You are on page 1of 15

Applied Geochemistry 41 (2014) 34–48

Contents lists available at ScienceDirect

Applied Geochemistry
journal homepage: www.elsevier.com/locate/apgeochem

Deciphering the impact of land-uses on terrestrial organic matter


and mercury inputs to large boreal lakes of central Québec using lignin
biomarkers
Matthieu Moingt ⇑, Marc Lucotte, Serge Paquet, Bassam Ghaleb
GEOTOP, Institut des Sciences de l’Environnement, Université du Québec à Montréal, C.P.8888, suc. Centre-Ville, Montréal, Québec H3C 3P8, Canada

a r t i c l e i n f o a b s t r a c t

Article history: To evaluate watershed impacts of anthropogenic activities on terrestrial organic matter (TOM) and total
Received 17 June 2013 mercury (THg) dynamics in large boreal lake ecosystems, we studied sediment cores retrieved in eight
Accepted 17 November 2013 large lakes of Québec (Canada). Two lakes with pristine watersheds were considered as reference lakes
Available online 1 December 2013
and six lakes with watersheds affected by different types of anthropogenic activities (e.g. logging and/
Editorial handling by JoAnn M. Holloway
or mining activities) were used to illustrate the influence of land-use on TOM and Hg cycling in lakes.
A Geographical Information System (GIS) approach was used to correlate the evolution of anthropogenic
land-uses from 1979 to 2010 (e.g. logging and mining activities) to TOM and THg contents measured in
sediment cores. In each core, THg concentrations gradually increased over the recent years. Using lignin
biomarkers, we noticed that the presence of both intense logging and mining activities in the watershed
does not necessarily correspond to noticeable changes in the relative amount of terrestrial organic matter
(TOM) exported from the watershed to the sediments and by extension to the level of THg measured in
sediments. Apparently large-scale watersheds show some ‘‘buffering’’ capacity to land-use disturbance.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction It is recognized that terrestrial organic matter (TOM) plays a


major role in the transport of contaminants from the watershed
Mercury (Hg) dynamics in lakes in central Québec are of special to the lake (Coquery and Welbourn, 1995; Moingt et al., 2010;
concern as local populations frequently consume wild fish and are Ouellet et al., 2009; Teisserenc et al., 2010). Lake sediments repre-
thus chronically exposed to the contaminant. For over a century, sent a powerful tool to reconstruct both Hg and TOM transfers
the release of Hg to the atmosphere associated with anthropogenic from the watersheds to the aquatic systems. Lignin biomarkers
activities resulted in an increase in the heavy metal deposition in have been used to evaluate both TOM fluxes and quality reaching
boreal lakes ecosystems (Pacyna et al., 2006). Dry and wet Hg sediments (Dittmar and Lara, 2001; Hedges et al., 1997; Hu et al.,
deposition from anthropogenic sources can reach areas very 1999; Ouellet et al., 2009) and lignin degradation biomarkers have
distant from their point source emission (Fitzgerald et al., 1998; been widely exploited for the study of TOM dynamics in soils,
Pacyna et al., 2006), thus remote areas such as central Québec water column, and sediments (Chabbi and Rumpel, 2004; Houel
can then be impacted. Hg is then transferred to the lake and et al., 2006; Ouellet et al., 2009). In comparison with bulk carbon
accumulated in sediments where increases in total mercury analysis, lignin biomarkers in sediment cores bring new informa-
(THg) contents have been reported (Lucotte et al., 1995; Rognerud tion on the nature of TOM fluxes from watersheds to lakes and help
and Fjeld, 2001). In parallel, central Quebec watersheds are to reconstruct historical watershed modifications. Moreover,
frequently perturbed by multiple anthropogenic activities such as because Hg is known to have affinity for TOM (Kainz et al., 2003;
logging, mining, road construction and urbanization. Those pertur- Kolka et al., 1999; Yang et al., 2007), lignin biomarkers could yield
bations modify the natural Hg fluxes from the watersheds to the new information on Hg dynamics in boreal lakes. Lignin alkaline
aquatic systems and could be responsible for supplementary oxidation by copper oxide (CuO) produces short-chain phenols
amounts of Hg reaching the lakes, which have the potential to be comprising ketone, aldehyde and/or carboxylic acid functionalities
incorporated into the food web (Ikingura and Akagi, 1996; Lavoie present in the parent lignin macropolymer. These major oxidation
et al., 2010; Soto et al., 2011; Watras et al., 1998). products are grouped into four major monomer families: vanillyl
(V), syringyl (S), cinnamyl (C) and p-hydroxyphenols (P) (Goñi
and Hedges, 1992; Goñi et al., 1993; Hedges and Ertel, 1982).
⇑ Corresponding author. Tel.: +1 514 987 3000x3523; fax: +1 514 987 3635. The relative proportions of these monomeric subunits allow dis-
E-mail address: matthieumoingt@yahoo.fr (M. Moingt). crimination between different types of vascular plant tissues, and

0883-2927/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.apgeochem.2013.11.008
M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48 35

