Do decades of tropical rainfall affect the chemical compositions of

Seediscussions,stats,andauthorprofilesforthispublicationat:
https://www.researchgate.net/publication/223456334
Dodecadesoftropicalrainfallaffectthe
chemicalcompositionsofbasalticlava
flowsinMountCameroon?
ArticleinJournalofVolcanologyandGeothermalResearch·March2005
DOI:10.1016/j.jvolgeores.2004.10.008
CITATIONS
READS
16
62
9authors,including:
CatherineChauvel
AlineDia
UniversityJosephFourier-Grenoble1
FrenchNationalCentreforScientificR…
188PUBLICATIONS5,783CITATIONS
133PUBLICATIONS2,671CITATIONS
SEEPROFILE
Allin-textreferencesunderlinedinbluearelinkedtopublicationsonResearchGate,
lettingyouaccessandreadthemimmediately.
SEEPROFILE
Availablefrom:CatherineChauvel
Retrievedon:17September2016
Journal of Volcanology and Geothermal Research 141 (2005) 195 – 223
www.elsevier.com/locate/jvolgeores
Do decades of tropical rainfall affect the chemical
compositions of basaltic lava flows in Mount Cameroon?
C. Chauvela,*, A.N. Diaa, M. Bulourdea, F. Chabauxb, S. Durandb,
P. Ildefonsec,F, M. Gerardd, B. Deruellee, I. Ngounounof
a
LGCA, Observatoire de Grenoble, 1381 rue de la Piscine, 38041 Grenoble Cedex 09, France
b
CGS, 1 rue Blessig, 67084 Strasbourg Cedex, France
c
LMCP, Université de Paris 6, 4 place Jussieu, 75052 Paris Cedex 05, France
d
LFS, IRD, Bondy, France
e
LGIS, Université de Paris 6, 4 place Jussieu, 75052 Paris Cedex 05, France
f
Université de Ngaoundéré, Ngaoundéré, Cameroun
Received 12 February 2004; accepted 25 October 2004
Abstract
To evaluate the effects of tropical rainfall on the compositions of basaltic lavas, we studied 20th century lava flows from
Mount Cameroon in Africa. Weathering conditions are extreme because the climate is particularly warm and humid, and
vegetation grows extremely quickly on the flows. The high rainfalls and dense vegetation contribute to rapid and intense
degradation of the volcanic rocks and should cause significant changes in chemical composition. Such effects need to be
quantified to constrain how young a lava flow must be so that its trace element and isotopic composition remains representative
of the original magma.
Fresh inner parts and altered flowtops of four different lava flows were sampled and analysed for major and trace elements as
well as O, U, Sr, Nd and Pb isotopic compositions. Four samples of the 1999 eruption were also analysed to constrain the
composition of fresh basalts.
Almost all major and trace elements display similar concentrations in inner and outer parts of the same flow. This is notably
the case for elements such as K, Rb and Sr, which are highly mobile during weathering. The lack of variation suggests that the
overall composition of the lava flows has not been significantly affected. However, some systematic chemical changes are
observed: Loss-on-ignition (LOI) and d 18O increase slightly from inner parts of flows to near surface samples; Na and, to a
lesser extent, U display significant losses in the outer samples. We interpret the Na loss in terms of hydration leading to
exchange between Na+ and H+ ions. This process, associated with oxidation of Fe2+ to Fe3+, accounts for the larger loss-onignition in the outer parts of flows. A change in U contents is only observed in the 1922 flow, which is covered by dense
vegetation. This emphasizes the role that complexation by organic ligands plays in U mobility. While U is not completely
immobile, all volcanic rocks are in secular equilibrium ((234U/238U)c1), indicating limited interaction between meteoric waters
* Corresponding author. Tel.: +33 4 76 63 59 12; fax: +33 4 76 51 40 58.
E-mail address: [email protected] (C. Chauvel).
F
Deceased.
0377-0273/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jvolgeores.2004.10.008
196
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
and basalts. Sr and Nd isotopic compositions remain constant and although 206Pb/204Pb ratios vary from inner to outer parts of
the lavas, the changes are not systematic and cannot be attributed to weathering.
The absence of significant chemical mobility and substantial isotopic exchange suggests very limited interaction between
water and lavas at the centimeter scale, due, most probably, to their relatively young age. This work shows that over a period of
about 100 years, no detectable geochemical changes are observed. These results are very promising for petrological and
geochemical studies of ocean island basalts located in tropical areas since they provide direct evidence of very limited trace
element mobility at the century time scale.
D 2004 Elsevier B.V. All rights reserved.
Keywords: Mount Cameroon; basalt; major and trace element data; isotopic compositions; chemical weathering; water/rock interaction
1. Introduction
Geochemical studies of basalts provide essential
information about mantle compositions. Their isotopic
signature is considered to represent that of their source
and their trace element characteristics are used to
constrain both source compositions and conditions
during partial melting. However, the effects of
alteration on the chemical composition of lava flows
have to be evaluated before any petrogenetic interpretation can be made. This is particularly true for
glassy samples which are known to be very sensitive
to alteration. Studies of the effects of alteration on
volcanic glasses, both in field and laboratory conditions (Cerling et al., 1985; Crovisier et al., 1990;
Daux, 1992; Daux et al., 1994, 1997; Le Gal et al.,
1999; Techer et al., 2001), have shown that: (1) an
alteration layer forms rapidly by precipitation of
secondary phases from solution after breakdown of
the glassy network; (2) the formation of secondary
phases controls the permeability of the altered sample;
and (3) elements such as the rare earths (REE) and Th
are readily mobilized during the weathering of basaltic
glass (Daux, 1992).
Element mobility is also well documented during
weathering of holocrystalline basaltic rocks (Price et
al., 1991; Nesbitt and Wilson, 1992; Prudêncio et al.,
1993; Cotten et al., 1995; Gı́slason et al., 1996;
Louvat, 1997; Aiuppa et al., 2000). In these cases,
studies were performed on old and altered samples
containing abundant secondary minerals. Nesbitt and
Wilson (1992), for example, showed through a
detailed investigation of the Baynton basalt profile
in Australia that most elements do not maintain their
original magmatic concentrations during weathering.
Elements are strongly fractionated from each other
because alkali and alkali-earth elements are preferentially leached out of the weathering profile while
less mobile elements such as Fe, Ti and Al remain.
Similarly, Feigenson et al. (1983) demonstrated
preferential mobility of K, Cs and Rb, which led
to abnormal alkali abundances and ratios in basalts
located on the wet northeastern side of Hawaii.
Other authors (Price et al., 1991; Prudêncio et al.,
1993; Cotten et al., 1995) showed that the REE,
which are very often used to identify mantle sources,
are soluble and fractionated by chemical weathering
in certain climatic conditions. According to Price et
al. (1991) and Cotten et al. (1995), this REE
mobility is independent of other chemical criteria
that are used to discriminate between fresh and
weathered samples (i.e., unusual alkali-element
ratios, Sr isotopic mobility or increase in loss-onignition [LOI]).
The time scale during which chemical changes
associated with weathering occur is, of course, a very
important factor. Most analyzed samples were old
(usually thousands or many millions of years, e.g.,
Price et al. (1991) and Nesbitt and Wilson (1992))
and few papers have been published on the alteration
of historic volcanic rocks. One exception is the work
of Dahlgren et al. (1999) who demonstrated that a
period of 4 years was sufficient to leach Si, Ca, Mg,
K and Na out of tephras from Mount St. Helens.
However, to our knowledge, no systematic study of
trace element mobility and isotopic exchange has
been conducted on recent lava flows, and the rate at
which trace elements are removed from basalts
shortly after lava emplacement is unknown. Constraining these parameters is particularly important to
quantify the contribution of alteration of basaltic
rocks to the regulation of atmospheric CO2 through
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
197
geological time (Dessert et al., 2001). It is also of
prime importance for studies centered on magmatic
processes because isotopic compositions and trace
element contents play a key role in all source and
melting models (Hofmann, 1997).
We therefore investigated the first stages of
alteration of 20th century lava flows. Basaltic rocks
from Mount Cameroon were selected because their
eruption ages are accurately known and because of the
extreme climatic conditions that prevail on the
volcano. In addition, study of recent soil profiles
developed on the flank of the volcano show that large
and systematic element mobilities occur during soil
formation (Bulourde, 2001). The hot and very wet
climate promotes weathering and leads to rapid
changes in chemical compositions. In our study, we
analysed major and trace element contents as well as
O, U, Sr, Nd and Pb isotopic compositions in samples
from different parts of several basaltic lava flows and
from a series of basaltic to picritic tephras. The lavas
are less than 100 years old, while soils developed on
tephras are less than 800 years old.
temperature drop is associated with a decrease in
rainfall. Extremely high rainfalls are recorded on the
southwest flank of Mount Cameroon where precipitation locally reaches 12 m per year. Lower rainfalls
occur on the opposite flank (e.g., 1.8 m/year at Ekona,
Fig. 1) which is partially sheltered from the oceanic
influence.
The climatic conditions promote the growth of
abundant vegetation on Mount Cameroon. The
vegetation cover depends both on the age of the
volcanic substrate and the altitude. The most recent
lava flows are colonized by pioneer species such as
mosses and lichens. On Mount Cameroon, most
basaltic terranes are covered by grasses, shrubs or
trees depending on the altitude, and rain forest is
present to altitudes of about 2000 m. Benedetti et al.
(1994) and Moulton and Berner (1998) showed that
this type of vegetation has a drastic effect on the
alteration rate of basalts; they demonstrated that
fluxes of dissolved material transported by rivers are
two to five times higher in vegetated areas than in
bare areas.
2. Geologic setting and climatic conditions
3. Samples
Mount Cameroon belongs to the Cameroon Line,
an intraplate volcanic alignment extending from the
Gulf of Guinea into the African continent. This chain
has been active for at least 65 Ma (Lee et al., 1994)
and includes 12 major volcanic centers that straddle
the African continent–ocean margin (Fig. 1). Mount
Cameroon is the only currently active volcano of the
Cameroon Line: Its volcanic activity started 11 Ma
ago and still persists. The most recent eruption was in
May 2000 and seven other eruptions have been
reported since the beginning of the 20th century
(Déruelle et al., 1987). Each eruption led to the
deposition of tephra and massive basaltic lava flows
on the flanks of the volcano.
Mount Cameroon is located in a humid tropical
area characterized by extreme rainfall and elevated
temperatures all year long. However, climatic conditions are not uniform on the volcano: The combination of high relief (4095 m) and proximity of the sea
leads to strong local climatic contrasts (Fig. 1). Mean
annual temperatures decrease from 26 – 29 8C at sea
level to 0 8C at the top of the mountain and this large
Of the eight eruptions that took place during the
20th century, we selected four largely massive lavas
dated at 1909, 1922, 1959 and 1982. Unweathered
tephra and lavas from the 1999 eruption were
collected soon after their cooling in order to characterize fresh basalt chemistry. Additional samples were
also collected along a soil profile formed through the
degradation of tephritic units deposited less than 800
years ago in order to quantify element mobility during
intense alteration.
