Download Supporting Online Material for

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Geophysical Research Letters
Supporting Information for
Temporal variations of riverine dissolved lithium isotopic signatures unveil contrasting
weathering regimes in low-relief Central Africa
Soufian Henchiri1, Jérôme Gaillardet1, Mathieu Dellinger1,2, Julien Bouchez1, Robert G.M.
Spencer3
1Institut
de Physique du Globe de Paris, Sorbonne Paris Cité, Univ Paris Diderot, CNRS, F-75005 Paris, France
of Earth Sciences, University of Southern California, Los Angeles, CA 90089, USA
3Department of Earth, Ocean & Atmospheric Science, Florida State University, Tallahassee, FL 32309, USA
2Department
Contents of this file
Text S1 to S5
Figures S1 and S2
Tables S1 to S3
Introduction
Supporting Information consists of 5 texts, 2 figures and 3 tables that includes the following
sections:
Text S1. Analytical methods
Text S2. Atmospheric inputs
Text S3. Lithium isotope composition of the average bedrock drained by the Oubangui
River
Text S4. Quantitative models for the Li isotopic signature of the end members
Text S5. Soils and suspended matter data
FIGURE S1. Rayleigh distillation models for calculating the shift between the solution and
the bedrock (Δ7Liw-bedrock = 7Liw - 7Libedrock) and between the secondary weathering
products in contact with the solution and the bedrock (Δ7Lisec-bedrock = 7Lisec - 7Libedrock) as a
function of the fraction of Li remained in solution after secondary mineral formation (fLi).
FIGURE S2. Open flow-through reactor models for calculating the shift between the solution
and the bedrock (Δ7Liw-bedrock = 7Liw - 7Libedrock) and between the secondary weathering
products in contact with the solution and the bedrock (Δ7Lisec-bedrock=7Lisec - 7Libedrock) as a
function of the fraction of Li remained in solution after secondary mineral formation (fLi).
1
TABLE S1. Major cations, anions, Sr and Li concentrations, and Sr and Li isotopic
composition of the riverine dissolved load.
TABLE S2. Concentrations corrected for atmospheric inputs and proportion coming from the
atmosphere calculated for Na, Sr and Li.
TABLE S3. Concentrations of Li and Na and Li isotopic compositions of soil samples from
Betou and suspended sediments from Congo at mouth.
2
Text S1. Analytical methods
1. Major and trace elements analyses
Concentrations of dissolved cations and anions were measured respectively by
HPLC Dionex 300 and 120 at the Institut de Physique du Globe de Paris (IPGP), with a
precision better than 5%. Concentrations of dissolved chlorine for samples “C89-24 8
pK345” and “C89-5a pK87” were measured by Quadrupole ICP-MS (Thermo Scientific,
precision 5%).
2. Lithium purification and Li isotopes measurements
For dissolved Li isotope measurements, volumes of water samples containing 20 ng
of Li (or 10 ng for the “blackwater rivers” samples because of available volumes) were first
evaporated and the residue was digested by 16 N HNO3 in order to remove the residue of
organic matter. After drying and uptake in HCl, Li was then separated from the river water
matrix by ion-exchange chromatography using resin Bio-Rad AG50-X12 resin (method
modified from James and Palmer [2000] and described in Dellinger et al., [2014]). The
procedure for analyzing Li isotopes of solid samples is the same as Dellinger et al. [2014]
and is described therein. The volume of the resin was 2.7 mL and Li was eluted in 0.2 N HCl.
The elution fraction containing Li was evaporated at a temperature of 90°C and the
residue Li was kept as a salt until uptake in HNO3 shortly before the measurement.
Lithium isotope ratios (7Li/6Li) were measured by multicollector ICP-MS (Neptune,
Thermo Scientific) at IPGP using the Apex sample inlet system. The ratios were measured
three times for each sample and normalized to the L-SVEC standard (with the same
concentration as the sample, within 10% error) [Flesch et al.,1973] using the standardsample bracketing method [Dellinger et al., 2014]. The Li isotope composition is expressed
using the delta notation (in ‰):
7
Li
⁄6 )
Li sample
(
δ7 Li =
7
(
(
Li
⁄6 )
Li L−SVEC
− 1 × 103
)
The reproducibility of the Li chemical purification procedure was checked using the
seawater standard reference NASS5 (long-term average 7Li = 31.06 ± 0.91 ‰, 2σ, n = 34
separations and measurements).
