The Impact of Climatic Factors on CО2 Emissions from Soils of Middle-Taiga Forests in Central Siberia: Emission as a Function of Soil Temperature and Moisture

Abstract

Soil CO2 emission is one of the most important components of the global carbon cycle. This study analyzes the seasonal dynamics of soil emission for various land cover types in the middle taiga subzone of central Siberia during five growing seasons. It is shown that, throughout a vast area covered by pine forests and their derivatives formed on sandy soils, seasonal CO2 emission values are determined primarily by the moisture conditions and only secondarily by the temperature regime and ecosystem type. The effect of the forest type is manifested only under the most favorable moisture conditions. A new approach is proposed: divide the growing season into dry and moist periods depending on the threshold soil moisture for areas with
different vegetation types.

Full text

46
ISSN 1067-4136, Russian Journal of Ecology, 2020, Vol. 51, No. 1, pp. 46–56. © Pleiades Publishing, Ltd., 2020.
Russian Text © The Author(s), 2020, published in Ekologiya, 2020, No. 1, pp. 51–61.
The Impact of Climatic Factors on CО2 Emissions from Soils
of Middle-Taiga Forests in Central Siberia: Emission
as a Function of Soil Temperature and Moisture
A. V. Makhnykina
a,
b,
*, A. S. Prokushkin
a, b
, O. V. Menyailo
b
, S . V. Ver kh ovets
a
, I. I. Tychkov
a
,
A. V. Urban
b
, A. V. Rubtsov
a
, N. N. Koshurnikova
a
, and E. A. Vaganov
a, b
aSiberian Federal University, Krasnoyarsk, 660041, Russia
bSukachev Institute of Forest, Siberian Branch, Russian Academy of Sciences, Krasnoyarsk, 660036, Russia
*e-mail: amakhnykina@sfu-kras.ru
Received March 29, 2019; revised May 16, 2019; accepted June 18, 2019
Abstract—Soil CO
2 emission is one of the most important components of the global carbon cycle. This study
analyzes the seasonal dynamics of soil emission for various land cover types in the middle taiga subzone of
central Siberia during five growing seasons. It is shown that, throughout a vast area covered by pine forests
and their derivatives formed on sandy soils, seasonal CO2 emission values are determined primarily by the
moisture conditions and only secondarily by the temperature regime and ecosystem type. The effect of the
forest type is manifested only under the most favorable moisture conditions. A new approach is proposed:
divide the growing season into dry and moist periods depending on the threshold soil moisture for areas with
different vegetation types.
Keywords: carbon cycle, soil CO2 emission, soil temperature, soil moisture, middle taiga, central Siberia
DOI: 10.1134/S10 67413620010063
The interaction between climatic changes and ter-
restrial carbon cycle is a central issue of modern ecol-
ogy and biogeochemistry. Soil CO2 emission into the
atmosphere is a result of autotrophic root respiration
and heterotrophic microbial respiration [1, 2]. The
temporal CO2 emission dynamics is determined by
phenological features of the plants, concentrations of
labile carbon and nitrogen in the soil [3], the microbial
community, and climatic factors of the area, primarily
soil temperature and moisture [4–6]. Parameteriza-
tion of the interaction between the CO2 emission and cli-
matic factors controlling this process is critical for the
carbon balance modeling and understanding of the reac-
tion of biogeocenoses to climate change [7–9].
An increase in CO2 emission with warming is the
key positive feedback mechanism between ecosystems
and global climate changes [10–13]. The parameter-
ization of the relationship between CO2 fluxes emitted
from soil and temperature is the primary approach to the
modeling of the ecosystem carbon balance [5, 14–16]
described normally by exponential response functions
[9, 17]. It has been shown that the CO2 emission in
biogeocenoses of the boreal and temperate zones can
be described exclusively by the temperature [18, 19],
while in precipitation-deficient ecosystems, soil mois-
ture must be included into the equation as well [20].
The role of moisture supply in the formation of emis-
sion fluxes in boreal and temperate forests is currently
increasing due to the growing frequency and duration
of dry periods [21, 22].
The purpose of this study was to assess the impact
of climatic factors on the CO2 efflux from the soil
(based on interannual variability) in several types of
forest biogeocenoses in the middle-taiga subzone of
central Siberia and examine the dependence of the
CO2 emission on soil temperature and moisture. We
assumed that the dependence of the CO2 flux on tem-
perature differs between forest types and varies during
the growing season. The objectives of this study were
as follows: (1) evaluate the influence of hydrothermic
conditions of the growing season on the CO2 flux for-
mation; (2) describe temporal variations in soil CO2
emission in areas with different vegetation types; and
(3) assess the dependence of the soil emission on the
soil temperature and moisture throughout the season.
The boreal zone of Siberia includes vast areas occu-
pied by pine forests and their derivatives growing on
sandy soils; seasonal CO2 emission values are deter-
mined there by wetting conditions of the growing sea-
son. We propose a new approach to CO2 emission
modeling in such ecosystems by dividing the growing
season into dry and moist periods.

RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
THE IMPACT OF CLIMATIC FACTORS 47
MATERIAL AND METHODS
Description of the Climate and Meteorological
Conditions in the Study Period
Studies were carried out in the middle-taiga sub-
zone of central Siberia, in the southern part of the
Turukhansk district, Krasnoyarsk krai (60°26′ N,
89°24′ E). The climate in the region is sharply conti-
nental; the average annual temperature is –3.5°C
(1936–2017). The sum of temperatures above 10°C
varies in the range of 1200–1400°C. The warmest
mo nth in the yea r is July wi th an av erage m ont hly tem-
perature of 18.1°C; the average monthly temperature
of the coldest month (January) is –23.8°C. The f luc-
tuation amplitude of average monthly temperatures
may reach 42°C [23]. The long-term average annual
precipitation in the growing season (June–September)
amounts in this area to 261 mm, while the total annual
precipitation is 594 mm (continuous measurements
since 1966). Overall, the rain precipitation constitutes
some 44% of the total precipitation.
It is necessary to note that the average annual tem-
perature in the study area (Bor weather station;
http://www.meteo.ru) has increased by 3.2°C during
the observation period (1936–2017). 2015 was the
warmest year with an average annual temperature of
–0.5°C. The air temperatures of the five studied grow-
ing seasons (2012, 2013, and 2015–2017) (Fig. 1) also
exceeded the long-term average annual value (13.2°C)
by 0.7–3.4°C: the average air temperature in 2012 was
16.5 ± 0.5°C; in 2013, 14.2 ± 0.6°C; in 2015, 14.2 ± 0.6°C;
in 2016, 16.6 ± 0.5°C; and in 2017, 13.9 ± 0.6°C.
A distinctive feature of the last decade was a high
variability of annual precipitation, with its amount dif-
fering between some years by more than 350 mm: for
instance, 385 mm in 2016 compared to 743 in 2017.
Furthermore, the precipitation variability was mani-
fested most clearly during the growing season. For
instance, significant precipitation deficits in compari-
son with the long-term average annual value were
observed in 2012 and 2013 (54 and 65% of the average
value, respectively). A higher precipitation amount
was recorded in the growing season of 2015: 126% of
the long-term average annual value. The growing sea-
son of 2017 was the closest to the average values by its
precipitation amount (291 and 261 mm respectively)
and air temperature (13.9 and 11.6°C) (Fig. 1).
Study Objects
Four experimental plots (EP) with different land
cover types (lichen pine forest, green-moss pine forest,
mixed forest, and open anthropogenically-disturbed
soil without vegetation) located less than 1.5 km from
each other were selected on a geomorphologically
homogeneous surface (hilltop). The ecosystem
descriptions are provided in Table 1.
Soils of the study region have been formed on gla-
ciofluvial deposits and feature the predominance of
sand in the upper part of the profile. Clayey horizons
(lenses) are usually noted at depths over 1 m. Soils in
areas with undisturbed vegetation cover are illuvial–
ferrous podzols with varying depths of the organic
horizon (thin in the lichen pine forest and mixed for-
est). According to the World Reference Base (WRB)
soil classification system, soils of the experimental
plots are Podzols. The disturbed plot has no upper
genetic soil horizons. Carbon stocks in soils of forest
biogeocenoses are relatively small and constitute,
according to our estimates, about 4 kg C m–2 in a 2-m-
deep layer. The organic horizon contains over 30% of
the total soil organic matter (OM) [24]. The root phy-
tomass constitutes 30–60% of the soil OM; the detri-
tus content is about 10%.
Measurements of Soil CO2 Emission,
Temperature, and Moisture
Soil CO2 emission was measured from June to Sep-
tember during five years: 2012, 2013, 2015, 2016, and
2017 (beginning from the middle of the season). In
spring 2012, polyvinylchloride (PVC) rings 20 cm in
diameter were installed in each EP 1–1.5 m from each
other: five rings were installed in each forested plot
and three rings in the disturbed one. CO2 fluxes from
the soil surface were measured using an LI 8100A
infrared gas analyzer (LI-COR Inc., United States)
during the daytime, in the period from 11:00 to 16:00.
The measurements were made in three replicates; the
mean values were subsequently calculated on the basis
of these data; the measurement time was 2 min, the
intervals between measurements were 30 s. A detailed
description of this methodology is provided in our
previous study [25]. The total numbers of measure-
ments per plot during each season varied from
33 (2017) to 189 (2013). The total numbers of mea-
Fig. 1. Average temperatures (1) and total precipitation
amounts (2) for each measurement season (June–Septem-
ber) of the 5-year period and long-term average annual
values of these parameters according to Bor meteorologi-
cal station (observation periods: 1936–2017 for the air
temperature and 1966–2017 for the precipitation). Data on
the air temperature are provided with standard errors.
Precipitation
amount, mm
1
2
Average 2012 2013 2015 2016 2017
0
100
200
300
400
0
3
6
9
12
15
18
Air temperature, °C

48
RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
MAKHNYKINA et al.
surements taken in each experimental plot over all sea-
sons were as follows: 144 measurements in the lichen
pine forest, 139 in the green-moss pine forest, 140 in
the mixed forest, and 87 in the disturbed plot (three
seasons).
During every CO2 flux measurement, soil tempera-
ture was measured at depths of 5, 10, and 15 cm using
a Soil Temperature Probe Type E (Omega, United
States) and soil water content (SWC) was measured at
a depth of 5 cm using a Theta Probe Model ML soil
moisture sensor (Delta T Devices Ltd., Great Britain).
Assessment of the Relationship between CO2 Emission
Flows and Hydrothermal Parameters
and Segregation of Dry Periods
The analysis of this relationship included examina-
tion of its three key components: soil CO2 emission on
the one hand and soil temperature and moisture cor-
responding to each emission measurement on the
other hand. The soil CO2 emission values and soil
temperatures over the five measurement seasons were
arranged in series of increasing soil moisture values,
separately for each ecosystem. An exponential rela-
tionship between the emission and soil temperature
was identified on the basis of the obtained data selec-
tion. Then the data selection volume was reduced
step-by-step in the emission and soil temperature
series, with the minimum data selection volume con-
sisting of three values. Changes in the determination
coeff icient (R2) between the values were assessed
through sequential exclusion of values from higher to
lover moisture levels (i.e. the data selection was con-
sistently characterized by the increasing soil moisture
deficit). The correlation coefficient (rfact) was calcu-
lated for the dependence of CO2 emission on soil tem-
perature on the basis of Fisher’s test for various data
selection volumes. Statistically significant values of
the correlation coefficient (rsign) (P < 0.05) were deter-
mined for all the examined data selection volumes
(Nis the number of measurements in the ecosystem).
The procedure for determining the threshold
soil water content involved identification of a spe-
cific minimum rsign value satisfying the requirement:
rfact – rsign > 0 (Fig. 2). The groups of measurements
were assessed with regard to the difference between
CO2 fluxes for the dry period (moisture below the
threshold) and the moist period (moisture above the
threshold). Analysis of the CO2 flux rates without lim-
iting moisture conditions made it possible to estimate
the temperature sensitivity for each biogeocenosis
type.
Table 1. Description of studied ecosystems, mean values and standard deviations (SD) over the five measurement seasons
Parameter Lichen pine forest Green-moss pine
forest Mixed forest Disturbed plot
Coordinates 60°47′57.3′′N,
89°21′22.7′′ E
60°48′00.8′′ N,
89°21′04.7′′ E
60°47′52.2′′ N,
89°21′13 .1′′ E
60°47′57.4′′ N,
89°21′01.7′′ E
Soil temperature, °C 13.81 ± 3.74 11.47 ± 2.42 12.98 ± 2.55 20.04 ± 7.23
Soil water content, m3 m–3 0.21 ± 0.09 0.26 ± 0.12 0.25 ± 0.10 0.31 ± 0.09
Carbon content, g C kg–1 soil
(to the depth of 50 cm)
13.0 ± 1.5 7.0 ± 3.6 14.9 ± 1.4 7.3
Proportion of roots, % of the total
Corg deposit (to a depth of 50 cm)
42 48 41 –
Nitrogen content,
g N kg–1 soil (to a depth of 50 cm)
0.6 ± 0.1 0.3 ± 0.1 0.6 ± 0.2 0.4
Forest stand age, years 73 119 27 –
Dominant tree species Pinus sylvestris L. Pinus sylvestris L. Pinus sylvestris L.,
Betula pendula Roth,
Populus tremula L.
–
Dominants of the living ground
cover
Cladonia stellaris
(Opiz) Pouzar et
Vezda, Cl. arbuscula
(Wallr) Flot
Pleurozium schre-
beri (Brid.) Mitt,
Dicranum polyse-
tum Michx.
Cladonia stellaris
(Opiz) Pouzar et
Vezda, Cl. arbuscula
(Wallr) Flot, Pleuro-
zium schreberi (Brid.)
Mitt, Dicranum poly-
setum Michx
–

RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
THE IMPACT OF CLIMATIC FACTORS 49
Q10 Calculation for Moist and Dry Periods
The temperature sensitivity coefficient (Q10) was
computed using the van’t Hoff equation (26:)
(1)
where T1 and T2 are soil temperatures, while R1 and R2
are CO2 fluxes from the soil at temperatures T1 and T2.
Exponential equations for CO2 emission as a function
of soil temperature were constructed for the two
groups (the dry and moist periods) and for each eco-
system separately. Based on these equations, flux val-
ues (R1 and R2) were determined for the temperatures
of 5 and 15°C. The difference between T1 and T2 is
exactly 10°C; therefore, Q10 coefficient was calculated
as the ratio between fluxes at 5 and 15°C.
Statistical Data Processing
A three-way ANOVA with repeated measures (eco-
system type, year, and period of the season (based on
the Julian day; DOY is the day of the year)) was used
for the assessment of the seasonal soil CO2 emission


−


=

21
10
2
10
1
,
TT
R
QR
dynamics. In addition, two-way ANOVA was used to
assess the cumulative effect of climatic conditions (i.e.
soil temperature and moisture) on the seasonal CO2
flux. Thereafter, multiple pairwise (post hoc) compar-
ison of the values were made using the least significant
difference (LSD) method.
The temperature sensitivity coefficient (Q10) was
computed for each ecosystem without replications but
separately for the dry and moist periods. This made it
possible to calculate significance of differences
between Q10 values for the dry and moist periods (Stu-
dent’s t-test), with ecosystems being regarded as repli-
cations.
RESULTS AND DISCUSSION
Soil Hydrothermal Regime
The snow cover disappeared between May 4 (in
2012) and May 14 (in 2015 and 2016). Accordingly, the
soil temperature usually reached above-zero values in
the first 10-day period of May. The maximum soil
temperatures in all experimental plots were recorded
in the third 10-day period of July. With respect to tem-
perature values (Table 2), the plots can be arranged in
Fig. 2. Dependence of correlation coefficients (r) between the soil CO
2
emission and soil temperature on the soil moisture and
data selection volume for plots with different vegetation types: (a) lichen pine forest, (b) green-moss pine forest, (c) mixed forest,
and (d) disturbed plot; (r
sign
) distribution of the statistically significant correlation coefficient between the soil emission and tem-
perature, (r
fact
) measured values.
(c)
Soil water content, m3 m−3
Soil water content, m3 m−3
0.24
0 0.2 0.4 0.6 0.8
0
0.2
0.4
0.6
0.8
1.0
(d)
0.23
0 0.2 0.4 0.6 0.8
(a)
Number of measurements
0.30
0 50 100 150 0 50 100 150
0.2
0.4
0.6
0.8
1.0
(b)
Number of measurements
0.14
rfact
rsign