The 1909 lava flow, whose scoriaceous upper part
is covered by moss, was sampled high on the north
flank of Mount Cameroon (Fig. 1) at an altitude of
2300 m. The 1922 lava flow is located on the
western flank of the volcano at sea level near
Bibundi, the site with the highest rainfall (Fig. 1).
This flow is covered by shrubs and herbaceous
vegetation and plant roots have penetrated the
scoriaceous and brecciated upper part of the flow.
The 1959 lava flow was sampled near Ekona on the
northeast flank of the volcano at an altitude of 485
m. Here, the vegetation growing on the flow consists
198
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Fig. 1. Map of Mount Cameroon showing sample location, altitude and rainfall. Inset: Maps show the location of the Cameroon Line in West
Africa (modified from Lee et al. (1994)).
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
mainly of shrubs, ferns, orchids, mosses and lichens.
The 1982 samples were collected at mid-slope (2400
m), on the southwest flank of the massif, where
lichens and ferns already colonize the top part of the
flow. Fresh basaltic lava from the 1999 eruption was
sampled in two different locations, the first at an
altitude of 2700 m near the 1982 crater and the
199
second near Bakingili at almost sea level. The 800year-old tephritic unit degraded into soil was
sampled at an altitude of 2540 m on the northern
flank of the volcano. Here, only herbaceous vegetation covers the outcrop.
Most of the basaltic lava erupted as tabular sheet
flows with extremely irregular and scoriaceous upper
margins (Fig. 2a and b). The proportion of vesicules
varies from 40% or more in the flow tops to 5–10% in
the massive flow interiors. The lavas are porphyritic;
they contain predominant augite phenocrysts accompanied by olivine and Ti-magnetite in a glassy matrix of
plagioclase microlites, Fe–Ti oxides and small grains
of augite.
Very few signs of alteration were observed in
both the flow interiors and upper scoriaceous
margins. The exception is the 1959 flow, in which
the upper 3 cm have a greenish colour that contrasts
with the dark grey of the interior of the flow. On the
basis of their appearance in thin section, glass,
phenocrysts and microcrysts seem to be fresh and,
in particular, there is no iddingsite or bowlingite in
the olivine cracks.
For each flow, we separated the fresh inner part
and the potentially altered outermost part. Samples
of the outermost parts of the flows are representative
of the uppermost five centimeters of crust of the
lava flow. In the case of the 1959 surficial sample,
three 15-mm-thick slices were made in order to
examine in detail potential chemical mobility
between surface and inner parts of the lava flow.
In particular, the grey-greenish uppermost part of the
lava flow was separated from the dark grey underlying part.
The soil sequence that formed through the alteration of tephritic deposits is characterized by two
layers, a 20-cm-thick horizon rich in organic components and invaded by numerous roots, and a 20-cmthick brownish andepts that contains millimeter-sized
lapilli and ash.
4. Analytical methods
Fig. 2. (a) Picture of the 1922 lava flow in Mount Cameroon
showing its structure and the plants growing on its surface; (b)
scoriaceous texture of the 1959 lava flow.
Before analysis, organic material such as roots and
mosses was removed from the basalt surface by handpicking. All rock samples were finely powdered in an
agate grinder.
200
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
4.1. Major elements
Major elements were measured on fused discs by
X-ray fluorescence (XRF) at Géosciences Rennes.
Analytical errors are less than 2% for SiO2, Al2O3,
Fe2O3, MgO, CaO, Na2O and TiO2 and 5–10% for
MnO, K2O and P2O5. Loss-on-ignition was calculated
after calcination at 1000 8C of 1 g of powder dried at
110 8C.
4.2. Trace elements
Trace elements were measured by ICP-MS (inductively coupled plasma-mass spectrometry) both at the
CGS (Centre de Géochimie de la Surface) in
Strasbourg and at the CRPG (Centre de Recherches
Pétrographiques et Géochimiques) in Nancy. Estimated analytical errors for both laboratories are 5–
10% for all trace elements. Some trace elements were
also determined on pressed-powder pellets by XRF at
Géosciences Rennes. This is the case for Nb, Zr, Y, Sr,
Rb, Co, V, Ni, Cr, Ba, Zn and Cu for which quoted
precisions are better than 10%. Four samples were
also analyzed by INAA (instrumental neutron activation analysis) in Saclay by J.-L. Joron and the
following trace elements were determined: U, Th,
Zr, Hf, Mo, Br, Ta, Ba, Sr, Cs, Rb, Cr, Co, Ni, Sc, La,
Ce, Sm, Eu and Yb. All the INAA analyses were
duplicated; data reproducibility is better than 7% for
Zr and 5% for the other elements. A comparison of
these four sets of data is given in Appendix A.
4.3. Sr, Nd and Pb isotope data
For each sample, 300 mg of unpowdered rock chips
was dissolved in a mixture of HNO3 and HF. Chips
were used to minimize potential lead contamination
from crushing and grinding. Conventional chromatographic methods described by Richard et al. (1976)
and by Manhès et al. (1984) were used to separate Pb,
Nd and Sr. Total blanks for Pb and Sr are 170 and 150
pg, respectively. For Nd, they vary between 20 and 260
pg. These blanks are negligible compared to the
amounts of Sr, Pb and Nd in the rocks.
Isotopic analyses were obtained at Géosciences
Rennes using a FinniganR MAT 262 multicollector
mass spectrometer. Analyses of the NBS 981 lead
standard gave the following mean isotopic composi-
tions: 2 06 Pb/ 2 04 Pb=16.90, 2 0 7 Pb/ 2 04 Pb=15.44,
208
Pb/204Pb=36.55. Comparison with the certified
isotopic compositions [206Pb/204Pb=16.9373, 207Pb/
204
Pb=15.4925, 208Pb/204Pb=36.7054 (Todt et al.,
1996)] gives a fractionation factor of 1.1x per
amu. For each sample, measurements were done in
duplicate; the reported values correspond to the
average of the two values after correction for
fractionation.
A 87Sr/86Sr of 0.710265F6 (2r, n=14) was
measured for the NBS 987. Since high-precision Sr
isotopic data were necessary to detect small withinflow variations, special measuring procedures were
developed for the project. Samples were loaded on
double Re-filaments and analyzed five times in static
mode. The reported Sr isotope ratios represent the
averages of the five measurements. Using this
technique, an external precision of F9106 (2r,
n=5) is obtained, based on sample and standard
reproducibility. Nd isotopic measurements were performed in peak-jumping mode and a 143Nd/144Nd of
0.511963F5 (2r, n=5) was obtained for the Rennes
Nd standard (this corresponds to a value of 0.511860
for the La Jolla standard).
4.4. Oxygen isotope data
Oxygen was extracted from rock samples through
fluorination with BrF5 (Clayton and Mayeda, 1963)
and reduced to CO2 for isotope analyses. Measurements were carried out at Géosciences Rennes with a
VGR SIRA 10 triple collector instrument. d 18O
values are expressed relative to the standard VSMOW. They are normalized to the in-house basaltic
glass standard Circé 93 whose d 18O value=5.68x; the
corresponding value for the NBS 28 standard is 9.6x.
For any given sample, two duplicated extractions
were done and the maximum deviation recorded is 0.1
d unit.
4.5. Uranium isotopic measurements
For the U isotopic measurements, 200 mg of
powdered rock was digested with HF, then slowly
evaporated and redissolved in a mixture of boric acid
and 7.5 N HNO3. U was separated and purified using
anion exchange resins (Chabaux et al., 1995). 234U
and 235U were analyzed on a VGR Sector TIMS
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
using a Daly detector in analogue mode. (234U/238U)
activity ratios are calculated from measured 234U/235U
ratios assuming 238U/235U=137.88 and using the
decay constants listed in Riotte and Chabaux
(1999). The analytical error for the (234U/238U)
activity ratio is about 0.5%. The reproducibility and
reliability of U isotopic analyses were tested by
regular measurement of the HUI U standard, supposed to be in secular equilibrium. A mean activity
ratio of 1.003F0.002 (2r error) is obtained, instead of
the theoretical value of 1 (Riotte and Chabaux, 1999).
Results were corrected for this deviation. Duplicate
analyses were obtained for several samples (see Table
3). Differences always remain within analytical
uncertainties.
5. Results
Major element concentrations in massive lava
flows are given in Table 1. Since trace element
contents were measured by multiple techniques, a
detailed comparison of the data obtained by INAA,
XRF and ICP-MS is given in Appendix A and in
Table A1, and only trace element data selected on the
basis of accuracy are given in Table 2. Isotopic data
are reported in Table 3.
the more incompatible trace elements (Cs, Rb, Ba, Th,
U and Nb) are less enriched (Chauvel et al., 1992).
The inner parts of flows have homogeneous Sr, Nd
and Pb isotopic compositions. 87Sr/86Sr ranges from
0.70331 to 0.70335, 143Nd/144Nd from 0.51278 to
0.51280 and 206Pb/204Pb from 20.28 to 20.39. These
values are similar to values reported previously for the
Cameroon Line (Halliday et al., 1988; Ballentine et al.,
1997) and approach those of HIMU-type basalts
(Palacz and Saunders, 1986; Chauvel et al., 1992). As
reported earlier by Halliday et al. (1988) for Mount
Cameroon basalts, the measured whole rock d 18O
values for the inner parts of the flows are slightly lower
than the usually accepted mantle value [5.1–5.2x, cf.
5.7x in fresh MORB (Harmon and Hoefs, 1995)].
(234U/238U) activity ratios vary between 0.999 and
1.003, indicating that inner parts of flows are, within
analytical uncertainty, in secular equilibrium.
5.2. Comparison between outer and inner parts of
flows
In general, element concentrations vary little
between outer and inner parts of the 1909, 1922 and
1959 flows (Fig. 4). There are, however, several
exceptions:
–
5.1. Inner parts of flows
The inner parts of flows have very similar chemical
characteristics. As expected for alkali basalts, Si
contents are low (average of 45%) and alkali element
contents are variable and high (between 3.5% and
6.3%). Loss-on-ignition values range from 0.51% to
0.31%. Mg numbers vary between 49 and 63 and
correlate with variations in major and trace element
contents. For example, a decrease in MgO is generally
accompanied by a pronounced increase in Al2O3,
Na2O, CaO and incompatible trace elements (i.e., Rb,
Ba and Nb).
Primitive-mantle-normalized element patterns are
also typical of alkaline magmas (Fig. 3). They show
progressive enrichment from Lu to Ta, relative
depletion in the most incompatible elements (Cs to
U) and marked negative K and Pb anomalies. These
characteristics resemble those of HIMU-type alkali
basalts, where enrichment is maximum for Nb–Ta, and
201
–
–
–
–
The outer parts of the 1959 lava flow are slightly
enriched in Mg, Ni and Cu relative to the inner
part. A significantly higher Ni content characterizes the surface sample from the 1909 flow and Cu
is enriched in the surface sample from the 1922
flow.