3. Strontium purification and Sr isotopes measurements
We used Sr isotopes to trace the solute sources of the Congo River. The automated
purification of Sr from river samples was performed by high performance ion
chromatography (HPLC Dionex 300) according to the procedure described in Meynadier et
al. [2006]. Sr isotope ratios (87Sr/86Sr) were measured on the multicollector ICP-MS
(Neptune, Thermo Scientific) at IPGP. Interferences of Kr (84Kr and 86Kr) were corrected by
measuring 83Kr/84Kr and 83Kr/86Kr of the Neptune background (i.e. in Ar gas and HNO3) at the
beginning of the measurement session. Interferences of 87Rb were corrected by measuring
85
Rb/87Rb of a low-concentration Rb solution at the beginning of the measurement session.
Instrumental mass bias was corrected using the exponential law and the 86Sr/88Sr ratio,
assumed to be equal to 0.1194. The Sr isotope ratios were measured three times for each
3
sample, and the reported uncertainty is the reproducibility (95%-confidence interval)
calculated over these three replicates. Due to the dilute nature of some rivers and to the
limited sample volume available, amounts as low as a few ng of Sr were analyzed for some
samples, resulting in relatively high uncertainty on the order of 10-3, which was nevertheless
sufficient for the purpose of this study. The accuracy of measurements was checked using
the international isotope standard SRM 987. Because of low available volumes for samples
“C89-5a pK 87” and “C89-24 8 pK 345”, their Sr isotopes were not measured. The 87Sr/86Sr
ratio of these samples is assumed to be close to that of the Oubangui from Négrel et al.
[1993].
References :
Dellinger, M., Gaillardet, J., Bouchez, J., Calmels, D., Galy, V., Hilton, R. G., ... & FranceLanord, C. (2014). Lithium isotopes in large rivers reveal the cannibalistic nature of modern
continental weathering and erosion. Earth and Planetary Science Letters, 401, 359-372.
Flesch, G. D., Anderson, A. R., & Svec, H. J. (1973). A secondary isotopic standard for 6Li/7Li
determinations. International Journal of Mass Spectrometry and Ion Physics, 12(3), 265-272.
James, R. H., & Palmer, M. R. (2000). The lithium isotope composition of international rock
standards. Chemical Geology, 166(3), 319-326.
Meynadier, L., Gorge, C., Birck, J. L., & Allègre, C. J. (2006). Automated separation of Sr
from natural water samples or carbonate rocks by high performance ion chromatography.
Chemical geology, 227(1), 26-36.
Négrel, P., Allègre, C. J., Dupré, B., & Léwin, E. (1993). Erosion sources determined by
inversion of major and trace element ratios and strontium isotopic ratios in river water: the
Congo Basin case. Earth and Planetary Science Letters, 120(1), 59-76.
4
Text S2. Atmospheric inputs
In regions characterized by low chemical weathering rates, atmospheric inputs can
significantly influence the chemistry of the dissolved load [Négrel et al., 1993]. Sources of
chemical elements in rainwater are dust and seawater. Atmospheric dusts are assumed here
to be mainly derived from the drainage basin and are therefore not considered as external
inputs. The concentration of Li derived from seasalts can then be calculated using the
seawater molar Li/Cl ratio (5.10-5, [Millot et al., 2010]):
Li
[Li]rain = [Cl]rain × ( )
Cl seawater
Li
( )
= 5 ∙ 10−5 [Millot et al. , 2010]
Cl seawater
As inputs from evaporitic rocks are relatively low in the Congo rivers system [Négrel
et al., 1993], virtually all riverine Cl is derived from the atmosphere ([Cl]rain). The same type
of equation can be written for Na and Sr. Results highlight that the proportion of Li coming
from rainwaters in the analyzed samples never exceed 2% (Table S2). It is remarkable that
even in the Congo Basin where chemical weathering rates are amongst the lowest on Earth
[Gaillardet et al., 1995], 98% of dissolved Li is sourced from the bedrock. According to our
results, the proportion of atmospheric Na is generally below 30% but can reach 53% or more
in the organic-rich rivers, whereas the proportion of Sr originating from atmospheric inputs
does not exceed 5% (Table S2). The proportions of atmospheric Na and Sr calculated here
are consistent with those determined by a mixing model solved using an inverse method
[Négrel et al., 1993].
References :
Gaillardet, J., Dupré, B., & Allègre, C. J. (1995). A global geochemical mass budget applied
to the Congo Basin rivers: erosion rates and continental crust composition. Geochimica et
Cosmochimica Acta, 59(17), 3469-3485.
Millot, R., Vigier, N., & Gaillardet, J. (2010). Behaviour of lithium and its isotopes during
weathering in the Mackenzie Basin, Canada. Geochimica et Cosmochimica Acta, 74(14),
3897-3912.
Négrel, P., Allègre, C. J., Dupré, B., & Léwin, E. (1993). Erosion sources determined by
inversion of major and trace element ratios and strontium isotopic ratios in river water: the
Congo Basin case. Earth and Planetary Science Letters, 120(1), 59-76.
