–742.

Le Bas M. J., Le Maitre R. W., Streckeien A., and Zanettin B.(1986) A chemical classification of volcanic rocks based on thetotal alkali-silica diagram. J. Petrol. 27, 745–750.

Le Roux L. J. and Glendenin L. E. (1963) Half-life of 232Th. InProceedings of the National Meeting on Nuclear Energy,

Pretorica, South Africa, pp. 83–94.Lee D. C., Halliday A. N., Davies G. R., Essene E. J., Fitton J. G.,

and Temdjim R. (1996) Melt enrichment of shallow depletedmantle: a detailed petrological, trace element and isotopic studyof Mantle-Derived xenoliths and megacrysts from the Camer-oon line. J. Petrol. 37(2), 415–441.

Lu Y.-H., Makishima A., and Nakamura E. (2007) Coprecipitationof Ti, Mo, Sn and Sb with fluorides and application todetermination of B, Ti, Zr, Nb, Mo, Sn, Sb, Hf and Ta by ICP-MS. Chem. Geol. 236, 13–26.

Lundstrom C. C., Hoernle K., and Gill J. (2003) U-seriesdisequilibria in volcanic rocks from the Canary Islands: plumeversus lithospheric melting. Geochim. Cosmochim. Acta 67,

4153–4177.

Macdonald R., Rogers N. W., Fitton J. G., Black S., and Smith M.(2001) Plume-lithosphere interactions in the generation of thebasalts of the Kenya Rift, East Africa. J. Petrol. 42(5), 877–900.

Makishima A., and Masuda A. (1994) Ce isotope ratios of N-typeMORB. Chem. Geol. 118, 1–8.

Makishima A., and Nakamura E. (2006) Determination of major,minor and trace elements in silicate samples by ICP-QMS andICP-SFMS applying isotope dilution-internal standardization(ID-IS) and multi-stage internal standardization. Geostandard.

Geoanalytic. Res. 30, 245–271.

Marzoli A., Piccoirillo E. M., Renne P. R., Bellieni G., IacuminJ. B., Nyobe J. B., and Tongwa A. T. (2000) The Cameroonvolcanic line revisited: petrogenesis of continental basalticmagmas from lithospheric and asthenospheric mantle source.J. Petrol. 41, 87–109.

McDonough W. F. (1990) Constraints on the composition of thecontinental lithospheric mantle. Earth Planet. Sci. Lett. 101,

1–18.

McKenzie D. (1985) 230Th–238U disequilibrium and the meltingprocesses beneath ridge axes. Earth Planet. Sci. Lett. 72, 149–157.

McKenzie D. (2000) Constraints on melt generation and transportfrom U-series activity ratios. Chem. Geol. 162, 81–94.

Nakamura E., Makishima A., Moriguti T., Kobayashi K., Sakag-uchi C., Yokoyama T., Tanaka R., Kuritani T., and Takei H.(2003) Comprehensive geochemical analyses of small amounts(<100 mg) of extraterrestrial samples for the analytical compe-tition related to the sample return mission MUSES-C. Inst.

Space Astronaut. Sci. Rep. SP 16, 49–101.

Pietruszka A. J., Hauri E. H., Carlson R. W., and Garcia M. O.(2006) Remelting of recently depleted mantle within theHawaiian plume inferred from the 226Ra–230Th–238U disequi-libria of Pu’O’o eruption lavas. Earth Planet. Sci. Lett. 244,

155–169.

Rankenburg K., Lassiter J. C., and Brey G. (2005) The role ofcontinental crust ant lithospheric mantle in the genesis ofCameroon volcanic line lavas: constraints from isotopic vari-ations in lavas and megacrysts from the Biu and Jos Plateaux. J.

Petrol. 46, 169–190.

Reiners P. W. (2002) Temporal-compositional trends in intraplatebasalt eruptions: implications for mantle heterogeneity and

melting processes. Geochem. Geophys. Geosyst. 3. doi:10.1029/

2001/GC000250.

Rudnick R. L., and Gao S. (2003) Compositioin of the continentalcrust. In Treatise on Geochemistry (eds. H.D. Holland and K.K.Turekian), vol. 3. Elsevier-Pergamon, pp. 1–64.

