PUBLICATIONS
Journal of Geophysical Research: Biogeosciences
RESEARCH ARTICLE
10.1002/2015JG002928
Key Point:
• Similar temperature sensitivity across
systems and within the seasonal scale
Supporting Information:
• Supporting Information S1
Correspondence to:
C. Gudasz,
[email protected]
Citation:
Gudasz, C., S. Sobek, D. Bastviken,
B. Koehler, and L. J. Tranvik (2015),
Temperature sensitivity of organic
carbon mineralization in contrasting
lake sediments, J. Geophys. Res.
Biogeosci., 120, 1215–1225, doi:10.1002/
2015JG002928.
Received 22 JAN 2015
Accepted 17 JUN 2015
Accepted article online 21 JUN 2015
Published online 15 JUL 2015
Temperature sensitivity of organic carbon mineralization
in contrasting lake sediments
Cristian Gudasz1,2, Sebastian Sobek1, David Bastviken3, Birgit Koehler1, and Lars J. Tranvik1
1
Limnology, Department of Ecology and Evolution, Uppsala University, Uppsala, Sweden, 2Now at Department of Ecology
and Evolutionary Biology, Princeton University, Princeton, New Jersey, USA, 3Department of Thematic Studies—Water and
Environmental Studies, Linköping University, Linköping, Sweden
Abstract
Temperature alone explains a great amount of variation in sediment organic carbon (OC)
mineralization. Studies on decomposition of soil OC suggest that (1) temperature sensitivity differs
between the fast and slowly decomposition OC and (2) over time, decreasing soil respiration is coupled with
increase in temperature sensitivity. In lakes, autochthonous and allochthonous OC sources are generally
regarded as fast and slowly decomposing OC, respectively. Lake sediments with different contributions of
allochthonous and autochthonous components, however, showed similar temperature sensitivity in
short-term incubation experiments. Whether the mineralization of OC in lake sediments dominated by
allochthonous or autochthonous OC has different temperature sensitivity in the longer term has not been
addressed. We incubated sediments from two boreal lakes that had contrasting OC origin (allochthonous
versus autochthonous), and OC characteristics (C/N ratios of 21 and 10) at 1, 3, 5, 8, 13, and 21°C for five
months. Compared to soil and litter mineralization, sediment OC mineralization rates were low in spite of low
apparent activation energy (Ea). The fraction of the total OC pool that was lost during five months varied
between 0.4 and 14.8%. We estimate that the sediment OC pool not becoming long-term preserved was
degraded with average apparent turnover times between 3 and 32 years. While OC mineralization was
strongly dependent on temperature as well as on OC composition and origin, temperature sensitivity was
similar across lakes and over time. We suggest that the temperature sensitivity of OC mineralization in lake
sediments is similar across systems within the relevant seasonal scales of OC supply and degradation.
1. Introduction
Soils and lake sediments are well-recognized sites for the mineralization and storage of organic carbon (OC).
In the boreal biome, lake sediments store large amounts of OC [Kortelainen et al., 2004; Benoy et al., 2007]. The
settling OC that reaches the sediment surface can either originate from internal primary production
(autochthonous) or is imported from the drainage area (allochthonous). For lakes in the boreal zone,
allochthonous dissolved organic carbon (DOC) that is leaching from wetlands and forest soils in their
catchments [Laudon et al., 2011] is partially flocculated in the water column and subsequently settled onto
the sediments [von Wachenfeldt et al., 2008]. A significant fraction of the terrestrial primary production in
boreal landscapes can end up on the bottom of lakes [von Wachenfeldt and Tranvik, 2008]. Given this
translocation of carbon across the terrestrial-aquatic continuum it is valid to ask: Does the temperature
sensitivity of the sediment mineralization of allochthonous OC differs from that of the soil where it
originates and with that of the sediment autochthonous OC? Such a perspective is currently lacking.
©2015. The Authors.
This is an open access article under the
terms of the Creative Commons
Attribution-NonCommercial-NoDerivs
License, which permits use and distribution in any medium, provided the
original work is properly cited, the use is
non-commercial and no modifications
or adaptations are made.
GUDASZ ET AL.
Lake sediments dominated by autochthonous or allochthonous OC sources have contrasting properties
[Meyers and Teranes, 2001] that is reflected in their susceptibility to microbial decomposition. Thus,
sediments dominated by autochthonous material decay faster compared to more slowly decomposing
sediments dominated by allochthonous inputs [Gudasz et al., 2012].
The OC fraction that is stored in the sediment is constrained by microbial decomposition. Temperature has
been shown to be a main controlling factor of sediment OC mineralization [Bergström et al., 2010; Gudasz
et al., 2010; Cardoso et al., 2014]. While differences in temperature sensitivities for the mineralization of fast
and slowly decomposing soil OC are intensely debated [von Lützow and Kögel-Knabner, 2009], the
temperature sensitivity for the mineralization of sediment OC of allochthonous or autochthonous origin is
virtually unknown. A difficulty in elucidating the temperature sensitivity of fast versus slowly decomposing
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1215
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
OC lies in the fact that what we measure is an apparent temperature response (i.e., apparent Ea). The apparent
Ea is the expression of the overall temperature sensitivity of OC decomposition by the microbial community
[e.g., Craine et al., 2010] and results from the integration of various temperature-dependent processes at
different scales [Ågren and Wetterstedt, 2007; Conant et al., 2011].