have been used to decipher sources and relative degradation state 2.2. Landscape and historical analysis
of organic matter in several natural ecosystems (Dittmar and Lara,
2001; Hedges and Mann, 1979; Hernes and Benner, 2003). More- Landscape analyses were performed following the method de-
over, several studies have shown that lignin quality is influenced scribed in details by Moingt et al. (2013). Briefly, landscape charac-
by hydrologic and soil processes in watersheds (Dalzell et al., terization was performed with raster satellite images from Landsat
2007; Hernes et al., 2007). 7 satellite imagery with image processing using GIS Geographic
In this paper, we combined the geochemical study of sediment Resources Analysis Support System (GRASS 6.4; http://grass.-
data (210Pb dating, THg concentrations, lignin biomarkers) with fbk.eu/) software in order to decipher the morphology and the land
watershed characteristics identified through GIS (land-use and cover of watersheds. The Canadian digital elevation data (scale: 1/
vegetation cover) to evaluate the impact over three decades of 50,000), extracted from the hypsographic and hydrographic ele-
anthropogenic watershed perturbations relative to the increase in ments of the National Topographic Data Base (Tahvanainen and
Hg transfers from watersheds to sediments in large lakes ecosys- Haraguchi, 2013) were also used. For the purposes of our study,
tems. We sampled sediment cores in eight large boreal lakes with the watershed is defined as the land area draining either directly
surface areas ranging from 15 to 210 km2, six of them presenting into a lake or into the first lake upstream corresponding to an order
specific disturbances in their watershed such as intensive logging 3 watershed level which has been described as the level influenc-
or mining, and two unperturbed lakes chosen as reference lakes. ing the most the associated aquatic ecosystem (Beaulne et al.,
Several studies point to anthropogenic perturbations occurring in 2012; Slauenwhite and Wangersky, 1996). In order to determine
the watershed (e.g. logging and mining activities) as a contributor the mean watershed slope and to identify different slope classes
to the increasing amount of THg reaching lakes sediments. How- (intervals of 2% from 0% to 20%, e.g. 0–2%) within each watershed,
ever, these studies have focused solely on small lakes ecosystems the slope of each pixel was calculated. Land cover was estimated
with surface areas ranging from 0.4 to 1.3 km2 (Garcia and Cari- by applying the Maximum Likelihood Classification (MLC) algo-
gnan, 2000, 2005; Porvari and Verta, 2003; Porvari et al., 2003). rithm modified by Neteler and Mitasova (2008) to the entire set
Large lake ecosystems have been ignored despite the fact that they of visible spectral bands. Historical reconstructions were based
represent a better picture of socio-economic-environmental land- on observations coming from satellite images over a 30 year period
uses of the boreal forest region (fishery, mining and logging activ- (1979–2011).
ities). Moreover, small lakes only allow the study of one process
one at a time whereas large lake ecosystems integrate a sum of 2.3. Sediment sampling
various processes and include the inertia of the system. Such ap-
proaches will help to gain a more holistic understanding of the cy- We used a pneumatic Mackereth corer to sample lake sedi-
cling of Hg between terrestrial and aquatic systems and the ments during sample missions in 2009 and 2010. This type of corer
influence of anthropogenic effects and environmental changes in presents the advantage of reducing both perturbations at the
the transport of Hg into aquatic ecosystems. water–sediment interface and sediment compaction. During the
sampling, a 1.5-m long Plexiglas tube (diameter 10 cm) is slowly
pushed into the sediment using compressed air. Then, sediment
2. Material and methods cores were sub-sampled in 20 mL glass vials (pre-combusted at
500 °C for 3 h to avoid carbon contamination of the vessels and
2.1. Study sites capped with TeflonÒ liners) every centimeter using a TeflonÒ spat-
ula. For each slice we only conserved the center of the core to avoid
Matagami, Ouescapis, Rodayer, Waswanipi, Dickson and cross-contamination between samples. Finally, samples were
Chibougamau lakes are located in the administrative region of freeze-dried prior to analysis. In order to integrate spatial varia-
Northern Quebec (Fig. 1). Lakes characteristics (location, altitude, tions in watershed to lake TOM and Hg transfers, coring was per-
lake area, drainage area, ratio of drainage area to lake area (DA/ formed at the focal point of the lake (Hakanson and Jansson,
LA)) are specified in Moingt et al. (2013). Matagami Lake is a 1983; Teisserenc et al., 2010).
43 km long and 14 km wide lake in a swampy region of Northern
Quebec and represents the confluence point of Allard, Bell, Goua- 2.4. Chemical analyses
ult and Waswanipi rivers. Perturbations in the watershed include
road construction, logging, mining and golf activities. Ouescapis 2.4.1. 210Pb radiometric measurements
and Rodayer lakes are located approximately 70 and 120 km Sedimentation rates were determined by radiometric measure-
north of Matagami Lake, respectively. None of these lakes show ment of 210Pb activity using alpha spectrometric measurement of
any substantial perturbations in their watershed. Waswanipi Lake the activity of the daughter product 210Po (Flynn, 1968) and
is situated about 100 km south-east from Matagami Lake to assuming secular radioactive equilibrium between the two iso-
which it is linked upstream by the Waswanipi River. Waswanipi topes. Aliquots between 0.2 and 0.4 g of sediments were spiked
Lake presents a watershed perturbed with logging and mining with 209Po yield tracer and digested in Teflon vials using a concen-
activities and several gravel roads. Dickson Lake is located about trated 5:4:1 mixture of HNO3:HCl:HF. Addition of H2O2 was per-
180 km south-east from Matagami Lake. This lake has a history of formed to eliminate organic carbon (OC) in the sample. The
intense logging activity in its watershed. Chibougamau Lake is lo- residue was then converted to a chloride salt by repeated evapora-
cated approximately 10 km south of the town of the same name. tion with 6 M HCl. Before the last evaporation was complete, the
This lake is the source of the Chibougamau River and is inter- remaining solution was centrifuged and the supernatant was
spersed with several islands, deep and large bays. The watershed collected in order to get rid of black carbon. Then, approximately
of Chibougamau Lake has been intensely used for mining activi- 100 mL of 0.5 M HCl was added to the collected supernatant to de-
ties (Petit et al., 2011). Des Jardins East and West lakes (total area crease the concentration of the solution. Then 1.5 mL of 1 M ascor-
of 15.02 km2) are located in the Témiscamingue region further bic acid was added to reduce Fe3+ into Fe2+. Po isotopes where
south (Fig. 1). Des Jardins East Lake flows into Des Jardins West deposited on a spinning Ag disk (Hamilton and Smith, 1986) and
Lake and their watersheds are covered by mature mixed wood the activity measured by a-spectrometry using ORTEC silicon
forests. Perturbations in the watershed include gravel roads and surface barrier detectors coupled with a PC running under Maestro
logging activities. data acquisition software. Blanks were run through the same
36 M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48

Fig. 1. Location of studied lakes on a Quebec map.

procedure. Chronology was established using downcore unsup- nitrogen analyses using a CE-Instruments model NC2500™ ele-
ported 210Pb, a constant initial concentration model (CIC), and mental analyzer with a relative precision of ±5% (1r). Replicate
dividing nonlinear profiles into linear segments assuming constant measurements on samples, treated and untreated by vapor acidifi-
sedimentation for each of these segments (Appleby and Oldfield, cation, were performed to verify that the inorganic carbon fraction
1992). 210Pbex corresponds to the excess of 210Pb in sediments was negligible. Acidified sample measurements systematically fell
(i.e. the difference between the total 210Pb activity at depth (i) within the range of the non-acidified sample, thus demonstrating
minus the value of 210Pb supported activity). 210Pb supported activ- that total carbon content can be considered as OC content.
ity corresponds to the average of measured values at the bottom of Sediment samples were subjected to alkaline cupric oxide (CuO)
the core when 210Pb activity does not vary anymore. oxidation according to a modified protocol from the initial method
described by Goñi and Montgomery (2000). For each sample,
2.4.2. Total mercury analyses 3.0 ± 0.1 mg of OC and CuO (330 ± 4 mg) were mixed in about
Replicate THg analyses were performed by cold vapor atomic 3.2 mL of 2 N NaOH in a reaction bomb purged with N2. Oxidations
fluorescence spectrometry (CV-AFS) following the protocol devel- were conducted with a Hewlett Packard 5890A™ gas chromato-
oped by Bloom and Fitzgerald (1988) and adapted by Pichet et al. graph oven modified by PRIME FOCUS Inc. (Seattle, WA) and fitted
(1999). Briefly, a combination of 16 M HNO3:6 M HCl (10 mL:1 mL) with 12 Monel-400 alloy (Ni:Cu:Fe 65:32:3 wt%) reaction vessels.
is added to approximately 250 mg of ground, freeze-dried sedi- Reaction bombs were heated to 150 °C for 150 min with an initial
ment and then heated to 120 °C for 6 h. The remaining solution temperature gradient of 4.2 °C/min for 30 min. After oxidation,
is brought back to a volume of 30 mL with NANOpureÒ water 50 lL of the internal standard cinnamic acid and ethyl-vanillin
and analyzed by atomic fluorescence. Hg calibration was done by was added to the bomb and the supernatant was decanted and
injecting known quantities of Hg (II) (400–1000 pg of Hg). The acidified to pH 1 with HCl (2 N). The organic phase was then li-
detection limit for a 250 mg sample was 0.1 ng/g. The accuracy quid–liquid extracted using ethyl acetate and dried by rotary evap-
of the method was verified using the Mess-3 certified standard oration. Finally, the extract was dissolved another time in pyridine
(NRC Canada). With an average value of 87 ± 3 ng/g for seven ali- and derivatized with N,O-bis(trimethylsilyl)trifluoroacetamide
quots, our results fell within the certified value (92 ± 9 ng/g). (BSTFA) and trimethylchlorosilane (TCMS; 99:1). A 2-lL extraction
fraction was analyzed by GC/MS (Varian 3800/Saturn 2000™)
2.4.3. Lignin biomarkers analyses fitted with a fused capillary column (DB-1 from J&W, 60 m,
In order to conduct lignin biomarkers analyses, percentage of 320 lm). The injector and the detector were held at 300 °C and
OC for each sample was measured performing total carbon and He was used as carrier gas in splitless mode. The initial column
M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48 37

Table 1
Comparison and variability of lignin CuO oxidation parameters measured from a reference sediment: SAG 05 (estuarine sediment). Values from Louchouarn et al. (2000) and
Houel et al. (2006) were generated using conventional oven and microwave oven respectively.