Cs and Rb contents of the 1959 surface samples
are greater than those of the inner sample, with the
exception of the uppermost part of the flow.
A strong enrichment in Pb is observed in the 1909
surface sample compared to the inner part of the
same flow.
Na contents are markedly lower in the 1909 and
1959 surface samples.
Loss-on-ignition increases systematically from
inner to outer parts of flows. The largest differences (0.86% and 0.59%) are found for the 1909
and the 1959 lava flows.
The differences in concentrations between outer
and inner parts of the 1982 flow contrast sharply with
202
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Table 1
Major element concentrations expressed in percent
Sample 1909 Lava flow 1922 Lava flow 1959 Lava flow
Inner Outer
Inner
C10F CA9C09 C8N
SiO2
46.61 46.08
(%)
TiO2
3.21 3.24
Al2O3 16.45 16.31
Fe2O3 11.49 11.51
MnO
0.2
0.2
MgO
4.76 4.84
CaO
9.7
9.37
Na2O
4.5
4.14
K2O
1.83 1.82
P2O5
0.85 0.83
LOI
0.42 0.44
Total
99.18 98.78
Mg#
49.1 49.5
43.22
Outer
Inner Outer
CA8N2 C8C
42.93
3.18
3.33
12.11 12.47
13.89 13.91
0.21
0.21
10.28
9.85
11.73 11.71
3.15
3.08
1.44
1.44
0.75
0.73
0.51 0.2
99.45 99.46
63.3
62.3
1982 Lava flow 1999 Eruption
Inner
Outer
Massive Tephra
lava
Massive Tephra
lava
CA8C1B CA8C1S3 CA8C1S2 C10W CA9C82 MC9902 MC9902P MC9903 MC9903P
46.03 45.69
46.06
45.75
44.13 44.19
3.25
16.00
11.86
0.2
5.2
10.04
4.43
1.85
0.83
0.31
99.38
50.5
3.24
15.94
11.88
0.2
5.39
9.96
4.24
1.8
0.82
0.18
99.71
51.4
3.24
15.59
12.01
0.2
5.72
9.92
4.08
1.76
0.8
0.28
99.35
52.6
3.48
15.2
13.57
0.2
5.77
11.32
3.6
1.4
0.61
0.5
98.78
49.8
3.26
15.93
11.89
0.21
5.27
9.92
4.25
1.79
0.81
0.06
98.96
50.8
3.42
14.86
13.73
0.2
6.07
11.86
3.34
1.26
0.53
0.4
99.06
50.7
47.02
46.97
45.96
45.92
2.98
17.26
11.08
0.19
4.79
9.2
4.98
2.04
0.75
0.28
100.01
50.2
2.99
17.32
11.14
0.19
4.71
9.27
4.96
2.03
0.77
0.36
99.99
49.6
3.09
16.05
12.1
0.2
6.21
10.28
4.24
1.67
0.63
0.37
100.06
54.5
3.13
15.63
12.29
0.2
6.51
11.03
4.07
1.58
0.59
0.74
100.21
55.2
For the 1959 lava flow, outer samples are arranged according to their distance to the surface, the most surficial sample being CA8C1S2. LOI
corresponds to the loss-on-ignition at 1000 8C. Mg# represents the Mg number (Mg # = MgO/(MgO+FeO)).
those of the other flows. Trace and major element
contents decrease systematically with decreasing
elemental compatibility. This raises questions about
the magmatic link between the two 1982 samples.
These samples were not collected in a single vertical
section and although they are related to the same
eruptive event, they might have recorded the chemical
composition of the lava at two different stages of the
1982 eruption. Therefore, their trace and major
element composition cannot be used to discuss
short-term weathering effects.
Sr and Nd isotopic compositions of the outer and
inner parts of the flows vary slightly. However, the
differences are within analytical error (Fig. 5). Similarly, the U activity ratios show no variation larger than
analytical uncertainties (Fig. 6b). Although the Pb
isotopic compositions display substantial within-flow
variations, these variations define no systematic trends
(Fig. 7). Whole rock d 18O values of upper parts of the
1959 and 1909 lava flows are 0.2 and 0.4 d units higher
than those of the corresponding inner parts (Fig. 6c).
5.3. Soils on tephrite
In contrast to the subtle differences observed in the
basaltic lava flows, the chemical compositions of the
tephritic units altered into soils have changed dramat-
ically. To monitor these differences, we compared the
compositions of the soils with those of fresh lavas and
tephrites. Preserved fragments of tephrites in the soils
contain relicts of abundant olivine and pyroxene
phenocrysts indicating that the original magmas had
picritic compositions. None of the massive lava flows
analyzed in this study has a picritic composition, but
such a composition has been reported by Déruelle et
al. (1987) for several other lava flows. We therefore
chose the composition of a picrite published by these
authors to evaluate element mobilities due to the
alteration and soil genesis.
In Fig. 8, we show the differences, in percent,
between the compositions of two 800 year old soils
produced through the degradation of a tephrite and the
composition of the reference picrite. In the lower, less
altered, sample, the abundances of most of the major
elements have not changed much, but the alkalis and
most trace elements are moderately to extremely
depleted. In the more altered upper sample, the major
elements are depleted by 10–50%, the alkalis are
severely depleted while the trace elements are not
much more affected than in the lower sample. It
appears that under the climatic conditions of Mount
Cameroon, the compositions of tephritic mafic volcanic rocks change rapidly and dramatically. The
difference in behaviour of the lava flows and the
Table 2
Trace element concentrations (expressed in ppm) measured by ICP-MS (see Appendix A for a detailed explanation)
Sample
1909 Lava flow
Inner
C10F*
Cs (ppm)
0.36
0.34
Rb
35.9
33.4
Ba
517
536
Sr
1065
1082
Nb
100.4
96.7
Ta
6.26
6.14
Hf
8.04
7.85
Zr
393
390
Pb
3.44
Pb #
2.27
U
2.27
2.23
Th
8.43
8.10
Cu
59.3
Co
30.4
30.0
Ni
30.9
Y
35.8
35.3
La
77.5
80.8
Ce
153
165
Pr
19.0
19.6
Nd
73.6
76.0
Sm
12.9
13.2
Eu
3.97
4.05
Gd
10.19
10.33
Tb
1.42
1.45
Dy
7.31
7.58
Ho
1.33
1.38
Er
3.17
3.26
Tm
0.47
0.48
Yb
2.53
2.61
Lu
0.36
0.37
1922 Lava flow
1959 Lava flow
1982 Lava flow
1999 Eruption
Inner
Inner
Inner
Massive Tephra
lava
% Difference CA9C09 C8N
5.6
7.0
3.7
1.6
3.7
1.9
2.4
0.8
1.8
3.9
1.3
1.4
4.3
7.8
3.2
3.3
2.3
2.0
1.4
2.1
3.7
3.8
2.8
2.1
3.2
2.8
Outer
CA8N2 C8C
Outer
Outer
Massive Tephra
lava
CA8C1B CA8C1S3 CA8C1S2 C10W CA9C82 MC9902 MC9902P MC9903 MC9903P
0.35
0.45
0.41
0.32
0.42
34.3
32.9
32.0
33.6
36.2
522
440
443
530
519
1078
1004
1017
1071
1081
97.2
89.5
88.4
100.4
98.8
6.15
5.54
5.51
6.31
6.18
7.90
7.45
7.26
7.94
7.81
400
351
345
385
381
4.62
4.04
3.90
4.32
4.32
5.4
2.22
2.12
1.94
2.27
2.27
8.33
7.59
7.33
8.68
8.62
58.9
65.1
74.7
65.7
71.5
31.8
51.8
50.8
32.6
33.0
37.8
177
166
46.9
45.9
36.6
31.6
31.2
35.0
34.8
78.0
73.6
74.0
81.8
80.5
154
146
147
165
161
19.0
17.9
18.1
19.4
19.2
73.4
69.1
69.4
74.7
74.0
12.8
11.8
11.9
12.8
12.6
3.92
3.54
3.56
3.93
3.87
10.14
9.29
9.39
10.12
10.02
1.4
1.25
1.25
1.4
1.38
7.24
6.29
6.30
7.24
7.18
1.31
1.14
1.14
1.32
1.30
3.17
2.69
2.69
3.14
3.11
0.46
0.39
0.39
0.46
0.45
2.47
2.11
2.12
2.49
2.49
0.35
0.30
0.30
0.35
0.35
0.44
36.5
509
1064
97.0
5.97
7.59
380
4.06
0.31
33.0
510
1041
97.5
6.11
7.78
385
4.49
0.44
0.40
32.9
30.3
433
400
995
927
81.9
75.8
5.18
5.06
7.36
7.12
325
307
3.32
3.13
2.16
8.26
94.9
33.8
49.6
34.8
78.9
157
18.8
72.3
12.4
3.82
9.85
1.35
6.99
1.28
3.05
0.45
2.43
0.34
2.22
8.4
68.0
35.9
70.4
35.1
79.4
159
19.0
72.9
12.5
3.84
10.00
1.37
7.09
1.30
3.09
0.45
2.44
0.35
1.80
1.63
7.18
6.42
105
105
43.2
45.3
62.7
66.6
32.2
31.4
66.2
60.8
137
127
16.1
15.0
62.8
59.2
11.2
10.8
3.37
3.27
8.82
8.48
1.28
1.24
6.68
6.59
1.22
1.19
2.89
2.80
0.42
0.41
2.32
2.26
0.32
0.32
0.57
0.59
49.1
49.2
595
605
1193
1216
113
115
7.48
7.59
9.06
9.24
433
441
4.45
4.64
2.67
10.0
57.6
30.1
30.2
36.2
88.6
181
20.2
75.9
13.0
3.91
10.4
1.34
7.46
1.31
3.48
0.44
2.65
0.38
2.72
10.3
57.6
29.8
28.0
36.5
89.9
184
20.6
77.1
13.3
4.00
10.6
1.37
7.55
1.33
3.56
0.45
2.72
0.39
0.46
0.44
39.1
37.9
486
469
1052
1035
91.5
88.8
5.99
5.85
7.97
8.03
365
360
3.56
3.61
2.14
8.14
69.7
37.5
66.9
31.8
72.2
149
16.9
64.4
11.4
3.47
9.22
1.20
6.66
1.17
3.08
0.39
2.34
0.34
2.09
7.91
79.7
40.0
75.0
32.1
70.5
146
16.6
63.7
11.4
3.51
9.35
1.21
6.71
1.17
3.11
0.39
2.35
0.34
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
C10F
Outer
For the 1959 lava flow, outer samples are arranged from right to left according to their distance to the surface, the most surficial sample being CA8C1S2. Relative variations between
C10F and the duplicate C10F* are indicated in italics and expressed in %. The # symbol corresponds to lead contents determined by isotope dilution.