5
Text S3. Lithium isotope composition of the average bedrock drained by the Oubangui
River
Considering a basin-scale steady state between the soil formation rate by chemical
weathering and their destruction rate by mechanical erosion (a reasonable hypothesis for
the Congo Basin [Gaillardet et al., 1995]) and particularly for the periphery of the Basin, we
can estimate the Li isotopic signature and the molar Li/Na ratio of the mean bedrock
Li
(denoted as δ7 Libedrock and (Na)
bedrock
, respectively) drained by the Oubangui River by
using the following mass budget equations:
Li
7
7
δ7 Libedrock = αLi
dissolved ∙ δ Lidissolved + (1 − αdissolved ) ∙ δ Lisolid
where αLi
dissolved =
[Li]*dissolved
(2)
[Li]solid ∙ SP + [Li]*dissolved
Li
Li ∗
Li
( )
=( )
αNa
)
(1 − αNa
dissolved + (
dissolved )
Na bedrock
Na dissolved
Na solid
where αNa
dissolved =
(1)
(3)
(4)
[Na]*dissolved
[Na]solid ∙ SP + [Na]*dissolved
In these equations:
αXdissolved is the proportion of the element X (Li or Na) in the dissolved load of the Oubangui
River near Bangui
[X]*dissolved is the concentration of dissolved X (Li or Na) in the Oubangui River near Bangui,
corrected for atmospheric inputs (see Text S2 and Table S2)
[X]solid is the concentration of X (Li or Na) in the residual solids of the Oubangui River near
Bangui
SP is the suspended sediment concentration in the Oubangui River near Bangui
δ7 Lidissolved is the dissolved Li isotope composition of the Oubangui River near Bangui
δ7 Lisolid is the Li isotope composition of the residual solids in the Oubangui River near
Bangui
Li ∗
Na dissolved
( )
is the molar Li/Na ratio of the dissolved load of the Oubangui River near
Bangui, corrected for atmospheric inputs (see Text S2 and Table S2)
Li
(Na)
solid
is the molar Li/Na ratio in the residual solids of the Oubangui River near Bangui
Data and numerical applications:
The dissolved Li concentration and the Li isotope composition of the Oubangui
River near Bangui (“C89-5 a pK87”) is 0.42 ppb and 25.8‰, respectively (Table S1). The
correction for Li atmospheric inputs is negligible (Text S2 and Table S2). The dissolved Na
concentration of the Oubangui River near Bangui, corrected for atmospheric inputs, is 966
ppb (Table S4). The suspended sediment concentration at mouth and by extrapolation in
the Oubangui River is 25 ± 5 mg/L [Gaillardet et al., 1995]. To estimate the composition of
the solids, we use two types of solid material derived from weathering: river suspended load
and soils. The average Li concentration and Li isotope composition in suspended sediments
at mouth are 26 ± 4.5 ppm and -5.6 ± 0.8 ‰ (n=2), respectively (Table S3 and Text S5). The
average Li concentration and 7Li of soils sampled at Betou (around 150 km downstream of
6
Bangui) are 33 ± 7 ppm (2σ, n=4) and -5.51 ± 0.6 ‰ (2σ, n=5), respectively (Table S3). The
average Na concentration of suspended sediments at mouth and of soils sampled at Betou
is 2385 ± 613 ppm (2σ, n=4) and 501 ± 93 ppm (2σ, n=4), respectively (Table S3).
Na
7
Using bulk soil data, we obtain αLi
river =34%, αriver = 99%, δ Libedrock = 5.2‰ and
Li
≈ 4.1×10-3 mol/mol. Using suspended sediments, our data lead to αLi
river =39%,
Na bedrock
Li
7
αNa
=3.5×10-3 mol/mol. Taking into account these
river =94%, δ Licrust =6.8‰ and (Na)
bedrock
𝐋𝐢
two estimates, we argue that δ7 Libedrock=6 ± 2‰ and molar (𝐍𝐚)
= 3.7 ± 0.8 × 10-3
𝐛𝐞𝐝𝐫𝐨𝐜𝐤
( )
for the crust drained by the Oubangui River.
The inferred rock Li isotopic composition is higher than the mean Upper
Continental Crust (UCC) [Teng et al., 2004; Sauzéat et al., 2015] value but lies in the range
of Archean rocks. Indeed, 7Li values reported so far in the literature shows that 7Li values
of Archean TTG range from 0.2 to 11.7‰ with an average value of 5.1 ± 2.4‰ (1σ, n=51)
[Teng et al., 2008; Wimpenny et al., 2010; Qiu, 2011]. The estimated molar Li/Na ratio is
also compatible with molar Li/Na of Archean TTG found in the literature (e.g. 3.2 ± 2.5×10-3
(1σ) (n=21) for South African Barberton TTG [Qiu, 2011]).