SAG 05
This study (n = 9) Louchouarn et al. (2000) (n = 11) Houel et al. (2006) (n = 6)
Average SD Average SD Average SD
Lambda 3.50 0.23 3.53 0.17 2.81 0.31
Sigma 8 7.41 0.48 ND ND ND ND
S 0.70 0.20 0.56 0.05 0.44 0.03
V 2.57 0.29 2.73 0.11 2.11 0.29
C 0.23 0.04 0.23 0.04 0.26 0.04
P 0.30 0.03 0.54 0.10 0.34 0.04
S/V 0.28 0.11 0.21 0.02 0.21 0.03
C/V 0.09 0.02 0.08 0.01 0.13 0.02
P/(V + S) 0.09 0.01 ND ND 0.14 0.02
3,5-Bd/V 0.10 0.02 0.08 0.02 0.11 0.01
(Ac/Ad)S 0.20 0.07 0.38 0.05 0.47 0.13
(Ac/Ad)V 0.37 0.03 0.38 0.05 0.48 0.09

oven temperature was set at 100 °C with a temperature gradient of (Fig. 4). Indeed, we can observe an inflexion point around 6 cm
4 °C/min to 320 °C followed by a holding time of 10 min. depth where the sedimentation rate rapidly shifts from 0.11 to
Replicate analyses of the SAG 05 ‘‘standard’’ sediment sample 0.31 cm/year. Due to the unusual 210Pb profile in Matagami Lake’s
(n = 9) showed that the analytical variability of the alkaline CuO sediment core, 137Cs measurements have been performed in order
oxidation products and major indicators range from 6% to 38% (Ta- to verify the estimated sedimentation rates in this lake. Fig. 4 pre-
ble 1). The intercomparison of this standard sediment analyses sents 137Cs activity and 210Pb age previously estimated vs. depth in
shows that the composition parameters and the yields of the pres- the sediment core. 137Cs is an anthropogenic radionuclide which
ent study are comparable to those obtained by both microwave has been released in important amounts because of nuclear tests.
137
digestion and traditional oven method (Table 1). Cs started to fallout on land surfaces circa 1952 (Davis et al.,
1984) and peaks of 137Cs are generally observed circa 1963 in sed-
2.4.4. Statistical analyses iments cores (Ali et al., 2008). In Fig. 4, we notice that the peak of
137
Statistical tests were applied to the regression observed Cs corresponds to a 210Pb age of approximately 45 years which
between the enrichment in THg from the deeper sections of the is in agreement with the date of 1963. Thus, 137Cs confirms the
cores to the surface ones. Such tests were also applied to the mor- ages previously estimated by 210Pb dating.
phological characteristics of the watersheds. In all cases, normal Based on sedimentation rates estimated by 210Pb dating, a
distribution of residuals was tested by performing a Shapiro–Wilk straight horizontal line symbolizes an age of 100 years in each sed-
goodness of fit test, using a statistical analysis program (JMP 7) and iment core in Figs. 5, 6, 10 and A1 and A2 from the Appendix A.
accepted W values >0.05. The F-ratio was then computed in order
to evaluate the effectiveness of the model and the student para- 3.2. Spatial, historical analysis of lakes and drainage areas and THg
metric test (t-test) was processed to evaluate the pertinence of contents in lakes sediment cores
regression parameters (p < 0.05).
Spatial and historical analysis of lakes, drainage areas and THg
3. Results contents in lakes sediment cores of the present study have been pre-
viously performed, measured and described by Moingt et al. (2013).
3.1. 210
Pb dating Data from Moingt et al. (2013) summarized the principal bio-mor-
phological parameters and THg results of the studied lakes are pre-
Figs. 2–4 present the results for radiometric measurements sented respectively in Tables A1 and A2 from the Appendix A.
realized on lake sediment cores. In surface sediments, 210Pb Based on 210Pb dating, the increases in THg in reference lakes
activities are 32, 42, 18, 37, 14 and 15 dpm/g for Ouescapis, Roda- correspond to about the start of the industrial era (Appendix A,
yer, Waswanipi, Chibougamau, Dickson and Matagami lakes Fig. A1). The Hg anthropogenic sedimentary enrichment factor
respectively. (ASEF) values come to 2.7 and 2.8 for Ouescapis and Rodayer lakes
Water content of each sample was determined by weighing respectively. These ASEF values are similar to those frequently re-
sediment before and after freeze-drying in order to correct the sed- ported in the literature for pristine North American lakes of the
iment accumulation rate because of possible compaction of the boreal domain (Lindberg et al., 2007; Lucotte et al., 1995).
core during sampling. After calculation, sedimentation rates ob- Three out of six lakes (Des Jardins East, Des Jardins West and
tained taking or not into account dry bulk density of each sediment Waswanipi) with perturbed watersheds are characterized by Hg
layer were not different suggesting a minor impact of compaction ASEF values inferior to 3 since the beginning of the industrial era.
in our case. Thus, sedimentation rates were directly estimated i.e. similar to those of the two pristine lakes (Appendix A,
using the slope of the relationship between depth (cm), the natural Fig. A2). The three remaining lakes (Chibougamau, Matagami and
logarithm of 210Pbex and the 210Pb decay constant. Dickson) present much higher Hg ASEF values (3.8, 5.4 and 15
Desjardins Lake has an estimated sedimentation rate of respectively; Appendix A, Table A2).
0.22 ± 00.6 (Lucotte et al., 1995; Teisserenc et al., 2010). Estimated
sedimentation rates are 0.11, 0.05, 0.05, 0.03, 0.05 and 0.31 cm/ 3.3. Lignin biomarkers measurements
year for Ouescapis, Rodayer, Waswanipi, Chibougamau, Dickson
and Matagami lakes respectively. Only Matagami Lake seems to Lambda and Sigma 8 indicators are commonly used to estimate
present a modification of its sedimentation rate through time the relative amount of TOM in sediments (Houel et al., 2006;
38 M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48

Fig. 2. Total 210Pb activity (full black rhombs) and natural logarithm of unsupported 210Pb activity (white rhombs) vs. depth in reference lakes sediment cores (i.e. Ouescapis
and Rodayer lakes).

Teisserenc et al., 2011). Lambda corresponds to the sum of vanillyl, the exception of Matagami Lake, which present important Lambda
syringyl and cinnamyl phenols per 100 mg of OC whereas Sigma 8 variations through time, all the lakes have Lambda values that are
represents the same sum of compounds but normalized for 10 g of relatively stable during the last 100 year period (Fig. 7).
sample. Thus, Lambda corresponds to the relative amount of terrig- S/V ratios are generally used to estimate the contribution of
enous OM with respect to the total OC content of a sediment sam- angiosperm species to TOM contents in lake sediments (Hedges
ple and Sigma 8 corresponds to the relative amount of terrigenous and Mann, 1979; Tesi et al., 2008). Mean S/V ratio values vary from
OM to a sediment sample. Fig. 5 presents Lambda and Sigma 8 0.22 ± 0.07 to 0.39 ± 0.02 (Fig. 8). P/(V + S) ratio is used as indicator
indicators profiles for reference lakes (i.e. Ouescapis and Rodayer of the degradation state of TOM (Dittmar and Lara, 2001; Opsahl
lakes) in the first 40 cm of the sediment cores. Lambda profiles and Benner, 1995) since demethoxylation leads to the selective
are similar for Ouescapis and Rodayer lakes with values ranging loss of OCH3 group in S and V families whereas p-hydroxyphenols
from 0.71 to 1.51 mg/100 mg of OC and from 0.68 to 1.76 mg/ are not affected (Hedges and Ertel, 1982). 3,5-Bd/V is an indicator
100 mg of OC respectively. Sigma 8 ranges from 1.71 to 3.76 mg/ of OM maturation in soils (Houel et al., 2006; Prahl et al., 1994).
10 g of sample and from 1.17 to 4.08 mg/10 g of sample for Oues- Mean P/(V + S) values range from 0.16 ± 0.05 to 1.07 ± 0.08
capis and Rodayer lakes respectively. whereas mean 3,5-Bd/V values are comprised between
Fig. 6 presents Lambda and Sigma 8 indicators profiles in lakes 0.14 ± 0.06 and 0.70 ± 0.15. Fig. 9 shows that when all lakes are
presenting various types of perturbations in their watersheds. Con- considered together, P/(V + S) and 3,5-Bd/V ratios are not corre-
sidering Lambda values, only Matagami Lake shows major devia- lated (r = 0.53; p = 0.2825). If Chibougamau Lake is excluded, we
tion from the 0.5 to 2.0 mg/100 mg of OC range observed in the then find a strong correlation between those two ratios (r = 0.95;
two reference lakes with values as high as 7.94 mg/100 mg of OC. p = 0.0150).
On the other hand, Sigma 8 values present a wide range of values Lignin phenol results are summarized in Table A3 in the Appen-
comprised between 0.02 and 48.14 mg/10 g of sample (Fig. 6). dix A.
In Figs. 7–9, we present the mean values of lignin biomarkers
and the corresponding standard deviation for the last 100 year per- 4. Discussion
iod, estimated using 210Pb dating, of each sediment core. Mean
Lambda values are comprised between 0.77 ± 0.07 and Although sampling one core per site at the focal point of a given
3.59 ± 2.10 mg/100 mg of OC (Fig. 7) which is in agreement with large lake does not allow observing site-specific inhomogeneities,
values reported for large lake sediments in the literature (Hyodo it remains an efficient way to obtain a global signal integrating
et al., 2008; Orem et al., 1997; Petit et al., 2011; Tareq et al., both TOM and Hg dynamics at the watershed scale (Hakanson
2011). Reference lakes present the lowest Lambda values. With and Jansson, 1983; Moingt et al., 2013; Teisserenc et al., 2010).
M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48 39

Fig. 3. Total 210Pb activity (full black rhombs) and natural logarithm of unsupported 210Pb activity (white rhombs) vs. depth in Dickson, Chibougamau and Waswanipi lakes
sediment cores.