Salters V. J. M., and Longhi J. (1999) Trace element partitioningduring the initial stages of melting beneath mid-ocean ridges.Earth Planet. Sci. Lett. 166(1–2), 15–30.

Salters V. J. M., Longhi J. E., and Bizimis M. (2002) Near mantlesolidus trace element partitioning at pressures up to 3.4 GPa.Geochem. Geophys. Geosyst. 3. doi:10.1029/2001GC000148.

Sigmarsson O. (1996) Short magma chamber residence time at anIcelandic volcano inferred from U-series disequilibria. Nature

382, 440–442.

Sigmarsson O., Condomines M., and Bachelery P. (2005) Magmaresidence time beneath the Piton de la Fournaise Volcano,Reunion Island, from U-series disequilibria. Earth Planet. Sci.

Lett. 234, 223–234.

Sims K. W. W., DePaolo D. J., Murrell M. T., Baldridge W. S.,Goldstein S., Clague D., and Jull M. (1999) Porosity of themelting zone and variations in the solid mantle upwelling ratebeneath Hawaii: inferences from 238U–230Th–226Ra and235U–231Pa disequilibria. Geochim. Cosmochim. Acta 63, 4119–

4138.

Sims K. W. W., DePaolo D. J., Murrell M. T., Baldridge W. S.,Goldstein S. J., and Clague D. A. (1995) Mechanisms of magmageneration beneath Hawaii and mid-ocean ridges: uranium/thorium and samarium/neodymium isotopic evidence. Science

267, 508–512.

Sims K. W. W., and Hart S. R. (2006) Comparison of Th, Sr, Ndand Pb isotopes in oceanic basalts: implications for mantleheterogeneity and magma genesis. Earth Planet. Sci. Lett. 245,

743–761.

Spiegelman M., and Elliott T. (1993) Consequences of melttransport for uranium series disequilibrium in young lavas.Earth Planet. Sci. Lett. 118, 1–20.

Stracke A., Bizimis M., and Salters V. J. M. (2003) Recyclingoceanic crust: quantitative constraints. Geochem. Geophys.

Geosyst. 4. doi:10.1029/2001GC000223.

Stracke A., Hofmann A. W., and Hart S. R. (2005) FOZO, HIMU,and the rest of the mantle zoo. Geochem. Geophys. Geosyst. 6.

doi:10.1029/2004GC000824.

Stracke A., Salters V. J. M., and Sims K. W. W. (1999) Assessingthe presence of garnet–pyroxenite in the mantle sources ofbasalts through combined hafnium–neodymium–thorium iso-tope systematics. Geochem. Geophys. Geosyst. 1. doi:10.1029/

1999GC000013.

Suh C. E., Sparks R. S. J., Fitton J. G., Ayonghe S. N., Annen C.,Nana R., and Luckmen A. (2003) The 1999 and 2000 eruptionsof Mount Cameroon: eruption behaviour and petrochemistryof lava. Bull. Volcanol. 65, 267–281.

Sun S.S. and McDonough W.F. (1989) Chemical and isotopicsystematics of oceanic basalts: implications for mantle compo-sition and processes. In Magmatism in the ocean basins, vol. 42(eds. A. D. Sau nders and M. J. Norry). Geological SocietySpecial Publication, pp. 313–345.

Tanaka R., Makishima A., Kitagawa H., and Nakamura E. (2003)Suppression of Zr, Nb, Hf, and Ta coprecipitation in fluoridecompounds for determination in Ca-rich materials. J. Anal.

Atom. Spectrom. 18, 1458–1463.

Thirlwall M. F. (1997) Pb isotopic and elemental evidence for OIBderivation from young HIMU mantle. Chem. Geol. 139, 51–74.