203
204
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Table 3
Sr, Nd, Pb, U and O isotope data
143
206
207
208
234
d 18O
0.703310F9
0.703309F5
0.512804F4
0.512789F5
20.394
20.394
15.645
15.655
40.124
40.152
1.002F0.005
1.002F0.005
5.11F0.05
5.48F0.05
1922 Flow
C8N
CA8N2
CA8N2*
0.703353F5
0.703360F8
0.512776F6
0.512769F6
20.281
20.384
15.637
15.643
40.028
40.133
1.003F0.005
1.006F0.005
1.001F0.005
1959 Flow
C8C
CA8C1B
CA8C1S3
CA8C1S2
CA8C1S2*
0.703316F5
0.703326F5
0.703323F5
0.703323F3
0.512790F6
0.512792F5
0.512779F4
0.512783F5
20.313
20.389
20.364
20.082
20.103
15.640
15.640
15.657
15.640
15.625
40.043
40.109
40.132
39.839
39.818
0.999F0.005
5.20F0.02
1.005F0.005
1.004F0.005
5.41F0.01
1982 flow
C10W
C10W*
CA9C82
CA9C82*
0.703321F5
0.703331F5
0.703312F6
0.512784F5
0.512790F6
0.512775F6
20.342
20.376
20.286
15.642
15.653
15.631
40.099
40.159
40.03
Sample
1909 Flow
C10F
CA9C09
87
Sr/86Sr
Nd/144Nd
Pb/204Pb
Pb/204Pb
Pb/204Pb
U/238U
0.999F0.005
1.003F0.005
1.001F0.005
Asterisks correspond to duplicates. d 18O data are expressed in x with respect to V-SMOW.
tephrites is explained by the differences in rock
composition, in texture and by the moderate difference in age.
6. Discussion
Evaluation of the extent and speed of alteration in
volcanic rocks is of prime importance both for
petrological and weathering studies. To understand
the nature of the source of volcanic lavas requires that
the isotopic and chemical composition of samples is
representative of the erupted magmas and has not
been affected by alteration. Similarly, to evaluate the
intensity of chemical weathering, the original composition of lavas needs to be well constrained. To
address both issues, we concentrated this study on the
comparison of outer and inner parts of very recent
lavas flows.
Experimental studies on synthetic and natural
glasses (Oelkers and Gislason, 2001; Gislason and
Oelkers, 2003) and studies of natural volcanic
glasses have shown that an alteration front grows
at a relatively high rate in the first years of exposure
(up to 20 Am/100 years) but that in older flows the
apparent rates are always lower (down to 0.001 Am/
100 years) (see Crovisier et al., 2003 for a review of
available studies). In the case of our Mount
Cameroon lavas, which are less than 100 years old,
the highest rate can be used. Nonetheless, we
calculate that alteration would affect only the outmost 2–20 Am of the basalt. In a highly vesicular
and fractured rock, however, the outermost 20 Am
may constitute a significant proportion of the total
volume. For example, in a lava containing 40%
vesicles, the total volume of 10-Am-thick alteration
zones on the walls of 1-mm-diameter vesicles
constitutes about 1.2% of the total volume of basalt.
If the altered zones contain, for example, 10 times
the Cs of fresh basalt, the presence of the alteration
zones will increase the Cs content of the lava by
about 18%.
In addition, to change the composition of a
vesicular and fractured rock does not require the
replacement of fresh minerals or glass by secondary
phases. The precipitation of secondary phases as thin
films on the inner walls of vesicles or along fractures
also influences the bulk composition of the sample. In
the same lava with 40% vesicles, 10-Am-thick films
on the inner walls of the vesicles would also constitute
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
205
Fig. 3. Trace element patterns of the inner parts of flows together with lavas from the 1999 eruption. Primitive mantle normalizing values are
from Hofmann (1988).
only slightly less than 1.2% of the total volume of the
basalt. Again, if these films contained 10 times the Cs
content of the basalt, their presence would also result
in an 18% increase in the Cs content of the sample.
Precipitation along fractures would contribute further
to the change in composition.
206
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Fig. 4. Trace element patterns of the outer parts of flows normalized to the corresponding inner parts. In the case of the 1959 flow, the three
samples representative of the surface layer are normalized to the inner sample.
The extent of alteration therefore depends crucially
on the texture of the lava, particularly its vesicularity
and abundance of fractures, which control the total
surface area of lava susceptible to alteration, and the
abundance and geometry of the fractures, which
control the permeability. Only if the vesicles and the
fractures are interconnected will water be able to
penetrate into the lava flow, to cause alteration of the
magmatic minerals or the precipitation of secondary
phases. The presence of vegetation accelerates the
process, by increasing the acidity of the fluids and,
through the growth of roots, by mechanically disrupting the upper part of the flow (Cochran and Berner,
1996). The potential for alteration is dramatically
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
207
Fig. 5. 143Nd/144Nd vs. 87Sr/86Sr ratios for Mount Cameroon samples. The arrows link samples from the same lava flow. Filled diamonds
represent inner samples and open diamonds outer samples (duplicates are surrounded by a field). The direction of the arrows indicates the
upward paths from the interior towards the surface. Typical external reproducibility is shown with bars. In the inset, new data are compared to
literature data for the various Cameroon Line volcanoes represented by fields (Halliday et al., 1988, 1990; Lee et al., 1994).
illustrated by the poorly consolidated, fragmental and
highly vesicular tephrites, whose compositions were
drastically altered on the hot, wet slopes of Mount
Cameroon in a period of only 800 years (see Fig. 8).
The question then is to establish the influence of such
conditions on the compositions of massive basalt, the
type of lava normally sampled during geochemical
studies.
6.1. How fresh are the inner parts of flows?
First, we need to make sure that samples from
the interior of the flow represent fresh basaltic
lavas. Highly incompatible elements are particularly
useful in this respect because during partial melting
and crystal fractionation, their relative concentrations remain practically unchanged and similar to
those of their mantle source. This is the case of
ratios such as Cs/Rb, Ba/Rb and Nb/U which are
nearly constant in fresh oceanic basalts worldwide
(Hofmann and White, 1983; Hofmann et al., 1986).
These ratios can therefore be used to establish
whether the inner parts of flows have been affected
by alteration. Among the highly incompatible
elements, some are mobile (Cs, Rb, Ba, U) but
others are relatively immobile even in severe
alteration conditions (Th, Nb). Ratios of the two
types of elements (e.g., Th/Ba, Th/Rb and Th/U)
will be most affected by weathering. Moreover,
these changes may take place during the very first
stages of weathering because these highly incompatible elements are concentrated in the glass which
is particularly sensitive to alteration.
Th/Ba, Th/Rb and Th/U ratios in the inner samples
are almost constant at 0.016–0.017, 0.23–0.26 and
3.57–3.85, respectively. Data of Déruelle et al. (1987)
for Mount Cameroon lavas display small variations in
Th/Ba, Th/Rb and Th/U ratios but remain within the
range of our samples (Fig. 9). Both our data and those
of Déruelle et al. (1987) deviate slightly from the
mean Th/Ba and Th/Rb ratios calculated by Fitton and
Dunlop (1985) for the whole Cameroon Line (Fig. 9),
but all the ratios are similar and coincide with values
reported for the island of Pagalu in the oceanic section
208
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Fig. 6. (a) Na2O content, (b) (234U/238U) activity ratio and (c) d 18O vs. loss-on-ignition (LOI) of Mount Cameroon lava flows. Filled circles
represent inner samples and open circles represent outer samples (duplicates are surrounded by a field); arrows link inner and outer samples
from the same flow and indicate the upward path from the center to the surface. d 18O values for MORB (Harmon and Hoefs, 1995) and other Mt
Cameroon lavas (Halliday et al., 1988) are also shown (hatched line and grey squares). Because Halliday et al. (1988) did not report LOI values,
the grey squares only indicate d 18O results. Analytical uncertainties are shown by crosses.
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
209
Fig. 7. Pb isotope compositions for Mount Cameroon lavas (207Pb/204Pb and 208Pb/204Pb vs. 206Pb/204Pb). The arrows link samples from the
same lava flow. Symbols are as in Fig. 5. The direction of the arrows indicates the upward paths from the inner parts towards the surface.
Dashed lines represent the mass fractionation correction applied to all samples. Both insets show new data represented by symbols and literature
data shown by fields for several volcanoes from the Cameroon Line (Halliday et al., 1988, 1990; Lee et al., 1994).
of the Cameroon Line (Lee et al., 1994). The values
are also similar to those of fresh, directly sampled
lavas including the 1999 Mount Cameroon eruption
(Fig. 9) and fresh HIMU basalts: Th/Ba=0.013–0.020,
Th/Rb=0.20–0.29, Th/U=2.65–3.61 (Weaver et al.,
1987; Dupuy et al., 1988; Sun and McDonough,
210
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Fig. 8. Composition of two soils developed on a tephritic layer deposited less than a thousand years ago. The lower part of the soil has almost
unchanged major element concentrations while most trace elements are highly depleted relative to the concentrations reported for an equivalent
picrite. Data for the picrite are from Déruelle et al. (1987) while data for the soils are from Bulourde (2001). The reproducibility of
measurements for both major and trace elements are shown by the grey fields (5% for major elements and 10% for trace elements).
1989). We are therefore confident that the inner parts
of flows represent the initial magmatic compositions.
6.2. Evidence of weathering in recent Mount
Cameroon lava flows
A first argument for element mobility is the
difference in LOI between inner and outer parts of
the flows (Table 1). LOI values are systematically
greater in the outer samples, reflecting slight hydration of the surficial parts of the flows. The three slices
made in the 1959 sample show that LOI increases
gradually towards the surface.
Several processes can increase the LOI during the
alteration of basaltic lavas. For example, Cerling et
al. (1985), Petit et al. (1990) and Grambow and
Müller (2001) have shown that hydration of volcanic
glass is accompanied by substantial exchange
between H+ ions from the solution and alkali ions
(Na+, K+) from the glass. The gain of H+ ions
commonly results in the formation of hydroxyl
groups that contribute to the LOI. Oxidation of Fe
from Fe2+ to Fe3+ also contributes to shifts from
negative to positive LOI. The samples in the interior
of the flows are completely anhydrous and for these,
the mass increase during determination of the LOI
produces negative values. In samples from the
margin of the flows, LOI varies from negative in
the least altered samples to positive in more altered
samples. In the latter samples, the positive LOI is due
to (a) minor hydration and (b) oxidation of FeO to
Fe2O3 during alteration of this part of the flow.
The largest differences in LOI are found between
the inner and the outer parts of the 1909 and the 1959
lava flows (Table 1). These flows also show significant changes in Na concentrations (Fig. 6a). Na2O
contents in the surficial parts of the 1909 and 1959
lava flows are 8% lower than in the corresponding
inner parts. This Na loss combined with the LOI
increase appears comparable to Cerling et al. (1985)
observation and can be interpreted as resulting of
hydration of the glass accompanied by substantial
exchanges between H+ and Na+ ions. However, this
process did not affect the other alkali elements (K, Cs,
Rb) whose concentrations are unchanged implying a
preferential substitution between Na+ and H+ ions as
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
documented by Cerling et al. (1985) in several
examples of glass hydration.