Based on soil studies, different competing hypotheses have emerged to explain apparent temperature
sensitivity of microbial OC decomposition [von Lützow and Kögel-Knabner, 2009; Conant et al., 2011]. The
temperature sensitivity of slowly decomposing OC has attracted particular interest due to the large
contribution of this pool to the soil carbon stocks [e.g., Davidson et al., 2000; Knorr et al., 2005]. According
to the Arrhenius equation [Arrhenius, 1889], the temperature sensitivity of microbial decomposition should
increase with increasing activation energy (Ea) of a reaction [Bosatta and Ågren, 1999; Davidson and
Janssens, 2006]. Hence, the carbon-quality-to-temperature hypothesis postulated by Bosatta and Ågren
[Bosatta and Ågren, 1999] argues that the net Ea for the decomposition of complex molecules is higher
due to a higher number of enzymatic steps needed to release carbon as carbon dioxide. Complex
molecules (i.e., lower quality) should exhibit higher temperature sensitivity than simple monomers
(i.e., higher quality). Although allochthonous and autochthonous OC sources are complex mixtures, in
general, they can be considered as having lower and higher quality (i.e., slow and fast decomposing)
[Koehler et al., 2012; Guillemette et al., 2013]. Assuming that temperature sensitivity of decomposition
at the molecular level also holds at the soils sample level [Janssens and Vicca, 2010], the slowly
decomposing OC (which exhibits higher Ea) should respond stronger to increasing temperature, than
fast decomposing OC, which exhibits lower Ea [Conant et al., 2008; Craine et al., 2010].
However, contradicting empirical studies [e.g., von Lützow and Kögel-Knabner, 2009] makes it difficult to
demonstrate a universal relationship between soil OC decomposition and temperature. Various factors
associated with the physicochemical environment (e.g., soil pH, humidity, nutrients, and oxygen) and
different effects associated with longer incubation periods can lead to differential effects of temperature
on soil OC decomposition.
The time scale is important when considering the temperature sensitivity of OC mineralization. The depletion
of short-lived fast decomposing OC, at higher temperatures during incubation of confined samples over
longer time, could lead to an increase in the apparent temperature sensitivity. Hence, increasing
incubation time lead to a decrease in soil respiration, which was coupled to an increase in apparent Ea,
but only after 150 days [Craine et al., 2010]. Alternatively, longer incubation time may result in decreased
soil respiration, coupled to decreased microbial biomass and thermal adaptation of the microbial
community [Bradford et al., 2008] and thus decreasing apparent Ea. Short time scale responses might
reveal only transient effects. It is virtually unknown how temperature sensitivity of sediment OC
mineralization changes over longer time scale, which is critical if we try to understand future responses to
climate warming.
It is not evident to what extent that the patterns of temperature sensitivity described for soils are valid also for
lake sediment OC mineralization. Considering the large amounts of OC stored in inland water sediments
[Dean and Gorham, 1998] and the high global rates of accumulation [Tranvik et al., 2009], it is important to
find out whether temperature sensitivity is different for the allochthonous OC, being washed out from
soils and settled to the lake bottoms, compared to the autochthonous OC originating from aquatic
primary production. The temperature sensitivity of OC mineralization rates is not only important for the
total balance between OC mineralization and preservation but may also influence how quickly different
types of OC are mineralized and could therefore have major impact on OC cycling and composition over
time. For example, if slowly decomposing OC has high-temperature sensitivity, a greater mineralization
and loss of presently preserved carbon stocks could be expected in a warmer climate.
In this study we explore the relationship between sediment OC mineralization and temperature during a fivemonth incubation of two contrasting boreal lake sediments, with allochthonous versus autochthonous
dominated OC sources, as well as of properties (i.e., slow versus fast decomposing OC). We show that for
the duration of incubation, which is equivalent to seasonal scale, temperature sensitivity of OC
mineralization was similar over time and among lakes with contrasting sediment OC characteristics. We
suggest that for the seasonal cycle of supply and degradation of OC, temperature sensitivity of sediment
OC mineralization is similar and encompasses differences in organic matter composition.
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1216
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
2. Methods
2.1. Study Lakes and Experimental Design
Intact sediment cores were sampled with a gravity corer (inner diameter = 59 mm; Uwitech, Mondsee,
Austria) at the deepest point of two lakes in south-central boreal Sweden under ice cover in February
2010. Svarttjärn is a small oligotrophic lake rich in humic matter and with a low contribution of
autochthonous OC in the sedimentation flux [von Wachenfeldt and Tranvik, 2008; Gudasz et al., 2012].
Vallentunasjön is a eutrophic lake with high nutrient concentrations and high productivity of
phytoplankton and macrophytes [Boström et al., 1989; Brunberg and Boström, 1992; Rydin et al.,
2010; Gustafsson and Rydin, 2015]. In a recent water quality survey [Gustafsson and Rydin, 2015],
the average between 2007 and 2014 of total phosphorus (TP) concentration in Vallentunasjön was
63.9 ± 5.7 μg L 1, while a seasonal survey in Svarttjärn showed TP of 15.1 ± 6.1 μg L 1 [Gudasz et al.,
2012], most of which is organically bound to dissolved humic matter. The dissolved organic carbon
(DOC) concentration was similar in both lakes, with 26 and 28 mg L 1 in Vallentunasjön and Svarttjärn,
respectively. However, the DOC-specific absorbance at 250 nm (A250/DOC), an indicator of terrestrial
humic material [McKnight and Aiken, 1998], was about 5 times higher in Svarttjärn (for details see
Gudasz et al. [2010]). More limnological characteristics of the two lakes can be found in Table S1 in the
supporting information.
At the time of sampling, bottom temperature was 4.8°C and the sediment surface was anoxic in both lakes.
The sediment cores were transported in insulated boxes to the laboratory and kept at in situ temperatures. In
the laboratory and within the same day of sampling, the upper 5 cm of the sediment was sliced off and
transferred to polycarbonate incubation tubes (length = 400 mm and inner diameter = 54 mm), keeping the
sediment structure with minimum disturbance. The method we have developed to transfer the sediment
does not entirely eliminate sediment disturbance during sampling but it minimizes disturbance and
allowed sampling without any visible mixing of the sediment. Approximately 0.6 L of water was kept
above the sediment. The incubation tubes were equilibrated and acclimated at the incubation
temperature for a period of 10 days, open to the atmosphere and with continuous mixing of the water
column. The water column mixing was achieved through a magnetic stirring system. A magnet-containing
buoyant Eppendorf vial was placed in each tube, and the tubes were immersed in an insulated, dark water
bath arranged in a circular pattern. Four arms connected to an axle and equipped with magnets were
rotating outside the tubes, moving the Eppendorf vials inside the sediment tubes, to gently mix the water
column without resuspending the sediment. For both lakes, we then incubated three sediment tubes at
each of the temperatures.
