188
- The hypolimnion between 15 and 20 m and ca. 60 m
depth - The mesolimnion, where is located the chemocline,
between about 60 and 70 m depth - The bottom layer between 70 m depth and the bottom.
Monimolimnion includes both mesolimnion and bottom
layer.
A hydrologic budget was presented by Assayag et al.
( 2008 ), updated by constraints discussed in this paper (Fig.
11.1). New estimations lead to a water residence time of
about 9 years for mixolimnion and 100 years for monimo-
limnion. The evaporation amount is still questionable, but as
it has a low influence on the C budget, it is no further dis-
cussed in the present paper. Figure 11.1 presents also termi-
nology and transport parameters (Kz) used in the discussion.
11.3 Methods
Temperature, dissolved oxygen, conductivity and pH were
measured with a Seabird SBE 19 probe; turbidity was mea-
sured with a STBD 300 nke probe.
Water samples were obtained from a 1-l syringe actuated
via a messenger. Samples were divided into different ali-
quots and carefully stored with classical procedures (Viollier
et al. 1995 ):
- Major cations are measured either by AAS (before 2000)
or by ICP-AES - Anions by ionic chromatography
- Alkalinity by the spectrophotometric method of Podda
and Michard ( 1994 ), or by Gran ( 1952 ) titration with
hydrochloric acid, with care to avoid any oxygenation of
the water. - DIC was calculated in 1987 (Camus et al. 1993 ) from
alkalinity and pH. pH was obtained by CEA (Commissariat
à l’Energie Atomique) group using an home made pH
electrode suitable for in situ measurement at great depth.
Since that time, many other DIC profiles were either cal-
culated with the same technique, but with commercially
available pH probes (Michard et al. 1994 ; Viollier et al.
1995 , 1997 ; Aeschbach-Hertig et al. 1999 , 2002 ,...) or
measured directly as DIC associated with δ^13 C determina-
tion (Assayag et al. 2008 ).
For CH 4 concentration, water from the oxic mixolimnion
(from the surface to 55 m) were collected into serum bottles
(115 mL) and analyzed as described in Abril and Iversen
( 2002 ). In the methane rich monimolimnion (from 55 m to
90 m), 30 mL of water from the sampling syringe were rap-
idly transferred into sealed 115 mL serum bottles previously
flushed with nitrogen gas. With this procedure, most of the
depressurization occurred in the vial and the loss of CH 4
could be minimized. The methane concentrations were quan-
tified using a gas chromatograph (GC) equipped with flame
ionisation detector (FID). Sediment cores from 40 to 92 m
depth were sampled by means of UWITEC corer
(DL 96 × 0 cm) in order to determine methane concentration
profiles in pore waters for June 2007. For this purpose, lat-
eral sub-sampling of the core were performed at 2 cm step,
using 2.5 mL syringe with open end that was introduced in
the core tube in pre-drilled holes (initially closed by tape).
The small sediment cores were then transferred quickly in
pre-filled serum bottles containing 10 mL of 0.5 M
NaOH. Bottles were immediately capped with a septum plus
aluminium capsule.
Carbone dioxide exchange rate at the air-water interface
was determined in June 2010, June 2011 and December 2012
by the floating chamber method (Abril et al. 2005 ), measur-
ing gaseous CO 2 in the air volume enclosed in the chamber
with a LICOR LI-820 Gas Analyzer.
Sediment traps (Uwitec) have been deployed at four
depths in the water column (23, 58, 70 and 88 m). At each
depth were two traps each consisting of an internal tube
diameter of 85.7 mm and 60 cm long completed by a remov-
able 2 L bottle, wherein the particles accumulate. After a
deployment time of about 1 month between each campaign,
the particles are recovered by filtration over glass fiber filters
(GF/F Whatman), dried and weighed before being analysed
(C and N on analyser Thermo Scientific Flash elementary
CHNS 2000, major, minor and trace elements after digestion
by ICP-AES and MS). The traps of the anoxic zone were
treated under nitrogen to prevent the precipitation of dis-
solved iron in contact with air.
Determination of the flow for both springs around the lake
and the outlet was performed almost monthly by the salt
(NaCl) addition and conductivity measurement method, or
electrochemical gauging (May 2006–Aug. 2007). Moreover,
since June 2007, the outlet was equipped with a water level
piezoelectric probe (OTT Orpheus Mini) in order to monitor
the outflow (after calibration of the level data vs. flow deter-
mination). Water level of the lake itself has been monitored
since February 2009 with the same kind of probe.
δ^13 CDIC determinations were performed on gas chroma-
tography– isotope ratio mass spectrometer; details are devel-
oped in Assayag et al. ( 2008 ). In brief, an aliquot of the water
sample is injected in an Exetainer Labco tube purged with a
stream of helium and charged with 1 mL of phosphoric acid.
After overnight equilibration, the gas phase is sampled and
introduced into a gas chromatograph coupled to a mass spec-
trometer (AP 2003). The calibration is made with respect to
the standard of CaCO 3 in the PDB scale.
δ^13 CCH4 and δDCH4: In order to extract CH 4 dissolved in
water, an aliquot of water was loaded through a septum into
an evacuated volume. Water and CO 2 were then trapped atD. Jézéquel et al.