References :
Teng, F. Z., McDonough, W. F., Rudnick, R. L., Dalpé, C., Tomascak, P. B., Chappell, B. W.,
& Gao, S. (2004). Lithium isotopic composition and concentration of the upper continental
crust. Geochimica et Cosmochimica Acta, 68(20), 4167-4178.
Teng, F. Z., Rudnick, R. L., McDonough, W. F., Gao, S., Tomascak, P. B., & Liu, Y. (2008).
Lithium isotopic composition and concentration of the deep continental crust. Chemical
Geology, 255(1), 47-59.
Qiu, L. (2011). Lithium and 7Li behavior during metamorphic dehydration processes and
crustal evolution.
Sauzéat, L., Rudnick, R. L., Chauvel, C., Garçon, M., & Tang, M. (2015). New perspectives on
the Li isotopic composition of the upper continental crust and its weathering signature.
Earth and Planetary Science Letters, 428, 181-192.
Wimpenny, J., James, R. H., Burton, K. W., Gannoun, A., Mokadem, F., & Gíslason, S. R.
(2010). Glacial effects on weathering processes: new insights from the elemental and
lithium isotopic composition of West Greenland rivers. Earth and Planetary Science Letters,
290(3), 427-437.
7
Text S4. Quantitative models for the Li isotopic signature of the end members
Chemical weathering is classically described by a simple two-step process:
dissolution of source minerals, and precipitation of secondary phases that re-incorporate
poorly soluble (such as Li) and insoluble elements. Li isotopes are portioned between
secondary products (6Li-enriched) and the solution (7Li-enriched).
Simple mass-budget models can be used to calculate the shift between the solution
and the bedrock (Δ7Liw-bedrock = 7Liw - 7Libedrock) as a function of the extent of Li
precipitation in secondary weathering products [Bouchez et al., 2013 ; Dellinger et al.,
2015]. These models allow us to predict the 7Li of the weathering solution under different
assumptions and using different isotopic fractionation factors. Thus, such models enable us
to interpret the isotopic signature of the two end members identified by the hydrologic
mixing model. Such a prediction is possible provided that the parent rock 7Li is known.
An open flow-through reactor model (equation 1) [Bouchez et al., 2013] or a
Rayleigh distillation model (equation 2) can account for the Li isotopic signature of the
residual dissolved phase (δ7 Liw) and assuming no isotopic fractionation during dissolution
of bedrock:
Δδ7 Liw−bedrock = δ7 Liw − δ7 Libedrock = −(1 − 𝑓𝐿𝑖 ) ∙ εsec−w
(1)
Δδ7 Liw−bedrock = δ7 Liw − δ7 Libedrock = εsec−w ∙ ln(𝑓𝐿𝑖 )
(2)
where 7Liw is the 7Li of the residual dissolved phase and fLi the remaining fraction of Li in
solution after the precipitation of secondary weathering products. The isotope fractionation
associated to secondary mineral (7Lisec) formation is quantified by an isotope fractionation
factor noted as εsec-w = 7Lisec - 7Liw ≈ 103.ln(sec-w) where sec-w = (7Li/6Li)sec/(7Li/6Li)w.
Δδ7 Liw−bedrock = δ7 Liw − δ7 Libedrock is displayed in Figures S1 (Rayleigh
distillation model) and S2 (open flow-through reactor) as a function of the extent of Li
precipitation in secondary phases (fLi).
The shift between the secondary minerals in contact with the solution and the
bedrock (Δ7Lisec-bedrock = 7Lisec - 7Libedrock) can also be predicted by such models (equation 3
for the instantaneous secondary minerals in the case of an open flow-through reactor
model and equations 4 and 5 for instantaneous and cumulated secondary minerals in the
case of a Rayleigh distillation model, respectively):
Δδ7 Lisec−bedrock = δ7 Lisec − δ7 Libedrock = 𝑓𝐿𝑖 ∙ εsec−w
(3)
Δδ7 Lisec−bedrock = δ7 Lisec − δ7 Li bedrock = (1 + ln(𝑓𝐿𝑖 )) ∙ εsec−w
(4)
Δδ7 Lisec−bedrock = (δ7 Libedrock + 103 ) (
1 − 𝑓𝐿𝑖 αsec−w
) − 103 − δ7 Libedrock
1 − 𝑓𝐿𝑖
(5)
Bedrocks underlying the rim of the Congo Basin are dominated by TTG, mainly of
Archean and Proterozoic ages [de Wit and Linol, 2015]. 7Li values reported so far in the
literature shows that 7Li values of Archean TTG range from 0.2 to 11.7‰ with an average
value of 5.2±2.5‰ (1σ, n=52) [Teng et al., 2008; Wimpenny et al., 2010 ; Qiu, 2011].