Thomas L. E., Hawkesworth C. J., Van Calsteren P., Turner S. P.,and Rogers N. W. (1999) Melt generation beneath oceanislands: a U–Th–Ra isotope study from Lanzarote in theCanary Islands. Geochim. Cosmochim. Acta 63, 4081–4099.

http://dx.doi.org/10.1029/2001/GC000250

http://dx.doi.org/10.1029/2001/GC000250

http://dx.doi.org/10.1029/2001GC000148

http://dx.doi.org/10.1029/2001GC000223

http://dx.doi.org/10.1029/2004GC000824

http://dx.doi.org/10.1029/1999GC000013

http://dx.doi.org/10.1029/1999GC000013


Turner S., Hawkesworth C., Rogers N., and King P. (1997) U–Thisotope disequilibria and ocean island basalt generation in theAzores. Chem. Geol. 139, 145–164.

van Westrenen W., Blundy J., and Wood B. (1999) Crystal-chemical controls on trace element partitioning between garnetand anhydrous silicate melt. Am. Mineral. 84(5-6), 838–847.

van Westrenen W., Blundy J. D., and Wood B. J. (2000) Effect ofFe2+ on garnet-melt trace element partitioning: experiments inFCMAS and quantification of crystal-chemical controls innatural systems. Lithos 53(3–4), 189–201.

Vigier N., Bourdon B., Joron J. L., and Allegre C. J. (1999) U-decay series and trace element systematics in the 1978 eruptionof Ardoukoba, Asal rift: timescale of magma crystallization.Earth Planet. Sci. Lett. 174, 81–97.

Willbold M., and Stracke A. (2006) Trace element composition ofmantle end-members: implications for recycling of oceanic andupper and lower continental crust. Geochem. Geophys. Geosyst.

7. doi:10.1029/2005GC001005.

Williams R. W., and Gill J. B. (1989) Effects of partial melting onthe uranium decay series. Geochim. Cosmochim. Acta 53, 1607–

1619.

Williams R. W., and Gill J. B. (1992) Th isotope and U-seriesdisequilibria in some alkali basalts. Geophys. Res. Lett. 19,

139–142.

Wood B. J., Blundym J. D., and Robinson J. A. C. (1999) The roleof clinopyroxene in generating U-series disequilibrium duringmantle melting. Geochim. Cosmochim. Acta 63, 1613–1620.

Workman R. K., Hart S. R., Jackson M., Regelous M., FarleyK. A., Blusztajn J., Kurz M., and Staudigel H. (2004)Recycled metasomatized lithosphere as the origin of theenriched mantle II (EM2) end-member: evidence from the

Samoan volcanic chain. Geochem. Geophys. Geosyst. 5.

doi:10.1029/2003GC000623.

Yokoyama T., Kobayashi K., Kuritani T., and Nakamura E.(2003) Mantle metasomatism and rapid ascnet of slab compo-nents beneath island arc: evidence from 228U–230Th–226Radisequilibria of Miyakejima volcano, Izu arc, Japan. J.

Geophys. Res. 108, No. B7. doi:10.1029/2002JB002103.

Yokoyama T., Kuritani T., Kobayashi K., and Nakamura E.(2006) Geochimical evoluiton of a shallow magama plumbingsystem during the last 500 years, Miyakejima volcano, Japan:constraints from 238U–230Th–226Ra systematics. Geochim. Cos-

mochim. Acta 70, 2885–2901.

Yokoyama T., Makishima A., and Nakamura E. (1999a) Evalu-ation of the coprecipitation of incompatible trace elements withfluoride during silicate rock dissolution by acid digestion.Chem. Geol. 157, 175–187.

Yokoyama T., Makishima A., and Nakamura E. (1999b) Separa-tion of thorium and uranium from silicate rock samples usingtwo commercial extraction chromatographic resins. Anal.

Chem. 71, 135–141.

Yokoyama T., Makishima A., and Nakamura E. (2001) Preciseanalysis of 234U/238U ratio using UO2

þ ion with thermalionization mass spectrometry for natural samples. Chem. Geol.

181, 1–12.

Yokoyama T., and Nakamura E. (2004) Precise analysis ofradium isotope ratio for short-lived U-series disequilibria innatural samples by total evaporation thermal ionization massspectrometry (TE-TIMS). J. Anal. Atom. Spectrom. 19,

717–727.

Associate editor: Martin A. Menzies

http://dx.doi.org/10.1029/2005GC001005

http://dx.doi.org/10.1029/2003GC000623

http://dx.doi.org/10.1029/2002JB002103

Date post:	16-Nov-2023
Category:	Documents
Upload:	okayama-u
View:	0 times
Download:	0 times

Plume–lithosphere interaction beneath Mt. Cameroon volcano, West Africa: Constraints from 238U–...

Documents