We used six temperature treatments for the experiment. To maintain constant temperatures, we used
insulated water baths connected to external refrigerated circulators (Julabo, F 25 ED, Seelbach, Germany).
The temperature inside the water baths was logged at 5 min intervals (Fluke 2625A, Hydra Series II, Everett,
USA) with fast-response Fluke 80PK-1 thermocouples. The thermocouples were adjusted against a
reference thermometer (Hart Scientific 1522, American Fork, USA), equipped with a reference probe (Hart
Scientific PRT 5613). Mean temperatures (±SD) throughout the duration of the experiment (150 days) were
0.99 ± 0.03, 2.97 ± 0.07, 5.04 ± 0.06, 7.99 ± 0.08, 13.01 ± 0.03, 21.06 ± 0.06°C.
2.2. Sediment OC Mineralization
OC mineralization is defined as production of dissolved inorganic carbon (DIC) and methane. Methane
release was tested at 21°C for both lakes, at the end of the experiment, and was below 1% of the DIC
production. Thus, we use DIC production as a measure of OC mineralization. Methane concentrations were
determined using a gas chromatograph with a flame ionization detector (GC-FID, Agilent Tech 7890A with
methanizer). Incubations were made in the dark and the sediment OC mineralization rate was measured as
the change in DIC (dissolved inorganic carbon) concentration in the mixed water overlying the sediment,
4, 17, 35, 58, 86, 100, and 150 days into the experiment. During each rate measurement occasion, the
incubation tubes were closed headspace-free, with gas-tight stoppers for 45 to 192 h (depending on the
temperature, with the shortest measurement times at 13 and 21°C), which were the durations needed to
confidently measureable differences in DIC concentrations. Two replicate measurements in Svarttjärn at 3°C
and during the fourth rate measurement occasion were lost. Oxygen concentration at the end of the
closure period always exceeded 3 mg L 1, with the exception of one core out of 252 on day 4 and another
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1217
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
core on day 17 of the experiment when it was ~1 mg L 1. At the start and end of tube closure, water samples
(20 mL) were withdrawn through sampling ports while pushing down the stoppers into the tubes to avoid
headspace formation. Once each rate measurement closure was finished, the tubes were kept open and
with continuous water column mixing in order to maintain oxic conditions until the next mineralization
rate measurement and surface lake water was added to keep water volume constant. All data on sediment
OC mineralization can be found in the supporting information.
Sediment incubations are likely have a level of uncertainty due to the experimental manipulations. The
sediment incubation measurements we have carried out were meant to measure the potential, apparent
mineralization rates of widely different sediments under comparable experimental conditions. It is
questionable, however, if the rates are representative for in situ sediment OC mineralization in an absolute
sense, but the comparison of the apparent OC mineralization we observed is robust and representative
with respect to the widely different types of sediment.
2.3. Temperature Sensitivity
The temperature sensitivity was calculated as the slope of a linear regression between the logarithm of the
sediment OC mineralization rate and temperature (°C). Apparent Ea (kJ mol 1) representing another
expression of temperature sensitivity was calculated as described in Craine et al. [2010].
2.4. Dissolved Oxygen Measurements
During field sampling, dissolved oxygen concentration and temperature at the lake bottom were measured
with a WTW Oxi 340i (Weilheim, Germany). During laboratory incubation, dissolved oxygen concentrations
was monitored with an optical sensor system (Oxy-10 mini) and PSt3 oxygen sensor spots (PreSens–Precision
Sensing GmbH, Regensburg, Germany).
2.5. Dissolved Inorganic Carbon
DIC concentrations were measured using a Sievers 900 total organic carbon analyzer (GE Analytical
Instruments, Manchester, UK), with a precision of <1.5% relative standard deviation and an accuracy of
±2% or 0.5 ppb, whichever is greater. During DIC measurements, water samples were kept cool at 3°C.
2.6. Elemental Analyses
At the end of the experiment, each incubated core was sliced into 1 cm sections, frozen, and freeze-dried.
Total carbon and total nitrogen content were determined with a CHN analyzer (ECS 4010, Costech
International, Milan, Italy). Instrument performance was checked using an acetanilide standard. Sediment
C/N atomic ratios were calculated by dividing percent carbon with percent nitrogen and then converted
to atomic ratios by multiplying with 1.167 (the ratio of atomic weights of nitrogen and carbon). All data on
sediment elemental analyses can be found in Table S2 in the supporting information.
2.7.
13
C-nuclear Magnetic Resonance Spectroscopy
The chemical composition of bulk organic matter in the 0–1 cm layer of sediment was analyzed by solid-state
C cross-polarization magic angle spinning (CP/MAS) nuclear magnetic resonance (NMR) spectroscopy
(Bruker DSX 200). A 13C resonance frequency of 50.32 MHz and a spinning speed of 6.8 kHz were applied.
Between 28,000 and 31,000 scans were accumulated, and a line broadening of 25 Hz was applied. The 13C
chemical shifts were calibrated relative to tetramethylsilane (0 ppm). The relative contributions of the
various C groups were determined by integration of the signal intensities in their respective chemical shift
regions. The region from 10 to 45 ppm was assigned to alkyl C, from 45 to 110 ppm to O-alkyl C, from
110 to 160 ppm to aromatic C, and from 160 to 220 ppm to carboxyl, carbonyl, and amide C. The one time,
snapshot NMR analysis of the sediment was used to provide a qualitative description of the sediment. All
data on sediment NMR spectra can be found in the supporting information.
13
2.8. Statistical Analyses
Factors regulating sediment OC mineralization were tested using a random-slope linear mixed-effects model
[Schielzeth and Forstmeier, 2009], which accounted for the correlation between repeated measurements on
the same subject and lack of spatial independence among the sampled sediment cores. DIC production
data were log-transformed, while temperature was mean centered and standardized by z transformation
according to Schielzeth [2010]. The model contained the linear and interaction terms of temperature, time
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1218
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
and lake as fixed effects, and sediment
cores nested in time as random
effects. A first-order temporal autoregressive process, which assumes that
correlation between measurements
decreases with increasing time distance, was included in the model.