8
Moreover, a first order mass budget calculation for the Oubangui River can be attempted to
independently estimate the 7Li of the local continental crust of the northwestern part of
the Congo Basin, yielding 7Libedrock = 6 ± 2‰ (Text S3), and thus Δ7Liw-bedrock≈ 19 ± 2‰ for
the Oubangui River.
By assuming that Li and Na have the same behavior during dissolution of bedrock
and that only Li is then reincorporated into secondary weathering products, the proportion
of dissolved Li that has been reincorporated in secondary weathering minerals (fLi) can be
independently estimated by using the ratio between the Li/Na ratio in the residual dissolved
phase corrected for atmospheric inputs (Li/Na)∗w (see Text S2 and Table S2) and that of
bedrock (Li/Na)bedrock [Millot et al., 2010; Dellinger et al., 2015; Liu et al., 2015]:
fLi =
(Li/Na)∗w
(Li/Na)bedrock
(6)
As for 7Libedrock, a first order mass budget calculation which assume a basin-scale
steady state leads to (Li/Na)bedrock = 3.7 ± 0.8×10-3 (mol/mol) (Text S3). This value is
compatible with molar Li/Na of Archean TTG found in the literature (e.g. 3.2 ± 2.5×10-3 (1σ)
(n=21) for South African Barberton TTG [Qiu, 2011]). Accordingly, we predict that in order
to explain the high 7Li end member found in the Oubangui River, 40 ± 10% of the Li initially
released by rock dissolution is required to remain in solution (equation 6).
By considering a TTG bedrock with a 7Libedrock value of around 6 ± 2‰ and a value
for the residual fraction of Li deduced by Li/Na ratios (fLi ≈ 40 ± 10%), an isotopic
fractionation factor εsec-w of around -20.7‰ is needed so that the 7Li value of the
weathering solution predicted by a Rayleigh distillation model correspond to the ones
measured in the Oubangui River (Figure S1 and Table S1). Such a εsec-river value is low but is
still within the range of values found in the literature from studies on laterite profiles and
small catchment, ranging from -28‰ to -1‰ [Kisakürek et al., 2004; Vigier et al., 2009].
Interestingly, εsec-w≈-20.7‰ is similar to the value of εsec-w between 7Li associated with Feoxyhydroxydes and 7Li in the ultrafiltered riverwater from the West Greenland (i.e. -20‰)
[Wimpenny et al. 2010]. Besides, we note that this value is consistent with the εsec-w value (20‰) used by Wanner et al. [2014] in their reactive transport modeling approach for
simulating Li isotopic fractionation during pedogenetic processes. The open flow-through
reactor model (Figure S2) requires, to explain both 7Li of the Oubangui River and soils
data, a too high εsec-W value (≈-31.7‰) compared to the ones reported in the literature and is
therefore not retained to explain the end-member with high 7Liw.
The relevance of the mass balance models can be tested by comparing predicted
7Lisec values with secondary weathering products data. For the Congo Basin, bulk soil data
are representative of secondary weathering products as primary minerals are not present in
such very intensely weathered soils (Text S5 and Table S3). The average 7Li of soils at
Betou is -5.5 ± 0.6‰ (2σ, n=4 ) and Δ7Lisec-bedrock≈ -11.5 ± 2‰. The 7Li value of cumulated
secondary weathering products predicted by a Rayleigh distillation model based on a εsec-w
value of -20.7‰ is compatible with bulk soils data at Betou (Figure S1).
Large uncertainties remain on the bedrock Li/Na ratio and isotopic composition but
to a first order, the relative enrichment in 7Li of the Oubangui waters (end-member with
high 7Li) is due to the reincorporation of Li into secondary minerals. This result suggests
that the lateritic regions covering the borders of the Congo River are still active and produce
secondary weathering products.
The models can also be applied to the dilute blackwater rivers of the swampy
central depression. The bedrock of the central part of the Basin consists of Eocene or
Paleocene sedimentary rocks ranging from claystones and silty claystones to grainstones of
9
lacustrice origin [Mbandaka well, Guillocheau et al., 2015]. Their 7Li value is not known but
it can reasonably be predicted to be close to values typical of shales (-0.5 ± 2‰) [Dellinger
et al., 2014]. The molar Li/Na corrected for atmospheric inputs of blackwater rivers (e.g.