The change in Na contents (expressed in mol/g
rock) between the inner and the outer samples of the
same flow is:
2MNa =MNa2 O
Na2 O
2O
DNa ¼
X Na
outer X inner
MNa
where M is the molecular weight (in g/mol) and X the
Na2O mass fraction. If Na+ ions are replaced by H+
ions, which are lost during heating, the LOI will
increase. Assuming a complete substitution between
Na+ and H+ ions, the change in LOI (expressed in g/g
rock) due to this process is given by:
DLOI ¼ DNa ðMH þ 0:5MO Þ
D LOI values calculated following this procedure
vary between 0.02% for the 1982 lava flow and
0.11% for the 1909 lava flow. The latter value is
significantly smaller than the 0.86% difference of LOI
in the 1909 flow which indicates that the formation of
hydroxyl groups by Na+–H+ substitution cannot
explain the entire LOI variation. Cerling et al.
(1985) demonstrated that up to 75% of the water in
hydrated glass is present as molecular water (H–O–H)
rather than hydroxyl species. The difference between
the calculated and measured LOI variation could
therefore result from a contribution of molecular
water.
Oxygen isotopic compositions obtained on the
1909 and 1959 lava flows show that the LOI increase
is accompanied by a moderate increase in d 18O
towards the surface (Fig. 6c). Post eruptive 18O
enrichment due to low-temperature water/rock interaction has been documented by Martinez and Turi
(1978), Cerling et al. (1985) and Kyser et al. (1986) in
both submarine and subaerial settings. In volcanic
rocks, the matrix contains glass and fine-grained
minerals that are particularly susceptible to oxygen
isotopic exchange (Lawrence and Taylor, 1972).
Kyser et al. (1986) found in glassy boninites from
Cape Vogel, New Caledonia and Cyprus that olivine
and pyroxene phenocrysts retain their pristine oxygen
isotope composition whereas glass is greatly enriched
in 18O with d 18O changes of up to 16x. These high
d 18O values in the glass are interpreted as the result of
interaction between boninite and seawater at low
temperature (b150 8C). Similarly, in sub-aerially
211
exposed tuffs from East Africa, Cerling et al. (1985)
observed that hydration of the glass by meteoric water
causes a significant d 18O increase of over 20x. Thus,
it seems likely that the d 18O shifts in the Mount
Cameroon lava flows result from isotopic exchange
between the phases present in the groundmass (glass
shards, microliths) and meteoric water. Nevertheless,
the effects on d 18O for lavas erupted less than 100
years ago remain extremely limited (+0.4 and +0.2 d
units for the 1909 and 1959 lava flows, respectively).
U is the only element other than Na whose
concentration changes can be attributed to weathering.
While no significant change is seen for the other
flows, variations between inner and outer parts of the
1922 lava flow reveal a U loss of about 10% in the
surface sample (Fig. 4). However, since this difference
is close to the analytical error, the U loss should be
treated with caution. The presence of luxuriant
vegetation on the 1922 lava flow associated with the
high local rainfalls might create unique conditions that
favour U mobility in this lava flow. Similar observations have been reported by Halbach et al. (1980) and
Daux (1992) who attributed the U losses to the
presence of organic matter which increases U solubilization and complexation. U mobility should be
accompanied by changes in U isotopic compositions
since 238U and 234U are usually fractionated during
weathering: since 234U is preferentially leached into
waters, these have (234U/238U) activity ratios greater
than 1 and residual weathered materials have ratios
lower than 1 (Ivanovich and Harmon, 1982; Chabaux
et al., 2003). Our samples have (234U/238U) activity
ratios always close to 1 and no significant difference
between inner and outer parts of flows is observed
(Fig. 6b). In addition, the slight U mobility noticed in
the 1922 lava flow is not accompanied by significant
U isotopic fractionation. (234U/238U) measurements in
Mount Cameroon waters (Chabaux et al., 1998; Riotte
et al., 2003) are consistent with this observation.
Indeed, considering that rainwater contains virtually
no uranium, that fresh lavas are in secular equilibrium
(234U/238U)=1, and that the maximum U loss in the
weathered basalts is about 10%, mass balance
calculations show that a 1% decrease in (234U/238U)
in the altered rock will be balanced by a maximum
increase of 10% in the waters and a corresponding
(234U/238Uwaters)b1.1. Such ratios were obtained for
the high altitude Mount Cameroon water springs that
212
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
essentially drain young basalts (Riotte et al., 2003).
The (234U/238U) data for Mount Cameroon lavas are
therefore consistent with the slight U mobility
observed in the 1922 lava flow.
6.3. No major chemical changes due to weathering at
a century time scale
Most element concentrations do not change
between inner and outer parts of the flows. This is
particularly true for high field strength elements (Zr,
Hf, Nb, Ta, Th) (see Table 2 and Fig. 4). These trace
elements are known for their low solubility in natural
waters (Cramer and Nesbitt, 1983) and are usually
considered as immobile during weathering (Middelburg et al., 1988; Nesbitt and Wilson, 1992;
Venturelli et al., 1997). Less predictable is the
behaviour of rare earth elements (REE) in lowtemperature conditions. Middelburg et al. (1988)
argues that REE are immobile while other authors
report clear mobility in various weathering contexts
(Alderton et al., 1980; Price et al., 1991; Prudêncio et
al., 1993; Daux et al., 1994; Cotten et al., 1995; Guy
et al., 1999). The most conspicuous evidence of REE
mobility is the occurrence of Ce anomalies relative to
La and Pr. Such anomalies develop because Ce does
not behave like other trivalent lanthanides when
present as Ce4+. In weathering profiles, a proportion
of Ce occurs as Ce4+ and enters cerianite (CeO2)
(Braun et al., 1990), the presence of which leads to
positive Ce anomalies (Ce/Ce*N1, with Ce/Ce*=CeN/
(LaNPrN)1/2). Negative Ce anomalies (Ce/Ce*b1)
have also been described when secondary phosphate
minerals such as rhabdophane precipitate from
groundwaters (Cotten et al., 1995). No significant
Ce anomalies are observed in Mount Cameroon recent
lava flows. Ce/Ce* ratios range from 0.96 to 0.98 in
the outer parts of the flows and values are identical for
corresponding inner parts. Inner and outer parts of the
flows also share similar LaN/YbN ratios and we
conclude that no fractionation occurred among REE.
The magmatic REE patterns appear to be preserved
throughout the flows.
K and Sr, which are generally very mobile during
weathering (Middelburg et al., 1988; Nesbitt and
Wilson, 1992), have almost identical concentrations in
the inner and outer parts of flows and show no
evidence of mobility. Furthermore, with the exception
of the 1959 lava flow, within-flow changes in Cs and
Rb contents never exceed the analytical uncertainties.
In the case of the 1959 lava flow, Cs and Rb
enrichments probably result from element mobility
during cooling of the lavas (see Section 6.4) and
cannot be interpreted in terms of weathering. Indeed,
because of their high solubility, alkali elements
generally show severe depletions rather than enrichments within weathered rocks. For example, in altered
Hawaiian basalts, Feigenson et al. (1983) reported
extremely low alkali abundances due to the leaching
of K, Cs and Rb. These elements are leached from the
basalts in unequal proportions leading to higher K/Rb
ratios in the weathered lavas (Feigenson et al., 1983).
As a result, Hawaiian basalts display K/Rb ratios
which vary greatly from 400 to 6000 according to the
weathering intensity. In the Mount Cameroon lavas,
K/Rb ratios are almost constant both at the flow scale
and between flows and remain within a magmatic
range (350bK/Rbb460). This strengthens our conclusion on the absence of alkali mobility (with the
exception of Na) during the very first stages of
alteration and contrasts with the observations reported
for the Hawaiian lavas by Feigenson et al. (1983).
However, these Hawaiian basalts are much older than
one hundred years (F.J. Spera, personal communication) and suffered heavy rainfalls for long periods of
time.
Within lava flows, Sr and Nd isotopic variations do
not exceed analytical uncertainties (Fig. 5). In
particular, the difference in 87Sr/86Sr ratios between
inner and outer parts of flows is always smaller than
external precision (i.e., 2105). However, it cannot
be excluded that minor mobility of Sr occurred,
Fig. 9. Th/Ba, Th/Rb and Th/U vs. Th contents for the inner parts of flows (filled diamonds). Grey circles represent lavas from the 1999
eruption. Other Mount Cameroon samples are reported for comparison (Déruelle et al., 1987) and shown by white squares. The black and white
stars indicate mean compositions of continental and oceanic sectors of the Cameroon Line, respectively (Fitton and Dunlop, 1985). Tubuai
island data (Pacific Ocean (Chauvel et al., 1992)) and Pagalu data (oceanic portion of the Cameroon Line (Lee et al., 1994)) are also shown for
comparison (small black dots for Tubuai and small white dots for Pagalu). Average Th/Ba, Th/Rb and Th/U ratios for St. Helena and HIMUtype basalts are also represented by a line (Weaver et al., 1987; Dupuy et al., 1988; Sun and McDonough, 1989).
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
213
214
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
provided that this mobility led to changes in 87Sr/86Sr
smaller than 2105. Using this maximum shift on
the 87Sr/86Sr ratio, interaction between water and rock
can be evaluated and a maximum water/rock ratio can
be calculated using a slightly modified version of
Langmuir (1978) mixing equations:
87
Sr
86 Sr
altered bas
¼
87
87
Sr
Sr
Xwat Cwat 86
þ Xbas Cbas 86
Sr wat
Srbas
87
87
Sr
Sr
Xwat 86
þ Xbas 86
Sr wat
Sr bas
87
with: 86Sr
Sr bas =average87composition of the inner parts of
flows (0.703325); 86 Sr
Sr altered bas =composition of a theo87
5
retical weathered basalt ( 86 Sr
Sr bas +2 10 =0.703345);
87
Sr
=composition of the meteoric waters (average
86 Sr
wat
value from four Mount Cameroon rainfalls: 0.706762;
Benedetti et al., 2003); X bas=mass fraction of basalt;
X wat=mass fraction of water; C bas=average Sr concentration of the inner parts of flows (1050 ppm);
C wat=average Sr concentration of meteoric waters
(value calculated from four Mount Cameroon rainfalls: 2.05 ppb; Benedetti et al., 2003) and where
Xwat þX bas ¼ 1 and the water/rock ratio= XXwat
.
bas
Using this equation, the maximum possible proportion of water that interacted with the basalt (the
water/rock ratio) is 3000. This value is extremely low
compared to ratios of 121,000 and 587,000 reported
by Innocent et al. (1997) for tropical laterites in
Brazil. Higher values—between 10,000 and 30,000—
were also calculated by Daux (1992) for basaltic
hyaloclastites in Iceland. In the case of Mount
Cameroon, the low ratio clearly indicates that little
water interacted with the basalt, as is to be expected
considering the more massive rock texture and the
extremely limited chemical mobilities.