4.1. Lambda and Sigma 8 indicators event. In each perturbed lake, Lambda and Sigma 8 core profiles are
similar suggesting that the variations observed for the Lambda
In reference lakes, the relative stability of Lambda along the indicator are due to variations in TOM fluxes coming from the wa-
sediment cores (Fig. 5) indicates that little changes occurred in tershed and not to fluctuations of the primary production into the
the ratio of allochthonous vs. autochthonous organic matter over water column of the lake (Houel et al., 2006; Rezende et al., 2010).
time. The magnitude of variation in Sigma 8 along the core profiles
is higher than that of Lambda (Fig. 5). Given that these two indica- 4.2. Impact of mining and logging activities on TOM fluxes: coupling
tors have the same trends along the profiles, their variations are lignin biomarkers and a GIS approach
unlikely due to variations in the amount of inorganic matter com-
ing from the watershed but rather due to changes in TOM fluxes. Using a GIS approach allowed us to confirm that the two refer-
Contrary to sediment cores of reference lakes, Lambda profiles ences lakes have not been intensely exposed to logging and/or
in sediment cores of Matagami and Dickson lakes present marked mining activities since 1979 whereas Waswanipi, Chibougamau,
variations (Fig. 6) suggesting shifts in the ratio of allochthonous vs. Matagami, Dickson and Des Jardins lakes watersheds have been
autochthonous sedimentary organic matter. Matagami Lake’s sed- widely logged and/or used by mining industries over the same per-
iment core presents sporadic and important variations of the iod of time (Moingt et al., 2013).
Lambda indicator suggesting that specific events induced tempo-
rary variations in sedimentary OM fluxes trough the time. Because 4.2.1. Impact of logging activities
those changes are observed before and after the onset of the indus- Lignin biomarkers in lake sediments reflect the nature of vege-
trial era, we cannot attribute those modifications to either natural tation cover and soil types in the lake watershed and thus can be
events or anthropogenic activities over the last 100 years. The used to identify the impact of logging activities on TOM fluxes
Lambda profile of Dickson Lake shows a modification of the indica- reaching the lakes (Brandenberger et al., 2011; Gordon and Goñi,
tor occurring at 20 cm depth (Fig. 6) which, by extrapolation of the 2003; Hedges and Mann, 1979; Louchouarn et al., 1999). Sedi-
sedimentation rate, would correspond to the year 1600. This mod- ments impacted by logging activities are characterized by higher
ification is sudden and permanent suggesting an important salient Lambda values than those of pristine lakes because of increased
event, which modified the sedimentary regime of OM in the lake. transfers of TOM from watersheds to aquatic systems in deforested
Similar observations have been reported in other lakes (Hyodo areas (Brandenberger et al., 2011; Loh et al., 2012). Fig. 7 presents
et al., 2008; Louchouarn et al., 1993; Teisserenc et al., 2013). More- the mean Lambda values measured in sediments covering the last
over, because this event happened before the European settlement 100 year period for the studied lakes. Des Jardins E. and W. lakes
of this region, we can assume that it corresponds to a major natural are not considered because the difference in latitude between
40 M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48

Fig. 4. Total 210Pb activity vs. depth (A) and natural logarithm of unsupported 210Pb activity vs. depth (B) in Matagami Lake. 137Cs activity vs. depth (C) and 137Cs activity and
210
Pb age vs. depth (D) in the sediment core of Matagami Lake.

Fig. 5. Lambda (full black rhombs) and Sigma 8 (white diamonds) indicators vs. depth in reference lakes. Error bars are not visible at this scale. The straight black lines
represent an age of 100 years for each core based on 210Pb dating.

those lakes and reference lakes could explain a difference in the re- the part of the core older than 100 years (Fig 6). Those observations
gime of the lakes as suggested by the high Sigma 8 values encoun- suggest minor impact of logging activities on TOM fluxes at the
tered for Desjardins Lakes (Fig. 6). Sediments of Waswanipi and watershed scale despite the fact they have been intensely impacted
Chibougamau lakes present mean Lambda values over the last by logging activities since 1979 (Moingt et al., 2013). In contrast,
100 year period similar to those measured in reference lakes sediments of Dickson and Matagami lakes display mean Lambda
(Fig. 7) and not different from the Lambda values measured in values over the last 100 years significantly higher than those of
M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48 41

Fig. 6. Lambda (full black rhombs) and Sigma 8 (white diamonds) indicators vs. depth in lakes presenting various types of anthropogenic perturbations occurring in their
watersheds. Error bars are not visible at this scale. The straight black lines represent an age of 100 years for each core based on 210Pb dating.

reference lakes suggesting at first that logging activities in these In Matagami Lake we can observe strong variations in the
lakes may have increased TOM fluxes reaching the lake. However, Lambda values both in recent sediments and in sediments older
in Dickson Lake, there is no difference between measured Lambda than 100 years. In the last 100 years, we can observe two major
values in recent sediments and in sediments older than 100 years peaks of Lambda at ten and three centimeters. The beginning of
once again suggesting a minor impact, at the watershed scale, of the deeper peak corresponds to an event older than 100 years
logging activities on the amount of TOM reaching lake sediments. and then could not be related to logging activities which start in
42 M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48

Fig. 9. Mean P/(V + S) ratio vs. mean 3,5-Bd/V ratio values and associated standard
deviation measured in sediments covering the last 100 year period (full black
Fig. 7. Mean Lambda values and associated standard deviation measured in rhombs) and the period before the industrial era (white squares) for the lakes of the
sediments covering the last 100 year period for the lakes of the study based on 210Pb study based on 210Pb dating.
dating. Gray area distinguishes reference lakes from the others lakes of the study.