9.36×10-3 for C89-44 Sangha) (Table S2) is relatively high and is within the range spanned
by molar Li/Na of sedimentary bedrocks (5.10-3 for the weighted average of chemical
compositions of arenaceous rock types in central East China [Gao et al., 1998] to 17.10-3 for a
typical shale [Dellinger et al., 2014]). Despite large uncertainties due to the poor knowledge
of the bedrock, if we consider an isotopic fractionation factor εsec-w of -20.7‰, the
remaining fraction of Li in solution (fLi) should be comprised between 60 and 90% according
to the mass-budget models considered and the range of 7Libedrock (-0.5 ± 2‰) (Figures S1
and S2).
References:
Bagard, M. L., West, A. J., Newman, K., & Basu, A. R. (2015). Lithium isotope fractionation
in the Ganges–Brahmaputra floodplain and implications for groundwater impact on
seawater isotopic composition. Earth and Planetary Science Letters, 432, 404-414.
Bouchez, J., Von Blanckenburg, F., & Schuessler, J. A. (2013). Modeling novel stable isotope
ratios in the weathering zone. American Journal of Science, 313(4), 267-308.
Dellinger, M., Gaillardet, J., Bouchez, J., Calmels, D., Louvat, P., Dosseto, A., ... & Maurice,
L. (2015). riverine Li isotope fractionation in the Amazon river basin controlled by the
weathering regimes. Geochimica et Cosmochimica Acta.
Gao, S., Luo, T. C., Zhang, B. R., Zhang, H. F., Han, Y. W., Zhao, Z. D., & Hu, Y. K. (1998).
Chemical composition of the continental crust as revealed by studies in East China.
Geochimica et Cosmochimica Acta, 62(11), 1959-1975.
Georg, R. B., Reynolds, B. C., West, A. J., Burton, K. W., & Halliday, A. N. (2007). Silicon
isotope variations accompanying basalt weathering in Iceland. Earth and Planetary Science
Letters, 261(3), 476-490.
Guillocheau, F., Chelalou, R., Linol, B., Dauteuil, O., Robin, C., Mvondo, F., ... & Colin, J. P.
(2015). Cenozoic Landscape Evolution in and Around the Congo Basin: Constraints from
Sediments and Planation Surfaces. In Geology and Resource Potential of the Congo Basin
(pp. 271-313). Springer Berlin Heidelberg.
Kısakürek, B., Widdowson, M., & James, R. H. (2004). Behaviour of Li isotopes during
continental weathering: the Bidar laterite profile, India. Chemical Geology, 212(1), 27-44.
Liu, X. M., Wanner, C., Rudnick, R. L., & McDonough, W. F. (2015). Processes controlling δ7Li
in rivers illuminated by study of streams and groundwaters draining basalts. Earth and
Planetary Science Letters, 409, 212-224.
Millot, R., Vigier, N., & Gaillardet, J. (2010). Behaviour of lithium and its isotopes during
weathering in the Mackenzie Basin, Canada. Geochimica et Cosmochimica Acta, 74(14),
3897-3912.
Qiu, L. (2011). Lithium and 7Li behavior during metamorphic dehydration processes and
crustal evolution.
10
Vigier, N., Gislason, S. R., Burton, K.W., Millot, R., & Mokadem, F. (2009). The relationship
between riverine lithium isotope composition and silicate weathering rates in Iceland. Earth
and Planetary Science Letters, 287(3), 434-441.
Wanner, C., Sonnenthal, E. L., & Liu, X. M. (2014). Seawater 7Li: A direct proxy for global
CO2 consumption by continental silicate weathering? Chemical Geology, 381, 154-167.
Wimpenny, J., James, R. H., Burton, K. W., Gannoun, A., Mokadem, F., & Gíslason, S. R.
(2010). Glacial effects on weathering processes: new insights from the elemental and
lithium isotopic composition of West Greenland rivers. Earth and Planetary Science Letters,
290(3), 427-437.
de Wit, M. J., & Linol, B. (2015). Precambrian Basement of the Congo Basin and Its Flanking
Terrains. In Geology and Resource Potential of the Congo Basin (pp. 19-37). Springer Berlin
Heidelberg.
11
Text S5. Soils and suspended matter data
Two lateritic soil profiles were sampled 100 km downstream from Bangui in November
1989 at Betou. They comprise typical iron-rich duricrust, iron oxide rich levels and white
kaolinitic horizons. In addition, suspended matter, recovered from the 0.2 μm-porosity
acetate cellulose filters during a cruise on the Congo River in November 1989 [Négrel et al.,
1993 ; Dupré et al., 1996] between Bangui and Brazzaville were also analysed.