We used the lme function from the
R-package nlme [Pinheiro et al.,
2014]. Model adequacy was checked
using residual plots [Crawley, 2009;
R Development Core Team, 2010].
P values were multiplicity adjusted
using the single-step method provided in the R-package multcomp
[Hothorn et al., 2008].
2.9. Apparent Turnover Time
of Sediment Organic Carbon
We estimated the apparent turnover
time of sediment OC based on the
fraction that was lost through mineralization, accounting for the fact that
not all the OC will be mineralized. Burial efficiency (i.e., OC burial/OC deposition) averages 66% and 22% in
boreal lakes dominated by allochthonous and autochthonous OC, respectively [Sobek et al., 2009].
Assuming that these averages apply to the study lakes, this means that 34 and 78%, respectively, of the
OC in Svarttjärn and Vallentunasjön will not be buried and can be regarded as a “labile” carbon pool.
Hence, for this remainder fraction, we calculated the carbon pool size for the 0–5 cm layer. We have then used
the average sediment OC mineralization throughout the 150 day duration, at each of the temperatures, to
calculate the apparent turnover time of this nonburied organic carbon pool. There is uncertainty in estimating OC burial efficiency as described by Sobek et al. [2009], and the true organic carbon burial efficiency of OC
in Vallentunasjön and Svarttjärn may be different from the applied averages. However, our calculations highlight overall differences in apparent turnover times of OC in lake sediments with predominantly autochthonous and allochthonous sources. Data and calculations on the apparent turnover time can be found in the
supporting information.
13
Figure 1. C CP/MAS NMR spectra of the 0–1 cm layer of the studied
sediments.
3. Results and Discussion
3.1. Sediment Characteristics
The sediments of the two studied lakes represent end members along a spectrum of OC origin and reactivity.
The average C/N ratio in the 0–5 cm layer of the incubated sediment (n = 72 for each lake) was 21.2 ± 2.8 in the
humic Svarttjärn sediment indicating a predominantly allochthonous OC source (i.e., more slowly
decomposing) and 9.5 ± 3.7 in the eutrophic Vallentunasjön sediment, indicating the larger proportion of
autochthonous OC (i.e., more rapidly decomposing). The complete data on sediment C/N ratios can be
found in Table S2. The C/N ratio is a reliable general source indicator with high organic content [Meyers
and Teranes, 2001], which mirrors the general characteristics of a eutrophic versus dystrophic lake (see
Methods). The 13C NMR spectra (Figure 1) corroborate the allochthonous character of the Svarttjärn
sediment showing a strong terrestrial imprint, with lignin signals at 150 ppm and 56 ppm, and a high share
of aliphatic C (Table 1), possibly derived from cutin or suberin in the cuticula and cell wall of terrestrial
plants. Although Vallentunasjön sediment also indicated the presence of a terrestrial imprint, it was richer
in polysaccharides (peak at 72 ppm) as well as carboxylic acids and proteins (172 ppm), indicating that the
OC is mainly derived from aquatic sources and more easily degraded than Svarttjärn sediment. In addition,
Vallentunasjön sediment is comparatively rich in O-Alkyl C (45–110 ppm), indicating the presence of
polysaccharides; Svarttjärn has a higher content of alkyl-C ( 10–45 ppm), i.e., methyl and methylene
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1219
Journal of Geophysical Research: Biogeosciences
Table 1. Relative Contribution of Major Chemical Shift Regions in the
Sediment in the Studied Lakes
Lake
Vallentunasjön
Svarttjärn
10.1002/2015JG002928
13
C CP/MAS NMR Spectra of the 0–1 cm Layer of
Alkyl C
0–45 ppm
O-Alkyl C
45–110 ppm
Aromatic C
110–160 ppm
Carboxyl C
160–210 ppm
31%
43%
45%
36%
9%
11%
12%
9.5%
groups in aliphatic rings and chains (Table 1 and Figure 1). While total aromatic C (110–160 ppm) was similar
in both lake sediments, Svarttjärn sediment displays a phenol peak at 150 ppm, which Vallentunasjön does
not. Similarly, the shoulder at 56 ppm, indicative of methoxyl C in lignin, was more pronounced in
Svarttjärn than in Vallentunasjön. On the other hand, Vallentunasjön sediment has a higher proportion of
carboxyl C and amide C (172 ppm) than Svarttjärn. Even though the two study lakes are very different in
character (see Methods), their sediment will still be a mixture of allochthonous and autochthonous OC.
Since the 13C NMR method we have used is qualitative or semiquantitative, it will detect similar
compounds and therefore the spectra are likely to look rather similar. Moreover, structures that are
difficult to degrade (such as aromatics derived from vascular plants) will not be degraded in the sediment
and will accumulate such that they will be visible in a qualitative analysis, even in a lake with large input of
readily degradable OM such as Vallentunasjön. Furthermore, Vallentunasjön has a well-developed
macrophytes belt [Rydin et al., 2010] and thus a vascular plant organic matter source to the sediment.
3.2. Temperature Sensitivity
Sediment OC mineralization increased with increasing temperature in both lakes (Figure 2). There was no
significant change in sediment OC mineralization and its temperature sensitivity between lakes and over
time (Figure 3 and Table 2). Temperature sensitivity of sediment OC mineralization over longer term
(five months; Figure 2) was similar to that derived from short-term incubations (between 37 and 94 h)
previously measured in the same lakes [Gudasz et al., 2010]. Temperature sensitivity obtained from pooled
data of Svarttjärn and Vallentunasjön was also similar with that observed for a larger set of different lakes,
geographically widely distributed [Gudasz et al., 2010]. Hence, this and previous studies suggest that
apparent temperature sensitivity of OC mineralization was similar for sediments of varying composition
and in contrast to the carbon-qualityto-temperature hypothesis [Bosatta
and Ågren, 1999].
Figure 2. Temperature sensitivity of sediment OC mineralization. Circles
represent mineralization rates measured during the 150 d experiment,
n = 18 for each lake and for n = 7 rate measurement occasions. The fitted
solid line corresponds to a linear regression estimate. The y axis is represented
on a log-scale. Statistical effect analysis was conducted using linear mixedeffects model (see Table 2).