ratios of pure angiosperm sources (Goñi and Hedges, 1992; Hedges


and Ertel, 1982; Hedges and Mann, 1979), suggesting a significant
contribution of gymnosperm species. However, Teisserenc (2009)
showed that a comparison of S/V ratio in sediments to S/V ratio
in soils would be a better tool for historical reconstruction of ter-
restrial inputs to aquatic systems since soils are a better integrator
of vegetation cover diversity, degradation and adsorption pro-
cesses affecting TOM composition (Hernes et al., 2007; Louchouarn
et al., 1999; Prahl et al., 1994; Sánchez-García et al., 2009). Our
mean S/V ratio values for the last 100 year period are in agreement
with the values reported by Teisserenc (2009) in boreal upper soils
suggesting then that the organic horizon of soils is the major con-
tributor of sedimentary TOM. It is also suggesting that in boreal
ecosystems, although the landscape is dominated by gymno-
sperms, angiosperms are important contributors to sedimentary
TOM since gymnosperms do not produce syringyls (Hedges and
Mann, 1979; Hedges and Parker, 1976). Unfortunately, because of
the strong variability of S/V values in the sediment core, this lignin
indicator cannot be used to confirm that Matagami Lake sediments
are actually significantly impacted by logging activities occurring
in their watershed.
Fig. 8. Mean S/V ratio values and associated standard deviation measured in P/(V + S) and 3,5-Bd/V ratios are used to estimate the level of
sediments covering the last 100 year period for the lakes of the study based on 210Pb degradation and the degree of maturation of TOM respectively
dating. Gray area distinguishes reference lakes from the others lakes of the study. (Dittmar and Lara, 2001; Houel et al., 2006). Moreover, 3,5-Bd/V ra-
tios could be used to distinguish TOM from organic and inorganic
soil horizons (Houel et al., 2006). Fig. 9 presents mean P/(V + S)
the middle of the 1960s in the Matagami region. The beginning of and 3,5-Bd/V ratios measured in sediments over the last 100 year
the second peak corresponds to the nineties which can be period for the lakes of the study. Chibougamau Lake probably does
associated with the most intense period of forest cutting in the not fit into the regression observed in Fig. 9 for two reasons: (i)
Matagami Lake watershed based on GIS results (Moingt et al., p-hydroxyphenols amount measured in sediments over the last
2013). Then, due to the high natural variability of the Lambda val- 100 years is similar to the one measured in the other sediment
ues, it is not possible to definitively relate any peak to anthropo- cores, and (ii) both syringyl and vanillyl contents are the lowest
genic activities. compared to the other lakes cores examined in this study
Logging activities are known to trigger the lixiviation of TOM (0.06 ± 0.04 and 0.24 ± 0.04 mg/100 mg of OC respectively). Chibo-
normally found in soils (Farella et al., 2001). Transfers of TOM from ugamau Lake has the lowest sedimentation rate of all the lakes of
logged watersheds to aquatic systems are then characterized by this study which could explain really low levels of syringyl phenols
lower S/V signatures, characteristic of humified TOM after pedo- since they are more sensitive to degradation processes than the
genesis (Teisserenc, 2009). Sediments in lakes impacted by logging others phenol families (Houel et al., 2006; Opsahl and Benner,
activities in their watersheds should then also record lower S/V ra- 1995). Both Matagami and Dickson lakes present significantly low-
tios than those of pristine lakes. Fig. 8 presents the mean S/V ratio er P/(V + S) and 3,5-Bd/V ratios suggesting that sedimentary TOM
values measured in sediments covering the last 100 year period for in those lakes comes from the upper soil organic horizon (Houel
the lakes of the study. Those values are low when compared to S/V et al., 2006; Teisserenc, 2009), which is in agreement with the fact
M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48 43

that draining processes are more superficial in clear-cut land in from the watershed to the water column (Fraser et al., 1999; Klee-
comparison to natural forest cover (Farella et al., 2001). However, berg et al., 2008; Sharpley et al., 2001). These additional nutrient
despite this observation, in the case of large lakes boreal ecosys- loads then enhance the production of aquatic biota. A change in
tems, at a watershed scale our data suggest that logging activities the primary production of the lake can be detected by the study
globally does not seem to have a noticeable impact on sedimentary of the Lambda indicator since it is proportional to the allochtho-
TOM signatures and fluxes. Then large-scale watersheds show nous/autochthonous ratio of OM in a sediment sample. Over the
some ‘‘buffering’’ capacity to logging disturbances since about last 100 year period, only Matagami Lake presents variations in
30% of cut in a watershed is needed to start showing an impact the Lambda indicator (Fig. 6). In this case, Lambda variations along
on sedimentary TOM (Moingt et al., 2013). Our study is based on the core cannot be associated with changes in primary production
large lakes with large watersheds (watershed areas comprised be- as explained previously in Section 4.1. The onset of mining activi-
tween 58 and 1084 km2), then the system’s inertia is most likely ties in the watershed could also induce a change in TOM sources.
responsible for this ‘‘buffering’’ capacity since it increases the First, excavation processes bring to the surface important amounts
amount of time needed for the perturbation to reach the system. of TOM from deeper soil horizon with lignin signatures impacted
Indeed, during the transfer of organic matter from the terrestrial by pedogenesis (i.e. high P/(V + S) and 3,5-Bd/V ratios). Second,
system to the aquatic system, several processes could contribute mining activities enhance drainage and erosion rates of surface soil
to mask and/or dilute the signal of the perturbation (e.g. OM deg- horizons (Mäkinen et al., 2010) characterized by less degraded
radation, OM incorporation into soils, fractionation of the TOM TOM (i.e. low P/(V + S) and 3,5-Bd/V ratios). However, studied lakes
pool). with mining activities in their watersheds do not present any
change of the P/(V + S) vs. 3,5-Bd/V signatures from the pre-indus-
4.2.2. Impact of mining activities trial period to present (Fig. 9). Biomarkers data then suggest that in
The presence of mine tailing in the watershed normally in- large lakes with large watersheds, the impact of mining activities
creases the amount of inorganic component of sediment reaching on TOM fluxes is negligible when considering sediment cores inte-
the lake (Petit et al., 2011). This increase should then induce a de- grating the whole watershed.
crease in the Sigma 8 in lake sediments by diluting the TOM signal
in the sediment sample. However, Fig. 6 shows that for the four 4.3. Combined transfers of TOM and Hg from watersheds to lakes
lakes presenting mining activities in their watersheds, Sigma 8 is
210
stable or slightly increases. Moreover, soil erosion normally trig- Pb dating allows us to confirm that for all the lakes of the
gers higher rates of transfer of dissolved and particulate nutrients present study the significant increase in THg contents in sediment

Fig. 10. THg (ng/g) and Sigma 8 (mg/10 g of sample) vs. depth in Matagami and Dickson lakes sediment cores. The straight black line represents an age of 100 years based on
210
Pb dating.

Fig. A1. Total mercury contents vs. depth in reference lakes (i.e. Ouescapis and Rodayer lakes). The straight black lines represent an age of 100 years for each core based on
210
Pb dating. Figure modified from Moingt et al. (2013).
44 M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48

Fig. A2. Total mercury contents vs. depth in lakes presenting various type of anthropogenic perturbations in their watershed. The straight black lines represent an age of
100 years for each core based on 210Pb dating. Figure modified from Moingt et al. (2013).

coincide with the beginning of the industrial era (Appendix A, In Dickson Lake, an important increase in the Sigma 8 indicator
Figs. A1 and A2). at 20 cm depth is observed and corresponds to a sharp increase in
THg contents increase in the last 100 years is not necessarily THg in sediments (r = 0.87, p < 0.0001). An extrapolation of 210Pb
associated with an increase in the Lambda and Sigma 8 indicators dating results allows us to situate this event 400 years ago and
(Figs. 5 and 6), which would indicate an increase in TOM inputs thus conclude that this anomaly corresponds to a natural event.
reaching lake sediments. Then, the results of the present study Moreover, after this event, both Sigma 8 and THg did not come
show that in large boreal lake ecosystems there is no clear evi- back to their initial levels suggesting a drastic change in the wa-
dence of an impact of land-uses occurring in the watershed on both tershed lake regime after natural flooding or landslide for example.
sedimentary TOM and THg contents. This lack of correlation be- Teisserenc et al. (2013) observed the same pattern in flooded lakes
tween the evolution of Lambda and Sigma 8 indicators and THg for both Lambda and THg in sediment cores and concluded that
contents along a sediment core could be due to the fact that the this pattern was also evidence of a close association between
amount of TOM naturally transferred to the lake is widely suffi- TOM and THg loading. Dickson Lake is also a good example show-
cient to scavenge and bind the amount of Hg being deposited from ing the importance of sedimentation rates to better understand the
the atmosphere even with this amount increasing over the last Hg cycle in boreal lakes ecosystems. In this lake, the ratio between
150 years due to the industrial era while TOM fluxes remain rela- THg baseline concentration and THg peak at the surface is 15 (i.e.
tively constant. However, this observation is not valid for Mata- ASEF = 15). However, knowing that most of the THg increase ob-
gami and Dickson lakes (Fig. 10). served in the sediment core is pre-industrial era, probably due to
Considering Matagami Lake, sporadic variations in the Sigma 8 a natural change of the regime of the lake, for a representative ASEF
indicator are observed before and after the onset of the industrial we have to change the THg baseline content previously considered.
era. Before that era, Sigma 8 variations are not correlated to THg The new ASEF becomes 3.2, which is more typical of other North
fluctuations whereas over the last century, trends in Sigma 8 and American lakes reported in the literature (Kamman and Engstrom,
THg contents do correlate (r = 0.93, p = 0.0187). This pattern sug- 2002; Lucotte et al., 1995; Swain et al., 1992).
gests that THg peaks measured in sediments over the last 100 year
period are due to transfers of recently deposited atmospheric Hg
from the watersheds to the lakes. Moreover, because an increase 5. Conclusion
in TOM fluxes induces an increase in THg contents in sediments,
any contemporary watershed perturbation (natural and/or anthro- Over the last decade, several publications based on the study of
pogenic) influencing TOM fluxes will affect sediment THg levels. small lakes point to anthropogenic perturbations occurring in the
M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48 45

Table A1
Biogeomorphological characteristics of the lakes of the present study as presented in Moingt et al. (2013). %TWA = percentage of the total watershed area, Gr = gravel roads,
L = logging activities, U = urbanization, G = golf.