50
RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
MAKHNYKINA et al.
the following descending series: disturbed plot >
lichen pine forest > mixed forest > green-moss pine
forest (P < 0.05).
Soil moisture reached a peak value during active
snow melting in early spring (the first 10-day period of
May) and also in September. According to its con-
tents, the plots ranked as follows: the disturbed plot in
the first place, 0.31 ± 0.09 m3 m–3 (Table 2); then fol-
low mixed forest and green-moss pine forest plots,
from 0.24 to 0.28 m3 m–3 in each; the lichen pine forest
is the least moistened, 0.21 ± 0.09 m3 m–3. The peak
of the seasonal moisture content for the June–Sep-
tember period was regularly recorded at the end of the
season (September). The season of 2012 was distin-
guished by specific moisture conditions in all forested
experimental plots: soil water content in June–July were,
on average, 60% lower than in August–September.
Seasonal Dynamics of Soil CO2 Emission
As a rule, the seasonal emission dynamics in the
studied land cover types had the pattern typical for for-
est ecosystems of the boreal and temperate zones: low
emission rates in spring and late fall, and maximum
rates in late July and early August (Fig. 3), which is
determined by the dynamics of soil temperature and its
effect on activities of autotrophs and heterotrophs [16,
20, 27].
Average emission values varied significantly (P <
0.05) in different experimental plots. The disturbed
plot featured the lowest emission flux (1.1 ± 0.1 μmol
CO2 m
–2 s
–1) due to the absence of vegetation and
organic soil horizons, the primary CO2 production
sources. As it is known, the availability of soil OM at
large depths is fairly low, and it cannot provide high
soil emission rates; in addition, the activity of hetero-
trophs decreases with depth. At the same time, it is
Table 2. Soil temperature (Ts, °C) and water content (SWC, m3 m–3) in all experimental plots, mean values with standard-
errors
Month
Lichen pine forest Green-moss pine forest Mixed forest Disturbed plot
TsSWC TsSWC TsSWC TsSWC
2012
June 16.5 ± 1.2 0.11 ± 0.01 10.4 ± 1.6 0.11 ± 0.03 13.3 ± 1.0 0.17 ± 0.03 31.1 ± 0.9 0.21 ± 0.01
July 18.3 ± 1.1 0.10 ± 0.01 13.2 ± 1.7 0.07 ± 0.01 15.9 ± 1.0 0.11 ± 0.01 22.3 ± 0.2 0.17 ± 0.01
August 11.9 ± 0.3 0.22 ± 0.02 9.7 ± 0.1 0.20 ± 0.03 11.5 ± 0.2 0.21 ± 0.02 16.9 ± 0.1 0.20 ± 0.01
September 10.2 ± 0.2 0.34 ± 0.20 9.7 ± 0.4 0.29 ± 0.04 10.5 ± 0.4 0.37 ± 0.05 No data No data
2013
June 13.5 ± 0.3 0.23 ± 0.01 8.6 ± 0.2 0.30 ± 0.01 11.1 ± 0.2 0.32 ± 0.01 22.3 ± 0.3 0.42 ± 0.01
July 15.0 ± 0.3 0.18 ± 0.01 11.4 ± 0.2 0.23 ± 0.01 13.4 ± 0.1 0.24 ± 0.01 25.9 ± 0.2 0.28 ± 0.01
August 15.2 ± 0.2 0.24 ± 0.01 12.7 ± 0.1 0.24 ± 0.01 14.2 ± 0.1 0.24 ± 0.01 20.7 ± 0.2 0.27 ± 0.01
September 8.2 ± 0.1 0.28 ± 0.01 8.9 ± 0.04 0.41 ± 0.02 8.9 ± 0.04 0.37 ± 0.01 8.9 ± 0.1 0.39 ± 0.01
2015
June 15.2 ± 0.7 0.18 ± 0.01 12.1 ± 0.7 0.27 ± 0.02 13.6 ± 0.5 0.24 ± 0.02 24.5 ± 0.1 0.39 ± 0.01
July 17.1 ± 0.4 0.23 ± 0.01 14.9 ± 0.5 0.24 ± 0.02 15.9 ± 0.4 0.24 ± 0.02 25.4 ± 0.2 0.27 ± 0.01
August 14.4 ± 0.2 0.28 ± 0.01 14.1 ± 0.4 0.38 ± 0.02 14.5 ± 0.2 0.34 ± 0.03 19.2 ± 0.1 0.28 ± 0.01
September 8.9 ± 0.2 0.27 ± 0.02 9.7 ± 0.1 0.38 ± 0.03 9.4 ± 0.1 0.34 ± 0.02 9.2 ± 0.1 0.33 ± 0.02
2016
June 16.1 ± 0.7 0.23 ± 0.01 11.8 ± 0.8 0.22 ± 0.03 14.6 ± 0.8 0.26 ± 0.02 No data No data
July 17.7 ± 0.4 0.24 ± 0.01 14.3 ± 0.3 0.23 ± 0.02 15.7 ± 0.4 0.26 ± 0.01
August 14.9 ± 0.2 0.26 ± 0.02 13.6 ± 0.3 0.27 ± 0.03 13.7 ± 0.5 0.26 ± 0.02
September 10.9 ± 0.1 0.24 ± 0.01 11.5 ± 0.2 0.31 ± 0.03 11.4 ± 0.2 0.24 ± 0.02
2017
July 15.6 ± 0.4 0.27 ± 0.01 13.2 ± 0.5 0.39 ± 0.04 13.9 ± 0.5 0.34 ± 0.02 No data No data
August 14.9 ± 0.3 0.29 ± 0.02 13.6 ± 0.2 0.38 ± 0.03 13.8 ± 0.2 0.35 ± 0.01
September 8.7 ± 0.2 0.35 ± 0.02 9.8 ± 0.1 0.56 ± 0.05 9.2 ± 0.1 0.47 ± 0.02

RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
THE IMPACT OF CLIMATIC FACTORS 51
necessary to note that during lengthy droughts (e.g.
in 2012 and 2013), no distinct seasonal dynamics of
emissions was observed in the disturbed EP, while its
average CO2 fluxes amounted to only 0.40 ±
0.04 μmol CO2 m–2 s–1 (2013). In moist years (e.g. the
season of 2015), the seasonal emission dynamics was
manifested more clearly, and higher soil respiration
values (up to 4.4 μmol CO2 m
–2 s
–1) were observed
both in the second half of June and late July.
In the lichen pine forest, CO2 fluxes over the grow-
ing season averaged 3.4 μmol CO2 m
–2 s
–1, ranging
from 0.9 to 9.8 μmol CO2 m–2 s–1. It is necessary to
note that lower soil CO2 emission rates were previously
recorded during the growing season in artificial south-
ern-taiga pine forests of central Siberia: from 0.5 to
4μmol CO
2 m
–2 s
–1 [28]. The authors of that study
have concluded that the heterotrophic component
predominates in the soil CO2 emission f lux.
The green-moss pine forest is similar to the lichen
pine forest in the seasonal CO2 fluxes, which aver-
aged 4.3 μmol CO2 m
–2 s
–1 and varied from 1.4 to
8.2 μmol m–2 s–1 in early August. In pine forests of the
United States [19, 26], an increase in soil CO2 emis-
sion (up to 12 μmol CO2 m–2 s–1) was also observed in
August due to higher precipitation.
The maximum soil CO2 emission and variations
were recorded in the mixed forest plot (Fig. 3): the
average seasonal CO2 fluxes differed significantly
(P< 0.05) from other plot types, amounting to 7.0 ±
1.2 μmol CO2m–2 s–1 and varying from 1.8 to 20.9 μmol
CO2 m–2 s–1. The high flux rates in this ecosystem are
supposedly determined not only by the hydrothermal
conditions, but also by the role of forest-forming spe-
cies, diversity of ground vegetation layer, and activity
of soil microorganisms. A comparative assessment of
soil emission in dark conifer and mixed forests of Can-
ada (Quebec) [29] has shown that the emission in the
mixed forest throughout the growing season is higher
by 45%. In mixed forests of France [30], the maxim um
fluxes reach 10 μmol CO2 m–2 s–1; however, the emis-
sion peak occurs in the beginning of the summer sea-
son (June).
Year-to-Year Variability of CO2 Emissions
Consideration of interannual variability makes it
possible to assess the impact of climatic factors on
changes in the CO2 flux from the soil, provided these
factors vary significantly (Table 3). In our case, the
studied seasons differed significantly in moisture con-
ditions; this was the main factor determining yearly
variations of the total soil CO2 emission.
The effect of different climatic factors on soil CO2
emission varies depending on EP type (Table 3). For
instance, the total seasonal precipitation has the main
effect on the emission rate in the lichen pine forest,
but this correlation is statistically insignificant. The
correlation with soil temperature (r = 0.999) is the
Fig. 3. Seasonal dynamics of soil CO
2
emissions in plots with different vegetation types during the 5-year measurement period.
Mean values with standard errors (N = 5) are shown.
Mixed forest
2012
DOY
140 175 210 245 280
CO2 flux, μmol m−2 s−1
0
7
14
21
Disturbed plot
140 175 210 245 280
Lichen pine forest
0
7
14
21
Green-moss pine forest
2013
2015
2016
2017