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
The duration of the incubation can
play an important role when studying
temperature sensitivity. Short-term
soil-warming experiments in the
laboratory (e.g., <100 days) have
been criticized [von Lützow and
Kögel-Knabner, 2009] because they
mainly reflect the mineralization of
the most active, labile carbon
compounds, possibly masking the
temperature response of the larger,
more slowly decomposing OC pool
[Davidson et al., 2000; Knorr et al.,
2005]. In soil-warming experiments in
the field, respiration rates often
return to prewarming conditions after
extended periods of incubation.
Different mechanisms have been
proposed to explain this pattern,
from depletion of labile carbon
to thermal adaptation of microbial
1220
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
respiration [e.g., Davidson and Janssens,
2006; von Lützow and Kögel-Knabner,
2009]. Our experiment lasted for
150 days, similar to seasonal length in
boreal lakes (e.g., ice cover and summer
stratification). Temperature sensitivity of
sediment OC mineralization derived
from long-term incubations at constant
temperatures may result in confounding
the influence of biochemical composition
with microbial acclimation processes.
However, the effect of time on sediment
OC mineralization was not significant
across temperatures and lakes in our
study (see Table 2), indicating that the
temperature sensitivity of sediment OC
mineralization was similar over a relevant
time scale encompassing annual events
of sedimentation and mineralization.
Figure 3. Temporal variation in temperature sensitivity of sediment OC
Furthermore, a negative relationship
mineralization. Circles represent the slope of linear regression between
between respiration rate and temperature
temperature and sediment OC mineralization (n = 18), log-transformed.
sensitivity, as apparent in soils [Craine
et al., 2010], was not observed for the
sediment mineralization, neither over time nor in lakes with contrasting OC composition (Table 2). Possibly,
incubations extending over several years in absence of input of new substrate from sedimentation may
reveal a higher temperature sensitivity of the more slowly decomposing substrate pools [Bosatta and Ågren,
1999]. However, this time scale is not relevant for the dynamics of input and decay of organic matter in
boreal lake sediments and for the seasonal temperature dynamics.
The fraction of the total OC pool that was lost during 150 days of incubation between 1 and 21°C averaged
between 0.4 and 2.8% (i.e., 5.9–32 mg C or 2.6–14 g C m 2) in the humic Svarttjärn and 3.1 to 14.8%
(i.e., 18.7–77.1 mg C or 8.1–33.7 g C m 2) in the eutrophic Vallentunasjön. The total carbon loss at 21°C
in Vallentunasjön was only 2.4 times (average across temperatures 2.8 ± 0.4) higher than the loss in
Svarttjärn, even though Vallentunasjön sediment is rich in readily decomposable autochthonous OC and
Svarttjärn sediment is largely made of slowly decomposing allochthonous OC. In comparison, the
respiration rate at 20°C of a wider variety of soils (root-free in the laboratory) and plant litter varied by
factors of 70 and 77 and between soil and litter in total by a factor of 103, respectively [Craine et al., 2010].
Hence, mineralization rates exhibit far less variability between contrasting lake sediments compared to the
variation across soils and litter.
Table 2. Results of the Linear Mixed Effects Model on the Relationships Between Sediment OC Mineralization Rates (Log-Transformed Before Analysis) and
a
Standardized Temperature, Time, Lake (Svarttjärn, Sv and Vallentunasjön, Va) and Their Interactions
Full Model Test Statistics
Group Mean Parameter Estimates
Predictor
DF
t value
Unadjusted P Value
Multiplicity Adjusted P Value
Parameter
Mean
±SE
Intercept
Lake
Temperature
Temperature × lake
Time
Time × lake
Temperature × time
Temperature × time × lake
209
28
4
28
209
209
209
209
66.98
21.06
10.51
1.13
0.84
1.02
1.08
1.57
<0.001
<0.001
0.0005
0.2676
0.4013
0.3069
0.2803
0.1180
<0.001
<0.001
<0.001
0.873
0.972
0.920
0.896
0.580
Intercept of Sv
Intercept of Va
Temperature slope of Sv
Temperature slope of Va
Time slope of Sv
Time slope of Va
Temperature × time of Sv
Temperature × time of Va
1.54
1.99
0.24
0.22
0.01
0.01
0.02
0.01
0.02
0.02
0.02
0.01
0.02
0.01
0.02
0.01
a
The left part gives the t table of the full model with the univariate and multiplicity-adjusted P values. Sv was the reference group against which contrasts were
tested. Group mean parameter estimates are given in the right part. Parameter estimates differ from those in Figure 2 because in the lme models standardized
temperature was used.
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1221
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
The Ea of OC mineralization varied within a narrow range in the two lakes with no trend over incubation time
(see Table 2), from 47 and 64 kJ mol 1 in Svarttjärn and between 45 and 54 kJ mol 1 in Vallentunasjön. The
average activation energy values in Svarttjärn, Ea = 56 ± 6 kJ mol 1, and Vallentunasjön, Ea = 50 ± 4 kJ mol 1,
were close to the lowest observed Ea for soil OC and litter decomposition (Ea = 51 kJ mol 1) [Craine et al.,
2010]. Hence, while the Ea for a wide range of root-free soils and plant litter decomposition varied by a factor
of 2.9 and 1.5, respectively [Craine et al., 2010], it was remarkably similar in our contrasting lake sediments.
The similarity of Ea across contrasting lake sediments, dissimilar to the wider range of Ea for OC
decomposition among different soils, suggests that either the decomposing OC is more similar among
sediments than among soils, or that different mechanisms control mineralization and temperature
sensitivity in soils and in lake sediments. These findings point toward fundamental differences in OC
mineralization in sediments as compared to soils [e.g., Hedges and Oades, 1997] and most likely to
differences in the integration of the various temperature dependent processes such as substrate release in
the environment, diffusion and uptake [Ågren and Wetterstedt, 2007]. We argue in the following that the
physical characteristics of the aquatic environment, in general, and of the sediment water interface in
particular set boundaries to sediment OC mineralization and its temperature dependence.