6.4. Changes that are not related to weathering
In contrast to most elements, Ni, Cu, Rb, Cs, Pb and
to a lesser extent Mg show some significant withinflow variations. Changes in Ni contents are greater than
20% in the 1909 lava flow and 50% in the 1959 lava
flow. In the latter case, the increase in Ni content in the
surface sample is accompanied by a marked increase in
MgO content (+10%) suggesting a magmatic origin to
the fluctuation. Mass balance calculation shows that an
additional c1% of phenocrysts is sufficient to explain
the difference in MgO. Such a small excess of olivine
would not be measurable in thin section but is sufficient
to account for the differences in MgO and Ni contents.
Since all lavas contain olivine phenocrysts, small
changes in their distribution at the centimeter scale
could explain the Ni and Mg variabilities. Similarly, the
observed changes in Cu contents in the 1922 and 1959
flows (Fig. 4) could be explained by uneven distribution of sulphides. Most of the copper contained in
igneous rocks is held by sulphides such as chalcopyrite
(CuFeS2) (Hall, 1987) and small heterogeneities in the
spatial distribution of sulfide grains could account for
the Cu variation. White and Hochella (1992) also
reported Cu enrichments at the surface of Hawaiian
basalt flows. During the cooling process, Cu and
volatile elements such as F, Cl and S can precipitate
from the gas phase, forming sublimate minerals such as
kroehnkite (Na2Cu(SO4)2d 2H2O) that change basalt
surface chemistry (White and Hochella, 1992). The
higher Cu contents of the 1922 outer sample therefore
could result from the presence of sulfate deposits at the
surface of the flow or within the vesicles. The uppermost sample of the 1959 lava flow (CA8C1S2) does
not display a similar Cu enrichment (Fig. 4) but the two
lower samples (CA8C1S3 and CA8C1B), collected
respectively 15 and 30 mm below the surface, show an
increase in Cu content accompanied by large Cs and Rb
enrichments and smaller Na losses. Surficial sublimate
minerals incorporating alkali elements include kroehnkite and thenardite (Na2SO4) (White and Hochella,
1992) and Cs and Rb are likely to substitute to Na in
such compounds. Thus, Cu, Cs and Rb enrichments in
the outer parts of the 1959 lava flow could have their
origin in the condensation of sublimate minerals on the
vesicle walls. The fact that the uppermost sample
CA8C1S2 does not exhibit Cs, Rb and Cu enrichment
might reflect the high sensitivity of the sublimate
minerals to weathering because these minerals are
rapidly leached at the basalt surface (White and
Hochella, 1992). A few centimeters below the surface,
however, they might be preserved in vesicles isolated
from the atmospheric conditions.
Whereas no systematic trend is seen for Pb
concentrations in three flows, the difference between
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
215
6.5. Factors explaining the limited effect of rainfalls
in Mount Cameroon recent lava flows
it was shown that Na, K, Mg and Ca were partially
leached from the lavas during the first year of
weathering. However, tephra are fragmental and
highly permeable materials into which water penetrates. The large surface area of the fragments then
allows efficient water/rock reaction. In the case of
Mount Cameroon lavas, the more massive flow
texture is certainly a key factor inhibiting the weathering process. The flowtops are characterized by the
presence of abundant fractures and vesicles (up to
40% of the volume) leading to a high total porosity.
However, high values of total porosity do not directly
imply high permeability of the material because
numerous independent factors contribute to the final
permeability of the material. The permeability, a
measure of the relative ease with which a medium
can transmit a water by advection under a potential
gradient, depends only on the shape and the size of the
pores and not on the nature and the amount of water,
nor on the size of the potential gradient Recently, Saar
and Manga (1999) studied the relationship between
permeability and porosity in vesicular basalts and
concluded that for a porosity higher than about 10%,
the permeability was nearly constant at values of
about 1012, values that are about 1000 times lower
than could be expected in fragmental volcanic rocks.
The permeability of the top of the Mount Cameroon
lava flows can be expected to be quite similar to the
vesicular lavas studied by Saar and Manga (1999). In
low-permeability material, the chemical diffusivity,
which would also influence the rate of reaction, would
also be low. With relatively low permeability for the
top of the flows, meteoric waters mainly flow along
the surface of lava flows without readily percolating
through the flows. The interaction between lava and
water therefore remains limited and alteration can not
progress quickly. In addition, the steep slopes of
Mount Cameroon tend to decrease water residence
time. Water/rock interaction is consequently very
limited, explaining the lack of significant isotopic
exchanges between water and lavas.
6.5.1. Lava textures
Considering the extreme climatic conditions at
Mount Cameroon, the absence of substantial chemical
mobility is surprising. This contrasts with mobilities
reported by Dahlgren et al. (1999) for the first stages
of weathering in Mount St. Helens tephra. In that case,
6.5.2. Age of lava flows and sample scale
Since lavas are subaerial, they are in direct contact
with rain. Meteoric waters are known to be slightly acid
and extremely poor in most elements. They are
chemically aggressive solutions and it is likely that
dissolution kinetics are particularly high, even more so
the inner and the outer part of the 1909 sample is large.
The first set of concentrations was obtained by ICPMS and the difference between the two parts of the
flow was confirmed by isotopic dilution measurements
(Table 2). Even though the two samples have
extremely different Pb contents (2.3 and 5.4 ppm),
they share the same Pb isotopic composition (Table 3)
suggesting that contamination by foreign Pb cannot
explain the higher concentration in the surface sample.
The additional Pb must be of magmatic origin and we
conclude that the difference in Pb contents must be due
to inhomogeneous distribution of Pb-rich primary
phases within the 1909 flow. Unlike Sr and Nd
isotopic compositions, which are identical in inner
and outer parts of flows (Fig. 5), Pb isotopic ratios vary
significantly within the 1922, 1959 and 1982 lava
flows (Fig. 7). In addition, a very low Pb isotopic
composition characterizes the 1959 surface sample.
However, this low value probably does not result from
anthropic contamination because it is not associated to
a significant increase in lead content. For all four lavas,
Pb isotopic changes cannot be interpreted in terms of
weathering effects because of the lack of systematic
trend between inner and surficial samples and the
within-flow Pb isotopic heterogeneity remains unclear.
To summarize, chemical variations due to weathering are extremely subtle. Only Na and may be U
show significant mobility but it is not systematic
among the studied flows. Moreover, their occurrence
and extent are not related to the age of the flows. Na is
mobile in the 1909 and 1959 flows but not in the 1922
flow and Na loss is not larger in the 1909 surface
sample than in the 1959 surface sample. Similarly, U
seems to be mobile in the 1922 flow but not in the
1909 flow. Local weathering conditions such as the
presence of vegetable cover and surface characteristics of the flows (e.g., the presence of cracks) are
thought to control Na and U mobilities.
216
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
because Mount Cameroon lavas contain easily alterable primary phases such as glass or olivine. However,
we have shown that chemical mobilities were limited
and only affected Na and possibly U contents. This
raises questions about the choice of sample size for the
study. Our results were obtained on few centimeter
thick samples located at the top of the flow for the
bouterQ samples and well inside the flow for the binnerQ
samples. Our study shows that, at that scale, weathering
is extremely limited. If the surface samples had been
thinner, the weathering effects might have been larger.
In the case of the 1959 flow, the size of samples was
determined by the colour of the lava (from dark grey to
grey-greenish at the surface). Even such separation
based on the external aspect of the rock does not let us
highlight marked mobilities. In any case, if weathering
had affected a layer only few millimeters thick in a time
period of 50–100 years, the main conclusion remains
that lava flows are essentially fresh and undisturbed by
alteration processes.
–
–
elements such as Cs, Rb, K and Sr which are
usually considered as mobile during weathering.
Non-systematic within-flow changes in Pb isotopic compositions occur. However, there is no
convincing evidence for these relatively small
variations to be due to either incipient weathering
or contamination by a foreign lead component.
Sr and Nd isotopic compositions show no significant within-flow variation. Despite several decades
of extreme rainfalls, Sr isotopes are not affected by
post-magmatic alteration processes and they are
still representative of the mantle isotopic signature.
This study clearly demonstrates that the effects of
rainfalls are negligible on the chemical and isotopic
compositions of these massive basaltic rocks, even
though extreme tropical conditions affected lava flows
for almost 100 years. The chemical characteristics of
these recent basalts (b100 years old) remain unchanged
and still represent the original magmatic compositions.
7. Conclusion
Acknowledgements
Our study of the very first alteration stage of
Mount Cameroon recent basalts shows that:
We are grateful to both S. Fourcade and F.
Martineau for oxygen isotope analyses. M. Le CozBouhnik is acknowledged for her support in the XRF
analyses. We thank J. Samuel for supplying the ICPMS analyses in Strasbourg, J.-L. Joron for the INAA
analyses, and J. Macé and N. Morin for their analytical
assistance during the isotope analyses. This work was
supported by the French programs PEGI (Programme
Environnement Géosphère Intertropicale) and PROSE
(Programme Sol et Erosion) funded by the INSU/
CNRS and the IRD agencies. Careful reviews by S.R.
Gı̀slason, A. White and E. Oelkers greatly helped
improving a first version of this manuscript while
numerous suggestions made by N.T. Arndt helped
improving this version of the manuscript. Constructive
comments by two anonymous reviewers helped to
improve the final version of this manuscript.
–
–
–
LOI and d 18O increase systematically from the
inner parts of the flows to the surface. This reflects
a slight hydration of the surficial parts of lava
flows.
Over a period shorter than one century, weathering
can produce changes in Na abundances. Lower
Na2O concentrations (and maybe U) are obtained
on the surficial parts of lavas. The largest differences still remain subtle at about 8%. Variations in
Na2O content are certainly related to hydration
processes while the possible U mobility might be
controlled by the presence of organic ligands in
solution. Changes in U content are not accompanied by variations in the (234U/238U) activity
ratio since all analysed lavas are still in secular
equilibrium.
Most elements do not show any concentration
change between internal and outermost parts of
flows. This is the case for elements such as Ti, Th
and Nb which are known to be robust even in
severe alteration conditions. This is also true for
Appendix A. Comparison between INAA, XRF
and ICP-MS data
Because weathering of very recent lava flows
might create only limited changes in trace element
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
concentrations, detecting these shifts requires accurate
analyses. To constrain the data quality, trace element
contents were analysed using three different techniques: INAA, XRF and ICP-MS. Comparison of the
ICPMS data from Nancy and Strasbourg (Table A1),
combined with reference to analyses of international
rock standards, shows that for most elements, differences are smaller than deviations based on the
reproducibility of the analyses. Normally, such differences would be considered significant only if they
reach 20%, which corresponds to the sum of errors of
both measuring techniques. Even though most elements have concentrations that are comparable using
the three different techniques, significant discrepancies exist for Cs, Rb, Ba and some REE.
Cs and Rb concentrations determined by ICP-MS
in Nancy are systematically higher than those
measured in Strasbourg. Differences vary from
+80% to +200% for Cs (see Fig. A1) and from +5%
to +40% for Rb. The most likely interpretation for the
higher Cs contents obtained in Nancy is that the very
low Cs contents of our samples are very close to the
detection limit of the method used in Nancy and in
this situation, the relative contribution of the blank
(0.15F0.15 ppm in Nancy, J. Carignan, personal
communication) becomes significant. Cs concentrations measured by INAA are generally similar to the
Strasbourg ICP-MS data (Table A1; Fig. A1).
