Lake Pavin History, geology, biogeochemistry, and sedimentology of a deep meromictic maar lake

(Chris Devlin) #1

208


a previous study, focused on the methane biogeochemical
cycle in Lac Pavin (Lopes et al. 2011 ), and applied to^14 C
profiles modeling of organic and inorganic carbon pools
(Albéric et al. 2013 ) and is briefly summarized below.
In the model, dissolved and particulate species are trans-
ported through the water column by vertical mixing, but par-
ticles are also transported by sedimentation. Dissolved Fe
and P species are supplied to the lake by inflows from subla-
custrine input. Dissolved and particulate species are allowed
to react in multiple biogeochemical pathways such as oxida-
tion, adsorption, precipitation and dissolution. Two partial
differential transport-reaction equations implemented in the
AQUASIM code were used to describe these biochemical
and physical processes. They were solved for dissolved and
particulate matter, respectively, and can be written as
follow:




= ∂









− ∂

C ++
tAz

AK C
zAz

QC r q
A
zC() Cin



= ∂








−



++∂

X
tAz

AK X
zAz

QX r
z
zX() ()VXsed

where C is the concentration of dissolved species (mol.
m−3), X is the concentration of particulate species (mol.
m−3), t is the time (s), z is the vertical coordinate pointing
downwards (m), A is the cross-sectional area of the lake
(m^2 ), Kz is the vertical mixing coefficient (m^2 .s-1), rC and
rX are the transformation rates of dissolved or particulate
species (mol.m−3.s−1), Q is the vertical discharge induced by
water inflows in the lake depth (m^3 .s−1), Cin is the inflow
concentration of dissolved species (mol.m−3), Vsed is the sed-
imentation velocity (m.s−1), q is the water discharge at depth
into the lake (m^2 .s−1). Both equations were solved numeri-
cally with a vertical resolution of 1 m (Reichert 1998 ). Four
chemical parameters from the various depth profiles obtained
in the water column were considered herein and fitted in our
modeling approach: dissolved Fe and P concentrations, par-
ticulate Fe concentration and Fe isotope composition of the
dissolved Fe(II). In the model, we focused on the processes
occurring in the monimolimnion and at the redox transition
zone, where the majority of Fe is cycled.


12.3 Results


12.3.1 Chemical and Isotopic Compositions
in the Water Column


Dissolved Fraction Figure 12.1 presents chemical and Fe
isotope profiles in a depth range from the redox transition
zone down to the bottom of the lake (data from July 2007
reported in Table 12.1). Dissolved Fe and P concentrations


are low in the oxic zone, respectively <30 nM and ~1 μM,
but increase dramatically below the redox transition zone
(~60 m) up to values of ~1200 and 370 μM at the lake bottom
(Fig. 12.1b). Dissolved Mn concentration also shows a pro-
gressive increase with depth but in a lower range, from 0.3 to
26 μM. Manganese concentration increases at a shallower
depth level in the redox transition zone, with a distinct peak
at ~61 m depth (Fig. 12.1a), suggesting rapid water column
dissolution of Mn oxides. Dissolved sulfate (SO 4 2−) concen-
tration decreases from ~15 μM in the oxic zone to <1 μM in
the anoxic zone, due to sulfate reduction into sulfide (Fig.
12.1a). Total sulfide (ΣH 2 S) in the dissolved fraction (water
filtered at 0.2 μm) corresponds to a limited amount of “free”
dissolved sulfide (H 2 S/HS−) but is dominated by FeS colloids
(about 80 % of the total sulfide; Bura-Nakic et al. 2009 ). Iron
isotope composition shows a large increase with depth
from−2.14 to +0.31 ‰, with the largest variation near the
redox transition zone (Fig. 12.1c).
The relationship between dissolved Fe and P in Lac Pavin
water column is illustrated in Fig. 12.2a. It shows a compila-
tion of data from two different periods of time (September
1992 and July 2007). Dissolved Fe and P concentrations dis-
play a remarkable linear correlation, with a mean slope of
3.45 passing through the origin of the graph. This suggests a
significant connection between Fe and P cycles in the lake,
also supported by suspended particulate matter data (see
below). Although this correlation suggest that Fe/P ratio is
roughly constant in dissolved fraction of water samples at
Lac Pavin, a detailed analysis shows some variability along
depth profile (Fig. 12.2b). The Fe/P molar ratio is very low
(<0.05) in the mixolimnion (0–60 m depth) due to the pres-
ence of oxygen, driving Fe oxidation and precipitation.
Between 60 and 65 m depth, the Fe/P ratio increases strongly
up to values of 5. From 65 m to lake bottom, the Fe/P ratio
decreases progressively from about 3.8 to 3.2 (Fig. 12.2b).

Suspended Particulate Matter Suspended particulate mat-
ters (SPM) collected in October 2010 were analyzed and
compared to turbidity profile. The variation with depth of Fe
concentrations of SPM (expressed relative to the volume of
water that has been collected and filtered) is consistent with
the results of a previous study (Viollier et al. 1995 ). It shows
a good agreement with turbidity variations at the redox inter-
face and in the anoxic zone (Fig. 12.3a). This suggests that
(1) the variations of turbidity around the redox boundary are
largely related to Fe-bearing particles in SPM at depth > 50 m
in Lac Pavin and (2) turbidity measurements can be used as
a proxy of these Fe-rich particles. In detail, we observed
sometimes that Fe particles collected on SPM extend deeper
in the water column than what turbidity measurement would
suggest. It may result from the fact that turbidity is a good

V. Busigny et al.

http://www.ebook3000.com

Free download pdf