The supply rate of materials to the sediment surface and to microbial cells is regulated by physical transport
mechanisms including advection, turbulence, and diffusion [Fenchel et al., 1998]. Among these, molecular
diffusion is the most important regulator of substrate flux to cells, even in highly turbulent water
[Jørgensen, 2001]. Temperature can to a large degree alter the transport characteristics of various gasses,
ions, and molecules not only within the sediment (i.e., diffusion in porewater to cell surfaces) but also
across the diffusive boundary layer (DBL), a water layer of up to a few mm in thickness on top of the
sediment surface, where molecular diffusion is the main transport mechanism of solutes [Boudreau and
Jørgensen, 2001]. The transport of solutes across the DBL is an important regulator of biogeochemical
processes in the sediment [Brand et al., 2009] and depends on the concentration gradient, the thickness of
the DBL (which in turn depends on turbulent movements of the bottom water), and on the molecular
diffusion coefficient of the solute (which in turn depends on temperature) [Jørgensen, 2001]. Hence, the
constraints imposed by the transport through the DBL together with the temperature effects on diffusion
make the DBL to act as a temperature-dependent barrier, limiting solute exchange between water and
sediment, and affecting the temperature response of sediment OC mineralization. Similarly, a reduced
temperature effect due to diffusive constraints has been suggested for aerobic respiration in marine
sediments [Thamdrup et al., 1998]. As a solute’s molecular diffusion coefficient is typically about 4 orders of
magnitude greater in air than in water, the contribution to the overall temperature sensitivity from
physical transport processes presumably is substantially greater in sediments and waterlogged soils
compared to aerated soils.
Our finding of the similar temperature sensitivity in contrasting lakes sediments with longer incubation time,
and previously shown by Gudasz et al. [2010] with short-term incubations, may be explained by the interplay
of several factors including the diffusion-controlled transport at the sediment-water interface and in
subsurface sediments and the availability of oxygen from overlaying water [Thamdrup et al., 1998]. These
factors cause the apparent Ea to be lower than in aerated soils.
Physico-chemical environmental variables can influence decomposition of fast and slowly decomposing OC
and their response to temperature differently [von Lützow and Kögel-Knabner, 2009]. Factors affecting the
catalytic activity of microbial enzymes and OC availability are likely to play an important role [Conant et al.,
2011]. Protective mineral sorption affects OC availability in both soils and marine sediments [Hedges and
Oades, 1997] but is not of major importance for OC burial efficiency (an indirect measure of OC
mineralization) in freshwater lake sediments [Sobek et al., 2009]. Instead, anoxia exerts strong control on
the mineralization of lake sediments. Unlike soils, lake sediments are anoxic below the very few top mm or
cm, and the lack of oxygen has a strong impact on the sediment organic matter stabilization, by slowing
down decomposition and in particular or allochthonous OC [Kristensen et al., 1995; Bastviken et al., 2003;
Sobek et al., 2009]. We suggest that the constraints of an anoxic environment upon mineralization of OC
in lake sediments contribute to the relative similarity in both rate and temperature sensitivity of
mineralization, compared to the highly variable decomposition rates and temperature sensitivity observed
in soils [Craine et al., 2010].
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1222
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
3.3. Apparent Turnover Time
Figure 4. The apparent turnover time of the nonburied sediment OC
fraction. Relationship between turnover time and temperature. The
nonburied sediment OC is defined as the fraction that is lost through
mineralization, accounting for the fact that not all the OC will be
mineralized (see Methods and supporting information). Circles
represent the apparent turnover time, n = 18 for each lake. The fitted
solid line corresponds to a linear regression estimate. The y axis is
represented on a log-scale.
A certain fraction of the OC that is
deposited onto the sediment will escape
microbial decomposition and eventually
be buried over geological time scales.
The fraction that is mineralized can be
regarded as the “labile” sediment OC
pool, i.e., the fraction of the sediment OC
that escapes burial. We calculated the
apparent turnover time for the top-most
sediment layer (0–5 cm, see Methods),
which harbors most of the biological
activity [Haglund et al., 2003]. The apparent turnover time of this labile pool
(Figure 4) varied on average between 5
and 32 years in Svarttjärn and between 3
and 11 years in Vallentunasjön, for the
range of temperatures between 1 and
21°C. The turnover time of OC is even
longer in the anoxic layers of the sediment, where mineralization is slower.
This is particularly important in sediments
rich in more slowly decomposing OC
(e.g., allochthonous OC [Kristensen et al.,
1995; Bastviken et al., 2003; Sobek
et al., 2009]).
Lakes differ greatly in OC delivery to the sediments, in productivity and watershed OC supply. Burial efficiency
integrates the effect of a number of variables such as differences in lake morphometry [Ferland et al., 2014],
oxygen exposure time [Sobek et al., 2009] and temperature [Gudasz et al., 2010]. However, we have shown
that OC mineralization and in particular temperature has a strong control over the burial efficiency [Gudasz
et al., 2010]. If we assume a similar temperature sensitivity of sediment OC mineralization among lakes and
over time, as we have shown here, increasing mineralization rates as a result of a potential increase in
temperature translates into a reduced OC burial efficiency and hence reduced burial, even if the amount
of OC delivered to the sediments will be different between lakes.
4. Concluding Remarks
Acknowledgments
The study was part of the research
environment Lake Ecosystem Response
to Environmental Change (LEREC),
financially supported by FORMAS
(the Swedish Research Council for
Environment, Agricultural Sciences and
Spatial Planning). Financial support for
laboratory equipment was obtained
from the Knut and Alice Wallenberg
Foundation. We also acknowledge
additional support from the Malméns
Foundation to C.G., and from FORMAS
to S.S. We thank Ingrid Kögel-Knabner
13
and Markus Steffens for the C NMR
measurements and María MoralesPineda, Erik Geibrink, and Ashraf Haque
for help during sampling. To access the
data in this article refer to the
corresponding author.
GUDASZ ET AL.
The focus of the present study was to compare the temperature sensitivity of sediment OC mineralization in
lake sediments with a wide difference in organic matter composition, largely due to differences in dominant
sediment OC source and composition (i.e., allochthonous versus autochthonous), as well as to elucidate
whether different patterns in temperature sensitivity emerge beyond that of short time scale responses.