The average 7Li of soils sampled at Betou (around 150 km downstream of Bangui) is 5.51 ± 0.6 ‰ (n=4) with no significant differences between the different soil horizons (Table
S3). Values are lower than that those generally reported in the literature for soil samples
[eg. Pistiner et al., 2003; Lemarchand et al., 2010; Liu et al., 2013; Ryu et al., 2014] but are
consistent with values measured in a soil profile in Guadeloupe [Clergue et al., 2015] and
similar to the lowest 7Li value of bulk soils corresponding to iron-rich horizons in the Bidar
laterite profile developed upon the Deccan Traps flood basalt in India [Kısakürek et al.,
2004]. The average 7Li of suspended particulate matter of the Congo River near its mouth
is -5.63 ± 0.8 ‰ (n=2) (Table S3). This value is lower than that reported for the Amazon,
Mackenzie and Ganga-Brahmaputra River systems by Dellinger et al. [2014] (ranging from 3.6 for surface sediments to +1.5‰ for bottom suspended particulate matter).
References :
Clergue, C., Dellinger, M., Buss, H. L., Gaillardet, J., Benedetti, M. F., Dessert, C. (accepted
in Chemical Geology). Influence of atmospheric deposits and secondary minerals on Li
isotopes budget in a highly weathered catchment, Guadeloupe (Lesser Antilles).
Dellinger, M., Gaillardet, J., Bouchez, J., Calmels, D., Galy, V., Hilton, R. G., ... & FranceLanord, C. (2014). Lithium isotopes in large rivers reveal the cannibalistic nature of modern
continental weathering and erosion. Earth and Planetary Science Letters, 401, 359-372.
Kısakürek, B., Widdowson, M., & James, R. H. (2004). Behaviour of Li isotopes during
continental weathering: the Bidar laterite profile, India. Chemical Geology, 212(1), 27-44.
Lemarchand, E., Chabaux, F., Vigier, N., Millot, R., & Pierret, M. C. (2010). Lithium isotope
systematics in a forested granitic catchment (Strengbach, Vosges Mountains, France).
Geochimica et Cosmochimica Acta, 74(16), 4612-4628.
Liu, X. M., Rudnick, R. L., McDonough, W. F., & Cummings, M. L. (2013). Influence of
chemical weathering on the composition of the continental crust: insights from Li and Nd
isotopes in bauxite profiles developed on Columbia River Basalts. Geochimica et
Cosmochimica Acta, 115, 73-91.
Pistiner, J. S., & Henderson, G. M. (2003). Lithium-isotope fractionation during continental
weathering processes. Earth and Planetary Science Letters, 214(1), 327-339.
Ryu, J. S., Vigier, N., Lee, S. W., Lee, K. S., & Chadwick, O. A. (2014). Variation of lithium
isotope geochemistry during basalt weathering and secondary mineral transformations in
Hawaii.
Geochimica
et
Cosmochimica
Acta,
145,
103-115.
12
FIGURE S1. Rayleigh distillation models for calculating the shift between the solution and
the bedrock (Δ7Liw-bedrock = 7Liw - 7Libedrock) and between the secondary weathering
products and the bedrock (Δ7Lisec-bedrock = 7Lisec - 7Libedrock) as function of the fraction of Li
remained in solution after secondary mineral formation (fLi). The black solid curves
represent Δ7Liw-bedrock values as function of fLi for different values of isotopic fractionation
factor (εsec-w = 7Lisec - 7Liw). The grey solid curves and the dotted grey curves represent
Δ7Lisec-bedrock values as a function of fLi for instantaneous and cumulated (for 7Libedrock = 6‰)
secondary minerals, respectively. The blue areas represent Δ7Liw-bedrock values for the
samples of the Oubangui and the blackwater rivers (see Text S4). The green line represents
the average Δ7Lisec-bedrock of soil samples from Betou, Congo Basin (see Texts S3, S4 and S5
and Table S3). For comparison, εsec-w deduced from large river systems dominated by
floodplain processes (Amazon and Ganges-Brahmaputra) found in the litterature range
from -4‰ to -9‰ [Dellinger et al., 2015; Bagard et al., 2015]. The value of εsec-w ≈ -20.7‰
explains both Δ7Liw-bedrock of the Oubangui River (blue star and purple area) and Δ7Lisecbedrock of soil samples from Betou (green area) (see Text S4).