52
RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
MAKHNYKINA et al.
strongest in the green-moss pine forest. In the mixed
forest, the amount of precipitation also has the main
effect on CO2 flux. Soil temperature is the primary
factor determining CO2 emission in the plot with dis-
turbed soil cover (r = 0.998). It is necessary to note
that a negative correlation of emission with the air
temperature is observed in all experimental plots, but
it lacks statistical significance. The influence of soil
moisture on CO2 fluxes is significant in the lichen pine
forests and mixed forest, with the correlation coeffi-
cients of 0.579 and 0.606, respectively.
The effect of ecosystem type lacked statistical sig-
nificance (P = 0.810) (Table 4), but the experimental
plots significantly differed from each other in some
years; i.e. the main factors such as “ecosystem type”
and “year” turned out to be interrelated (P < 0.010).
The combination of seasonal temperatures and
moisture conditions statistically significantly (P < 0.05)
affects the f lux rate (Fig. 3). The disturbed plot is an
exception; the impact of these factors was insignificant
in it (P = 0.235).
The maximum fluxes in the lichen pine forest were
recorded in the growing season of 2015 (Fig. 3): the
average soil emission was 5.3 μmol CO2 m–2 s–1 reach-
ing 9.8 μmol CO2 m–2 s–1 in the middle of the season. In
the green-moss pine forest, all seasons could be divided
into two groups by the soil emission rates (i.e. CO2 flux
values): group 1 (up to 5 μmol CO2 m–2 s–1) and group 2
(5 to 10 μmol CO2 m–2 s–1). The first group includes
two seasons: precipitation-deficient 2012 and 2013
with average seasonal soil emission rates of 2.9 and
2.7 μmol CO2 m–2 s–1 respectively. The second group
includes three seasons: 2015, 2016, and the second half
of the 2017; the seasonal CO2 f luxes in this group were
higher by 35–55% in comparison with group 1.
During the dry season, soil CO2 emission in Bel-
gian pine forests with moss ground layer formed on
sandy soils [31] reached 2.3 μmol CO2 m–2 s–1 in late
July, while the peak of the f luxes coincided with the
period of higher precipitation.
Similarly with other forested plots, the mixed forest
EP demonstrated the maximum CO2 fluxes in the
Table 3. Correlation between average seasonal fluxes (kg C m–2) and climatic variables on each EP. Statistically significant
correlation coefficients (p < 0.05) are boldfaced
(Tair) air temperature, (Tsoil) soil temperature, (P) seasonal precipitation amount, and (SWC) soil water content.
EP
Correlation coeff icient and p values (in bparentheses)
Tair, °CTsoil, °CP*, mm SWC, m3 m–3
Lichen pine forest –0.579 0.856 0.991 0.579
(0.607) (0.345) (0.084) (0.607)
Green-moss pine forest –0.127 0.999 0.813 0.127
(0. 919) (0.033) (0.396) (0.919)
Mixed forest –0.606 0.837 0.995 0.606
(0.585) (0.367) (p = 0.063) (0.585)
Disturbed plot –0.138 0.998 0.819 0.138
(0.912) (0.040) (0.389) (0.912)
Table 4. Results of three-way ANOVA for CO2 emission rate and two-way ANOVA for annual CO2 flux
(F) Fisher’s criterion, (P) signif icance level, (–) the effect of that factor or interaction was not determined.
Factors and their
interactions
CO2 flux rate Annual CO2 flux
degrees of freedom FPdegrees of freedom FP
Ecosystem 2 2.75 <0.010 2 3.60 0.810
Year 4 6.52 <0.001 4 10.12 <0.001
Day 10 46.35 <0.001 – – –
Ecosystem × year 8 2.18 0.010 8 2.12 0.008
Ecosystem × day 20 2.86 0.032 – – –
Day × year 40 3.11 0.045 – – –
Ecosystem × year × day 80 2.55 0.002 – – –

RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
THE IMPACT OF CLIMATIC FACTORS 53
growing season of 2015 (Fig. 3): the average seasonal
CO2 fluxes significantly differed from other plot types
and amounted to 9.6 ± 0.7 μmol CO2 m–2 s–1 reaching
20.9 μmol CO2 m–2 s–1 in the middle of the growing
season of 2015. The minimum fluxes were noted in
2012 and 2013 with average seasonal CO2 fluxes of 3.7
and 4.5 μmol CO2 m–2 s–1 respectively.
On the whole, significant differences in soil emis-
sion were observed between the ecosystems in moist
seasons (Fig. 4). Under the precipitation deficit con-
ditions in 2012 and 2013, the CO2 flux decreased by
43% on an average, while the differences in soil emis-
sion values between the forest biogeocenoses were
nullified.
The highest average seasonal emission were
recorded in 2015: 0.7 ± 0.1 kg C m–2. In the driest sea-
son of 2012, CO2 fluxes in some ecosystems decreased
by a factor of 2–2.5 in comparison with 2015. The
studied ecosystems turned out to be much more sus-
ceptible to droughts than, for instance, meadow eco-
systems [32] where the CO2 flux decreases by 8–20%
depending on the drought severity and duration.
A previous study by Timokhina [33] provides eco-
system respiration for a lichen pine forest growing in
the study area for the period of 1998–2000: about
372 g C m–2 yr–1 (1364 g CO2 m–2 yr–1). According to
these estimates based on the data collected by the Zot-
ino Tall Tower Observatory (ZOTTO), the emission
fluxes from ecosystems located within the coverage
zone of the tall tower varied in 2009–2013 from 331
(2010) to 398 g C m–2 per season (2012), which is con-
sistent with our results. According to Shibistova et al.
(16), the soil respiration constitutes some 60% of the
annual ecosystem respiration (23.1–23.4 mol m–2 yr–1 ~
280 g C m–2 yr–1).
Correlation between Soil Temperature
and Soil CO2 Emissions
As noted in previous studies on pine forests [16], if
the effect of droughts is excluded from the analysis of
the relationship between the soil emission and tem-
perature, then the correlation between these factors
(R2) increases from 0.06 to 0.59 and becomes signifi-
cant. To assess the relationship between the soil emis-
sion and temperature, we have divided all available
data into two groups by soil moisture conditions (i.e.
into dry and moist periods). Fig. 2 shows the distribu-
tion of the correlation coefficient (r) in the course of
the step-by-step exclusion of emission values recorded
on days with high soil water content: from higher to
lower moisture levels and down to the last three pairs
of values. The intersection point of the two relation-
ships indicates the threshold soil water content level;
starting from it, the correlation between the CO2 flux
and soil temperature becomes significant. The thresh-
old soil water content (m3 m
–3) dividing statistically
significant r values from insignificant ones in the series
of increasing soil moisture and numbers of measure-
ments are as follows: 0.30 for the lichen pine forest,
0.14 for the green-moss pine forest, 0.24 for the mixed
forest, and 0.23 for the disturbed plot. The soil emis-
sion rates in the lichen pine forest, green-moss pine
forest, and mixed forest in the group with moisture
levels exceeding their threshold values were higher
than those in the group limited by the water content by
8, 25 and 14% respectively (Fig. 5).
The following trend was observed in all forested
experimental plots under the extremely low soil mois-
ture conditions: the relationship between the CO2 flux
and soil temperature was very strong, and it was statis-
tically significant (Fig. 2). Apparently, this correlation
manifests the stress reaction to absolute drought con-
ditions [15, 22].
Temperature Sensitivity (Q10)
The temperature sensitivity of the soil OM miner-
alization process is expressed through a coefficient
(Q10) [17] indicating how many times the process rate
increases with an increase in the temperature by 10°C
(30). The coefficient was computed separately for the
two distinguished soil moisture groups (Fig. 6).
The temperature sensitivity of autotrophic respira-
tion is higher in comparison with microbial respiration
(18, 34, 35); this determines higher Q10 values in for-
ested plots. In the group with moisture levels exceed-
ing the threshold value, the mixed forest features the
highest Q10 value (Q10 = 4.3). Q10 coefficients in the
lichen and green-moss pine forests are close to each
other: 3.6 and 3.4 respectively. In the disturbed plot
(no vegetation), Q10 values had varied from 1.5 to 1.8
depending on the year. The Q10 values computed for
this EP are close to published estimates made for het-
Fig. 4. CO
2
fluxes from the soil surface in the measure-
ment period (2012–2017): (1) lichen pine forest, (2) green-
moss pine forest; (3) mixed forest, and (4) disturbed plot.
The diagram shows mean fluxes from all experimental
plots for each growing season ± standard deviations (SD).
1
2012 2013 2015 2016 2017
0
0.5
1.0
1.5
2.0
C–CO2 emission, kg C m−2
2
3
4

54
RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
MAKHNYKINA et al.
erotrophs (Q10 = 2 [7]), which confirms the heterotro-
phic origin of emitted CO2.
In dry periods, the temperature sensitivity of all
forest ecosystems remains at the same and consider-
ably lower level (Q10 = 1.9). The highest differences in
the temperature sensitivity were noted in the lichen
pine forest: Q10 decreases by 65% in the group with dry
conditions; in the mixed forest and green-moss pine
forest, Q10 values decrease by 58 and 43% respectively.
The minimum differences in Q10 values between the
two groups with different moisture conditions were
recorded in the disturbed plot: in the group with dry
conditions, the soil emission was lower by 18% than
that in the group with higher soil moisture.
The temperature sensitivity of soil emissions
changes under different moisture conditions. In the
conditionally high-moisture group, Q10 in forested
experimental plots is higher by 55% on an average than
that under dry conditions. The CO2 emission rate in
the disturbed plot was less sensitive to rises of the tem-
perature; in addition, Q10 differed insignificantly in
moist and dry periods.
CONCLUSIONS
Total CO2 f luxes from the studied ecosystems
amounted to 0.3–08 kg C m–2 per season. The mixed
forest featured the maximum fluxes, while the dis-
turbed plot, the minimum ones: the average fluxes in
the mixed forest was 5 times higher. Soil CO2 emission
reach the maximum in late July and beginning of
August; however, in dry seasons, an increase in the
fluxes was observed only since the second half of
August.
Fig. 5. Relationships between the soil CO2 emission and soil temperature for two groups distinguished by the soil water content
(SWC) on plots with different vegetation types: (a) lichen pine forest, (b) green-moss pine forest; (c) mixed forest, and (d) dis-
turbed plot.
SWC < 0.24
SWC > 0.24
0 5 10 15 20 25 30 35
CO2 flux, μmol m−2 s−1
7
14
21
(c)
y = 1.0475e0.1411x
Soil temperature, °C
R2 = 0.49
y = 1.8796e0.0722x
R2 = 0.08
SWC < 0.23
SWC > 0.23
0 5 10 15 20 25 30 35
(d)
y = 0.184e0.0609x
R2 = 0.22
y = 0.3816e0.0412x
R2 = 0.22
SWC < 0.30
SWC > 0.30
0
7
14
21 (а)
y = 0.7152e0.1272x
R2 = 0.80
y = 2.0752e0.0233x
R2 = 0.03
SWC < 0.14
SWC > 0.14
(b)
y = 0.9058e0 .1218x
R2 = 0.53
y = 1.235e0.0657x
R2 = 0.17
Fig. 6. Temperature sensitivity coefficient (Q
10
) values
computed for plots with different vegetation types. The
data are presented for two groups distinguished by the soil
water content: (1) SWC above the threshold value, and
(2) SWC below the threshold value.
Lichen
pine forest
Green-moss
pine forest
Mixed
forest
Disturbed
plot
0
2
4
6
Q10
1
2

RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
THE IMPACT OF CLIMATIC FACTORS 55
The climatic conditions, primarily moisture condi-
tions, significantly affect the seasonal CO2 flux; in
other words, the studied forest biogeocenoses of cen-
tral Siberia formed on well-drained sandy soils are
highly sensitive to droughts. The effect of the forest
type is manifested only in moist years. The precipita-
tion deficit in the seasons of 2012 and 2013 (54 and
65% of the average value respectively) caused a
decrease in the CO2 flux by 46% on an average in com-
parison with seasons with suff icient precipitation. In
the driest season of 2012, CO2 fluxes in some ecosys-
tems had decreased by 2–2.5 times in comparison with
the moist season of 2015.
FUNDING
This study was supported by the Russian Foundation for
Basic Research, project nos. 17-05-01257 and 18-34-00736.
COMPLIANCE WITH ETHICAL STANDARDS
Conf lict of Interest
The authors declare that they have no conflict of interest.
REFERENCES
1. Buchmann, N., Biotic and abiotic factors controlling
soil respiration rates in Picea abies stands, Soil Biol. Bio-
chem., 2000, vol. 32, pp. 1625–1635.
2. IPCC Special Report: Global Warming of 1.5°C, Switzer-
land, IPCC, 2018.
3. Matvienko, A.I., Effect of nitrogen on carbon mineral-
ization in soils under Siberian larch and Scots pine,
Cand. Sci. (Biol.) Dissertation, Krasnoyarsk, 2017.
4. Menyailo, O. and Huwe, B.C., N-mineralization and
denitrification as function of temperature and water po-
tential in organic and mineral horizons of forest soil,
J. Plant Nutr. Soil Sci., 1999, vol. 162, pp. 527–531.
https://doi.org/10.1002/(SICI)1522-2624(199910)162
5. Raich, J.W. and Schlesinger, W.H., The global carbon-
dioxide flux in soil respiration and its relationship to
vegetation and climate, Te ll us, vo l. 44 , n o. 2, p p. 81– 99 .
https://doi.org/10.1034/j.1600-0889.1992.t01-1-00001.x
6. Wiseman, P.E. and Seiler, J.R., Soil CO2 efflux across
four age classes of plantation loblolly pine (Pinus taeda L.)
on the Virginia Piedmont, For. Ecol. Manag., 2004,
vol. 192, pp. 297–311.
https://doi.org/10.1007/s10533-004-0566-3
7. Czimczik, C.I., Schmidt, M.W.I., and Schulze, E.-D.,
Effects of increasing fire frequency on black carbon and
organic matter in podzols of Siberian Scots pine forests,
Eur. J. Soil Sci., 2005, vol. 56, pp. 417–428.
https://doi.org/10.1111/j.1365-2389.2004.00665.x
8. Lindroth, A., Grelle, A., and Morén, A.-S., Long-term
measurements of boreal forest carbon balance reveal
large temperature sensitivity, Glob. Change Biol., 1998,
pp. 443–450.
https://doi.org/10.1046/j.1365-2486.1998.00165.x
9. Lloyd, J. and Taylor, J.A., On the temperature depen-
dence of soil respiration, Funct. Ecol., 1994, vol. 8,
pp. 315–323.
https://doi.org/10.2307/2389824
10. Grace, J. and Rayment, M., Respiration in the balance,
Nature, 1999, vol. 404, pp. 819–820.
https://doi.org/10.1038/35009170
11. Rodeghiero, M. and Cescatti, A., Main determinants of
forest soil respiration along an elevation/temperature
gradient in the Italian Alps, Glob. Change Biol., 2005,
vol. 11, pp. 1024–1041.
h tt p s :/ / d oi . or g / 10 .1111 /j .1 36 5 -2 4 8 6. 2 0 05 . 00 9 6 3 .x
12. Sánchez, M.L., Ozores, M.I., López, M.J., et al., Soil
CO2 fluxes beneath barley on the Central Spanish Pla-
teau, Agric. For. Meteorol., 2003, pp. 85–95.
https://doi.org/10.1016/S0045-65 35(02)00071-1
13. Schlesinger, W. and Andrews, J., Soil respiration and
the global carbon cycle, Biogeochemistry, 2000, vol. 48,
pp. 7–20.
14. Davidson, E.A. and Janssens, I.A., Temperature sensi-
tivity of soil carbon decomposition and feedbacks to cli-
mate change, Nature, 2006, vol. 440, pp. 65–173.
https://doi.org/10.1038/nature04514
15. Lee, X., Wu, H.J., Sigler, J., et al., Rapid and transient
response of soil respiration to rain, Glob. Change Biol.,
2004, vol. 10, no. 6, pp. 1017–1026.
htt ps: / /d o i.o rg/ 10 .1111 / j.1529 -8817.20 03. 007 87.x
16. Shibistova, O., Lloyd, J., Evgrafova, S., et al., Seasonal
and spatial variability in soil CO2 eff lux rates for a cen-
tral Siberian Pinus sylvestris forest, Tel lus , 2002,
vol. 54B, pp. 552–567.
https://doi.org/10.3402/tellusb.v54i5.16687
17. Kirschbaum, M.U.F., Will changes in soil organic carbon
act as a positive or negative feedback on global warming,
Biogeochemistry, 2000, vol. 48, no. 1, pp. 21–51.
https://doi.org/10.1023/A:1006238902976
18. Boone, R.D., Nadelhoffer, K.J., Canary, J.D., and
Kaye, J.P., Roots exert a strong influence on the tem-
perature sensitivity of soil respiration, Nature, 1998,
vol. 396, pp. 570–572.
19. Irvine, J. and Law, B.E., Contrasting soil respiration in
young and old-growth ponderosa pine forests, Glob.l
Change Biol., 2002, vol. 8, no. 12, pp. 1183–1194.
https://doi.org/10.1046/j.1365-2486.2002.00544.x
20. Yuste, J.C., Janssens, I.A., Carrara, A., et al., Interac-
tive effects of temperature and precipitation on soil res-
piration in a temperate maritime pine forest, Tree
Physiol., 2003, vol. 23, pp. 1263–1270.
21. Lia, J.-T., Wang, J.-J., Zeng, D.-H., et al., The inf lu-
ence of drought intensity on soil respiration during and
after multiple drying–rewetting cycles, Soil Biol. Bio-
chem., 2018, vol. 127, pp. 8 2–89. do i
https://doi.org/10.1016/j.soilbio.2018.09.018
22. Silva, B.O., Moitinho, M.R., Araújo Santos, G.A.,
et al., Soil CO2 emission and short-term soil pore class
distribution after tillage operations, Soil Tillage Res.,
2019, vol. 186, pp. 224–232.
https://doi.org/10.1016/j.still.2018.10.019
23. Lesnye ekosistemy Eniseiskogo meridiana (Forest Eco-
systems on the Yenisei Meridian), Pleshikov, F.I., Ed.,
Novosibirsk: Nauka, 2002.
24. Polosukhina, D.A. and Prokushkin, A.S., Comparative
characterization of soil organic matter stocks and iso-
tope composition in forest biogeocenoses within the