We show that the temperature sensitivity of OC mineralization in lake sediments is strikingly similar across a
wide range of organic matter composition and within the relevant seasonal time scales of supply and
degradation of OC. Hence, the temperature sensitivity may not change significantly over time, even
though the OC supply and quality, hence the absolute reaction rates, may change. However, factors other
than temperature, such as DOC or nutrient load, may affect future mineralization rates in boreal lakes. We
also suggest that the physical characteristics of the aquatic environment and in particular of the sediments
and sediment-water interface largely govern sediment temperature sensitivity of the OC mineralization.
References
Ågren, G. I., and J. A. M. Wetterstedt (2007), What determines the temperature response of soil organic matter decomposition?, Soil Biol.
Biochem., 39(7), 1794–1798.
Arrhenius, S. (1889), Über die Reaktionsgeschwindigkeit bei der Inversion von Rohrzucker durch Säuren, Z. Phys. Chem., 4, 226–248.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1223
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
Bastviken, D., M. Olsson, and L. Tranvik (2003), Simultaneous measurements of organic carbon mineralization and bacterial production in
oxic and anoxic lake sediments, Microb. Ecol., 46(1), 73–82, doi:10.1007/S00248-002-1061-9.
Benoy, G., K. Cash, E. McCauley, and F. Wrona (2007), Carbon dynamics in lakes of the boreal forest under a changing climate, Environ. Rev.,
15, 175–189, doi:10.1139/a07-006.
Bergström, I., P. Kortelainen, J. Sarvala, and K. Salonen (2010), Effects of temperature and sediment properties on benthic CO2 production in
an oligotrophic boreal lake, Freshwater Biol., 55, 1747–1757, doi:10.1111/j.1365-2427.2010.02408.x.
Bosatta, E., and G. I. Ågren (1999), Soil organic matter quality interpreted thermodynamically, Soil Biol. Biochem., 31(13), 1889–1891.
Boström, B., A.-K. Pettersson, and I. Ahlgren (1989), Seasonal dynamics of a cyanobacteria-dominated microbial community in surface
sediments of a shallow, eutrophic lake, Aquat. Sci., 51(2), 153–178, doi:10.1007/BF00879300.
Boudreau, B. P., and B. B. Jørgensen (Eds.) (2001), The Benthic Boundary Layer: Transport Processes and Biogeochemistry, Oxford Univ. Press,
New York.
Bradford, M. A., C. A. Davies, S. D. Frey, T. R. Maddox, J. M. Melillo, J. E. Mohan, J. F. Reynolds, K. K. Treseder, and M. D. Wallenstein
(2008), Thermal adaptation of soil microbial respiration to elevated temperature, Ecol. Lett., 11(12), 1316–1327, doi:10.1111/
j.1461-0248.2008.01251.x.
Brand, A., C. Dinkel, and B. Wehrli (2009), Influence of the diffusive boundary layer on solute dynamics in the sediments of a seiche-driven
lake: A model study, J. Geophys. Res., 114, G01010, doi:10.1029/2008JG000755.
Brunberg, A.-K., and B. Boström (1992), Coupling between benthic biomass of Microcystis and phosphorus release from the sediments of a
highly eutrophic lake, in Sediment/Water Interactions, pp. 375–385, Springer, Dordrecht, Netherlands.
Cardoso, S. J., A. Enrich-Prast, M. L. Pace, and F. Roland (2014), Do models of organic carbon mineralization extrapolate to warmer tropical
sediments?, Limnol. Oceanogr., 59(1), 48–54, doi:10.4319/lo.2014.59.1.0048.
Conant, R. T., R. A. Drijber, M. L. Haddix, W. J. Parton, E. A. Paul, A. F. Plante, J. Six, and J. M. Steinweg (2008), Sensitivity of organic matter
decomposition to warming varies with its quality, Global Change Biol, 14(4), 868–877, doi:10.1111/j.1365-2486.2008.01541.x.
Conant, R. T., et al. (2011), Temperature and soil organic matter decomposition rates—Synthesis of current knowledge and a way forward,
Global Change Biol, 17(11), 3392–3404, doi:10.1111/J.1365-2486.2011.02496.X.
Craine, J. M., N. Fierer, and K. K. McLauchlan (2010), Widespread coupling between the rate and temperature sensitivity of organic matter
decay, Nat. Geosci., 3(12), 854–857, doi:10.1038/Ngeo1009.
Crawley, M. J. (2009), The R book, John Wiley, Chichester, England.
Davidson, E. A., and I. A. Janssens (2006), Temperature sensitivity of soil carbon decomposition and feedbacks to climate change, Nature,
440(7081), 165–173, doi:10.1038/nature04514.
Davidson, E. A., S. E. Trumbore, and R. Amundson (2000), Biogeochemistry - Soil warming and organic carbon content, Nature, 408(6814),
789–790.
Dean, W. E., and E. Gorham (1998), Magnitude and significance of carbon burial in lakes, reservoirs, and peatlands, Geology, 26(6), 535–538.
Fenchel, T., G. M. King, and T. H. Blackburn (1998), Bacterial Biogeochemistry: The Ecophysiology of Mineral Cycling, Elsevier Acad. Press,
San Diego, Calif.
Ferland, M.-E., Y. T. Prairie, C. Teodoru, and P. A. del Giorgio (2014), Linking organic carbon sedimentation, burial efficiency and long-term
accumulation in boreal lakes, J. Geophys. Res. Biogeosci., 119, 836–847, doi:10.1002/2013JG002345.
Gudasz, C., D. Bastviken, K. Steger, K. Premke, S. Sobek, and L. J. Tranvik (2010), Temperature-controlled organic carbon mineralization in lake
sediments, Nature, 466(7305), 478–481, doi:10.1038/nature09186.
Gudasz, C., D. Bastviken, K. Premke, K. Steger, and L. J. Tranvik (2012), Constrained microbial processing of allochthonous organic carbon in
boreal lake sediments, Limnol. Oceanogr., 57(1), 163–175, doi:10.4319/lo.2012.57.1.0163.