Lake Localization Altitude Lake area Drainage DA/LA Mean Year Water Coniferous Mixed forest, Unforested Number of Major
lat./long. (m) (km2) area (km2) slope (%) (%TWA) (%TWA) deciduous >50% (%TWA) mines in the disturbances
(TWA%) watershed
Des Jardins 46°390 N 320 15.02 207.39 13.81 5.85 2001 9.62 0.00 33.66 3.57 x Gr, L
78°150 W 2004 10.04 0.00 18.46 0.37
2010 9.56 0.00 43.36 2.27
Ouescapis 50°150 N 278 36.41 107.94 2.96 4.84 1979 0.00 0.00 7.01 2.35 x Gr
76°600 W 1990 0.00 0.00 1.80 1.80
2010 0.02 0.00 5.12 13.98
Rodayer 50°510 N 249 22.31 57.7 2.59 1.34 1979 1.93 0.00 13.94 3.65 x Gr
77°410 W 1990 2.12 0.00 5.33 0.95
2010 2.43 0.00 3.27 2.44
Waswanipi 49°330 N 267 200.69 1483.66 7.39 2.02 1979 1.41 0.00 12.05 7.35 3 Gr, L
76°270 W 1990 0.65 0.00 5.47 20.38
2010 1.36 0.00 7.16 15.61
Chibougamau 49°500 N 378 201.06 769.87 3.83 4.45 1999 8.68 0.00 18.66 16.31 14 Gr, L, U
74°130 W 2002 8.76 0.00 21.13 5.82
2009 8.61 0.00 19.17 10.18
Dickson 49°380 N 342 10.00 227.59 22.76 3.23 1979 0.85 0.00 11.63 5.65 1 Gr, L
75°110 W 1990 1.58 0.00 2.63 12.65
2010 1.30 0.00 10.37 7.78
Matagami 49°500 N 248 209.86 951.36 4.53 2.84 1979 2.15 0.00 12.94 8.72 7 Gr, L, U, Go
77°370 W 1990 0.24 0.00 8.46 27.27
2010 1.14 0.00 12.15 8.28

Table A2 of the TOM pool, incorporation of Hg into wetlands, Hg adsorption


Mercury characteristics in lake sediment cores. Data are from Moingt et al. (2013). on the soil mineral phase, which could contribute to mask and/or
Lake THg baseline in THg peak in ASEF EHg dilute the signal of the perturbation.
sediment (ng/g) sediment (ng/g) Irrespective of the presence or absence of an impact from land-
(ng/g) use, this study underlines the importance of TOM, based on lignin
Des Jardins 169 364 2.2 196 biomarkers results, as a vector of Hg transfer from the watershed
Ouescapis 33 89 2.7 56 to the receiving lakes since sedimentary Hg enrichment seems to
Rodayer 24 67 2.8 43
be directly proportional to the amount of TOM coming from the
Waswanipi 40 98 2.5 58
Chibougamau 40 150 3.8 110 watershed (Sigma 8 indicator). This is critical since TOM flows nat-
Dickson 6 90 15 84 urally from watershed soils and has thus contributed to a marked
Matagami 16 87 5.4 (surface) 71 increased accumulation of atmospheric Hg in lake sediments. Then,
15.0 (peak)
any natural and/or anthropogenic modification increasing TOM
fluxes from the watershed to the lake can increase the amount of
Hg reaching the lake, this supplementary Hg having the potential
watershed (e.g. logging and mining activities) as a contributor to
to be methylated and then incorporated into the aquatic food
the increasing amount of THg reaching lakes sediments (Garcia
chain. Because most of the atmospheric/deposited Hg is associated
and Carignan, 2000, 2005; Porvari and Verta, 2003; Porvari et al.,
with ground vegetation and with soils (Hintelmann et al., 2002;
2003). In contrast to these findings on watersheds for small lakes,
Ouellet et al., 2009) a decrease in Hg emissions should ultimately
the present study demonstrates that the impact of major anthropo-
lead to a decrease in the amount of Hg transferred from the wa-
genic activities (e.g. extensive logging, mining) occurring in the
tershed to the lake. A solid knowledge of both bio-morphological
watershed of large boreal lakes is not necessarily discernible in
parameters (e.g. mean slope, vegetation cover; Moingt et al.,
Hg and TOM concentrations in sediment cores integrating the
2013) and TOM inputs to the lake could help to identify water-
whole watershed. Indeed, in most of the lakes with anthropogenic
sheds at higher risk of exporting Hg to the lake in order to limit
perturbations in their watersheds, THg profiles are typical of atmo-
the human health risk by fish consumption.
spheric/deposition impacted sediments (i.e. ASEF around 3)
whereas the sedimentary lignin signature does not change after
the beginning of those activities in the watershed. However, in Acknowledgments
highly impacted watersheds (e.g. forest clearing in the order of
30% the watershed surface over 10 years) both Hg and organic mat- This project was financed through the CARA (Clean Air Regula-
ter contents can be influenced. This study indicates that large-scale tory Agenda) program of Environment Canada. We would like to
watersheds are characterized by some ‘‘buffering’’ capacity to land- thank Sophie Chen for her assistance in the laboratory and Jean-
use disturbance, which would be more likely due to the system’s Sébastien Beaulne for his help during the sampling work.
inertia since it increases the amount of time needed for a perturba-
tion to reach the system. Indeed in large watershed boreal ecosys- Appendix A. Appendix
tems, THg and TOM are more likely to be impacted by processes,
such as OM degradation, OM incorporation into soils, fractionation See Figs. A1 and A2 and Tables A1–A3.
Table A3

46
Primary lignin–phenol results. In each core, the gray area corresponds to the portion of the core younger than 100 years. allo = allochthonous, auto = autochthonous.