Rb concentrations obtained by ICP-MS in Nancy
are also systematically higher than those measured in
Strasbourg (see Table A1 and Fig. A1). This excess is
more difficult to explain by contamination from
lithium metaborate or Rb contents too close to the
detection limit because the measured concentrations
are not so low. Furthermore, a good agreement exists
between the Nancy data and the XRF data (see Fig.
A1). The INAA analyses give results intermediate
between the Nancy and the Strasbourg data but their
deviations from the Strasbourg values never exceed
20%.
Rb and Cs concentrations measured in Strasbourg
in the international standard BR (basalt) were 48.4
and 0.88 ppm, respectively (Table A1). These values
are similar to the certified values reported by
Govindaraju (1994): Rb=47 ppm, Cs=0.8 ppm. Rb
concentrations in BEN and in another basaltic geostandard, BHVO, were also determined by XRF.
Measurements gave values of 50 ppm for BEN and
217
11 ppm for BHVO. These values agree with those of
Govindaraju (1994): RbBEN=47 ppm, RbBHVO=11
ppm. However, in the case of the standard BHVO,
several authors reported Rb concentrations lower than
11 ppm. These measurements were obtained by XRF
[RbBHVO=8.9 ppm (Elliott et al., 1997)], by ICP-MS
[RbBHVO=9.2 ppm (Jenner et al., 1990); RbBHVO=9.1
ppm (Hollocher et al., 1995)] and by isotopic dilution
[RbBHVO=9.0 ppm (Hergt et al., 1989); RbBHVO=9.25
ppm (Rautenschlein et al., 1985); RbBHVO=9.3 ppm,
average of four isotopic dilution measurements made
in Géosciences Rennes]. Since isotopic dilution mass
spectrometric method gives the most accurate data, we
suggest that the 11 ppm Rb content in BHVO used to
calibrate the XRF is too high. This would explain why
XRF values are higher than the INAA and Strasbourg
ICP-MS data. The differences between the ICPMS
data from Nancy and Strasbourg values are difficult to
explain. Indeed, when the basaltic standard BR,
whose Rb content is similar to that of our samples,
was measured at Nancy, it gave a Rb value of 46.8
ppm which is the same as the certified value of 47
ppm (Govindaraju, 1994).
Very similar Ba contents are determined by ICPMS in Nancy and in Strasbourg and by INAA (see
Fig. A1). In contrast, the XRF data are 21–43% lower
than the Strasbourg ICP-MS data. When the standards
BHVO and BEN were analyzed as unknowns their Ba
contents were 137 and 774 ppm, respectively (see
Table A1), compared with certified values of 139 and
1025 ppm (Govindaraju, 1994). The low value for
BEN suggests that Ba measured by XRF on our
samples are inaccurate and significantly too low.
Although the REE generally reproduce well, in a
few samples, Ce, Tm and Lu contents differ by up
14% between the Nancy and Strasbourg ICP-MS data
(see Table A1). While the differences in Tm and Lu
could be attributed to low signals, this is not the case
for Ce whose abundances is high (c150 ppm).
In summary, despite the minor discrepancies
discussed above, the data obtained using the different
methods agree well. In order to avoid the small
differences that do exist, in our interpretation, we used
a single set of data. Because accurate data for the
alkali elements—potentially the most mobile—was
important, we avoided the Nancy ICP-MS and the
XRF data. We therefore selected the ICP-MS data
from Strasbourg, complemented by the Pb, Cu and Ni
218
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Table A1
Calculated relative variations (expressed in %) between the trace element concentrations determined by ICP-MS at CGS-Strasbourg and those determined in three other laboratories
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Variations in content greater than 20% are shaded. The concentrations measured for the geostandards BR, BEN and BHVO are also reported.
219
220
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Fig. A1. Cs, Rb and Ba concentrations determined by ICP-MS at CRPG-Nancy (filled triangles), INAA (open circles) and XRF (open and filled
squares) vs. concentrations obtained by ICP-MS at CGS-Strasbourg on the same rock powders. The line bSlope=1Q corresponds to equal
concentrations and the shaded area defines the field where concentration differences are less than 20%.
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
concentrations from Nancy because these elements
were not measured elsewhere.
References
Aiuppa, A., Allard, P., d’Alessandro, W., Michel, A., Parello, F.,
Treuil, M., Valenza, M., 2000. Mobility and fluxes of major,
minor and trace metals during basalt weathering and groundwater transport at Mt Etna volcano (Sicily). Geochim. Cosmochim. Acta 64, 1827 – 1841.
Alderton, D.H.M., Pearce, J.A., Potts, P.J., 1980. Rare earth element
mobility during granite alteration: evidence from southwest
England. Earth Planet. Sci. Lett. 49, 149 – 165.
Ballentine, C.J., Lee, D.-C., Halliday, A.N., 1997. Hafnium isotopic
studies of the Cameroon line and new HIMU paradoxes. Chem.
Geol. 139, 111 – 124.
Benedetti, M.F., Menard, O., Noack, Y., Carvalho, A., 1994. Water–
rock interactions in tropical catchments: field rates of weathering and biomass impact. Chem. Geol. 118, 203 – 220.
Benedetti, M.F., Dia, A., Riotte, J., Chabaux, F., Gérard, M.,
Boulègue, J., Fritz, B., Chauvel, C., Bulourde, M., Déruelle, B.,
Ildefonse, P., 2003. Chemical weathering of basaltic lava flows
undergoing extreme climatic conditions: the water geochemistry
record. Chem. Geol. 201, 1 – 17.
Braun, J.-J., Pagel, M., Muller, J.-P., Bilong, P., Michard, A.,
Guillet, B., 1990. Cerium anomalies in lateritic profiles.
Geochim. Cosmochim. Acta 54, 781 – 795.
Bulourde, M., 2001. Processus d’altération des basaltes du Mont
Cameroun: approche géochimique. PhD thesis, Université de
Rennes, Rennes, 271 pp.
Cerling, T.H., Brown, F.H., Bowman, J.R., 1985. Low-temperature
alteration of volcanic glass: hydration, Na, K, 18O and Ar
mobility. Chem. Geol., Isot. Geosci. Sect. 52, 281 – 293.
Chabaux, F., Cohen, A.S., O’Nions, R.K., Hein, J.R., 1995.
238
U–234U–230Th chronometry of Fe–Mn crusts. Growth processes and recovery of Thorium isotopic ratios of sea water.
Geochim. Cosmochim. Acta 59, 633 – 638.
Chabaux, F., Riotte, J., Benedetti, M.F., Boulègue, J., Gérard, M.,
Ildefonse, P., 1998. Uranium isotopes in surface waters from the
Mount Cameroon: tracing water sources or basalt weathering?
Mineral. Mag. 62, 296 – 297.
Chabaux, F., Riotte, J., Dequincey, O., 2003. U–Th–Ra fractionation during weathering and river transport. Rev. Mineral.
Geochem. 52, 533 – 576.
Chauvel, C., Hofmann, A.W., Vidal, P., 1992. HIMU-EM: the
French Polynesian connection. Earth Planet. Sci. Lett. 110,
99 – 119.
Clayton, R.N., Mayeda, T.K., 1963. The use of bromine pentafluoride in the extraction of oxygen from oxydes and silicates
for isotopic analysis. Geochim. Cosmochim. Acta 27, 43 – 52.
Cochran, M.F., Berner, R.A., 1996. Promotion of chemical weathering by higher plants: field observations on Hawaiian basalts.
Chem. Geol. 132, 71 – 77.
Cotten, J., Le Dez, A., Bau, M., Caroff, M., Maury, R.C., Dulski, P.,
Fourcade, S., Bohn, M., Brousse, R., 1995. Origin of anomalous
221
rare-earth element and yttrium enrichments in subaerially
exposed basalts: evidence from French Polynesia. Chem. Geol.
119, 115 – 138.
Cramer, J.J., Nesbitt, H.W., 1983. Mass-balance relations and traceelement mobility during continental weathering of various
igneous rocks. Sci. Géol., Mém. 73, 63 – 73.
Crovisier, J.-L., Atassi, H., Daux, V., Eberhart, J.-P., 1990.
Hydrolyse d’un verre basaltique tholéiitique à 60 8C. Dissolution sélective puis congruente par élévation du pH. Compte
Rendus de l’Académie des Sciences de Paris t.310 (Série II),
941 – 946.
Crovisier, J.-L., Advocat, T., Dussossoy, J.-L., 2003. Nature and
role of natural alteration gels fromed on the sruface of ancient
volcanic glasses (natural analogs of waste containment glasses).
J. Nucl. Mater. 321, 91 – 109.
Dahlgren, R.A., Ugolini, F.C., Casey, W.H., 1999. Field weathering
rates of Mt. St. Helens tephra. Geochim. Cosmochim. Acta 63
(5), 587 – 598.
Daux, V., 1992. Comportement du strontium, des terres rares, de
l’uranium et du thorium pendant l’altération de verres basaltiques d’Islande. 3ème Cycle thesis, Strasbourg, 135 pp.
Daux, V., Crovisier, J.-L., Hémond, C., Petit, J.-C., 1994. Geochemical evolution of basaltic rocks subjected to weathering:
fate of the major elements, rare earth elements and thorium.
Geochim. Cosmochim. Acta 58, 4941 – 4954.
Daux, V., Guy, C., Advocat, T., Crovisier, J.-L., Stille, P., 1997.
Kinetic aspects of basaltic glass dissolution at 90 8C: role of
aqueous silicon and aluminium. Chem. Geol. 142, 109 – 126.
Déruelle, B., N’ni, J., Kambou, R., 1987. Mount Cameroon: an
active volcano of the Cameroon Line. J. Afr. Earth Sci. 6 (2),
197 – 214.
Dessert, C., Dupré, B., François, L.M., Schott, J., Gaillardet, J.,
Chakrapani, G., Bajpai, S., 2001. Erosion of Deccan Traps
determined by river geochemistry: impact on the global climate
and the 87Sr/86Sr ratio of seawater. Earth Planet. Sci. Lett. 188,
459 – 474.
Dupuy, C., Barsczus, H.G., Liotard, J.M., Dostal, J., 1988. Trace
element evidence for the origin of ocean island basalts: an
example from the Austral Islands (French Polynesia). Contrib.
Mineral. Petrol. 98, 293 – 302.
Elliott, T., Plank, T., Zindler, A., White, W., Bourdon, B., 1997.
Element transport from slab to volcanic front at the Mariana arc.
J. Geophys. Res. 102 (B7), 14991 – 15019.
Feigenson, M.D., Hofmann, A.W., Spera, F.J., 1983. Case studies
on the origin of basalt: II. Th transition from tholeiitic to alkalic
volcanism on Kohala volcano, Hawaii. Contrib. Mineral. Petrol.