Guillemette, F., S. L. McCallister, and P. A. Giorgio (2013), Differentiating the degradation dynamics of algal and terrestrial carbon within
complex natural dissolved organic carbon in temperate lakes, J. Geophys. Res. Biogeosci., 118, 963–973, doi:10.1002/jgrg.20077.
Gustafsson, A., and E. Rydin (2015), Vattenkvalitet och Plankton i Vallentunasjön 2014, Rapport 2015:14, Naturvatten i Roslagen AB, Norrtälje.
Haglund, A.-L., P. Lantz, E. Törnblom, and L. Tranvik (2003), Depth distribution of active bacteria and bacterial activity in lake sediment, FEMS
Microbiol. Ecol., 46(1), 31–38, doi:10.1016/S0168-6496(03)00190-9.
Hedges, J. I., and J. M. Oades (1997), Comparative organic geochemistries of soils and marine sediments, Org. Geochem., 27(7–8), 319–361.
Hothorn, T., F. Bretz, and P. Westfall (2008), Simultaneous inference in general parametric models, Biom J., 50(3), 346–363, doi:10.1002/
bimj.200810425.
Janssens, I. A., and S. Vicca (2010), Biogeochemistry soil carbon breakdown, Nat. Geosci., 3(12), 823–824, doi:10.1038/Ngeo1024.
Jørgensen, B. B. (2001), Life in the diffusive boundary layer, in The Benthic Boundary Layer: Transport Processes and Biogeochemistry, edited by
B. P. Bodreau and B. B. Jørgensen, pp. 348–373, Oxford Univ. Press, New York.
Knorr, W., I. C. Prentice, J. I. House, and E. A. Holland (2005), Long-term sensitivity of soil carbon turnover to warming, Nature, 433(7023),
298–301, doi:10.1038/nature03226.
Koehler, B., E. von Wachenfeldt, D. Kothawala, and L. J. Tranvik (2012), Reactivity continuum of dissolved organic carbon decomposition in
lake water, J. Geophys. Res., 117, G01024, doi:10.1029/2011JG001793.
Kortelainen, P., H. Pajunen, M. Rantakari, and M. Saarnisto (2004), A large carbon pool and small sink in boreal Holocene lake sediments,
Global Change Biol, 10(10), 1648–1653, doi:10.1111/j.1365-2486.2004.00848.x.
Kristensen, E., S. I. Ahmed, and A. H. Devol (1995), Aerobic and anaerobic decomposition of organic matter in marine sediment: Which is
fastest?, Limnol. Oceanogr., 40(8), 1430–1437.
Laudon, H., M. Berggren, A. Agren, I. Buffam, K. Bishop, T. Grabs, M. Jansson, and S. Kohler (2011), Patterns and dynamics of dissolved
organic carbon (DOC) in boreal streams: The role of processes, connectivity, and scaling, Ecosystems, 14(6), 880–893, doi:10.1007/
S10021-011-9452-8.
McKnight, D., and G. Aiken (1998), Sources and age of aquatic humus, in Aquatic Humic Substances, Ecol. Stud., vol. 133, edited by D. O. Hessen
and L. J. Tranvik, pp. 9–39, Springer, Berlin.
Meyers, P. A., and J. L. Teranes (2001), Sediment organic matter, in Tracking Environmental Change Using Lake Sediments, vol. 2, edited by
W. M. Last and J. P. Smol, pp. 239–270, Kluwer Acad., Dordrecht, Netherlands.
Pinheiro, J., D. Bates, D. Sarkar, and R Core Team (2014), _nlme: Linear and nonlinear mixed effects Models_R package version 3.1-118.
[Available at http://CRAN.R-project.org/package=nlme>.]
R Development Core Team (2010), R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna,
Austria.
Rydin, E., M. Arvidsson, and A. Gustafsson (2010), Vallentunasjön 2008–2009 - Vattenkemi, Plankton och Vattenväxter. Rapport 2010:2,
Naturvatten i Roslagen AB, Norrtälje.
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1224
Journal of Geophysical Research: Biogeosciences
10.1002/2015JG002928
Schielzeth, H. (2010), Simple means to improve the interpretability of regression coefficients, Methods Ecol. Evol., 1(2), 103–113, doi:10.1111/
j.2041-210X.2010.00012.x.
Schielzeth, H., and W. Forstmeier (2009), Conclusions beyond support: Overconfident estimates in mixed models, Behavioral Ecol., 20(2),
416–420, doi:10.1093/beheco/arn145.
Sobek, S., E. Durisch-Kaiser, R. Zurbrugg, N. Wongfun, M. Wessels, N. Pasche, and B. Wehrli (2009), Organic carbon burial efficiency in lake
sediments controlled by oxygen exposure time and sediment source, Limnol. Oceanogr., 54(6), 2243–2254.
Thamdrup, B., J. W. Hansen, and B. B. Jørgensen (1998), Temperature dependence of aerobic respiration in a coastal sediment, FEMS
Microbiol. Ecol., 25(2), 189–200, doi:10.1016/S0168-6496(97)00095-0.
Tranvik, L. J., et al. (2009), Lakes and reservoirs as regulators of carbon cycling and climate, Limnol. Oceanogr., 54(6), 2298–2314.
von Lützow, M., and I. Kögel-Knabner (2009), Temperature sensitivity of soil organic matter decomposition—What do we know?, Biol. Fertil.
Soils, 46(1), 1–15, doi:10.1007/s00374-009-0413-8.
von Wachenfeldt, E., and L. J. Tranvik (2008), Sedimentation in boreal lakes - The role of flocculation of allochthonous dissolved organic
matter in the water column, Ecosystems, 11(5), 803–814, doi:10.1007/S10021-008-9162-Z.
von Wachenfeldt, E., S. Sobek, D. Bastviken, and L. J. Tranvik (2008), Linking allochthonous dissolved organic matter and boreal lake sediment
carbon sequestration—The role of light-mediated flocculation, Limnol. Oceanogr., 53(6), 2416–2426.
GUDASZ ET AL.
LAKE SEDIMENT TEMPERATURE SENSITIVITY
1225