M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48


M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48 47

References Houel, S., Louchouarn, P., Lucotte, M., Canuel, R., Ghaleb, B., 2006. Translocation of
soil organic matter following reservoir impoundment in boreal systems:
implications for in situ productivity. Limnol. Oceanogr. 51, 1497–1513.
Ali, A.A., Ghaleb, B., Garneau, M., Asnong, H., Loisel, J., 2008. Recent peat
Hu, F.S., Hedges, J.I., Gordon, E.S., Brubaker, L.B., 1999. Lignin biomarkers and pollen
accumulation rates in minerotrophic peatlands of the Bay James region,
in postglacial sediments of an Alaskan lake. Geochim. Cosmochim. Acta 63,
Eastern Canada, inferred by 210Pb and 137Cs radiometric techniques. Appl.
1421–1430.
Radiat. Isot. 66, 1350–1358.
Hyodo, F., Tsugeki, N., Azuma, J.-I., Urabe, J., Nakanishi, M., Wada, E., 2008. Changes
Appleby, P.G., Oldfield, F., 1992. Application of 210Pb to sedimentation studies. In:
in stable isotopes, lignin-derived phenols, and fossil pigments in sediments of
Uranium Series Disequilibrium. Application to Earth, Marine and
Lake Biwa, Japan: implications for anthropogenic effects over the last 100 years.
Environmental Sciences. Oxford University Press, pp. 731–778.
Sci. Total Environ. 403, 139–147.
Beaulne, J.-S., Lucotte, M., Paquet, S., Canuel, R., 2012. Modeling mercury
Ikingura, J.R., Akagi, H., 1996. Monitoring of fish and human exposure to mercury
concentrations in northern pikes and walleyes from frequently fished lakes of
due to gold mining in the Lake Victoria goldfields, Tanzania. Sci. Total Environ.
Abitibi-Témiscamingue (Québec, Canada): a GIS approach. Boreal Environ. Res.
191, 59–68.
17, 277–290.
Kainz, M., Lucotte, M., Parrish, C.C., 2003. Relationships between organic matter
Bloom, N.S., Fitzgerald, W.F., 1988. Determination of volatile mercury species at the
composition and methyl mercury content of offshore and carbon-rich littoral
picogram level by low temperature gas chromatography with cold-vapor
sediments in an oligotrophic lake. Can. J. Fish. Aquat. Sci. 60, 888–896.
atomic fluorescence detection. Anal. Chim. Acta 208, 151–161.
Kamman, N.C., Engstrom, D.R., 2002. Historical and present fluxes of mercury to
Brandenberger, J., Louchouarn, P., Crecelius, E., 2011. Natural and post-urbanization
Vermont and New Hampshire lakes inferred from 210Pb dated sediment cores.
signatures of hypoxia in two basins of Puget sound: historical reconstruction of
Atmos. Environ. 36, 1599–1609.
redox sensitive metals and organic matter inputs. Aquat. Geochem. 17, 645–670.
Kleeberg, A., Schapp, A., Biemelt, D., 2008. Phosphorus and iron erosion from non-
Chabbi, A., Rumpel, C., 2004. Decomposition of plant tissue submerged in an
vegetated sites in a post-mining landscape, Lusatia, Germany: impact on
extremely acidic mining lake sediment: phenolic CuO-oxidation products and
aborning mining lakes. CATENA 72, 315–324.
solid-state 13C NMR spectroscopy. Soil Biol. Biochem. 36, 1161–1169.
Kolka, R.K., Grigal, D.F., Verry, E.S., Nater, E.A., 1999. Mercury and organic carbon
Coquery, M., Welbourn, P.M., 1995. The relationship between metal concentration
relationships in streams draining forested upland/peatland watersheds. J.
and organic matter in sediments and metal concentration in the aquatic
Environ. Qual. 28, 766–775.
macrophyte Eriocaulon septangulare. Water Res. 29, 2094–2102.
Lavoie, R.A., Hebert, C.E., Rail, J.-F., Braune, B.M., Yumvihoze, E., Hill, L.G., Lean,
Dalzell, B.J., Filley, T.R., Harbor, J.M., 2007. The role of hydrology in annual organic
D.R.S., 2010. Trophic structure and mercury distribution in a Gulf of St.
carbon loads and terrestrial organic matter export from a midwestern
Lawrence (Canada) food web using stable isotope analysis. Sci. Total Environ.
agricultural watershed. Geochim. Cosmochim. Acta 71, 1448–1462.
408, 5529–5539.
Davis, R.B., Hess, C.T., Norton, S.A., Hanson, D.W., Hoagland, K.D., Anderson, D.S.,
Lindberg, S., Bullock, R., Ebinghaus, R., Engstrom, D., Feng, X., Fitzgerald, W., Pirrone, N.,
1984. 137Cs and 210Pb dating of sediments from soft-water lakes in New England
Prestbo, E., Seigneur, C., 2007. A synthesis of progress and uncertainties in attributing
(U.S.A.) and Scandinavia, a failure of 137Cs dating. Chem. Geol. 44, 151–185.
the sources of mercury in deposition. AMBIO: A J. Hum. Environ. 36, 19–33.
Dittmar, T., Lara, R.J., 2001. Molecular evidence for lignin degradation in sulfate-
Loh, P.S., Chen, C.-T.A., Anshari, G.Z., Wang, J.-T., Lou, J.-Y., Wang, S.-L., 2012. A
reducing mangrove sediments (Amazônia, Brazil). Geochim. Cosmochim. Acta
comprehensive survey of lignin geochemistry in the sedimentary organic
65, 1417–1428.
matter along the Kapuas River (West Kalimantan, Indonesia). J. Asian Earth Sci.
Farella, N., Lucotte, M., Louchouarn, P., Roulet, M., 2001. Deforestation modifying
43, 118–129.
terrestrial organic transport in the Rio Tapajós, Brazilian Amazon. Org.
Louchouarn, P., Lucotte, M., Mucci, A., Pichet, P., 1993. Geochemistry of mercury in
Geochem. 32, 1443–1458.
two hydroelectric reservoirs in Quebec, Canada. Can. J. Fish. Aquat. Sci. 50, 269–
Fitzgerald, W.F., Engstrom, D.R., Mason, R.P., Nater, E.A., 1998. The case for
281.
atmospheric mercury contamination in remote areas. Environ. Sci. Technol.
Louchouarn, P., Lucotte, M., Farella, N., 1999. Historical and geographical variations
32, 1–7.
of sources and transport of terrigenous organic matter within a large-scale
Flynn, W.W., 1968. The determination of low levels of polonium-210 in
coastal environment. Org. Geochem. 30, 675–699.
environmental materials. Anal. Chim. Acta 43, 221–227.
Lucotte, M., Mucci, A., Hillaire-Marcel, C., Pichet, P., Grondin, A., 1995.
Fraser, A.I., Harrod, T.R., Haygarth, P.M., 1999. The effect of rainfall intensity on soil
Anthropogenic mercury enrichment in remote lakes of northern Québec
erosion and particulate phosphorus transfer from arable soils. Water Sci.
(Canada). Water Air Soil Pollut. 80, 467–476.
Technol. 39, 41–45.
Louchouarn, P., Opsahl, S., Benner, R., 2000. Isolation and Quantification of
Garcia, E., Carignan, R., 2000. Mercury concentrations in northern pike (Esox lucius)
Dissolved Lignin from Natural Waters Using Solid-Phase Extraction and GC/
from boreal lakes with logged, burned, or undisturbed catchments. Can. J. Fish.
MS. Anal. Chem. 72, 2780–2787.
Aquat. Sci. 57, 129–135.
Mäkinen, J., Kauppila, T., Loukola-Ruskeeniemi, K., Mattila, J., Miettinen, J., 2010.
Garcia, E., Carignan, R., 2005. Mercury concentrations in fish from forest harvesting
Impacts of point source and diffuse metal and nutrient loading on three
and fire-impacted Canadian Boreal lakes compared using stable isotopes of
northern boreal lakes. J. Geochem. Explor. 104, 47–60.
nitrogen. Environ. Toxicol. Chem. 24, 685–693.
Moingt, M., Bressac, M., Bélanger, D., Amyot, M., 2010. Role of ultra-violet radiation,
Goñi, M.A., Hedges, J.I., 1992. Lignin dimers: structures, distribution, and potential
mercury and copper on the stability of dissolved glutathione in natural and
geochemical applications. Geochim. Cosmochim. Acta 56, 4025–4043.
artificial freshwater and saltwater. Chemosphere 80, 1314–1320.
Goñi, M.A., Montgomery, S., 2000. Alkaline CuO oxidation with a microwave
Moingt, M., Lucotte, M., Paquet, S., Beaulne, J.-S., 2013. The influence of
digestion system: lignin analyses of geochemical samples. Anal. Chem. 72,
anthropogenic disturbances and watershed morphological characteristics on
3116–3121.
Hg dynamics in Northern Quebec large boreal lakes. Adv. Environ. Res. 2, 81–98.
Goñi, M.A., Nelson, B., Blanchette, R.A., Hedges, J.I., 1993. Fungal degradation of
Neteler, M., Mitasova, H., 2008. Open Source GIS a GRASS GIS Approach. Springer,
wood lignins: geochemical perspectives from CuO-derived phenolic dimers and
New-York, p. 383.
monomers. Geochim. Cosmochim. Acta 57, 3985–4002.
Opsahl, S., Benner, R., 1995. Early diagenesis of vascular plant tissues: lignin and
Gordon, E.S., Goñi, M.A., 2003. Sources and distribution of terrigenous organic
cutin decomposition and biogeochemical implications. Geochim. Cosmochim.
matter delivered by the Atchafalaya River to sediments in the northern Gulf of
Acta 59, 4889–4904.
Mexico. Geochim. Cosmochim. Acta 67, 2359–2375.
Orem, W.H., Colman, S.M., Lerch, H.E., 1997. Lignin phenols in sediments of Lake Baikal,
Hakanson, L., Jansson, M., 1983. Principles of Lake Sedimentology. Springer, New-
Siberia: application to paleoenvironmental studies. Org. Geochem. 27, 153–172.
York.
Ouellet, J.-F., Lucotte, M., Teisserenc, R., Paquet, S., Canuel, R., 2009. Lignin
Hamilton, T.F., Smith, J.D., 1986. Improved alpha energy resolution for the
biomarkers as tracers of mercury sources in lakes water column.
determination of polonium isotopes by alpha-spectrometry. Int. J. Radiat.
Biogeochemistry 94, 123–140.
Appl. Instrum. Part A. Appl. Radiat. Isot. 37, 628–630.
Pacyna, E.G., Pacyna, J.M., Steenhuisen, F., Wilson, S., 2006. Global anthropogenic
Hedges, J.I., Ertel, J.R., 1982. Characterization of lignin by gas capillary
mercury emission inventory for 2000. Atmos. Environ. 40, 4048–4063.
chromatography of cupric oxide oxidation products. Anal. Chem. 54, 174–178.
Petit, S., Lucotte, M., Teisserenc, R., 2011. Mercury sources and bioavailability in
Hedges, J.I., Mann, D.C., 1979. The characterization of plant tissues by their lignin
lakes located in the mining district of Chibougamau, eastern Canada. Appl.
oxidation products. Geochim. Cosmochim. Acta 43, 1803–1807.
Geochem. 26, 230–241.
Hedges, J.I., Parker, P.L., 1976. Land-derived organic matter in surface sediments
Pichet, P., Morrison, K., Rheault, I., Tremblay, A., 1999. Analysis of total mercury and
from the Gulf of Mexico. Geochim. Cosmochim. Acta 40, 1019–1029.
methylmercury in environmental samples. In: Schetagne, R., Lucotte, M.,
Hedges, J.I., Keil, R.G., Benner, R., 1997. What happens to terrestrial organic matter
Thérien, N., Langlois, C., Tremblay, A. (Eds.), Mercury in the Biogeochemical
in the ocean? Org. Geochem. 27, 195–212.
Cycle, Natural Environments and Hydroelectric Reservoirs of Northern Québec.
Hernes, P.J., Benner, R., 2003. Photochemical and microbial degradation of dissolved
Springer, pp. 41–52.
lignin phenols: implications for the fate of terrigenous dissolved organic matter
Porvari, P., Verta, M., 2003. Total and methyl mercury concentrations and fluxes
in marine environments. J. Geophys. Res. 108, 3291.
from small boreal forest catchments in Finland. Environ. Pollut. 123, 181–191.
Hernes, P.J., Robinson, A.C., Aufdenkampe, A.K., 2007. Fractionation of lignin during
Porvari, P., Verta, M., Munthe, J., Haapanen, M., 2003. Forestry practices increase
leaching and sorption and implications for organic matter ‘‘freshness’’. Geophys.
mercury and methyl mercury output from boreal forest catchments. Environ.
Res. Lett. 34, L17401.
Sci. Technol. 37, 2389–2393.
Hintelmann, H., Harris, R., Heyes, A., Hurley, J.P., Kelly, C.A., Krabbenhoft, D.P.,
Prahl, F.G., Ertel, J.R., Goni, M.A., Sparrow, M.A., Eversmeyer, B., 1994. Terrestrial
Lindberg, S., Rudd, J.W.M., Scott, K.J., St.Louis, V.L., 2002. Reactivity and mobility
organic carbon contributions to sediments on the Washington margin.
of new and old mercury deposition in a boreal forest ecosystem during the first
Geochim. Cosmochim. Acta 58, 3035–3048.
year of the METAALICUS study. Environ. Sci. Technol. 36, 5034–5040.
48 M. Moingt et al. / Applied Geochemistry 41 (2014) 34–48