13
FIGURE S2. Open flow-through reactor models for calculating the shift between the solution
and the bedrock (Δ7Liw-bedrock = 7Liw - 7Libedrock) and between the secondary weathering
products and the bedrock (Δ7Lisec-bedrock = 7Lisec - 7Libedrock) as a function of the fraction of Li
remained in solution after secondary mineral formation (fLi). The black solid curves
represent Δ7Liw-bedrock values as a function of fLi for different values of isotopic fractionation
factor (εsec-w = 7Lisec - 7Liw). The grey solid curves represent Δ7Lisec-bedrock values as function
of fLi for secondary minerals, respectively. The blue areas represent Δ7Liw-bedrock values for
the samples of the Oubangui and the blackwater rivers (see Text S4). The green line
represents the average Δ7Liw-bedrock of soil samples from Betou, Congo Basin (see Texts S3,
S4 and S5 and Table S3). For comparison, εsec-w deduced from large river systems
dominated by floodplain processes (Amazon and Ganges-Brahmaputra) found in the
litterature range from -4‰ to -9‰ [Dellinger et al., 2015 ; Bagard et al., 2015]. The value of
εsec-w ≈ -20.7‰ explains both Δ7Liw-bedrock of the Oubangui River and Δ7Lisec-bedrock of soil
samples from Betou by using a distillation fractionation model (see Text S4). With this
model, a much lower and unrealistic εsec-w (-31.7‰) would be needed to explain the data.
14
TABLE S1. Major cations, anions, Sr and Li concentrations, and Sr and Li isotopic composition of the riverine dissolved load.
Date of sampling
Discharge
(m3/s)
Li
(nM)
Na
(μM)
Sr
(nM)
Cl
(μM)
δ7Li
(‰)
2σ Neptune
(‰)
Congo_140110
01/14/10
51880
111
64
134
18
14.66
0.08
0.718422
0.000010
Congo_180310
03/18/10
39910
101
79
137
17
18.79
0.12
0.719412
0.000229
Congo_240510
05/24/10
41260
100
71
156
18
17.54
0.24
0.719290
0.000137
Congo_130710
07/13/10
34280
97
75
151
7
20.61
0.10
0.719029
0.000017
Congo_210910
09/21/10
36300
88
82
136
8
22.13
0.09
0.718215
0.000079
Congo_251110
11/25/10
51920
127
72
131
3
16.16
0.17
0.718698
0.000175
C89-19 a pK288 Motaba
33
6
46
13
5.62
0.714237
0.000126
C89-44 Sangha
111
25
70
15
5.70
1.16
0.55
0.716236
0.000021
58
52
201
14
25.15
0.40
0.7196821
0.000029
0.10
1
0.000029
Samples
87
Sr/86Sr
95%CI
Time series
Blackwater rivers
Oubangui
C89-24 8 pK 345
11/01/89
C89-5 a pK87
61
61
187
17
25.84
0.719682
1 87
Sr/86Sr data for Oubangui from Négrel et al. [1993]
15
TABLE S2. Concentrations corrected for atmospheric inputs and proportion coming from the atmosphere calculated for Na, Sr and Li (Text S2).
Samples
Date of sampling
Na*
(μM)
Naatm/Naw
(%)
Sr*
(μM)
Sratm / Srw
(%)
Li*
(nM)
Liatm / Liw
(%)
Time series
Congo_140110
Congo_180310
Congo_240510
Congo_130710
Congo_210910
Congo_251110
01/14/10
03/18/10
05/24/10
07/13/10
09/21/10
11/25/10
49
64
55
70
74
69
23
18
22
8
9
3
0.13
0.13
0.15
0.15
0.14
0.13
2.18
2.05
1.94
0.74
1.03
0.36
111
100
99
96
88
127
0.79
0.84
0.91
0.35
0.48
0.11
[Na]atm > [Na]river
12
53
0.04
0.07
4.84
3.60
33
110
2.03
0.69
40
42
24
31
0.20
0.18
1.19
1.47
57
60
1.24
1.37
Blackwater rivers
C89-19 a pK288 Motaba
C89-44 Sangha
Oubangui
C89-24 8 pK 345
C89-5 a pK 87
11/01/89
16
TABLE S3. Concentrations of Li and Na and Li isotopic compositions of soil samples from Betou and suspended sediments from Congo at
mouth.
δ7Li L-SVEC e1
(‰)
δ7Li L-SVEC e2
(‰)
δ7Li (average)
(‰)
2σ
(‰)
-5.55
-5.57
-5.04
-5.55
-5.30
0.75
Samples
Type
Soils
Betou 1A
Betou 2A
Betou 1B
Betou 2B
Betou 1C
Soil
Soil
Soil
Soil
Soil
-5.27
-5.92
Suspended sediments
C89-63 pK 1105
C89-65 pK 1105
MES (Congo @mouth)
MES (Congo @mouth)
-6.20
-5.05
-5.27
-5.92
Li
(ppm)
Na
(ppm)
1
1
41
38
27
26
31
371
519
519
594
1
1
29
23
2998
1771
n separation
1
2
17