84, 390 – 405.
Fitton, J.G., Dunlop, H.M., 1985. The Cameroon line, West
Africa, and its bearing on the origin of oceanic and
continental alkali basalt. Earth Planet. Sci. Lett. 72, 23 – 38.
Gislason, S.R., Oelkers, E.H., 2003. Mecanism, rates and consequences of basaltic glass dissolution: II. An experimental study
of the dissolution rates of basaltic glasses as a function of pH and
temperature. Geochim. Cosmochim. Acta 67, 3817 – 3832.
Gı́slason, S.R., Arnórsson, S., Ármannsson, H., 1996. Chemical
weathering of basalt in southwest Iceland: effects of runoff, age
of rocks and vegetative/glacial cover. Am. J. Sci. 296, 837 – 907.
222
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Govindaraju, K., 1994. Geostand. Newsl. 18 (Special Issue) 331 pp.
Grambow, B., Mqller, R., 2001. First-order dissolution rate law and
the role of surface layers in glass performance assessment. J.
Nucl. Mater. 298, 112 – 124.
Guy, C., Daux, V., Schott, J., 1999. Behaviour of rare earth elements
during seawater/basalt interactions in the Mururoa massif.
Chem. Geol. 158, 21 – 35.
Halbach, P., Von Borstel, D., Gundermann, K.-D., 1980. The uptake
of uranium by organic substances in a peat bog environment on
a granitic bedrock. Chem. Geol. 29, 117 – 138.
Hall, A., 1987. Igneous Petrology. Longman Scientific and
Technical 573 pp.
Halliday, A.N., Dickin, A.P., Fallick, A.E., Fitton, J.G., 1988.
Mantle dynamics: a Nd, Sr, Pb and O isotopic study of the
Cameroon Line Volcanic Chain. J. Petrol. 29 (part 1), 181 – 211.
Halliday, A.N., Davidson, J.P., Holden, P., DeWolf, C., Lee, D.-C.,
Fitton, J.G., 1990. Trace-element fractionation in plumes and the
origin of HIMU mantle beneath the Cameroon line. Nature 347
(347), 523.
Harmon, R.S., Hoefs, J., 1995. Oxygen isotope heterogeneity of the
mantle deduced from global 18O systematics of basalts from
different geotectonic settings. Contrib. Mineral. Petrol. 120,
95 – 114.
Hergt, J.M., Chappell, B.W., McCulloch, M.T., McDougall, I.,
Chivas, A.R., 1989. Geochemical and isotopic constraints on the
origin of the jurassic dolerites of Tasmania. J. Petrol. 30 (Part 4),
841 – 883.
Hofmann, A.W., 1988. Chemical differenciation of the Earth: the
relationship between mantle, continental crust, and oceanic
crust. Earth Planet. Sci. Lett. 90, 297 – 314.
Hofmann, A.W., 1997. Mantle geochemistry: the message form
oceanic volcanism. Nature 385, 219 – 229.
Hofmann, A.W., White, W.M., 1983. Ba, Rb and Cs in the Earth’s
mantle. Zeitschrift fqr Naturforschung 38a, 256 – 266.
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.
Hollocher, K., Fakhry, A., Ruiz, J., 1995. Trace element determinations for USGS Basalt BHVO-1 and NIST standard
reference materials 278, 688 and 694 by inductively coupled
plasma-mass spectrometry. Geostand. Newsl. 19 (1), 35 – 40.
Innocent, C., Michard, A., Malengreau, N., Loubet, M., Noack, Y.,
Benedetti, M., Hamelin, B., 1997. Sr isotopic evidence for ionexchange buffering in tropical laterites from the Paranà, Brazil.
Chem. Geol. 136, 219 – 232.
Ivanovich, M., Harmon, R.S., 1982. Uranium Series Disequilibrium: Application to Environmental Problems. Oxford University Press. 571 pp.
Jenner, G.A., Longerich, H.P., Jackson, S.E., Fryer, B.J., 1990. ICPMS—a powerful tool for high-precision trace-element analysis
in Earth sciences: evidence from analysis of selected U.S.G.S.
reference samples. Chem. Geol. 83, 133 – 148.
Kyser, T.K., Cameron, W.E., Nisbet, E.G., 1986. Boninite
petrogenesis and alteration history: constraints from stable
isotope compositions of boninites from Cape Vogel,
New Caledonia and Cyprus. Contrib. Mineral. Petrol. 93,
222 – 226.
Langmuir, D., 1978. Uranium solution–mineral equilibria at low
temperatures with applications to sedimentary ore deposits.
Geochim. Cosmochim. Acta 42, 547 – 569.
Lawrence, J.R., Taylor, H.P., 1972. Hydrogen and oxygen isotope
systematics in weathering profiles. Geochim. Cosmochim. Acta
36, 1377 – 1393.
Le Gal, X., Crovisier, J.-L., Gauthier Lafaye, F., Honnorez, J.,
Grambow, B., 1999. Meteoric alteration of Icelandic volcanic
glass: long-term changes in the mecanism. Comptes Rendus de
l’Académie des Sciences de Paris 329, 175 – 199.
Lee, D.C., Halliday, A.N., Fitton, J.G., Poli, G., 1994. Isotopic
variations with distance and time in the volcanic islands of the
Cameroon line: evidence for a mantle plume origin. Earth
Planet. Sci. Lett. 123, 119 – 138.
Louvat, P., 1997. Etude géochimique de l’érosion fluviale des ı̂les
océaniques à l’aide des bilans d’éléments majeurs et traces.
Unpublished PhD thesis, Paris 7 University, Paris.
Manhès, G., Allègre, C.J., Provost, A., 1984. U–Th–Pb systematics of the eucrite bJuvinasQ: precise age determination and
evidence for exotic lead. Geochim. Cosmochim. Acta 48,
2247 – 2264.
Martinez, M.P., Turi, B., 1978. The isotopic composition of oxygen
and carbon in the hyaloclastites from the Mt. Iblei volcanic
area, Eastern Sicily: a preliminary study. Bull. Volcanol. 41 (3),
168 – 174.
Middelburg, J.J., Van Der Weijden, C.H., Woittiez, J.R.W., 1988.
Chemical processes affecting the mobility of major, minor and
trace elements during weathering of granitic rocks. Chem. Geol.
68, 253 – 273.
Moulton, K.L., Berner, R.A., 1998. Quantification of the effect of
plants on weathering: studies in Iceland. Geology 26, 895 – 898.
Nesbitt, H.W., Wilson, R.E., 1992. Recent chemical weathering of
basalts. Am. J. Sci. 292, 740 – 777.
Oelkers, E.H., Gislason, S.R., 2001. Mecanism, rates and consequences of basaltic glass dissolution: I. An experimental study
of the dissolution rates of basaltic glasses as a function of
aqueous Al, Si and oxalic acid concentration at 25 8C and pH=3
and 11. Geochim. Cosmochim. Acta 65, 3671 – 3681.
Palacz, Z.A., Saunders, A.D., 1986. Coupled trace element and
isotope enrichment in the Cook–Austral–Samoa islands,
southwest Pacific. Earth Planet. Sci. Lett. 79, 270 – 280.
Petit, J.-C., Della Mea, G., Dran, J.C., Magontier, M.-C., Mando,
P.A., Pacagnella, A., 1990. Hydrated-layer formation during
dissolution of complex silicate glasses and minerals. Geochim.
Cosmochim. Acta 54, 1941 – 1955.
Price, R.C., Gray, C.M., Wilson, R.E., Frey, F.A., Taylor, S.R.,
1991. The effects of weathering on rare-earth element, Y and Ba
abundances in Tertiary basalts from southeastern Australia.
Chem. Geol. 93, 245 – 265.
Prudêncio, M.I., Braga, M.A.S., Gouveia, M.A., 1993. REE
mobilization, fractionation and precipitation during weathering
of basalts. Chem. Geol. 107, 251 – 254.
Rautenschlein, M., Jenner, G.A., Hertogen, J., Hofmann, A.W.,
Kerrich, R., Schmincke, H.U., White, W.M., 1985. Isotopic and
trace element composition of volcanic glasses from the Akaki
Canyon, Cyprus: implications for the origin of the Troodos
ophiolite. Earth Planet. Sci. Lett. 75, 369 – 383.
C. Chauvel et al. / Journal of Volcanology and Geothermal Research 141 (2005) 195–223
Richard, P., Shimizu, N., Allègre, C.J., 1976. 143Nd/146Nd, a natural
tracer: an application to oceanic basalts. Earth Planet. Sci. Lett.
31, 269 – 278.
Riotte, J., Chabaux, F., 1999. (234U/238U) activity ratios in freshwaters: tracing hydrological processes or bed-rock weathering?
Case of Strengbach watershed (Vosges, France). Geochim.
Cosmochim. Acta 63, 1263 – 1275.
Riotte, J., Chabaux, F., Benedetti, M.F., Dia, A., Gérard, M.,
Boulègue, J., Etamé, J., 2003. Uranium colloidal transport and
origin of the 234U–238U fractionation in surface waters: new
insights from Mount Cameroun. Chem. Geol. 202, 365 – 381.
Saar, M.O., Manga, M., 1999. Permeability–porosity relationship
in vesicular basalts. Geophys. Res. Lett. 26 (1), 111 – 114.
Sun, S.-S., McDonough, W.F, 1989. Chemical and isotopic
systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A.D., Norry, M.J. (Eds.),
Magmatism in the Ocean Basins. Spec. Pub. Geol. Soc. Lond.,
pp. 313 – 345.
Techer, I., Advocat, T., Lancelot, J., Liotard, J.M., 2001.
Dissolution kinetic of basaltic glasses: control by solution
223
chemistry and protective effect of the alteration film. Chem.
Geol. 176, 235 – 263.
Todt, W., Cliff, R.A., Hanser, A., Hofmann, A.W., 1996.
Evaluation of a 202Pb–205Pb double spike for high-precision
lead isotope analysis. In: Basu, A., Hart, S.R. (Eds.), Earth
Processes: Reading the Isotopic Code. AGU Geophysical
Monograph, pp. 429 – 437.
Venturelli, G., Contini, S., Bonazzi, A., 1997. Weathering of
ultramafic rocks and element mobility at Mt. Prinzera, Northern
Apennines, Italy. Mineral. Mag. 61, 765 – 778.
Weaver, B.L., Wood, D.A., Tarney, J., Joron, J.-L., 1987. Geochemistry of ocean island basalts from the South Atlantic:
ascension, Bouvet, St. Helena. Gough and Tristan da Cunha. In:
Fitton, J.G., Upton, B.G.J. (Eds.), Alkaline Igneous Rocks.
Spec. Pub. Geol. Soc. Lon., pp. 253 – 267.
White, A.F., Hochella, M.F., 1992. Surface chemistry associated
with the cooling and subaerial weathering of recent basalt flows.
Geochim. Cosmochim. Acta 56, 3711 – 3721.

Download Report

Do decades of tropical rainfall affect the chemical compositions of

Paperzz.com

Your Paperzz