56
RUSSIAN JOURNAL OF ECOLOGY Vol. 51 No. 1 2020
MAKHNYKINA et al.
ZOTTO coverage zone, in Lomonosovskie chteniya na
Altae: fundamental’nye problemy nauki i obrazovaniya:
Izbr. trudy mezhdun. konf. (Lomonosov Lectures in Al-
tai: Fundamental Problems in Science and Education,
Selected Papers of Int. Conf.), Moscow, 2017, p. 55.
25. Makhnykina, A.V., Prokushkin, A.S., Vaganov, E.A.,
et al., Dynamics of CO2 fluxes from the soil surface in
pine stands of Central Siberia, Zh. Sib. Fed. Univ., Ser.
Biol., 2016, no. 3, pp. 338–357.
https://doi.org/10.17516/1997-1389-2016-9-3-338-357
26. Tang, J., Misson, L., Gershenson, A., et al., Continu-
ous measurements of soil respiration with and without
roots in a ponderosa pine plantation in the Sierra Neva-
da Mountains, Agric. For. Meteorol., 2005, vol. 132,
pp. 212–227.
https://doi.org/10.1016/j.agrformet.2005.07.011
27. Laganiere, J., Pare, D., Bergeron, Y., et al., The effect
of boreal forest composition on soil respiration is medi-
ated through variations in soil temperature and C qual-
ity, Soil Biol. Biochem., 2012, vol. 53, pp. 18–27.
https://doi.org/10.1016/j.soilbio.2012.04.024
28. Matvienko A.I., Makarov M.I., and Menyailo, O.V.,
Biological sources of soil CO2 under Larix sibirica and
Pinus sylvestris, Russ. J. Ecol., 2014, vol. 45, no. 3,
pp. 174–180.
https://doi.org/10.7868/S0367059715020079
29. Vanэt Hoff, J.H., Lectures on Theoretical and Physical
Chemistry, vol. 1: Chemical Dynamics, London: E. Ar-
nold, 1899, pp. 224–229.
30. Vincent, G., Shahriari, A.R., Lucot, E., et al., Spatial
and seasonal variations in soil respiration in a temperate
deciduous forest with f luctuating water table, Soil Biol.
Biochem., 2006, vol. 38, pp. 2527–2535.
https://doi.org/10.1016/j.soilbio.2006.03.009
31. Harper, C.W., Blair, J.M., Fay, P.A., et al., Increased
rainfall variability and reduced rainfall amount de-
creases soil CO2 efflux in a grassland ecosystem, Glob.
Change Biol., 2005, vol. 11, pp. 322–334.
htt ps: / /d o i.o rg/ 10 .1111 / j.1365 -24 86. 200 5 .0 0 89 9 .x
32. Tchebakova, N.M., Vygodskaya, N.N., Arneth, L.,
et al., Energy and mass exchange and the productivity
of main Siberian ecosystems (from eddy covariance
measurements): 2. Carbon exchange and productivity,
Biol. Bull. (Moscow), 2015, vol. 42, no. 6, pp. 579–588.
https://doi.org/10.7868/S00023329140100 44
33. Timokhina, A.V., Prokushkin, A.S., Onuchin, A.A.,
et al., Variability of ground CO2 concentration in the
middle taiga subzone of the Yenisei region of Siberia,
Russ. J. Ecol., 2015, vol. 46, no. 2, pp. 143–151.
https://doi.org/10.7868/S0367059715020122
34. Matteucci, M., Gruening, C., Goded, BallariI., et al.,
Components, drivers and temporal dynamics of ecosys-
tem respiration in a Mediterranean pine forest, Soil Biol.
Biochem., 2015, vol. 88, pp. 224–235.
https://doi.org/10.1016/j.soilbio.2015.05.017
35. Ruehr, N.K. and Buchmann, N., Soil respiration flux-
es in a temperate mixed forest: Seasonality and tem-
perature sensitivities differ among microbial and root-
rhizosphere respiration, Tree Physiol., 2009, vol. 30,
pp. 165–176.
https://doi.org/10.1093/treephys/tpp106
Translated by L. Emeliyanov