Rezende, C.E., Pfeiffer, W.C., Martinelli, L.A., Tsamakis, E., Hedges, J.I., Keil, R.G., 2010. Tareq, S.M., Kitagawa, H., Ohta, K., 2011. Lignin biomarker and isotopic records of
Lignin phenols used to infer organic matter sources to Sepetiba Bay – RJ, Brasil. paleovegetation and climate changes from Lake Erhai, southwest China, since
Estuar. Coast. Shelf Sci. 87, 479–486. 18.5 ka BP. Quatern. Int. 229, 47–56.
Rognerud, S., Fjeld, E., 2001. Trace element contamination of Norwegian lake Teisserenc, R., 2009. Dynamique de la matière organique terrigène et du mercure
sediments. AMBIO: A J. Hum. Environ. 30, 11–19. dans les lacs et réservoirs boréaux. Institut des Sciences de l’Environnement.
Sánchez-García, L., De Andrés, J.R., Martín-Rubí, J.A., Louchouarn, P., 2009. Ph.D. UQAM, Montréal, 2009, p. 182.
Diagenetic state and source characterization of marine sediments from the Teisserenc, R., Lucotte, M., Houel, S., Carreau, J., 2010. Integrated transfers of
inner continental shelf of the Gulf of Cádiz (SW Spain), constrained by terrigenous organic matter to lakes at their watershed level: a combined
terrigenous biomarkers. Org. Geochem. 40, 184–194. biomarker and GIS analysis. Geochim. Cosmochim. Acta 74, 6375–6386.
Sharpley, A.N., Mcdowell, R.W., Kleinmann, P.J.A., 2001. Phosphorus loss from land Teisserenc, R., Lucotte, M., Houel, S., 2011. Terrestrial organic matter biomarkers as
to water: integrating agricultural and environmental management. Plant Soil tracers of Hg sources in lake sediments. Biogeochemistry 103, 235–244.
237, 287–307. Teisserenc, R., Lucotte, M., Canuel, R., Moingt, M., Obrist, D., 2013. Combined
Slauenwhite, D.E., Wangersky, P.J., 1996. Extraction of marine organic matter on dynamics of mercury and terrigenous organic matter following impoundment
XAD-2: effect of sample acidification and development of an in situ pre- of Churchill Falls Hydroelectric Reservoir, Labrador. Biogeochemistry.
acidification technique. Mar. Chem. 54, 107–117. Tesi, T., Langone, L., Goñi, M.A., Turchetto, M., Miserocchi, S., Boldrin, A., 2008.
Soto, D.X., Roig, R., Gacia, E., Catalan, J., 2011. Differential accumulation of mercury Source and composition of organic matter in the Bari canyon (Italy): dense
and other trace metals in the food web components of a reservoir impacted by a water cascading versus particulate export from the upper ocean. Deep Sea Res.
chlor-alkali plant (Flix, Ebro River, Spain): implications for biomonitoring. Part I 55, 813–831.
Environ. Pollut. 159, 1481–1489. Watras, C.J., Back, R.C., Halvorsen, S., Hudson, R.J.M., Morrison, K.A., Wente, S.P.,
Swain, E.B., Engstrom, D.R., Brigham, M.E., Henning, T.E., Brezonik, P.L., 1992. 1998. Bioaccumulation of mercury in pelagic freshwater food webs. Sci. Total
Increasing rates of atmospheric mercury deposition in midcontinental North Environ. 219, 183–208.
America. Science 257, 784–787. Yang, Y.-K., Zhang, C., Shi, X.-J., Lin, T., Wang, D.-Y., 2007. Effect of organic matter
Tahvanainen, T., Haraguchi, A., 2013. Effect of pH on phenol oxidase activity on and pH on mercury release from soils. J. Environ. Sci. 19, 1349–1354.
decaying Sphagnum mosses. Eur. J. Soil Biol. 54, 41–47.

You might also like